Analysis of Composite Wipe Samples
for Lead Content
Final Report
For U.S. Environmental Protection Agency
Technical Programs Branch, 7404
Chemical Management Division
Office of Pollution Prevention and Toxics
Battelle Subcontract No. 103639-G002421
Task Order No. 2-5
MRI Project No. 5021 -A(05)
July 1,1996
-------
This document is printed on recycled paper.
-------
July 1, 1996
EPA 747-R-96-003
Analysis of Composite Wipe Samples
for Lead Content
Prepared by
Midwest Research Institute
for
Technical Programs Branch
Chemical Management Division
Office of Pollution Prevention and Toxics
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
Recycled/Recyclable Printed with Vegetable Based Inks on Recycled Paper (20% Postconsumer)
-------
Disclaimer
This document has been prepared for the Office of Pollution Prevention and Toxics
(OPPT), U.S. Environmental Protection Agency. The material in this document has been
subject to Agency technical and policy review and approved for publication as an EPA
report. Mention of trade names, products, or services does not convey, and should not be
interpreted as conveying official EPA approval, endorsement, or recommendation.
-------
Contributing Organizations
The study described in this report was funded by the U.S. Environmental Protection
Agency and the U.S. Department of Housing and Urban Development. The study was
managed by the U.S. Environmental Protection Agency. The study was conducted
collaboratively by two organizations under contract to the U.S. Environmental Protection
Agency, Midwest Research Institute and Battelle Memorial Institute. Each organization's
responsibilities are listed below.
Midwest Research Institute
Midwest Research Institute (MRI) was responsible for the study including the overall
production of the Test Plan, Quality Assurance Project Plan, the experimental design, the
laboratory analysis of the wipe samples, the statistical analysis, and writing the technical
report.
Battelle Memorial Institute
Battelle Memorial Institute reviewed the Test Plan and provided input for the
experimental design and laboratory analysis, assisted with the development of the report,
and provided management oversight for the project.
U.S. Environmental Protection Agency
The U.S. Environmental Agency (EPA) cofunded the study and was responsible for
managing the study, for reviewing study documents, and for arranging for the peer review
of the final report. The EPA Work Assignment Manager was Mr. Todd Holderman.
Barbara Leczynski and Brad Schultz also provided substantial direction in the study design.
The EPA Project Officer was Jill Hacker. Cindy Stroup, Branch Chief of the Technical
Programs Branch, initiated this study.
U.S. Department of Housing and Urban Development
The Department of Housing and Urban Development (HUD) cofunded the study.
in
-------
Contents
Contributing Organizations iii
Figures vii
Tables viii
Executive Summary ix
1 Introduction 1-1
1.1 Background 1-1
1.2 Objectives 1-1
1.3 Scope of the Investigation 1-2
1.4 Overview of Report 1-3
2 Conclusions 2-1
2.1 Method Performance 2-1
2.2 Effect of Number of Wipes 2-2
2.3 Analytical Method 2-2
2.4 Cost 2-2
3 Study Design 3-1
3.1 Study Objectives 3-1
3.2 Method Selection 3-4
3.3 Wipe Selection 3-6
3.4 Standard Reference Material (SRM) Selection 3-7
3.5 Experimental Design 3-8
4 Sample Handling and Identification 4-1
4.1 Wipe Loading Procedure 4-1
4.2 Sample Identification 4-2
4.3 Sample Containers, Storage, and Shipment 4-2
5 Sample Digestion and Analysis 5-1
5.1 Instrumentation 5-1
5.2 Calibration Standard Sources 5-2
5.3 Laboratory Observations 5-2
6 Quality Assurance/Quality Control 6-1
6.1 Quality Assurance 6-1
6.2 Quality Control 6-2
6.3 Other Data Quality Indicators 6-11
-------
7 Statistical Analysis and Results 7-1
7.1 Data Tracking, Data Entry, and Data Processing 7-1
7.2 Statistical Analysis of the Analytical Data 7-1
7.3 Evaluation of Process Parameter Data 7-18
8 References and Bibliography 8-1
Appendices
A. Method Review SummaryProtocols for Sample Preparation and Analysis
of Lead in Single-Surface and Composite Wipe Samples
B. Modified SW-846 Method 3050A Acid Digestion Procedure for Single-Wipe
Samples
C. Modified SW-846 Method 3050A Acid Digestion Procedure for Two-Wipe
Composite Samples
D. Modified SW-846 Method 3050A Acid Digestion Procedure for Four-Wipe
Composite Samples
E. WOHL Acid Digestion Procedure for Single-Wipe Samples
F. WOHL Acid Digestion Procedure for Two- and Four-Wipe Composite
Samples
G. Preparation of Reference Material Wipe Samples for the Composite Wipe
Investigation
H. Laboratory Data and Lead (Pb) Recovery Results
I. ANOVA Results for Four-way and Three-way Mixed Models
VI
-------
Figures
3-1. Design Layout 3-3
3-2. Hot Plate Configuration for the Modified EPA SW-846 Method 3050A 3-12
3-3. Hot Plate Configuration for the Wisconsin Occupational Health Laboratory
(WOHL) Method 3-13
6-1. Distribution Plots of Gravimetric Data for Single Samples and Composites .... 6-3
6-2. LCS Percent Recovery by FAA for EPA 3050A and WOHL Methods 6-9
6-3. LCS Percent Recovery by TCP for EPA 3050A and WOHL Methods 6-10
6-4. Distribution Plots of FAA Percent Recovery Data by Laboratory, Digestion
Method, and Type of Composite 6-12
6-5. Distribution Plots of TCP Percent Recovery Data by Laboratory, Digestion
Method, and Type of Composite 6-13
6-6. Distribution Plots of FAAICP Percent Recovery Data by Laboratory,
Digestion Method, and Type of Composite 6-16
7-1. Basic FAA Recovery Statistics for Each Laboratory, Preparation Method,
Composite Type, and SRM Loading 7-3
7-2. Basic ICP Recovery Statistics for Each Laboratory, Preparation Method,
Composite Type, and SRM Loading 7-4
7-3. Recovery Statistics by Analytical Method, Digestion Method, SRM Loading
Level, and Composite Type 7-12
7-4. Between and Within Laboratory Precision by Analytical Method, Digestion
Method, and Composite Type 7-16
vn
-------
Tables
3-1. Design Layout for the Modified EPA SW-846 Method 3050A 3-14
3-2. Design Layout for the Wisconsin Occupational Health
Laboratory (WOHL) Method 3-15
4-1. Wipe Sample Containers 4-2
6-1. Method Blank Sample Layout and Analytical Results 6-5
6-2. Wipe Blank Sample Layout and Analytical Results 6-7
6-3. Number of Wipe Blank Detects by Wipe Brand 6-7
6-4. Laboratory Control Sample Layout 6-8
6-5. LCS Percent Recovery Statistics 6-11
6-6. Wipe Samples with Extreme Recovery Results 6-14
7-1. Average Percent Recovery by Wipe Brand 7-7
7-2. Overall Recovery Statistics by Analytical and Digestion Method 7-9
7-3. Recovery Statistics by Analytical Method, Digestion Method, SRM
Loading Level, and Composite Type 7-11
7-4. Within and Between Laboratory Precision Results by Digestion Method,
Analytical Method, and Composite Type 7-15
7-5. Comparison of FAA and TCP Recoveries by Digestion Method, SRM Level,
and Composite Type 7-18
7-6. Relative Effort Comparison Among Types of Composite for Laboratory 1 ... 7-21
7-7. Relative Cost Factor Comparison Across Composite Types, Separately for
Each Digestion Method 7-21
7-8. Relative Cost Factor Comparison Between Digestion Methods, Separately
for Each Composite Type 7-22
Vlll
-------
Executive Summary
The United States government has responded to the existing hazard posed by the
presence of lead-based paint in the nation's housing stock by enacting Title X (the
Residential Lead-based Paint Hazard Reduction Act, of federal housing legislation). This
Act of Congress also enacted Title IV of the Toxic Substances Control Act (TSCA).
Title X directs the U.S. Environmental Protection Agency (EPA), the U.S. Department of
Housing and Urban Development (HUD), and other agencies to develop programs that
would ultimately reduce lead hazards in housing. It is envisioned that Title X and Title IV
programs will result in increased dust sampling in residences across the country. One of
the preferred methods for sampling residential dust for lead uses baby or hand wipes.
"Compositing" multiple wipes during sample collection has been suggested as a means to
reduce the costs of both the sampling and analysis aspects of these activities. In fact,
compositing of wipes is mentioned for clearance testing and risk assessment in the HUD
Guidelines for the Evaluation and Control of Lead-Based Paint Hazards in Housing,1 and
in various EPA documents.
The term composite is used when two or more physical samples are combined for
laboratory analysis. In the case of wipe sampling, two or more wipes collected from
common components (e.g., floors or window sills) in a dwelling are combined in the field
and then analyzed as a single sample. Prior to this study, the feasibility of modifying
current sample preparation protocols to handle the increased mass of wipes and dust that
result from compositing had not been demonstrated. In addition, it was uncertain whether
any cost savings could be realized by compositing. Further, no standard method currently
exists for the digestion of composite wipe samples for the analysis of lead in dust.
For those reasons, this study was undertaken to (1) investigate the feasibility of
developing further the existing sample preparation methods to analyze composite wipe
samples while meeting basic data quality objectives for accuracy and precision; and
(2) determine whether compositing of wipes, if acceptable on a technical or performance
basis, could reduce the cost of sample preparation and analysis relative to single-wipe
methods.
From a number of available sample preparation methods, two were selected for
evaluation based on discussions with experts in the lead analysis field and a literature
review. The modified EPA Method 3050A and the Wisconsin Occupational Health
Laboratory (WOHL) method were selected based on their anticipated ability to adequately
handle the added mass of materials and dust expected in a composite of wipes. In addition,
modified EPA 3050A is a commonly used procedure in many laboratories and is consistent
1 Guidelines for the Evaluation and Control of Lead-based Paint Hazards in Housing.,
June 1995. The National Center for Lead-Safe Housing, U.S. Department of Housing and
Urban Development, Washington, D.C.
IX
-------
with the ASTM method titled "Emergency Standard Practice for Hot Plate Digestion of
Dust Wipe Samples for Determination of Lead by Atomic Spectrometry" (ASTM
Designation ES 36-94). The WOHL method is consistent with the method described in
Appendix 14.2 of the HUD Guidelines and is applicable to composite wipes. The WOHL
method also appeared to employ a more rapid digestion procedure, which could provide
additional cost savings.
The experimental design for this study involved the following elements: (1) three
off-the-shelf wipe materials (one hand-wipe product and two brands of baby wipes) were
used; (2) a single-wipe sample, a two-wipe composite sample, and a four-wipe composite
sample were used; (3) each wipe was spiked with one standard reference material (NIST
SRM 2704) at either a low dust loading level of 0.5 g or a high level of 2 g. These dust
loading levels correspond to lead levels of 80.5 jig and 322 jig per wipe, respectively;
(4) three separate laboratories involved in lead testing and participating in the National
Lead Laboratory Accreditation Program (NLLAP) performed the analysis; (5) each
laboratory prepared the samples using both the EPA 3050A and the WOHL preparation
methods; and (6) each sample was analyzed with two different analytical techniques, flame
atomic absorption spectrometry (FAA) and inductively-coupled plasma spectrometry
(ICP).
Performance specifications were met for EPA Method 3050A under all experimental
conditions tested (average spike recoveries across all three laboratories ranged from 89% to
101%). Acceptable recoveries were achieved for the WOHL method for single-wipe
samples and two-wipe composites at the low and high SRM loading levels, using either
FAA or ICP analysis (average spike recoveries across all three laboratories ranged from
82% to 94%). However, difficulties were sometimes encountered with the WOHL method
when compositing four wipes. This was the case for FAA and ICP results at the high SRM
level (average spike recovery across all three laboratories was 78% for FAA and 65% for
ICP). This was also the case for ICP results at the low SRM level (average spike recovery
across all three laboratories was 79%). Acceptable recovery (average of 83% across all
laboratories) was achieved at the low SRM loading level using the four-wipe WOHL
method and FAA.
Based on individual laboratory results, again, all three laboratories met the
performance specification for recovery under all experimental conditions using EPA
Method 3050A. Acceptable performance was also achieved by each laboratory for the
single-wipe samples and the two-wipe composites using the WOHL method and FAA.
However, some laboratories experienced difficulties with the four-wipe WOHL method
followed by FAA analysis. None of the three laboratories had consistently acceptable
performance using the WOHL method followed by ICP analysis, regardless of the type of
composite. Thus, it was found that acceptable results using the WOHL method with four-
wipe composites could be achieved, although difficulties may be experienced by some
laboratories.
x
-------
On average, EPA Method 3050A yielded consistently higher recoveries than the
WOHL method. In addition, FAA yielded higher recoveries on average than ICP for either
digestion method. In all cases, intra-laboratory precision, based on the results from the
three laboratories in the study, met the data quality objectives of less than 20% relative
standard deviation for both the EPA 3050A and WOHL methods. The results from the
four-wipe composites were consistently the least precise, although always acceptable.
The time required to perform each digestion procedure and composite type was
recorded by each laboratory participating in the investigation. These results indicate that
the time requirements for a given digestion method are fairly consistent across each
composite type for a given laboratory. The multiwipe composites required slightly more
time to digest, relative to the single-wipe samples. However, that increase in time was not
proportional to the number of wipes in a composite: the two- and four-wipe composites did
not take two and four times as long to digest as the single-wipe samples, respectively.
A sample preparation and analysis cost comparison was performed based on
information provided by the three laboratories. Each laboratory also evaluated the cost
aspects of compositing from a retrospective viewpoint. Using the single wipe as the basis
for cost comparisons, an approximate cost reduction of 64% was estimated for the four-
wipe composites using the EPA Method 3050A. An approximate cost reduction of 67%
was estimated for the four-wipe composites using the WOHL method. That is, laboratory
analysis of a four-wipe composite was approximately 65% less expensive than that of four
single-wipe analyses for each method. Across all laboratories and composite types,
approximate cost reductions of 25% to 70% were estimated for the WOHL method as
compared to the EPA Method 3050A.
Acceptable performance can be achieved using the EPA Method 3050A for
composites containing two or four wipes based on the recovery and precision results
obtained with the standard reference material used in this study. The results were
especially promising using FAA analysis. The WOHL method was found to provide
acceptable performance for single wipes and two-wipe composites. It is possible to
achieve acceptable performance with the WOHL method for four-wipe composites,
although some difficulties might be encountered. Furthermore, the WOHL method was
affected by the SRM loading level when ICP was used. A statistically significant,
estimated loss in recovery of 1.75% for each gram of SRM was found when preparing the
samples with the WOHL method and analyzing them by ICP. This limitation of the
WOHL method may be a consideration in sampling applications where high dust loading
levels are expected, such as for window sills and troughs in risk assessment. It is
recommended that every laboratory would need to validate its performance with any
method for the analysis of composite wipes as well as single-wipe samples. A significant
cost benefit was derived from compositing. In summary, this study found that compositing
two or four wipes into a single sample is a viable alternative for the analysis of lead in dust
wipe samples and does result in cost savings.
XI
-------
1-1
-------
2-1
-------
5-1
-------
4-1
-------
5-1
-------
6-1
-------
7-1
-------
3-1
-------
Section 1
Introduction
1.1 Background
Household dust is recognized as a major pathway for conveying lead to children and
adults. Numerous studies in residential settings have collected dust and analyzed it for lead
to determine the lead exposure faced by the occupants in response to mandates cited in
Title X.1 The collection of dust in selected areas of the microenvironment using wipes and
then compositing these wipes for analysis is a general methodology with potentially broad
application because it is technically feasible and may result in cost savings. Compositing
of wipes is mentioned for dust sampling in risk assessments, lead hazard screens, and
clearance testing in Chapter 5 and Chapter 15 of the HUD Guidelines,2 and in various EPA
documents. There is, however, essentially no standard method available for the analysis of
composited wipes. The present work assignment is intended to assess such a protocol.
Specific criteria on composite wipe sampling and analysis relative to clearance testing
and other uses need to be set. A number of wipe sampling and analysis methods focus on
single-wipe samples. The analysis of a composite wipe sample encompasses sample
preparation and subsequent spectrometric analysis. Essentially no difference exists in the
instrumental analysis of a single and a composite wipe sample, provided the same
analytical method is used. The difference lies in the preparation of the sample.
Although several sample preparation methods are available, to our knowledge, none
have been thoroughly evaluated for composite wipe samples. For example, the Wisconsin
Occupational Health Laboratory (WOHL) developed a method for composite samples
containing up to four wipes, but published evaluations of this method have been limited. It
should be noted that the WOHL method was included as Appendix 14.2 in both the
February 1995 draft and June 1995 versions of the HUD Guidelines*'2
Prior to this study, it was expected that with further development, modified EPA
SW-846 Method 3050A,3 a hot plate digestion method, could produce an acceptable,
practical method for preparing composite samples of up to four wipes. This method is a
commonly used laboratory method for single wipes but is not typically used for composites
wipes. As part of this study, modified SW-846 Method 3050A was further modified for
two- and four-wipe composites.
1.2 Objectives
The two objectives of this investigation were to evaluate whether (a) preparation
methods can be developed for composite wipes so that acceptable recoveries can be
achieved and (b) these methods will reduce costs relative to the single-wipe methods.
1-1
-------
This exploratory project was limited to investigating the feasibility of further
developing the existing modified EPA SW-846 Method 3050A to achieve a recovery of
80% to 120% and a precision of 20% relative standard deviation, and estimating the
precision and recovery of the WOHL method. National Institute of Standards and
Technology (NIST) Standard Reference Material (SRM) 2704 (Buffalo River Sediment)
with a certified value of 161 jig Pb/g of material was used to evaluate both digestion
methods.
1.3 Scope of the Investigation
The primary efforts for this study were conducted at the contractor's main laboratories,
where staff prepared all samples. The sample digestion and analysis efforts were
conducted at a total of three laboratories. The inclusion of two commercial laboratories in
this investigation provided for an assessment of digestion and analysis variability across
multiple laboratories.
Two laboratories selected for participation in the study were involved in lead testing
with the EPA National Lead Laboratory Accreditation Program (NLLAP).5 The third
laboratory was in the process of seeking NLLAP recognition at the time of the study. As
part of the NLLAP, all three laboratories had successfully participated in the quarterly
Environmental Lead Proficiency Analytical Testing Program (ELPAT) and had maintained
a proficient rating for several years. At the onset of this study, two of the three laboratories
had been accredited by the American Industrial Hygiene Association (AIHA), a
professional association. The third laboratory applied for the Environmental Lead Analysis
Program (ELAP) accreditation from the American Association for Laboratory
Accreditation (A2LA), and recently received final approval from the A2LA Accreditation
Council for NLLAP recognition.
The investigation of the two preparation methods was limited to testing three off-the-
shelf wipe materials under two dust/lead-loading conditions using a single NIST SRM.
Two brands of baby wipes and one brand of moist disposable towelettes were evaluated in
the study. These types of products are not intended for this application, but historically
have been suitable for lead analysis. Refer to Section 3.3 for more information on the
selection of the wipes.
The digestion methods were evaluated using two different analytical methods each.
The first analytical method (EPA SW-846 Method 6010A) determines metals in the
digestates using inductively coupled plasma atomic emission spectrometry (ICP). The
determination of lead by direct aspiration flame atomic absorption spectrometry (FAA) was
the second analytical method (EPA SW-846 Methods 7000A [general] and 7420).
1-2
-------
1.4 Overview of Report
The audience for this report are technical management personnel who are responsible
for activities involving lead dust sampling and analysis. This may include laboratory
management, risk assessors, inspectors, representatives of public housing authorities,
and/or state/local health department representatives. The remaining sections of this report,
which are described briefly below, present the results, the study design and objectives, and
the methods.
Section 2 presents the discussion and conclusions made from the results of the study.
Section 3 presents the study objectives, experimental design, and discusses the selection of
the methods, wipe materials, SRM, and SRM loadings used in the study.
Section 4 describes the procedure used for loading the wipe samples with SRM, the
sample identifiers, sample containers, and sample storage/shipment.
Section 5 presents the sample analysis instrumentation, calibration standard sources, and
method references used for the analysis. The laboratory observations made
during the conduct of sample preparation and analysis also are presented in this
section. The specific sample digestion procedures used are included in the
appendices.
Section 6 describes the quality assurance and quality control (QA/QC) aspects of the
study, including the data audit and data assessment. The results of the
gravimetric data generated from loading the wipes with SRM, along with the QC
sample results, also are presented in this section.
Section 7 describes the procedures used for processing the data. This section presents the
results for the analytical data and the statistical analysis procedures. The method
accuracy, method precision, and process parameter data also are included in this
section.
The appendices contain the following information:
A Method Review SummaryProtocols for Sample Preparation and Analysis of
Lead in Single-Surface and Composite Wipe Samples
B Modified SW-846 Method 3050A Acid Digestion Procedure for Single-Wipe
Samples
C Modified SW-846 Method 3050A Acid Digestion Procedure for Two-Wipe
Composite Samples
D Modified SW-846 Method 3050A Acid Digestion Procedure for Four-Wipe
Composite Samples
E WOHL Acid Digestion Procedure for Single-Wipe Samples
1-3
-------
F WOHL Acid Digestion Procedure for Two- and Four-Wipe Composite Samples
G Preparation of Reference Material Wipe Samples for the Composite Wipe
Investigation
H Laboratory Data and Lead (Pb) Recovery Results
I ANOVA Results for Four-way and Three-way Mixed Models
1-4
-------
Section 2
Conclusions
This exploratory study was undertaken to investigate the feasibility of preparing and
analyzing wipe composites from a performance and cost standpoint. Three laboratories
were recruited to digest and analyze single-wipe samples, two-wipe composite samples,
and four-wipe composite samples, using two digestion methods and two analytical
methods. The three laboratories provided information regarding time, cost, and effort
required for sample preparation and analysis, using the two digestion methods tested.
Highlights of the results obtained from the statistical analysis of the analytical data, and
tentative conclusions based on an evaluation of the process parameter information provided
by the three laboratories are presented below.
2.1 Method Performance
The data quality objectives (DQOs) for the analytical results were those specified for
the EPA Method 3050A: (1) the accuracy must be between 80% and 120% of the known
spiked amount; and (2) the intra-laboratory precision must be no more than 20% relative
standard deviation (%RSD). No DQOs are established for the WOHL method.
The results of this study, using NIST SRM 2704, show that the modified EPA
Method 3050A met the DQOs for accuracy and precision, and yielded superior results over
the WOHL method, regardless of SRM level, composite type, or analytical method. It was
found to be possible that the DQO for recovery could be met using the WOHL method for
single wipes and two-wipe composites. The WOHL method met the DQO for accuracy
(80% to 120%) for the single wipes and two-wipe composites, regardless of SRM level or
analytical method. However, for the four-wipe composites, average accuracies met the
DQO only at the low SRM level and with FAA analysis. Otherwise, the method accuracy
fell short of the DQO at the high SRM level (78% using FAA; 65% using ICP) and at the
low SRM level using ICP (79%). Furthermore, the WOHL method was affected by the
SRM loading level when ICP was used. A statistically significant, estimated loss in
recovery of 1.75% for each gram of SRM was found when preparing the samples with the
WOHL method and analyzing them by ICP. This limitation of the WOHL method may be
a consideration in sampling applications where high dust loading levels are expected, such
as for window sills and troughs in risk assessment. The WOHL method provided
acceptable intra-laboratory precision results. Both preparation methods resulted in %RSD
below 10% (compared to the 20% RSD DQO) in all cases.
While the modified EPA Method 3050A is an established procedure (for single-wipe
samples only), the WOHL method is relatively new and the laboratories may not have been
familiar with this method prior to the study. Furthermore, digesting composite wipes may
have been a new procedure for these laboratories.
2-1
-------
2.2 Effect of Number of Wipes
The EPA Method could be performed acceptably regardless of the number of wipes
per sample: average recoveries ranged from 88% to 100% for single-wipe samples; from
91% to 101% for two-wipe composites; and from 94% to 100% for four-wipe composites.
Similarly, the WOHL method provided acceptable results for single-wipe samples and
two-wipe composites. Average recoveries ranged from 82% to 91% for single-wipe
samples and from 82% to 94% for two-wipe composites. However, difficulties were
sometimes encountered with the WOHL method when compositing four wipes. This was
the case for both analytical methods (TCP and FAA) at the high SRM level (average
recoveries of 65% and 78%, respectively) and at the low SRM level for ICP (average
recovery of 79%). The WOHL method performed acceptably only for FAA at the low
SRM level, with an average recovery of 83%. On average, the four-wipe composites
provided percent recoveries approximately 12% lower than both the single-wipe and the
two-wipe composites for the WOHL digestion method. In addition, the four-wipe
composites consistently yielded the least precise intra-laboratory results, although in all
cases the %RSD was less than 9%.
2.3 Analytical Method
In general, recovery differences between the two analytical methods were affected only
slightly by the two digestion methods, with no apparent pattern between the two. Overall,
across all factors in the study design, slightly higher recoveries were obtained with FAA
than with ICP (average difference of approximately 7%, though not statistically
significant). However, in most cases, differences in recoveries between FAA and ICP were
larger when using the WOHL method (average difference of 8%) than when using the EPA
3050A Method (average difference of 3%), all other conditions held constant. In addition,
ICP recoveries showed a slight decrease with increasing SRM loading for both digestion
methods, while SRM loading levels had no effect on FAA recoveries under similar
conditions. The implication is that either method will suffice, but there may be modest
advantages of using FAA over ICP. Further, since the FAA instrument is generally less
expensive to operate, it may be the preferred instrument when analyzing composites with
these methods, especially if modified EPA 3050A is being used.
2.4 Cost
The time required to perform each digestion procedure and composite type
combination was recorded by each laboratory participating in the investigation. The three
laboratories also evaluated their costs associated with digestion and analysis of composite
samples from a retrospective viewpoint. A limited sample preparation and analysis cost
comparison was performed based on this information.
2-2
-------
Based on this cost comparison, a cost-benefit can be derived from compositing. Using
the single wipe as the basis for cost comparisons, an average cost reduction of 64% per
wipe was estimated for the four-wipe composites using EPA Method 3050A. For example,
suppose that a single wipe costs $20 to analyze. Based on the reduction of effort obtained
from this study, a four-wipe composite would cost $28.50, a 64% cost reduction over $80,
the cost of four single wipes. Similarly, a reduction of 67% was estimated for the four-
wipe composites using the WOHL method. In general, the multiwipe composites required
slightly more time to digest relative to the single-wipe samples; however, that increase was
not proportional (i.e., the two- and four-wipe composites did not take two and four times as
long to digest as the single-wipe samples, respectively).
The three laboratories involved in the study also estimated per-wipe cost reductions
ranging from 25% to 70% for the WOHL method as compared to the EPA Method 3050A.
Some laboratories may not achieve acceptable performance with the four-wipe composites
using the WOHL method; in this case, the per-wipe cost reduction would not be realized.
In addition, all three laboratories indicated that FAA required less time than ICP analysis,
reporting an estimated 2 min per sample for FAA versus 3 to 5 min per sample for ICP.
Finally, since FAA instruments are less expensive to purchase than ICP instruments, an
additional cost reduction can be achieved from using FAA rather than ICP.
2-3
-------
Section 3
Study Design
This section provides the overall study design for the development of the composite
wipe method. The study objectives and the selection of the methods, wipe materials, and
standard reference material (SRM) are discussed, along with the experimental design. A
detailed description of the laboratory methods and data analysis is provided in subsequent
sections and the appendices.
3.1 Study Objectives
Compositing of wipes, if acceptable, could have important cost containment
implications in the effort to reduce childhood lead poisoning. Cost containment is
important if lead hazard control and clearance testing following those efforts are to take
place in millions of dwellings with lead-based paint hazards. However, specific criteria on
composite wipe sampling and analysis relative to clearance testing and other uses need to
be set. As part of this work, two digestion methods were investigated: (a) the modified
EPA SW-846 Method 3050A, a hot plate digestion method commonly used for single
wipes, and (b) the Wisconsin Occupational Health Laboratory (WOHL) method developed
for up to four-wipe composites.
The two objectives of this investigation were to evaluate whether (a) preparation
methods can be developed for composite wipes so that acceptable recoveries can be
achieved and (b) these methods will reduce costs relative to the single-wipe methods.
The scope of this project was limited. First, the feasibility of further modifying the
EPA SW-846 Method 3050A for two- and four-wipe composites to achieve a recovery of
80% to 120% and a precision of 20% relative standard deviation (RSD) was investigated.
Second, the precision and recovery of the WOHL method were estimated when using
composite wipe samples. Only a single reference material, NIST SRM 2704 with a
certified value of 161 jig Pb/g of material, was used to evaluate both methods. The
optimization or validation of a composite wipe preparation method was not included in the
scope of this project.
During the planning phase of the study, a Test Plan was prepared that addressed the
background, objective, and scope of the investigation. The Test Plan also presented the
experimental design, laboratory methodology, and the QA/QC aspects for the study. The
Test Plan was then submitted to six organizations for peer review, and comments were
received from nine reviewers. Their comments fell into the following general categories:
Statistical design and analysis
5-1
-------
SRM selection and loading levels
Wipe brands and the number of brands to be tested
Analytical methods and potential interferences
Digestion methodology (e.g., beaker sizes, final digest volumes, and reagent
volumes)
Types of composites to include in the investigation
Type and number of quality control samples to be included in the sample batches
Number of laboratories to be included in the study
All review comments were discussed in detail during the planning phase of the study.
Where appropriate and feasible, the Test Plan was revised to reflect the reviewers'
comments and suggestions.
As a result, the investigation of the two preparation methods encompassed the
following:
Testing three off-the-shelf wipe materials under two dust/lead-loading conditions
using a single NIST standard reference material. An attempt was made to obtain
wipes from two different manufacturers' lots for each brand.
Performing all sample digestion and analysis work at three laboratories to capture
digestion and analysis variability across multiple laboratories.
Using two analytical methodsinductively coupled plasma atomic emission
spectroscopy (ICP) and flame atomic absorption spectroscopy (FAA)to analyze
the wipes for lead at each laboratory.
Estimating the digestion times and relative cost factors for each composite type and
each digestion procedure at the three laboratories.
Documenting all equipment, facilities, and materials used at each laboratory.
Comparing the performance and relative costs of the two methods within and
among the laboratories.
Figure 3-1 shows, in a schematic form, the layout of this study.
5-2
-------
Selected three brands of wipes
Defined sample types:
- single wipe
- two-wipe composite
- four-wipe composite
Spiked individual wipes with SRM 2704
at one of two levels:
- 0.5 g (low)
- 2 g [high)
Selected three laboratories
Prepared samples using two different
methods at each laboratory
Analyzed each sample using two different
analytical techniques at each laboratory
Brand
A
Brand
B
Brand
c
a
/\
low
high
Lab 1 Lab 2 Lab 3
EPA 3050A
WOHL
Figure 3-1. Design Layout
3-3
-------
3.2 Method Selection
A literature search was previously conducted to identify existing sample preparation
and analysis methods for lead in single surface and composite wipes. The methods found
were presented in a report titled, "Protocols for the Collection and Analysis of Composite
Wipe Samples."6 A summary of this report, along with other methods evaluated, is
presented in Appendix A. The advantages and disadvantages of each method were
weighed before method selection and performance of this investigation. The methods
found during the literature search are briefly described below:
Hot plate digestion using nitric acid (HNO3) and hydrogen peroxide (H2O2): This is
a commonly used digestion procedure that could be adapted to multiwipe
composites and provide suitable lead recoveries. Existing methods that are
applicable to or have been adapted to digesting single-wipe samples include the
modified EPA SW-846 Method 3050A, ASTM Method ES 36-94, NIOSH Method
7082, and a method used at Azimuth Laboratories.
Hot plate digestion using HNO3 and hydrochloric acid (HC1): This method is
applicable to single surface and multiwipe composites. Existing methods include
the WOHL method, which is equivalent to the method included in Appendix 14.2 of
the HUD Guidelines (February 1995 draft and June 1995 versions).4'2
Hot plate digestion using HNO3 and perchloric acid (HC1O4): NIOSH Method 7300
exists, but a perchloric acid fume hood is required for this type of digestion due to
the potential explosion hazard associated with HC1O4.
Room temperature extraction and hot plate digestion using 1 M HNO3: This
method is used by the University of Cincinnati Medical Center, Department of
Environmental Health, for hand wipe composites for the analysis of lead and
arsenic.
Hot plate and microwave digestion: Two methods were found that used a
combination of hot plate and microwave oven for heating the digests. These
methods also used perchloric acid, which is potentially explosive.
Room temperature extraction: This extraction method uses 0.1 N HC1 at room
temperature.
