United States
Environmental Protection
Agency
Office of
Toxic Substances
Washington, DC 20460
EPA 560/5-88-002
May, 1988
Toxic Substances
&EPA Assessing Asbestos
Exposure In Public Buildings
T.
-------
May, 1988
ASSESSING ASBESTOS EXPOSURE IN PUBLIC BUILDINGS
Prepared by:
Battelle Columbus Division
Washington Operations
2030 M Street, N.W.
Washington, D.C. 20036
EPA Contract No. 68-02-4294
Price Associates
1825 K Street, N.W.
Washington, D.C. 20006
EPA Contract No. 68-02-4294
Alliance Technologies Corporation
213 Burlington Road
Bedford, Massachusetts 07130
EPA Contract -6«-e2
R. J. Lee Group, Inc.
350 Hochberg Road
Monroeville, Pennsylvania 15146
EPA Contract No. 68-03-3406
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
EPA Contract No. 68-02-4252
for the:
Exposure Evaluation Division
Office of Toxic Substances
Office of Pesticides and Toxic Substances
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
-------
This document has been reviewed and approved for
publication by the Office of Toxic Substances, Office of
Pesticides and Toxic Substances, U.S. Environmental Protection
Agency. The use of trade names or commercial products does not
constitute Agency endorsement or recommendation for use.
11
-------
AUTHORS AND CONTRIBUTORS
This study pf asbestos in public buildings represents
the combined efforts of several organizations and many
individuals. The names of the principal authors and contributors
of the various organizations, along with the role of each
organization, are summarized below.
Battelle study design, planning, Quality Assurance Plan (QAP)
preparation, building and site selection, external analyses of
air samples, two laboratory audits of R. J. Lee Group, Inc., data
processing and management, statistical analyses, study report
preparation. Key Battelle staff included:
Jeff Hatfield Julius Ogden
Jerry Stockrahm Barbara Leczynski
Fred Todt
Price Associates, Inc. study design, planning, help with QAP,
placement of air sampling pumps, statistical analyses, study
report preparation. Key Price Associates staff included:
Bertram Price Jean Chesson
James Russell
Alliance Technologies Corporation provided the two "core"
raters, bulk sample collection and analyses, field work for air
sampling, provided portions of QAP and study report. Key
Alliance Technologies Corporation staff included:
Patrick Ford James Thomas
John Fitzgerald Richard Roat
R. J. Lee Group, Inc. analyses of air samples using
transmission electron microscopy, provided portions of QAP and
study report. Key R. J. Lee Group, Inc. staff included:
Rich Lee Drew Van Orden
George Dunmyre
Midwest Research Institute external analyses of bulk samples,
5 field audits of Alliance Technologies Corporation. Key Midwest
Research Institute staff included:
Paul Constant James McHugh
111
-------
Georgia Institute of Technology provided training course for
raters. Key Georgia Institute of Technology staff included:
Dave Mayer
Bill Ewing
William Spain
Steve Hays
McCrone Environmental Services
provided the building
inspector. Key McCrone Environmental Services staff included:
Rich Hatfield
Anthony Claveria
EPA, OTS, Exposure Evaluation Division supervised all aspects
of this study including design, planning, QAP, analyses of bulk
and air samples, study report. Key EPA staff were:
EPA Task Managers:
EPA Project Officers:
Joan Blake
Elizabeth Dutrow
Brad Schultz
Cindy Stroup
Mary Frankenberry
Joseph J. Breen
iv
-------
TABLE OF CONTENTS
AUTHORS AND CONTRIBUTORS
ACKNOWLEDGEMENTS
EXECUTIVE SUMMARY
1.0 INTRODUCTION ..................... 1
1.1 BACKGROUND .................... 1
1.2 OBJECTIVES .................... 3
1.3 ORGANIZATION OF REPORT .............. 4
2.0 CONCLUSIONS ...................... 5
3.0 QUALITY ASSURANCE ................... 15
3.1 INTRODUCTION ................... 15
3.2 BULK SAMPLE AND POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE .............. 15
3.2.1 Side-by- Side Duplicates .......... 16
3.2.2 External Analyses ........ * . . . . 16
3.2.3 Replicate Analyses ............ 16
3.3 AIR SAMPLE AND TRANSMISSION ELECTRON MICROSCOPY
QUALITY ASSURANCE .............. 17
3.3.1 Production Lot Blanks ........... 17
3.3.2 Field Blanks ............... 17
3.3.3 Field Audits ............... 18
3.3.4 Laboratory Audits ............. 18
3.3.5 Replicate and External Analyses ...... 18
3.3.6 Examination of Additional Grid Openings . . 19
4.0 STUDY DESIGN ..................... 21
5.0 BUILDING SELECTION, INSPECTION AND ASSESSMENT
FIELD METHODS ..................... 25
5.1 BUILDING SELECTION ................ 25
5.2 BUILDING INSPECTION ............... 29
5.3 ASSESSMENT .................... 30
6.0 RESULTS OF THE FIELD TEST OF THE ASSESSMENT METHOD . . 35
6.1 DATA ANALYSIS .................. 35
6.2 ASSESSMENT RESULTS ................ 36
6.3 CONCLUSIONS OF TEST OF ASSESSMENT METHOD ..... 39
-------
7.0 BULK SAMPLE ANALYSIS AND POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE 45
7.1 BULK SAMPLE ANALYSIS 45
7.2 BULK SAMPLE AND POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE . 45
7.2.1 Side-by-Side Duplicates 48
7.2.2 External Analyses 48
7.2.3 Replicate Analyses 49
7.3 Building Classification 49
8.0 AIR MONITORING 53
8.1 FIELD METHODS 53
8.2 AIR SAMPLE ANALYSIS 53
8.3 AIR SAMPLE AND TRANSMISSION ELECTRON MICROSCOPY
QUALITY ASSURANCE 54
8.3.1 Production Lot Blanks 55
8.3.2 Field Blanks 55
8.3.3 Flow Rate Calibration 56
8.3.4 Field Audits 56
8.3.5 Laboratory Audits 57
8.3.6 Replicate and External Analyses 57
8.3.7 Examination of Additional Grid Openings . . 59
8.4 ANALYSIS OF AIR MONITORING DATA 61
8.4.1 Methods 61
8.4.2 Results 64
REFERENCES 68
APPENDIX A RESPONSES OF INDIVIDUAL RATERS IN EACH ASSESSED
AREA WITHIN EACH REGION TO CONDITION, POTENTIAL
FOR DISTURBANCE, AND AIR FLOW FACTORS 71
APPENDIX B COUNTS OF THE RESPONSES OF THE RATERS IN EACH
ASSESSED AREA WITHIN EACH REGION FOR CONDITION,
POTENTIAL FOR DISTURBANCE, AND AIR FLOW
FACTORS 97
APPENDIX C CLASSIFICATION OF ACM CONDITION 133
APPENDIX D AIR SAMPLING FIELD METHODS 137
APPENDIX E AIR SAMPLE PREPARATION AND SUMMARY OF TEM
ANALYTICAL PROTOCOL 147
APPENDIX F ANALYSIS OF TEM GRID OPENING DATA 155
APPENDIX G AIR MONITORING DATA LISTING 161
APPENDIX H GLOSSARY 181
Vi
-------
LIST OF FIGURES
Figure 2-1 Scatter Plots and Medians of the Average
Airborne Asbestos structure Concentrations for
Each Building Category and Outdoors 6
Figure 2-2 The Average Agreement Index, Also Called the
Average A-Value, for Condition, Disturbance, and
Air Flow in Each of the 5 Study Regions .... 9
Figure 2-3 Average A-Values for Condition with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 10
Figure 2-4 Average A-Values for Disturbance with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 11
Figure 2-5 Average A-Values for Air Flow with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 12
.y
Figure 5-1 Overview of Field Methods 26
Figure 5-2 Form Used While at GSA Regional Off-ices to
Collect Information About the ACM in GSA
Buildings 27
Figure 5-3 Form Used While at GSA Regional Offices to
Collect Information About the ACM Within
Specific Areas in GSA Buildings 28
Figure 5-4 Chain-of-Custody Form for This Study 31
Figure 5-5 Assessment Form for Recording Information About
the ACM in a Given Area 33
Figure 6-1 The Average Agreement Index, Also Called the
Average A-Value, for Condition, Disturbance, and
Air Flow in Each of the 5 Study Regions .... 38
Figure 6-2 Average A-Values for Condition with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 40
Figure 6-3 Average A-Values for Disturbance with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 41
vii
-------
Figure 6-4 Average A-Values for Air Flow with Core and
Local Raters Plotted Separately in Each of the 5
Study Regions 42
Figure 7-1 Analytical Data Form for Reporting Bulk Sample
Analysis 46
Figure 8-1 Scatter Plots and Medians of the Average
Airborne Asbestos Structure Concentrations for
Each Building Category and Outdoors 65
Figure D-l Pump Diagram 141
Figure D-2 Field Data Form Used for Air Monitoring .... 144
LIST OF TABLES
Table 1 Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc) xv
Table 2-1 Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc) 7
Table 5-1 List of Study Regions, Week . of Rating, and
Number of Buildings and Sites Within Buildings
Rated in Each Study Region 32
Table 6-1 Percentages of Sites Which Showed Total
Agreement Among Raters and the Percentages of
Sites with Minimum Agreement Among Raters ... 37
Table 7-1 Final Classification of Buildings in Categories
1, 2, and 3 in Each Study Region 51
Table 8-1 Period of Air Sampling Within Each Study Region 54
Table 8-2 Comparison of Airborne Asbestos Concentrations
Estimated by the Original, Replicate (Same
Laboratory), and External (Different Laboratory)
TEM Analysis 58
Table 8-3 Estimated Mean, Variance, and Value of k for the
Number of Structures Counted Per Grid Opening
Based on Examination of 50 Openings 61
Table 8-4 Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc) 66
viii
-------
Table 8-5 Results of Randomization Test Indicating the
Statistical P-Values for Differences between
Median Airborne Asbestos Concentrations in Each
of the Three Building Categories and Outdoor
Concentrations 67
Table A-l Responses of Raters to Overall Condition
Variable Separated by Region, Building, and
Area 73
Table A-2 Responses of Raters to Potential for Disturbance
Separated by Region, Building, and Area .... 81
Table A-3 Responses of Raters to Air Flow Separated by
Region, Building, and Area 89
Table B-l Responses of Raters to Overall Condition
Variable, Separated by Region, Building, and
Area 99
Table B-2 Responses of Raters to Potential for
Disturbance, Separated by Region, Building, and
Area 114
Table B-3 Responses of Raters to Air Flow, Separated by
Region, Building, and Area 123
Table G-l Air Monitoring Data Listing Showing the Asbestos
Structure Concentration (s/cc) at Each Site that
was Air Sampled 163
-------
ACKNOWLEDGEMENTS
Many people are to be acknowledged for their time and
effort towards the successful completion of this study. In
particular, from the U.S. Environmental Protection Agency we
thank Martin P. Halper, Director of the Exposure Evaluation
Division, Susan F. Vogt, Deputy Director of the Office of Toxic
Substances, and David Kling, Chief of the Regulatory and
Technical Assistance Section of the Chemical Control Division's
Hazard Abatement Assistance Branch.
The study could not have been performed without the
full cooperation of the General Services Administration (GSA).
At GSA, we thank Robert J. DiLuchio, Assistant Commissioner for
Real Property Management and Safety, and Henry J. Singer,
Director of the Safety and Environmental Management Division.
Special thanks go to the many regional and local GSA staff who
participated in the study, in a variety of ways.
We are grateful to all the individuals who served as
raters in the field in each of the study regions. They are to be
commended for the many hours of service that each provided for
the assessment portion of this study.
From Battelle, we acknowledge Jeanette Hochstedler for
assistance in the building selection process, Dean Margeson for
help with site selection, Nick Sasso for set-up of computer
files, Dennis Haney for field coordination, Jill Daffer for help
with data processing, and Jan Clark and Pat Lyday for data entry.
Finally, we thank Steve Jones, Katherine Raeder, and Karen
Krasner for their many hours of word processing and
administrative support.
-------
EXECUTIVE SUMMARY
The U.S. Environmental Protection Agency (EPA)
conducted a field study during 1987 to address a number of
asbestos exposure issues. The study was designed and conducted
during a period when information needs concerning asbestos in
buildings were rapidly changing and expanding. Late in 1985,
while administering the Asbestos School Hazard Abatement Act
(ASHAA) (Public Law 98-377), EPA was asked to conduct an air
monitoring study of asbestos levels in buildings. The request,
submitted to the 99th Congress in a report by the House Committee
on Appropriations (House Report 99-212), appropriated $500,000
for air monitoring studies when signed on November 25, 1985.
These air monitoring studies were intended to provide data on
asbestos exposure levels inside buildings with asbestos-
containing materials (ACM) and ambient outdoor levels. In
response to this request, EPA began planning an air monitoring
study to be conducted in the latter part of 1986 and early 1987.
Simultaneously, EPA was engaged in the development of
an assessment method for differentiating areas of ACM requiring
immediate abatement action from areas where abatement action
could be deferred. An assessment approach had been proposed and
refined during a two-day workshop by consultants, administrators,
and others with asbestos management experience (USEPA 1986a).
Plans were being developed to field test the assessment method.
Plans for the two studies were reconsidered when, in
October, 1986, the Asbestos Hazard Emergency Response Act (AHERA)
(Public Law 99-160) was signed into law. AHERA introduced
assessment and abatement concepts and terminology that were
different in some major respects from the concepts and
terminology that were about to be field tested. AHERA required
EPA to promulgate regulations within one year, addressing
inspections, abatement, and management of ACM in schools.
In addition to regulations pertaining to asbestos in
schools, Section 213 of AHERA required EPA to report to Congress
within one year on regulatory issues for public and commercial
buildings. Specifically, the report to Congress was to:
assess the extent to which asbestos-containing
materials are present in public and commercial
buildings;
assess the condition of asbestos-containing
materials in commercial buildings and the
likelihood that persons occupying such buildings,
including service and maintenance personnel, are,
or may be, exposed to asbestos fibers;
XI
-------
consider and report on whether public and
commercial buildings should be subject to the same
inspection and response action requirements that
apply to school buildings;
assess whether existing federal regulations
adequately protect the general public,
particularly abatement personnel, from exposure to
asbestos during renovation and demolition of such
buildings; and
include recommendations that explicitly address
whether there is a need to establish standards
for, and regulate asbestos exposure in public and
commercial buildings.
Section 213 of AHERA placed significant new demands on
EPA to collect and analyze data relating to asbestos in
buildings. Due to the short time frame to meet the AHERA
requirements, EPA had to rely primarily on existing data for the
Section 213 report to Congress. For example, to gain insight on
the extent and condition of ACM in public and commercial
buildings, additional analyses of data collected in a 1983-84 EPA
survey of asbestos - in buildings were conducted (Rogers 1987). In
addition to using existing information, EPA conducted a field
study which included air monitoring. The study was designed with
three objectives:
to determine if airborne asbestos levels are
elevated in buildings that have ACM;
to field test an assessment method developed to
facilitate abatement decision making in the
context of an asbestos management program; and
to gather data to help EPA make recommendations to
Congress on future regulation of public and
commercial buildings (in order to meet the AHERA
Section 213 requirements).
In order to satisfy these objectives, inspection,
assessment, and air monitoring were conducted in three types of
buildings: (1) buildings without ACM; (2) buildings with all or
most of the ACM in good condition allowing for a limited number
of areas of moderate damage; and (3) buildings which had at least
one area of significantly damaged ACM or numerous areas of
moderately damaged ACM. These three building types are
subsequently referred to as Categories 1, 2, and 3, respectively.
Severe resource and time constraints precluded doing a
national survey of public and commercial buildings, and
necessitated the identification of an appropriate subset of
xii
-------
buildings. During 1987, when the study was being planned, the
General Services Administration (GSA) was in the process of
implementing a national asbestos management program in federally
owned buildings. The population of GSA owned buildings provided
an efficient laboratory for this study. GSA buildings had been
inspected and the ACM had been identified and evaluated.
Buildings were selected based on prior evaluation and
identification with one of the building categories. Without this
information, an expensive and time-consuming preliminary study
involving a much larger number of buildings would have been
required to identify buildings that contained asbestos. Also,
building access was assured in GSA buildings, avoiding the access
problems typically encountered in privately-owned buildings.
Based on GSA asbestos building records, 67 buildings
distributed across five study regions were chosen for study and
initially classified into the three building categories defined
above. These buildings had been inspected for ACM previously by
GSA. The buildings were reinspected, bulk samples collected and
analyzed, and ACM condition rated in four or more sites per
building by four inspectors. The bulk samples and assessment
data that were collected by the field teams were used to field
test the assessment method and to verify the initial
classification of the building categories. Forty-nine buildings
were chosen for air monitoring. Among the 49 buildings six had
no ACM (Category 1), six had ACM primarily in good condition
(Category 2), and 37 had at least one area of significantly
damaged ACM or numerous areas of moderately damaged ACM (Category
3). A total of 387 air samples were collected from the 49
buildings (an average of eight per building) including 48 samples
of outdoor air (one per building with one exception where outdoor
sampling was not possible).
Buildings initially classified as Category 1 (no ACM)
or Category 2 (ACM primarily in good condition) were entirely
reinspected by an experienced building inspector to confirm the
classifications. During each of these inspections, bulk samples
were collected and analyzed by polarized light microscopy (PLM).
The ACM in buildings in Category 2 and Category 3 was evaluated
by an assessment team consisting of four raters. Two "core"
raters evaluated four or more sites in each Category 2 and 3
building in each of the five study regions. The other two raters
were "local" (regional EPA, GSA, or local city government staff)
and evaluated buildings in their own region only. With the
exception of one study region, the raters attended a professional
training course before the field work began. The rating data
were used to confirm the classification of Category 2 and 3
buildings and to measure the consistency of the rating method.
Air monitoring was conducted in an average of seven
areas inside each building and one area outside. Half of the
inside samples were collected near the most damaged ACM in each
xiii
-------
building, if damaged material was present. ("Most damaged" could
mean ACM in good or moderately damaged condition if no
significantly damaged ACM was found in the building.) The
remaining samples were collected in public areas. Each sample
was taken over a two-day period, eight hours per day, during
periods of normal building activity. The samples were analyzed
by transmission electron microscopy (TEM) using a direct filter
preparation method to estimate the airborne asbestos structure
concentrations within the buildings. Asbestos structures include
asbestos fibers and asbestos bundles, clusters, and matrices.
A comprehensive quality assurance (QA) program governed
all field data collection and laboratory analysis activities. A
variety of standard QA samples were collected and analyzed. The
results indicated a high level of precision and accuracy in the
data.
Conclusions;
Objective (1): Determine if airborne asbestos structure
levels are elevated in buildings that have asbestos-
containing materials.
While the differences in airborne asbestos levels are
small in absolute magnitude, the results of this study indicate a
tendency for average airborne asbestos levels in buildings with
ACM to be higher than average levels in buildings without ACM
(comparing the medians of the building averages). Airborne
asbestos concentrations in Category 1 buildings (no ACM) have the
smallest median level; the median level in Category 2 buildings
(all or most of the ACM in good condition allowing for a limited
number of areas of moderate damage) is higher than the median
level in Category 1; and the median level in Category 3 buildings
(at least one area of significantly damaged ACM or numerous areas
of moderately damaged ACM) is highest. The air monitoring
results are summarized in Table 1. The difference between
Category 1 and Category 3 medians is statistically significant at
the 0.02 level. The remaining comparisons among building
categories are not statistically significant (each has p-values
of 0.18 or greater).
The median of building averages in Category 3 buildings
is higher than the median ambient outdoor level. The evidence
for a significant difference is not strong, but is suggestive of
a trend: the difference is statistically significant at the 0.09
level of significance (i.e, the p-value is 0.09). The other two
building categories when compared to ambient outdoor levels
suggest no difference (that is, they have p-values greater than
0.60). Estimates of indoor asbestos levels are more precise than
estimates of outdoor levels because indoor levels are based on
several samples per building. Outdoor levels are based on one
xiv
-------
Table 1. Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc)
Statistic
Standard
deviation
ACM
Outdoor
Category I Category 2 Category 3
Median
Mean
Sample
size
<0. 00001
0.00039
48
(sites)
0.00010
0.00099
6
(buildings)
0.00040
0.00059
6
(buildings)
0.00058
0.00073
37
(buildings)
0.00096
0.00198
0.00052
0.00072
Notes:
1. The data points used in the calculation of each
statistic are the average concentration within a building (for
indoor samples) or the concentration outside each building (for
outdoor samples).
2. The mean for Category 1 is heavily influenced by
one sample in one building which produced an unexplained large
s/cc value. The Category 1 mean, excluding this one value, is
0.00020 s/cc.
sample per building. Thus, an observed difference between two
building categories corresponds to a smaller p-value than the
same observed difference between a building category and
outdoors.
The buildings in this study were selected from three
building categories in order to investigate relationships between
building category and airborne asbestos levels. It is important
to note that the method of selection does not allow formal,
statistical projection of the total number of buildings which
have characteristics measured in the study to the population of
GSA buildings, federally owned buildings, or public and
commercial buildings. However, since the buildings were selected
without prior knowledge of airborne asbestos levels or without
prior knowledge of any other variable measured in the study, the
resulting relationships are suggestive of true relationships in
buildings similar to those studied.
xv
-------
Objective (2); Field test an assessment method developed to
facilitate abatement decision making in the context of an
asbestos management program.
Using rater consistency as an evaluation criterion, the
definitions of material condition, potential for disturbance, and
air flow used in this study, show promise as assessment tools for
use in the field. P-values less than 0.05 indicated that
consistency among raters was greater than expected if ratings
were applied at random. Two hundred fifty seven areas within 60
buildings in five study regions were assessed by a team of two
core raters. In addition, in each of the five study regions a
team of two local raters evaluated the buildings in their study
region.
Assessment of material condition was the factor most
consistently rated across the five study regions. Assessment of
potential for disturbance was less consistent. Assessment of air
flow showed the greatest level of variability. Trends observed
in the data suggest that lack of consistency among raters can be
attributed, in part, to imprecision in definitions and lack of
training.
Objective (3): Gather data to help EPA make recommendations
under AHERA to Congress on future regulation of public and
commercial buildings (in order to meet the AHERA Section 213
requirements).