Ultrasonification: This procedure uses 10% HNO3 and a commercial ultrasonic
cleaner to extract the wipes.
Dry-ashing: Dry-ashing procedures use furnace heating to oxidize the sample. The
resulting ash is solubilized in HNO3 or HC1 and is diluted to volume with water.
5-4
-------
The methods were reviewed to determine if they could be adapted to multiwipe
composites. Experts in lead analysis were consulted from NIOSH, Azimuth Laboratories,
the University of Cincinnati Medical Center Department of Environmental Health, and
EPA\Atmospheric Research Exposure Assessment Laboratory (AREAL) to discuss the
basic characteristics of a wipe method that must respond to the challenge of handling an
added mass of wipe material and dust. Based on the literature review and further
discussions, two types of hot plate digestion methods were selected.
The first method selected for inclusion in the study was the modified EPA SW-846
Method 3050A. This method was selected because it is a commonly used procedure that
had a strong likelihood of achieving the accuracy and precision objectives of the study.
The ASTM method titled "Emergency Standard Practice for Hot Plate Digestion of Dust
Wipe Samples for Determination of Lead by Atomic Spectrometry" (ASTM Designation
ES 36-94) is consistent with the modified Method 3050A for single surface samples. The
modified Method 3050A is also very similar to a method used by Azimuth Laboratories.
The modified EPA SW-846 Method 3050A was routinely used for digesting single-
wipe samples using HNO3 and H2O2. As part of this study, further modifications were
made to the method to accommodate the two- and four-wipe composite samples.
Additional safety measures were required to deal with the larger quantities of acids needed
to perform composite sample digestion. The modified EPA SW-846 3050A methods used
for single surface, two-, and four-wipe composites are presented in Appendices B, C, and
D, respectively.
The second method selected was the WOHL method, which is the method presented in
Appendix 14.2 of the HUD Guidelines (February 1995 draft and June 1995 versions). As
presented in the HUD Guidelines, this method is applicable to the analysis of single-wipe
samples or composite samples of up to four wipes. This method was selected because it
appeared to be a more rapid digestion procedure; however, it has been only used for a
limited number of tests. The WOHL method uses HNO3 and HC1 for the digestion. The
WOHL digestion procedures for single surface and multiwipe composites are presented in
Appendices E and F, respectively.
It should be noted that approximately 70% to 80% of the laboratories participating in
the ELPAT program routinely use hotplate digestion methods. These laboratories could
thus readily adopt a proposed composite sample digestion technique. Most of the
remaining participating laboratories use microwave digestion techniques. To our
knowledge, there is no commercially available microwave digestion vessel that can handle
multiple wipes. Therefore, microwave digestion would be precluded as an alternative
digestion procedure for composite sample digestion unless some method involving
predigestion of the wipe media could be developed.
The two types of digestion methods were evaluated using two different analytical
methods each. The first analytical method (EPA SW-846 Method 6010A) determines
metals in the digestates using ICP. The determination of lead by direct aspiration FAA
3-5
-------
(EPA SW-846 Methods 7000A [general] and 7420) was the second analytical method.
These methods of analysis were selected because they are techniques commonly used by
commercial laboratories for lead analysis. The two methods were included in the study to
capture the analysis variability for each digestion method and analysis method
combination. It should be noted that about 60% of ELPAT participating laboratories use
FAA analysis and about 35% of them use ICP-AES analysis, so the two selected
instrumental techniques cover the most commonly used instrumental method used in the
ELPAT program for lead wipes.
3.3 Wipe Selection
The evaluation criteria used for the selection of candidate wipes included in the
composite wipe study were taken from Appendix 13.1 of the HUD Guidelines, and are
summarized below:
The wipes should contain low background lead levels (< 5 jig/wipe). The analysis
of wipe blanks was incorporated into the test design to provide background levels
for each brand and each lot of wipes.
The wipes must be durable, of a single thickness, and should not tear easily.
Aloe vera must be avoided to minimize lead background levels.
The wipes must be the premoistened type for field sampling purposes; dry materials
such as gauze and filter paper are excluded (wipes containing alcohol may be used
as long as they do not dry out).
The wipes selected should yield 80% to 120% recovery rates from samples spiked
with leaded dust.
It should be mentioned that a specification for wipes was recently adopted by ASTM.
The Standard Specification for Wipe Sampling Materials for Lead in Surface Dust (ASTM
El792-96) establishes limits for background lead in the wipe, ruggedness when used on
various typical surfaces, lead recovery, and collection efficiency. The specification is
presently undergoing further change in ASTM to improve the specification in regard to the
uniformity of moisture and the wipe thickness range specified.
The three brands of wipes selected for inclusion in the composite wipe investigation
were Wash a-bye Baby wipes (Scott Paper Company), Baby Wipes with Lanolin (Rockline,
Inc.), and Wash'n Dri moist disposable towelettes (Colgate-Palmolive). These wipes were
selected based on the above criteria. The brands were also selected because they
historically have been used by practitioners in the field for sample collection of settled
leaded dust, and/or are listed in the HUD Guidelines.
5-6
-------
The wipe manufacturers were contacted during the planning phase of the study to
gather information on manufacturers' lot designations and other characteristics for each
brand of wipe. The inquiries were made because several sources involved in lead testing
suggested that potential differences in background lead levels in wipes may exist, since the
chemical constitution of the wipes is not exactly the same.
In an effort to capture data on potential differences among different manufacturers' lots
for wipes of the same brand, the contractor purchased two brands from different
geographical regions of the United States because lot designations were not obtainable
from the manufacturers. One set of Wash a-bye Baby wipes and one set of Wash'n Dri
moist disposable towelettes were purchased in the metropolitan Kansas City area, and these
sets were assigned a lot code of 1. Another set of wipes for these two brands was
purchased in North Carolina, and these sets were assigned a lot code of 2. This approach
was used following one wipe manufacturer's suggestion to purchase the product from
different stores and locations to acquire different lots.
The third brand, Baby Wipes with Lanolin, was a replacement product for a brand
formerly known as Little Ones Baby Wash Cloths, supplied from a mass merchandiser.
The replacement brand was used because the original supplier discontinued distribution of
Little Ones Baby Wash Cloths in many regions of the country. The replacement brand was
selected because it was a brand often used by HUD in public housing dust sampling.
Different lots could not be evaluated since this brand could only be purchased from a single
supplier.
3.4 Standard Reference Material (SRM) Selection
NIST SRM 2704 (Buffalo River Sediment) with a certified lead concentration of
161 jig Pb/g was selected for applying lead onto the wipes. NIST SRM 2704 was selected
because a dust matrix SRM was not available for purchase at the time of the study, and the
schedule for the study did not permit time for preparing evaluation materials for lead in
dust. In addition, a large database of lead percent recovery results for NIST SRM 2704 and
single-wipe samples existed for reference to the results of this study.
The wipes were loaded with NIST SRM 2704 at one of two levels, either 0.50 g or 2 g
per wipe. The resulting lead-loading levels were 80.5 and 322 jig/wipe. The EPA has the
same interim dust clearance standards as the HUD Guidelines, which range from 100 to
800 |ig/ft2, depending on the surface sampled (Interim 403 Guidance, Memorandum of
July 14, 1994, from Lynn Goldman to Regional Division Directors). Assuming a 1 ft2 area
is wiped during sample collection, the lower of the two loading levels selected is
approximately 80% of the 100 |ig/ft2 value, and the higher of the two is in the mid-range of
the interim HUD dust clearance standards. Two loading levels were selected to assess the
impact of the quantity of material applied to the wipe on percent recovery of lead. The
high loading level also provided a dust loading challenge to the methods for all sampling
applications (i.e., risk assessments, lead hazard screens, and clearance testing).
3-7
-------
3.5 Experimental Design
To meet the study objectives, a laboratory experiment was designed based on a
number of parameters. However, due to the exploratory nature of this project, only a small
number of parameters were varied, while other experimental conditions were defined and
held constant. Furthermore, since this investigation only considers the sample preparation
steps for each method, the only relevant parameters are those pertaining to the sample
preparation activities. These parameters can be classified into two major categories:
method and process parameters. A description of the parameters held constant throughout
the experiment, and the method and process parameters that were varied are discussed
next.
Design constants: To minimize the size of the experiment, the following parameters
were held constant throughout the study. Therefore, the effect of these parameters on
method performance was not tested.
A single SRM was used for spiking. NIST SRM 2704 with a concentration of 161
jig Pb/g of material was selected for the purpose of comparing the study results with
past study results.
A single dust-loading procedure (see Appendix G, 3.6) for applying the SRM on the
wipes was followed.
Work at each laboratory was performed by laboratory technicians with comparable
skill levels. It was assumed that the skill level required for the WOHL method was
the same as that required for the EPA SW-846 Method 3050A.
Three commercially available wipe brands were tested: Wash a-bye Baby wipes,
Baby Wipes with Lanolin, and Wash'n Dri moist disposable towelettes. These
brands of wipes were selected because they are commonly used by practitioners in
the field.
Method parameters: The effect of the following parameters (along with their levels,
which are indicated parenthetically) on lead recovery from wipe composites was investi-
gated:
Sample preparation method (2)
Number of wipe(s) per composite (3)
Dust loading (2)
Laboratory (3)
Sample preparation method (2): Two methods were tested: the modified EPA
SW-846 Method 3050A for single-wipe samples, further modified to accommodate two-
and four-wipe composites, and the WOHL methods for single and multiwipe composites.
The modified EPA SW-846 Method 3050A is described in Appendix B. This method,
-------
further modified for two- and four-wipe composites, is described in Appendices C and D,
respectively. The WOHL methods for single and multiwipe composites are provided in
Appendices E and F, respectively.
Number of wipe(s) per composite (3): Three levels were considered in this
exploratory study: a single wipe, a composite of two wipes, and a composite of four wipes.
This applies to each of the three brands of wipes. The single-wipe samples were included
in the study to provide a reference base. The maximum of four wipes per composite was
chosen for practical purposes. In addition, four wipes are the maximum number
recommended for compositing according to the HUD Guidelines.
Dust loading (2): Two dust loading levels were tested: one standard dust material,
NIST SRM 2704 with a lead concentration of 161 jig Pb/g, was spiked onto the wipes in
amounts of 0.5 g/wipe or 2 g/wipe. This resulted in lead loadings of approximately
80.5 jig and 322 jig per single wipe. The maximum composite lead loading was therefore
1,288 jig for the four-wipe composite with a 2 g/wipe SRM loading.
Laboratory (3): Three laboratories were selected to perform the work in an identical
fashion. A contractor prepared and shipped the wipes to two commercial laboratories and
then participated in the study as the third laboratory performing the analyses. This
approach incorporated the variability associated with digestion and analysis procedures
among laboratories during the final data evaluation. A comparison of overall performance
among laboratories per se was not the intent of this study.
Process parameters: To evaluate the effectiveness of using wipe composites,
information on the following process parameters was recorded:
Material costs (e.g., reagents and waste disposal)
Time of sample preparation and relative cost factors
Time of sample analysis
Equipment and facility needs
Information on these process parameters was recorded and was an integral part of the
set of evaluation factors, along with method precision and accuracy which were determined
from the chemical analysis results.
A full factorial experimental design was used to assess the performance of the two
digestion methods. This design was based on the experimental parameters discussed
above. In addition, the following laboratory/technician constraints were carefully
considered:
Each method was tested separately in order to keep different digestion procedures
separate for the laboratory technician.
Only one type of composite was handled in a given batch. This facilitated the
addition of different volumes of reagents to different composite types.
3-9
-------
Batch size considerations included the number of beakers that could fit on the hot
plates. For example, Figures 3-2 and 3-3 show the number of beakers that could fit
on a single hot plate (1 ft x 2 ft) for the modified EPA SW-846 Method 3050A and
the WOHL method, respectively.
All samples within a given hood would be handled by a single laboratory
technician.
Each batch included the required number and type of QC samples, as described in
Section 6.
Number of replicates: Considering the three experimental parameters (sample
preparation method, number of wipes per composite, and dust loading) and their respective
levels, a total of 2 x 3 x 2 = 12 unique combinations were tested for each of the three types
of wipes at the three laboratories. To assess method variability, a number of replicate runs
were performed. The number of replicates were two or four, depending on the method
used and the number of wipes per composite. Tables 3-1 and 3-2 provide the complete
layout for the two methods. The number of replicates in each case was guided by the size
of each batch and a minimum of two replicate runs for each unique combination of
parameters. The size of digestion beakers and hot plates determined how many beakers
could fit on one hot plate (1 ft x 2 ft). Two hot plates could be used simultaneously in one
hood; therefore, one sample batch consisted of samples processed by a single technician
using one or two hot plates (see Figures 3-2 and 3-3).
Lots within wipe brands: As mentioned earlier (see Section 3.3), wipes from two
different lots were only obtained for two of the three wipe brands (i.e., because one brand
was only available from one location). This was done to account for potential lot-to-lot
variability. This decision did not affect the design layouts shown in Tables 3-1 and 3-2.
The wipes that were loaded with SRM were selected from the two lots in a balanced
fashion; that is, equal numbers of wipes from each lot were assigned to the two loading
treatments (0.5 g and 2 g). In addition, all wipe samples in a given composite sample
consisted of wipes from a same lot. Wipe blanks were randomly selected from the two lots
so as to obtain approximately equal numbers of blank wipes of each lot.
The three brands of wipes were selected based on the HUD Guidelines and other
criteria listed in Section 3.3. These three brands of wipes were included in the study to
take into account variability among wipes. The intent was not to compare the brands but to
include a replication scheme encompassing replicates within a brand and across qualifying
brands available on the market. Therefore, for analysis purposes, the type of wipe was
considered a replicate level rather than a parameter in the design. In summary, from
Tables 3-1 and 3-2, the following number of samples were prepared for each laboratory,
requiring the following number of wipes per laboratory:
5-10
-------
Method/QC
EPA SW-846 Method 3050A
QC
Total
WOHL Method
QC
Total
Grand total
Number of samples
60a
24C
84
72a
18C
90
174
Number of wipes
156b
33d
189
168b
21d
189
378
This number is the sum of the 3 entries in row r5 in Table 3-1 or Table 3-2,
respectively, for the two methods.
This number is the sum of the 3 entries in row r6 in Table 3-1 or Table 3-2,
respectively, for the two methods.
This number is the sum of the 3 entries in row r15 in Table 3-1 or Table 3-2,
respectively, for the two methods.
This number is the sum of the 3 entries in row r17 in Table 3-1 or Table 3-2,
respectively, for the two methods.
This experimental design applied to all three laboratories participating in the study.
Each laboratory followed identical digestion and analysis procedures. Aliquots of each
prepared sample were analyzed at each laboratory by both ICP and FAA.
3-11
-------
250 mL
Griffin Beakers
400 mL
Griffih Beaker
r
800 mL
Griffin Beaker
Figure 3-2. Hot Plate Configuration for the Modified EPA SW-846 Method 3050A
3-12
-------
250 mL
Phillips Beaker
Single Wipe,
2-Wipe Composite, -*-
or 4-Wipe Composite
Figure 3-3. Hot Plate Configuration for the Wisconsin Occupational Health
Laboratory (WOHL) Method
3-13
-------
Table 3-1. Design Layout for the Modified EPA SW-846 Method 3050A
Row No.
r1
r2
r3
r4
r5
r6
r7
r8
r9
r10
r11
r12
r13
r14
r15
r16
r17
r18
r19
SRM-loaded wipe samples
Number of wipes per composite
Type of wipe (brand)
SRM loading per wipe
Number of replicate runs
Number of composite wipe samples (r2xr3xr4)
Number of wipes (r5xr1)
1 wipe
A, B, and C
0.5 g and 2 g
2
12
12
2-wipe composite
A, B, and C
0.5 g and 2 g
4
24
48
4-wipe composite
A, B, and C
0.5 g and 2 g
4
24
96
Laboratory configuration
Size of digestion beaker
Maximum number of beakers per hot plate
Number of hot plates needed
Number of batches (maximum 2 hot plates/batch)
250 mL
21
1
1
400 mL
18
2
1
800 mL
10
4
2
Quality control samples per batch (see Section 6)
Method blank (MB)
Wipe blank (WB)
Laboratory control sample (LCS)
Number of QC samples per batch (r11+...+r13)
Total number of QC samples (r1 Oxr1 4)
Number of QC wipes per batch (r1xr12)
Total number of QC wipes (r1 6xr1 0)
1a
A, B, and C
2b
6
6
3
3
1
A, B, and C
2
6
6
6
6
1
A, B, and C
2
6
12
12
24
Number of wipes and samples (analyses)
Total number of wipes needed (r6+r17)
Total number of samples (i.e., beakers)
r5+r15
15
18
54
30
120
36
a One laboratory prepared one extra method blank. See Section 6.2.2.
b One laboratory prepared two extra laboratory control samples. See Section 6.2.2.
-------
Table 3-2. Design Layout for the Wisconsin Occupational Health Laboratory (WOHL) Method
Row No.
r1
r2
r3
r4
r5
r6
r7
r8
r9
r10
r11
r12
r13
r14
r15
r16
r17
r18
r19
SRM-loaded wipe samples
Number of wipes per composite
Type of wipe (brand)
SRM loading per wipe
Number of replicate runs
Number of composite wipe samples (r2xr3xr4)
Number of wipes (r5xr1)
1 wipe
A, B, and C
0.5 g and 2 g
4
24
24
2-wipe composite
A, B, and C
0.5 g and 2 g
4
24
48
4-wipe composite
A, B, and C
0.5 g and 2 g
4
24
96
Laboratory configuration
Size of digestion beaker
Maximum number of beakers per hot plate
Number of hot plates needed
Number of batches (maximum 2 hot plates/batch)
250 mL
21
2
1
250 mL
21
2
1
250 mL
21
2
1
Quality control samples per batch (see Section 6)
Method blank (MB)
Wipe blank (WB)
Laboratory control sample (LCS)
Number of QC samples per batch (r11+...+r13)
Total number of QC samples (r1 Oxr1 4)
Number of QC wipes per batch (r1xr12)
Total number of QC wipes (r1 6xr1 0)
1a
A, B, and C
2b
6
6
3
3
1
A, B, and C
2
6
6
6
6
1
A, B, and C
2
6
6
12
12
Number of wipes and samples (analyses)
Total number of wipes needed (r6+r17)
Total number of samples (i.e., beakers)
r5+r15
27
30
54
30
108
30
a One laboratory prepared one extra method blank. See Section 6.2.2.
b One laboratory prepared two extra laboratory control samples. See Section 6.2.2.
-------
Section 4
Sample Handling and Identification
This section addresses the aspects of loading the wipes with NIST SRM, laboratory
sample handling, identification, containers, storage, and shipping. The sample handling
procedure was carried out in accordance with the sequence specified by the test design
outlined in Section 3. All samples analyzed in this investigation were either wipes spiked
with a NIST SRM or various QC samples. All wipe samples were loaded with NIST 2704
(Buffalo River Sediment) at the contractor laboratory.
4.1 Wipe Loading Procedure
A planning session was held with project personnel to determine an efficient order, or
loading sequence, that would reduce the chance for error in loading the wipes. This
planning session was important because approximately 1,000 wipes were loaded with
NIST SRM for the study. The test design was then sorted into the optimum order, and the
bar code numbers were sequentially assigned. NIST SRM 2704 (Buffalo River Sediment)
with a certified lead concentration of 161 jig Pb/g was used for loading lead onto the
wipes.
A Mettler Model AE200 analytical balance was used for weighing the NIST SRM
aliquots that were applied to the wipes. The wipes were loaded with NIST SRM 2704 at
one of two levels, either 0.50 g or 2 g per wipe, as specified in the test design. The
resulting lead-loading levels were 80.5 and 322 jig Pb/wipe. The weighed portion of
reference material was placed on the wipe, the wipe was folded, and the SRM was pressed
into the wipe as it was folded, and the loaded wipes were transferred to labeled Nalgene
bottles for storage until sample preparation. The specific procedure used for loading the
reference material onto the wipe(s) is presented in Appendix G.
The samples were randomly assigned to the three laboratories. This was done within
the strata defined by SRM levels, composite types, wipe brands, digestion methods, and
number of replicates. As a result, the make-up of samples with respect to these factors was
identical for the three laboratories. The samples that were loaded with NIST SRM 2704
were segregated into the appropriate batches, incorporating the required number of wipe
blanks into each batch, as specified in the test design. The samples were batched and
placed in secure storage, pending transfer to the laboratories for sample digestion and
analysis. Each laboratory participating in the study was responsible for incorporating the
method blank and laboratory control samples into the batches as specified in the digestion
procedures. The resulting wipe samples were digested using either the modified EPA
SW-846 Method 3050A or the WOHL method.
4-1
-------
4.2 Sample Identification
Each sample generated for the study was assigned a unique bar code identifier for
laboratory processing. The bar code identifier was correlated to components of the
experimental design as presented below:
Laboratory code (Nos. 1, 2, and 3)
Preparation method (EPA 3050A and WOHL)
Type of composite (one-, two-, and four-wipe composite)
Sample type (loaded wipe, QC sample type: laboratory control sample, method
blank, wipe blank)
Nominal SRM loading (0.5 g or 2 g per wipe)
Wipe brand (A, B, or C)
Lot code (1 and 2 for brands A and B only)
4.3 Sample Containers, Storage, and Shipment
As shown in Table 4-1, the sample containers used for storing the wipe samples were
wide-mouth, high-density polyethylene Nalgene bottles. These containers were selected
based on the volume capacity needed for the two- and four-wipe composite samples. A
hard-walled container was used instead of a plastic bag for sample storage because this
type of container can be rinsed to ensure a quantitative transfer of the dust sample to the
digestion vessel.
Table 4-1. Wipe Sample Containers
Number of
wipes per sample
1
2
4
Nalgene
bottle size
125 ml
250 ml
250 ml
The wipe samples and resulting digests were stored at room temperature. The holding
time from the date of loading the wipes with NIST SRM 2704 through instrumental
analysis was 6 months, as specified in EPA SW-846 for metals analysis. The holding time
was met for all analyses.
The wipe samples were packaged and shipped to the designated laboratories via next
day delivery. Each laboratory participating in the study was provided with the
documentation for each method and a work plan describing the project overview, scope of
work, schedule, and required deliverables for the study.
4-2
-------
Section 5
Sample Digestion and Analysis
The sample preparation and analysis efforts were conducted according to the methods
listed in the Quality Assurance Project Plan (QAPjP).7 The sample preparation methods
are presented in the appendices, as outlined in Section 1.4. The digestion methods were
evaluated using two different analytical methods each. The first analytical method (EPA
SW-846 Method 6010A) determines metals in the digestates using ICP. The determination
of lead by direct aspiration FAA (EPA SW-846 Methods 7000A [general] and 7420) was
the second analytical method.
5.1 Instrumentation
The spectrometers of each laboratory used for the ICP and FAA analyses are listed
below.
ICP Analysis Spectrometers
Laboratory 1: Thermo Jarrell Ash ICAP 6 IE simultaneous vacuum path
spectrometer with computer-controlled background and interelement corrections
and an analytical wavelength of 220.353 nm.
Laboratory 2: Perkin Elmer Optima 3000 XL ICP simultaneous nitrogen purged
path spectrometer with computer-controlled background correction and an
analytical wavelength of 220.353 nm.
Laboratory 3: Thermo Jarrell Ash ICAP 61 simultaneous air path spectrometer with
computer-controlled background and interelement corrections and an analytical
wavelength of 220.353 nm.
FAA Analysis Spectrometers
Laboratory 1: ARL GBC Model 902 FAA Spectrometer with deuterium
background correction and an analytical wavelength of 283.3 nm.
Laboratory 2: Thermo Jarrell Ash Video 12 E FAA Spectrometer with
Smith-Hieftje background correction and an analytical wavelength of 217.0 nm.
Laboratory 3: Perkin Elmer 5000 FAA Spectrometer with deuterium background
correction and an analytical wavelength of 283.3 nm.
5-1
-------
5.2 Calibration Standard Sources
The standard stocks used for the analyses at each laboratory were purchased from
commercial suppliers. Working calibration standards were prepared by dilution of the
concentrated calibration stocks. The standards used for instrument calibration (primary
source) and the independent check standard (secondary source) were either prepared from
stocks purchased from different suppliers, or were prepared from different lots of stocks
purchased from the same supplier. The sources of the standard stocks used for the analyses
at each laboratory are listed below. Spex Industries, Inc., is listed twice under instrument
calibration standard sources because this supplier was used by two of the three laboratories.
Instrument Calibration Standard Sources (primary)
Spex Industries, Inc.
Spex Industries, Inc.
Plasmachem Associates, Inc.
Independent Check Standard Sources (secondary)
Inorganic Ventures, Inc.
Spex Industries, Inc.
Fisher Scientific Chemical Division
5.3 Laboratory Observations
The following observations were made by the laboratories conducting the sample
digestion procedures:
Although the hot plate time was low for the WOHL digestion methods, these
methods were labor-intensive due to several steps in the procedure (e.g., cutting up
the wipes, centrifuging, and filtering). The 3050A method had more hot plate time
and seemed to provide a more complete digestion of the wipe material.
The larger volume beakers and volumetric flasks, required for the modified SW-846
Method 3050A for the four-wipe composites, were difficult to handle in the
laboratory and also were more difficult to clean.
Some of the SRM material adhered to the scissors during the cutting step of the
WOHL method. Attempts were made to recover and quantitatively transfer this
material to the digestion vessel by wiping the scissors with a small piece of clean
wipe material contained in the sample.
An adequate rinse could not be obtained for the WOHL single-wipe method due to
the small final dilution volume (50 mL) used for this method.
Additional observations made regarding the digestion time requirements and material
costs for each digestion procedure are presented in Section 7.3 (Process Parameter Data).
5-2
-------
Section 6
Quality Assurance/Quality Control
An independent evaluation of the sample preparation and analysis results was
performed under the direction of the program QA Officer (QAO). In addition, a technical
review of the data reported by the outside laboratories was conducted by the task leader,
and the final QC data were statistically analyzed by the project statistician.
The results are presented in the following subsections.
6.1 Quality Assurance
Performance audits and complete systems audits were not conducted at all three
laboratories because of their involvement with the NLLAP. The ELPAT analysis was
considered to be equivalent to a performance audit, thus additional performance audits
were deemed unnecessary. The benefit of conducting an on-site systems audit at each of
the laboratories was not considered to outweigh the associated travel costs; however, a
partial systems audit was conducted on work performed by the contractor laboratory in
which all the records were examined.
A complete list of all sample identification and calculated recovery results is shown in
Appendix H. These data pertain to all QC samples as well as all loaded wipe samples used
during the project. The hard copy and computer records for the work performed by the
contractor laboratory, including the tables for the final report, were audited. All sample
identification codes and data were verified throughout the data handling process; no errors
were found. The items verified during the audit are listed below.
Accuracy and completeness of the data packet.
Completeness, compliance, and accuracy of the laboratory notebook pages,
including the method of analysis, the project number, date of analysis, the analyst's
name, the standards used for calibration, the steps performed in the preparation of
the standards, and the dilutions.
Accuracy of the value used for NIST SRM 2704, based on the Certificate of
Analysis information.
Accuracy of the sample identification codes in the report tables, using the sample
preparation inventory listing from the test design.
Accuracy of the instrument responses presented as |ig/mL for ICP and peak height
absorbance for FAA in the data calculation tables, using the instrumental raw
analysis data.
6-1
-------
Accuracy of the Initial Calibration Verification (ICV) control charts, using the raw
data and data calculation tables.
Accuracy of the computer entry of the weight data and the analytical results in the
SAS computer output file.
The data reports provided by the outside laboratories also were examined. Several
discrepancies were noted for Laboratory 2 between the results reported in the QC summary
tables, the control chart data report, the detection limit interpretation, and the results. The
FAA and TCP data columns were interchanged in several preliminary data tables, and a few
data discrepancies also were noted. These problems were resolved during the technical
review and data audit, and a revised report was submitted. The reported sample
identification codes and data for the outside laboratories were verified after entry into the
database.
The only deviation from the QAPjP was one laboratory's use of a smaller than
specified capacity (rpm) centrifuge for the WOHL methods. The purpose of the centrifuge
was to eliminate particles that would clog the nebulizer during analysis. The centrifuge
used was set at approximately 4,000 rpm, which was adequate for this purpose, and the
deviation did not impact data quality. The deviation was properly documented in the
records.
6.2 Quality Control
This section presents the analyses of the gravimetric quality control results; the
analytical quality control results for method blanks, wipe blanks, and laboratory control
samples; the spiked wipe extreme recovery values; and the detection limits.
6.2.1 Gravimetric Data
NIST SRM 2704 was spiked at two nominal dust levels onto 972 single wipes: low
(0.5 g) and high (2 g). Actual SRM weights were measured for each individual wipe. The
actual weight of each single wipe and each composite wipe sample was then compared to
the nominal levels, for a total of 396 comparisons. The discrepancies between actual and
nominal dust levels were computed and compared against the specifications stated in the
QAPjP: 0.5 ± 0.05 g for the 0.5-g loading and 2.0 ± 0.1 g for the 2-g loading. The
discrepancies between actual and nominal dust levels are shown in the form of box (or
box-and-whisker) plots in Figure 6-1. The data distributions are shown separately for each
nominal level on a per single-wipe or composite basis (i.e., 0.5 g, 1 g, 2 g, 4 g, and 8 g).
Since the loading of 2 g on a single-wipe sample is the same as the total loading of four
times 0.5 g on a four-wipe composite sample, the results were further separated into the
6-2
-------
two loading levels, low and high, in Figure 6-1. All box plots in this report were generated
using a commercially available software package.8
#000 o
Figure 6-1. Distribution Plots of Gravimetric Data for Single Samples and
Composites
Following is a description of how to read the box plots in Figure 6-1:
The left and right edges of a box are located at the sample 25th and 75th
percentiles; thus 50% of the data fall within the box. The center vertical line is
drawn at the sample median. The horizontal lines, called "whiskers," extend from
the box to a distance of at most 1.5 interquartile ranges. (An interquartile range is
the distance between the 25th and the 75th percentiles, the length of the box.)
Thus, approximately 99% of the values will fall between the left- and right-side
ends of the whiskers. Any value more extreme than this is marked with an
asterisk "*" if it is within three interquartile ranges of the box, or with a small
letter "o" if it is still more extreme.
As shown in Figure 6-1, the accuracy of weighing out the nominal levels decreases
with increasing SRM loading at a given loading level because the errors in weighing are
additive as the individual wipes are composited. The weight discrepancies exhibit a slight
positive bias. However, the specifications stated above were only exceeded on a total of
6-3
-------
8 out of 972 individual wipes. After compositing, the specifications were not met for only
1 out of 396 composited wipe samples. The amount of SRM applied to Sample
No. 906423 (a two-wipe composite at the high SRM level) was 4.2803 g, when it should
have been 4.0 ± 0.2 g. These results confirmed that the weighing procedure achieved the
target accuracy called for in the QAPjP.
6.2.2 Quality Control Sample Data
Three types of quality control (QC) samples, for a total of 132 samples, were analyzed
by both analytical methods (FAA and TCP) along with the composite wipe samples accord-
ing to the QAPjP (see Section 4.1.3 of QAPjP). These included:
23 method blanks (MBs) to demonstrate absence of laboratory contamination
63 wipe blanks (WBs) to determine background levels, if any, in the wipe material
46 laboratory control samples (LCSs) to monitor method performance in the
absence of any wipe matrix effect
Detailed analytical results for these 132 QC samples are included in Appendix H. It
should be noted that Laboratory 1 analyzed more QC samples than the other two
laboratories because three samples that had been diluted or spilled during the original
preparation were discarded and new samples were prepared. This situation resulted in two
additional batches, one each for the EPA Method 3050A and WOHL single-wipe batch
types.
Method Blanks. One method blank was analyzed with each sample batch to monitor
background levels in the reagents used for digestion and to check for potential laboratory
contamination. The QC sample layout and analytical results are shown in Table 6-1. The
last two columns provide the results (or range of results) for method blanks when
normalized to jig/wipe. Normalized results for each batch type were calculated by
multiplying the analytical result (|ig/mL) by the digest volume (mL) used for that
method/batch type, then dividing by the number of wipes of the corresponding batch. If the
analytical result was below the instrument detection limit, then the instrument detection
limit (|ig/mL) was multiplied by the digest volume and then divided by the number of
wipes, and the result was flagged as less than (i.e., <). The digest volume varied by
method and composite type.
6-4
-------
Table 6-1. Method Blank Sam
Laboratory
code
1
2
3
Digestion
method
EPA
3050A
WOHL
EPA
3050A
WOHL
EPA
3050A
WOHL
Composite
batch type
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
pie Layout and Analytical Results
Number
of
samples
2
1
2
2
1
1
1
1
2
1
1
1
1
1
2
1
1
1
Number of
detects found
using
FAA
0
0
0
0
0
0
1
1
1
0
1
0
0
0
0
0
0
0
ICP
0
0
0
0
0
0
1
1
2
1
1
1
0
0
0
0
0
0
Normalized
results
(ug/wipe)a
FAA
<9
<9
< 12
<3
<6
<3
5
2
1.3
<1
1.5
<0.3
<4
<4
<5
<2
<2
< 1
ICP
<6
<6
<7
<2
<1
<0.4
4
5
5b
2
1.5
0.8
<10
<10
< 13
<5
<5
<2.5
a The normalized results indicate detection limits or actual levels detected.
b This result is the average of 2 identical values of 20 ug for 4 wipes, normalized to
5 ug/wipe.
Of the three laboratories, only Laboratory 2 reported lead levels above the detection
limit in the method blanks. Lead levels were detected by ICP in all seven method blanks
(four digested by the EPA Method 3050A and three digested by the WOHL method).
When FAA was used, lead levels were detected in four of the seven method blanks (three
digested by the EPA 3050A and one digested by the WOHL method).
Laboratory 2 did not use the same approach in establishing detection limits as did
Laboratories 1 and 3. This situation was noted during the evaluation of the method blank
and wipe blank data; however, as discussed later in Section 6.3.2 (Detection Limits), this is
not considered a problem due to the low levels detected.
6-5
-------
Wipe Blanks. Wipe blanks were used to determine the background levels, if any, in
the wipe material. One blank wipe of each brand was included in each preparation batch,
according to the test design. The blank wipes were analyzed by both FAA and TCP.
Table 6-2 summarizes the layout and analytical results for each digestion procedure
and analytical method. The last two columns provide the results (or range of results) for
the wipe blanks, when normalized to jig/wipe, based on the number of wipes contained in
the sample.
A number of wipes were found to have lead levels above the detection limit. A
summary of the number of wipe blank detects for each brand of wipe is presented in
Table 6-3.
To investigate whether wipe brands vary in their number of wipe blank detects and
nondetects, Fisher's exact test was applied to the four 3x2 contingency tables defined by the
combinations of digestion and preparation methods and to the two 3x2 contingency tables
defined by each analytical method, as shown in Table 6-3. The statistical test results are
shown in the last row of Table 6-3. Since none of the tests was significant at the two-sided
95% confidence level, there is no significant correlation between the number of wipe blank
detects and wipe brand.
Of the 40 detects out of a total of 126 analyses shown in Table 6-3, 33 were attributed
to Laboratory 2. The number of detects can be broken down by laboratory and analytical
method as follows:
Laboratory 1 0/21 detect by FAA
2/21 detects by ICP and WOHL
Laboratory 2 13/21 detects by FAA (5 by EPA 3050A, 8 by WOHL)
20/21 detects by ICP (12 by EPA 3050A, 8 by WOHL)
Laboratory 3 4/21 detects by FAA (3 by EPA 3050A, 1 by WOHL)
1/21 detect by ICP and WOHL
As noted earlier for method blanks, a disproportionate number of detects on the wipe
blanks were attributed to Laboratory 2. The detects are discussed later in Section 6.3.2,
Detection Limits.
Laboratory Control Samples (LCSs). LCSs were used to monitor the method
performance in the absence of wipe matrix effects. These samples were prepared at the
same NIST SRM 2704 loading levels as the wipe samples for each composite batch type
(i.e., one low and one high level). For example, two-wipe composites required a low level
loading of 0.5 g/wipe (total of 1.0 g) and a high level loading of 2.0 g/wipe (total of 4.0 g).
The two LCSs for this composite type batch consisted of 1.0-g (low) and 4.0-g (high)
aliquots of NIST SRM 2704, spiked into empty beakers (i.e., no wipe). The low and high
6-6
-------
LCSs were digested with each batch of samples. The sample layout for these LCSs is
shown in Table 6-4.
6-7
-------
Table 6-2. Wipe Blank Sample Layout and Analytical Results
Laborato-
ry code
1
2
3
Diges-
tion
method
EPA
3050A
WOHL
EPA
3050A
WOHL
EPA
3050A
WOHL
Compos-
ite type
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
Number
of
samples
3
3
6
3
3
3
3
3
6
3
3
3
3
3
6
3
3
3
Number of
detects found
using
FAA
0
0
0
0
0
0
1
1
3
2
3
3
1
0
2
0
0
1
ICP
0
0
0
0
0
2
3
3
6
3
2
3
0
0
0
0
0
1
Normalized results
(ug/wipe)a
FAA
<9
<9
< 12
<3
<6
<3
< 1 to 3
< 0.5 to 6
< 0.25 to 4
< 1 to 2
1 to 2
0.3 to 1
<4to5
<4
< 5 to 7. 5
<2
<2
< 1 to 4.8
ICP
<6
<6
<7
<2
<1
< 0.4 to 1.2b
4 to 5
4 to 5
4 to 5
1 to 2
0.5
0.5 to 3
<10
< 10
<13
<5
<5
< 2. 5 to 4. 3
a The normalized results indicate detection limits or actual levels detected.
b The range of results obtained is reported for several method/composite types. The
actual values, prior to normalization, are provided in Appendix H.
Table 6-3. Number of Wipe Blank Detects by Wipe Brand
Wipe brand
A
B
C
All brands
Fisher's
exact test:
p-value
FAA method
EPA
3050A
1/12a
2/12
5/12
8/36
0.21
WOH
L
2/9
3/9
4/9
9/27
0.87
Both
3/21
5/21
9/21
17/63
0.14
ICP method
EPA
3050A
4/12
4/12
4/12
12/36
1.0
WOH
L
2/9
4/9
5/9
11/27
0.49
Both
6/21
8/21
9/21
23/6
3
0.72
All
9/42
13/42
18/42
40-
/126
Number of detects/number of samples analyzed.
6-8
-------
Table 6-4. Laboratory Control Sam
Laboratory code
1
2
3
Digestion
method
EPA 3050A
WOHL
EPA 3050A
WOHL
EPA 3050A
WOHL
Nominal
SRM loading (g)
0.5
1.0
2.0
4.0
8.0
0.5
1.0
2.0
4.0
8.0
0.5
1.0
2.0
4.0
8.0
0.5
1.0
2.0
4.0
8.0
0.5
1.0
2.0
4.0
8.0
0.5
1.0
2.0
4.0
8.0
pie Layout
Nominal
Pb loading (|jg)
80.5
161
322
644
1,288
80.5
161
322
644
1,288
80.5
161
322
644
1,288
80.5
161
322
644
1,288
80.5
161
322
644
1,288
80.5
161
322
644
1,288
Number of
samples
2
1
4
1
2
2
1
3
1
1
1
1
3
1
2
1
1
2
1
1
1
1
3
1
2
1
1
2
1
1
The data quality objective for these samples was to obtain a recovery of 80% to 120%.
Percent recovery statistics for the 46 LCSs are plotted in Figures 6-2 and 6-3 for FAA and
TCP, respectively. Within each figure, the results are shown separately for the two
digestion methods, using the laboratory code to represent the data.
All results obtained by FAA fall within the range of 80% to 120% recovery. When the
samples were analyzed by TCP, all recovery results obtained from the samples digested by
the EPA 3050A method met the recovery objectives. Of the results obtained by TCP on the
samples digested by the WOFIL method, three results (two from Laboratory 3 and one from
Laboratory 2) fell below the lower limit of 80%. Two of these results were slightly above
79%, while the third result was approximately 73%. The plot at the bottom of Figure 6-3
shows a decreasing trend of percent recovery with increasing SRM for ICP results when
6-10
-------
using the WOHL digestion method. This estimated trend is statistically significant
(p = 0.06) with a decrease of 1.75% recovery for each gram of SRM (slope = 1.75), and
6-11
-------
FAA Recovery (%)
CD O -* N
0 O O C
[
FAA Laboratory Control Sample Results
using EPA 3050A Method
1 -
1
i i i i i i i i
1123456785
Actual SRM Loading (g)
Code = Laboratory code
)
£ 110
0)
8 100
CL
ft 90
C
FAA Laboratory Control Sample Results
using WOHL Method
_ 3 1 ? a 3
1 ^
232 ^ 2
"
i i i i i i i i
) 1 2 3 4 5 6 7 8 £
Actual SRM Loading (g)
Code = Laboratory code
)
Figure 6-2. LCS Percent Recovery by FAA for EPA 3050A and WOHL Methods
6-12
-------
o:
o_
o
130
120
1 10
100
90
80
70
- 2
ICP Laboratory Control Sample Results
using EPA 3050A Method
4 5
Actual SRM Loading (g)
Code = Laboratory code
ICP R
130
120
-*
O
O
O
CD
O
80
70
ICP Laboratory Control Sample Results
using WOHL Method
4 5
Actual SRM Loading (g)
Code = Laboratory code
Figure 6-3. LCS Percent Recovery by ICP for EPA 3050A and WOHL Methods
6-13
-------
may be a consideration in sampling applications where high dust loading levels are
expected, such as for window sills and troughs in risk assessment.
LCS recovery statistics are further summarized in Table 6-5, separately for each
laboratory, for each of the four digestion and analytical method combinations, across SRM
levels. The precision, expressed as relative standard deviation, was less than 13% in all
cases, ranging from approximately 2% to 7% for FAA and from 2% to 12% for ICP.
Table 6-5. LCS Percent Recovery Statistics
Digestion
method
EPA 3050A
WOHL
Laboratory
code
1
2
3
1
2
3
Number of
samples
10
8
8
8
6
6
FAA recovery (%)
Mean
96.99
98.99
100.23
98.15
93.62
100.31
RSDa
6.97
3.73
4.59
3.11
1.90
3.56
ICP recovery (%)
Mean
94.96
105.67
88.94
90.24
100.02
82.20
RSDa
6.02
1.95
3.97
4.13
12.10
7.18
Relative standard deviation: RSD (%) = 100 x standard deviation/mean.
6.3 Other Data Quality Indicators
6.3.1 Extreme Percent Recovery Data
Percent recovery data were calculated for each spiked wipe composite sample
separately for the FAA and ICP methods as:
fo/. Amount Found (ug)
Recovery (%) = ^-^L
Amount Spiked (jig)
100
Figures 6-4 and 6-5 show the recovery data distribution in the form of box plots, for
FAA and ICP, respectively. The data (box plots) within each figure are organized from the
bottom up by laboratory, digestion method, and number of wipes per composite. For
example, 1-3050A2 denotes Laboratory 1, digestion Method 3050A, and a two-wipe
composite. Any value falling between 1.5 and 3 times the interquartile range from the
outside of the box is marked with an asterisk "*", or with a small letter "o" if it is still more
extreme. The 11 FAA recovery values and the 6 ICP recovery values marked as either "*"
or "o" were further investigated and are summarized in Table 6-6. The gravimetric and
analytical data for each sample were verified, and no data reporting errors or anomalies
were noted. As a result, the data were not discarded from the statistical analyses.
6-14
-------
ffi
0
a)
o #
Figure 6-4. Distribution Plots of FAA Percent Recovery Data by Laboratory,
Digestion Method, and Type of Composite
6-15
-------
0
05
a
0
fo/
Figure 6-5. Distribution Plots of ICP Percent Recovery Data by Laboratory,
Digestion Method, and Type of Composite
6-16
-------
Table 6-6. Wipe Samples with Extreme Recovery Results
Sample
bar code
Laborator
y code
Digestion
method
Composit
e type
Wipe
brand
FAA
recovery
(%)
ICP
recovery
(%)
Samples with Recoveries Flagged as "*"
906223
906296
906179
906350
906279
906101
906354
906389
906224
906295
906351
906372
906077
1
2
3
1
2
WOHL
WOHL
WOHL
3050A
WOHL
3050A
3050A
3050A
WOHL
WOHL
3050A
3050A
3050A
4-wipe
4-wipe
2-wipe
2-wipe
1-wipe
2-wipe
2-wipe
4-wipe
4-wipe
4-wipe
2-wipe
4-wipe
4-wipe
B
B
B
C
B
A
C
C
B
B
C
C
A
62.41
62.89
101.86
110.73
84.26
110.68
110.20
118.03
41.70
51.56
113.57
112.26
111.52
Samples with Recoveries Flagged as "o"
906224
906295
906333
906357
1
3
WOHL
WOHL
3050A
3050A
4-wipe
4-wipe
1-wipe
2-wipe
B
B
C
C
46.18
56.94
121.96
124.01
6-17
-------
In addition, 9 of the 17 samples yielded recoveries within the 80% to 120% data objective.
Figures 6-4 and 6-5 also show that a considerable portion of recoveries are below the 80%
lower limit for Laboratories 1 and 3 when using the WOHL method and either analytical
technique (FAA or ICP) for four-wipe composites.
The difference between FAA and ICP percent recoveries was calculated for each
individual sample to compare the two analytical methods. These data are shown in the box
plots in Figure 6-6. These data are organized in the same fashion as Figures 6-4 and 6-5
and provide a means of comparing the two analytical methods on a per sample basis.
6.3.2 Detection Limits
As mentioned earlier, Laboratory 2 found more method blank and wipe blank detects
than the other two laboratories. Laboratory 2 did not use the same approach in establishing
detection limits as did Laboratories 1 and 3. The study only allowed for incorporating
variability in digestion and analysis procedures among laboratories into the final data
evaluation. A comparison of overall performance among laboratories per se was not the
intent of this study.
Several reasons why detection limits can vary among laboratories are given:
The analytical methods used do not elaborate on how to determine a detection limit;
therefore, detection limits can be estimated using different procedures
The use of different instrumentation may provide different detection limits
Even the same instrumentation will provide variable detection limits over time; for
example, many standard procedures specify that instrument detection limits should
be determined every three calendar months
The use of different analytical parameters, such as the analytical wavelength, can
provide different detection limits from the same instrument (e.g., for FAA,
Laboratory 2 used a more sensitive wavelength than Laboratories 1 and 3)
Of the detection limits reported by the three laboratories, the highest detection limits
reported for FAA and ICP were 12 jig/wipe and 13 jig/wipe, respectively. In each case, the
highest detection limits reported were those for the EPA 3050A digestion method and the
four-wipe composite sample type. This digestion method/composite type combination
yielded the highest method detection limits (MDL) because the MDL is dependent on the
final digest volume, as discussed in Section 6.2.2.
The EPA has the same interim dust clearance standards as the HUD Guidelines, which
range from 100 to 800 |ig/ft2, depending on the surface sampled (Interim 403 Guidance,
Memorandum of July 14, 1994, from Lynn Goldman to Regional Division Directors).
6-18
-------
Figure 6-6. Distribution Plots of FAAICP Percent Recovery Data by Laboratory,
Digestion Method, and Type of Composite
6-19
-------
Assuming a 1-ft2 area is wiped during typical sample collection, then the lowest clearance
standard can be calculated as 100 jig/wipe. From the results reported by the three
laboratories, both digestion methods tested provided sufficient sensitivity for the purpose
of clearance applications.
Appendix 13.1 of the HUD Guidelines states that "Blank media should contain no
more than 25 jig/wipe (the detection limit using Flame Atomic Absorption)."
Appendix 13.1 also specifies that wipe media should contain less than 5 jig/wipe. As
discussed above, some of the detection limits reported by the laboratories in this study were
greater than 5 jig/wipe. For the methods with adequate sensitivity at the 5 jig/wipe level,
the highest levels of lead detected in the wipe blanks above 5 jig/wipe were the following:
Laboratory 2, Method 3050A, FAA analysis, two-wipe composite: 6 jig/wipe
Laboratory 3, Method 3050A, FAA analysis, four-wipe composite: 7.5 jig/wipe
These values are only slightly above the criterion of 5 jig/wipe. Overall, approximately
68% (86 out of 126) of the wipe blanks were found to have levels below the detection
limit. However, Laboratory 2 only found 21% (9 out of 42) of the wipe blanks to have
levels below the detection limit. In contrast, Laboratories 1 and 3 combined found 92%
(77 out of 84) of the wipe blanks to have levels below the detection limit. It was noted that
the wipe blank levels reported as detects by Laboratory 2 were in most cases at or below
the detection limits reported by Laboratories 1 and 3. More sensitive analytical techniques,
such as graphite furnace atomic absorption spectrometry, could be used for screening blank
media. This technique would provide more quantitative data at the 5 jig/wipe level.
The wipe blank results reported by all three laboratories also were evaluated using a
worst-case scenario (i.e., using the maximum detection limit reported for each digestion
method, composite type, and analytical method combination). If the maximum detection
limits are used for the comparison, then the highest values detected are the wipe blanks for
the WOHL four-wipe composites listed below:
Laboratory 2: ICP at 3.0 jig/wipe
Laboratory 3: ICP at 4.3 jig/wipe
Laboratory 3: FAA at 4.8 jig/wipe
These values were all less than 5 jig/wipe.
Based on this evaluation, blank correction was not used because all three brands of
wipe media used in this study were found to have a minimal number of detects and, in
those instances where detects were reported, the lead levels were negligible. For future
work, it is recommended that detection limits be established by the same procedure for all
laboratories so that the results are more comparable.
6-20
-------
Section 7
Statistical Analysis and Results
In accordance with the study design described in Section 3, two types of data were
obtained from this study: analytical results, in units of total amount of lead per sample
(jig), and information on all relevant process parameters from three selected laboratories.
This section describes in detail the data processing, the treatment of each of the two types
of data, and an approach to the final assessment of the two preparation methods for
composite wipes.
7.1 Data Tracking, Data Entry, and Data Processing
All MRI project records, including sample control, sample weighing data, instrument
outputs, wipe lead-loading data, instrumental parameters, and final summary data tables,
were assembled and stored in central project files at MRI. The data reports submitted from
the two commercial laboratories also were retained in the files. All laboratory results were
traceable to the unique sample identifier (bar code). Unusual experimental observations
were documented in laboratory notebooks and in the reports received from the two outside
laboratories. All project records were technically reviewed and audited as described in
Section 6. Checklists were used to assure that the analytical data packets were complete.
The analytical results for the single- and composite-wipe sample analyses (ICP and
FAA) conducted at MRI were checked, by separate laboratory personnel, using a double
key entry procedure. This provided a 100% check of the raw data entered into
spreadsheets. MRI staff also entered all analytical results reported from the other two
laboratories into spreadsheets using the double key entry procedure. Following the
technical review and the QA audit, all data results were released for statistical analysis.
The individual data spreadsheets from all the laboratories were then uploaded to an
IBM mainframe for analysis using SAS,9 a commercial statistical software package,
available to MRI at the University of Missouri-Columbia. Basic error-checking procedures
were performed to identify possible data errors. Appendix H provides a complete list of all
gravimetric and analytical data generated and statistically analyzed during this project.
7.2 Statistical Analysis of the Analytical Data
At each of three laboratories and for each of two digestion methods, lead
measurements, in total jig Pb/sample, were obtained from both the ICP and FAA analyses
of each sample as defined in Section 3. Percent recovery, calculated as the ratio of
measured over spiked amount of lead per sample, was the dependent variable in all
7-1
-------
statistical analyses and provided a measure of accuracy. The lead data obtained by FAA or
TCP were analyzed separately, using identical approaches.
Average percent recoveries were calculated for each laboratory, separately for each
composite type, SRM loading, digestion method, and analysis method. Upper and lower
95% confidence limits for individual recovery data were computed as well. These statistics
(average and confidence limits) are based on 12 measurements except for single-wipe
samples prepared by the EPA 3050A method, in which case only 6 measurements were
available. These basic statistics are shown in Figures 7-1 and 7-2 for FAA and TCP
analysis, respectively.
The aim of this experimental study was to assess the effect of the chosen factors on
method recovery. The factors investigated, either fixed or random from a statistical point
of view, were:
Fixed effect factors:
1. Number of wipes per composite (one, two, or four wipes)
2. SRM loading level (low at 0.5 g or high at 2 g per wipe)
3. Digestion method (EPA 3050A or WOHL)
Random effect factors:
4. Laboratory (Laboratory 1, 2, or 3)
5. Wipe brand (three commercially available brands, A, B, or C)
6. Wipe brand lot (two lots; does not apply to Brand C because it was purchased from a
single location)
The first three factors were selected at the onset of the study which was limited to
investigating these main factors and their levels. Statistically speaking, these factors are
called fixed factors. On the other hand, the three laboratories can be considered as having
been selected from among all laboratories performing comparable lead analyses. The
objective of the study was to assess the performance of the two composite methods across
laboratories in general, not for these three laboratories in particular. Wipe brand and lot
within brands (Brands A and B only) were two additional factors in this study. The three
wipe brands represent a sample of commercially available wipe brands adequate for dust
sampling. The purpose for including more than one brand of wipes and, to the extent
possible, wipes obtained from different manufacturing lots, was to take into account
potential variability among lots and/or brands of wipes. Again, the intended purpose of the
study was not to assess method performance for individual wipe brands. Statistically
speaking, the factors, laboratory, wipe brand, and wipe brand lot are called random factors.
7-2
-------
-------
140
120
y
* 100
HE
0. 80
o
60
40
ICP - EPA 3050A
Laboratory 1 Laboratory 2
A A A
A A
* * * A""'*"
Low SRM High SRM Low SRM
Laboratory 3
~*^ 1_^ A A A
^
« *
Hqh SRM Low SRM High SRM
1-w 2-w 4-w 1-w 2-w 4-w 1-w 2-w 4-w 1-w 2-w 4-w t-w 2-w 4-w 1-w 2-w 4-w
-B- Average Lower 95% Umit A Upper 95% Limit
ICP Recovery (%)
S S S 8 S S
ICP - WOHL
Laboratory 1 Laboratory 2
A A
" - - . ^/- :
Low SRM High SRM Low SRM
1 _w 2-w 4-w 1-w 2-w 4-w 1-w 2-w 4-w 1-w
-(- Average Lower 96% Ur
Laboratory 3
*
*
H^h SRM LowSRU High SRM ^
i i i i i i i 1 i i 1
2-w 4-w 1-w 2-w 4-w 1-w 2-w 4-w
lit A Upper 95% Limit
Figure 7-2. Basic ICP Recovery Statistics for Each Laboratory, Preparation Method,
Composite Type, and SRM Loading
-------
A model including both fixed and random effects is called a mixed model in which all
estimates of the fixed effects will be adjusted for the random effects in the model and broad
inference will be made across laboratories. Since only three laboratories were included in
the design, treating them as a random factor will result in less sensitive tests of significance
as compared to a fixed factor model. This is reflected in the relatively wide confidence
intervals computed for means, based on three laboratories only. All mixed model
ANOVAs were performed using the PROC MIXED procedure of SAS.10'11
An assessment of method accuracy and method precision for both digestion methods is
presented next.
7.2.1 Method Accuracy Analysis
A statistical model was developed to estimate the effect of the above main factors
(fixed and random) and their interactions on percent recovery (accuracy) using FAA and
TCP analyses. As stated earlier, three brands of wipes and two lots within brand (Brands A
and B only) were considered in this study in an attempt to broaden the method performance
evaluation of the two preparation methods. Various wipe brands are commercially
available for the purpose of lead dust sampling. Rather than evaluating the two digestion
methods and the two analysis methods based on a single brand of wipe, it was decided that
one or two lots of three wipe brands each would provide an appropriate basis for evaluating
method performance in the context of this study.
This approach was in no means an attempt to rank commercially available wipe brands
with respect to method recovery of lead dust. However, the potential variability between
lots within a brand and that among brands were examined to assess the need to incorporate
these random factors into subsequent statistical models. The alternative approach chosen
in the present case was not to include the two factors (brand and lot), whether random or
fixed, into the model, but to simply treat all wipes as replicate samples. This decision is
supported by the outcome of the following analyses.
Investigation of Potential Wipe Manufacturers' Lot Differences: Potential
differences between lots of Brands A and B were investigated before simply treating brands
and lots as design replicates. Since the design was balanced with respect to the fixed
factors (number of wipes per composite and SRM loading) across brands, a nested random
effects two-way analysis of variance (ANOVA) was performed, where wipe brand and lot
nested within brands were the main effects. This analysis was done separately for each
digestion method using percent recovery results from FAA and TCP. None of the factors
was found to have a statistically significant effect on percent recovery in either case
(smallest p-value = 0.38). It was concluded that the variability between lots of Brand A
and Brand B was negligible and that the manufacturers' lot information can be ignored in
subsequent analyses.
7-5
-------
Investigation of Potential Wipe Brand Differences: In light of the results from the
above analysis, the wipes of Brands A and B were each pooled across lots. Under this
configuration and using the data from all three brands A, B, and C, the design was again
balanced with respect to the fixed factors (number of wipes per composite and SRM
loading) across brands. A one-way ANOVA was then performed separately for FAA and
TCP recoveries to assess whether percent recovery varied significantly among the three
wipe brands.
The ANOVA results are shown in Table 7-1. The p-values in the fourth column
indicate the significance level of the overall wipe brand effect. The wipe brand effect is
significant in all but one caseTCP recovery when using the EPA 3050A digestion method
(p=0.2191). Average recoveries were then computed for each wipe brand, and these are
shown in columns 5 through 7 in Table 7-1. The F-test for brand effect was followed by a
multiple comparison test to identify which brands vary among themselves. In all cases,
there was no statistically significant difference between Brands A and B. However, except
for ICP and EPA 3050A, percent recovery for Brand C differs significantly from that of
either Brand A or B. The last two columns of Table 7-1 summarize the results of
comparing Brand C recovery with the average recovery of Brands A and B. Average
differences and their standard error are shown.
The outcome of this analysis would suggest that the random effect for brand should be
included in the main statistical models analyzed subsequently. However, it was decided
not to included brand as a factor in light of the original intention to treat wipes of different
brands simply as replicate samples. (It should be noted that the statistical models discussed
next were also investigated while including the random effect for brand. This approach
had no impact on the mean recoveries but resulted in slightly wider confidence intervals for
the estimated means.) In summary, neither manufacturing lot nor type of wipe brand was
considered as a design factor in this study.
Statistical Modeling Approach: FAA and ICP recovery results from a total of
396 wipe composites were available for the final evaluation. In light of the above analysis
results for wipe brands and wipe lots, the four factors remaining to be analyzed were:
Fixed effect factors:
1. Number of wipes per composite (one, two, or four wipes)
2. SRM loading level (low at 0.5 g or high at 2 g per wipe)
3. Digestion method (EPA 3050A or WOHL)
Random effect factor:
4. Laboratory (Laboratory 1, 2, or 3)
7-6
-------
Table 7-1. Average Percent Recovery by Wipe Brand
Analytical
method
FAA
ICP
Digestion
method
EPA 3 050 A
WOHL
EPA 3 050 A
WOHL
Total
number
of wipes
180
216
180
216
Overall
F-test for
brand effect:
p-value
0.0183
0.0003
0.2191
0.0039
Average recovery (%)
Brand Aa
99.33
86.06
93.94
77.23
Brand Ba
98.39
86.10
93.62
78.31
Brand Cb
101.66
91.62
96.19
83.88
Recovery difference (%)
C-(A+B)/2C
Mean
2.80
5.54
2.41d
6.11
Standard error
of mean
1.02
1.34
1.38
1.83
a Percent recoveries from Brands A and B are not statistically significant from each other.
b Percent recovery from Brand C is significantly different from that of Brand A or B, except for ICP and EPA Method 3050A.
0 Estimated percent recovery difference between Brand C and the average of Brands A and B.
d Difference is not significantly different from zero at the 5% significance level.
-------
The resulting design is a full factorial design for both FAA and TCP recovery statistics. In
a preliminary analysis, the effect on percent recovery (FAA and TCP) of all main factors
and all their interactions was investigated in a four-way mixed model ANOVA. This linear
model, in which all fixed effect estimates are adjusted for random effects, relates percent
recovery to the following factors and their interactions:
Digestion Method
SRM Level
Number of Wipes
Digestion Method x SRM Level
Digestion Method x Number of Wipes
SRM Level x Number of Wipes
Digestion Method x SRM Level x Number of Wipes
Laboratory
Laboratory x Digestion Method
Laboratory x SRM Level
Laboratory x Number of Wipes
Laboratory x Digestion Method x SRM Level
Laboratory x Digestion Method x Number of Wipes
Laboratory x SRM Level x Number of Wipes
Laboratory x Digestion Method x SRM Level x Number of Wipes
All terms involving Laboratory as a factor were treated as random in the model. All other
terms were considered fixed.
The results from this analysis, performed separately for FAA and TCP recovery, are
shown in Tables 1-1 and 1-2 in Appendix I, respectively. Of all the fixed effect terms
considered in the model, only the following three were statistically significant at the 95%
confidence level:
Digestion method for FAA recovery (p=0.0431)
Digestion method for ICP recovery (p=0.0191)
The interaction between digestion method and number of wipes per composite for ICP
recovery (p=0.0206)
These analyses clearly show the recovery differences between the two digestion
methods, with the EPA Method 3050A yielding higher recoveries, on the average, than the
-------
WOHL method. In addition, FAA yields higher recoveries on the average than ICP for
either digestion method. Average recoveries and their standard error are summarized in
Table 7-2.
Table 7-2. Overall Recovery Statistics by Analytical and
Digestion Method
Analytical
method
FAA
ICP
Digestion
method
EPA 3050A
WOHL
Difference
EPA 3050A
WOHL
Difference
Mean
recovery
(%)
99.7
87.9
11.8
94.3
79.8
14.5
Standard
error of
mean3 (%)
2.84
2.84
2.53
5.56
5.56
2.03
a Note that these standard errors, computed in a mixed
model setting, have only 2 degrees of freedom (the
number of laboratories, 3 minus 1).
At a reduced confidence level of 90%, the following additional fixed effect terms
would be statistically significant:
Number of wipes per composite for FAA recovery (p=0.086)
The interaction between digestion method and number of wipes per composite for
FAA recovery (p=0.0849)
SRM loading level for ICP recovery (p=0.069)
Following these preliminary significant analysis results, it was decided to analyze
recovery data separately for each digestion method and each analytical method. This
approach provided a means to report method performance separately for each of the
methods tested.
Investigation of Fixed Effect FactorsComposite Type and SRM Loading Level:
The effect on percent recovery (FAA and ICP) of the remaining main factors and all their
interactions was investigated separately for each digestion method in a three-way mixed
model ANOVA. This linear model, in which all fixed effect estimates are adjusted for
random effects, relates percent recovery to the following factors and their interactions:
SRM Level
Number of Wipes
SRM Level * Number of Wipes
7-9
-------
Laboratory
Laboratory x SRM Level
Laboratory x Number of Wipes
Laboratory x SRM Level x Number of Wipes
As before, all terms involving Laboratory as a factor were treated as random in the model.
All other terms were considered fixed. The model parameters were estimated for the four
combinations defined by the two digestion methods and the two analytical methods. The
results from this analysis performed for FAA and EPA 3050A and WOHL are shown in
Tables 1-3 and 1-4 in Appendix I. The TCP results from analysis using EPA 3050A and
WOHL are shown in Tables 1-5 and 1-6 in Appendix I.
The highlights in Tables 1-3 through 1-6 pertaining to the effect of the fixed factors
(SRM loading level and number of wipes per composite) and their interaction on percent
recovery can be summarized as follows:
The interaction between SRM level and type of composite is never significant
SRM loading level by itself is never a significant factor
Composite type (number of wipes per composite) is only significant for the
combination ICP and WOHL method (p=0.048). This factor is also significant at the
90% confidence level for the FAA and WOHL combination (p=0.0789). More
specifically:
The difference of 11.9% between two-wipe and four-wipe composite recoveries
(averaged across SRM levels) for FAA and WOHL is statistically significant
(p=0.0409). The difference of 10.1% between one-wipe and four-wipe composite
recoveries for FAA and WOHL is slightly less statistically significant (p=0.0661)
(see contrast estimates in Table 1-4 under heading "Estimate Statement Results").
The difference of 13.1% between one-wipe and four-wipe composite recoveries
(averaged across SRM levels) for ICP and WOHL is statistically significant
(p=0.0222). The difference of 9.9% between two-wipe and four-wipe composite
recoveries for ICP and WOHL is statistically significant (p=0.0531). (See contrast
estimates in Table 1-6 under heading "Estimate Statement Results.")