This objective was met indirectly by the total body of
information collected in this study which was combined with
information from a variety of sources and considered in the
preparation of the AHERA Section 213 report to Congress. While
the air monitoring data collected to satisfy this study's first
objective are not necessarily representative of air levels in all
public and commercial buildings, they provide information
pertinent to exposure issues raised in AHERA Section 213. The
conclusions regarding the second objective, assessment method
evaluation, are pertinent to management programs in public and
commercial buildings as well as the schools that are covered by
the AHERA regulations. This information will be considered in
the preparation of future guidance documents, training programs,
and regulations for public and commercial buildings.
xvi
-------
1.0 INTRODUCTION
1.1 BACKGROUND
This study, referred to as the Public Buildings Study,
was designed and conducted during a period when information needs
concerning asbestos in buildings were rapidly changing and
expanding. Late in 1985, while administering the Asbestos School
Hazard Abatement Act (ASHAA) (Public Law 98-377), EPA was asked
to conduct an air monitoring study of asbestos levels in
buildings. The request, submitted to the 99th Congress in a
report by the House Committee on Appropriations (House Report 99-
212), appropriated $500,000 for air monitoring studies when
signed on November 25, 1985. These air monitoring studies were
intended to provide data on asbestos exposure levels inside
buildings with asbestos-containing materials (ACM) and ambient
outdoor levels. In response to this request, EPA began planning
an air monitoring study to be conducted in the latter part of
1986 and early 1987.
Simultaneously, EPA was engaged in the development of
an assessment method for differentiating areas of ACM requiring
immediate abatement action from areas where abatement action
could be deferred. An assessment approach had been proposed and
refined during a two-day workshop by consultants, administrators,
and others with asbestos management experience (USEPA 1986a).
Plans were being developed to field test the assessment method.
Plans for the two studies were reconsidered when, in
October, 1986, the Asbestos Hazard Emergency Response Act (AHERA)
(Public Law 99-160) was signed into law. AHERA introduced
assessment and abatement concepts and terminology that were
different in some major respects from the concepts and
terminology that were about to be field tested. AHERA required
EPA to promulgate regulations within one year, addressing
inspections, abatement, and management of ACM in schools.
In addition to regulations pertaining to asbestos in
schools, Section 213 of AHERA required EPA to report to Congress
within one year on regulatory issues for public and commercial
buildings. Specifically, the report to Congress was to:
assess the extent to which asbestos-containing
materials are present in public and commercial
buildings;
assess the condition of asbestos-containing
materials in commercial buildings and the
likelihood that persons occupying such buildings,
including service and maintenance personnel, are,
or may be, exposed to asbestos fibers;
-------
consider and report on whether public and
commercial buildings should be subject to the same
inspection and response action requirements that
apply to school buildings;
assess whether existing federal regulations
adequately protect the general public,
particularly abatement personnel, from exposure to
asbestos during renovation and demolition of such
buildings; and
include recommendations that explicitly address
whether there is a need to establish standards
for, and regulate asbestos exposure in public and
commercial buildings.
Section 213 of AHERA placed significant new demands on
EPA to collect and analyze data relating to asbestos in
buildings. Due to the short time frame to meet the AHERA
requirements, EPA had to rely primarily on existing data for the
Section 213 report to Congress. For example, to gain insight on
the extent and condition of ACM in public and commercial
buildings, additional analyses of data collected in a 1983-84 EPA
survey of asbestos in buildings were conducted (Rogers 1987). In
addition to using existing information, EPA conducted a field
study which included air monitoring.
In order to meet the objectives of the public buildings
study, which are formally stated and discussed in Section 1.2,
inspection, assessment, and air monitoring were conducted in
three categories of buildings: (1) buildings without ACM; (2)
buildings with all or most of the ACM in good condition allowing
for a limited number of areas of moderate damage; and (3)
buildings which had at least one area of significantly damaged
ACM or numerous areas of moderately damaged ACM. A number of
practical problems prevented EPA from constructing a
comprehensive list of all public and commercial buildings from
which to draw a national probability sample of buildings with the
desired characteristics. To develop such a listing would have
required more resources than were available for this study and
the additional time necessary would have precluded the collection
of information in time to consider it in the preparation of the
AHERA Section 213 report, due to Congress in October 1987.
The General Services Administration (6SA) was in the
process of implementing a national asbestos management program in
federally owned buildings. The population of GSA owned buildings
provided an efficient laboratory for this study. GSA buildings
had been inspected and the ACM had been identified and evaluated.
Buildings were selected based on prior evaluation and
identification with one of the building categories. Without this
information, an expensive and time-consuming preliminary study
-------
involving a much larger number of buildings would have been
required to identify buildings that contained asbestos. Also,
building access was assured in GSA buildings/ avoiding the access
problems typically encountered in privately-owned buildings.
Details concerning the study design, including the method used to
select buildings and data collection techniques, are presented in
subsequent sections of this report.
1.2 OBJECTIVES
The Public Buildings Study had three objectives:
to determine if airborne asbestos levels are
elevated in buildings that have ACM;
to field test an assessment method developed to
facilitate abatement decision making in the
context of an asbestos management program; and
to gather data to help EPA make recommendations to
Congress on future regulation of public and
commercial buildings (in order to meet the AHERA
Section 213 requirements).
A brief discussion of each objective follows.
Objective (1); Determine if airborne asbestos levels are
elevated in buildings that have asbestos-containing
materials.
GSA buildings were selected from three categories of
buildings for air monitoring. Indoor and ambient outdoor air was
monitored for asbestos at each building. Airborne asbestos
levels measured in buildings with ACM were compared to levels in
buildings without ACM. Indoor and ambient outdoor levels were
also compared. These comparisons were used to determine if
airborne asbestos levels in buildings with ACM are elevated. The
buildings in this study were selected from three building
categories in order to investigate relationships between building
category and airborne asbestos levels. It is important to note
that the method of selection does not allow formal, statistical
projection of the total number of buildings which have
characteristics measured in the study to the population of GSA
buildings, federally owned buildings, or public and commercial
buildings. However, since the buildings were selected without
prior knowledge of airborne asbestos levels or without prior
knowledge of any other variable measured in the study, the
resulting relationships are suggestive of true relationships in
buildings similar to those studied.
-------
Objective (2); Field test an assessment method developed to
facilitate abatement decision making in the context of an
asbestos management program.
The assessment factors described in the November 7,
1986 draft of "Guidance for Assessing and Managing Exposure to
Asbestos in Buildings" (USEPA, 1986a) were tested in the field.
Consistency among different raters assessing the same sites was
evaluated. The buildings and areas in buildings used in the
study were selected to ensure that a range of ACM materials and
conditions would be rated. The rating data collected to test and
analyze consistency among raters, therefore, do not provide
information about the characteristics of ACM that can be
projected to the population of all federal buildings, The
results, however, can be used to suggest ways in which
consistency among raters can be improved.
Objective (3); Gather data to help EPA make recommendations
under AHERA to Congress on future regulation of public and
commercial buildings (in order to meet the AHERA Section 213
requirements).
Section 213 of AHERA requires that information be
developed to address the five specific issues previously listed.
Most of the information required for Section 213 has been
developed in other studies. Information collected in the current
study may be used to supplement these other sources, as
appropriate.
1.3 ORGANIZATION OF REPORT
The report consists of eight sections. This section,
Section 1, includes the background and objectives, which provide
an introduction to the study. Conclusions are summarized in
Section 2. Quality assurance, the study design, and the field
methods are presented in Sections 3, 4, and 5, respectively. The
sample analysis methods, statistical analysis, and results are
presented in Sections 6, 7 and 8. Section 6 addresses the field
test of the assessment factors, Section 7 addresses bulk
sampling, and Section 8 is the air monitoring portion of the
study.
-------
2.0 CONCLUSIONS
The study conclusions are organized and discussed in
relationship to the three study objectives. During 1987, when
the study was conducted, GSA was in the process of implementing
an asbestos management program in buildings. The impact of the
management program, if any, is not addressed in these
conclusions. Results of statistical tests are indicated by "p-
values." A p-value is the probability of obtaining a result as
extreme or more extreme than the result observed under the null
hypothesis of no difference or relationship between the factors
being studied. A small p-value indicates that the magnitude of
the observed result is unlikely under the null hypothesis, and
therefore lends support to the alternative hypothesis, namely
that the difference or relationship is real. Detailed
presentations of the results supporting the conclusions are found
in Sections 6, 7, and 8.
Objective (1); Determine if airborne asbestos structure
levels are elevated in buildings that have asbestos-
containing materials.
While the differences in airborne asbestos levels are
small in absolute magnitude, the results of this study indicate a
tendency for average airborne asbestos levels in buildings with
ACM to be higher than average levels in buildings without ACM
(comparing the medians of the building averages). Airborne
asbestos concentrations in Category 1 buildings (no ACM) have the
smallest median level; the median level in Category 2 buildings
(all or most of the ACM in good condition allowing for a limited
number of areas of moderate damage) is higher than the median
level in Category 1; and the median level in Category 3 buildings
(at least one area of significantly damaged ACM or numerous areas
of moderately damaged ACM) is highest. The air monitoring
results are presented in Figure 2-1 and summary statistics are
given in Table 2-1. The difference between Category 1 and
Category 3 medians is statistically significant at the 0.02
level. The remaining comparisons among building categories are
not statistically significant (each has p-values of 0.18 or
greater).
The median of building averages in Category 3 buildings
is higher than the median ambient outdoor level. The evidence
for a significant difference is not strong, but is suggestive of
a trend: the difference is statistically significant at the 0.09
level of significance (i.e., the p-value is 0.09). The other two
building categories when compared to ambient outdoor levels
suggest no difference (that is, they have p-values greater than
0.60). Estimates of indoor asbestos levels are more precise than
estimates of outdoor levels because indoor levels are based on
several samples per building. Outdoor levels are based on one
-------
0.0054
0.004
0.003
S
T
R
U
C
T
U
R
E
C
0
N
C
E
N
T
R
A
T
I
0
N
s/cc
0.000
0.002
0.001
A
C
A
B
Outdoor
(48 sites)
Figure 2-1.
Building
category 1
(6 buildings)
Building
category 2
(6 buildings)
Building
category 3
(37 buildings)
Scatter plots*and medians of the average airborne
asbestos structure concentrations for each building
category and outdoors.
*The data points for each scatter plot are the average concentration
within a building (for indoor samples) or the concentration outside each
building (for outdoor samples).- A=l data point, B=2 data points, ...,
J=10 data points, and Z=41 data points. The diamond represents the median
of the data points in each scatter plot.
-------
Table 2-1. Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc).
Statistic
Standard
deviation
ACM
Outdoor Category 1 Category 2
0.00096
0.00198
0.00052
Category 3
f f '
Median
Mean
Sample
size
<0. 00001
0.00039
48
(sites)
0.00010
0.00099
6
(buildings)
0.00040
0.00059
6
(buildings)
0.00058
0.00073
37
(buildings)
0.00072
Notes:
1. The data points used in the calculation of each
statistic are the average concentration within a building (for
indoor samples) or the concentration outside each building (for
outdoor samples).
2. The mean for Category 1 is heavily influenced by
one sample in one building which produced an unexplained large
s/cc value. The Category 1 mean, excluding this one value, is
0.00020 s/cc.
sample per building. Thus, an observed difference between two
building categories corresponds to a smaller p-value than the
same observed difference between a building category and
outdoors.
The buildings in this study were selected from three
building categories in order to investigate relationships between
building category and airborne asbestos levels. It is important
to note that the method of selection does not allow formal,
statistical projection of the total number of buildings which
have characteristics measured in the study to the population of
GSA buildings, federally owned buildings, or public and
commercial buildings. However, since the buildings were selected
without prior knowledge of airborne asbestos levels or without
prior knowledge of any other variable measured in the study, the
-------
resulting relationships are suggestive of true relationships in
buildings similar to those studied.
Objective (2); Field test an assessment method developed to
facilitate abatement decision making in the context of an
asbestos management program.
Using rater consistency as an evaluation criterion, the
definitions of material condition, potential for disturbance, and
air flow used in this study, show promise as assessment tools for
use in the field. P-values less than 0.05 indicated that
consistency among raters was greater than expected if ratings
were applied at random. Two hundred fifty-seven areas within 60
buildings in five study regions were assessed by a team of two
core raters. In addition, in each of the five study regions a
team of two local raters evaluated the buildings in their study
region. The study regions are numbered in the order in which
they were sampled, (i.e., Study Regions 1 to 5). This numbering
scheme is not related to the regional classification used by
either GSA or EPA.
Figure 2-2 shows that assessment of material condition
was the factor most consistently rated across the five study
regions. Assessment of potential for disturbance was less
consistent. Assessment of air flow showed the greatest level of
variability. Trends observed in Figures 2-3, 2-4 and 2-5 suggest
that lack of consistency among raters can be attributed, in part,
to imprecision in definitions and lack of training, both of which
can be remedied. This conclusion is based on the following
results:
There is greater consistency among raters when
assessing condition than when assessing potential
for disturbance. In this study, condition was
defined in quantitative terms (e.g., terms such as
greater than 10% damage) whereas the definitions
for disturbance were more qualitative.
Consistency in air flow ratings varies from study
region to study region, with low consistency in
some study regions and high consistency in others.
The present two-part scale does not distinguish
significant air flow from very slight air flow. A
three-part scale (high, moderate, and low/none)
for air flow may increase consistency.
8
-------
All Raters
Condition
Study Region
Disturbance
Air Row
Figure 2-2.
The average agreement index, also called the average A-
value, for condition, disturbance, and air flow in each
of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement. The basis for A-values is
explained in Section 6.
-------
Condition
Study Region
Local raters
* Core raters
Figure 2-3. Average A-values for condition with core and local raters
plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scbring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement. The basis for A-values is
explained in Section 6.
-------
1
Disturbance
0.9 -
0.8 -
a>
o»
0.7 -
§ 0.6 -
< 0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0
1
Study Region
Local raters
* Core raters
Figure 2-4. Average A-values for disturbance with core and local
raters plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement. The basis for A-values is
explained in Section 6.
-------
Air Flow
10
Study Region
Local raters
Core raters
Figure 2-5. Average A-values for air flow with core and local raters
plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement. The basis for A-values is
explained in Section 6.
-------
Consistency between the core raters is greater
than consistency between the local raters. The
core raters had more experience in applying the
assessment method.
Region 5 local raters, who did not attend the
training, showed the least consistency in
assessing condition and potential for disturbance.
Objective (3); Gather data to help EPA make recommendations
under AHERA to Congress on future reflation of public and
commercial buildings (in order to meet the AHERA Section 213
requirements).
This objective was met indirectly by the total body of
information collected in this study which was combined with
information from a variety of sources and considered in the
preparation of the AHERA Section 213 report to Congress. While
the air monitoring data collected to satisfy this study's first
objective are not necessarily representative of air levels in all
public and commercial buildings, they provide information
pertinent to exposure issues raised in AHERA Section 213.
The conclusions regarding the second objective,
assessment factor evaluation, are pertinent to management
programs in public and commercial buildings as well as the
schools that are covered by the AHERA regulations. This
information will be considered in the preparation of future
guidance documents, training programs, and regulations for public
and commercial buildings.
13
-------
3.0 QUALITY ASSURANCE
3.1 INTRODUCTION
This section presents an overview of quality assurance
(QA) procedures followed in this study and the results of those
procedures. The details on the statistical analysis of QA data
are presented in Sections 7 and 8. Guidelines for the QA of the
data collected in this study were set forth in a comprehensive
Quality Assurance Plan (Hatfield et al. 1987). All procedures
were employed, and unavoidable deviations were documented.
Two types of measurements were collected in the field:
bulk samples and air samples. Bulk samples were analyzed with
polarized light microscopy (PLM) to determine the percentage of
asbestos present. For the purposes of this study, only the
presence (">1%") or absence ("none detected" or "trace") of
asbestos from a given site was utilized. Air samples were
analyzed with transmission electron microscopy (TEM) direct
filter preparation to estimate the asbestos structure
concentration for each site. QA procedures were performed for
all aspects of data collection and analysis.
Chain-of-custody procedures were implemented for all
samples collected during the project. Field custody procedures
were used to document the location and handling of each sample
from the time of collection until received by the analytical
laboratory. At this point, internal laboratory records were used
to document the chain-of-custody of each sample through to its
final disposition.
3.2 BULK SAMPLE AND POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE
As specified in the QA plan, various types of quality
control (QC) samples were collected and analyzed to determine the
accuracy and precision of asbestos content estimates. Included
were: side-by-side duplicates, and external (referee laboratory)
analyses and replicate analyses. A side-by-side duplicate is a
sample collected in the immediate area of the original sample but
handled separately. The degree of agreement of the analyses of
the original sample with its duplicate indicates the level of
precision in the sample collection and field handling procedures.
An external analysis is one in which the sample is analyzed a
second time by another analytical laboratory. This type of
analysis is performed as a QC check on the performance of the
method by the primary laboratory. The degree of agreement of the
original analysis with the external analysis indicates the
consistency of the method performance. A replicate analysis is
one in which the same sample is analyzed twice by the same
analytical laboratory. The degree of agreement of the two
15
-------
analyses indicates the level of precision in the laboratory
analysis procedures.
All three of the QC procedures described above were
used for the analyses of the bulk samples in this study. The
results of these QC analyses indicated a very high level of
precision and accuracy in the bulk sample collection and PLM
analyses.
3.2.1 Side-by-Side Duplicates
A total of 279 bulk samples were collected by the
building inspector and rating team in the field and analyzed for
asbestos content. Of these 279 samples, 20 were collected in the
field as side-by-side duplicates (20/259 = 8%). With respect to
presence or absence of asbestos, all 20 of the side-by-side
duplicates agreed with their original sample (100% agreement).
3.2.2 External Analyses
From the 279 bulk samples, 31 were randomly chosen for
external analysis by a second laboratory (31/279 = 11%). With
respect to the presence or absence of asbestos, 30 of the 31
externals agreed with their original samples (30/31 = 97%
agreement). The one disagreement did not result in the
misclassification of the building category as verified by
additional bulk samples collected at that area by the rating
team.
The primary analytical laboratory also performed its
own QC checks. The laboratory participated in the EPA Asbestos
Bulk Sample Analysis Quality Assurance Program. Sixteen bulk
samples (for which the "true" percentage of asbestos is known)
were submitted as blind QC samples along with the field samples.
With respect to the presence or absence of asbestos, all 16 of
these results agreed with the original determination (100%
agreement).
3.2.3 Replicate Analyses
The analytical laboratory randomly chose 33 of the 279
field-collected samples and resubmitted them for replicate bulk
sample analysis (33/279 = 12%). With respect to the presence or
absence of asbestos, 32 of the 33 replicates agreed with the
original result (32/33 97% agreement). The one disagreement
did npt result in the misclassification of the building category
because that building's classification was based on another area
where several samples were used to verify the presence of
asbestos.
16
-------
3.3 AIR SAMPLE AND TRANSMISSION ELECTRON MICROSCOPY
QUALITY ASSURANCE
The QA procedures used to ensure the accuracy and
precision of the air sample collection and TEM analyses included
the collection of production lot and field blanks, field audits,
laboratory audits, replicate and external analyses, and a study
to further evaluate the results obtained by the TEM method.
Production lot blanks are filters chosen prior to the
start of field work. They are analyzed by the analytical
laboratory to check for filter contamination. Field blanks are
filters taken into the field and handled in the same manner as
exposed air sample filters. Their purpose is to check for
contamination which might occur in the field but not as a result
of air sampling. Field audits determine whether the field team
is following set procedures. Laboratory audits determine the
same for the analytical laboratory personnel. Replicate and
external analyses serve the same purpose as discussed above for
PLM analyses.
3.3.1 Production Lot Blanks
Blank filters from prescreened production lots were
randomly selected three times during the project: at the
beginning of field activities, in the middle, and near
completion. Each time, two filter cassettes were randomly
selected from a previously unopened box of 50 filters. A total
of 26 production lot blanks were selected in this way for
analysis. The analysis of the production lot blanks indicated
that there was not a problem with background filter
contamination.
3.3.2 Field Blanks
During the pump set-up, preloaded filter cassettes were
selected as field blanks. These filters were labeled and handled
in an identical manner as were the sample filters, except that
they were not attached to the sampling pump. The filters were
capped during active sampling periods and open faced during the
non-run hours when the actual sample cassettes were also open
faced. Field blanks were collected in 30 of the buildings
sampled. The purpose of the field blanks was to measure
contamination which might occur during periods when the pumps
were not running.
Of the 30 field blanks collected, 19 were selected for
analysis. If a high level of contamination was found from the
17
-------
analysis results of the 19 blanks, the remaining 11 blanks would
have been analyzed. The 19 blanks that were analyzed were chosen
at random from the 30 blanks collected. No structures were
detected in 18 of the 19 field blanks that were analyzed. A
single fiber was counted on the remaining blank. This level of
blank contamination corresponds to an airborne asbestos structure
concentration of 0.00015 s/cc when 5,000 L of air is collected, a
very low level of contamination. Thus, it was not necessary to
analyze the remaining blanks.
3.3.3 Field Audits
Five field audits were conducted by an independent
field auditor, one audit in each of the study regions. The field
auditor accompanied the field crew during pump set-up in several
buildings per study region. He checked to be sure that the field
crew was following the guidelines set forth in the Quality
Assurance Plan (Hatfield et al. 1987), and documented any
violations in procedures so they could be corrected. For
example, an air hose on one pump was found to be punctured. This
was noted and immediately corrected. The field auditor also
measured 216 flow rates in pumps in these buildings. This was
done in order to estimate the relative accuracy of the flow
rates, defined as [(field value-audit value)/(audit value)] x
100. The percentage of flow rates within + 20% relative accuracy
was 99%.
3.3.4 Laboratory Audits
To ensure the accuracy of the air sample analyses using
TEM, two laboratory audits were performed. An independent
laboratory auditor visited the TEM analytical laboratory to
verify that all procedures specified in the Quality Assurance
Plan (Hatfield et al. 1987) were followed. He audited the
analytical laboratory twice, at the beginning of the analyses and
at the end.
3.3.5 Replicate and External Analyses
Twenty air samples to be used for replicate and
external TEM analyses were chosen at random from a total of 387
air samples (20/387 = 5%). These samples were receded and
submitted to the original analytical laboratory for replicate
analyses. They were then sent to a second TEM laboratory for
external analyses. Thus, for these 20 samples, 3 measurements
were .collected for each sample: the original, the replicate, and
the external.
18
-------
Very few asbestos structures were counted in any of the
original, replicate, or external analyses. No asbestos
structures were detected on 13 of the 20 samples (65%). A single
structure was detected by one or more of the three analyses on
the remaining seven samples. Statistical analysis (Section
8.3.6) of these data indicated that there is no evidence of
inconsistency between the original, replicate, and external
analyses.