Average (model least squares means) recoveries and their standard errors are shown in
Table 7-3 for each analytical method, digestion method, SRM loading level, and composite
type. (These and other statistics can be found in Tables 1-3 through 1-6 under the heading
"Least Squares Means.") These recovery results, along with their lower and upper 95%
confidence limits, are plotted in Figure 7-3. It should be noted that the confidence
7-10
-------
Table 7-3. Recovery Statistics by Analytical Method, Digestion
Method, SRM Loading Level, and Composite Type
Analytical
method
FAA
ICP
Digestion
method
EPA 3050A
WOHL
EPA 3050A
WOHL
SRM
level
Low
High
Low
High
Low
High
Low
High
Composite
type
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
Mean
recovery
(%)
100.2
100.6
98.4
99.7
98.3
100.8
100.0
99.7
91.2
94.2
83.1
89.5
90.1
90.8
78.1
86.4
97.3
96.5
98.7
97.5
88.7
90.7
94.0
91.1
88.6
82.3
79.4
83.4
81.9
81.7
64.9
76.2
Standard
error of
mean (%)
3.40
3.31
3.31
3.24
3.40
3.31
3.31
3.24
4.19
4.19
4.19
3.24
4.19
4.19
4.19
3.24
5.18
5.14
5.14
5.08
5.19
5.14
5.14
5.08
6.81
6.81
6.81
6.27
6.81
6.81
6.81
6.27
7-11
-------
110
100
£ 90
a:
ii 70
60
50
110
100
2 90
S 70
60
50
FAA - EPA 3050A
~ ~ -EH E3
-
-
-
_
LowSRM HlghSRM
1-wipe 2-wipe 4-wipe 1-wipe 2^wipe 4-wipe
B Mean Lower 95% Limit A Upper 95% Limit
FAA - WOHL
_
A A A
" ^^ B xi
-
-
_
LowSRM HlghSRM
1-wipe 2-wipe 4-wipe 1-wipe 2^wipe 4-wipe
e Mean Lower 95% Limit A Upper 95% Limit
110
100
₯
S.
0-
Q 70
60
50
ICP - EPA 3050A
_
A
A A
~ Q Q
~ B
. »
* »
»
-
Low SRM High SRM
l^wipe 2-wipe 4^wipe 1-wipe 2-wipe 4^vipe
B Mean » Lower 95% Limit A Upper 95% Limit
110
100
g 90
»
a?
Q 70
60
50
ICP- WOHL
A
A A A
A
^"^^
\
* ^
- ^
LowSRM High SRM *
1 -wipe 2-wipe 4-wipe 1 -wipe 2-wipe 4-wipe
B Mean * Lower 95% Limit A Upper 95% Limit
Figure 7-3. Recovery Statistics by Analytical Method, Digestion Method, SRM Loading Level, and Composite Type
to
-------
intervals, calculated from the means and their standard errors, are relatively large. This is
due to the fact that these statistics are based on only 4 degrees of freedom defined by the
3-way interaction: (3-l)Laboratoiy * (2-l)SRMlevel * (3-l)Compositetype, as the model is treated as a
mixed model with the random factor, Laboratory.
Comparison of Method Recovery with Data Quality Objectives: One of the
objectives of this study was to assess whether, upon compositing and digesting two or four
wipes as single samples, acceptable recoveries could still be achieved. The data quality
objectives were to obtain recoveries between 80% and 120%. The average recoveries
shown in Table 7-3 provide a comparison for each situation tested:
Regardless of the SRM loading level and the number of wipes per composite,
acceptable recoveries were achieved for EPA Method 3050A, using either FAA or ICP
analysis. Average recoveries ranged from a low of 98.3% to a high of 100.8% for
FAA, and from a low of 88.7% to a high of 98.7% for ICP.
Acceptable recoveries were achieved for the WOFIL method for all three types of
composite at the low SRM level (0.5 g per wipe) using FAA analysis. Average
recoveries ranged from a low of 83.1% (four-wipe composite) to a high of 94.2%
(two-wipe composite).
Some difficulties were encountered with the WOFIL method when compositing four
wipes. This was the case for FAA and ICP results at the high SRM level (2 g per
wipe), and for ICP results at the low SRM level. The average recoveries in these cases
were 79.4% (ICP and low SRM level); 78.1% (FAA and high SRM level); and 64.9%
(ICP and high SRM level). Of these three results, only one is considerably outside the
data quality objectives.
7.2.2 Method Precision Analysis
Lead recovery results were obtained for various combinations of fixed factors based on
the results from three laboratories. The analysis results presented in the previous section
pertain to the accuracy of these recovery results, considering the effects of all factors. In
this section, the precision of the recovery results is evaluated, with a focus on between and
within laboratory precision components. Although some recovery results appeared to be
outliers (see Section 6.3.1) for some laboratories, all data were included in the precision
analyses as none of the data could be discarded for any plausible reasons (e.g., data
reporting errors, laboratory anomalies).
Between and Within Laboratory Precision by Individual Composite Type: To
estimate these two types of variability, the following approach was used. A two-way fixed
model ANOVA including Laboratory, SRM loading level, and the interaction of these two
factors was performed separately for each composite type and digestion method. As in
previous analyses, this was done individually for FAA and ICP recovery results. Thus a
7-13
-------
total of 12 ANOVAs were performed. This approach deviates from that previously used
for assessing method accuracy, where Laboratory was treated as a random effect. In the
present case, inference was made for the variability between and within the three
laboratories.
For each analysis, the mean square error provided an estimate of the within laboratory
variability, and the mean square associated with Laboratory provided an estimate of the
between laboratory variability. The square root of each of these quantities provided an
estimate of the corresponding standard deviation. Dividing the standard deviation by the
mean recovery in each case provided an estimate of the corresponding relative standard
deviations. Within and between precision estimates, expressed both in absolute standard
deviation and relative standard deviation, are summarized in Table 7-4. In each row, the
table shows the average recovery across all samples of a given type analyzed by all three
laboratories using the corresponding digestion and analytical methods. Between and
within laboratory precision estimated, expressed in percent relative standard deviations
(% RSD), are plotted in Figure 7-4 separately for the two analytical methods.
Within laboratory precision is good, with ICP yielding slightly better precision
estimates (3.18% to 8.12%) than FAA (4.01% to 7.93%). Except for ICP and WOHL, the
two-wipe samples provided the best precision. The four-wipe samples consistently yielded
the least precise results, although in all cases, the % RSD was less than 9%. In all cases,
within laboratory precision, based on the results from the three laboratories included in the
study, met the data quality objectives of being less than 20% RSD for both the EPA 3050A
and the WOHL methods.
Between laboratory precision is, in many cases, an order of magnitude worse than
within laboratory precision. This is especially true for some of the four-wipe composite
samples. Except for FAA and EPA Method 3050A results, between laboratory precision
considerably exceeded 20% relative standard deviation. As seen in Figure 7-4, no
particular pattern is apparent between the one-, two-, and four-wipe composites.
7.2.3 Comparison of FAA and ICP Recovery Results
Each loaded wipe composite sample was analyzed by both FAA and ICP, two
analytical methods commonly used by commercial laboratories for lead determination. In
this section, the recoveries of the two methods are compared, separately for each digestion
method, and within and between each laboratory.
In total, 396 samples were analyzed under different combinations of SRM levels,
composite types, and lead-loading levels. For each individual sample, the paired difference
between FAA and ICP recoveries was computed. The distribution of these pairwise
differences was plotted previously in Figure 6-6 in Section 6, in the form of box plots,
organized by laboratory, digestion method, and composite type. Overall, differences in
7-14
-------
Table 7-4. Within and Between Laborat
Analytical Method, and Com
Digestion
method
Composite
type
Number
of wipes
Mean
recovery
(%)
ory Precision Results by Digestion Method,
posite Type
Absolute standard
deviation (%)
Between
laboratories
Within
laboratories
Relative standard
deviation3 (%)
Between
laboratories
Within
laborator-
ies
FAA Precision results
EPA 3050A
WOHL
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
36
72
72
72
72
72
99.25
100.67
99.18
90.65
92.52
80.60
16.21
21.99
16.97
20.57
25.45
46.52
4.75
4.03
5.22
4.87
4.53
6.39
16.33
21.85
17.11
22.70
27.51
57.72
4.79
4.01
5.26
5.38
4.90
7.93
ICP Precision Results
EPA 3050A
WOHL
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
36
72
72
72
72
72
93.01
93.61
96.35
85.28
81.99
72.14
28.01
43.25
42.85
45.58
37.20
73.51
3.70
2.98
4.11
4.34
4.45
5.86
30.12
46.20
44.48
53.45
45.37
101.89
3.98
3.18
4.26
5.09
5.43
8.12
' Relative standard deviation = 100*Absolute standard deviation/mean.
-------
FAA-EPA-lw FAA-EPA-4W FAA-WOHL-1W FAA-WOHL-4W
FAA-EPA-2W FAA-WOHL-2W
B Between labs Within Labs
110
100 -
.0
;:
ICP-EPA-lw ICP-EPA-4W
ICP-EPA-2W
ICP-WOHL-lw ICP-WOHL-4W
ICP-WOHL-2W
9 Between Labs Within Labs
Figure 7-4. Between and Within Laboratory Precision by Analytical
Method, Digestion Method, and Composite Type
7-16
-------
recoveries between FAA and TCP ranged from -15.73% to 31.00%, with an average of
6.80% and a standard deviation of 9.36%.
The differences between FAA and TCP recovery results were analyzed in a way similar
to that used in previous ANOVAs (see section 7.2.1). A mixed ANOVA model was used
to investigate the effect of the fixed factors (composite type and SRM loading level) and
the random factor (laboratory) on paired differences in analytical recovery results, sepa-
rately for each digestion method. On the average, higher recoveries were obtained with
FAA than with ICP, regardless of the digestion method used. However, in most cases,
differences in recoveries were larger when using the WOFIL method, all other conditions
held constant. Of the two fixed effects and their interaction, only composite type signifi-
cantly (p=0.027) affected the difference in recovery between FAA and ICP when using the
EPA 3050A digestion method. For EPA Method 3050A, average (model least square
means) differences between FAA and ICP recoveries, across SRM levels, were as follows:
Single-wipes: 6.24%
Two-wipe composites: 7.06%
Four-wipe composites: 2.84%
where the difference for the four-wipe composites is significantly lower (at the 5% level)
than either that for the single wipes or the two-wipe composites.
Average differences and their lower and upper 95% confidence limits obtained from
the mixed model ANOVAs are shown in Table 7-5, for each level of the fixed factors and
their interaction. As shown in the table, none of the average differences is significantly
different from zero as none of the relatively wide confidence intervals includes zero. This
is in part due to the choice of the mixed model as an analysis approach and the fact that
laboratories are the single largest contributor to the variability in these results.
Figure 6-6 clearly shows the variability among laboratories. Although the differences
between the recoveries of the two analytical methods were consistently high at Laboratory
3 using either digestion method, these differences exhibit a relatively small variation. In
comparison, Laboratory 1 showed consistently higher FAA results with very little
variability for the WOHL method. On the other hand, when using the EPA
Method 3050A, Laboratory 1 showed no consistent differences between the two analytical
methods. This pattern is practically the opposite of that shown by the Laboratory 2 results.
7-17
-------
Table 7-5. Comparison of FAA and ICP Recoveries by Digestion
Method, SRM Level, and Composite Type
Digestion
method
EPA
3050A
WOHL
SRM
level
Low
High
All
Low
High
All
Composite
type
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
All
1-wipe
2-wipe
4-wipe
FAA-ICP recovery difference (%)
Mean
2.92
4.12
-0.37
2.22
9.56
10.01
6.05
8.54
6.24
7.06
2.84
2.61
11.89
3.70
6.07
8.13
9.16
13.22
10.17
5.37
10.53
8.46
95% lower
limit
-13.50
-12.22
-16.71
-22.87
-6.86
-6.33
-10.29
-16.55
-7.86
-6.99
-11.22
-12.09
-2.81
-11.01
-11.40
-6.58
-5.54
-1.48
-7.30
-6.78
-1.63
-3.69
95% upper
limit
19.34
20.45
15.96
27.31
25.98
26.34
22.38
33.63
20.34
21.12
16.89
17.32
26.60
18.40
23.54
22.84
23.87
27.93
27.65
17.52
22.68
20.61
7.3 Evaluation of Process Parameter Data
The second objective of this study was to determine whether compositing of wipes, if
acceptable on a technical or performance basis, could reduce the cost of sample preparation
and analysis relative to single-wipe methods. To help assess the relative costs of the
methods, each laboratory was asked to record the time, materials, and equipment
requirements for each sample digestion and analytical procedure included in the study. It
should be noted that the time required to conduct analytical methods is not a parameter
typically reported by commercial laboratories.
7-18
-------
7.3.1 Initial Process Parameter Data
The cost information initially provided by the laboratories was fairly limited since the
digestion times reported by each laboratory were based only on one or two sample batches
for each digestion method and composite type. Only one batch of single wipes and one
batch of two-wipe composites were prepared using the EPA Method 3 050 A and the
WOHL method. Only one batch of four-wipe composites was prepared using the WOHL
method, while two batches of four-wipe composites were prepared using the EPA method.
Until new digestion methods become routine, the actual time required to use the methods
cannot be defined with certainty.
The approximate time required to perform each digestion method and composite type
combination was recorded by each laboratory participating in the investigation. A relative
sample preparation and analysis cost comparison was then made. A general evaluation of
the results indicated that, within a given laboratory, the time requirements for a given
method were fairly consistent across composite types. The multiwipe composites required
slightly more time to digest relative to the single-wipe samples; however, that increase was
not proportional (i.e., the two- and four-wipe composites did not take two and four times as
long to digest as the single-wipe samples, respectively). The single-wipe samples digested
using the WOHL method at Laboratory 1 actually required more time to digest than the
multiwipe composites, but this is most likely attributable to the lack of familiarity with that
method.
An interlaboratory comparison was made for the digestion times reported by each
laboratory; it was found that the data were not very comparable. Although each laboratory
reported the total hours required for each digestion method and composite type
combination, differences in other time parameters were reported. For example,
Laboratory 1 reported the total hours to complete the digestions, and reported also the
number of labor hours required (e.g., the hours when staff were actively working on the
digestion procedures, including glassware cleaning, filtering samples, centrifuging samples,
etc.). Laboratory 2 reported the total hours and the digestion times required for each
method. Laboratory 3 simply reported the total hours and the number of hours that the
samples were on the hot plate for each digestion procedure.
The total hours reported by Laboratory 1 and Laboratory 3 were somewhat
comparable; however, the total hours reported by Laboratory 2 were an order of magnitude
lower than the digestion times reported by Laboratories 1 and 3. Additional information
was requested from Laboratory 2, but the revised data were similar to those originally
reported. It was noted that the digestion times reported by Laboratory 2 for the WOHL
method corresponded to those specified in the procedures for the digestion. The total hours
reported by Laboratory 2 for the EPA Method 3050A were considered to be unrealistic, in
comparison to the results from Laboratories 1 and 3. Based on these observations, it was
concluded that Laboratory 2 did not accurately record the required digestion times for the
study. Thus, the times reported by that laboratory were not included in the evaluation.
Because Laboratory 1 recorded all work steps in the necessary detail and since all
7-19
-------
laboratories followed the same steps in conducting the digestion procedures, only the total
hour and labor hour data provided by Laboratory 1 were evaluated. The data reported by
Laboratory 1 for each digestion method and composite type combination were considered
representative of all three laboratories and are presented in Table 7-6.
The percent relative effort and estimated percent reduction in effort per wipe were
calculated for each digestion method (EPA 3050A and WOHL). These results are shown
also in Table 7-6 and provide a means for comparing the digestion times to evaluate the
cost efficiency of the composite wipe methods. Using the single-wipe batch type as the
basis for cost comparisons, a reduction of 70% was estimated for the four-wipe composites
using the EPA Method 3050A. A reduction of 77% was estimated for the four-wipe
composites using the WOHL method. That is, laboratory analysis of a four-wipe
composite was estimated to be 70% and 77% less time consuming than four single-wipe
analyses for the EPA Method 3050A and the WOHL method, respectively. Similarly for
the two-wipe composites, time reductions of 39% and 53% were estimated for the EPA
3050A and WOHL methods, respectively.
7.3.2 Supplementary Process Parameter Data
The three laboratories also evaluated the cost aspects of compositing from a
retrospective viewpoint, based on the experience gained from the study. Each laboratory
was asked to estimate the relative cost factors that could be applied to the digestion
methods, assuming someone approached the laboratory with a need to analyze a sizable
number of composite samples for lead. A limited sample preparation and analysis cost
comparison was performed based on the information provided by the three laboratories.
Comparison across composite types, separately for each digestion method: The
relative cost factors that could be applied to two- and four-wipe composites for a given
digestion method were estimated, assuming the single-wipe methods represented 100%.
For example, if two-wipe composites digested with the EPA method required 1.5 times the
effort needed to digest single-wipe samples using the same method, then the relative cost
factor would be 150% for two-wipe composites. The data reported by each laboratory from
this comparison are presented in Table 7-7.
Again, using the single-wipe batch type as the basis for cost comparisons, an average
per-wipe cost reduction of 64% was estimated for the four-wipe composites using EPA
Method 3050A. For example, suppose that a single wipe costs $20 to analyze. Based on
the reduction of effort obtained from this study, a four-wipe composite would cost $28.50,
a 64% cost reduction over $80, the cost of four single wipes. A reduction of 67% was
estimated for the four-wipe composites using the WOHL method. Similarly for the two-
wipe composites, per-wipe cost reductions of 42% and 38% were estimated for the EPA
3050A and WOHL methods, respectively. The estimated cost reductions from the
retrospective data are therefore in line with the cost reduction data obtained from
Laboratory 1 (Table 7-6).
7-20
-------
Table 7-6. Relative Effort Comparison Among Types of Composite for Laboratory 1
Digestion
method
EPA
3050A
WOHL
Composite
type
1-wipe
2-wipe
4-wipe
1-wipe
2-wipe
4-wipe
Number of
analysis
batches
1
1
2
1
1
1
Total hours
18
17
40
21
19
19
Labor hours
14
17
34
16
15
15
Labor hours
per batch
14
17
17
16
15
15
Relative effort (%)
Per batch3
100
121
121
100
94
94
Per wipe"
100
61
30
100
47
23
%
Reduction
in effort per
wipe0
NA
39
70
NA
53
77
3 Within each cell, % ratio of labor hours per batch (2- or 4-wipe composite) to labor hours per batch (1 wipe).
b Within each cell, % relative effort per batch divided by the number of wipes in the composite.
c The % reduction in effort per wipe is calculated as 100% minus the % relative effort per wipe.
Table 7-7. Relative Cost Factor Comparison Across Composite Types, Separately for Each Digestion Method
Method
EPA
3050A
WOHL
Laboratory
1
2
3
1
2
3
Single-wipe
weighting (%)
100
100
100
100
100
100
2-wipe composite
Relative cost
factor (%)
120
130
100
110
110
150
% Relative
effort per
wipe3
60
65
50
55
55
75
% Reduction
in effort per
wipe"
40
35
50
45
45
25
4-wipe composite
Relative cost
factor (%)
150
150
125
125
120
150
% Relative
effort per
wipe3
38
38
31
31
30
38
%
Reduction
in effort per
wipe"
62
62
69
69
70
63
i 3 The % relative effort per wipe is calculated by dividing the relative cost factor by the number of wipes in the composite.
^ b The % reduction in effort per wipe is calculated as 100% minus the % relative effort per wipe.
-------
Based on this limited cost comparison, an approximate cost benefit can be derived from
compositing.
Comparison between digestion methods, separately for each composite type: The
cost-effectiveness of the EPA method versus the WOHL method was also of interest. The
relative cost factors that could be applied to the WOHL digestion method, for a given
composite type, were estimated assuming the EPA method represented 100% for each
composite type. For example, if the WOHL Method for two-wipe composites could be
performed in half the time required by the EPA Method for two-wipe composites, then the
relative cost factor would be 50% for the WOHL method. The data from this comparison
are presented in Table 7-8.
Table 7-8. Relative Cost Factor Comparison Between Digestion
Methods, Separately for Each Composite Type
Composite
type
1-wipe
2-wipe
4-wipe
Laboratory
1
2
3
1
2
3
1
2
3
EPA method
weighting (%)
100
100
100
100
100
100
100
100
100
WOHL method
weighting (%)
60
30
50
60
40
75
50
40
50
% Reduction
in effort
WOHL vs.
EPA
40
70
50
40
60
25
50
60
50
Across all laboratories and composite types, cost reductions of 25% to 70% were
estimated for the WOHL method as compared to the EPA Method 3050A. Thus, based on
these performance data alone, cost saving can be derived from using the WOHL method
instead of EPA Method 3050A. It should be noted that the process parameter data do not
take into consideration the analytical performance of the methods.
7.3.3 Other Cost Data
The costs of reagents, waste disposal, and associated increased safety were higher for
the multiwipe composite methods relative to the single-wipe methods since larger
quantities of acids are needed to perform composite sample digestion. However, these
costs also were compared to the total cost per sample to determine whether that increase
was significant. The reagent and waste disposal costs for the EPA Method 3050A were
7-22
-------
estimated at $1, $2, and $4 per sample for the single-, two-, and four-wipe composite types,
respectively. If the wipe samples are composited, then the reagent and waste disposal costs
can be normalized to $ I/wipe. The reagent and waste disposal costs were a negligible part
(less than $0.50 per sample) for the WOHL single- and composite-wipe methods. Thus,
the WOHL method is less expensive than the EPA Method 3050A in all cases.
The times required for the TCP and FAA analyses were reported also by each
laboratory. As expected, the analysis times did not vary by composite type for a given
analytical technique. All three laboratories indicated that an FAA analysis required less
time than an ICP analysis, with an estimated 2 min per sample for FAA versus 3 to 5 min
per sample for ICP. However, since many laboratories use autosamplers, the importance of
instrument analysis time is not very important in the context of cost savings. In addition,
since FAA instruments are also less expensive to purchase than ICP instruments, a cost
advantage can be achieved from using FAA rather than ICP.
7-23
-------
Section 8
References and Bibliography
1. Title X, Residential Lead-Based Paint Hazard Reduction Act.
2. Guidelines for the Evaluation and Control of Lead-based Paint Hazards in Housing,
June 1995. The National Center for Lead-Safe Housing, U.S. Department of Housing
and Urban Development, Washington, DC.
3. "Test Methods for Evaluating Solid Waste." Volume 1 A: Laboratory Manual
Physical/Chemical Methods, SW-846 Methods 3050A, 6010A, 7000A, and 7420,
November 1986, Third Edition and updates, U.S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response, Washington, D.C.
4. Draft Guidelines for the Evaluation and Control of Lead-based Paint Hazards in
Housing, May 1994 and February 1995. The National Center for Lead-Safe Housing,
U.S. Department of Housing and Urban Development, Washington, D.C.
5. National Lead Laboratory Accreditation Program Presentation, final draft report,
August 18, 1995. EPA Contract 68-DO-0137, MRI Project No. 9804-10-01.
6. "Protocols for the Collection and Analysis of Composite Wipe Samples," report,
September 30, 1994. EPA Contract 68-DO-0137, MRI Project No. 9803-01-10.
7. "Analysis of Composite Wipe Samples for Lead Content: Quality Assurance Project
Plan," May 18, 1995. EPA Contract No. 68-DO-0137, MRI Project No. 9804-1, WA
5-07.
8. SYSTATfor Windows: Graphics, Version 5. SYSTAT, Inc., Evanston, Illinois, 1992.
9. SAS/STAT User's Guide, Volume 2, Version 6, Fourth Edition. SAS Institute, Inc.,
Cary, North Carolina, 1989.
10. SAS/STATSoftware: Changes and Enhancements, Release 6.07, SAS Technical
Report P-229. SAS Institute, Inc., Cary, North Carolina, 1992.
11. Latour, D., Latour, L., and Wolfmger, R.D., "Getting Started with PROC MIXED."
Software Sales and Marketing Department, SAS Institute, Inc., Cary, North Carolina,
1994.
3-1
-------
Appendix A
Method Review SummaryProtocols for
Sample Preparation and Analysis of Lead in
Single-Surface and Composite Wipe Samples
-------
Method Review SummaryProtocols for
Sample Preparation and Analysis of Lead in
Single-Surface and Composite Wipe Samples
1 Introduction
The contractor completed a review of available protocols for the preparation and
analysis of lead (Pb) in single-surface and composite wipe samples, and found that few
protocols were available for composite wipes. Sample preparation techniques typically use
acid(s) and water to solubilize (digest) the Pb contained in the samples, prior to analysis.
Sample preparation procedures written specifically for single-surface samples may require
modifications for multiwipe composites, because of the increased mass of wipe and dust
material contained in the samples. As the total mass of a sample increases (composite
samples), the total amount of solution needed to solubilize the sample increases. Because
sample preparation protocols for lead are designed and validated with specific upper limits
for sample mass, most of the currently available preparation protocols are not intended for
composite wipes.
The analysis of composite samples is dependent upon the preparation of sample
digestate, which is likely the most important step in the analysis process. The effectiveness
of a given method is measured by a number of analytical parameters including accuracy
and precision. Of course, each laboratory would need to validate its performance with any
method for the analysis of composite wipes as well as single-wipe samples.
The mass of the dust collected on a single-wipe sample is generally unknown and
cannot be determined in the laboratory, because the tare weight of the wipe prior to sample
collection cannot be measured. Using a blank wipe to obtain a tare weight is not effective
since the mass between wipes is variable as a result of inconsistencies in manufacturing
and varying degrees of wetness among individual wipes. Therefore, research to develop or
validate a laboratory method for preparing composite wipe samples should include an
assessment of the maximum level of dust loading that can be tolerated by the method, to
ensure that the method is suitable for the intended sampling application (e.g., risk
assessments, lead hazard screens, or clearance testing).
This appendix is divided into four sections. Section 1 provides an introduction, and
Section 2 provides sample preparation and analysis protocols, including descriptions of
sample preparation and analytical techniques currently available in the literature for single-
surface and composite wipe samples. Section 3 discusses the advantages and
disadvantages of the methods, and Section 4 provides a list of references reviewed in
preparing this appendix.
A-l
-------
2 Protocols
The sample preparation and analysis protocols found in the literature search are
summarized in Sections 2.1 and 2.2, respectively. Most of the methods cited apply to
single-surface samples, but may be adaptable to composite wipes.
2.1 Sample Preparation Techniques
Several preparation techniques were reviewed; however, few were designed for
composite samples. A summary of the available information on sample preparation
techniques for the determination of lead in wipe samples is given in Table A-l. The
superscript number next to the method type identifies the corresponding document
referenced in Section 4.
2.2 Analytical Techniques
The most commonly used techniques to analyze lead in the laboratory are flame and
graphite furnace atomic absorption spectrometry (AAS) and inductively coupled plasma
atomic emission spectrometry (ICP-AES). Other techniques include the polarographic,
colorimetric, and anodic stripping voltammetry methods. The analysis protocols reviewed
are discussed below and their advantages and disadvantages are summarized in Table A-2.
The superscript refers to the Section 4 references.
Flame Atomic Absorption Spectrometry (FAA).5'6'11 An oxidizing air-acetylene
flame-type is recommended. The working range is dependent on the analytical line
selected for analysis. The 217.0-nm line with a slit width of 1.0 nm has an optimum
working range of 2.5 to 20 |ig/mL with a sensitivity of 0.06 |ig/mL. The 283.3-nm line is
often preferred for routine analysis using a slit width of 0.5 nm and a working range of 7.0
to 50 |ig/mL with a sensitivity of 0.16 |ig/mL. Few interferences are reported. The use of
deuterium background correction is recommended at the 217.0-nm wavelength to correct
for non-atomic absorbances. For method of quantitation, standards, blanks, and samples
are aspirated and absorbances are recorded. Standards of known concentration are plotted
versus the obtained absorbance. Using the measured sample absorbances, the
corresponding sample concentrations (|ig/mL) are calculated from the standard curve
plotted. The lead concentration (|ig/ft2) is calculated by multiplying the sample
concentration by the sample dilution volume (mL) and then dividing the product by the
area sampled in square feet. FAA is a readily available laboratory method that is easy to
operate. The method is capable of quantifying lead from 0.1 |ig/mL and recoveries (when
digested with HNO3:H2O2) can be expected to be from 95% to 105%.
A-2
-------
Table A-l. Summary of Sample Preparation Methodologies for Wipe Samples
Method type
Summary description
Hot Plate2'6
(HNO3:H2O2)
Hot Plate7
(HNO3:H2O2)
Hot Plate and
Microwave3
(HCI:HNO3:HCIO4)
Advantages: Commonly used laboratory method. Possibly adaptable for composite
wipes. Lead recovery is expected to be adequate.
Disadvantages: Limited to 1 wipe. Method not easily adaptable to field use since it
requires a laboratory fume hood.
Typical Digestion Vessel Size: 150-mL or 250-mL beaker
Primary Chemicals Used: HNO3, H2O2
Final Dilution Volume: 100 ml
Procedure: Transfer the single-wipe sample to a clean empty beaker. Add 25 mL
1:1 HNO3, gently swirl to mix, and cover with a watch glass. Gently heat the sample
at 85° to 100°C and reflux for 10 to 15 min without boiling. Remove from hot plate
and allow the sample to cool. Add 10 mL concentrated HNO3, replace the watch
glass and reflux for 30 min. Repeat this last step to ensure complete oxidation.
Allow the solution to evaporate to approximately 10 mL without boiling, while
maintaining a covering of solution over the bottom of the beaker. Cool the sample,
add 5 mL of Type I water and 5 mL of 30% H2O2. Cover the beaker with a watch
glass and return the covered beaker to the hot plate for warming. Heat until the
effervescence subsides and cool the beaker. Continue to add 30% H2O2 in 1-mL
aliquots with warming until the effervescence is minimal or until the general sample
appearance is unchanged. Do not add more than a total of 10 mL of 30% H2O2,
even if the effervescence has not been reduced to a minimal level. Heat the acid-
peroxide digestate carefully until the volume has been reduced to approximately
10 mL. Allow the digestate to cool, rinse the beaker walls and bottom of the watch
glass with Type I water, and quantitatively transfer it to a 100-mL volumetric flask.
Dilute to volume with Type I water.
Advantages: Relatively simple, commonly used laboratory method. Possibly
adaptable for composite wipe samples. Lead recovery is expected to be adequate.
Disadvantages: Currently written for the preparation of air particulate on cassette
filters. Modifications must be made to use for wipe samples. Modifications would
need to be substantial to use on composite wipe samples. Possibly adaptable for
composite wipes. This method is not easily adaptable to field use since it requires a
laboratory fume hood.
Typical Digestion Vessel Size: 50-mL beaker
Primary Chemicals Used: HNO3, H2O2
Final Dilution Volume: 10 mL
Procedure: Transfer one filter sample to a clean, empty beaker. AddSmL
concentrated HNO3 acid and 1 mL 30% H2O2 and cover with a watch glass. Gently
heat the sample at 140°C until the volume is reduced to about 0.5 mL. Repeat two
more times using 2 mL HNO3 and 1 mL 30% H2O2 each time. Heat on the hot plate
(140°C) until about 0.5 mL remains. When the sample is dry, rinse the beaker walls
and bottom of the watch glass with 3 to 5 mL 10% HNO3. Allow the solution to
evaporate to dryness. Cool each beaker and dissolve the residue in 1 mL HNO3.
Transfer the solution to a 10-mL volumetric flask and dilute to volume with distilled
water.
Advantages: Relatively simple laboratory method. Possibly adaptable for composite
wipes. Lead recovery is expected to be adequate.
Disadvantages: Limited to 1 wipe. This method not easily adaptable to field use
since it requires a laboratory fume hood. The use of HCIO4 is potentially explosive.
Typical Digestion Vessel Size: 50-mL beaker
Primary Chemicals Used: HNO3, H2O2, HCIO4
Final Dilution Volume: 50 mL
Procedure: Place each wipe in a 50-mL beaker, add the optimum mixture (15 mL)
6:2:5 HCI:HNO3:HCIO4, and carry out digestion at 100°C until no brown gas evolves.
Decant the acid into a beaker after centrifugation at 700 rpm for 10 min, evaporate it
just until it is dry, and solubilize the residue in 10 mL 10% HNO3.
A-3
-------
Table A-l (Continued)
Method type
Summary description
Hot Plate and
Microwave Preparation3
Room Temp Elution4
(HCI)
Ultrasonifi cation8
(HNO3) Preparation
Dry-Ashing Procedures9
Advantages: Written for multiple wipes.
Disadvantages: Lead recovery information is not known. This method not easily
adaptable to field use since it requires a laboratory fume hood. Evaporation step is
expected to be extremely time consuming and is prone to losses caused by
spattering of the sample at the indicated temperature. The use of HCIO4 is
potentially explosive.
Typical Digestion Vessel Size: 800-mL beaker
Primary Chemicals Used: HNO3, HCIO4
Final Dilution Volume: 10 ml
Procedure: Up to six wipes are placed in a 800-mL beaker to which 5:4
HNO3:HCIO4 acid (100 ml) was added. The microwave is operated at 10% power
for 15 min and at 50% power for an additional 15 min, after which the contents are
swirled manually and then processed again for 15 min at 50% power. The beaker is
transferred to a hot plate, and the acid evaporated to dryness at 250°C. After
cooling, the residue is redissolved in 10 ml 10% HNO3 for lead analysis.
Advantages: Simple procedure. Possibly adaptable to field use. Possibly
adaptable for composite wipe samples.
Disadvantages: Limited to 1 wipe sample. Lead recovery information is not known.
Extraction using 0.1 N HCI at room temperature is not expected to achieve full lead
recoveries from all samples. Contamination potential exists from drying step.
Typical Digestion Vessel Size: 30-mL test tube
Primary Chemicals Used: HCI
Final Dilution Volume: 20 ml
Procedure: Flatten the sample on a clear surface, allow it to dry, roll it up and insert
into a test tube. Add 20 mL of 0.1 N HCI and allow the sample to stand at room
temperature for 10 to 15 hours. Decant the acid solution (leachate) and take 15-mL
aliquots for analysis.
Advantages: Simple procedure that is readily usable in field. Possibly adaptable for
composite wipe samples.
Disadvantages: Limited to 1 wipe. Lead recovery information is not known.
Typical Digestion Vessel Size: 50-mL beaker
Primary Chemicals Used: HNO3
Final Dilution Volume: 30 mL
Procedure: Transfer wipe samples to a small glass beaker, treat with 30 mLof 10%
HNO3 and then sonicate in a commercial ultrasonic cleaner for 60 minutes.
Advantages: Relatively simple procedure. Not adaptable to field use. Can be used
for composite wipe samples.
Disadvantages: Dry-ashing procedures have potential problems with convection
losses, volatility loss of lead, contamination problems, and limited batch sizes.
Typical Digestion Vessel Size: 100 mL porcelain dish
Primary Chemicals Used: HNO3
Final Dilution Volume: 100 mL
Procedure: Dry-ashing procedures employ furnace heating to oxidize a sample and
convert sample components to mostly oxides. The resulting ash is solubilized in
HNO3 or HCI and diluted to volume with water. An ashing agent is commonly used
to aid in retaining the elements of interest during the heating process.
A-4
-------
Table A-l (Continued)
Method type
Summary description
Room Temperature and
Hot Plate13 (1 M HNO3)
Advantages: Relatively simple laboratory method. Applicable to six-wipe
composites from children's hands.
Disadvantages: This method may not accommodate the increased sample mass
and particle sizes expected from the collection of settled dust.
Typical Digestion Vessel Size: 800-, 250-, and 100-mL beakers
Primary Chemicals Used: 1 M HNO3
Final Dilution Volume: 10 ml
Procedure: Transfer each sample to a labeled 800-mL beaker. Add 100 ml of 1 M
HNO3 to each sample, and swirl each sample for 10 sec with a glass stirring rod to
thoroughly wet the wipes. Cover the samples with watch glasses and allow the
samples to extract at room temperature for 2 hr. Decant the acid solution from the
800-mL beaker into a clean, labeled 250-mL beaker. Add 50 ml of 1 M HNO3 to the
hand wipes in the 800-mL beaker, and swirl the wipes with a glass rod for 10 sec.
Decant the acid solution into the same 250-mL beaker to composite the acid rinse.
Again, add 50 mL of 1 M HNO3 to the hand wipes in the 800-mL beaker and swirl the
wipes with a glass rod for 10 sec. Decant the acid rinse solution into the same 250-
mL beaker to composite the acid rinse for a total volume of about 200 mL. Place
the samples with watch glass covers on a hot plate at a setting that does not exceed
100°C (if lead is the only analyte the temperature can be 140°C). Rinse the stirring
rods into the beakers and discard them. Evaporate the samples until the sample
volume is approximately 25 mL. Remove the samples from the hot plate, cool, and
filter the samples collecting the filtrate in labeled 100-mL beakers. Place the 100-
mL beakers with watch glass covers on a hot plate and reduce the volume to
approximately 5 mL. Remove the samples from the hot plate and cool. Transfer
each sample to a graduated test tube with stopper and dilute to 10 mL with distilled
deionized water.
A-5
-------
Table A-2. Laboratory Analytical Protocols
Protocol
Flame Atomic
Absorption
Spectrometry5'6'11
Inductively
Coupled Argon
Plasma Atomic
Emission
Spectrometry69
Graphite
Furnace1'6
Polarographic4
Colorimetric8
Anodic Stripping
Voltammetry10
Functio
n
Analysis
Analysis
Analysis
Analysis
Analysis
Analysis
Field
or
lab use
Lab
Lab
Lab
Lab
Field/
Lab
Field
Advantages and disadvantages
A (Advantage): Readily available and capable of quantifying lead from 0.1 ug/mL;
recoveries can be expected to be from 95% to 105%.
D (Disadvantage): Interferences: high concentrations of calcium, sulfate,
carbonate, phosphate, iodine, fluoride, or acetate.
A: Readily available and capable of quantifying lead from 0.03 ug/mL; recoveries
can be expected to be from 95% to 105%.
D: Spectral interferences.
A: Readily available and capable of quantifying lead from 0.002 ug/mL; recoveries
can be expected to be from 95% to 105%.
D: High concentrations of calcium sulfate, phosphate, iodide, fluoride, or acetate
can interfere with lead determination.
A: Recoveries of soluble lead standard from towels using the HCI room
temperature elution method gave recoveries of approximately 90%.
D: Selective and sensitive method. Not widely used.
A: Method can be used in the field. Lead can be extracted from the samples
using the ultrasonic technique, then measured with the Hach method.
D: Limited sample size.
A: Method can be used in the field.
D: Not widely available.
>
-------
Inductively Coupled Argon Plasma Atomic Emission Spectrometry (ICP).6 9 The
working range (at the lead wavelength of 220.3 nm) for this method is 0.03 to 20.0 |ig/mL.
Spectral interferences are the primary interferences encountered in ICP analysis. These are
minimized by interelement correction factors and background correction. For method of
quantitation, standards, blanks, and samples are aspirated and emissions are recorded.
Standards of known concentration are plotted versus the obtained emission. Using the
measured sample emissions, the corresponding sample concentrations (|ig/mL) are
calculated from the standard curve plotted. The lead concentration (jig/ft2) is calculated by
multiplying the sample concentration by the sample dilution volume (mL), then dividing
the product by the area sampled in square feet. The ICP method is a readily available
laboratory application that is easy to operate. The method is capable of quantifying lead
from 0.03 |ig/mL and recoveries (when samples are digested with HNO3:H2O2) can be
expected to be from 95% to 105%.
Graphite Furnace Atomic Absorption Spectrometry (GFAAS).1'6 The instrument
is equipped with background correction and a lead hollow cathode lamp or equivalent and
is capable of making lead absorption measurements at the 283.3-nm absorption line. For
instrument setup, the GFAAS spectrometer should be set for the analysis of lead at
283.3 nm according to the instructions given by the manufacturer. The hollow cathode
lamp should have at least a 30-minute warmup prior to starting calibration and analysis.
The calibration curve for GFAAS should consist of a minimum of three calibration
standards and a blank. Calibration and quantitative lead measurement of sample digestates
and instrumental quality control are performed in sequential order. A calibration curve is
prepared to convert instrument response to concentration using a linear regression fit. All
instrumental measurements are converted on instrumental quality control standards and
sample digests to lead concentration using the calibration.
Polarographic.4 Polarographic waves are recorded from -0.24 to -0.69 volts against
a saturated calomel electrode (SCE). A 0.4-mL portion of 0.2% gelatin solution is added
for suppression of the polarographic maxima. For method quantitation, standards of
known concentration made from stock solutions are analyzed, and a calibration curve is
drawn relating the lead concentration (|ig/mL) and the limiting current (|iA) of
polarographic wave. Sample concentrations are calculated from the current (jiA) of
polarographic wave measured versus the calibration curve (|ig/mL). The lead
concentration (|ig/ft2) is calculated by multiplying the sample concentration by the sample
dilution volume (mL) and then dividing the product by the area sampled in square feet.
Polarography is a selective and sensitive method for the determination of lead. It is not
widely used.
Colorimetric8 (Hach Method). An aliquot of the extract is added to 100 mL
deionized water. The solution is passed through an extraction column and lead is removed
from the solution. Lead is then eluted from the column, and a color-forming reagent is
added. The resultant solution is divided in two; in one of the halves, the colored product of
the lead and color-forming reagent are dispersed by the addition of a decolorizing reagent.
The colorimeter is blanked with the decolorized sample, and the other half is placed in the
A-7
-------
colorimeter to read concentration. The linear range of the method is approximately 1 to
150|igPb/L.
Metals that have been reported to interfere with lead analysis by this method are:
aluminum, barium, calcium, copper, iron, magnesium, manganese, and zinc. Several other
inorganic ions are also reported to interfere with lead analysis including: chloride, fluoride,
ammonium, nitrate, and sulfate. For method of quantitation, standards, blanks, and
samples are treated with the color-forming reagent, and the percent transmittance of the
samples is recorded. Standards of known concentration are plotted versus the obtained
percent transmittance. Using the measured sample percent transmittances, the
corresponding sample concentrations (|ig/mL) are calculated from the standard curve
plotted. The lead concentration (|ig/ft2) is calculated by multiplying the sample
concentration by the sample dilution volume and then dividing the product by the area in
square feet. This method can be taken into the field for measurement of lead in samples
collected. Lead can be extracted from the samples using the microwave technique and then
measured with the Hach Method. Recoveries of 90% to 110% can be expected for samples
using this method.
Anodic Stripping Voltammetry (ASV).10 The ASV technique employs disposable
electrodes to measure lead in solutions. Voltametric methods are based on the current-
voltage curves arising from a microelectrode when diffusion is the rate-determining step in
the electrochemical reaction. The ASV device is compact and portable, making it ideal for
field analysis.
3 Summary of Advantages and Disadvantages of Methods
There are advantages and disadvantages to the sample preparation and analytical
methods discussed in Section 2 (see Tables A-l and A-2) for use with composite wipe
samples. Note that most of these methods and protocols were originally designed for the
preparation and analysis of single-wipe samples.
3.1 Laboratory Preparation Methods
Hot plate digestion methods are widely used procedures in which lead is solubilized
with liquid oxidizing reagents facilitated by heat using a hot plate. Hot plate procedures,
which are generally not subject to the loss and contamination problems common to other
preparation procedures, can only be performed in a laboratory. Although the room
temperature elution and ultrasonification preparation methods can be performed in the
field, the sample size is limited.
A-8
-------
3.2 Laboratory Analytical Methods or Protocols
Flame atomic absorption spectrometry is readily available and capable of quantifying
lead from 0.1 |ig/mL and recoveries can be expected to be from 95% to 105%. However,
interferences that can affect results include high concentrations of calcium, sulfate,
carbonate, phosphate, iodine, fluoride, or acetate. Inductively coupled argon plasma
atomic emission spectrometry is readily available and capable of quantifying lead from
0.03 |ig/mL. Recoveries can be expected to be from 95% to 105%. There are, however,
spectral interferences. Polarographic analysis yielded recoveries for towels spiked with
soluble lead standard of approximately 90% when using the HC1 room temperature elution
method. This analytical method, which is selective and sensitive, is not in wide use.
Although colorimetric analysis can be used in the field, the method was originally designed
for water testing.
4 References
MRI reviewed the documents listed below. The materials listed contained valuable
and useful information.
1. Midwest Research Institute, "Quality Assurance Project Plan for Comparative Field
Study of Methodologies Used to Detect Lead in Paint," for the Renovation and
Remodeling Exposure Study, EPA Contract 68-DO-0137, WA 57 (July 1993).
2. Battelle and Midwest Research Institute, "Quality Assurance Project Plan for
the Comprehensive Abatement Performance Study, Appendix B," EPA Contract
Nos. 68-DO-0126 and 68-DO-0137, WAs 2-3 and WA 19 (February 1992).
3. Que Hee, S., et al., "Evolution of Efficient Methods to Sample Lead Sources, Such as
House Dust and Hand Dust, in the Homes of Children," Envir. Res., 38:77-95 (1985).
4. Vostal, J., et al., "Lead Analysis of House Dust: A Method for the Detection of
Another Source of Lead Exposure in Inner City Children," Envir. Health Per sped.,
May 1974, 91-97.
5. OSHA, Method for the Analysis of Metal and Metalloid Particulates in Workplace
Atmospheres by Atomic Absorption, ID-121.
6. ASTM, "Emergency Standard Practice for Hot Plate Digestion of Dust Wipe Samples
for Determination of Lead by Atomic Spectrometry," Designation: ES 36-94.
7. NIOSH, Method 7082, Issue 2, Lead by Flame AAS, August 1994.
A-9
-------
8. Luk, K. K., et al., "Standard Operating Procedure for Field Analysis of Lead in Paint,
Bulk Dust, and Soil by Ultrasonic Acid Digestion and Colorimetric Measurement,"
NTIS Publication, PB94-121738 (1993).
9. QuanTech XRF, Test Kit, Lab Evaluation Study, Draft Final Report, July 1994.
10. Rose, Donna, "Emerging Technologies for Detection of Lead," EPA Contract
68-DO-0137, WA 3-59, Final Report (September 1993).
11. Athanasopoulos, Nick, "Flame Methods Manual for Atomic Absorption," GBC
Scientific Equipment PTY Ltd., Victoria, Australia, pp. 9-11.
12. NIOSH, Method 7300, Issue 2, Elements by TCP, August 1994.
13. Handwipe Preparation Procedure for Lead and Arsenic, University of Cincinnati
Medical Center, Department of Environmental Health, Cincinnati, Ohio.
A-10
-------
Appendix B
Modified SW-846 Method 3050A Acid Digestion
Procedure for Single-Wipe Samples
-------
Modified SW-846 Method 3050A Acid Digestion
Procedure for Single-Wipe Samples
1 Summary of Method
This method is an acid digestion procedure based on SW-846 Method 3050A. The
published method was revised to accommodate dust samples collected on a single wipe.
This method is used for the acid digestion of dust samples and associated quality control
(QC) samples for analysis of lead. It should be noted that this procedure does not use
hydrochloric acid for digestion. The entire wipe sample is digested in nitric acid (HNO3)
and hydrogen peroxide (H2O2). The digestates are diluted to a final volume of 100 mL
following the final cookdown step. The sample digestate will contain approximately 10%
(v/v) HNO3 following digestion.
2 Apparatus and Materials
2.1 Beakers: 250-mL Griffin
2.2 Watch glasses appropriate for designated beakers
2.3 Forceps: polyethylene, or equivalent
2.4 Volumetric flasks with stoppers: 100-mL, Class A
2.5 Funnels: plastic or glass, sized to fit into the 100-mL volumetric flasks
2.6 Thermometers: red alcohol, range of 0° to 110°C
2.7 Hot plates: capable of maintaining a temperature of approximately 95 °C
2.8 Centrifuge (optional)
2.9 Centrifuge tubes: polyethylene with screw caps, 50-mL capacity
2.10 Kimwipes, or equivalent
2.11 Syringes: plastic, 10-mL disposable with Luer-lok fittings
2.12 Gelman Acrodisc filters, female Luer lok, 0.45-|im pore size, or equivalent
2.13 Filter paper: Whatman No. 541, or equivalent
B-l
-------
2.14 Foodservice towel s
2.15 Gloves: disposable, powderless, vinyl
2.16 Pipettes with disposable tips
2.17 Repipet reagent dispensers
2.18 Analytical balance: Capable of weighing to the nearest 0.0001 g
2.19 Drying oven: Capable of maintaining a temperature of ~ 110°C
2.20 Desiccator
3 Reagents/Standard Stock
3.1 Milli-Q reagent water: minimum resistance of 16.67 MQ-cm, or equivalent
3.2 Concentrated nitric acid (HNO3)70% to 71%: Baker Instra-Analyzed grade, or
equivalent
3.3 Hydrogen peroxide (H2O2)30%: A.C.S. reagent grade
3.4 NIST SRM 2704Buffalo River Sediment
4 Quality Control
Quality control (QC) samples are processed with each digestion batch. The QC
samples to be included in each batch are summarized in Table B-l.
B-2
-------
Table B-l. Quality Control Samples Prepared With Each Batch
Description
Method blank
Wipe Blanks
Laboratory control sample (NIST
SRM 2704 spiked into an empty
beaker, i.e., no wipe matrix)
Abbre-
viation
MB
WB
LCSL
LCSH
Purpose
Demonstrate absence of
laboratory contamination
Determine background
levels in the wipe material
Monitor method
performance in the
absence of wipe matrix
effects
Frequency
One with each sample batch
Incorporated into the batch as
specified in the test design.
One of each wipe type per
batch.
One low and one high NIST
SRM 2704 loading level per
batch. The levels are
equivalent to the single-wipe
loadings.
5 Procedure
5.1 Record all reagent sources, lot numbers, and expiration dates used for sample
preparation in the laboratory notebook. In addition, record any inadvertent
deviations to this procedure, unusual happenings, or observations on a real-time
basis as the samples are processed. If any deviation or unusual happenings occur,
notify the Work Assignment Leader or Facility Manager in a manner that permits
corrective action as close to real-time as possible.
5.2 Treat all samples (including QC samples) in a processing batch equally. For
example, if one sample requires additional hydrogen peroxide, then all samples in
that batch must receive the same volume of additional hydrogen peroxide.
5.3 Take care during the execution of each step of the following procedure to ensure
that sample losses do not occur due to splattering or spillage and that samples are
not contaminated or cross-contaminated.
5.4 Document the condition of the samples received in a laboratory notebook.
5.5 Label 250-mL Griffin beakers for each wipe sample and associated quality control
sample to be processed.
5.6 Do not weigh the wipe samples. Open the sample container and carefully transfer
the wipe to a labeled Griffin beaker, using a new pair of plastic gloves and\or plastic
forceps.
5.7 Using a pipette or repipet dispenser, transfer 5 mL of Milli-Q reagent water to the
empty sample container. Cap the bottle with the original cap, shake the bottle to
remove any remaining dust residue from the inside surface of the bottle and cap.
Open the bottle and transfer the Milli-Q rinse solution to the beaker.
5.8 Repeat step 5.7 two times to ensure a quantitative transfer of the dust sample to the
digestion beaker.
B-3
-------
5.9 Transfer all wipe samples in the batch to labeled beakers. A low-level and a high-
level laboratory control sample are processed with each sample batch. The control
samples are designated LCSL and LCSH, respectively. Using an analytical balance
transfer 0.5 ± 0.05 g and 2.0 ± 0.1 g aliquots of NIST SRM 2704 to tared beakers
designated for the LCSL and LCSH samples, respectively. Record the actual
sample weight used.
NOTE: NIST SRM 2704 must be dried according to the procedure listed on the
Certificate of Analysis, and allowed to cool to room temperature in a desiccator
prior to use.
5.10 Carefully, add 25 mL of 1:1 HNO3 to each beaker, gently swirl to mix, and cover
with a watch glass. Gently heat the sample to approximately 95°C and reflux for 10
to 15 min without boiling.
5.11 Allow the sample to cool.
5.12 Add 10 mL of concentrated HNO3, replace the watch glass, and reflux for 30 min
without boiling.
5.13 Repeat step 5.12 to ensure complete oxidation of the sample.
5.14 Remove each watch glass and allow the digests to evaporate to approximately
10 mL without boiling, while maintaining a covering of solution over the bottom of
the beaker.
NOTE: As the watch glasses are removed, rinse the condensate on the watch glass
into the respective sample beaker with a minimum amount of Milli-Q reagent water.
Place the rinsed watch glasses upside down on new clean food service towels or
Kimwipes.
5.15 Following the cookdown step (5.14), replace the watch glass cover, and allow the
sample to cool.
5.16 Add 5 mL of Milli-Q reagent water and 5 mL of 30% hydrogen peroxide (H2O2).
Cover the beaker with a watch glass and return the covered beaker to the hot plate
for warming and to start the peroxide reaction. Care must be taken to ensure that
losses do not occur due to excessively vigorous effervescence. Heat until the
effervescence subsides, then allow the digests to cool.
5.17 Continue to add 30% H2O2 in 1-mL aliquots with warming and cooling, as needed,
until the effervescence is minimal or until the general sample appearance is
unchanged. Do not add more than a total of 10 mL of 30% H2O2, even if
effervescence is still observed.
B-4
-------
5.18 Carefully remove the watch glass and continue heating the acid-peroxide digestate
until the volume has been reduced to approximately 10 mL (see note in step 5.14).
5.19 Allow the digestate to cool; rinse the beaker walls and underside of the watch glass
with Milli-Q reagent water into the beaker. Using a funnel, quantitatively transfer
the digestate to a 100-mL volumetric flask. Dilute to volume with Milli-Q reagent
water and mix thoroughly. Do not filter at this step.
5.20 Remove the particulates in the digestate by filtration, by centrifugation, or by
allowing the sample to settle prior to analysis. A disposable syringe equipped with
an Acrodisc filter can be used to filter a portion of the sample digest prior to
analysis.
5.21 The diluted digest contains approximately 10% (v/v) HNO3. The calibration
standards used for analysis should be made in 10% (v/v) HNO3 to match the sample
digests.
B-5
-------
Appendix C
Modified SW-846 Method 3050A Acid Digestion
Procedure for Two-Wipe Composite Samples
-------
Modified SW-846 Method 3050A Acid Digestion
Procedure for Two-Wipe Composite Samples
1 Summary of Method
This method is an acid digestion procedure based on SW-846 Method 3050A. The
published method was revised to accommodate composite dust samples containing two
wipes. This method is used for the acid digestion of dust samples and associated quality
control (QC) samples for analysis of lead. It should be noted that this procedure does not
use hydrochloric acid for digestion. The entire composite wipe sample is digested in nitric
acid (HNO3) and hydrogen peroxide (H2O2). The digestates are diluted to a final volume of
200 mL following the final cookdown step. The sample digestate will contain approximat-
ely 10% (v/v) HNO3 following digestion.
2 Apparatus and Materials
2.1 Beakers: 400-mL Griffin
2.2 Watch glasses appropriate for designated beakers
2.3 Forceps: polyethylene, or equivalent
2.4 Volumetric flasks with stoppers: 200-mL, Class A
2.5 Funnels: plastic or glass, sized to fit into the 200-mL volumetric flasks
2.6 Thermometers: red alcohol, range of 0° to 110°C
2.7 Hot plates: capable of maintaining a temperature of approximately 95 °C
2.8 Centrifuge (optional)
2.9 Centrifuge tubes: polyethylene with screw caps, 50-mL capacity
2.10 Kimwipes, or equivalent
2.11 Syringes: plastic, 10-mL disposable with Luer-lok fittings
2.12 Gelman Acrodisc filters, female Luer lok, 0.45-|im pore size, or equivalent.
2.13 Filter paper: Whatman No. 541, or equivalent
C-l
-------
2.14 Foodservice towel s
2.15 Gloves: disposable, powderless, vinyl
2.16 Pipettes with disposable tips
2.17 Repipet reagent dispensers
2.18 Analytical balance: Capable of weighing to the nearest 0.0001 g
2.19 Drying oven: Capable of maintaining a temperature of ~ 110°C
2.20 Desiccator
3 Reagents/Standard Stock
3.1 Milli-Q reagent water: minimum resistance of 16.67 MQ-cm, or equivalent
3.2 Concentrated nitric acid (HNO3)70% to 71%: Baker Instra-Analyzed grade, or
equivalent
3.3 Hydrogen peroxide (H2O2)30%: A.C.S. reagent grade
3.4 NIST SRM 2704Buffalo River Sediment
4 Quality Control
Quality control (QC) samples are processed with each digestion batch. The QC
samples to be included in each batch are summarized in Table C-l.
Table C-l. Quality Control Samples Prepared With Each Batch
Description
Method blank
Wipe Blanks
Laboratory control sample (NIST
SRM 2704 spiked into an empty
beaker, i.e., no wipe matrix)
Abbre-
viation
MB
WB
LCSL
LCSH
Purpose
Demonstrate absence of
laboratory contamination
Determine background levels
in the wipe material
Monitor method performance
in the absence of wipe matrix
effects
Frequency
One with each sample batch
Incorporated into the batch
as specified in the test
design. One of each wipe
type per batch.
One low and one high NIST
SRM 2704 loading level per
batch. The levels are
equivalent to the two-wipe
composite loadings.
C-2
-------
5 Procedures
5.1 Record all reagent sources, lot numbers, and expiration dates used for sample
preparation in the laboratory notebook. In addition, record any inadvertent
deviations to this procedure, unusual happenings, or observations on a real-time
basis as the samples are processed. If any deviation or unusual happenings occur,
notify the Work Assignment Leader or Facility Manager in a manner that permits
corrective action as close to real-time as possible.
5.2 Treat all samples (including QC samples) in a processing batch equally. For
example, if one sample requires additional hydrogen peroxide, then all samples in
that batch must receive the same volume of additional hydrogen peroxide.
5.3 Take care during the execution of each step of the following procedure to ensure
that sample losses do not occur due to splattering or spillage and that samples are
not contaminated or cross-contaminated.
5.4 Document the condition of the samples received in a laboratory notebook.
5.5 Label 400-mL Griffin beakers for each composite wipe sample and associated
quality control sample to be processed.
5.6 Do not weigh the composite wipe samples. Open the sample container and
carefully transfer both wipes to a single labeled Griffin beaker, using a new pair of
plastic gloves and\or plastic forceps.
5.7 Using a pipette or repipet dispenser, transfer 5 mL of Milli-Q reagent water to the
empty sample container. Cap the bottle with the original cap, shake the bottle to
remove any remaining dust residue from the inside surface of the bottle and cap.
Open the bottle and transfer the Milli-Q rinse solution to the beaker.
5.8 Repeat step 5.7 two times to ensure a quantitative transfer of the dust sample to the
digestion beaker.
5.9 Transfer all wipe samples in the batch to labeled beakers. A low-level and a high-
level laboratory control sample are processed with each sample batch. The control
samples are designated LCSL and LCSH, respectively. Using an analytical balance
transfer 1.0 ± 0.05 g and 4.0 ± 0.1 g aliquots of NIST SRM 2704 to tared beakers
designated for the LCSL and LCSH samples, respectively. Record the actual
sample weight used.
NOTE: NIST SRM 2704 must be dried according to the procedure listed on the
Certificate of Analysis, and allowed to cool to room temperature in a desiccator
prior to use.
C-3
-------
5.10 Carefully, add 50 mL of 1:1 HNO3 to each beaker, gently swirl to mix, and cover
with a watch glass. Gently heat the sample to approximately 95°C and reflux for 10
to 15 min without boiling.
5.11 Allow the sample to cool.
5.12 Add 20 mL of concentrated HNO3, replace the watch glass, and reflux for 30 min
without boiling.
5.13 Repeat step 5.12 to ensure complete oxidation of the sample.
5.14 Remove each watch glass and allow the digests to evaporate to approximately
20 mL without boiling, while maintaining a covering of solution over the bottom of
the beaker.
NOTE: As the watch glasses are removed, rinse the condensate on the watch glass
into the respective sample beaker with a minimum amount of Milli-Q reagent water.
Place the rinsed watch glasses upside down on new clean food service towels or
Kimwipes.
5.15 Following the cookdown step (5.14), replace the watch glass cover, and allow the
sample to cool.
5.16 Add 10 mL of Milli-Q reagent water and 10 mL of 30% hydrogen peroxide (H2O2).
Cover the beaker with a watch glass and return the covered beaker to the hot plate
for warming and to start the peroxide reaction. Care must be taken to ensure that
losses do not occur due to excessively vigorous effervescence. Heat until the
effervescence subsides, then allow the digests to cool.
5.17 Continue to add 30% H2O2 in 2-mL aliquots with warming and cooling, as needed,
until the effervescence is minimal or until the general sample appearance is
unchanged. Do not add more than a total of 20 mL of 30% H2O2, even if
effervescence is still observed.
5.18 Carefully remove the watch glass and continue heating the acid-peroxide digestate
until the volume has been reduced to approximately 20 mL (see note in step 5.14).
5.19 Allow the digestate to cool; rinse the beaker walls and underside of the watch glass
with Milli-Q reagent water into the beaker. Using a funnel, quantitatively transfer
the digestate to a 200-mL volumetric flask. Dilute to volume with Milli-Q reagent
water and mix thoroughly. Do not filter at this step.
5.20 Remove the particulates in the digestate by filtration, by centrifugation, or by
allowing the sample to settle prior to analysis. A disposable syringe equipped with
an Acrodisc filter can be used to filter a portion of the sample digest prior to
analysis.
C-4
-------
5.21 The diluted digest contains approximately 10% (v/v) HNO3. The calibration
standards used for analysis should be made in 10% (v/v) HNO3 to match the sample
digests.
C-5
-------
Appendix D
Modified SW-846 Method 3050A Acid Digestion
Procedure for Four-Wipe Composite Samples
-------
Modified SW-846 Method 3050A Acid Digestion
Procedure for Four-Wipe Composite Samples
1 Summary of Method
This method is an acid digestion procedure based on SW-846 Method 3050A. The
published method was revised to accommodate composite dust samples containing four
wipes. This method is used for the acid digestion of dust samples and associated quality
control (QC) samples for analysis of lead. It should be noted that this procedure does not
use hydrochloric acid for digestion. The entire composite wipe sample is digested in nitric
acid (HNO3) and hydrogen peroxide (H2O2). The digestates are diluted to a final volume of
500 mL following the final cookdown step. The sample digestate will contain approximat-
ely 10% (v/v) HNO3 following digestion.
2 Apparatus and Materials
2.1 Beakers: 800-mL Griffin
2.2 Watch glasses appropriate for designated beakers
2.3 Forceps: polyethylene, or equivalent
2.4 Volumetric flasks with stoppers: 500-mL, Class A
2.5 Funnels: plastic or glass, sized to fit into the 500-mL volumetric flasks
2.6 Thermometers: red alcohol, range of 0° to 110°C
2.7 Hot plates: capable of maintaining a temperature of approximately 95 °C
2.8 Centrifuge (optional)
2.9 Centrifuge tubes: polyethylene with screw caps, 50-mL capacity
2.10 Kimwipes, or equivalent
2.11 Syringes: plastic, 10-mL disposable with Luer-lok fittings
2.12 Gelman Acrodisc filters, female Luer lok, 0.45-|im pore size, or equivalent
2.13 Filter paper: Whatman No. 541, or equivalent
D-l
-------
2.14 Foodservice towel s
2.15 Gloves: disposable, powderless, vinyl
2.16 Pipettes with disposable tips
2.17 Repipet reagent dispensers
2.18 Analytical balance: Capable of weighing to the nearest 0.0001 g
2.19 Drying oven: Capable of maintaining a temperature of ~ 110°C
2.20 Desiccator
3 Reagents/Standard Stock
3.1 Milli-Q reagent water: minimum resistance of 16.67 MQ-cm, or equivalent
3.2 Concentrated nitric acid (HNO3)70% to 71%: Baker Instra-Analyzed grade, or
equivalent
3.3 Hydrogen peroxide (H2O2)30%: A.C.S. reagent grade
3.4 NIST SRM 2704Buffalo River Sediment
4 Quality Control
Quality control (QC) samples are processed with each digestion batch. The QC
samples to be included in each batch are summarized in Table D-l.
D-2
-------
Table D-l. Quality Control Samples Prepared With Each Batch
Description
Method blank
Wipe Blanks
Laboratory control sample (NIST
SRM 2704 spiked into an empty
beaker, i.e., no wipe matrix)
Abbr
e-
viatio
n
MB
WB
LCSL
LCSH
Purpose
Demonstrate absence of
laboratory contamination
Determine background
levels in the wipe material
Monitor method
performance in the
absence of wipe matrix
effects
Frequency
One with each sample batch
Incorporated into the batch as
specified in the test design.
One of each wipe type per
batch.
One low and one high NIST
SRM 2704 loading level per
batch. The levels are
equivalent to the four-wipe
composite loadings.
5 Procedure
5.1 Record all reagent sources, lot numbers, and expiration dates used for sample
preparation in the laboratory notebook. In addition, record any inadvertent
deviations to this procedure, unusual happenings, or observations on a real-time
basis as the samples are processed. If any deviation or unusual happenings occur,
notify the Work Assignment Leader or Facility Manager in a manner that permits
corrective action as close to real-time as possible.
5.2 Treat all samples (including QC samples) in a processing batch equally. For
example, if one sample requires additional hydrogen peroxide, then all samples in
that batch must receive the same volume of additional hydrogen peroxide.
5.3 Take care during the execution of each step of the following procedure to ensure
that sample losses do not occur due to splattering or spillage and that samples are
not contaminated or cross-contaminated.
5.4 Document the condition of the samples received in a laboratory notebook.
5.5 Label 800-mL Griffin beakers for each composite wipe sample and associated
quality control sample to be processed.
5.6 Do not weigh the composite wipe samples. Open the sample container and
carefully transfer the four wipes to a single labeled Griffin beaker, using a new pair
of plastic gloves and\or plastic forceps.