3.3.6 Examination of Additional Grid Openings
In most of the original 387 TEM analyses, 10 grid
openings were counted per sample filter to estimate the number of
asbestos structures on each filter. These results were used to
compute structure concentrations. To determine whether 10 grid
openings per sample provided a sufficiently precise estimate of
the number of structures on the filter, 40 additional grid
openings were counted on 16 randomly-selected air samples, for a
total of 50 grid openings per filter. Statistical analysis
(Section 8.3.7) indicated that, in general, examination of 10
grid openings is sufficient.
19
-------
4.0 STUDY DESIGN
The objectives of the Public Buildings Study, which are
formally stated and discussed in Section 1.2, call for inspection
and air monitoring in three types of buildings: (1) buildings
without ACM; (2) buildings with all or most of the ACM in good
condition (allowing for a limited number of areas of moderate
damage); and (3) buildings with at least one area of
significantly damaged ACM or numerous areas of moderately damaged
ACM. These are referred to subsequently as Categories 1, 2, and
3, respectively. The objectives focus on relationships between
building categories and airborne asbestos levels and ratings of
ACM characteristics in different types of buildings.
Buildings were selected for the study from the
population of federally owned buildings in five geographically
dispersed regions of the United States (two cities on the east
coast, one midwestern city, one western city, and a west coast
region consisting of two cities). In this report the study
regions are identified as Regions 1 to 5 according to the order
in which they were sampled. This numbering scheme is not related
to the regional classification used by either the General
Services Administration (GSA) or EPA.
A target quota of 20 buildings was specified for
evaluation in each study region: four buildings in Category 1;
four buildings in Category 2; and 12 buildings in Category 3.
The initial classification of these buildings into categories was
based on inspection and evaluation information available from GSA
records. The classification of each building was to be confirmed
by the project team, and ten buildings (two in Category 1, two in
Category 2, and six in Category 3) were to be selected in each
study region for the air monitoring portion of the study.
By pooling data across the study regions, estimates for
Category 3 would be based on measurements in 30 buildings (i.e.,
five study regions, six buildings per study region). Estimates
for the other categories were to be based on measurements in 10
buildings (i.e., five study regions, two buildings per study
region). When buildings are selected randomly and the
coefficient of variation of individual measurements is between 1
and 1.25, a range observed in previous studies, the likelihood of
detecting a five-fold difference between Category 3 and one of
the other categories using Student's t-test with a significance
level of 0.05 is at least 0.90 (i.e., the statistical power is at
least 0.90). Relatively low airborne asbestos levels were
anticipated in Categories 1 and 2 (i.e., buildings with no ACM or
buildings with ACM primarily in good condition), and therefore,
sample sizes sufficient to detect a five-fold differential were
considered adequate.
21
-------
The study was intended to investigate two
relationships: (i) the relationship between airborne asbestos
levels and building category as defined previously; and (ii) the
relationship between assessments conducted by different raters.
It is important to determine if these relationships can be used
to differentiate among extremes i.e., damaged ACM and ACM in
good condition, or damaged ACM and no ACM. Therefore, buildings
with these conditions must be included in the study regardless of
how frequently or infrequently these conditions occur in the
population of all buildings.
The population of GSA owned buildings provided an
efficient laboratory for this study. GSA buildings had been
inspected and the ACM had been identified and evaluated.
Buildings were selected based on prior evaluation and
identification with one of the building categories. Without this
information, an expensive and time-consuming preliminary study
involving a much larger number of buildings would have been
required to identify buildings that contained asbestos. Also,
building access was assured in GSA buildings, avoiding the access
problems typically encountered in privately-owned buildings.
By selecting buildings in the manner described, the
results regarding the relationships studied apply, from a formal
statistical perspective, only to the buildings in the study. The
approach used to select the buildings, however, is similar to
many experimental studies where the experimental units selected
satisfy predetermined specified criteria. Under these
circumstances, projecting the total number of buildings which
have characteristics measured in the study to an appropriate
target population is not possible. However, since the buildings
were selected without prior knowledge of airborne asbestos levels
or without prior knowledge of any other variable that was
measured and analyzed in the study, the resulting relationships
are suggestive of true relationships in buildings similar to
those studied. [In concept, the study circumstances are typical
of "analytical" studies, which are differentiated from
"enuraerative" studies by Deming (1950).]
As discussed in detail in Section 6, differences in
airborne asbestos levels among building categories and outdoors
are indicated by plots and tables of summary statistics. The
measured airborne asbestos levels, and consequently the
statistics calculated from them, are subject to various sources
of statistical error including air sampling and analytical error.
A statistical test was applied to provide a quantitative measure
of the strength of evidence associated with the observed
differences (i.e., probabilities that the observed differences
may have occurred only by chance were estimated).
For planning purposes, specifically the determination
of sample size discussed above, airborne asbestos levels were
22
-------
assumed to follow a lognormal distribution and to be amenable to
standard analysis of variance techniques used in previous studies
(USEPA 1985b, 1986b; Tuckfield et al. 1987). Once the data were
collected it was apparent that standard methods of analysis were
not appropriate for this study because of the large number of
zero observations. Therefore a permutation (also referred to as
randomization) approach was used. The permutation test, which is
based on the null hypothesis that all measured levels are
independent observations from the same underlying statistical
distribution, is consistent with the objectives and design of the
study. Buildings were selected according to condition of ACM and
without any knowledge of airborne asbestos levels. Therefore, in
the absence of any relationship between condition and airborne
asbestos levels, the measured values will be equivalent
observations from a single distribution. To compare building
categories, a "p-value," the level of significance, is estimated
for each comparison. The p-value is the probability of obtaining
a difference as great or greater than the difference observed
under the hypothesis that no true difference exists. A small p-
value indicates that the magnitude of the observed difference is
unlikely under the hypothesis of no true difference, and
therefore lends support to the alternative hypothesis, namely
that the difference is real.
23
-------
5.0 BUILDING SELECTION, INSPECTION AND ASSESSMENT
FIELD METHODS
Sixty-seven buildings were selected for this study from
a population of several thousand GSA buildings. Selection was
based on GSA's asbestos building records. The assessment factors
were tested in 60 of the buildings. Forty-nine buildings were
selected from the original 67 for air monitoring. This section
describes the field methods used to select and categorize the
buildings by asbestos condition. It also describes the
procedures used to test the assessment factors and to collect
information to select the air monitoring sites. An overview of
these methods is provided in Figure 5-1.
5.1 BUILDING SELECTION
Initial selection of buildings was achieved by
reviewing the existing asbestos building records maintained by
GSA in each of the five study regions. Buildings were chosen
based upon the following criteria:
Each building must be GSA-owned to ensure easy access.
Each building must contain occupied office space (e.g.,
storage sheds were eliminated).
Each building must have adequate asbestos building
records indicating whether or not an assessment had
been performed and whether or not ACM was present.
Buildings with information on condition of the ACM were
preferred.
Buildings with surfacing ACM were preferred.
All buildings within a given study region must be
within a small enough area to facilitate sampling
logistics. The exception to this was Study Region 4,
which consisted of 2 cities, each sampled separately
but counted as one study region.
Buildings were excluded if more than a 3-day security
clearance was required for the field personnel prior to
gaining entry.
Information on each building was collected using the
forms shown in Figures 5-2 and 5-3. Selected portions of the
asbestos records for each building were photocopied.
Only 67 buildings were found to satisfy the above
criteria. Therefore, the target quota of 20 buildings per study
region could not be obtained. However, there were sufficient
25
-------
i
I
Review GSA's
building records
Select 67
buildings and classify
as Categories
1. 2. or 3
Building Inspector
inspects 28 buildings
from Categories
1 and 2
Reclassify buildings
if necessary and
select 49 buildings
for air sampling
I
Rating team
assesses selected
material In 60 buildings
from Categories
2 and 3
i
6 buildings from
Category 1
6 building* from
Category 2
!
37 buildings from
Category 3
Select air monitoring
sites within each
of 49 buildings
approx 7 inside
1 outside
Ajseismant data
from approximately
4 sites within each
of 60 buildings
Figure 5-1. Overview of field methods
26
-------
U.S. EPA Study of Airborne Asbestos Levels in Buildings
Initial Building Survey - Overall Building
Identification and Location
Building ID
Building Name
Address
City/State/Zip
Contact person
GSA owned?
GSA inspected?
If yes, date?
Eligibility Verification
Yes NO
Phone
Yes
No
Bldg type okay?
Within geog. area?
Building and ACM Information
Group letter
Assessment number
Exposure index
Asbestos-containing surfacing
material present?
Yes No _
If yes, condition?
Don't know
Asbestos-containing
TSI present?
If yes, condition?
Yes
No
Don't know
Number of ACM areas identified
Bldg height (number of stories)
Year built
Comments
Figure 5-2.
Form used while at GSA regional offices to collect
information about the ACM in GSA buildings.
27
-------
U.S. EPA Study of Airborne Asbestos Levels in Buildings
Initial Building Survey - ACM Containing Areas
Building ID
Building Name
Location
Area Number __________
Type of ACM: Surfacing
Condition: Surfacing ________
TSI
Other
TSI
Have abatement procedures been implemented?
Comments:
Yes
No
Area Number ____________
Type of ACM: Surfacing
Condition: Surfacing
Location
TSI
Other
TSI
Hava abatement procedures been implemented?
Comments:
Yes
No
Figure 5-3.
Form used while at GSA regional offices to collect
information about the ACM within specific areas in
GSA buildings.
28
-------
buildings to allow air sampling in 10 buildings per study region
as planned. The breakdown of the number of buildings selected in
each study region is as follows: Study Region 1 17 buildings,
Study Region 2 10 buildings, Study Region 3 11 buildings,
Study Region 4 13 buildings, and Study Region 5 16
buildings.
Based on GSA's building records, each building was
initially classified as Category 1, Category 2, or Category 3.
Buildings which contained ACM, but had little or no information
on the condition of the material were placed in Category 2.
Based on subsequent field work performed in this study the
classification of buildings into each category was confirmed
prior to air monitoring. These results are discussed in
Section 7-
5.2 BUILDING INSPECTION
In order to verify the classification of the Category 1
and Category 2 buildings, a qualified, independent building
inspector inspected the buildings which were initially placed in
these categories. The verification of the classification of
buildings in Category 3 was accomplished by the assessment team.
In each Category . 1 building, the building inspector
performed a one-day inspection and collected bulk samples from
any areas containing friable surfacing material and thermal
systems insulation (TSI). In each Category 2 building, the
inspector bulk sampled any friable surfacing material and TSI in
areas that were indicated in the GSA building records as
containing damaged ACM. He also searched for and sampled any
other friable material he determined to be in worse condition.
Building inspection was performed in each study region during the
weeks of: Study Region 1 March 16, Study Region 2 March
23, Study Region 3 April 6, Study Region 4 April 20, and
Study Region 5 May 4.
Bulk sampling was performed following the random
sampling procedures described in USEPA (1985a). In addition,
convenience samples were collected in the following
circumstances:
Pipe wrap in fan/boiler room;
Limited access to sprayed-on material due to
piping, ductwork, etc.; and
Specific requests by the building escort or
management not to collect more than one sample per
site.
29
-------
Undamaged TSI, except for exposed material, was not sampled since
this would violate GSA's asbestos management program guidelines.
Only friable surfacing materials and TSI were sampled (e.g.,
ceiling and floor tiles were not sampled). Detailed notes of the
sampled areas, including the condition of the material, were
maintained by the building inspector. A chain-of-custody form
was completed for all samples (Figure 5-4). The bulk samples
were mailed to the laboratory for analysis at the end of the
inspection period in each study region.
Resulting from the building inspection is the final
classification of each building in Category 1 and Category 2.
These results are presented in Section 7.
5.3 ASSESSMENT
This section provides the methods for the field test of
the assessment method for asbestos described in USEPA (1986a).
The "field test" involved training individuals to apply the
assessment factors and determining the degree of consistency
among different individuals assessing the same material.
Another goal, which was achieved by using the
individuals applying the assessment factors, was to further
verify the condition of ACM in the buildings which were
classified (based on GSA's asbestos building records) as
Category 2 and Category 3. The assessment team, also called the
rating team, collected bulk samples of damaged friable material
to verify either the presence or absence of damaged ACM in these
buildings. This information was then used to reclassify these
buildings, when necessary, for air monitoring. The information
was also used in the selection of air monitoring sites for pump
placement.
The rating team consisted of two "core" raters who
visited every study region, and two "local" raters in each study
region. The local raters were regional EPA, GSA, or local city
government staff.
A professional, two-day training course was held for
the raters 1$ weeks before field work began. This produced a
group of individuals with experience and training typical of
those likely to be assessing ACM in real-world applications.
Study Region 5 local raters did not attend the training course
because the decision to include this city was made after the
training course had taken place.
The rating team assessed approximately four
predetermined sites in each of between 9 and 15 buildings in each
of the study regions. The sites were selected to represent a
variety of conditions, locations, and types of ACM. Table 5-1
30
-------
Sample Slle
Dal* Sampled.
Shipped By
DaU Shipped .
Carrier
ID Number
Invoice Number
Dale Recelvad_
Received By
Condition
. Test Number.
SAMPLE
NUMBER
SAMPLE
DESCRIPTION
SHIPPING
CONDITION
RECEIVING
CONDITION
Signature ol Sender .
Signature ol Receiver.
Dale.
Dale.
Figure 5-4. Chain-of-custody form for this study.
-------
Table 5-1. List of Study Regions, Week of Rating, and Number of
Buildings and Sites Within Buildings Rated in Each
Study Region
Study
region
1
2
3
4
5
Week of
rating
March 23
April 6
April 20
May 4
May 18
Number of
buildings
15
9
11
11
14
Number of
sites3
64
37
47
52
57
aRated by two or more raters
lists the study regions, week of rating, and the number of
buildings and sites. The study regions are numbered in the order
in which they were sampled. This numbering scheme is not related
to the regional classification used by either GSA or EPA. Each
rater completed the form shown in Figure 5-5 in order to obtain
the necessary assessment factors described in USEPA (1986a).
This form contains the key factors discussed in USEPA (1986a) for
assessing ACM (i.e., overall condition, potential for
disturbance, and air flow) as well as other information. The
exact definitions of these factors can be found in the Glossary
(Appendix H). The rating team did not discuss the sites before
or during rating in order to ensure that the ratings were
accomplished independently.
After all predetermined sites in a building were rated
by the team, additional information was collected in order to
verify the building category and to collect information to select
air monitoring sites. The location of the area thought to be the
most damaged ACM in the building, based on GSA building records
or information collected by the building inspector, was contained
in a sealed envelope and opened by the rating team at the
conclusion of the rating process. (Note that "most damaged"
could mean good condition if there were no moderately or
significantly damaged areas in the building.) The rating team
was instructed to bulk sample this area to verify the presence of
asbestos and any areas which they had visited and found to be in
32
-------
Assessment Form
Building ID
Area ID
Type of ACM
Surfacing Thermal System Insulation Other.
Description
Condition
Percent Damage _____ %
Distribution of Damage: Localized Even
Type of Damage:
Deterioration Physical Damage Water Damage
Description
Overall condition: Sig Damage Mod Damage Good.
Potential for Disturbance
Accessibility: High Low
Description
Vibration: High Low None..
Source
Air Erosion: High Low None_
Source
Overall Potential for Disturbance:
High Moderate Low_
Air Flow
In Air Flow? Yes No
Type of Flow
Comments
Signed Date
Figure 5-5. Assessment form for recording information about the
ACM in a given area.
33
-------
worse condition. Occasionally, the building escort suggested
additional areas that were also bulk sampled.
Bulk sampling was performed following the random
sampling procedures in USEPA (19853), although convenience
samples were collected in the following circumstances:
Pipe wrap in fan/boiler room;
Limited access to sprayed-on material due to
piping, ductwork, etc.; and
Specific requests by the building escort or
management not to collect more than one sample per
site.
Undamaged TSI, except for exposed material, was not sampled
because that would violate GSA's asbestos management program
guidelines. Only friable surfacing materials and TSI were
sampled (e.g., ceiling and floor tiles were not sampled). The
rating team kept detailed notes of the areas sampled and also
sketched the location of the material. The bulk samples were
carried back to the laboratory for analysis at the end of the
rating period in each study region.
The purpose of bulk sampling the areas described above
was to confirm the initial classification of buildings based on
GSA records as Category 2 or Category 3. Depending on whether
the areas rated contained asbestos or not, the buildings were
reclassified as necessary prior to air sampling. These results
are discussed in Section 7.
34
-------
6.0 RESULTS OF THE FIELD TEST OF THE ASSESSMENT METHOD
6.1 DATA ANALYSIS
ACM condition for a specific area within a building can
take one of three possible values: good, moderate damage, or
significant damage. These values are coded as 1, 2, and 3,
respectively. Likewise, potential for disturbance is rated as
low, moderate, or high. These are coded as 1, 2, and 3,
respectively. For the remainder of this report, potential for
disturbance will be referred to as "disturbance." Air flow takes
the values yes or no, coded as 1 and 0, respectively.
The data listing in Appendix A shows the responses of
the individual core and local raters for each of the three
assessment factors at each site. Table A-l gives the condition
of the sites, Table A-2 gives the disturbance rating of the
sites, and Table A-3 shows the air flow of the site. The
frequency of occurrence of each score is given in Appendix B.
Consistency of ratings may be measured in a variety of
ways. One simple measure is the percentage of sites where
perfect agreement occurs (i.e., all raters give the same rating).
Another measure is the percentage of sites where minimum
agreement occurs (i.e., at least one rater assigns the maximum
rating and at least one rater assigns the minimum rating). While
both measures have been calculated and summarized, neither
distinguishes between situations where just one rater disagrees
with the remaining raters from situations where there is little
agreement among any of the raters. An "agreement" index, A, was
developed to measure the overall degree of consistency among
raters taking into account both the number and magnitude of
disagreements. A is defined as 1 minus the sum of the
differences between the raters' scores, normalized so it takes
values between 0 and 1. It is calculated as:
A = 1 -
max
where i * n = number of raters which rated the site;
%'-. $£ &' . V
« X; = response of the i rater; and
max = the theoretical maximum that the numerator
can take for n raters.
When there is perfect agreement between raters at a
site, A=l. A=0 indicates minimum agreement. A similar
measurement of rater agreement was used in a previous EPA study
(USEPA 1981). In that case the measure was divided by the number
35
-------
of comparisons rather than normalized to range from 0 to 1.
Although the two measures are not quantitatively identical, the
qualitative results of the two studies can be compared.
A-values were calculated for each site and averaged
within study regions to provide a summary of rater consistency
for each of the three assessment factors.
A statistical test was developed to determine whether
the assessment factors generate more agreement than would be
expected if the raters were assigning ratings at random. The
probability distribution of A was calculated under the null
hypothesis of random assessment. From this, the probability
distribution of A averaged over sites was generated by computer
simulation for each combination of number of ratings, number of
raters, and number of sites. If the observed average is found in
the upper tail of the distribution, it is unlikely that the
ratings are being assigned at random. An observed average equal
to the (l-a)th percentile, i.e., a p-value of a, would be called
statistically significant with a significance level equal to oc.
6.2 ASSESSMENT RESULTS
Table 6-1 shows the percentages of sites with total
agreement among raters and the percentages of sites having
minimum agreement. Minimum agreement occurs at a site if at
least one rater scores the lowest value of a factor and at least
one rater scores the highest value of the factor. Note that all
the percentages are sensitive to the number of raters present.
Total agreement is more likely with fewer raters, while minimum
agreement is more likely when there are more raters.
Overall, there is greater total agreement and less
minimum agreement for condition than for disturbance, although
the relationship varies slightly from study region to study
region. The highest frequency of minimum agreement occurs in
Study Region 5. High total agreement and minimum agreement for
air flow are not surprising since there are only two ratings (0
and 1), and the 2 percentages must sum to 100.
In Figure 6-1, average A-values for condition,
disturbance, and air flow are plotted by study region with study
region numbered in chronological order. The A-value for a given
site is based on the ratings of between two and four raters
depending on how many raters were actually present at the site
and whether all of the raters completed all the assessment
factors at a site (there were occasional missing entries). Each
average measure of consistency shown in Figure 6-1 has a
significance level less than 0.05 when tested against random
ratings. The study, therefore, indicates a tendency for
consistency among raters.
36
-------
Table 6-1.
Percentages of Sites Which Shoved Total Agreement
Among Raters and the Percentages of Sites with
Minimum Agreement Among Raters
Study
region
Assessment
factor
% Total
agreement
% Minimum
agreement
All
Condition
Disturbance
Air flow
Condition
Disturbance
Air flow
Condition
Disturbance
Air flow
Condition
Disturbance
Air flow
Condition
Disturbance
Air flow
Condition
Disturbance
Air flow
31.3
35.9
55.6
48.6
13.5
29.7
29.8
25.5
72.3
38.5
15.4
75.0
22.8
17.5
66.7
33.1
22.6
61.3
0
4.7
44.4
0
8.1
70.3
6.4
4.3
27.7
1.9
7.7
25.0
19.3
19.3
33.3
5.8
8.9
38.7
Note: Minimum agreement occurs at a site if at least one rater
scores the lowest value of a factor and at least one rater scores
the highest value of a factor.
37
-------
All Ratera
oo
1
1
0
O»
Condition
Study Region
Disturbance
Air Row
Figure 6-1. The average agreement index, also called the average A-
value, for condition, disturbance, and air flow in each
of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement.
-------
The average A-value for condition tends to be higher
than the A-value for disturbance. This suggests that the
definition of condition used in this study can be applied more
consistently than the definition of potential for disturbance.
The average A-value for air flow varies from study region to
study region. In the first two study regions there is less
agreement on air flow than on condition and disturbance. In the
last three study regions the opposite is true.
A previous EPA study on assessing ACM also showed
greater consistency when rating condition (USEPA 1981). The
assessment method used in that study differs from the current
one, but factors related to condition (water damage, physical
damage) showed greater rating consistency than those related to
disturbance (accessibility, activity). The 1981 study considered
a factor "air-moving system" which was more narrowly defined than
in the present study and showed high consistency among raters.
A-values were also calculated for core and local raters
separately (Figures 6-2, 6-3 and 6-4). The core raters were the
same two individuals across all study regions, while the local
raters were represented by different individuals in each study
region. For condition and disturbance (Figures 6-2 and 6-3,
respectively), the amount of agreement between the core raters
begins at between 0.8 and 0.9 and remains fairly constant across
study regions (with some slight improvement over time). For the
core raters, all average A-values for condition and disturbance
have p-values less than 0.05. In Study Region 1, local raters
have about the same amount of agreement as the core raters.
Local raters' level of agreement tends to decrease for the
remaining study regions. For condition, the test of consistency
among local raters had a significance level less than 0.05 in all
study regions except Study Region 5. Study Region 5 raters did
not attend the training course. A-values for disturbance have
significance levels less than 0.05 only in Study Regions 1 and 4.