5.7 Using a pipette or repipet dispenser, transfer 10 mL of Milli-Q reagent water to the
empty sample container. Cap the bottle with the original cap, shake the bottle to
remove any remaining dust residue from the inside surface of the bottle and cap.
Open the bottle and transfer the Milli-Q rinse solution to the beaker.
D-3
-------
5.8 Repeat step 5.7 two times to ensure a quantitative transfer of the dust sample to the
digestion beaker.
5.9 Transfer all wipe samples in the batch to labeled beakers. A low-level and a high-
level laboratory control sample are processed with each sample batch, which are
designated LCSL and LCSH, respectively. Using an analytical balance transfer 2.0
± 0.05 g and 8.0 ± 0.1 g aliquots of NIST SRM 2704 to tared beakers designated for
the LCSL and LCSH samples, respectively. Record the actual sample weight used.
NOTE: NIST SRM 2704 must be dried according to the procedure listed on the
Certificate of Analysis, and allowed to cool to room temperature in a desiccator
prior to use.
5.10 Carefully, add 100 mL of 1:1 HNO3 to each beaker, gently swirl to mix, and cover
with a watch glass. Gently heat the sample to approximately 95°C and reflux for 10
to 15 min without boiling.
5.11 Allow the sample to cool.
5.12 Add 40 mL of concentrated HNO3, replace the watch glass, and reflux for 30 min
without boiling.
5.13 Repeat step 5.12 to ensure complete oxidation of the sample.
5.14 Remove each watch glass and allow the digests to evaporate to approximately
50 mL without boiling, while maintaining a covering of solution over the bottom of
the beaker.
NOTE: As the watch glasses are removed, rinse the condensate on the watch glass
into the respective sample beaker with a minimum amount of Milli-Q reagent water.
Place the rinsed watch glasses upside down on new clean food service towels or
Kimwipes.
5.15 Following the cookdown step (5.14), replace the watch glass cover, and allow the
sample to cool.
5.16 Add 20 mL of Milli-Q reagent water and 20 mL of 30% hydrogen peroxide (H2O2).
Cover the beaker with a watch glass and return the covered beaker to the hot plate
for warming and to start the peroxide reaction. Care must be taken to ensure that
losses do not occur due to excessively vigorous effervescence. Heat until the
effervescence subsides, then allow the digests to cool.
5.17 Continue to add 30% H2O2 in 4-mL aliquots with warming and cooling, as needed,
until the effervescence is minimal or until the general sample appearance is
unchanged. Do not add more than a total of 40 mL of 30% H2O2, even if
effervescence is still observed.
D-4
-------
5.18 Carefully remove the watch glass and continue heating the acid-peroxide digestate
until the volume has been reduced to approximately 50 mL (see note in step 5.14).
5.19 Allow the digestate to cool; rinse the beaker walls and underside of the watch glass
with Milli-Q reagent water into the beaker. Using a funnel, quantitatively transfer
the digestate to a 500-mL volumetric flask. Dilute to volume with Milli-Q reagent
water and mix thoroughly. Do not filter at this step.
5.20 Remove particulates in the digestate by filtration, by centrifugation, or by allowing
the sample to settle prior to analysis. A disposable syringe equipped with an
Acrodisc filter can be used to filter a portion of the sample digest prior to analysis.
5.21 The diluted digest contains approximately 10% (v/v) HNO3. The calibration
standards used for analysis should be made in 10% (v/v) HNO3 to match the sample
digests.
D-5
-------
Appendix E
WOHL Acid Digestion Procedure for
Single-Wipe Samples
-------
WOHL Acid Digestion Procedure for
Single-Wipe Samples
1 Summary of Method
This digestion procedure is based on the method used by the Wisconsin Occupational
Health Laboratory (WOHL) and Appendix 14.2 of the HUD Guidelines for digesting
samples containing a single wipe. This method is used for the acid digestion of dust
samples and associated quality control (QC) samples for analysis of lead. The wipe
samples are digested in nitric acid (HNO3) and hydrochloric acid (HC1). The digestates are
diluted to a final volume of 50 mL following a 20- to 30-min reflux at 100°C and a cooling
step. The sample digestates contain approximately 4% (v/v) HNO3 and 4% (v/v) HC1.
2 Apparatus and Materials
2.1 Beakers: 125-mL, Phillips conical
2.2 Watch glasses appropriate for designated beakers
2.3 Scissors: stainless steel, or equivalent
2.4 Volumetric flasks with stoppers: 50-mL, Class A
2.5 Funnels: plastic or glass, sized to fit into the 50-mL volumetric flasks
2.6 Thermometers: red alcohol, range of 0° to 110°C
2.7 Stirring rods: glass or plastic
2.8 Hot plates: capable of maintaining a temperature of approximately 100°C
2.9 Centrifuge
2.10 Centrifuge tubes: polyethylene with screw caps, 50-mL capacity
2.11 Filter paper: Whatman No. 541, or equivalent
2.12 Kimwipes, or equivalent
2.13 Foodservice towel s
2.14 Gloves: disposable, powderless, vinyl
E-l
-------
2.15 Pipettes with disposable tips
2.16 Repipet reagent dispensers
2.17 Analytical balance: Capable of weighing to the nearest 0.0001 g
2.18 Drying oven: Capable of maintaining a temperature of ~ 110°C
2.19 Desiccator
3 Reagents/Standard Stock
3.1 Milli-Q reagent water: minimum resistance of 16.67 MQ-cm, or equivalent
3.2 Concentrated nitric acid (HNO3)70% to 71%: Baker Instra-Analyzed grade, or
equivalent
3.3 Hydrochloric acid (HC1)36.5% to 38%: Baker Instra-Analyzed grade, or
equivalent
3.4 NIST SRM 2704Buffalo River Sediment
4 Quality Control
Quality control (QC) samples are processed with each digestion batch. The QC
samples to be included in each batch are summarized in Table E-l.
Table E-l. Quality Control Samples Prepared With Each Batch
Description
Method blank
Wipe blanks
Laboratory control sample (NIST
SRM 2704 spiked into an empty
beaker, i.e., no wipe matrix)
Abbre-
viation
MB
WB
LCSL
LCSH
Purpose
Demonstrate absence of
laboratory contamination
Determine background
levels in the wipe material
Monitor method
performance in the
absence of sample matrix
effects
Frequency
One with each sample batch
Incorporated into the batch as
specified in the test design. One
of each wipe type per batch.
One low and one high NIST
SRM 2704 loading level per batch.
E-2
-------
5 Procedure
5.1 Record all reagent sources, lot numbers, and expiration dates used for sample
preparation in the laboratory notebook. In addition, record any inadvertent
deviations to this procedure, unusual happenings, or observations on a real-time
basis as the samples are processed. If any deviation or unusual happenings occur,
notify the Work Assignment Leader or Facility Manager in a manner that permits
corrective action as close to real-time as possible.
5.2 Treat all samples (including QC samples) in a processing batch equally. Care
should be taken during the execution of each step of the following procedure to
ensure the sample losses do not occur due to splattering or spillage and that samples
are not contaminated or cross-contaminated.
5.3 Document the condition of the samples received in a laboratory notebook.
5.4 Label 125-mL Phillips conical beakers for each wipe sample and associated quality
control sample to be processed.
5.5 Do not weigh the wipe samples. Open the sample container and remove the wipe
using a new pair of plastic gloves and/or plastic forceps. Hold the wipe over the
neck of a 125-mL Phillips conical beaker designated for the sample, and using
scissors cut the wipe into smaller pieces collecting the wipe pieces in the labeled
beaker.
5.6 Clean and dry the scissors between samples using Kimwipes and Milli-Q reagent
water.
5.7 Quantitatively rinse the inside surfaces of the sample bottle and cap with Milli-Q
water, transferring the rinses into the respective sample beaker.
5.8 Transfer all wipe samples in the batch to labeled beakers as described above. A
low-level and a high-level laboratory control sample are processed with each sample
batch. The control samples are designated LCSL and LCSH, respectively. Aliquots
of NIST SRM 2704 are transferred to tared beakers designated for the LCSL and
LCSH samples, respectively. The amount of NIST SRM used for the LCSL and
LCSH are shown in the following table. Record the actual sample weight used.
Amount of NIST SRM Amount of NIST SRM
2704 spiked LCSL 2704 spiked LCSH
(9) (9)
0.5 ±0.05 2.0 ±0.1
E-3
-------
NOTE: NIST SRM 2704 must be dried according to the procedure listed on the
Certificate of Analysis, and allowed to cool to room temperature in a desiccator
prior to use.
5.9 Add 10 mL of Milli-Q water to each sample covering the wipe.
5.10 Add 2 mL of concentrated HNO3 and 2 mL of concentrated HC1 to each beaker,
gently swirl to mix, and cover with a watch glass. Gently heat the sample to
approximately 100°C and reflux for 20 to 30 min.
5.11 Allow the digestates to cool; rinse the beaker walls and underside of the watch glass
with Milli-Q reagent water into the beaker. Using a stirring rod, squeeze the wipe
material, and then quantitatively transfer the acid digestate and the bulk material to
a 50-mL volumetric flask.
5.12 To help ensure a quantitative transfer of the sample digest, if there is too much bulk
material remaining in the beaker, rinse the material remaining with Milli-Q water,
stir and squeeze the wipe material with the stirring rod, and transfer this rinsate to
the 50-mL volumetric flask. Dilute to volume with Milli-Q reagent water and mix
thoroughly. Do not filter at this step.
5.13 Filter a portion of the digest through a Whatman 541 filter paper. Then centrifuge
this portion at 9,000 rpm for 20 min. The supernatant is drawn off and analyzed by
AA, ICP, or other equivalent methodology.
5.14 The diluted digest contains approximately 4% (v/v) HNO3 and 4% (v/v) HC1. The
calibration standards used for analysis should be made in 4% (v/v) HNO3 to match
the sample digests.
E-4
-------
Appendix F
WOHL Acid Digestion Procedure for Two- and
Four-Wipe Composite Samples
-------
WOHL Acid Digestion Procedure for Two- and
Four-Wipe Composite Samples
1 Summary of Method
This digestion procedure is based on the method used by the Wisconsin Occupational
Health Laboratory (WOHL) and Appendix 14.2 of the Hud Guidelines for digesting
composite samples containing up to four wipes. This method is used for the acid digestion
of dust samples and associated quality control (QC) samples for analysis of lead. The wipe
composites are digested in nitric acid (HNO3) and hydrochloric acid (HC1). The digestates
are diluted to a final volume of 100 mL following a 5-min reflux at 100°C and cooling
step. The sample digestates contain approximately 8% (v/v) HNO3 and 8% (v/v) HC1.
2 Apparatus and Materials
2.1 Beakers: 250-mL, Phillips conical
2.2 Watch glasses appropriate for designated beakers
2.3 Scissors: stainless steel, or equivalent
2.4 Volumetric flasks with stoppers: 100-mL, Class A
2.5 Funnels: plastic or glass, sized to fit into the 100-mL volumetric flasks
2.6 Thermometers: red alcohol, range of 0° to 110°C
2.7 Stirring rods: glass or plastic
2.8 Hot plates: capable of maintaining a temperature of approximately 100°C
2.9 Centrifuge
2.10 Centrifuge tubes: polyethylene with screw caps, 50-mL capacity
2.11 Filter paper: Whatman No. 541, or equivalent
2.12 Kimwipes, or equivalent
2.13 Foodservice towel s
2.14 Gloves: disposable, powderless, vinyl
F-l
-------
2.15 Pipettes with disposable tips
2.16 Repipet reagent dispensers
2.17 Analytical balance: Capable of weighing to the nearest 0.0001 g
2.18 Drying oven: Capable of maintaining a temperature of ~ 110°C
2.19 Desiccator
3 Reagents/Standard Stock
3.1 Milli-Q reagent water: minimum resistance of 16.67 MQ-cm, or equivalent
3.2 Concentrated nitric acid (HNO3)70% to 71%: Baker Instra-Analyzed grade, or
equivalent
3.3 Hydrochloric acid (HC1)36.5% to 38%: Baker Instra-Analyzed grade, or
equivalent
3.4 NIST SRM 2704Buffalo River Sediment
4 Quality Control
Quality control (QC) samples are processed with each digestion batch. The QC
samples to be included in each batch are summarized in Table F-l.
Table F-l. Quality Control Samples Prepared With Each Batch
Description
Method blank
Wipe blanks
Laboratory control sample (NIST
SRM 2704 spiked into an empty
beaker, i.e., no wipe matrix)
Abbr
e-
viatio
n
MB
WB
LCSL
LCSH
Purpose
Demonstrate absence of
laboratory contamination
Determine background
levels in the wipe material
Monitor method
performance in the
absence of sample matrix
effects
Frequency
One with each sample batch
Incorporated into the batch as
specified in the test design.
One of each wipe type per
batch.
One low and one high NIST
SRM 2704 loading level per
batch. The levels are equivalent
to the two- and four-wipe
loadings.
F-2
-------
5 Procedure
5.1 Record all reagent sources, lot numbers, and expiration dates used for sample
preparation in the laboratory notebook. In addition, also record any inadvertent
deviations to this procedure, unusual happenings, or observations on a real-time
basis as the samples are processed. If any deviation or unusual happenings occur,
notify the Work Assignment Leader or Facility Manager in a manner that permits
corrective action as close to real-time as possible.
5.2 Treat all samples (including QC samples) in a processing batch equally. Care
should be taken during the execution of each step of the following procedure to
ensure the sample losses do not occur due to splattering or spillage and that samples
are not contaminated or cross-contaminated.
5.3 Document the condition of the samples received in a laboratory notebook.
5.4 Label 250-mL Phillips conical beakers for each wipe sample and associated quality
control sample to be processed.
5.5 Do not weigh the wipe samples. Open the sample container and remove the wipe
using a new pair of plastic gloves and/or plastic forceps. Hold the wipe over the
neck of a 250-mL Phillips conical beaker designated for the sample, and using
scissors cut the wipe into smaller pieces collecting the wipe pieces in the labeled
beaker. Repeat this procedure for each wipe contained in the two- and four-wipe
composite samples.
5.6 Clean and dry the scissors between samples using Kimwipes and Milli-Q reagent
water.
5.7 Using a pipette or repipet dispenser, transfer 10 mL of Milli-Q reagent water to the
empty sample container. Cap the bottle with the original cap, shake the bottle to
remove any remaining dust residue from the inside surface of the bottle and cap.
Open the bottle and transfer the Milli-Q rinse solution to the beaker.
5.8 Repeat step 5.7 two times to ensure a quantitative transfer of the dust sample to the
digestion beaker. Then add an additional 10 mL of Milli-Q reagent water to each
beaker.
5.9 Transfer all wipe samples in the batch to labeled beakers. A low-level and a high-
level laboratory control sample are processed with each sample batch. The control
samples are designated LCSL or LCSH, respectively. Aliquots of NIST SRM 2704
are transferred to tared beakers designated for the LCSL and LCSH samples,
respectively. The amount of NIST SRM used for the LCSL and LCSH is shown in
the following table. Record the actual sample weight used.
F-3
-------
No. of wipes
per sample
for the batch
2
4
Amount of NIST
SRM 2704
spiked LCSL
(g)
1.0 ±0.05
2.0 ±0.05
Amount of NIST
SRM 2704
spiked LCSH
(g)
4.0 ±0.1
8.0 ±0.1
NOTE: NIST SRM 2704 must be dried according to the procedure listed on the
Certificate of Analysis, and allowed to cool to room temperature in a desiccator
prior to use.
5.10 Add 8 mL of concentrated HNO3 and 8 mL of concentrated HC1 to each beaker,
gently swirl to mix, and cover with a watch glass. Gently heat the sample to
approximately 100°C and reflux for 50 min.
5.11 Allow the digestates to cool; rinse the beaker walls and underside of the watch glass
with Milli-Q reagent water into the beaker. Using a stirring rod, squeeze the wipe
material, and then quantitatively transfer the acid digestate to a 100-mL volumetric
flask.
5.12 To help ensure a quantitative transfer of the sample digest, add approximately
20 mL of Milli-Q reagent water to the wipe material remaining in the beaker, stir
and squeeze the wipe material with the stirring rod, and transfer this rinsate to the
100-mL volumetric flask. Dilute to volume with Milli-Q reagent water and mix
thoroughly. Do not filter at this step.
5.13 Filter a portion of the digest through a Whatman 541 filter paper. Then centrifuge
this portion at 9,000 rpm for 20 min. The supernatant is drawn off and analyzed by
AA, ICP, or other equivalent methodology.
5.14 The diluted digest contains approximately 8% (v/v) HNO3 and 8% (v/v) HC1. The
calibration standards used for analysis should be made in 8% (v/v) HNO3 to match
the sample digests.
F-4
-------
Appendix G
Preparation of Reference Material Wipe Samples
for the Composite Wipe Investigation
-------
1 Summary of Method
This procedure is used for loading lead (Pb) using a standard reference material (SRM)
onto wipes for the composite wipe study. The National Institute of Standards and
Technology (NIST) SRM 2704 Buffalo River Sediment with a certified Pb concentration
of 161 |ig/g is used for this loading. Three types of wipes are used: (1) Wash a-bye Baby
wipes, (2) Baby Wipes with Lanolin, and (3) Wash'n Dri wipes.
The wipes are loaded with NIST SRM 2704 at one of two levels, either 0.50 g or 2 g.
The weighed portion of SRM is placed on the wipe, and the wipe is carefully folded and
transferred to a labeled Nalgene bottle for storage until sample preparation.
2 Apparatus and Materials
2.1 Gloves: disposable, powderless, vinyl, or equivalent
2.2 Ziplock bags: quart and gallon size
2.3 Sample labels
2.4 NIST SRM 2704 Buffalo River Sediment
2.5 Wash a-bye Baby wipes
2.6 Baby Wipes with Lanolin
2.7 Wash'n Dri wipes
2.8 Analytical balance: capable of weighing to the nearest 0.0001 g
2.9 Weighing paper
2.10 Drying oven: capable of maintaining a temperature of ~ 110 °C
2.11 Beaker: appropriate size for drying NIST SRM 2704
2.12 Nalgene bottles: polyethylene, 125 mL and 250 mL
2.13 Scissors: stainless steel, or equivalent
2.14 Desiccator
G-l
-------
3 Procedure
3.1 Transfer a sufficient amount of NIST SRM 2704 to a beaker for drying. The
amount of NIST SRM 2704 required for a given weighing session is based on the
number of wipes to be loaded and the loading level (0.50 g or 2 g) for each wipe.
Place the beaker containing NIST SRM 2704 in a drying oven and dry the material
for a minimum of 2 h at ~ 110 °C.
3.2 Remove the beaker from the drying oven, place the beaker in a desiccator, and allow
the SRM to cool to room temperature.
3.3 Select the appropriate wipe material, as designated in the test design, which is
targeted for loading with the reference material.
3.4 Using clean gloves and a pair of scissors, cut open a new Ziplock bag (quart or
gallon size, whichever is appropriate based on the dimensions of the wipe). Open
the bag exposing the inside surface, and place the bag inside surface up on the lab
bench. Unfold a new wipe and place the wipe on the inside surface of the Ziplock
bag located on the lab bench.
3.5 Remove the pre-dried portion of NIST SRM 2704 from the desiccator and weigh
the designated aliquot, either 0.50 ± 0.05 g or 2.0 ± 0.1 g (as designated in the test
design), of the pre-dried NIST SRM 2704 onto a clean piece of tared weighing
paper. Record the weight of SRM used.
NOTE: Return the beaker containing the pre-dried NIST SRM 2704 to the
desiccator between each respective sample weighing.
3.6 With gloves on, carefully tap the weighed portion of the reference material from the
weighing paper into the center of the wipe. Carefully fold the wipe in half several
times, pressing the NIST SRM into the wipe while folding. Wipe off any residue
remaining on the weighing paper using the folded wipe. Care should be taken to
ensure that losses of the SRM do not occur during the folding process.
3.7 Transfer the folded wipe to a labeled Nalgene bottle according to the table below:
Number of
wipes per sample
1
2
4
Nalgene
bottle size
125 ml
250 ml
250 ml
G-2
-------
Place each Nalgene bottle and the additional bar code labels in a Ziplock bag for
secondary containment and seal the bag. NOTE: For multiwipe composites, the
same Ziplock bag used for the first wipe in the composite (Step 3.4) should be
reused for all wipes (2 or 4) contained in the composite.
3.8 When preparing a composite wipe sample of two or four wipes, repeat steps 3.3
through 3.7 for each wipe contained in the composite. It should be noted that for
the two- and four-wipe composite samples, all wipes that comprise the composite
sample should be transferred to the same labeled Nalgene bottle.
3.9 Record any unusual observations noted in the laboratory notebook.
3.10 Store the wipe samples loaded with the reference material at room temperature until
sample digestion.
G-3
-------
Appendix H
Laboratory Data and Lead (Pb) Recovery Results
-------
Laboratory Data and Lead (Pb) Recovery Results
All laboratory data and recovery results for 528 samples are presented in this
appendix. The data are organized in the following order:
laboratory code (Nos. 1, 2, and 3)
preparation method (EPA 3050A and WOHL)
type of composite (one-, two-, and four-wipe composite)
sample type (loaded wipe, QC sample type: laboratory control sample, method
blank, wipe blank)
total sample target (nominal) SRM loading (0.5 g to 8.0 g)
wipe brand (A, B, or C)
lot code (1 and 2 for brands A and B only)
Page breaks occur for each combination of laboratory, preparation method, and
composite type. The data tables were generated in SAS with the following headers:
SAS variable name Definition (unit)
OBS Running observation number (1 through 528)
LAB_ID Laboratory code
METHOD Preparation method
COMPOSIT Composite type
BRAND_W Wipe brand
LOT Wipe brand lot code
SPLETYPE Sample type
BARCODE Sample barcode ID
TOT_SRM Target (nominal) SRM loading (g)
TOT_LAB Actual SRM loading (g)
TOT_PB Total lead (Pb) loading (|ig)
TCP TCP lead (Pb) result (|ig)
ICP_Q TCP qualifier (ND = nondetect, Q = quantified)
FAA FAA lead (Pb) result (|ig)
FAA_Q FAA qualifier (ND = nondetect, Q = quantified)
ICP_REC TCP recovery (
FAA_REC FAA recovery
The values at the bottom of each page are column totals corresponding to the
variable(s) on the left-hand side of each page. The values are sequentially totaled across
composite type, preparation method, and laboratory.
H-l
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
1
0
B
S I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
L H
\ E
J T
H
r o
) D
3050A
3050A
3050A
3050A
3050A
30SOA
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
H
P
0
S
I
T
1-UIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
1-WIPE
1-WIPE
1-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
B
R
A
N
D
W
A
A
B
B
C
C
A
A
B
B
C
C
QC
QC
QC
QC
QC
QC
A
B
C
L
0
T
1
2
1
2
i
2
1
2
m
i
_
B
1
2
L
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-LCS
QC-LCS
QC-MB
QC-MB
OC-WB
QC-WB
QC-WB
AB_ID=1 Ml
B
A
R
C
0
D
E
906013
906460
906163
906241
906328
906329
906046
906121
906196
906274
906394
906395
905710
905770
905711
905771
905706
905769
906001
906229
906307
.iHOU=iu:
T
0
T
S
R
M
0.5
0.5
0.5
0.5
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
0.5
0.5
2.0
2.0
,
.
m
.
)UA LUMHUSI
T
0
T
L
A
B
0.5006
0.5032
0.5059
0.5054
0.5122
0.5088
2.0075
2.0082
2.0079
1.9833
2.0351
2.0128
0.5017
0.5015
2.0053
2.0004
.
.
,
.
=i-wirt -
' T
0
T
P
B
80.60
81.02
81.45
81.37
82.46
81.92
323.21
323.32
323.27
319.31
327.65
324.06
80.77
80.74
322.85
322.06
.
.
.
.
>
I
C
p
73.8
78.0
79.5
72.6
80.4
76.7
263.0
271.0
264.0
266.0
270.0
262.0
80.2
80.6
287.0
286.0
.
.
.
.
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
ND
-
F
A
A
73.5
73.5
73.5
75.6
81.8
73.5
305.0
313.0
311.0
303.0
315.0
305.0
75.6
81.8
328.0
324.0
.
.
.
.
f
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
ND
I
C
P
R
E
C
91.57
96.28
97.61
89.22
97.50
93.63
81.37
83.82
81.67
83.30
82.40
80.85
99.29
99.82
88.89
88.80
.
.
.
.
F
A
A
R
E
C
91.19
90.72
90.24
92.91
99.19
89.73
94.37
96.81
96.20
94.89
96.14
94.12
93.59
101.31
101.59
100.60
.
.
.
.
COMPOS IT
20.0 20.0998 3236.07 2790.8
3112.8
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
L M
A E
B T
H
I 0
D D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
1 3050A
C
0
M
P
0
S
I
T
2-WIPE
2-U1PE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
.
_
_
B
1
1
2
2
1
1
2
2
.
.
.
_
_
m
m
2
1
L
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
OC-LCS
OC-LCS
QC-MB
QC-WB
QC-WB
QC-WB
«B iu=i nt
B
A
R
C
0
D
E
906022
906023
906097
906098
906172
906173
906250
906251
906346
906347
906348
906349
906055
906056
906130
906131
906205
906206
906283
906284
906412
906413
906414
906415
905712
905713
905707
906079
906154
906313
:inuu=ou:
T
0
T
S
R
M
1.0
1.0
1.0
1.0
1.0
1.0
1.0
.0
.0
.0
.0
.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
.
.
.
>UA Lunrusi
T
0
T
L
A
B
.0098
.0110
.0141
.0106
.0320
.0224
.0155
.0197
.0418
.0064
.0114
.0428
4.0082
4.0671
4.0605
4.0363
4.0169
4.0193
4.1336
4.1156
4.0204
4.0008
3.9977
4.0205
1.0026
4.0008
B
.
,
=£-wire -
T
0
T
P
B
162.58
162.77
163.27
162.71
166.15
164.61
163.50
164.17
167.73
162.03
162.84
167.89
645.32
654.80
653.74
649.84
646.72
647.11
665.51
662.61
647.28
644.13
643.63
647.30
161.42
644.13
.
.
.
I
C
P
158.0
151.0
157.0
150.0
156.0
159.0
160.0
152.0
156.0
170.0
154.0
152.0
551.0
554.0
577.0
555.0
562.0
541.0
552.0
568.0
578.0
560.0
577.0
566.0
165.0
563.0
.
.
f
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
F
A
A
155.0
143.0
155.0
147.0
143.0
155.0
155.0
151.0
139.0
155.0
151.0
147.0
652.0
652.0
652.0
643.0
647.0
631.0
635.0
660.0
668.0
660.0
647.0
656.0
160.0
647.0
.
.
m
f
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
I
C
P
R
E
C
97.18
92.77
96.16
92.19
93.89
96.59
97.86
92.59
93.01
104.92
94.57
90.54
85.38
84.61
88.26
85.41
86.90
83.60
82.94
85.72
89.30
86.94
89.65
87.44
102.22
87.40
.
.
m
F
A
A
R
E
C
95.34
87.85
94.93
90.35
86.07
94.16
94.80
91.98
82.87
95.66
92.73
87.56
101.04
99.57
99.73
98.95
100.04
97.51
95.42
99.61
103.20
102.46
100.52
101.34
99.12
100.45
.
.
,
COMPOS IT
65.0 65.7378 10583.79 9344.0
10406
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
COMPOS IT
METHOD
L M
K E
B T
H
F 0
) D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
305 OA
C
0
M
P
0
S
I
T
4-W1PE
4-UIPE
4-WIPE
4-WIPE
4-UIPE
4-WIPE
4-UIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
QC
QC
QC
A
A
B
B
C
C
L
0
T
1
1
2
2
1
1
2
2
B
.
_
.
1
1
2
2
1
1
2
2
t
.
_
_
.
.
_
m
f
1
2
1
2
_
L
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-LCS
QC-LCS
QC-MB
QC-MB
QC-WB
QC-WB
QC-WB
QC-WB
QC-WB
QC-WB
«B ltl=l 1
B
A
R
C
0
D
E
906034
906043
906109
906118
906184
906193
906262
906271
906370
906371
906388
906389
906067
906076
906142
906151
906217
906218
906301
906302
906436
906437
906454
906455
905714
905715
905716
905717
905708
905709
906007
906085
906160
906238
906319
906325
it i nuu=ji
T
0
T
S
R
M
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
2.0
8.0
8.0
t
m
t
140.0
225.0
J3U« lunrua
T
0
T
L
A
B
2.0125
2.0264
2.0291
2.0264
2.0072
2.0527
1.9971
2.0321
2.0270
2.0236
2.0428
2.0418
8.0149
8.0444
8.0380
8.0403
8.0696
8.0072
8.0144
8.0794
8.1699
8.0875
8.0953
8.0274
2.0017
2.0020
8.0103
8.0031
i
.
t
t
.
141.0241
226.8617
i=<»-wire -
T
0
T
P
B
324.01
326.25
326.69
326.25
323.16
330.48
321.53
327.17
326.35
325.80
328.89
328.73
1290.40
1295.15
1294.12
1294.49
1299.21
1289.16
1290.32
1300.78
1315.35
1302.09
1303.34
1292.41
322.27
322.32
1289.66
1288.50
f
f
.
f
t
m
n
22704.88
36524.73
I
C
P
292.0
342.0
318.0
309.0
304.0
336.0
309.0
338.0
312.0
327.0
322.0
388.0
1050.0
1160.0
1170.0
1200.0
1210.0
1160.0
1150.0
1170.0
1170.0
1160.0
1260.0
1230.0
304.0
331.0
1200.0
1200.0
t
n
f
t
f
.
.
21022
33157
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
NO
ND
ND
ND
ND
ND
ND
ND
F
A
A
253.0
305.0
294.0
284.0
284.0
284.0
263.0
294.0
284.0
315.0
284.0
357.0
1190.0
1300.0
1300.0
1320.0
1360.0
1270.0
1270.0
1300.0
1380.0
1330.0
1350.0
1350.0
274.0
274.0
1320.0
1300.0
m
t
f
t
22389
35908
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
ND
ND
ND
ND
ND
ND
ND
ND
1
C
P
R
E
C
90.12
104.83
97.34
94.71
94.07
101.67
96.10
103.31
95.60
100.37
97.90
118.03
81.37
89.57
90.41
92.70
93.13
89.98
89.13
89.95
88.95
89.09
96.67
95.17
94.33
102.69
93.05
93.13
m
t
m
.
.
f
f
f
A
A
R
E
C
78.08
93.49
89.99
87.05
87.88
85.93
81.80
89.86
87.02
96.69
86.35
108.60
92.22
100.37
100.45
101.97
104.68
98.51
98.43
99.94
104.91
102.14
103.58
104.46
85.02
85.01
102.35
100.89
.
t
m
.
.
f
m
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
L M
A E
B T
0 H
B To
S 0 D
88 1 UOHL
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
107 1 WOHL
108 1 WOHL
109 1 WOHL
110 UOHL
111 WOHL
112 WOHL
113 WOHL
114 WOHL
115 WOHL
116 WOHL
117 WOHL
118 WOHL
119 WOHL
120 WOHL
C
0
H
P
0
S
I
T
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
B
R
A
N
D
W
A
A
A
A
B
8
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
f
.
.
1
1
2
2
1
1
2
2
,
.
m
f
m
t
B
.
m
1
2
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-LCS
QC-LCS
QC-HB
QC-MB
QC-WB
QC-WB
QC-WB
L«b_IU=l
B
A
R
C
0
0
E
906016
906017
906091
906092
906166
906167
906244
906245
906334
906335
906336
906337
906049
906050
906125
906461
906199
906200
906278
906462
906400
906401
906402
906403
905721
905773
905722
905774
905718
905772
906004
906232
906310
Pit 1 MUU=W
T
0
T
S
R
M
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
0.5
0.5
2.0
2.0
.
a
.
.
UML tOMHUbl
T
0
T
L
A
B
0.5040
0.5063
0.5116
0.5059
0.5127
0.5129
0.5036
0.5090
0.5154
0.5158
0.5111
0.5258
2.0096
2.0035
2.0051
2.0118
2.0165
2.0061
2.0853
2.0129
2.0085
2.0147
2.0227
2.0032
0.5018
0.5010
2.0028
2.0067
.
.
.
.
I=1-WIPE -
T
0
T
P
B
81.14
81.51
82.37
81.45
82.54
82.58
81.08
81.95
82.98
83.04
82.29
84.65
323.55
322.56
322.82
323.90
324.66
322.98
335.73
324.08
323.37
324.37
325.65
322.52
80.79
80.66
322.45
323.08
.
_
_
f
I
C
P
66.7
61.4
61.5
61.0
68.8
69.4
61.7
67.5
68.6
76.6
69.1
70.6
237.0
242.0
230.0
233.0
257.0
250.0
261.0
295.0
269.0
259.0
281.0
257.0
70.9
71.2
302.0
306.0
>
f
f
I
C
P
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
NO
ND
NO
ND
ND
F
A
A
73.3
66.2
68.3
65.2
76.4
76.4
66.2
72.3
73.3
84.5
74.3
76.4
255.0
258.0
244.0
251.0
270.0
270.0
287.0
308.0
287.0
272.0
295.0
275.0
78.4
74.3
324.0
318.0
_
F
A
A
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
ND
I
C
P
R
E
C
82.20
75.32
74.67
74.89
83.35
84.04
76.10
82.37
82.67
92.24
83.97
83.40
73.25
75.02
71.25
71.94
79.16
77.40
77.74
91.03
83.19
79.85
86.29
79.69
87.76
88.27
93.66
94.71
F
A
A
R
E
C
90.33
81.21
82.92
80.05
92.56
92.52
81.65
88.23
88.34
101.75
90.29
90.25
78.81
79.98
75.58
77.49
83.16
83.60
85.48
95.04
88.75
83.86
90.59
85.27
97.04
92.11
100.48
98.43
COMPOS IT
35.0 35.3463 5690.75 4624.0
4939.5
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
L H
A E
B T
0 H
B To
S D D
121 1 WOHL
122 1 WOHL
123
124
125
126
127
128
129
130
131
132
133
134
135
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
136 1 WOHL
137 1 WOHL
138 T WOHL
139 1 WOHL
140 1 WOHL
141 1 WOHL
142 1 WOHL
143 1 WOHL
144 1 WOHL
145 1 WOHL
146 1 WOHL
147 1 WOHL
148 1 WOHL
149 1 WOHL
150 1 WOHL
C
0
M
P
0
S
1
T
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
m
m
_
m
1
1
2
2
1
1
2
2
.
a
.
.
.
_
m
2
1
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-HB
QC-WB
QC-WB
QC-WB
LAB_1U=I 1
B
A
R
C
0
D
E
906028
906029
906103
906104
906178
906179
906256
906257
906358
906359
906360
906361
906061
906062
906136
906137
906211
906212
906289
906290
906424
906425
906426
906427
905723
905724
905719
906082
906157
906316
it i nuu=wi
T
0
T
S
R
M
.0
.0
.0
.0
.0
.0
.0
.0
.0
1.0
1.0
1.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
.
a
m
JHL Lunruai
T
0
T
L
A
B
1.0092
1.0133
1.0107
1.0131
1.0158
1.0122
1.0136
1.0109
1.0697
1.0157
1.0113
1.0079
4.0101
4.0342
4.0213
4.0104
4.0175
4.0258
4.0286
4.0222
4.0700
4.1911
4.0937
4.0301
1.0008
4.0330
.
B
m
=£-wire
T
0
T
P
B
162.48
163.14
162.72
163.11
163.54
162.96
163.19
162.75
172.22
163.53
162.82
162.27
645.63
649.51
647.43
645.67
646.82
648.15
648.60
647.57
655.27
674.77
659.09
648.85
161.13
649.31
.
f
m
I
C
P
122.0
119.0
130.0
127.0
120.0
145.0
127.0
131.0
144.0
141.0
141.0
132.0
509.0
522.0
547.0
538.0
548.0
540.0
501.0
505.0
555.0
589.0
542.0
546.0
144.0
612.0
.
e
n
I
C
p
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
F
A
A
136.0
138.0
150.0
146.0
134.0
166.0
144.0
144.0
158.0
154.0
152.0
148.0
556.0
572.0
592.0
580.0
594.0
584.0
538.0
550.0
588.0
618.0
592.0
592.0
162.0
655.0
.
.
B
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
I
C
P
R
E
C
75.09
72.94
79.89
77.86
73.37
88.98
77.82
80.49
83.61
86.22
86.60
81.34
78.84
80.37
84.49
83.32
84.72
83.31
77.24
77.98
84.70
87.29
82.24
84.15
89.37
94.25
.
,
a
F
A
A
R
E
C
83.70
84.59
92.18
89.51
81.94
101.86
88.24
88.48
91.74
94.17
93.36
91.20
86.12
88.07
91.44
89.83
91.83
90.10
82.95
84.93
89.73
91.59
89.82
91.24
100.54
100.88
.
.
B
COMPOSIT
65.0 65.7922 10592.54 8777.0
9543.0
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
COMPOS IT
METHOD
LAB ID
L M
\ E
B T
H
r o
5 D
1 WOHL
1 WOHL
1 WOHL
1 WOHL
1 WOHL
t WOHL
1 WOHL
1 WOHL
1 WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
C
0
M
P
0
S
I
T
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
.
B
m
m
1
1
2
2
1
1
2
2
.
.
B
.
m
f
.
1
2
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-MB
QC-WB
QC-WB
QC-WB
LHD_IU= 1
B
C
0
D
E
906037
906038
906112
906113
906187
906188
906265
906266
906376
906377
906378
906379
906070
906071
906145
906146
906223
906224
906295
906296
906442
906443
906444
906445
905725
905726
905720
906010
906235
906322
nc i nuLf-i
T
0
T
S
R
M
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
8.0
.
.
^
130.0
230.0
455.0
*unL Lunrua
T
0
T
L
A
B
2.0209
2.0138
2.0188
2.0287
2.0310
2.0257
2.0192
2.0435
2.0818
2.0328
2.0598
2.0362
8.0217
8.0433
8.0253
8.0455
8.0515
8.0435
8.0716
8.0296
8.1496
8.1349
8.1482
7.8996
2.0179
8.0165
.
.
B
131.1109
232.2494
459.1111
i-H-wirc -
T
0
T
P
B
325.36
324.22
325.03
326.62
326.99
326.14
325.09
329.00
335.17
327.28
331.63
327.83
1291.49
1294.97
1292.07
1295.33
1296.29
1295.00
1299.53
1292.77
1312.09
1309.72
1311.86
1271.84
324.88
1290.66
.
_
B
21108.85
37392.15
73916.89
I
C
P
214.0
210.0
207.0
204.0
216.0
211.0
211.0
208.0
247.0
236.0
239.0
229.0
879.0
827.0
877.0
858.0
733.0
540.0
670.0
742.0
868.0
940.0
981.0
962.0
273.0
1160.0
_
2.2
4.7
13949
27350
60507
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
Q
Q
F
A
A
243.0
243.0
247.0
239.0
249.0
249.0
247.0
243.0
281.0
275.0
273.0
261.0
978.0
922.0
966.0
958.0
809.0
598.0
740.0
813.0
966.0
1020.0
1090.0
1080.0
311.0
1290.0
.
.
m
15591
30074
65981
f
A
A
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
ND
ND
ND
ND
I
C
P
R
E
C
65.77
64.77
63.69
62.46
66.06
64.70
64.90
63.22
73.69
72.11
72.07
69.85
68.06
63.86
67.88
66.24
56.55
41.70
51.56
57.40
66.15
71.77
74.78
75.64
84.03
89.88
.
.
m
f
A
A
R
E
C
74.69
74.95
75.99
73.17
76.15
76.35
75.98
73.86
83.84
84.03
82.32
79.61
75.73
71.20
74.76
73.96
62.41
46.18
56.94
62.89
73.62
77.88
83.09
84.92
95.73
99.95
.
.
m
ffi
-------
oo
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
L
A
B
T
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
M
E
T
H
0
D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
M
P
0
S
I
T
-WIPE
-UIPE
-WIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
1-UIPE
B
R
A
N
D
U
A
A
B
B
C
C
A
A
B
B
C
C
QC
QC
QC
A
B
C
L
0
T
1
2
1
2
i
2
1
2
_
_
i
2
LA
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
OC-LCS
QC-MB
QC-WB
OC-UB
QC-WB
B_ID=
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
L
A
B
T
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
M
E
T
H
0
D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
305 OA
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
M
P
0
S
I
T
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-UIPE
2-UIPE
2-UIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
OC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
f
f
m
1
1
2
2
1
1
2
2
.
.
.
B
m
m
2
1
LI
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-HB
QC-WB
QC-WB
QC-WB
\U_lD=d Mt
B
A
R
C
0
D
E
906024
906025
906099
906100
906174
906175
906252
906253
906350
906351
906352
906353
906057
906058
906132
906133
906207
906208
906285
906286
906416
906417
906418
906419
905733
905734
905728
906080
906155
906314
THOO=iU5
T
0
T
S
R
M
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
_
_
JA COMPOS IT
T
0
T
L
A
B
.0098
.0035
.0103
.0080
.0185
.0065
1.0176
1.0172
1.0097
0.9844
1.0049
1.0160
4.0116
4.0004
4.0246
4.0202
4.0216
4.0174
4.0337
4.0138
4.0263
4.0460
4.0393
4.0212
1.0025
4.0050
m
f
=
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
0
B
S
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
COMPOS IT
METHOD
L
A
B
T
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
M
E
T
H
0
D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
M
P
0
S
1
T
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
OC
QC
OC
A
A
B
B
C
C
L
0
T
1
1
2
2
1
1
2
2
B
m
f
.
1
1
2
2
1
1
2
2
a
B
B
_
m
B
B
B
B
.
1
2
1
2
.
Lfl
S
P
L
E
T
Y
P
E
LOAD
LOAD .
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-LCS
QC-LCS
QC-MB
QC-MB
QC-WB
QC-WB
QC-WB
QC-WB
QC-WB
QC-WB
E
VJ
II m < a:
3
«'
C
0
D
E
906035
906044
906110
906119
906185
906194
906263
906272
906372
906373
906390
906391
906068
906077
906143
906152
906219
906220
906303
906304
906438
906439
906456
906457
905735
905736
905737
905738
905729
905730
906008
906086
906161
906239
906320
906326
:inuu=ju:
T
0
T
S
R
M
2.0
2.0
2.0
2.0
- 2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
2.0
8.0
8.0
f
.
.
.
B
.
.
140.0
222.5
JUA Lunrubi
T
0
T
L
A
B
2.0116
2.0183
2.0205
2.0546
2.0295
2.0357
2.0372
2.0475
2.0472
2.0297
2.0391
2.0347
8.0384
8.0199
8.0277
8.0313
8.0425
8.0591
8.0280
8.0571
8.0614
8.0372
8.0829
8.1545
2.0003
2.0068
8.0008
8.0077
B
.
.
.
m
m
.
141.0612
224.1436
I="»-W1KC --
T
0
T
P
B
323.87
324.95
325.30
330.79
326.75
327.75
327.99
329.65
329.60
326.78
328.30
327.59
1294.18
1291.20
1292.46
1293.04
1294.84
1297.52
1292.51
1297.19
1297.89
1293.99
1301.35
1312.87
322.05
323.09
1288.13
1289.24
f
f
.
.
m
.
.
22710.85
36087.12
I
C
P
330.0
335.0
340.0
345.0
330.0
335.0
340.0
350.0
370.0
350.0
350.0
345.0
1360.0
1440.0
1375.0
1380.0
1370.0
1325.0
1365.0
1365.0
1380.0
1385.0
1405.0
1425.0
340.0
340.0
1370.0
1370.0
20.0
20.0
20.0
20.0
15.0
20.0
20.0
20.0
24270
38072
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
F
A
A
315.0
295.0
335.0
335.0
310.0
345.0
300.0
345.0
350.0
325.0
365.0
355.0
1275.0
1385.0
1330.0
1305.0
1320.0
1260.0
1290.0
1315.0
1275.0
1295.0
1360.0
1360.0
295.0
320.0
1265.0
1285.0
5.0
m
15.0
15.0
.
.
10.0
22955
36512
F
A
A
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
Q
ND
Q
ND
ND
Q
I
C
P
R
E
C
101.89
103.09
104.52
104.30
100.99
102.21
103.66
106.17
112.26
107.11
106.61
105.32
105.09
111.52
106.39
106.73
105.80
102.12
105.61
105.23
106.33
107.03
107.97
108.54
105.57
105.23
106.36
106.26
B
.
m
f
n
m
.
F
A
A
R
E
C
97.26
90.78
102.98
101.27
94.87
105.26
91.47
104.66
106.19
99.45
111.18
108.37
98.52
107.26
102.90
100.93
101.94
97.11
99.81
101.37
98.24
100.08
104.51
103.59
91.60
99.04
98.20
99.67
B
.
.
.
m
f
.
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
10
0
B
S
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
L
A
B
T
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
H
E
T
H
0
D
WOHL
WOHL
UOHL
WOHL
WOHL
UOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
C
0
M
P
0
S
I
T
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
1-WIPE
1-WIPE
1-WIPE
1-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
m
f
f
m
1
1
2
2
1
1
2
2
m
t
m
f
_
B
t
1
2
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-MB
QC-WB
QC-WB
QC-WB
VI
II
3
m «*:
fi
<
-1
C
0
D
E
906018
906019
906093
906094
906168
906169
906246
906247
906338
906339
906340
906341
906051
906052
906126
906127
906201
906202
906279
906280
906404
906405
906406
906407
905742
905743
905739
906005
906233
906311
it i nuu=w
T
0
T
S
R
H
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
0.5
2.0
t
f
.
JHL Lunruai
T
0
T
L
A
B
0.5007
0.5026
0.5027
0.5059
0.5061
0.5109
0.5062
0.5057
0.5289
0.5192
0.5210
0.5088
2.0024
2.0103
2.0097
2.0156
2.0085
2.0090
1.9756
2.0132
2.0409
2.0413
2.0260
2.0093
0.5001
2.0005
.
.
.
=i-wirt -
T
0
T
P
B
80.61
80.92
80.93
81.45
81.48
82.25
81.50
81.42
85.15
83.59
83.88
81.92
322.39
323.66
323.56
324.51
323.37
323.45
318.07
324.13
328.58
328.65
326.19
323.50
80.52
322.08
.
_
1
C
P
77.0
82.0
78.0
84.0
81.0
84.0
82.0
82.0
89.0
84.0
90.0
90.0
289.0
276.0
290.0
281.0
303.0
294.0
268.0
298.0
295.0
309.0
305.0
304.0
88.0
308.0
2.0
1.0
1.0
2.0
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
F
A
A
70.0
73.0
73.0
74.0
77.0
80.0
75.0
74.0
77.0
75.0
78.0
80.0
287.0
281.0
292.0
287.0
307.0
295.0
268.0
297.0
295.0
314.0
303.0
303.0
75.0
311.0
_
_
2.0
2.0
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
Q
Q
1
C
P
R
E
C
95.52
101.34
96.37
103.13
99.41
102.12
100.62
100.72
104.52
100.49
107.29
109.87
89.64
85.28
89.63
86.59
93.70
90.90
84.26
91.94
89.78
94.02
93.50
93.97
109.29
95.63
.
m
_
F
A
A
R
E
C
86.83
90.21
90.20
90.85
94.50
97.26
92.03
90.89
90.43
89.72
92.99
97.66
89.02
86.82
90.25
88.44
94.94
91.20
84.26
91.63
89.78
95.54
92.89
93.66
93.15
96.56
.
m
f
COMPOSIT
32.5
32.7811
5277.76 4917.0
4825.0
-------
10
LABORATORY DATA AND LEAD RECOVERY RESULTS
11
0
B
S
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
L
A
B
I
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
H
E
T
H
0
D
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
C
0
M
P
0
S
1
T
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
.
.
.
m
1
1
2
2
1
1
2
2
.
.
.
.
m
f
2
1
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-HB
QC-UB
QC-WB
QC-WB
LAB_ID=i
B
A
R
C
0
D
E
906030
906031
906105
906106
906180
906181
906258
906259
906362
906363
906364
906365
906063
906064
906138
906139
906213
906214
906291
906292
906428
906429
906430
906431
905744
905745
905740
906083
906158
906317
HETHOD=W
T
0
T
S
R
M
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
.
_
_
DHL COMPOSIT=Z-WIPE
T
0 T
T 0
L
A
B
1.0000
1.0042
.0038
.0135
.0209
.0072
.0091
.0118
.0507
.0190
.0142
.0172
4.0091
4.0189
4.0118
4.0153
4.0131
4.0317
4.0265
4.0227
4.0208
4.0334
4.0067
4.0165
1.0006
4.0010
.
f
_
T
P
B
161.00
161.68
161.61
163.17
164.36
162.16
162.47
162.90
169.16
164.06
163.29
163.77
645.47
647.04
645.90
646.46
646.11
649.10
648.27
647.65
647.35
649.38
645.08
646.66
161.10
644.16
t
I
C
P
139.0
138.0
149.0
138.0
134.0
152.0
144.0
141.0
159.0
151.0
152.0
150.0
551.0
597.0
576.0
578.0
576.0
591.0
542.0
596.0
631.0
613.0
588.0
597.0
178.0
625.0
3.0
i!o
1.0
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
Q
Q
F
A
A
159.0
159.0
173.0
165.0
159.0
181.0
165.0
175.0
184.0
177.0
178.0
181.0
563.0
592.0
584.0
583.0
598.0
604.0
560.0
618.0
632.0
619.0
602.0
599.0
147.0
604.0
3.0
3.0
2.0
4.0
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
1
C
P
R
E
C
86.34
85.36
92.20
84.57
81.53
93.74
88.63
86.56
93.99
92.04
93.09
91.59
85.36
92.27
89.18
89.41
89.15
91.05
83.61
92.02
97.47
94.40
91.15
92.32
110.49
97.03
F
A
A
R
E
C
98.76
98.34
107.05
101.12
96.74
111.62
101.56
107.43
108.77
107.89
109.01
110.52
87.22
91.49
90.42
90.18
92.55
93.05
86.38
95.42
97.63
95.32
93.32
92.63
91.25
93.77
COMPOSIT
65.0 65.3997 10529.35 9591.0
9973.0
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
12
0
B
S
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
COMPOS IT
METHOD
LAB ID
L
A
B
T
D
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
M
E
T
H
0
D
UOHL
UOHL
UOHL
WOHL
WOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
C
0
M
P
0
S
I
T
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
B
R
A
N
D
U
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
.
i
1
2
2
1
1
2
2
.
_
t
.
i
2
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-HB
QC-UB
QC-WB
QC-UB
LAB_ID=i:
B
A
R
C
0
D
E
906039
906040
906114
906115
906189
906190
906267
906268
906380
906381
906382
906383
906072
906073
906147
906148
906225
906226
906297
906298
906446
906447
906448
906449
905746
905747
905741
906011
906236
906323
ntiHou=Vi
T
0
T
S
R
M
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
8.0
.
_
.
130.0
227.5
450.0
IUHL tUHHOS]
T
0
T
L
A
B
2.0156
2.0196
2.0195
2.0207
2.0287
2.0325
2.0264
2.0379
1.9903
2.0405
2.0341
2.0435
8.0188
8.0131
8.0598
8.0361
8.0392
8.0870
8.0363
8.1291
8.0727
8.2092
8.0661
8.0031
2.0030
8.0051
f
.
m
131.0879
229.2687
453.4123
,\=<*-Hlrt -
T
0
T
P
B
324.51
325.16
325.14
325.33
326.62
327.23
326.25
328.10
320.44
328.52
327.49
329.00
1291.03
1290.11
1297.63
1293.81
1294.31
1302.01
1293.84
1308.79
1299.70
1321.68
1298.64
1288.50
322.48
1288.82
m
m
f
21105.15
36912.26
72999.38
I
C
P
324.0
322.0
332.0
319.0
341.0
338.0
330.0
339.0
355.0
350.0
349.0
344.0
942.0
967.0
946.0
946.0
988.0
990.0
993.0
972.0
1009.0
1001.0
1001.0
994.0
350.0
1020.0
3.0
2.0
2.0
9.0
17178
31686
69758
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
F
A
A
292.0
292.0
313.0
283.0
302.0
300.0
290.0
299.0
307.0
307.0
309.0
309.0
1186.0
1225.0
1196.0
1200.0
1151.0
1158.0
1137.0
1105.0
1220.0
1201.0
1182.0
1198.0
299.0
1215.0
.
2.0
1.0
4.0
19283
34081
70593
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
Q
Q
Q
I
C
P
R
E
C
99.84
99.03
102.11
98.05
104.40
103.29
101.15
103.32
110.79
106.54
106.57
104.56
72.97
74.95
72.90
73.12
76.33
76.04
76.75
74.27
77.63
75.74
77.08
77.14
108.53
79.14
,
.
m
F
A
A
R
E
C
89.98
89.80
96.27
86.99
92.46
91.68
88.89
91.13
95.81
93.45
94.35
93.92
91.86
94.95
92.17
92.75
88.93
88.94
87.88
84.43
93.87
90.87
91.02
92.98
92.72
94.27
.
.
.
U>
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
13
0
B
S
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
COMPOS IT
L
A
B
I
D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
C
0
H H
E P
T 0
H S
0 I
D T
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
-WIPE
3050A 1-WIPE
3050A 1-WIPE
3050A 1-WIPE
3050A 1-WIPE
3050A 1-WIPE
3050A 1-WIPE
3050A 1-WIPE
B
R
A
N
D
W
A
A
B
B
C
C
A
A
B
B
C
C
QC
QC
QC
A
B
C
L
0
T
1
2
1
2
,
_
i
2
1
2
.
.
.
.
1
2
L
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
OC-LCS
QC-LCS
QC-MB
QC-WB
QC-W8
QC-WB
AB_IU=i PI
B
A
R
C
0
D
E
906015
906090
906165
906243
906332
906333
906048
906123
906198
906276
906398
906399
905752
905753
905748
906003
906231
906309
t 1 HUU=iU
T
0
T
S
R
M
0.5
0.5
0.5
0.5
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
0.5
2.0
t
m
t
17.5
3u« Lunrusi
T
0
T
L
A
B
0.5014
0.5063
0.5073
0.5012
0.4983
0.5042
2.0074
2.0068
2.0070
2.0245
2.0188
2.0299
0.5000
2.0000
m
m
B
17.6131
I=|-HIKC -
T
0
T
P
B
80.73
81.51
81.68
80.69
80.23
81.18
323.19
323.09
323.13
325.94
325.03
326.81
80.50
322.00
m
m
B
2835.71
I
C
P
72.0
79.0
75.0
73.0
73.0
83.0
280.0
270.0
270.0
260.0
280.0
260.0
74.0
280.0
m
f
m
2429.0
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
NO
ND
ND
ND
f
A
A
83.0
88.0
82.0
82.0
85.0
99.0
320.0
320.0
320.0
320.0
330.0
300.0
83.0
320.0
m
.
.
5.0
2837.0
f
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
ND
ND
ND
Q
I
C
P
R
E
C
89.19
96.92
91.83
90.47
90.99
102.25
86.64
83.57
83.56
79.77
86.15
79.56
91.93
86.96
f
.
f
f
A
A
R
E
C
102.82
107.96
100.40
101.62
105.95
121.96
99.01
99.04
99.03
98.18
101.53
91.80
103.11
99.38
m
.
.
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
14
0
B
S
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
L
A
B
T
D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
H
E
T
H
0
D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
M
P
0
S
I
T
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-WIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
2-UIPE
B
R
A
N
D
U
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
i
1
2
2
1
1
2
2
.
2
1
LAB_ID=3 METHOD=305
S
P B T
LAO
E R T
T C
YDS
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-MB
QC-UB
QC-WB
QC-UB
D
E
906026
906027
906101
906102
906176
906177
906254
906255
906354
906355
906356
906357
906059
906060
906134
906135
906209
906210
906287
906288
906420
906421
906422
906423
905754
905755
905749
906081
906156
906315
R
M
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
.
.
OA COMPOSITE-WIPE -
T
0 T
T 0
T
L
A
B
.0104
.0145
.0101
.0097
0.9974
.0159
.0153
.0179
.0145
.0288
1.0286
1.0017
4.0232
4.0173
4.0271
4.0234
4.0923
4.0295
4.0176
4.0741
4.0484
4.1514
4.0267
4.2803
1.0000
4.0000
.
.
.
P
B
162.67
163.33
162.63
162.56
160.58
163.56
163.46
163.88
163.33
165.64
165.60
161.27
647.74
646.79
648.36
647.77
658.86
648.75
646.83
655.93
651.79
668.38
648.30
689.13
161.00
644.00
.
.
.
I
C
p
140.0
150.0
150.0
150.0
140.0
140.0
150.0
150.0
150.0
150.0
140.0
150.0
560.0
530.0
550.0
550.0
540.0
550.0
550.0
520.0
540.0
590.0
520.0
540.0
150.0
550.0
.
.
.
I
C
P
Q
0
Q
0
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
ND
ND
ND
ND
F
A
A
170.0
170.0
180.0
170.0
160.0
170.0
170.0
170.0
180.0
180.0
160.0
200.0
660.0
640.0
680.0
690.0
670.0
660.0
660.0
660.0
660.0
690.0
660.0
670.0
170.0
660.0
.
.
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
I
C
P
R
E
C
86.06
91.84
92.24
92.27
87.18
85.60
91.76
91.53
91.84
90.56
84.54
93.01
86.46
81.94
84.83
84.91
81.96
84.78
85.03
79.28
82.85
88.27
80.21
78.36
93.17
85.40
.
.
F
A
A
R
E
C
104.50
104.08
110.68
104.58
99.64
103.94
104.00
103.73
110.20
108.67
96.62
124.01
101.89
98.95
104.88
106.52
101.69
101.73
102.04
100.62
101.26
103.24
101.80
97.22
105.59
102.48
.
.
COMPOSIT
65.0 65.9761 10622.15 9000.0
10910
-------
LABORATORY DATA AJID LEAD RECOVERY RESULTS
15
0
B
S
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
V38
COMPOS IT
METHOD
L
A
B
I
D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
H
E
T
H
0
D
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
3050A
C
0
M
P
0
S
I
T
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
4-UIPE
B
R
A
N
D
U
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
QC
QC
QC
A
A
B
B
C
C
L
0
T
1
1
2
2
1
1
2
2
_
f
.
i
1
2
2
1
1
2
2
_
_
_
.
.
<
1
2
1
2
.
L
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-LCS
QC-LCS
QC-MB
QC-MB
QC-UB
QC-UB
QC-UB
QC-UB
QC-UB
QC-UB
AB_ID=i f
B
A
R
C
0
D
E
906036
906045
9061 1 1
906120
906186
906195
906264
906273
906374
906375
906392
906393
906069
906078
906144
906153
906221
906222
906305
906306
906440
906441
906458
906459
905756
905757
905758
905759
905750
905751
906009
906087
906162
906240
906321
906327
ltlHUU=Jl
T
0
T
S
R
M
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
2.0
8.0
8.0
B
m
f
B
m
B
m
140.0
222.5
DUA LUMKUSI
T
0
T
L
A
B
2.0250
2.0077
2.0288
2.0296
2.0361
2.0486
2.0935
2.0349
2.0446
2.0400
2.0311
2.0564
8.0300
8.0208
8.1043
8.0317
8.0213
8.0303
8.0259
8.0255
8.1479
8.0892
8.0416
8.0753
2.0000
2.0000
8.0000
8.0000
.
m
m
t
.
141.1201
224.7093
l=<»-WIKt -
T
0
T
P
B
326.03
323.24
326.64
326.77
327.81
329.82
337.05
327.62
329.18
328.44
327.01
331.08
1292.83
1291.35
1304.79
1293.10
1291.43
1292.88
1292.17
1292.11
1311.81
1302.36
1294.70
1300.12
322.00
322.00
1288.00
1288.00
.
.
.
.
.
.
.
22720.34
36178.20
I
C
P
290.0
290.0
290.0
310.0
300.0
300.0
300.0
290.0
310.0
300.0
320.0
320.0
1100.0
1100.0
1100.0
1100.0
1100.0
1100.0
1100.0
1000.0
1100.0
1200.0
1100.0
1100.0
290.0
300.0
1100.0
1100.0
.
.
.
.
.
.
19610
31039
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
ND
ND
ND
ND
ND
ND
ND
ND
F
A
A
330.0
340.0
360.0
330.0
350.0
320.0
360.0
330.0
350.0
360.0
350.0
340.0
1300.0
1300.0
1300.0
1200.0
1300.0
1300.0
1200.0
1200.0
1300.0
1400.0
1200.0
1200.0
290.0
320.0
1300.0
1300.0
.
.
.
.
30.0
.
30.0
22590
36337
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
Q
ND
Q
ND
I
C
P
R
E
C
88.95
89.72
88.78
94.87
91.52
90.96
89.01
88.52
94.17
91.34
97.86
96.65
85.08
85.18
84.30
85.07
85.18
85.08
85.13
77.39
83.85
92.14
84.96
84.61
90.06
93.17
85.40
85.40
.
.
.
.
.
.
.
r
A
A
R
E
C
101.22
105.19
110.21
100.99
106.77
97.02
106.81
100.73
106.32
109.61
107.03
102.69
100.55
100.67
99.63
92.80
100.66
100.55
92.87
92.87
99.10
107.50
92.69
92.30
90.06
99.38
100.93
100.93
.
.
.
.
.
.
.
-
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
16
0
B
S
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
L
A
B
I
D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
C
0
M H
E P
T 0
H S
0 I
D T
UOHL 1-UIPE
WOHL 1-WIPE
UOHL 1-UIPE
UOHL 1-UIPE
UOHL 1-UIPE
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
UOHL
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
-UIPE
B
R
A
H
D
U
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
_
i
1
2
2
1
1
2
2
f
_
_
.
f
i
2
S
p
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
OC-LCS
QC-LCS
QC-MB
QC-WB
QC-UB
QC-WB
LAB_IU=i r
B
A
R
C
0
D
E
906020
906021
906095
906096
906170
906171
906248
906249
906342
906343
906344
906345
906053
906054
906128
906129
906203
906204
906281
906282
906408
906409
906410
906411
905763
905764
905760
906006
906234
906312
it i nuu=wt
T
0
T
S
R
M
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
0.5
2.0
f
m
f
JML Lunruai
T
0
T
L
A
B
0.5038
0.5025
0.5099
0.5094
0.5061
0.5102
0.5071
0.5092
0.5222
0.5078
0.5067
0.5089
2.0002
2.0056
2.0226
2.0148
2.0151
2.0058
2.0233
2.0069
1.9862
2.0054
2.0200
2.0107
0.5000
2.0000
f
,
m
=|-HIKC -
T
0
T
P
B
81.11
80.90
82.09
82.01
81.48
82.14
81.64
81.98
84.07
81.76
81.58
81.93
322.03
322.90
325.64
324.38
324.43
322.93
325.75
323.11
319.78
322.87
325.22
323.72
80.50
322.00
m
m
.
I
C
P
64.0
66.0
65.0
64.0
67.0
68.0
69.0
65.0
76.0
70.0
71.0
69.0
240.0
250.0
250.0
250.0
230.0
240.0
260.0
240.0
240.0
260.0
260.0
260.0
72.0
270.0
,
.
.
I
C
P
Q
Q
Q
Q
Q
0
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
F
A
A
72.0
78.0
73.0
72.0
74.0
76.0
78.0
73.0
84.0
81.0
79.0
78.0
280.0
330.0
320.0
330.0
280.0
290.0
310.0
310.0
290.0
320.0
330.0
320.0
82.0
330.0
,
.
.
F
A
A
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
0
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
0
0
ND
ND
ND
ND
I
C
P
R
E
C
78.90
81.58
79.18
78.04
82.23
82.78
84.51
79.29
90.40
85.62
87.03
84.22
74.53
77.42
76.77
77.07
70.89
74.32
79.82
74.28
75.05
80.53
79.95
80.32
89.44
83.85
.
.
.
F
A
A
R
E
C
88.77
96.41
88.92
87.79
90.82
92.52
95.54
89.04
99.91
99.08
96.84
95.20
86.95
102.20
98.27
101.73
86.30
89.80
95.16
95.94
90.69
99.11
101.47
98.85
101.86
102.48
.
.
.
'
COMPOSIT
32.5
32.7204
5267.98 4136.0
5040.0
ffi
-------
ffi
I>
oo
LABORATORY DATA AND LEAD RECOVERY RESULTS
17
0
B
S
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
L
A
B
I
D
3
3
3
3
3 -
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
H
E
T
H
0
D
WOHL
WOHL
WOHL
UOHL
UOHL
UOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
C
0
M
P
0
S
I
T
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
2-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
B
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
i
1
2
2
^
1
2
2
\
2
1
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-MB
QC-WB
QC-WB
QC-WB
LAB ID=3 METHOD=WOHL COMPOSITE-WIPE --
B T T
A 0 0 T
R T T 0
C T
0 S L
D
E
906032
906033
906107
906108
906182
906183
906260
906261
906366
906367
906368
906369
906065
906066
906140
906141
906215
906216
906293
906294
906432
906433
906434
906435
905765
905766
905761
906084
906159
906318
R
M
1.0
1.0
1.0
1.0
1.0
.0
.0
.0
.0
.0
.0
.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
1.0
4.0
.