For air flow (Figure 6-4), the relationship between the core and
local raters varied from study region to study region. In Study
Regions 2 and 4, the core raters show less agreement than the
local raters. In Study Regions 1, 3, and 5, the local raters
show less agreement. All air flow A-values have significance
levels less than 0.05 with the exception of Study Region 2 for
core raters and Study Region 1 for local raters.
6.3 CONCLUSIONS OF TEST OF ASSESSMENT METHOD
Based on the criterion of rater consistency, the
assessment factors, as defined in this study, show promise as
assessment tools for use in the field. Lack of consistency can
be attributed, in part, to imprecision in definitions and lack of
training, both of which can be remedied. This conclusion is
based on the following results:
39
-------
Condition
Study Region
Local raters
Core raters
Figure 6-2. Average A-values for condition with core and local raters
plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement.
-------
1
0.9-
0.8-
I
0.7-
0.6 -
0.5 -
0.4-
0.3 -
0.2 -
at-I
0
Disturbance
1
Study Region
Local ratera
Core ratera
Figure 6-3. Average A-values for disturbance with core and local
raters plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement.
-------
Air now
to
234
Study Region
Local raters + Core raters
Figure 6-4. Average A-values for air flow with core and local raters
plotted separately in each of the 5 study regions.
*The A-value is an agreement index developed for this analysis
to demonstrate consistency in scoring of the assessment
factors. A-values range from 1 for maximum agreement among
raters to 0 for minimum agreement.
-------
Consistency between raters is significantly
greater than would be predicted if assessment were
at random.
There is greater consistency among raters when
assessing condition than when assessing potential
for disturbance. In this study, condition was
defined in quantitative terms (e.g., terms such as
greater than 10% damage) whereas the definitions
for disturbance were more qualitative.
Consistency in air flow ratings varies from study
region to study region, with low consistency in
some study regions and high consistency in others.
The present two-part scale does not distinguish
significant air flow from very slight air flow. A
three-part scale (high, moderate, and low/none)
for air flow may increase consistency.
Consistency between the core raters is greater
than consistency between the local raters. The
core raters had more experience in applying the
assessment method.
Region 5 local raters, who did not attend the
training, showed the least consistency in
assessing condition and potential for disturbance.
43
-------
7.0 BULK SAMPLE ANALYSIS AND POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE
7.1 BULK SAMPLE ANALYSIS
All bulk samples collected by the building inspector
and rating team were analyzed with polarized light microscopy
(PLM), as per Appendix A of USEPA (1982). For the purposes of
this study, however, only the information about presence (">1%")
or absence ("none detected" or "trace") of asbestos was used.
The method described in USEPA (1982) is most commonly
used to identify and quantify asbestos fibers in bulk samples.
It can distinguish fibers of the serpentine group (chrysotile)
from those of the amphibole group (amosite, crocidolite,
anthophyllite, tremolite-actinolite), and is sensitive to
asbestos content as low as 1%. Asbestos fibers are reported as
area percentages of the total sample. This method is limited to
fiber sizes greater than 1 JOB in length.
Sample preparation involved gross examination to
characterize the sample and to determine homogeneity. If the
sample was not homogeneous, each phase was separated for
individual identification. At least one microscope slide was
prepared for each phase by teasing a small piece of sample from
the bulk and mounting it on the slide with a refractive index oil
(n = 1.54) and cover slip. All gross examinations and slide
preparations were performed in the glovebox to protect the
analyst.
Sample examination involved analysts trained in
classical crystallography techniques. The specific techniques
used depended upon the nature of the sample, but in general, the
lination utilized magnifications of 20X to 400Z, and at least
four fields were counted.
The analytical laboratory used standard asbestos data
forms (Figure 7-1), bound in a notebook format for recording all
analytical data. Data were recorded in duplicate, with one copy
remaining in the analyst's notebook, and the other being
submitted to the data entry technician.
7.2 BULK SAMPLE AHD POLARIZED LIGHT MICROSCOPY
QUALITY ASSURANCE
The collection of bulk samples by the building
inspector and rating team was a QA activity because this
information was used to verify the classification of the building
categories. The final classification of these buildings is
discussed in Section 7.3. In addition, specific QA procedures
45
-------
ASBESTOS ANALYSIS WORKSHEET
ASBESTOS ANALYSIS GROUP
PROJECT: CONTRACT i
FIELD ID * ID # PacUity ID *
SAMPLE LOCATION:
SAMPLE DESCRIPTION/CONDITION:
ASBESTOS FOUND: NONE DETECTED TRACE >1%
PLEASE COMPLETE FOR ALL ITEMS. ENTER "NONE DETECTED" IF NONE FOUND.
MATERIAL PERCENT IF A RANGE, SPECIFY HERE.
ASBESTOS;
CHRYSOTILE -
AMOSITE " - "
OTHER ASBESTOS
SPECIFY OTHER TYPE(S) OF ASBESTOS:
FIBROUS NON-ASBESTOS:
GLASS WOOL OR MINERAL WOOL
FIBERGLASS
CELLULOSE
OTHER FIBROUS NON-ASBESTOS
MATERIALS
SPECIFY OTHER TYPE(S) OF FIBROUS NON-ASBESTOS:_
NON-FIBROUS:
PERLITE
VERMICULITE
OTHER NON-FIBROUS
MATERIALS
SPECIFY OTHER TYPE(S) OF NON-FIBROUS:
DATE OF ANALYSIS: ANALYST:
Figure 7-1. Analytical data form for reporting bulk sample
analysis.
46
-------
were used for the bulk sample collection and analyses to ensure
their precision and accuracy.
Chain-of-custody procedures were implemented for all
samples collected during the project. Field custody procedures
were used to document the location of a sample from the time of
collection until its receipt by the analytical laboratory. At
this point, internal laboratory records were used to document the
handling of the sample through its final disposition.
Standard sample custody (traceability) procedures were
used during this project. Each sample was labeled with a unique
random identification number immediately after collection. This
number was recorded in the field logbook along with the following
information:
Name(s) of the sampler;
Date of collection;
Sample location;
Sketch of location (rating team only); and
Comments.
A chain-of-custody form was filled out in the field for
all samples. A copy of the form was included with each shipment
of samples to the analytical laboratory. Figure 5-4 is a
representative copy of the chain-of-custody form used during the
project.
Upon receipt of the samples at the analytical
laboratory, the following steps were taken:
Sample labels were cross-checked with the
accompanying custody form.
Samples were logged in a master sample logbook.
Prior to and after analysis, samples were stored
in a controlled area.
Samples handled by laboratory personnel were
traced by proper log-in/log-out procedures.
Specific QA procedures were performed to estimate the
precision and accuracy of the bulk sample collection and
analyses: side-by-side duplicates, and external and replicate
analyses. A side-by-side duplicate is a sample collected in the
immediate area of the original sample but handled separately.
The degree of agreement of the analyses of the original sample
47
-------
with its duplicate indicates the level of precision in the sample
collection and field handling procedures. An external analysis
is one in which the sample is analyzed a second time by another
analytical laboratory. This type of analysis is performed as a
quality control (QC) check on the performance of the method by
the primary laboratory. The degree of agreement of the original
analysis with the external analysis indicates the consistency of
the method performance. A replicate analysis is one in which the
same sample is analyzed twice by the analytical laboratory. The
degree of agreement of the two analyses indicates the level of
precision in the laboratory analysis procedures.
All three of the QA procedures described above were
used for the analyses of the bulk samples in this study. The
results of these QA samples indicated a very high level of
precision and accuracy in the bulk sample collection and PLM
analyses.
7.2.1 Side-bv-Side Duplicates
A total of 279 bulk samples were collected by the
building inspector and rating team in the field and analyzed for
asbestos content. Of these 279 samples, 20 were collected in the
field as side-by-side duplicates (20/259 = 8%). With respect to
presence or absence of asbestos, all 20 of the side-by-side
duplicates agreed with their original samples (100% agreement).
7.2.2 External Analyses
From the 279 bulk samples, 31 were randomly chosen for
external analysis by a second laboratory (31/279 = 11%). With
respect to the presence or absence of asbestos, 30 of the 31
externals agreed with their original samples (30/31 = 97%
agreement). The one disagreement did not result in the
misclassification of the building category as verified by
additional bulk samples collected at that area by the rating
team.
The primary analytical laboratory also performed its
own QC checks. The laboratory participated in the EPA Asbestos
Bulk Sample Analysis Quality Assurance Program. Sixteen of the
EPA bulk QA samples (for which the "true" percentage of asbestos
is known) were submitted as blind QC samples along with the field
samples. With respect to the presence or absence of asbestos,
all 16 of these results agreed with the original EPA
determination (100% agreement).
48
-------
7.2.3 Replicate Analyses
The analytical laboratory randomly chose 33 of the 279
field-collected samples and resubmitted them for replicate bulk
sample analysis (33/279 = 12%), With respect to the presence or
absence of asbestos, 32 of the 33 replicates agreed with the
original result (32/33 - 97% agreement). The one disagreement
did not result in the misclassification of the building category
because that building's classification was based on another area
where several samples were used to verify the presence of
asbestos.
7.3 Building Classification
The goal of the building inspector was to verify the
classification of the Category 1 and Category 2 buildings. A
building's final classification as Category 1 depended on GSA
records and the bulk sampling results. A Category 1 building is
defined as one in which no friable asbestos-containing surfacing
materials or TSI were noted in the GSA records and none was found
during the building inspection. Asbestos was defined to be
present in a bulk sample if the PLM analysis found greater than
1% asbestos. Absence of asbestos was defined as a result of
"none detected" or "trace". Using these definitions, 6 of the 7
potential Category 1 buildings remained in that category. Because
of the presence of asbestos in the remaining building, it was not
included in the study.
In Category 2 buildings, the building inspector
searched for and bulk sampled any friable surfacing material and
TSI that were indicated in the GSA building records to contain
damaged ACM and any areas in worse condition. The damaged areas
found in these buildings were subsequently visited by the rating
team. The functions of the rating team, in addition to field
testing the assessment method, were to help verify the
classification of buildings in Category 2 and 3 before air
sampling. Based on the assessments of the rating team and the
bulk sample results, the final classification of Category 2 and 3
buildings was achieved.
To assess the condition of the ACM in a given area, the
rating team used definitions from USEPA (1986a) for the various
conditions: good condition, moderate damage, and significant
damage. The definitions of these condition categories for a
given area are found in Appendix C. Only the assessments of the
rating team leader (Core Rater 1, the most experienced member of
the team) were used for the classification of building category.
A building was classified as Category 2 if all or most of the
areas with friable asbestos-containing surfacing materials or TSI
were in good condition with perhaps a few areas (less than four)
of moderate damage. Recall that the rating team visited the
49
-------
areas that were thought to be the most damaged ("most damaged"
can mean good condition if there are no moderately or
significantly damaged ACM areas in the building), and they bulk
sampled those areas as well as any areas in worse condition. If
an area contained greater than 1% asbestos in the bulk sample
results and the condition was rated as significantly damaged,
then the building would be reclassified as Category 3.
Similarly, a number of moderately damaged areas with positive
bulk sample results in one of these buildings would exclude the
building from Category 2.
The areas in the Category 3 buildings were chosen based
on the GSA building records. Since the rating team assessed what
was thought to be the most damaged ACM area in each building and
bulk sampled those and any areas in worse condition, this
information was used to verify the condition of Category 3
buildings for air sampling. A building's final classification
was defined as Category 3 if there was at least one significantly
damaged area of friable asbestos-containing surfacing material or
TSI, or there were numerous moderately damaged areas. Thus, most
buildings placed in Category 3 were found to have positive (>1%)
bulk sample results in areas with damaged ACM in the building.
However, in 4 buildings (1 in Study Region 2, 1 in Study
Region 3, 2 in Study Region 4), positive confirmation of damaged
ACM within the building was not obtained. These buildings were
classified as Category 3 because GSA records indicated the
presence of damaged ACM within the buildings.
Forty-nine of the initial 67 buildings were selected
for air monitoring. Table 7-1 shows the final classification of
the 49 buildings by region. One of the remaining 18 buildings
was not used for air sampling because it was initially classified
as Category 1, but was found to contain asbestos. The remaining
17 buildings were excluded from air sampling because of
geographic location (related to sampling logistics), uncertainty
about classification as a Category 2 or 3 building, or, in the
case of one building, because the ACM had been removed since the
GSA records had been reviewed.
50
-------
PLEASE RETURN TO?
NCIC/OTS CHEMICAL LIBRARY
401 M ST., S.W., TS-793
WASHINGTON, D.C. 20460
Table 7-1. Final Classification of Buildings in Categories 1, 2,
and 3 in Each Study Region
Number of Number of
Study Category 1 Category 2
region buildings buildings
1 2 2
2 1 1
300
4 21
512
Total : "6 ~6
Number of
Category 3
buildings
5
8
10
7
7
17
Note: Only the 49 buildings chosen for air sampling were
classified.
51
-------
8.0 AIR MONITORING
8.1 FIELD METHODS
Air samples were collected in 49 buildings (9 in
Study Region 1, 10 in each of Study Regions 2 through 5). Seven
areas were sampled inside each building and one area was sampled
outside. (One building, Building 44, could not be sampled
outside and one building, Building 18, could be sampled at only
six indoor sites.) About half of the sampling pumps were located
near the most damaged ACM in the building. "Most damaged" could
mean ACM in good condition if there were no moderately or
significantly damaged materials in the building. The rest were
placed in public areas nearby. Table 8-1 shows the five Study
Regions, period of air sampling, and the number of buildings and
sites at each building. A more detailed description of the air
sampling field methods appears in Appendix D.
The air samples were collected on cellulose ester
(Millipore) filters. Two side-by-side samples were drawn at each
site: one of 5,000 L and one of 2,500 L. If the high volume
sample contained too much debris for analysis, then the low
volume sample was available for analysis. Each pump ran for two
periods of approximately eight hours during consecutive weekdays
during normal building activity. The volume of air sampled was
determined from the flow rate readings taken at the beginning,
during, and at the end of the sampling period. One field blank
filter was collected in each building to check for contamination
from sources other than the sampled air.
An independent auditor accompanied the field crew to
each study region to ensure that all procedures were followed.
After sampling was completed in each region, the filters were
hand carried to the laboratory and analyzed by transmission
electron microscopy (TEM).
8.2 AIR SAMPLE ANALYSIS
The TEM protocol is given in the Quality Assurance Plan
(Hatfield et al. 1987). A summary appears in Appendix E of this
report. Sample preparation involved collapsing the filter,
plasma etching, and directly coating the filter with a thin layer
of carbon by evaporative deposition under vacuum. The samples
were cleared with acetone, leaving the particles attached to the
carbon film. The samples were analyzed at a magnification of
20,OOOX. A minimum of ten grid openings with a total area of
0.062 eg. mm were examined on each filter. The total structure
count includes asbestos fibers (structures with essentially
parallel sides and an aspect ratio of 3:1 or greater) and
asbestos bundles, clusters, and matrices as defined in the TEM
53
-------
Table 8-1. Period of Air Sampling Within Each Study Region
Study
region
1
2
3
4
5
Period of
air sampling
April 12-17
May 6-12
May 20-28
June 10-16
June 21-26
Number
Category 1
2(16)
1(8)
0
2(16)
1(8)
of buildinqs
Category 2
2(16)
1(8)
0
1(8)
2(16)
(sites)
Category 3
5(40)
8(63)
10(80)
7(56)
7(55)
Note: The number of sites within buildings, including outdoor
sites, is given in parentheses.)
protocol. Two samples from Building 43 were too heavily loaded
to be analyzed with the direct TEM method.
8.3 AIR SAMPLE AND TRANSMISSION ELECTRON MICROSCOPY
QUALITY ASSURANCE
Chain-of-custody procedures were implemented for all
air samples collected during the project. Field custody
procedures were used to document the existence of a sample from
the time of collection until received by the analytical
laboratory. At this point, internal laboratory records were used
to document the custody of the sample through its final
disposition.
Standard sample custody (traceability) procedures were
used during this project. Each sample was labeled with a unique
random identification number immediately after collection. This
number was recorded on the field data form along with the
following information:
Name(s) of the sampler;
Date of collection;
Sample location;
54
-------
Sketch of location; and
Comments.
A chain-of-custody form was filled out in the field for
all air samples. A copy of the form was included with each
shipment of samples to the analytical laboratory. Figure 5-4 is
a representative copy of 'the chain-of-custody form used during
the project.
Specific QA procedures used to ensure the accuracy and
precision of the air sample collection and TEM analyses included
the collection of production lot and field blanks, field audits,
laboratory audits, replicate and external analyses, and a study
to further evaluate the results obtained by the TEM method.
Production lot blanks are filters chosen prior to the
start of field work. They are analyzed by the analytical
laboratory to check for filter contamination. Field blanks are
filters taken into the field and handled in the same manner as
exposed air sample filters. Their purpose is to check for
contamination which might occur in the field but not as a result
of air sampling. Field audits determine whether the field team
is following set procedures. Laboratory audits determine the
same for the analytical laboratory personnel. Replicate and
external analyses serve the same purpose as discussed in
Section 7 for PLM analyses.
8.3.1 Production Lot Blanks
Blank filters from prescreened production lots were
randomly selected three times during the project: at the
beginning of field activities, in the middle, and near
completion. Each time, two filter cassettes were randomly
selected from a previously unopened box of 50 filters. A total
of 26 production lot blanks were selected in this way for
analysis. The analysis of the production lot blanks indicated
that there was not a problem with background filter
contamination.
8.3.2 Field Blanks
During the pump set-up, preloaded filter cassettes were
selected as field blanks. These filters were labeled and handled
in an identical manner as were the sample filters, except that
they were not attached to the sampling pump. The filters were
capped during active sampling periods and open faced during the
non-run hours when the actual sample cassettes were also open
faced. Field blanks were collected in 30 of the buildings
sampled. The purpose of the field blanks was to measure
55
-------
contamination which might occur during periods when the pumps
were not running.
Of the 30 field blanks collected, 19 were selected for
analysis. If a high level of contamination was found from the
analysis results of the 19 blanks, the remaining 11 blanks would
have been analyzed. The 19 blanks that were analyzed were chosen
at random from the 30 blanks collected. No structures were
detected in 18 of the 19 field blanks that were analyzed. A
single fiber was counted on the remaining blank. This level of
blank contamination corresponds to an airborne asbestos structure
concentration of 0.00015 s/cc when 5,000 L of air is collected, a
very low level of contamination. Thus, it was not necessary to
analyze the remaining blanks.
8.3.3 Flow Rate Calibration
All data collected in the field were transcribed from
the field data sheets onto a computer disk. Flow readings were
corrected to standard temperature and pressure (STP) via internal
calculations built into the computer spreadsheet.
Since the flow rate was controlled by limiting orifices
and no adjustment could have been made to the diaphragm vacuum
pump, the equipment limited to calibration were the rotameters,
barometric gauges, and thermometers. Two rotameters of differing
capacity were used to measure the flow rate under field
conditions. A 0 to 5 Lpm rotameter was used to monitor the
2.5 Lpm limiting orifice side, and a 0 to 20 Lpm rotameter was
used on the 5 Lpm orifice side. The procedure to calibrate
rotameters to STP used a bubble tube as a secondary standard.
The procedure is described in USEPA (1977).
Using STP, a calibration curve was developed for each
rotameter. Upon return from the field, the recordings made in
the field were compared to the calibration curve, and a STP flow
was achieved. The STP flow was then recalculated using the
computer to finalize the flow-to-field operating conditions. A
random set of final flow volumes was recalculated as a
confirmation check. Final flows, their matching random I.D.
numbers and locations were tabulated and matched by a computer
with the TEM results from the electron microscopy laboratory.
8.3.4 Field Audits
Five field audits were conducted by an independent
field auditor, one audit in each of the study regions. The field
auditor accompanied the field crew during pump set-up in several
buildings*per study region. He checked to be sure that the field
crew was following the guidelines set forth in the Quality
56
-------
Assurance Plan (Hatfield et al. 1987), and documented any
violations in procedures so they could be corrected. For
example, an air hose on one pump was found to be punctured. This
was noted and immediately corrected. The field auditor also
measured 216 flow rates in pumps in these buildings. This was
done in order to estimate the relative accuracy of the flow
rates, defined as [(field value-audit value)/(audit value)] x
100. The percentage of flow rates within + 20% relative accuracy
was 99%.
8.3.5 Laboratory Audits
To ensure the accuracy of the air sample analyses using
TEM, two laboratory audits were performed. An independent
laboratory auditor visited the TEM analytical laboratory to
verify that all procedures specified in the Quality Assurance
Plan (Hatfield et al. 1987) were followed. Two audits were
conducted, one at the beginning of the analyses and one at the
end. The auditor concluded that:
The sample identification was traceable from
sample acceptance through preparation, analysis
and reporting.
The sample preparation was done according to
protocol, with the exception of the use of
scissors instead of a scalpel for the cutting of
filters. .
The TEM analysis was done according to protocol.
The reporting procedure was implemented properly
and was accurate.
8.3.6 Replicate and External Analyses
Twenty air samples, four from each study region, were
selected at random to investigate within and between laboratory
performance. The samples were reanalyzed by the original
laboratory (replicate analysis) and by a second laboratory
(external analysis). These 40 QC analyses increase the total
number of analyses by just over 10%. Within each study region,
three samples were selected at random from sites within Category
2 and 3 buildings. The fourth sample was selected from the
outdoor and Category I sites. The samples were receded to avoid
analyst bias in the replicate analysis.
Table 8-2 presents the results of the original,
replicate, and external analyses. No asbestos structures were
57
-------
Table 8-2.
Comparison of Airborne Asbestos Concentrations
Estimated by the Original, Replicate (Same
Laboratory), and External (Different Laboratory) TEM
Analysis
Original
Number of
structures
0
1
0
0
1
0
0
0
I
0
0
0
0
0
0
0
0
0
0
0
s/cc
0.000
0.001
0.000
0.000
0.003
0.000
0.000
0.000
0.003
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Replicate
Number of
structures
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
0
0
0
s/cc
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.003
0.003
0.000
0.000
0.000
0.000
0.000
0.002
0.003
0.000
0.000
0.000
External
Number of
structures
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
s/cc
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.000
58
-------
detected on 13 of the 20 filters. One or more of the three
analyses detected a single structure on the remaining seven
filters.
There is no evidence of inconsistency among the three
sets of analyses. A Wilcoxon signed-rank test (Sokal and Rohlf
1969) did not detect any significant tendency for any one
analysis (original, replicate, or external) to give higher or
lower structure counts than any other.