.
.
A
B
.0072
.0073
.0053
.0112
.0112
.0179
.0009
.0017
.0204
.0062
.0150
.0777
4.0122
4.0474
4.0191
4.0204
4.0233
4.0157
4.0323
4.0250
4.0524
4.0415
4.0261
4.0749
1.0000
4.0000
.
.
.
P
B
162.16
162.18
161.85
162.80
162.80
163.88
161.14
161.27
164.28
162.00
163.42
173.51
645.96
651.63
647.08
647.28
647.75
646.53
649.20
648.03
652.44
650.68
648.20
656.06
161.00
644.00
.
.
.
I
C
P
120.0
120.0
110.0
110.0
120.0
120.0
130.0
130.0
140.0
140.0
130.0
150.0
450.0
440.0
460.0
470.0
460.0
460.0
460.0
490.0
480.0
490.0
460.0
490.0
130.0
510.0
.
.
.
I
C
P
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
F
A
A
140.0
130.0
130.0
130.0
140.0
140.0
140.0
140.0
150.0
160.0
150.0
170.0
600.0
560.0
590.0
600.0
570.0
600.0
570.0
620.0
600.0
600.0
600.0
610.0
150.0
650.0
.
.
.
F
A
A
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
ND
I
C
P
R
E
C
74.00
73.99
67.96
67.57
73.71
73.22
80.67
80.61
85.22
86.42
79.55
86.45
69.66
67.52
71.09
72.61
71.01
71.15
70.86
75.61
73.57
75.31
70.97
74.69
80.75
79.19
.
F
A
A
R
E
C
86.33
80.16
80.32
79.85
85.99
85.43
86.88
86.81
91.31
98.77
91.79
97.98
92.88
85.94
91.18
92.69
88.00
92.80
87.80
95.68
91.96
92.21
92.56
92.98
93.17
100.93
COMPOSIT
65.0 65.5723 10557.14 7770.0
9640.0
-------
LABORATORY DATA AND LEAD RECOVERY RESULTS
18
0
B
S
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
COMPOS IT
METHOD
LAB ID
L
A
B
I
D
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
M
E
T
H
0
D
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
WOHL
C
0
M
P
0
S
1
T
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
4-WIPE
B
R
A
N
D
W
A
A
A
A
B
B
B
8
C
C
C
C
A
A
A
A
B
B
B
B
C
C
C
C
QC
QC
QC
A
B
C
L
0
T
1
1
2
2
1
1
2
2
f
B
m
m
1
1
2
2
1
1
2
2
m
m
B
f
f
m
f
1
2
S
P
L
E
T
Y
P
E
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
LOAD
QC-LCS
QC-LCS
QC-HB
QC-WB
QC-WB
QC-WB
LHD_IU=O
B
A
R
C
0
D
E
906041
906042
906116
906117
906191
906192
906269
906270
906384
906385
906386
906387
906074
906075
906149
906150
906227
906228
906299
906300
906450
906451
906452
906453
905767
905768
905762
906012
906237
906324
nc I nuu=
T
0
T
S
R
M
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
8.0
B
.
.
130.0
227.5
450.0
1355.0
wunL uunrua
T
0
T
L
A
B
2.0201
2.0202
2.0521
2.0402
2.0299
2.0310
2.0417
2.0243
2.0441
2.0434
2.0337
2.0539
8.0268
8.0114
8.0465
8.0404
8.0425
8.0909
8.0234
8.0295
8.0154
8.1075
8.1224
8.0462
2.0000
8.0000
f
.
.
131.0375
229.3302
454.0395
1366.5629
i i-t-Mirc. -
T
0
T
P
B
325.24
325.25
330.39
328.47
326.81
326.99
328.71
325.91
329.10
328.99
327.43
330.68
1292.31
1289.84
1295.49
1294.50
1294.84
1302.63
1291.77
1292.75
1290.48
1305.31
1307.71
1295.44
322.00
1288.00
f
.
21097.04
36922.16
73100.36
220016.63
I
C
P
210.0
180.0
220.0
220.0
220.0
220.0
200.0
200.0
230.0
270.0
240.0
260.0
670.0
680.0
790.0
700.0
660.0
650.0
710.0
700.0
780.0
750.0
790.0
810.0
280.0
940.0
s
.
.
17.0
12597
24503
55542
185807
I
C
P
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
ND
ND
ND
Q
F
A
A
250.0
220.0
260.0
270.0
250.0
260.0
230.0
240.0
270.0
310.0
270.0
300.0
880.0
940.0
1000.0
960.0
880.0
850.0
940.0
900.0
1000.0
940.0
1000.0
1100.0
330.0
1300.0
B
.
.
19.0
16169
30849
67186
203760
f
A
A
Q
Q
Q
Q
0
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
0
Q
ND
ND
ND
Q
I
C
P
R
E
C
64.57
55.34
66.59
66.98
67.32
67.28
60.84
61.37
69.89
82.07
73.30
78.63
51.84
52.72
60.98
54.07
50.97
49.90
54.96
54.15
60.44
57.46
60.41
62.53
86.96
72.98
B
.
.
F
A
A
R
E
C
76.87
67.64
78.70
82.20
76.50
79.51
69.97
73.64
82.04
94.23
82.46
90.72
68.09
72.88
77.19
74.16
67.96
65.25
72.77
69.62
77.49
72.01
76.47
84.91
102.48
100.93
(
.
.
-------
Appendix I
ANOVA Results for Four-way and
Three-way Mixed Models
-------
Organization of Statistical Tables 1-1 Through 1-6
The following tables are reproduced computer outputs obtained when using the PROC
MIXED procedure of SAS.10'11 Each table presents the following information:
1. Class Level Information: Lists the class variables (fixed and random factors) and
their levels included in the model.
2. REML (REstricted Maximum Likelihood) Iteration History: These estimates are
the values that maximize the restricted likelihood function of the data, or alternatively
(as on this output), minimize the function, -2 times the logarithm of this likelihood
function, using a Newton-Raphson algorithm. For example, in Table 1-1,3 iterations
were needed to obtain the estimates. A criterion of zero indicates that the algorithm
converged.
3. Covariance Parameter Estimates (REML): For each random factor, interaction, and
residual (under header "Cov Farm"), this tables provides estimates of variance
components under the heading "Estimate". The column "Ratio" is the ratio of these
variance components over the residual variance component. The "STD Error"
provides approximate standard errors of the variance components, with associated
Z-statistics to test that the variance components equal zero, and the associated p-level.
4 Model Fitting Information for Dependent Variable (e.g., FAA_REC, ICP_REC):
This output provides various information about the fitted mixed model.
5. Tests of Fixed Effects: This table lists all fixed effect factors and their interaction,
numerator and denominator degrees of freedom (NDF and DDF), and p-level
associated with their F-test for significance. A p-level below 0.05 indicates that the
corresponding factor or interaction is significant at the 5% level.
6. Estimate Statement Results: This table lists estimates and associated statistics for
selected contrasts. For example, the estimate for "FAA: 3050A vs WOFIL" of
11.777% (Table 1-1) is the estimated difference of FAA recoveries between the EPA
3050A and the WOHL methods. Its standard error is 2.527% with 2 degrees of
freedom. This difference is statistically significant at the 5% level as indicated by the
p-level of 0.0431. Lower and upper 95% (alpha of 5%) are given in the last two
columns.
7. Least Squares Means: This information is shown in Tables 1-3 through 1-6 and
provides model least squares means for all main fixed effect factors and their interac-
tions. Identical information to that shown in item No. 6 is provided. Parameter
estimates in item No. 6 can be obtained from those in this table.
1-1
-------
Table 1-1. Analysis of Variance Results for the 4-way Mixed Model
for FAA Recovery
FAA: 4-UAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
The MIXED Procedure
Class Level Information
Class Levels Values
LAB ID 3123
METHOD 2 EPA3050A UOHL
SRMLEVEL 2 HIGH LOW
COHPOSIT 3 1-WIPE 2-WIPE 4-UIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
1957.8188777
1726.4612877
1726.4289827
1726.4285906
0.00003545
0.00000045
0.00000000
Convergence criteria met.
Cov Parm
LABJD
LAB_ID*METHOD
LAB ID*SRMLEVEL
LAB~ID*COMPOSIT
LAB~ID*METHOD*SRMLEV
LAB_ID*MET HOD*COMPOS
LAB ID*SRMLEV*COMPOS
LAB~*METH*SRML*COMPO
Residual
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error Z Pr > |Z|
0.54402482 13.84447502 19.55222070 0.71
0.00000000
0.00000000
0.00000000
0.46894918
0.29911154
0.17744576
0.15616808
0.00000000
0.00000000
0.00000000
11.93393183
7.61186263
4.51568263
3.97420288
.17062097
.20050721
4.95242324
4.63559043
1.30
1.23
0.91
0.86
1.00000000 25.44824142 1.89510158 13.43
0.4789
0.1931
0.2196
0.3619
0.3913
0.0001
Model Fitting Information for FAA_REC
Description Value
Observations 396.0000
Variance Estimate 25.4482
Standard Deviation Estimate 5.0446
REML Log Likelihood -1216.09
Akaike's Information Criterion -1225.09
Schwarz's Bayesian Criterion -1242.86
-2 REML Log Likelihood 2432.173
Source
Tests of Fixed Effects
NDF DDF Type III F Pr > F
METHOD 1 2 21.72
SRMLEVEL 1 2 0.45
COMPOSIT 2 4 4.82
METHOD*SRMLEVEL 1 2 0.52
METHOO*COMPOSIT 2 4 4.86
SRMLEVEL*COMPOSIT 2 4 0.00
METHCO*SRMLEV*COMPOS 24 1.51
0.0431
0.5733
0.0860
0.5463
0.0849
0.9994
0.3252
Parameter
FAA: 3050A VS UOHL
FAA: 3050A MEAN
FAA: UOHL MEAN
ESTIMATE Statement Results
Estimate Std Error DDF T Pr > |T| Alpha
11.77720504
99.70114814
87.92394310
2.52705050
2.84223932
2.83532227
4.66
35.08
31.01
0.0431
0.0008
0.0010
Lower
Upper
0.05
0.05
0.05
0.9042 22.6502
87.4720 111.9303
75.7245 100.1234
1-2
-------
Table 1-2. Analysis of Variance Results for the 4-way Mixed Model
for ICP Recovery
ICP: 4-UAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
Class
LABJD
METHOD
SRHLEVEL
COMPOSIT
The MIXED Procedure
Class Level Information
Levels Values
3123
2 EPA3050A UOHL
2 HIGH LOW
3 1-WIPE 2-WIPE 4-WIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
2175.7735891
1627.2329038
1626.9787910
1626.9369787
1626.9361081
Convergence criteria met.
0.00004732
0.00000105
0.00000000
Cov Parm
LAB ID
LAB_ID*METHOO
LAB ID*SRMLEVEL
LAB~ID*COMPOSIT
LAB_ID*METHOD*SRMLEV
LAB_ID*METHOO*COMPOS
LAB_ID*SRMLEV*COMPOS
LAB_*METH*SRML*COMPO
Residual
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error
Z Pr > Izl
4.48046522 86.11838349 89.73612688 0.96 0.3372
.00000000
.00000000
0.07596925
0.32401634
0.13790059
0.00000000
00000000
00000000
46019416
22787195
65056763
0.00000000
5.18508680
6.75406839
6.77970587
0.28
0.92
0.39
0.7782
0.3565
0.6958
0.59765443 11.48743067 6.99246107 1.64 0.1004
1.00000000 19.22085750 1.43178133 13.42 0.0001
Model Fitting Information for ICP_REC
Description Value
Observations 396.0000
Variance Estimate 19.2209
Standard Deviation Estimate 4.3842
REML Log Likelihood -1166.34
Akaike's Information Criterion -1175.34
Schuarz's Bayesian Criterion -1193.12
-2 REML Log Likelihood 2332.681
Source
Tests of Fixed Effects
NDF DDF Type III F Pr > F
METHOD 1 2 50.79 0.0191
SRMLEVEL 1 2 13.00 0.0690
COMPOSIT 2 4 3.14 0.1513
METHOO*SRMLEVEL 1 2 0.06 0.8327
METHOD*COMPOSIT 2 4 11.95 0.0206
SRMLEVEL*COMPOSIT 2 4 2.47 0.1999
METHOO*SRMLEV*COMPOS 2 4 3.44 0.1353
Parameter
ICP: 3050A VS WOHL
ICP: 3050A MEAN
ICP: WOHL MEAN
ESTIMATE Statement Results
Estimate Std Error DDF T Pr > |T| Alpha
14.51673604
94.32092798
79.80419193
2.03690605
5.56396402
5.56129786
7.13
16.95
14.35
0.0191
0.0035
0.0048
Lower
Upper
0.05
0.05
0.05
5.7526 23.2808
70.3811 118.2607
55.8759 103.7325
1-3
-------
Table 1-3. Analysis of Variance Results for the 3-way Mixed Model
for FAA Recovery using EPA Method 3050A
FAA: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=EPA3050A
The MIXED Procedure
Class Level Information
Class Levels Values
LAB ID 3123
SRMLEVEL 2 HIGH LOW
COMPOSIT 3 1-WIPE 2-WIPE 4-WIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
850.08630832
748.21339847
748.02978880
748.01067270
748.01035961
0.00004734
0.00000083
0.00000000
Convergence criteria met.
Cov Parm
LAB ID
LAB~ID*SRMLEVEL
LAB~ID*COMPOSIT
LAB^ID*SRMLEV*COMPOS
Residua I
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error
0.11331120
1 .27797303
0.00000000
0.02519946
1.00000000
2.48280130
28.00211443
0.00000000
0.55215415
21.91135012
22.29316410
28.96455200
1 .42823131
2.43164949
Z Pr > |Z|
0.11
0.97
0.39
9.01
0.9113
0.3337
0.6991
0.0001
Model Fitting Information for FAA_REC
Description Value
Observations 180.0000
Variance Estimate 21.9114
Standard Deviation Estimate 4.6810
REML Log Likelihood -533.900
Akaike's Information Criterion -538.900
Schwarz's Bayesian Criterion -546.798
-2 REML Log Likelihood 1067.801
1-4
-------
Table 1-3. (Continued)
FAA: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=EPA3050A
Source
Tests of Fixed Effects
NDF DDF Type III F Pr > F
Parameter
SRMLEVEL 1 2 0.00 0.9948
COMPOSIT 2 4 1.66 0.2992
SRMLEVEL*COMPOSIT 24 1.43 0.3398
ESTIMATE Statement Results
Estimate Std Error DDF T Pr > |T| Alpha
Lower
Upper
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
SRM HIGH VS LOU -0.
SRM LOW MEAN 99.
SRM HIGH MEAN
1-U VS
1-W VS
2-W VS
1-UIPE
2-WIPE
4-UIPE
2-W
4-W
4-W
MEAN
MEAN
MEAN
99.
-1.
0.
1.
99.
100.
99.
03241493
71735561
68494067
42520511
06386256
48906766
24736729
67257240
18350474
4.39679408
3.23936715
3.23936715
1.04738928
1.04738928
0.89033701
2.48903471
2.42713195
2.42713195
2
2
2
4
4
4
4
4
4
-0.01
30.78
30.77
-1.36
0.06
1.67
39.87
41.48
40.86
0.9948
0.0011
0.0011
0.2452
0.9543
0.1697
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
-18.9503
85.7795
85.7471
-4.3332
-2.8442
-0.9829
92.3367
93.9338
92.4447
18.8855
113.6552
113.6228
1.4828
2.9719
3.9610
106.1580
107.4114
105.9223
Level
SRMLEVEL HIGH
SRMLEVEL LOW
COMPOSIT 1-WIPE
COMPOSIT 2-WIPE
COMPOSIT 4-WIPE
SRMLEVEL*COMPOSIT HIGH 1-WIPE
SRMLEVEL*COMPOSIT HIGH 2-WIPE
SRMLEVEL*COMPOSIT HIGH 4-WIPE
SRMLEVEL*COMPOSIT LOW 1-WIPE
SRMLEVEL*COMPOSIT LOW 2-WIPE
SRMLEVEL*COMPOSIT LOW 4-WIPE
Least Squares Means
LSMEAN Std Error DDF
T Pr > |T| Alpha Lower
Upper
99.68494067
99.71735561
99.24736729
100.67257240
99.18350474
98.30015829
100.75141735
100.00324639
100.19457629
100.59372745
98.36376308
3.23936715
3.23936715
2.48903471
2.42713195
2.42713195
3.40043926
3.30973391
3.30973391
3.40043926
3.30973391
3.30973391
2
2
4
4
4
4
4
4
4
4
4
30.77
30.78
39.87
41.48
40.86
28.91
30.44
30.21
29.47
30.39
29.72
0.0011
0.0011
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
85.7471
85.7795
92.3367
93.9338
92.4447
88.8590
91.5621
90.8140
90.7534
91.4044
89.1745
113.6228
113.6552
106.1580
107.4114
105.9223
107.7413
109.9407
109.1925
109.6357
109.7830
107.5531
1-5
-------
Table 1-4. Analysis of Variance Results for the 3-way Mixed Model
for FAA Recovery using the WOHL Method
FAA: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=UOHL
The MIXED Procedure
Class Level Information
Class Levels Values
LAB ID 3123
SRMLEVEL 2 HIGH LOW
COMPOSIT 3 1-WIPE 2-WIPE 4-WIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
1101.2677265
966.05875487
966.04641441
966.04625725
0.00002386
0.00000032
0.00000000
Convergence criteria met.
Cov Parm
LAB ID
LAB~ID*SRMLEVEL
LAB~ID*COMPOS1T
LAB~ID*SRMLEV*COMPOS
Residual
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error
0.73515523
0.00000000
0.57398455
0.46531735
1.00000000
20.86376901
0.00000000
16.28973121
13.20574648
28.38008639
29.44089379
17.60699292
8.99200342
2.85230600
Z Pr > Z
0.71
0.93
1.47
9.95
0.4785
0.3549
0.1419
0.0001
Model Fitting Information for FAA_REC
Description Value
Observations 216.0000
Variance Estimate 28.3801
Standard Deviation Estimate 5.3273
REML Log Likelihood -676.000
Akaike's Information Criterion -681.000
Schwarz's Bayesian Criterion -689.368
-2 REML Log Likelihood 1352.000
1-6
-------
Table 1-4. (Continued)
FAA: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=UOHL
Source
Tests of Fixed Effects
NDF DDF Type III F Pr > F
Parameter
SRMLEVEL 1 2 2.87 0.2322
COMPOSIT 2 4 5.12 0.0789
SRMLEVEL*COMPOSIT 2 4 0.34 0.7287
ESTIMATE Statement Results
Estimate Std Error DDF T Pr > |T| Alpha
Lower
Upper
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
FAA:
SRM HIGH VS LOW -3
SRM LOW MEAN 89
SRM HIGH MEAN
1-U VS
1-W VS
2-W VS
1-WIPE
2-WIPE
4-WIPE
2-W
4-W
4-W
MEAN
MEAN
MEAN
86
-1
10
11
90
92
80
.15278030
.50033325
.34755295
.86546788
.05395272
.91942059
.65343804
.51890592
.59948533
1.86015254
3.23954373
3.23954373
4.00625412
4.00625412
4.00625412
3.87035214
3.87035214
3.87035214
2
2
2
4
4
4
4
4
4
-1.69
27.63
26.65
-0.47
2.51
2.98
23.42
23.90
20.82
0.2322
0.0013
0.0014
0.6657
0.0661
0.0409
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
-11.1564
75.5617
72.4089
-12.9886
-1.0692
0.7963
79.9076
81.7731
69.8537
4.8508
103.4390
100.2862
9.2577
21.1771
23.0426
101.3993
103.2647
91.3453
Level
SRMLEVEL HIGH
SRMLEVEL LOW
COMPOSIT 1-WIPE
COMPOSIT 2-WIPE
COMPOSIT 4-WIPE
SRMLEVEL*COMPOSIT HIGH 1-WIPE
SRMLEVEL*COMPOSIT HIGH 2-WIPE
SRMLEVEL*COMPOSIT HIGH 4-WIPE
SRMLEVEL*COMPOSIT LOW 1-WIPE
SRMLEVEL*COMPOSIT LOW 2-WIPE
SRMLEVEL*COMPOSIT LOW 4-WIPE
Least Squares Means
LSMEAN Std Error DDF
T Pr > |T| Alpha Lower
Upper
86.34755295
89.50033325
90.65343804
92.51890592
80.59948533
90.07055275
90.83250605
78.13960005
91.23632334
94.20530579
83.05937061
3.23954373
3.23954373
3.87035214
3.87035214
3.87035214
4.19222510
4.19222510
4.19222510
4.19222510
4.19222510
4.19222510
2
2
4
4
4
4
4
4
4
4
4
26.65
27.63
23.42
23.90
20.82
21.49
21.67
18.64
21.76
22.47
19.81
0.0014
0.0013
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
72.4089
75.5617
79.9076
81.7731
69.8537
78.4311
79.1930
66.5001
79.5968
82.5658
71.4199
100.2862
103.4390
101.3993
103.2647
91 .3453
101.7100
102.4720
89.7791
102.8758
105.8448
94.6989
1-7
-------
Table 1-5. Analysis of Variance Results for the 3-way Mixed Model
for ICP using EPA Method 3050A
ICP: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=EPA3050A
The MIXED Procedure
Class Level Information
Class Levels Values .
LABJD 3 123
SRMLEVEL 2 HIGH LOU
COMPOSIT 3 1-WIPE 2-UIPE 4-WIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
927.43114177
662.43820134
662.43129462
662.43121927
0.00002202
0.00000023
0.00000000
Convergence criteria met.
Cov Parm
LABJO
LAB ID*SRMLEVEL
LABJO*COMPOSIT
LABJD*SRMLEV*COMPOS
Residual
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error
Z Pr > |Z|
5.09913705
0.81293052
0.16184772
0.00000000
1.00000000
65.66017473
10.46788110
2.08406822
0.00000000
12.87672289
72.02885617
10.89665288
2.06566234
1.41488270
0.91
0.96
1.01
9.10
0.3620
0.3367
0.3130
0.0001
Model Fitting Information for ICP_REC
Description Value
Observations 180.0000
Variance Estimate 12.8767
Standard Deviation Estimate 3.5884
REML Log Likelihood -491.111
Akaike's Information Criterion -496.111
Schwarz's Bayesian Criterion -504.009
-2 REML Log Likelihood 982.2218
1-8
-------
Table 1-5. (Continued)
ICP: 3-UAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=EPA3050A
Source
Tests of Fixed Effects
HOF DDF Type III F Pr > F
SRMLEVEL 1 2 5.53 0.1431
COHPOSIT 2 4 3.45 0.1346
SRMLEVEL*COMPOSIT 2 4 3.33 0.1410
ESTIMATE Statement Results
Parameter
Estimate
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
SRM HIGH VS LOW
SRM LOU MEAN
SRM HIGH MEAN
1-U VS 2-U
1-W VS 4-U
2-W VS 4-W
1-WIPE MEAN
2-UIPE MEAN
4-UIPE MEAN
-6.35108986
97.49647291
91.14538304
-0.60513353
-3.34154735
-2.73641382
93.00536768
93.61050121
96.34691503
Std Error DDF
2.70120921
5.07607660
5.07607660
1.38777121
1.38777121
1.32176608
4.96827414
4.95024288
4.95024288
T Pr >
-2.35
19.21
17.96
-0.44
-2.41
-2.07
18.72
18.91
19.46
Alpha
Lower
Upper
0.1431
0.0027
0.0031
0.6853
0.0737
0.1072
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
-17.9735
75.6559
69.3048
-4.4582
-7.1946
-6.4062
79.2112
79.8664
82.6028
5.2713
119.3371
112.9860
3.2479
0.5115
0.9334
106.7995
107.3546
110.0910
Level
SRMLEVEL HIGH
SRMLEVEL LOW
COMPOS!T 1-WIPE
COMPOS IT 2-WIPE
COMPOSIT 4-WIPE
SRMLEVEL*COMPOSIT HIGH 1-WIPE
SRMLEVEL*COMPOSIT HIGH 2-WIPE
SRMLEVEL*COMPOSIT HIGH 4-WIPE
SRMLEVEL*COMPOSIT LOW 1-WIPE
SRMLEVEL*COMPOSIT LOW 2-WIPE
SRMLEVEL*COMPOSIT LOW 4-WIPE
Least Squares Means
LSMEAN Std Error DDF
91.14538304
97.49647291
93.00536768
93.61050121
96.34691503
88.73652837
90.74296245
93.95665832
97.27420700
96.47803997
98.73717175
5.07607660
5.07607660
4.96827414
4.95024288
4.95024288
5.17552717
5.14085545
5.14085545
5.17552717
.14085545
5.14085545
T Pr > |T| Alpha Lower
17.96
19.21
18.72
18.91
19.46
17.15
17.65
18.28
18.80
18.77
19.21
Upper
0.0031
0.0027
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
69.3048
75.6559
79.2112
79.8664
82.6028
74.3670
76.4697
79.6834
82.9046
82.2047
84.4639
112.9860
119.3371
106.7995
107.3546
110.0910
103.1061
105.0163
108.2300
111.6438
110.7513
113.0105
1-9
-------
Table 1-6. Analysis of Variance Results for the 3-way Mixed Model
for ICP Recovery using the WOHL Method
ICP: 3-UAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=WOHL
The MIXED Procedure
Class Level Information
Class Levels Values
LAB_ID 3 123
SRMLEVEL 2 HIGH LOW
COMPOSIT 3 1-WIPE 2-WIPE 4-WIPE
REML Estimation Iteration History
Iteration Evaluations Objective Criterion
0 1 1233.9150460
1 1 941.32300026
Convergence criteria met.
0.00000000
Cov Parm
LABJD
LAB ID*SRMLEVEL
LAB~ID*COMPOSIT
LAB~ID*SRMLEV*COMPOS
Residual
Covariance Parameter Estimates (REML)
Ratio Estimate Std Error
4.08088497
0.34261207
0.32377314
0.88843638
1.00000000
99.26100538
8.33349097
7.87526425
21.60979519
24.32340194
110.42337665
17.14295083
16.24058601
16.71494438
2.44459387
Z Pr > |Z|
0.90
0.49
0.48
1.29
9.95
0.3687
0.6269
0.6277
0.1961
0.0001
Model Fitting Information for ICP_REC
Description Value
Observations 216.0000
Variance Estimate 24.3234
Standard Deviation Estimate 4.9319
REML Log Likelihood -663.639
Akaike's Information Criterion -668.639
Schwarz's Bayesian Criterion -677.006
-2 REML Log Likelihood 1327.277
1-10
-------
Table 1-6. (Continued)
ICP: 3-WAY MIXED MODEL ANOVA WITH ALL INTERACTIONS
LABORATORY: RANDOM FACTOR
METHOD=WOHL
Tests of Fixed Effects
Source NDF DDF Type III F Pr > f
Tests of Fixed Effects
Source NDF DDF Type III F Pr > F
SRMLEVEL
COMPOSIT
SRMLEVEL*COMPOSIT
4.87 0.1580
7.12 0.0480
3.04 0.1575
Parameter
ESTIMATE Statement Results
Estimate Std Error DDF T Pr > |T| Alpha
Lower
Upper
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
ICP:
ICFh
ICP:
SRM HIGH VS LOU -7
SRM LOU MEAN 83
SRM HIGH MEAN
1-W VS
1-U VS
2-U VS
1-UIPE
2-UIPE
4-UIPE
2-U
4-U
4-U
MEAN
MEAN
MEAN
76
3
13
9
85
81
72
.25684385
.43261385
.17577001
.29175757
.14287877
.85112120
.28240404
.99064647
.13952528
3.28759346
6.27424629
6.27424629
3.62340879
3.62340879
3.62340879
6.40628306
6.40628306
6.40628306
2
2
2
4
4
4
4
4
4
-2.21
13.30
12.14
0.91
3.63
2.72
13.31
12.80
11.26
0.1580
0.0056
0.0067
0.4150
0.0222
0.0531
0.0002
0.0002
0.0004
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
-21.4022
56.4367
49.1799
-6.7684
3.0827
-0.2091
67.4957
64.2040
54.3528
6.8885
110.4285
103.1717
13.3520
23.2031
19.9113
103.0691
99.7773
89.9262
Level
SRMLEVEL HIGH
SRMLEVEL LOU
COMPOSIT 1-UIPE
COMPOSIT 2-UIPE
COMPOSIT 4-UIPE
SRMLEVEL*COMPOSIT
SRMLEVEL*COMPOSIT
SRMLEVEL*COMPOSIT
SRMLEVEL*COMPOSIT
SRMLEVEL*COMPOSIT
SRMLEVEL*COMPOSIT
Least Squares Means
LSMEAN Std Error DDF
T Pr > |T| Alpha Lower
Upper
HIGH 1-UIPE
HIGH 2-UIPE
HIGH 4-UIPE
LOU 1-UIPE
LOU 2-UIPE
LOU 4-UIPE
76.17577001
83.43261385
85.28240404
81 .99064647
72.13952528
81 .94296926
81.66936157
64.91497920
88.62183883
82.31193138
79.36407135
6.27424629
6.27424629
6.40628306
6.40628306
6.40628306
6.80946660
6.80946660
6.80946660
6.80946660
6.80946660
6.80946660
2
2
4
4
4
4
4
4
4
4
4
12.14
13.30
13.31
12.80
11.26
12.03
11.99
9.53
13.01
12.09
11.65
0.0067
0.0056
0.0002
0.0002
0.0004
0.0003
0.0003
0.0007
0.0002
0.0003
0.0003
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
49.1799
56.4367
67.4957
64.2040
54.3528
63.0369
62.7633
46.0089
69.7157
63.4058
60.4580
103.1717
110.4285
103.0691
99.7773
89.9262
100.8491
100.5755
83.8211
107.5279
101.2180
98.2702
1-11
-------
50272-101
REPORT DOCUMENTATION
PAGE
[.REPORT NO. 2.
EPA 747-R-96-003
4. TiUe and Subtitle
Analysis of Composite Wipe Samples for Lead Content
7. Author(s) Nancy Friederich, Karin Bauer
9. Performing Organization Name and Address
Battclle Memorial Institute
505 King Avenue
Columbus, Ohio 43201-2693
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
Office of Pollution Prevention and Toxics
401 M Street S.W.
Washington, D.C. 20460
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
3. Recipient's Accession No.
5 . Report Date
July 1996
6.
8. Performing Organization Rept.
No.
10. ProjeclTask/Work Unit No.
1 1 . Contract6 or Grant(G) No.
c 68-05-0137,68-05-0008
(G)
13. Type of Report & Period
Covered
Final Report
14.
15. Supplementary Notes
16. Abstract (Limit 200 words): This study was undertaken to (1) investigate the feasibility of developing further the existing sample preparation methods
analyze composite wipe samples while meeting basic data quality objectives for accuracy and precision; and (2) determine whether compositing of wipes,
acceptable on a technical or performance basis, could reduce the cost of sample preparation and analysis relative to single-wipe methods. From a numbei
of available sample preparation methods, two were selected: the modified EPA Method 3050A and the Wisconsin Occupational Health Laboratory (WOH
method.
The study found that acceptable performance can be achieved using the EPA Method 3050A for composites containing two or four wipes based
the recovery and precision results obtained with the standard reference material used in this study. The results were especially promising using FAA
analysis. The WOHL method was found to provide acceptable performance for single wipes and two-wipe composites. It is possible to achieve acceptable
performance with the WOHL method for four-wipe composites, although some difficulties might be encountered. Furthermore, the WOHL method was
affected by the SRM loading level when ICP was used. A statistically significant, estimated loss in recovery of 1.75% for each gram of SRM was found
when preparing the samples with the WOHL method and analyzing them by ICP. This limitation of the WOHL method may be a consideration in sampling
applications where high dust loading levels are expected, such as for window sills and troughs in risk assessment. It is recommended that every laboratory
would need to validate its performance with any method for the analysis of composite wipes as well as single-wipe samples. A significant cost benefit was
derived from compositing. In summary, this study found that compositing two or four wipes into a single sample is a viable alternative for the analysis of
lead in dust wipe samples and does result in cost savings.
17. Document Analysis a. Descriptors
Lead, laboratory analysis, compositing, dust, wipes
b. Identifiers/Open-Ended Terms
Lead, Lead-Based Paint, Lead-Based Paint Risk Assessment, Lead-Based Paint Clearance, Analysis Costs
c. COS ATI Field/Group
18. Availability Statement
Release Unlimited
19. Security Class (This Report)
Unclassified
20. Security Class (This Page)
Unclassified
21. No. of Pages
170
22. Price
(See ANSI-239.18)
OPTIONAL FORM 272 (4-77)
(Formerly NT1S-35)
Department of Commerce
------- |