A confidence interval for the mean structure count can
be constructed by assuming a statistical distribution for the
counts. A Poisson distribution gives a 95% confidence interval
of (0, 3.0) when zero structures are counted. Therefore, an
observation of zero structures is not inconsistent with a mean of
3 structures, indicating that the differences of a single
structure between analyses are not significant. This conclusion
is even stronger if structure counts follow a negative binomial
distribution. The confidence interval for a negative binomial is
wider than the corresponding interval for a Poisson distribution.
8.3.7 Examination of Additional Grid Openings
Sixteen air samples were selected for additional
analysis to determine if the 10 grid openings specified by the
TEM protocol provide estimates of sufficient precision for the
purposes of the study. The 16 samples were selected as follows
to provide a range of structure counts:
4 "indoor" samples which had structure counts of 3
or more in the first 10 grid openings counted;
8 "indoor" samples which had structure counts of 0
in the first 10 grid openings;
2 "outdoor" samples; and
2 field blanks.
An additional 40 grid openings, giving a total of 50,
were examined on each sample and the number of structures in each
opening recorded.
The precision of the TEM analysis was investigated by
fitting a negative binomial distribution to the number of
asbestos structures per grid opening. The negative binomial is a
discrete distribution which is often used to describe clumped or
aggregated populations. Javitz and Fowler (1981) found that the
negative binomial was superior to the Poisson for describing
asbestos structure counts obtained by electron microscopy. The
variance of the negative binomial is m (m + k)/k where m is the
59
-------
mean and k is a measure of aggregation. As k increases, the
variance decreases and, consequently, the precision of estimated
airborne asbestos concentrations increases. The Poisson
distribution is a limiting case of the negative binomial as k
becomes very large.
The parameter k was estimated for each filter with a
non-zero structure count. Given k, the precision of the
structure count can be determined as a function of the number of
grid openings counted. (Details are given in Appendix F.)
Estimates of the mean number of structures per grid
opening (m), the variance, and k for each filter are given in
Table 8-3. Estimates of k equal to infinity indicate that the
variance does not exceed the mean and that a Poisson or more
uniform distribution is more appropriate than the negative
binomial. This implies a small variance and hence increased
precision.
No asbestos structures were counted on eight of the 16
filters. The eight include the two field blanks and the two
outdoor samples. Of the eight filters with non-zero counts, five
have estimates of k equal to infinity. The remaining three
estimates of k are 0.6, 0.4, and 0.07.
For k equal to infinity, i.e., a Poisson distribution,
a 95% confidence interval for the true structure count when no
structures are counted in 10 grid openings is (0, 3.0). The size
of the confidence interval increases slightly to (0, 3.1) as k
decreases to 0.4. Thus, for values of k greater than or equal to
0.4 the examination of 10 grid openings in this study yields an
airborne asbestos concentration that is sufficiently precise to
distinguish 0 s/cc from 0.009 s/cc with high probability. (In
this study one structure corresponds to approximately
0.003 s/cc.)
The data in Table 8-3 indicate that k is usually
greater than 0.4, but that smaller values, such as k = 0.07 are
possible. The standard deviation of this estimate of k is 0.9.
For k = 0.07, a 95% confidence interval for the true structure
count when no structures are counted in 10 grid openings is
(0, 50). If the number of grid openings counted is increased to
50, the confidence interval shrinks to (0, 4.7).
Of the 16 filters examined, all but one indicate that
examination of 10 grid openings is sufficient to distinguish
0 s/cc from 0.009 s/cc with high probability. Although, without
additional data, it is difficult to predict how frequently
exceptions will occur, the results suggest that examination of
additional grid openings is generally unnecessary unless higher
precision*is required.
60
-------
Table 8-3.
Estimated Mean, Variance, and Value of k for the
Number of Structures Counted Per Grid Opening
Based on Examination of 50 Openings
Type of filter
Openings
1-10
Mean
Mean
Openings
1-50
Var
Indoor with 3
or more structures
in openings
1-10
Indoor with
0 structures in
openings 1-10
Outdoor
Field blank
1.1
0.3
0.4
0.3
0
0
0
0
0
0
0
0
0
0
0
0
0.36
0.18
0.12
0.1
0.02
0.02
0
0
0
0
0.02
0.04
0
0
0
0
0.52
0.27
0.11
0.21
0.02
0.02
0
0
0
0
0.02
0.04
0
0
0
0
0.60
0.41
00
0.07
00
00
-
00
00
-
^
8.4 ANALYSIS OF AIR MONITORING DATA
8.4.1 Methods
For each of the 387 air samples collected and analyzed,
an estimate of airborne asbestos concentration, c, in structures
per cubic centimeter (s/cc) is given by:
where
c = (ns * A/a)/V
ns - the number of structures counted in the
microscope;
A = the effective area of the filter;
61
-------
a = the area of filter examined; and
V = the volume of air collected (in cubic centimeters).
The area of filter examined is calculated by multiplying the
number of grid openings by the area of one grid opening. The
analytical sensitivity is the smallest value, other than zero,
that c can take. It corresponds to the observation of a single
structure and depends on the values of A, a and V. In this
study, the analytical sensitivity for most samples is
approximately 0.003 s/cc.
Individual buildings are the basic statistical units
for comparisons between different building categories and between
indoor and outdoor levels. The airborne asbestos level in each
building was estimated by the arithmetic mean of the samples.
The distribution of these building averages was plotted and
summary statistics (percentiles, mean, standard deviation)
calculated for each building category and for the outdoor
measurements.
Differences of distributions of airborne asbestos
levels between building categories are indicated by the plots and
tables of summary statistics. A statistical test was applied to
provide a quantitative measure of the strength of evidence
associated with observed differences (i.e., probabilities that
the observed differences may have occurred only by chance were
estimated). A "p-value," the level of significance, is reported
for each comparison. The p-value is the probability of obtaining
a difference as great or greater than the difference observed
under the hypothesis that no true difference exists between the
categories being compared. A small p-value indicates that the
magnitude of the observed difference is unlikely under the
hypothesis of no true difference, and therefore lends support to
the alternative hypothesis, namely that the difference is real.
A permutation (also referred to as randomization) test
(Cox and Hinckley 1974, Section 6.2) was used to test for
differences between building categories and outdoors. (See also
Edgington 1987 for a discussion of randomization tests.) Medians
were chosen to represent the distributions of airborne asbestos
levels. The median is appropriate for summarizing the location
of this type of data because it gives equal weight to large and
small data values and is not unduly influenced by a small number
of extreme values. The permutation approach, rather than
analysis of variance or Student's t-test, was used because the
data contained a large number of zero observations. In previous
air monitoring studies, where the majority of measured airborne
asbestos concentrations were greater than zero, a log
transformation was used to equalize variances prior to analysis
by standard analysis of variance techniques (USEPA 1985b, 1986b;
Tuckfield et al. 1987).
62
-------
Under the null hypothesis of the permutation test, each
of the 387 airborne asbestos measurements is an independent
observation from the same probability distribution. Therefore,
every possible permutation of the 387 values is equally likely.
The distribution of any given statistic (e.g., the difference
between two building category medians) under the null hypothesis
can be determined by calculating the value of the statistic for
each possible permutation and tabulating the frequency of
occurrence of each value. The enumeration of every possible
permutation is impractical for large data sets. Instead, the
distribution of the statistic is based on a random sample of all
possible permutations. The precision of any estimated percentile
is determined by the size of the random sample.
A random sample of 1,000 permutations of the 387
airborne asbestos measurements was used to estimate the
distribution of the differences between building category and
outdoor medians under the null hypothesis that all airborne
asbestos measurements are independent observations from the same
probability distribution. For each permutation, the difference
between medians was calculated in the same way as for the
original data, i.e., measurements were averaged within a
building, the median of each building category and outdoors was
determined, and the difference between medians calculated. With
1,000 replications, the width of an approximate confidence
interval for the 95th percentile is 0.02. The p-value for each
observed difference is read directly from the estimated
distribution. For example, if an observed difference
corresponded to the 97th percentile of the distribution estimated
under the null hypothesis, the p-value would be 0.03.
The p-values obtained in this way do not take into
account the fact that several comparisons are being made
simultaneously. Although the reported p-values may be smaller
than a p-value obtained from the joint probability distribution
governing all comparisons simultaneously, the reported values
indicate which observed differences are most consistent with the
null hypothesis of no true difference, and which provide support
for the alternative hypothesis that a true difference exists.
The permutation method can also be used to compare
different building categories using indoor-outdoor differences.
This can be thought of as adjusting each indoor measurement for
outdoor levels by subtracting the outdoor measurement. A test
using adjusted data will be more powerful if the adjustment
reduces the variance of the data. As will be seen below with
this particular data set, all but 7 of the outdoor values were
zero. For most of the data points, adjusting for outdoor levels
has no effect. For the 7 buildings with non-zero outdoor levels
(one structure observed), the adjustment results in a negative
building mean because none of these buildings has an average
63
-------
indoor level as high as one structure per sample. The adjusted
data have a higher variance than the unadjusted data. For this
data set, adjusting for outdoor levels conveys no advantage and
results in a less powerful statistical test. Therefore, this
approach was not used.
8.4.2 Results
The estimated airborne asbestos concentration for each
sample is given in Appendix 6. No asbestos structures were
detected in 83% of the 387 samples. For one sample from Building
23, the TEM result, a zero structure count, was reported after
the statistical analysis was completed. This 388th sample is not
included in any of the following tables or analyses. The maximum
number of structures counted on a single sample was 11,
corresponding to an airborne asbestos concentration at that site
of 0.033 s/cc. This sample was collected in Building 2, which
was categorized as containing no ACM (Category 1). No asbestos
structures were observed on five of the remaining six filters
collected in the building; the remaining filter had one
structure. Neither GSA, nor the building inspector for this
study, identified ACM despite thorough inspections of the
building and analyses of bulk samples. The reason for this
unusually high value is unknown. (The next highest airborne
asbestos concentration at a single site is 0.013 s/cc.) The
source of asbestos structures could not be found.
Figure 8-1 presents scatter plots of the average
airborne asbestos concentration in each building, by building
category, and for the 48 individual outdoor samples. The medians
of each category are also shown. With the exception of the one
Category 1 building, the highest average airborne asbestos
concentrations occur in buildings from Category 3. However, in
27% of the buildings in Category 3, no asbestos structures were
detected at the indoor sites. (No asbestos structures were
detected in 16% of the Category 2 buildings, 50% of the Category
1 buildings, and 85% of the outdoor sites.)
The medians and arithmetic means of the average
airborne asbestos concentrations for each building category and
outdoor samples are reported in Table 8-4. Outdoor samples have
the smallest median, followed by buildings from Category 1,
buildings from Category 2, and finally, buildings from
Category 3. The arithmetic mean of Category 1 is greatly
influenced by the one unusually high value in that category,
although the absolute magnitude of the mean is still very small.
The other means follow the same trend as the medians. The
outdoor means by building category, 0.00043, 0.00048, and 0.00036
s/cc for% Categories 1, 2, and 3, respectively, do not show any
apparent trend. Therefore, the trend in indoor means with
64
-------
Ui
o.oos
s
T
R
U
0.004
T
U
R
E
0.003
C
0
N
C
E 0002
N
T
R
A
T
j 0001
0
N
8/ec
0.000
A
A
A
A
fl
Z o
1 1
Outdoor
(48 sites)
A
A
ft *
A
A
A
A
A
A
B
e
B
Building Building Building
category 1 category 2 category 3
(6 buildings) (6 buildings) (37 buildings)
Figure 8-1. Scatter plots*and medians of the average airborne
asbestos structure concentrations for each building
category and outdoors.
*The data points for each scatter plot are the average concentration
within a building (for indoor samples) or the concentration outside each
building (for outdoor samples). A=l data point, B=2 data points, ...,
j=10 data points, and Z=41 data points. The diamond represents the median
of the data points in each scatter plot.
-------
Table 8-4. Summary Statistics for Average Airborne Asbestos
Structure Concentrations (s/cc)
Statistic
Standard
deviation
ACM
Outdoor
Category 1
Category 2
0.00096
0.00198
0.00052
Category 3
Median
Mean
Sample
size
<0. 00001
0.00039
48
(sites)
0.00010
0.00099
6
(buildings)
0.00040
0.00059
6
(buildings)
0.00058
0.00073
37
(buildings)
0.00072
Notes:
1. The data points used in the calculation of each
statistic are the average concentration within a building (for
indoor samples) or the concentration outside each building (for
outdoor samples).
2. The mean for Category 1 is heavily influenced by one
sample in one building which produced an unexplained large s/cc
value. The Category 1 mean, excluding this one value, is
0.00020 s/cc.
building category cannot be attributed to differences in outdoor
airborne asbestos concentrations.
The analytical sensitivity of a given mean is
approximately the analytical sensitivity of a single sample divided
by the number of air samples used to calculate that mean. Thus,
the analytical sensitivity of a building mean is approximately
0.0004 s/cc (0.003/7). The analytical sensitivity of the Category
3 mean is approximately 0.00001 s/cc (0.003/255). Similarly, the
analytical sensitivity for Category 1 and 2 means is approximately
0.00007 s/cc (0.003/42). For the outdoor mean the analytical
sensitivity is approximately 0.00006 s/cc (0.003/48).
The results of the permutation test are listed in
Table 8-5,. The difference between Category 3 buildings and
Category 1 buildings has the smallest p-value (p < 0.02). The next
smallest p-value is obtained for the comparison between buildings
66
-------
Table 8-5. Results of Randomization Test Indicating the
Statistical P-Values for Differences between Median
Airborne Asbestos Concentrations in Each of the Three
Building Categories and Outdoor Concentrations
Difference between
Comparison medians (s/cc) P-valuea
Category 1 versus 0.00010 p < 0.96
outdoor
Category 2 versus 0.00040 p < 0.65
outdoor
Category 3 versus 0.00058 p < 0.09b
outdoor
Category 2 versus 0.00030 p < 0.21
Category 1
Category 3 versus 0.00048 p < 0.02
Category 1
Category 3 versus 0.00018 p < 0.18
Category 2
Probability of obtaining a difference as great or
greater than the difference observed when there are no real
differences among building categories or between a building
category and the outdoor air.
This p-value is based on 2,000 replications to provide
additional precision.
in Category 3 and outdoors (p < 0.09). The p-values for the
remaining comparisons are 0.18 or greater. Estimates of indoor
asbestos levels are more precise than estimates of outdoor levels
because indoor levels are based on several samples per building.
Thus, an observed difference between two building category
medians corresponds to a smaller p-value than the same observed
difference between a building category median and the outdoor
median.
67
-------
REFERENCES
Bishop, YMM, Fienberg SB, Holland PW. 1980. Discrete
raultivariate analysis: Theory and practice. Cambridge, MA: MIT
Press.
Cox DR, Hinckley DV. 1974. Theoretical Statistics. London:
Chapman and Hall.
Deming WE. 1950. Some Theory of Sampling. New York: John
Wiley and Sons.
Edgington, ES. 1987. Randomization Tests (2nd edition). New
York: Marcel Dekker, Inc.
Javitz HS, Fowler DP. 1981. Statistical analysis of microscopic
counting data, in "Electron Microscopy and X-ray Applications,"
Russell PA (ed.), Ann Arbor Science.
Hatfield J, Leczynski B, Chesson J et al. 1987. Battelle
Columbus Division. Public buildings study quality assurance
plan. Final report. Washington, DC: Office of Toxic
Substances, U.S. Environmental Protection Agency. Contract No.
68-02-4243.
Miller, R6. 1981. Simultaneous statistical inference (2nd
edition). New York: Springer-Verlag.
Rogers, J. 1987. Westat, Inc. Additional analysis of data
collected in the asbestos in buildings survey. Draft final
report. Washington, D.C.: Office of Toxic Substances, U.S.
Environmental Protection Agency. Contract No. 68-02-4243.
Sokal, RR, Rohlf, FJ. 1969. Biometry. San Francisco: W.H.
Freeman.
Tuckfield RC, Chesson J, Tsay Y-L, et al. 1987. Battelle
Columbus Division. Evaluation of asbestos abatement techniques
phase III: removal. Draft final report. Washington, D.C.:
Office of Toxic Substances, U.S. Environmental Protection Agency.
Contract No. 68-02-4243.
USEPA. 1977. U.S. Environmental Protection Agency. Quality
assurance handbook for air pollution measurement systems, volume
II - ambient air specific methods. Washington, DC: Office of
Toxic Substances, U.S. Environmental Protection Agency. EPA
600/4-77-027a.
USEPA. 1981. U.S. Environmental Protection Agency. Asbestos in
schools.. Washington, DC: Office of Toxic Substances, U.S.
Environmental Protection Agency. EPA 560/5-81^002.
68
-------
USEPA. 1982. U.S. Environmental Protection Agency. Friable
asbestos-containing materials in schools: identification and
modifications. Washington, DC: Office of Toxic Substances, U.S.
Enviornmental Protection Agency. 40 CFR Part 763.
USEPA. 1985a. U.S. Environmental Protection Agency. Asbestos
in buildings: simplified sampling scheme for friable surfacing
materials. Washington, DC: Office of Toxic Substances, U.S.
Environmental Protection Agency. EPA 560/5-85-030a.
USEPA. 1985b. U.S. Environmental Protection Agency. Evaluation
of asbestos abatement techniques phase I: removal. Washington,
DC: Office of Toxic Substances, U.S. Enviornmental Protection
Agency. EPA 560/5-85-019.
USEPA. 1986a. U.S. Environmental Protection Agency. Guidance
for assessing and managing exposure to asbestos in buildings.
Draft Report. Washington, D.C.: Office of Toxic Substances,
U.S. Environmental Protection Agency. Contract No. 68-02-4243.
USEPA. 1986b. U.S. Environmental Protection Agency. Evaluation
of asbestos abatement techniques phase II: encapsulation with
latex paint. Washington, DC: office of Toxic substances, U.S.
Environmental Protection Agency. EPA 560/5-86-016.
Yamate G, Agarwal SC, Gibbons RD. 1984. Methodology for the
measurement of airborne asbestos by electron microscopy. Draft
report. Washington, DC: Office of Toxic Substances, U.S.
Environmental Protection Agency. Contract No. 68-02-3266.
69
-------
APPENDIX A
RESPONSES OF INDIVIDUAL RATERS IN EACH ASSESSED AREA
WITHIN EACH REGION TO CONDITION, POTENTIAL FOR DISTURBANCE,
AND AIR FLOW FACTORS
-------
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING. AND AREA
1=GOOD 2=MODERATE DAMAGE 3=SIGNIFICANT DAMAGE
REGIONe1
BUILDING
13
13
13
13
13
14
14
14
14
14
IE
16
IS
IE
16
16
16
16
16
17
' 17
17
17
17
Be
50
60
50
61
61
61
61
62
62
52
62
53
53
63
53
64
54
54
54
54
55
56
55
56
56
56
AREA
1
2
4
6
7
1
2
3
4
7
1
2
3
4
1
3
4
7
8
1
2
3
4
5
1
2 .
3
4
1
2
3
4
1
2
3
4
2
3
4
8
1
3
4
5
8
1
2
3
4
1
2
CORE RATER
ONE
1
3
1
3
3
3
1
1
2
3
2
2
3
3
2
3
2
3
2
2
2
2
2
1
2
2
2
2
2
2
3
1
2
2
3
2
2
2
2
2
2
3
2
2
3
1
2
2
3
2
2
CORE RATER
TWO
1
2
1
3
3
3
1
1
2
2
3
2
2
3
2
2
3
2
2
2
2
1
3
2
2
2
2
3
3
1
3
2
2
2
t
3
1
2
2
3
t
1
2
3
3
.
t
LOCAL RATER
ONE
2
3
2
3
3
3
1
2
2
3
3
3
3
3
2
2
3
2
3
2
2
,
2
2
2
2
2
3
3
1
3
3
3
1
3
3
2
2
2
3
3
3
;
,
.
,
LOCAL RATER
TWO
2
3
2
3
3
3
1
1
3
3 '
3
2
3
2
2
3
3
3
3
3
3
3
3
3
1
3
3
3
1
3
3
2
3
3
3
3
3
i
2
3
3
3
3
-------
BUILDING
0
56
66
67
67
67
67
7
7
7
7
7
7
8
8
8
8
AREA
3
4
1
2
4
6
1
2
3
4
7
8
1
2
3
4
CORE RATER
ONE
2
2
2
3
1
2
2
2
1
3
3
3
2
1
2
2
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
IsGOOD 2eMODERATE DAMAGE ^SIGNIFICANT DAMAGE
REGIONS1
ISl
CORE
TWO
RATER
2
2
3
3
1
1
1
2
1
3
3
3
2
1
1
1
LOCAL RATER
ONE
,
3
2
2
2
2
2
1
3
.
.
2
2
2
2
LOCAL RATER
TWO
3
3
,
.
,
.
.
.
.
.
.
.
.
.
.
.
-------
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
1=GOOD 2=MODERATE DAMAGE 3=SIGNIFICANT DAMAGE
BUILDING
18
18
16
18
18
19
19
19
19
19
19
19
20
20
20
20
20
If
15
21
21
22
22
22
22
23
23
23
23
24
24
24
24
25
25
25
25
9
9
9
9
AREA
1
2
3
4
8
1
2
3
4
7
8
9
1
2
3
4
5
8
]
2
3
4
1
2
3
4
1
2
3
4
1
2
4
5
1
2
3
4
1
2
3
4
CORE RATER
ONE
1
2
2
1
2
1
3
3
2
.
.
.
2
2
1
2
2
3
2
2
2
3
1
2
1
2
2
3
2
2
3
2
1
2
1
2
2
1
2
2
CORE RATER
TWO
1
3
2
1
3
2
3
3
2
3
3
2
3
3
2
2
2
3
1
2
1
2
2
3
3
(
3
2
1
(
2
3
2
1
2
3
LOCAL RATER
ONE
3
2
2
3
1
3
3
2
3
3
1
2
2
3
3
2
3
3
1
2
1
3
2
3
3
3
3
2
1
2
1
3
2
1
.
3
LOCAL RATER
TWO
1
2
2
1
2
2
3
2
2
3
3
1
2
i
3
2
2
2
3
1
2
1
2
2
3
3
3
3
2
1
2
1
2
1
1
2
3
-------
BUILDING
26
26
26
26
27
27
27
27
27
28
28
28
28
28
28
29
29
29
^ 29
II
30
30
30
30
31
31
31
31
32
32
32
32
32
33
33
33
33
34
34
34
34
35
35
35
35
58
58
58
58
AREA
1
2
3
4
1
2
3
4
8
1
2
3
4
5
6
1
2
3
4
1
3
4
6
8
1
2
3
4
1
2
4
6
9
1
4
5
9
1
2
3
4
1
2
3
4
1
2
3
4
CORE RATER
ONE
3
2
2
2
3
2
2
3
2
2
2
1
1
.
2
2
2
2
2
3
2
2
2
3
2
2
2
3
2
2
2
3
3
2
3
2
3
2
2
2
2
1
2
2
2
2
3
2
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
IsGOOD 2=MODERATE DAMAGE 3=SIGNIFICANT DAMAGE
REGION=3
CORE RATER
TWO
2
2
2
2
3
1
2
3
3
2
2
2
2
2
2
2
1
3
3
2
2
3
3
2
3
3
3
1
2
2
3
3
2
3
2
3
1
2
2
1
1
2
3
2
3
3
2
LOCAL RATER
ONE
2
2
2
1
2
3
2
3
2
1
1
1
1
3
2
2
2
2
,
2
3
2
2
2
2
3
3
2
2
1
2
2
2
2
2
2
1
3
2
1
1
2
2
2
3
3
2
LOCAL RATER
TWO
.
.
,
3
1
2
3
2
3
3
2
1
3
3
3
2
3
.
2
3
3
3
3
3
3
2
1
.
3
3
3
3
3
-------
BUILDING
10
10
10
10
36
36
36
36
36
37
37
37
37
37
38
38
38
38
-J 39
-J 39
39
39
39
40
40
40
40
40
40
41
41
41
41
41
41
41
41
42
42
42
42
59
59
59
59
59
60
60
60
60
61
AREA
1
2
3
6
1
2
3
4
7
1
2
3
5
6
1
2
3
4
1
2
3
4
6
1
2
3
4
7
13
1
2
3
4
5
6
7
8
1
2
3
4
1
2
3
4
6
1
2
3
4
2
CORE RATER
ONE
1
1
1
1
1
2
2
2
3
3
2
1
1
2
1
2
2
1
2
2
1
1
2
2
1
1
1
1
2
2
3
2
2
2
1
2
2
1
2
2
2
2
2
1
2
2
2
1
1
2
1
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
1=GOOD 2=MODERATE DAMAGE SeSIGNIFICANT DAMAGE
REGION=4
CORE RATER
TWO
1
1
1
1
1
2
2
2
3
3
2
1
2
2
2
2
1
2
2
3
1
2
2
2
1
1
1
2
2
3
2
2
2
1
2
2
1
2
2
2
1
2
1
1
2
2
1
2
2
LOCAL RATER
ONE
1
2
2
2
2
2
2
2
.
3
2
2
3
2
2
,
2
2
3
1
2
2
2
1
1
1
2
3
2
*
2
2
2
2
1
3
2
2
2
2
2
2
3
2
2
2
2
LOCAL RATER
TWO
1
1
1
1
2
3
2
2
2
3
.
2
2
1
2
1
2
2
3
1
1
2
2
1
1
1
3
3
3
2
3
2
2
3
2
1
3
3
1
3
1
1
3
1
2
2
1
-------
TABLE A-l. RESPONSE OP RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
IsGOOD 2=MODERATE DAMAGE 3=SIGNIFICANT DAMAGE
REGION**
BUILDING AREA CORE RATER CORE RATER LOCAL RATER LOCAL RATER
ONE TWO ONE TWO
61 31221
61 42222
61 61121
00
-------
TABLE A-I. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
1=GOOD 2&MODERATE DAMAGE ^SIGNIFICANT DAMAGE
REGION=6
sO
BUILDING
11
11
11
11
12
12
12
43
A3
43
43
44
44
44
44
46
46
46
46
46
46
46
46
46
47
47
47
47
48
48
48
48
49
49
49
49
62
62
62
63
63
63
63
64
64
64
64
66
66
66
66
AREA
1
2
3
4
1
2
3
1
2
3
6
1
2
3
4
2
3
4
7
2
2
4
6
6
1
2
3
4
1
2
3
4
1
2
3
4
1
7
8
1
2
3
8
2
3
4
7
1
2
6
6
CORE RATER
ONE
1
1
1
2
2
1
1
2
3
1
1
2
3
2
2
2
2
1
2
1
1
2
2
2
2
2
2
3
2
2
3
2
2
1
3
2
2
1
2
2
2
2
2
2
1
2
1
2
2
2
2
CORE RATER
TWO
1
1
1
2
2
1
1
3
2
1
1
2
3
2
2
1
2
2
2
2
2
2
2
2
2
2
1
3
2
2
3
2
2
1
3
3
2
2
1
2
2
2
2
2
1
2
1
2
2
2
1
LOCAL RATER
ONE
1
1
1
2
2
1
2
2
2
1
2
2
2
2
1
1
1
2
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
2
1
2
2
1
2
1
2
2
2
1
1
1
1
1
1
1
LOCAL RATER
TWO
1
2
1
1
,
,
2
3
1
2
.
,
,
,
3
3
2
2
2
2
2
3
3
3
2
3
3
3
2
3
2
3
1
3
3
3
3
2
3
2
3
3
,
,
.
,
2
2
2
1
-------
TABLE A-l. RESPONSE OF RATERS TO OVERALL CONDITION VARIABLE
SEPARATED BY REGION, BUILDING, AND AREA
1=GOOD 2=MODERATE DAMAGE 3=SIGNIFICANT DAMAGE
REGIONsB
BUILDING AREA CORE RATER CORE RATER LOCAL RATER LOCAL RATER
ONE TWO ONE TWO
66 11112
66 21112
66 31112
66 42112
66 51112
66 61112
CX)
O
-------
BUILDING
13
13
13
13
13
14
14
14
14
14
16
IB
16
16
16
16
16
16
16
00 17
~ 1?
17
17
17
60
60
60
60
61
61
61
61
62
E2
62
62
63
63
63
63
64
54
64
64
54
55
65
55
55
66
56
AREA
1
2
4
6
7
1
2
3
4
7
1
2
3
4
1
3
4
7
8
1
2
3
4
5
1
2
3
4
1
2
3
4
1
2
3
4
2
3
4
6
1
a
4
5
8
1
2
3
4
1
2
CORE RATER
ONE
1
3
2
3
3
2
2
2
2
2
2
2
3
2
2
3
3
3
3
2
2
3
31
2
3
2
2
2
2
2
2
1
2
2
2
1
3
3
3
2
2
3
3
3
3
2
3
3
2
2
2
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
REGION=1
CORE RATER
TWO
2
2
3
2
3
2
1
1
1
3
2
3
1
3
2
2
3
2
2
2
2
2
3
2
2
2
2
3
2
2
3
2
2
2
3
3
2
2
2
3
3
3
2
3
2
2
2
2
LOCAL RATER
ONE
2
3
2
3
3
3
1
2
2
3
3
3
2
3
2
2
3
2
2
3
2
,
3
2
2
3
2
3
3
1
2
3
3
1
3
3
2
2
2
3
3
3
f
t
,
,
LOCAL RATER
TWO
3
3
3
3
3
3
1
)
3
.
.
.
3
3
3
3
3
3
3
3
,
3
3
3
3
3
3
3
3
3
3
3
2
3
3
3
3
3
3
3
3
1
2
3
3
3
3
3
-------
BUILDING
» 56
66
67
67
67
67
7
7
7
7
7
7
8
8
8
8
AREA
3
A
I
2
4
6
1
2
3
4
7
8
1
2
3
4
CORE RATER
ONE
2
2
2
2
3
2
2
3
2
3
3
3
2
2
2
2'
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
REGIONsl
CORE
TWO
RATER
2
3
3
3
2
2
2
3
2
3
3
3
2
2
2
2
LOCAL RATER
ONE
.
3
2
2
2
2
3
2
3
.
,
2
2
2
2
LOCAL RATER
TWO
3
3
,
.
.
.
,
.
,
.
.
.
.
,
.
ffi
-------
BUILDING
18
18
18
18
18
19
19
19
19
19
19
19
20
20
20
20
20
20
21
~r 21
21
21
22
22
22
22
23
23
23
23
24
24
24
24
26
2E
26
26
9
9
9
9
AREA
1
2
3
4
8
1
2
3
4
7
8
9
1
2
3
4
6
8
1
2
3
4
1
2
3
4
1
2
3
4
1
2
4
6
1
2
3
4
1
2
3
4
CORE RATER
ONE
2
2
1
2
2
1
3
2
2
,
,
,
2
2
3
2
.
.
2
3
3
3
2
3
1
1
2
1
3
3
3
3
3
3
2
1
1
2
3
2
2
2
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
REGION=2
CORE RATER
TWO
2
3
2
3
3
1
3
2
2
3
3
3
2
2
3
2
3
2
3
1
1
2
1
3
3
2
3
2
2
2
2
3
2
3
2
3
2
LOCAL RATER
ONE
3
3
2
3
3
1
3
3
3
3
2
3
3
3
3
3
2
3
2
)
2
2
2
3
3
3
3
3
3
1
1
1
1
2
3
3
3
LOCAL RATER
TWO
2
2
1
1
2
1
2
2
2
3
2
3
3
2
2
2
2
2
2
1
1
1
1
3
3
2
2
2
2
2
1
2
2
3
1
3
3
-------
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
REGION&3
BUILDING
26
26
26
26
27
27
27
27
27
28
28
28
28
28
28
29
29
29
29
30
30
30
30
30
31
31
31
31
32
32
32
32
32
33
33
33
33
34
34
34
34
36
35
35
36
58
58
58
58
AREA
1
2
3
4
1
2
3
4
8
1
2
3
4
6
6
1
2
3
4
1
3
4
6
8
1
2
3
4
1
2
4
6
9
1
4
5
9
1
2
3
4
1
2
3
4
1
2
3
4
CORE RATER
ONE
2
2
2
1
1
3
2
2
2
2
2
2
2
2
2
2
3
2
2
2
2
2
3
2
2
2
2
2
3
3
2
2
2
2
2
3
2
2
2
2
2
2
2
2
2
IB a
CORE RATER
TWO
3
3
2
2
2
1
2
3
2
2
3
1
1
2
2
2
2
3
3
2
2
2
3
2
3
2
3
2
2
2
3
2
2
2
2
2
2
3
2
2
2
2
3
.
2
3
3
LOCAL RATER
ONE
3
3
1
1
2
1
1
2
2
1
2
2
3
2
1
2
1
2
.
1
2
1
2
2
2
2
3
1
1
1
3
2
2
2
2
2
3
3
3
3
2
1
2
1
1
2
2
LOCAL RATER
TWO
.
.
.
.
3
1
2
3
3
2
3
3
3
2
2
2
2
2
.
2
3
2
2
3
3
2
2
3
2
3
2
2
3
3
-------
BUILDING
10
10
10
10
36
36
36
36
36
37
37
37
37
37
38
36
38
38
39
00 39
in 39
39
39
40
40
40
40
40
40
41
41
41
41
41
41
41
41
42
42
42
42
59
59
59
59
59 ... - -
60
60
60
60
61
AREA
1
2
3
E
1
2
3
4
7
1
2
3
E
6
1
2
3
4
1
2
3
4
E
1
2
3
4
7
13
1
2
3
4
E
6
7
8
1
2
3
4
1
2
3
4
6
1
2
3
A
2
CORE RATER
ONE
2
3
2
2
3
2
3
3
2
2
2
2
2
1
2
1
2
1
3
3
2
2
2
2
2
1
1
1
3
2
2
2
3
3
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
1
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
REGION=4
CORE RATER
TWO
2
3
2
1
2
3
2
3
3
2
2
2
2
2
2
2
2
3
3
3
2
3
3
2
1
1
1
2
2
2
3
3
3
2
2
2
2
3
2
2
2
2
2
1
3
2
3
2
1
LOCAL RATER
ONE
3
2
3
2
2
2
3
3
3
2
2
2
2
2
3
2
2
3
3
1
2
2
3
1
2
2
3
2
3
2
3
3
2
3
3
2
3
2
2
2
2
2
2
3
2
3
2
3
LOCAL RATER
TWO
3
,
3
2
3
3
3
3
2
2
2
2
i
3
1 .
1
1
3
3
3
1
3
3
3
2
1
1
3
3
3
3
3
3
3
3
3
3
3
3
3
2
3
3
2
3
3
3
3
1
-------
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3*HIGH POTENTIAL
REGION**
BUILDING AREA CORE RATER CORE RATER LOCAL RATER LOCAL RATER
ONE TWO ONE TWO
61 31131
61 42232
61 62232
00
ON
-------
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIQH POTENTIAL
BUILDING
11
11
11
11
12
12
12
43
43
43
43
44
44
44
44
45
46
4B
46
°° 46
^ 46
46
46
46
47
47
47
47
48
48
48
48
49
49
49
49
62
62
62
63
63
63
63
64
64
64
64
66
66
65
65
AREA
1
2
3
4
1
2
3
1
2
3
5
1
2
3
4
2
3
4
7
2
2
4
6
6
1
2
3
4
1
2
3
4
1
2
3
4
1
7
8
1
2
3
8
2
3
4
7
1
2
5
6
CORE RATER
OKIE
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
1
2
2
2
3
3
1
2
3
2
2
3
2
2
2
3
2
2
2
2
1
1
2
2
2
2
REGIONsS
CORE RATER
TWO
2
2
2
3
2
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
1
2
2
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
LOCAL RATER
ONE
1
1
1
2
1
1
1
1
1
1
2
1
2
2
1
1
1
1
3
2
1
2
1
1
1
1
1
1
1
1
1
1
1
LOCAL RATER
TWO
1
1
1
2
2
2
1
1
2
3
2
3
2
2
2
2
3
2
2
3
3
2
2
2
2
3
2
3
3
2
2
3
1
2
2
2
2
2
1
-------
TABLE A-2. RESPONSE OF RATERS TO POTENTIAL FOR DISTURBANCE
SEPARATED BY REGION, BUILDING, AND AREA
1=LOW 2=MODERATE 3=HIGH POTENTIAL
BUILDING
* 66
66
66
66
66
66
AREA
1
2
3
4
B
6
CORE RATER
ONE
1
1
1
1
1
1
00
00
REGION=5
CORE RATER LOCAL RATER LOCAL RATER
TWO ONE TWO
1 1 1
1 1 1
1 1
1 1 1
1 1 1
1 1 1
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING. AND AREA
0=NO AIR FLOW 1=AIR FLOW
REGIONsl
BUILDING
13
13
13
13
13
14
14
14
14
14
16
16
16
16
16
16
16
16
~ 16
00 17
vo 1?
17
17
17
60
60
60
60
61
61
61
61
62
62
62
62
63
63
63
63
64
64
64
64
64
66
66
66
66
66
66
AREA
1
2
4
6
7
1
2
3
4
7
1
2
3
4
1
3
4
7
8
1
2
3
4
6
1
2
3
4
1
2
3
4
1
2
3
4
2
3
4
8
1
3
4
6
8
1
2
3
4
1
2
CORE RATER
ONE
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
1
0
0
1
1
1
1
1
1
3
1
0
0
0
0
0
0
0
0
1
1
0
0
1
1
1
1
0
0
0
0
1
0
e>
CORE RATER
TWO
0
0
1
0
0
0
0
0
0
0
1
0
0
1
1
1
1
1
1
0
0
0
0
0
^
^
0
0
0
,
,
1
1
1
0
1
0
0
0
0
0
0
1
0
0
LOCAL RATER
ONE
0
1
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
0
1
0
6
0
0
0
0
0
1
0
0
0
0
0
0
1
1
1
0
0
0
0
0
;
.
,
.
.
LOCAL RATER
TWO
0
1
1
0
0
0
0
0
0
;
.
.
B
0
i
i
i
0
i
i
i
*
i
i
0
i
0
0
0
1
1
1
1
0
1
1
1
0
1
1
1
1
0
0
1
1
e
e
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0eNO AIR FLOW 1=AIR FLOW
REGION*!
BUILDING
» 56
66
57
57
57
67
7
7
7
7
7
7
8
8
8
8
AREA
3
A
1
2
4
6
1
2
3
4
7
e
i
2
3
4
CORE RATER
ONE
1
0
0
0
0
0
1
0
0
0
1
0
1
0
0
0
CORE RATER LOCAL F
TWO ONE
1
0
0
0
1
0
1
0
0
1
1
1
t
0
1
0
*ATER LOCAL RATEF
TWO
:
:
0
0
i
0
0
0
1
0
.
,
0
0
0
0
{
I
vo
O
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0=NO AIR FLOW 1=AIR FLOW
REGION=2
BUILDING
18
18
18
18
18
19
19
19
19
19
19
19
20
20
20
20
20
20
21
vO 21
" 21
21
22
22
22
22
23
23
23
23
24
24
24
24
26
25
26
26
9
9
9
9
AREA
1
2
3
4
8
1
2
3
4
7
8
9
1
2
3
4
6
8
1
2
3
4
1
2
3
4
1
2
3
4
1
2
4
6
1
2
3
4
1
2
3
4
CORE RATER
ONE
1
1
0
0
0
1
0
0
0
.
,
.
1
0
1
0
.
.
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
1
1
0
1
1
1
0
0
0
CORE RATER
TWO
0
1
1
0
1
1
1
1
1
i
i
i
0
i
i
i
i
0
0
0
0
0
0
1
1
1
1
0
1
0
1
0
1
1
1
0
0
LOCAL RATER
ONE
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
LOCAL RATER
TWO
1
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0=NO AIR FLOW IsAIR FLOW
REGION=3
BUILDING
. 26
* 26
26
26
27
27
27
27
27
28
28
28
28
28
28
29
29
29
29
° 30
10 11
30
30
30
31
31
31
31
32
32
32
32
32
33
33
33
33
34
34
34
34
35
35
35
35
58
56
56
58
AREA
1
2
3
4
1
2
3
4
8
1
2
3
4
5
6
1
2
3
4
1
3
4
6
8
1
2
3
4
1
2
4
6
9
1
4
5
9
1
2
3
4
1
2
3
4
1
2
3
4
CORE RATER
ONE
1
1
0
0
0
0
0
0
0
0
0
0
1
.
,
0
1
0
0
0
1
1
0
1
0
0
1
0'
1
0
0
0
0
.
0
0
0
0
0
.
0
1
1
0
1
0
0
0
0
CORE RATER
TWO
1
1
0
1
0
1
0
1
0
0
1
0
0
0
0
0
0
0
1
1
0
1
0
0
1
0
1
0
0
0
0
0
0
0
0
1
1
0
1
1
0
0
0
0
0
0
LOCAL RATER
ONE
1
1
0
0
0
1
0
0
0
0
1
0
1
0
0
0
0
,
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
1
1
0
1
0
0
1
0
0
0
0
LOCAL RATER
TWO
.
.
.
.
0
0
0
0
0
0
0
0
1
0
0
0
0
0
.
1
e>
i
0
0
1
0
i
0
0
0
0
i
0
0
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0=NO AIR FLOW 1=AIR FLOW
REGIONS
BUILDING
10
10
10
10
36
36
36
36
36
37
37
37
37
37
38
38
38
38
39
39
vO 39
W 39
39
40
40
40
40
40
40
41
41
41
41
41
41
41
41
42
42
42
42
59
59
59
69
59
60
60
60
60
61
AREA
1
2
3
5
1
2
3
4
7
1
2
3
5
6
1
2
3
4
1
2
3
4
5
1
2
3
4
7
13
1
2
3
4
5
6
7
8
1
2
3
4
1
2
3
4
6
1
2
3
4
2
CORE RATER
ONE
0
1
0
0
0
0
0
1
0
0
1
1
0
0
0
0
0
0
1
0
0
0
0
1
1
0
.
0
0
1
1
1
1
1
0
0
0
0
1
1
1
0
0
0
1
0
0
1
0
0
0
CORE RATER
TWO
0
0
0
0
0
0
0
i
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
LOCAL RATER
ONE
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
,
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
1
1
1
1
0
0
0
1
0
0
e
e
e
LOCAL RATER
TWO
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
e
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
B=NO AIR FLOW 1=AIR FLOW
REGIONS
BUILDING AREA CORE RATER CORE RATER LOCAL RATER LOCAL RATER
ONE TWO ONE TWO
61 30000
61 41010
61 60000
\O
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0sNO AIR FLOW 1=AIR FLOW
REGIONS
BUILDING
11
11
11
11
12
12
12
43
43
43
43
44
44
44
44
46
46
46
46
vO 46
U) 46
46
46
46
47
47
47
47
48
48
48
48
49
49
49
49
62
62
62
63
63
63
63
64
64
64
64
66
66
66
66
AREA
1
2
3
4
1
2
3
1
2
3
6
1
2
3
4
2
3
4
7
2
2
4
6
6
1
2
3
4
1
2
3
4
1
2
3
4
1
7
8
1
2
3
8
2
3
4
7
1
2
6
6
CORE RATER
ONE
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
1
1
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
e
CORE RATER
TWO
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
(
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LOCAL RATER
ONE
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LOCAL RATER
TWO
0
0
0
1
,
,
f
0
0
0
0
,
,
,
,
1
1
1
J
0
0
1
0
0
1
1
1
1 .
1
1
1
.
0
0
0
0
0
1
0
0
0
0
0
.
.
.
.
0
0
0
0
-------
TABLE A-3.RESPONSE OF RATERS TO AIR FLOW
SEPARATED BY REGION, BUILDING, AND AREA
0=NO AIR FLOW 1=AIR FLOW
REGIONeB
BUILDING AREA CORE RATER CORE RATER LOCAL RATER LOCAL RATER
ONE ' TWO ONE TWO
66 10000
66 2000^
66 30000
66 40000
66 60000
66 60000
vo
ON
-------
APPENDIX B
COUNTS OF THE RESPONSES OF THE RATERS IN EACH ASSESSED
AREA WITHIN EACH REGION FOR CONDITION, POTENTIAL
FOR DISTURBANCE, AND AIR FLOW FACTORS
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING (AREA
13
14
16
17
1
2
4
6
7
1
2
CONDITION
GOOD |
COUNT |
.1
1
21
1
1
41
3 | 3|
4 -1
7
0
2
3
1
1
1
.1
4 .1
1
1
3 | -I
4 I -I
7
8
1
1
1
1
2 J
MODERATE | SIGNIFICANT
COUNT |
4
11
2|
1
1
1
1
H
3|
1
2|
11
11
11
2|
4|
1
4|
3|
COUNT
3
.
4
4
4
1
1
1
2
2
2
3
2
4
-
1
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
O
O
BUILDING |
17
50
\
u--mi,i
_ __ITJ
61 -- '"'
62
63
64
AREA
3
4
6
2
3
4
2
3
4
1
2
3
4
2 1
3
4
8
1
13
CONDITION
GOOD | MODERATE
COUNT | COUNT
1
3
2|
1 2
1 3
1 3
1 3
1 3
1 1
41
1 1
1 1
2| 2
1 1
11 3
1 3
1 3
1
SIGNIFICANT
COUNT
1
1
2
1
1
1
1
3
4
3
2
3
1
2
3
1 1
1 1
4
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
64
56
56
57
7
AREA
4
6
8
1
2
3
4
1
2
3
4
1
2
A
5
1
2
3
4
7
8
CONDITION
GOOD MODERATE
COUNT COUNT
1
1
3|
-1 3
,.:! I
1
1 1
1
-1 21
2|
1
1
2| 1|
1 2|
1 2|
1 3|
3| -1
1 -1
.1
| SIGNIFICANT
COUNT
2
1 2
1
2
3
1
1
1
1
2
2
3
2
2
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
8
AREA
1
2
3
4
GOOD
COUNT
2
1
I
CONDITION
MODERATE
COUNT
3
1
2
2
SIGNIFICANT
COUNT
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
O
OJ
BUILDING
IB
19
20
21
AREA
1
2
3
4
6
1
2
3
A
7
B
9
1
2
3
4
6
8
1
2
3
CONDITION
- GOOD 1 MODERATE
COUNT | COUNT
3
2
1 - 4
3| 1
2
2 2|
1
1
4
1
1
11
1|
3| 1|
3|
1
1
1 2|
1
31
| SIGNIFICANT
COUNT
2
2
-
A
. 3
1
1
3
3
1
1
1
'
4
1
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
BUILDING |
21 |
22
23
24
26
9
CONDITION
GOOD MODERATE | SIGNIFICANT
COUNT | COUNT COUNT
AREA
4 4
1 .31
2 .|.4
3 4|
4 .|4.
1 4|
2 .( 3 1
3 .| 4
4 .|.4
1 .| 1 3
2 > . 1| 2
4 4
E .| 4|
1 4
2 . 3| .
3 3| 1 .
4 . 2| 2
1 1| 3|
2 4
3 . 3| .
4 . 1| 3
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE.
SEPARATED BY REGION, BUILDING AND AREA.
REGION 3
O
Ln
BUILDING
26
27
28
29
30
(AREA
1
2
3
4
I
Z
3
4
8
1
2
3
4
5
6
1 1
2
3
4
1
3
CONDITION
GOOD MODERATE
COUNT COUNT
2
3
.| ; 3
1 2
1
2| 1|
4
3|
1 2|
H 21
2| 2|
3| 1|
1 -1
H
2|
3|
3|
H 3|
2|
1
(SIGNIFICANT
COUNT
1
.
3
1
.
A
1
1
1
2
1
1
2
2
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 3
BUILDING |
30
31
32
33
34
35
AREA
4
6
6
1
2
3
4
1
2
4
6
9
1
4
6
9
1
2
3
4
11
CONDITION
GOOD MODERATE
COUNT \ COUNT
4
2
2
1
3
2
1
1 2
3
1 2
1
1
3
1
3
1
2| 1
2
1 3
2| 2
SIGNIFICANT
COUNT
2
2
3
1
2
3
3
2
2
.
2
2
1 1
1
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 3
BUILDING
35
68
. ,...,.
AREA
2
3
4
1
2
3
4
CONDITION
GOOD | MODERATE
COUNT | COUNT
.*"-: |
«l
1 3
1 2
1 3
1 1
1
.1 3|
SIGNIFICANT
COUNT
2
1
3
4
I
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION A
O
00
BUILDING |
10
36
37
38
39
GOOD
COUNT
AREA
1
2
3
6
1 1
2 I
3 I
< 1
7 |
1 1
2
3 I
5 I
6
1 1
2 I
3
4
1 1
2 1
3 I
CONDITION
MODERATE (SIGNIFICANT
COUNT COUNT
< .
3 1
3 1
3 1
2 2|
3 1
4
1 «
.| 1 2
4
1 3
21 1
11
3 1
1 3
1 3
3
3 1
4
.| 3|
1 . 3
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
BUILDING (AREA
39
40
41
42
59
4
CONDITION
GOOD MODERATE | SIGNIFICANT
COUNT COUNT |
.1
COUNT
5 I r si
1 . 4|
2 1 3|
3
.
.
4 .1
4 1 4 .|
7
13
1
2
3
4
« -1
2|
3|
1
4|
.21
B | 4|
6 | 2| 2|
7
8
1
2
3|
.
1
1
4
1
1
.| 4|
4| .|
1 21
3 I .1 3|
4 i .i r\
1
21 2|
1
1
1
.
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
BUILDING |
59
60
61
AREA
2
3
4
6
1
2
3
4
2
3
4
6
CONDITION
GOOD MODERATE
COUNT COUNT
3
3 1
2| 2
1 1
2
2 2
2| 2
4
1 21 2
2| 2
4
3| 1
SIGNIFICANT
COUNT
1
2
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
BUILDING
11
12
43
44
46
46
GOOD
COUNT
(AREA
1 4
2 3
3 4
< I 1
1 1
2 3
3 | 2
1
2
3 I 4
6 2|
1 -I
2 -I
3 .|
4 | . I
2 | 2|
3 | 1|
4 | 2|
7 1 -1
2 I 3|
4 | 11
CONDITION
MODERATE (SIGNIFICANT
COUNT COUNT
11
3|
3|
1
1
3| 1
2 2
21
2|
1| 2
3|
3|
H 1
2| 1
2|
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 5
BUILDING (AREA
46
47
48
49
62
63
5
6
1
2
3
1
2
3
4
1
2
3
4
7
8 I
1
2
3
8
GOOD
COUNT
1
1
2|
1
1
1
1
1
1
'4
1
2
1
CONDITION
MODERATE (SIGNIFICANT
COUNT |
2|
2|
3|
1
2|
3|
1
3|
2|
1
11
3|
2|
2|
3|
3|
J 3|
3|
COUNT
1
1
1
1
3
1
3
1
3
2
1
1
1
1
1
-------
TABLE B-l. RESPONSES OF RATERS TO OVERALL CONDITION VARIABLE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 5
u>
BUILDING
64
66
66
AREA
2
3
4
7
I
2
S
6
1
2
3
4
5
6
GOOD
COUNT
3
I
3
1
1
1
3
3
3
3
2|
3
3|
CONDITION
MODERATE | SIGNIFICANT
| COUNT COUNT
3
1
2|
.
3|
3|
3|
1
1
1
11
2|
1
1
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING |
13
14
16
16
17
DISTURBANCE
LOW | MODERATE) HIGH
COUNT | COUNT | COUNT
AREA 1 I I
1 1 l| 2|
2 .| H
4 .| 2|
e -I H
' 1 -1 -I
1 | .| 2|
2 | 3| 1|
3 2| 2|
4 | 1| 2|
7 | .| 1|
1 .1 H
2 | .| 2|
3 I .| .1
4 | 1| 2|
1 | .| 1|
3 .| 2|
4 | .) 2|
7 .| .1
e 1 .1 -I
1 .1 3|
2 .| 3|
1
3
2
3
4
2
.
1
2
1
3
3
2
2
4
1
1
1
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
17
60
61
62
63
64
(AREA
3
4 |
6
1
2
3 I
4 |
1 1
2
3 I
4
1
2
3
4
2
3 I
« 1
e |
i 1
3 I
DISTURBANCE
LOW (MODERATE) HIGH
COUNT COUNT | COUNT
1| 3
2| 2
2|
1 ^
3| 1
1 3| 1
2| 2
3| 1
1 H 3
2| 2
2| H 1
.| 2| 2
.1 2| 2
1 2| 2
2| 2|
1 .1 <
1 4
2| 2
3| 1
3| 1
.1 -I 4
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
64
66
69
67
7
1
C(
(AREA
4
6 1
8
1 1
2 1
3 1
4 1
1 1
2 1
3
4 |
1 1
2
4
6 i
1 1
2
3
4 I
7 1
6 1
DISTURBANCE
LOW | MODERATE)
9UNT | COUNT |
I
1
1 -1
1 -I
1 3|
1 -I
H
1 2|
1 2|
1 21
2|
11
1 H
1 2|
2|
3|
1 31
1 -I
1 3|
1 -I
.1 -I
1 ' -1
HIGH
COUNT
4
4
1
3
2
1
1
1
1
2
2
1
1
3
3
2
2
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
6
IAREA
"i
2
3 I
4
DISTURBANCE
LOW | MODERATE) HIGH
COUNT | COUNT | COUNT
1 3|
1 3|
1 3|
1 3|
-------
TABLE B-2. EESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
L
CO
BUILDING (AREA
IB
19
20
21
1
2 1
3
4
e
i 1
2 1
3 1
« 1
7
B
9
1 1
2
3 1
« 1
5
8
1 1
2 1
3 I
DISTURBANCE
OW | MODERATE)
LJNT | COUNT |
3|
2)
2| 2|
»l H
2|
« -I
H
.1 3|
1 3)
H .1
1 .1
n
11
1 3)
.1
1 2|
1 H
1 H
3|
1 M
1 2)
HIGH
COUNT
1
2
.
2
2
.
3
1
1
.
3
1
4
2
1
3
2
DISTURBANCE
LOW | MODERATE) HIGH
COUNT
BUILDING (AREA
21 |4
22
23
24
26
9
COUNT | COUNT
2 2
1 .) 3) 1
2 | .| 2) 2
3 | 4 . .
< ) aj 1) .
1 | 1 3)
2 ) 3) 1)
3 I .) . 4
4 .| .) 4
1 1
2) 2
2 .13
4 2) 2
6
2) 2
1 1 H 3|
2 | 3) 1 .
3 | 2) 1 1
4 I H 3)
1 .13
2 1) 2) 1
3
H 3
4 | .| 2| 2
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 3
BUILDING |
28
2T
28
29
30
DISTURBANCE
LOW | MODERATE |
COUNT | COUNT |
AREA 1 1 I
1 1 .1 l|
2 i .1 .1
3 I If 2?
4 I ll 2l
i r .1 3i
2 1 4| .|
3 21 21
4 | -1 H
e .1 3|
1 1 11 3|
2 | .| 2|
3 | 1| 2|
4 I 11 H
6 I H -1
8 1 -I H
1 1 -1 «l
2 H 31
3 1 -1
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 3
oc
BUILDING (AREA
36
68
2
3
4
1
2
3
4
DISTURBANCE
LOW | MODERATE | HIGH
COUNT
1
1
1
*
COUNT
3
COUNT
1
3|
2| 2
21
31
2| 2
2| 2
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
vO
BUILDING
10
36
37
38
39
DISTURBANCE
LOW (MODERATE)
COUNT | COUNT |
AREA I I
I 2|
2 .1 .1
3 -1 31
6 | 1| 2|
I 1 -I 2|
2 | .) 2|
3 | .| 2|
4 1 .1 .1
7 -1 21
1 -1 31
2 -1
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
NJ
o
BUILDING
69
60
61
AREA
;
3
*
6
1
2
3
4
2
3
<
6
DISTURBANCE
LOW (MODERATE
COUNT COUNT
3
3
2| 2
1
1
3|
1
1 3
31
3|
1 3
1 3
| HIGH
COUNT
1
1
3
1
3
1
1
1
»
»
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
BUILDING |
11
12
43
44
46
46
DISTURBANCE
LOW (MODERATE I
COUNT | COUNT |
AREA 1 1
1 2| 2|
2 | 2| 2|
3 | 2| 2|
4 1 1 3|
1 H 2|
2 2| 1|
3 2| 1|
1 1 H 3)
2 | 1| 2|
3 | 2| 2|
6 1 H 3|
'l * ll «I
2 -I 3|
3 1 -1 »l
4 | 1| 2|
'2 * ll 3|
3 | 1| 2|
4 | 1| 3|
7 H 2|
2 ll «I
4 H 31
HIGH
COUNT
1
*
1
1
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
DISTURBANCE
LOW | MODERATE) HIGH
BUILDING (AREA
46
47
48
49
62
63
6
6
1
2 I
3 I
« I
1 1
2
3 1
4
1
2 I
3
* I
1 1
1 1
B I
1 1
2 I
3 I
8 I
COUNT |
ij
H
H
2|
1|
H
H
11
H
3|
H
H
11
n
.1
.1
11
.1
2|
n
n
COUNT | COUNT
3
2| 1
3|
2|
2| 1
2| 1
3|
H 2
H 2
1
2| 1
2| 1
n
H 2
1 4
-------
TABLE B-2. RESPONSES OF RATERS TO POTENTIAL FOR DISTURBANCE,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
to
BUILDING |
64
66
66
AREA
.
3
«
7
1
2
6
6
1
2
3
4
6
6
DISTURBING!
LOW (MODERATE
COUNT | COUNT
J 2
1| 2
2| 1
2| 1
1| 3
H 3
1| 3
1 21 2
3|
1 «l
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA,
REGION 1
to
co
I
1
Cl
BUILDING (AREA
13
14
16
16
17
1
2 1
* |
« 1
7 1
1 1
2
3 1
4 |
7 |
1
2
3
4 |
1
3
4 I
7 I
B I
1 1
2 1
MR FLOW
40 | YES
1UNT| COUNT
4
H 3
2| 2
<|
«!
4
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
N>
BUILDING |
64
66
66
67
7
AREA
4
6
8
1
2
3
4
2
3
4
1
2
4
6
1
2
3
4
7
8
AIR FLOW
NO | YES
COUNT (COUNT
2 2
2| 2
1
31
3|
2| 1
1 3
1 3| .
3|
1 3
1 21 1
1 3|
31
1 M 2
3|
1| 2
1 31
2| 1
| 2| 1
1 -1 2
1 1
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 1
BUILDING
e
AREA
7 1
2 1
3
< 1
AIR FLOW
NO | YES
COUNT | COUNT
ll I
3|
2| 1
3|
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 2
10
BUILDING
18
19
20
21
IAREA
i
2
3
4
8
1
2
3
4
7
B
9 1
1
2
3 I
4 I
6
B
1
2
3 1
AIR FLOW
NO | YES
COUNT (COUNT
2| 2
1 21 2
1 3| 1
«l
1 3| 1
2| 2
3| 1
3| 1
3| 1
H
H
H
1| 3
3| 1
1 «
«|
1 1
1 1
3| 1
3| 1
31 1
1
Cl
BUILDING (AREA
21 M
22
23
24
26
8
1 1
2
3 I
4
1 1
2 I
3
« 1
1 1
2 I
4 I
E
1
2 I
3 I
« 1
» 1
2 I
3 1
« 1
MR FLOW
W YES
JUNTI COUNT
3 1
4|
*|
4
4|
4
4
2| 2
1 3
3| 1
3| 1
4|
2| 2
3| 1
3| 1
3| 1
2| 2
1 4
3| 1
3| 1
2| 2
-------
TABLE B-l. RESPONSES OF RATERS TO AIR PLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION I
TABLE B-l. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 8
r\)
BUILDING |
36
27
21
29
IB
AREA
1
2
I
4
1
2
1
4
8
1
2
1
4
E
6
1
2
1
4
1
1
AIR FLOW
NO | YES
COUNT (COUNT
.| B
.1 >
>l
2| 1
4|
2| 2
4|
1 1
4|
4|
2| 2
4|
1 H
11
H
II
II 1
4|
41
1 4|
.1 2
BUILDING
se
si
12
II
14
IB
AREA
;
e
t
i
2
»
4
1
2
4
e
9
1
4
B
9
1
8
1
4
1
AIR FLOW
NO | YES
COUNT (COUNT
l| »
4|
H >
4|
4|
1 4
4|
1 »
2| 1
»l
1
1
21
»l
S|
I
1
11 2
1 2
II
1 4
-------
TABLE B-S. RESPONSES OP RATERS TO AIR PLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION I
to
BUILDING
IS
SI
AREA
1
1
4
1
2
I
4
AIR PLOW
NO | YES
COUNT (COUNT
2| 2
«l
2| 9
*\
*\ 1
<|
<|
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
* SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
TABLE 6-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
to
00
BUILDING |
10
36
37
38
39
AREA
1
2
3
6
1
2
3
4
7
1
2
3
B
8
1
2
3
4
1
2
3
AIR FLOW
NO | YES
COUNT (COUNT
4|
3| 1
4|
1 4
4|
3| 1
3| 1
1 H
4|
4|
1 4|
1 -1 3
1 3|
4|
BUILDING
39
40
41
42
69
|AREA
4
6
1
2
3
4
7
13
1
2
3
4
6
6
7
B
1
2
3
4
1
AIR FLOW
NO | YES
COUNT (COUNT
*\
«l
3| 1
3| 1
«l
3|
«l
«l
3| 1
3| 1
3| 1
I <
1 <
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 4
NJ
BUILDING
60
60
61
AREA
>.
3
4
6 1
1 1
2 1
a 1
4 I
2 1
a
< 1
6
AIR FLOW
NO | YES
COUNT (COUNT
4
4
2| 2
11
4
3| 1
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
' SEPARATED BY REGION, BUILDING AND AREA.
REGION S
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
10
O
BUILDING |
11
12
43
44
46
46
AREA
1
2
3
<
1
2
3
1
2
3
6
1
2
3
4
2
3
4
7
2
4
AIR FLOW
NO | YES
COUNT (COUNT
3| 1
-------
TABLE B-3. RESPONSES OF RATERS TO AIR FLOW,
SEPARATED BY REGION, BUILDING AND AREA.
REGION 6
BUILDING
64
66
66
AREA
>.
3
4
7
1
2
6 1
6 }
I
2
3 1
4 1
6
6 I
AIR FLOW
NO | YES
COUNT) COUNT
|
3| .
31
3|
3|
«l
-------
APPENDIX C
CLASSIFICATION OF ACM CONDITION
(USEPA 1986a)
-------
SURFACING MATERIAL
Significant Damage ACM with one or more of the following
characteristics: the surface crumbling or blistered over at
least one tenth of the area if the damage is evenly distributed,
or at least one quarter if the damage is localized; large areas
.of material hanging from the surface, delaminated, or showing
adhesive failure; at least one tenth of the surface water-stained
or heavily gouged, marred or abraded (or one quarter if the
damage is localized); large accumulation of powder, dust, or
debris on surfaces beneath the ceiling or wall.
Moderate Damage ACM with one or more of the following
characteristics: up to one tenth of the surface (if the damage
is evenly distributed) or up to one quarter of the surface (if
the damage is localized) blistered, crumbling, water-stained, or
gouged, marred or abraded; some accumulation of powder, dust or
debris on surfaces beneath the ceiling or wall.
Good^Condition ACM with no visible damage or deterioration, or
showing only very limited damage or deterioration.
THERMAL SYSTEM INSULATION
Significant Damage ACM with one or more of the following
characteristics: mostly missing jackets; water-damaged, crushed
or heavily gouged or punctured insulation on at least one tenth
of pipe runs/risers if the damage is evenly distributed, or at
least one quarter if the damage is localized; powder, dust, and
debris on surfaces beneath pipe/boilers/tanks, etc.
Moderate Damage ACM with one or more of the following
characteristics: a few water stains or sections of missing
jackets; crushed insulation or water stains, gouges, punctures,
or mars on up to one tenth of the insulation if the damage is
evenly distributed, or up to one quarter if the damage is
localized; some accumulation .of powder, dust, debris on surfaces
beneath pipes/boilers/tanks, etc.
Good Condition ACM with no visible damage or deterioration, or
showing only very limited damage or deterioration.
135
-------
APPENDIX D
AIR SAMPLING FIELD METHODS
-------
Asbestos air samples were collected on 0.45 /tra pore
size, cellulose acetate membrane filters enclosed in preassembled
37 minicassettes. Two side-by-side samples were collected at each
location, each at a different flow rate. Volumes were controlled
through the use of two limiting orifices, 5.0 1pm, and 2.5 1pm,
with air flow being drawn across the filter by a diaphragm vacuum
pump. Collection of side-by-side samples at two different flow
rates allowed for a backup filter, should the higher flow rate
sample become overloaded. Samples were logged in the field and
hand carried to the analytical laboratory.
A. Site Selection
Sampling locations in each of the monitored buildings
were based on the representativeness of the location, proximity
to the ACM, accessibility, potential for vandalism, and access to
power. In general, eight sampling locations were selected in
each of the monitored buildings: seven indoor locations and one
outdoor ambient location.
Pump placement involved locating two pumps directly in
the area with the most damaged ACM, 2 in the nearest public
access area, 2 in other assessed areas, one in a public access
area adjacent to another area; and one pump at an outdoor
location. Local circumstances may have required pump placement
at other sites in a few instances.
In general, a sampling survey in each study region was
split into two segments, with each segment involving sample
collection in 5 buildings. The exception to this was Study
Region 4, consisting of 2 cities, in which 5 buildings were
sampled in each city. Each sampling period contained an initial
set-up day for pump placement, a 2-day sampling period, and
approximately one day for breakdown and filter delivery.
B. Sampling Equipment
The sampling system used during the project consisted
of the following:
Two open-faced 37 mm cassettes, each containing a
0.4 fim. cellulose acetate membrane filter;
Two flow control orifices; one at 2.5 1pm and the
other at 5 1pm;
A pump with a muffler;
Associated plumbing and stand;
139
-------
A 7-day timer; and
A clock to record elapsed time.
The sampler setup it represented in Figure D-l with
two modifications. The 36-inch rod used to hold the filters in
place was attached to a separate laboratory stand and not to the
pump base. This modification served to minimize the effects of
pump vibrations on the filter. The second modification was the
use of a T-fitting with two orifices (5.0 1pm and 2.5 1pm) and
two separate filter cassettes. This allowed for the collection
of two simultaneous samples, each at a different flow rate.
C. Sample Collection Procedures
Sample collection in each building was conducted
during periods of maximum activity (daylight hours) over a 2-day
period. Generally, sample collection hours were between 7:00
a.m. and 5:00 p.m. on each day of the 2-day sampling period.
Exceptions occurred when timer malfunctions required alterations
of the sampling period.
Pumps were set up one day prior to the actual sample
collection and set to activate the morning of the following day
via an in-line, 7-day timer. Advance set-up was necessary to
ensure that all samples started at approximately the same time
each day, since geographic locations were dispersed and building
access in the early morning hours uncertain. The following
details the sampling procedures followed during the program.
Sampling Protocol
1. Visually inspect preloaded filter cassettes for
damage. Label filter cassette with random I.D.;
2. Place filter in cassette holder, clamp into
position, and attach pump tubing. Ensure that
filter holder (ring stand) is placed in such a
way as to minimize or eliminate vibration effects
caused by the pump;
3. Rotate filter holders to a vertical position
(perpendicular to the ground);
4. Check plumbing for any leaks;
5. Check flow rates with a flowmeter;
6. Set automatic timer to correct date and time;
140
-------
Hoidrt
Oflfict
DotlH
1 4" CODOtr
facing Wound
4
J»oot 14' i 318'
I-2-MMC4T
NOU Connwior to Mm
M' Mill PiD« 10
1 *" l.O
me
Mofl«i 107CA1I
*«< w 6" i iv Mm SUM)
Mi't Hoow 90*
' I M»it 0iO» "n'MO«C to
1 4 TuB»
200;.
Figure 0-1. Pump diagram.
141
-------
7. Make appropriate logbook entries;
8. Conduct sampling;
9. After sampling, check flows (leave pump running);
and
10. Stop pump and remove filters.
Filter Handling Procedures
1. Use preloaded filter cassettes to minimize
contamination;
2. After sampling, place the cover over the filter
holder, maintaining exposed side up during the
handling, and transport to the laboratory;
3. Hand deliver all samples at the end of each
sampling period to the electron microscopy
laboratory; and
4. Maintain the filter in a horizontal position
during handling, transport and storage. Handle
in such a way as to minimize dislodging
structures from the filter surface.
Post-Sampling Procedures
1. Measure the flow;
2. Check filter condition and location of sampler;
3. Record time position of timer clock and elapsed
time;
4. Record the relative humidity and temperature
inside and outside the building; and
5. Complete chain-of-custody record prior to
packaging and shipment to the laboratory.
Logbook/Data Form Entries
An important part of any field program are the
observations and accurate records of the field team. The
following information was recorded in the logbook and data forms
for each sampling location:
142
-------
1. Name of field program;
2. Date of record;
3. Site number and location;
4. Tag numbers of pump and timer;
5. Relative humidity and temperature inside and
outside the building;
6. Position of sampler within the site;
7. Brief site description;
8. Corresponding filter number;
9. Sample flow rate at the start of sampling;
10. Settings of timer clock;
11. Sample flow rate at end of sampling period;
12. Comments; and
13. Name(s) of samplers.
A copy of the data form used during this program is
included as Figure D-2.
Field Flow Measurement
At a minimum, flow rate measurements were taken twice
during a sample run: during sampler set up prior to initiation
of the run, and at the completion of sampling. If possible, a
mid-point flow measurement also was taken. The following
describes the procedures used to determine sample flow rates in
the field.
1. Turn on the sampling pump;
2. Set up the sampling system as shown below with
both rotameters in line between the filter and
the orifice;
143
-------
Project No.
Building I.D.
Location
Field Data Sheet
Start Date.
Stop Date.
Pump I.D. No.
Filter Lot »..
Box *.
Day 1:
Start
Day 2:
Start
Flow Checks:
Date _
Tine
Flow Rate (st?).
Date.
Time.
F'.ow Rate (sr,p)
Date
Time
Flow Rate (st?)
Picture Roll -
Comments :
Flow Control Device.
Type
Random I.D.
Stop
Stop
Temp..
Rctaseter reading.
Rotameter neading.
Temp
Rotameter Reading.
Elapsed
Elapsed
B.P.
V.?
Rctameter No..
v.?.
Rotameter No.
V.P.
Frame *
Rotameter No.
Figure D-2. Field data form used for air monitoring.
144
-------
FILTER 1
FILTERS
fl
o
ORIFICE
ROTA«TER
BOTAMETER
ft<>
PUMP
WITH
MUFFLER
TIMER
ELECTRICAL
POWER
SOURCE
3. Record the rotameter readings in the notebook;
4. Turn off the pump and remove the rotameters;
5. Reconnect all tubing and turn the filters to a
horizontal position;
6. Repeat procedures 1 through 5 at the end of the
sampling period; and
7. Calculate the flow as follows:
a. Using the calibration curve for the
rotameter, determine the flow rates for each
rotameter reading and record these values on
the data sheet.
b. Calculate the average flow rate for the
sampling period using the following
equation:
Average flow rate = (initial flow rate^ final flow rate)
c. Calculate the actual volume of sample
collected by multiplying the average sample
rate by the sampling time.
145
-------
APPENDIX B
AIR SAMPLE PREPARATION AND SUMMARY OF TEM
ANALYTICAL PROTOCOL
-------
AMfLl HUBtAlATIQM
Tht following it an abbrtviattd vtrtion of tht aamplt
preparation procedure utiliiod in thii projtot, A mort detailed
vtraion of tho procedure, along with tne referenoti rtlattd to
tht development of thii procedure, ia contained in Itotion 7 of
Yamati tt al. (1914).
Tho Low Ttmporaturt Aahor (LTA) uiod in thoin toata wai
calibrated through a aerioa of ttata to determine tht etching
rate of a mixtd colluloat tattr filttr, It WAI tht inttnt of
thii projtot to ttoh approximattly 1 0m of tht turfact to rtvtal
itruoturt dttail that may havt bttn hiddtn in tht rtplicatt,
rrootdurti
1. A atotion of tht mtmbrtnt filttr ! out with a ioalptl, and
plaotd on a eltan mioroioopt ilidt with tht lampltd ildt
laoing up.
a. Tht out Motion ia faattntd on all eidtt to tht ilidt with
narrow atripa of tranapartnt tapt.
I. Tht ilidt, with tht out itotion, it txpoitd to aettont vapor
(not liquid) for approximattly 10 minutti. Tht aottont
vapor oolltpata tht atruoturt of tht filttr and product
fuitd, rtlativtly amooth-aurfaotd film. Tht aiit of tht
aottont vapor bath and timt of filttr rtaponat to tht vapori
art oritioal in obtaining tht dtairtd amooth, fuatd iurfaot.
4. laeh oollapatd filttr atgmtnt with a known dtpoait arta io
oartfully plaotd in a oltan ttat tubt (II mm x 10 mm) uaing
a oltan twttitr.
B. With forotpa. tht tubta containing tht atmplt, and 1 lab
blank (unuitd filttr atgmtnt of tht aamt aiit and typt of
filttr material aa tht aamplt) art plaotd Itngthwiit, iidt
by aidt in tht ohambtr, with tht moutha of tht tubta facing
tnt optn tnd (door) of tht aahtr ohambtr. Tht tubta art
laid In tht otnttr of tht ohambtr within tht rtgion of tht
eoila aurrounding tht ohambtr. Up to four aamplt tubta and
1 blank can bt laid likt loga inaidt tht ohambtr
I. Tht powtr ia lowly and oartfully incrtaatd to prtvtnt
"flaaning" of tht filttr, which would rtault in loaa of
aamplt.
7. Tht filttr mtmbrant ia ttohtd for approximattly 30 atoonda.
Tht chamber ia lowly allowed to rtaon ambient preaaurt.
149
-------
8. The etched-collapsed filter section is placed on the
rotating stage of the vacuum evaporator for carbon-coating.
9. A 3-nun-diameter portion of the carbon-coated filter is
transferred to an EM grid in the modified Jaffe wick washer.
10. Acetone is used in dissolving the fused membrane filter.
11. Transfer to the properly labeled grid storage container.
SUMMARY OF TEM ANALYTICAL PROTOCOL
The following is a summary of the Analytical Protocol
utilized in this project. A more detailed description of the
protocol is contained in Appendix B to the Quality Assurance Plan
(Hatfield et al. 1987) for this project.
Procedure:
1. Start a new Count Sheet for each sample to be analyzed.
Record on that sheet: Client Name; Project or Job No.;
Sample No.; Volume of Air Analyzed (from TEM Working Log);
Microscope; Magnification for Analysis; Filter type, and
Diameter.
2. Start with the grid in capsule labelled No. 1 located in the
Specimen Box.
3. Determine Suitability of Grid
A. Look at grid in Low Mag mode (100X) to determine its
suitability for detailed study at higher mags.
B. Reject grid if:
i. Replica does not cover at least 15 full grid
openings with 0% holes in any grid, and < 15%
coverage maximum. Discount any grid opening that is
doubled or folded for counting.
ii. Specimen is too dark due to incomplete dissolution
of the filter.
iii. The average particulate loading exceeds 15%.
C. If grid is rejected, load grid from capsule No. 2, etc.
D. If grid is acceptable, continue on to next step.
150
-------
4. Scan the Grid
A. Set the magnification to 19,OOOX
B. Scan grid as follows:
i. At the appropriate magnification, make a series of
parallel traverses across the grid opening.
Traverse the grid opening (also referred to as a
field), starting at 1 corner (upper left or upper
right) and using the area defined by the small
square of the fluorescent screen (area of screen
that lifts up for photograph purposes) as a
"window".
ii. On reaching the end of 1 traverse, move the image 1
"window" width, and reverse the traverse. A slight
overlap should be used so as not to miss any part of
the opening.
iii. Make parallel traverses until the entire grid
opening has been scanned.
C. Ten good fields or grid openings or 100 structures need
to be counted (whichever comes first).
5. Identify each structure morphologically and analyze as it
enters the "window".
6. For morphology: appearance and size
A. Determine morphologically if the structure is a "fiber",
"bundle", "cluster", or "matrix".
B. If record "bundle", "cluster", or "matrix", then record
also how many figures are involved; i.e., Bundle 7,
Bundle > 50, etc.
C. Size each structure using the calibrated 20 mm rule on
the screen.
7. Selected area electron diffraction pattern (SAED)
A. Center structure, focus and obtain SAED pattern
B. From a visual examination of the electron diffraction
pattern (camera length (CL) of 22; through binoculars on
small screen), classify the observed structure as
151
-------
belonging to one of the following categories by comparing
it to known patterns:
i. Chrysotile: The chrysotile asbestos pattern has
characteristic streaks on layer lines other than the
central line and some streaking also on the central
alternate lines (2nd, 4th, etc.)* The repeat
distance between layer lines is about 0.53 mm.
ii. Amphibole Group (includes amosite, crocidolite,
anthophyllite, tremolite and actinolite): Amphibole
asbestos structure patterns who layer lines formed
by very closely spaced dots, and the repeat distance
between layer lines is also about 0.53 mm.
Streaking in layer lines is occasionally present due
to structure defects.
iii. Ambiguous (incomplete spot patterns).
iv. N, if there is no pattern present. (This should go
under SAED column).
C. If the pattern is a suspected chrysotile or amphibole,
then take picture of diffraction pattern as needed.
8. X-ray Analysis (EDS).
A. For each structure that chemistry is necessary, take
chemistry with EDS system.
B. If EDS signal is weak, take another spectrum, being sure
that spot is still on structure.
C. If EDS is used for confirmation, record structure
identification. Record a check mark or an "X" in EDS
column when chemistry is checked but not saved.
D. If EDS is used in case of unknown or ambiguous
structures, categorizing amphiboles or showing
representative structures on particular field, save
spectra to disk and record Disk No. and File No. on Count
Sheet under EDS Column.
9. After all necessary analyses of structure, continue scanning
until all structures are identified, measured, analyzed, and
categorized in the grid opening.
10. Select additional grid openings at low mag, scan at 19,OOOX
and analyze until the total number of asbestos structures
exceeds 100, or a minimum of 10 grid openings have been
examined, whichever comes first.
152
-------
11. Carefully record all data as it is being collected, and
check for accuracy.
12. After finishing with grid, remove from microscope,and
replace in appropriate polyethylene capsule.
153
-------
APPENDIX F
ANALYSIS OF TEM GRID OPENING DATA
-------
BACKGROUND
When air samples are analyzed for asbestos by TEM using
a direct preparation technique the spatial distribution of
asbestos structures on the electron microscope grid is similar to
their distribution on the filter at the time of collection.
(Some changes may take place during transport to the laboratory.)
Concerns have been raised over the uniformity of the spatial
distribution. Since only a small proportion of the filter is
examined, a highly clumped or non-uniform distribution may yield
low structure counts by chance, even though the average structure
density is high. Conversely, if the area of filter examined
happens to include an aggregation of asbestos structures the
airborne asbestos concentration will be overestimated.
No asbestos structures were detected on over 80% of the
samples collected in this study. This prompted additional
analyses to determine if the low structure counts reflected
actual low airborne asbestos concentrations or could be explained
by a non-uniform distribution of asbestos structures on the
surface of the filter. The objective of the investigation was to
characterize the spatial distribution of asbestos structures and
determine the effect of the distribution on the precision of
structure counts.
STUDY DESIGN
Sixteen air samples were selected for additional
analysis to determine if the 10 grid openings specified by the
TEM protocol provide estimates of sufficient precision for the
purposes of the study. The 16 samples were selected as follows
to provide a range of structure counts:
4 "indoor" samples which had structure counts of 3 or
more in the first 10 grid openings counted;
8 "indoor" samples which had structure counts of 0 in
the first 10 grid openings;
2 "outdoor" samples; and
2 field blanks.
Samples were selected at random within each category.
An additional 40 grid openings, giving a total of 50,
were examined on each sample and the number of structures in each
opening recorded.
157
-------
STATISTICAL MODEL
The negative binomial is a discrete distribution which
is often used to describe clumped or aggregated populations.
Javitz and Fowler (1981) found that the negative binomial was
superior to the Poisson for describing asbestos structure counts
obtained by electron microscopy. The variance of the negative
binomial is m(m+k)/k where m is the mean and k is a measure of
aggregation. As k increases the variance decreases and
consequently the precision of estimated airborne asbestos
concentrations increases. The Poisson distribution is a limiting
case of the negative binomial as k becomes very large.
If the number of asbestos structures in one grid
opening is assumed to be distributed according to a negative
binomial distribution with parameters m and k, then the number of
structures in n grid openings is distributed according to a
negative binomial with parameters nm and nk. This result assumes
that the number of structures in a grid opening is independent of
the number in any other grid opening. The assumption will hold
if grid openings are selected at random. If grid openings are
not selected at random then one must assume that there is no
spatial correlation between the number of structures in different
grid openings.
PARAMETER ESTIMATES
The maximum likelihood estimate of m is the sample mean
x. The maximum likelihood estimate of k is obtained by solving
the equation
x °°
nlog(l + P ) = Z f^l/k +l/(k+l)+..-+l/(k+j-l)),
j=l -1
where fj is the number of grid openings with j structures (Bishop
et al 1975). Estimates of m and k were obtained for each sample.
Large values of k mean higher precision, and hence narrower
confidence intervals for the total structure count on a filter.
CONFIDENCE INTERVALS
Exact 100(l-a)% confidence intervals, (HIL^U)' were
obtained for x> 0 by finding mL and my such that
F(x,mu)=
-------
«)% confidence interval is given by (Q,mq) where nty satisfies
F(x,mu)=a.
A negative binomial with parameters nra and nk is used
to obtain confidence intervals for the total count on a filter
when n grid openings are examined. Increasing k, and/or
increasing n, increases the precision of the count and reduces
the width of the confidence interval.
RESULTS
No asbestos structures were counted on eight of the 16
filters. The eight include the two field blanks and the two
outdoor samples. Of the eight filters with non-zero counts, five
have estimates of k equal to infinity. The remaining three
estimates of k are 0.6, 0.4, and 0.07.
For k=», i.e., a Poisson distribution, a 95% confidence
interval for the true structure count when no structures are
counted in 10 grid openings is (0,3.0). The size of the
confidence interval increases slightly to (0,3.1) as k decreases
to 0.4. Thus, for values of k greater than or equal to 0.4 the
examination of 10 grid openings in this study yields an airborne
asbestos concentration that is sufficiently precise to
distinguish 0 s/cc from 0.009 s/cc with high probability. (In
this study one structure corresponds to approximately 0.003
s/cc.)
The data indicate that k is usually greater than 0.4,
but that smaller values, such as k=0.07 are possible. The
standard deviation of this estimate of k is 0.9. For k-0.07, a
95% confidence interval for the true structure count when no
structures are counted in 10 grid openings is (0,50). If the
number of grid openings counted is increased to 50, the
confidence interval shrinks to (0,4.7).
Of the 16 filters examined, all but one indicate that
examination of 10 grid openings is sufficient to distinguish 0
s/cc from 0.009 s/cc with high probability. Although, without
additional data, it is difficult to predict how frequently
exceptions will occur, the results suggest that examination of
additional grid openings is generally unnecessary unless higher
precision is required.
159
-------
APPENDIX G
AIR MONITORING DATA LISTING
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING*IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =1 BUILDING CATEGORY
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.001
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
BUILDING NUMBER =2 BUILDING CATEGORY =1
u>
SITE STRUCTURE
CONCENTRATION
1
2
3
4
5
6
7
002
000
033
BUILDING NUMBER =3 BUILDING CATEGORY el
SITE
0
1
2
3
4
6
6
7
STRUCTURE
CONCENTRATION
.000
006
000
000
000
000
0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =4 BUILDING CATEGORY
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
5 0 000
6 0.000
7 0.000
BUILDING NUMBER =5 BUILDING CATEGORY =1
ON
SITE STRUCTURE
CONCENTRATION
0 0.003
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
BUILDING NUMBER =6 BUILDING CATEGORY si
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =7 BUILDING CATEGORY *2
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.001
2 0.000
3 0.000
4 0.003
6 0.000
6 0.003
7 0.000
BUILDING NUMBER =8 BUILDING CATEGORY =2
SITE STRUCTURE
CONCENTRATION
,_, 0 0.003
0% 1 0.004
Ui 2 0.000
3 0.003
4 0.003
5 0.000
6 0.000
7 0.000
BUILDING NUMBER =9 BUILDING CATEGORY =2
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.003
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =10 BUILDING CATEGORY =2
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.000
BUILDING NUMBER =11 BUILDING CATEGORY =2
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.002
3 0.000
4 0.000
E 0.000
6 0.000
7 0.000
BUILDING NUMBER =12 BUILDING CATEGORY =2
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.002
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =13 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.006
3 0.004
A 0.002
6 0.000
6 0.006
7 0.000
BUILDING NUMBER =14 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.004
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
BUILDING NUMBER =16 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1
2
3
4
6
6
.001
.003
.000
.001
.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =16 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.001
2 0.000
3 0.002
4 0.000
5 0.000
6 0.000
7 0.003
BUILDING NUMBER =17 BUILDING CATEGORY =3
ON
00
SITE
0
1
2
3
4
5
6
7
STRUCTURE
CONCENTRATION
,000
,000
,000
,000
,000
,000
BUILDING NUMBER =18 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0 «000
1 0.000
2 0.000
3 0.000
4 0.000
5 0.003
6 0.003
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =19 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.003
3 0 «000
4 0.000
o 0«000
6 0.000
7 0.000
BUILDING NUMBER =20 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
_ 0 0.000
£ 1 0.000
Co 2 0.000
3 0.003
4 0.000
6 0.000
6 0.000
7 0.000
BUILDING NUMBER =21 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0000
4 0.000
5 0 »000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =22 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0 000
3 0.000
4 0.000
E 0.000
6 0.000
7 0.000
BUILDING NUMBER =23 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0 000
1 0.003
2 0.013
3 0.000
4 0.000
6 0.003
6
7 0.000
- BUILDING NUMBER =24 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =25 BUILDING CATEGORY =3
SITE
0
1
2
3
4
6
6
7
STRUCTURE
CONCENTRATION
.000
.000
.000
.013
.000
.000
0.000
BUILDING NUMBER =26 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0 *000
1 0.003
2 0.000
3 0.000
4 0.000
O 0000
6 0.000
7 0.000
BUILDING NUMBER =27 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 .000
1 .000
2 .000
3 .000
4 .000
5 00o
0 000
7 .000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =28 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0 .000
1 0.000
2 0.000
3 0 *000
A 0.000
6 0.008
6 A AAA
0 0190
7 0.000
BUILDING NUMBER =29 BUILDING CATEGORY «3 ----
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.003
2 0.000
3 0.000
4 0.000
B 0.000
6 0.003
7 0.000
BUILDING NUMBER =30 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.007
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =31 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.009
2 0.000
3 0 00D
4 0.000
o 0.000
6 0 000
7 0.000
BUILDING NUMBER =32 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
h-> 1 0.000
-J 2 0.000
1° 3 0.000
4 0.000
5 0.000
6 0.000
7 0.
BUILDING NUMBER =33 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.003
1 0.000
2 0.000
3 0.000
4 0.003
5 0 000
6 0.003
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =34 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0 *000
7 0.000
BUILDING NUMBER =35 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
E 0.000
6 0.003
7 0.000
BUILDING NUMBER =36 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.003
-------
TABLE 6-1. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =37 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.006
3 0.005
4 0.000
5 0.000
6 0.000
7 0.000
BUILDING NUMBER =3B BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.002
,_ 1 0.000
Cj 2 0.004
Ui 3 0.002
4 0.002
6 0.000
6 0.003
7 0.003
BUILDING NUMBER =39 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.003
1 0 000
2 0.000
3 0.000
4 0.002
5 0.000
6 0.003
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
-- BUILDING NUMBER =40 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
5 0.002
6 0.000
7 0.000
BUILDING NUMBER =41 BUILDING CATEGORY =3
ON
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
BUILDING NUMBER =42 BUILDING CATEGORY =3
SITE
0
1
2
3
4
5
6
7
STRUCTURE
CONCENTRATION
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION fS/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =43 BUILDING CATEGORY *3
SITE STRUCTURE
CONCENTRATION
0 0.002
1
2
3 0.003
4 0.000
D 0 »000
6 0.000
7 0.000
BUILDING NUMBER =44 BUILDING CATEGORY »3 --------
SITE STRUCTURE
CONCENTRATION
i 0.000
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.000
BUILDING NUMBER =46 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.
1 0.000
2 0.000
3 0.003
4 0.000
5 0.003
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0" SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =46 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.002
1 0.000
2 0.000
3 0.000
4 0.000
5 0.000
6 0.000
7 0.005
BUILDING NUMBER =47 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.006
3 0.003
4 0.000
6 0.000
6 0.000
7 0.003
BUILDING NUMBER =48 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0.000
1 0.000
2 0.000
3 0.000
4 0.000
6 0.000
6 0.000
7 0.000
-------
TABLE G-l. AIR MONITORING DATA LISTING SHOWING THE ASBESTOS
STRUCTURE CONCENTRATION (S/CC) AT EACH SITE THAT WAS
AIR SAMPLED. THE "0« SITE AT EACH BUILDING IS
ALWAYS THE OUTDOOR LOCATION. SITES 1-7 DO NOT
CORRESPOND TO THE SITE NUMBERING USED IN APPENDICES
A AND B.
BUILDING NUMBER =49 BUILDING CATEGORY =3
SITE STRUCTURE
CONCENTRATION
0 0 000
1 0.002
2 0.000
3 , 0.000
4 0.000
5 0.000
D 0 .000
7 0.000
vO
-------
APPENDIX H
GLOSSARY
-------
ACM: asbestos-containing material.
Air sample: a filter through which a known volume of air has
passed in order to measure the asbestos structure concentration
in the air during the period of sampling.
Air flow: an air flow transports fibers from the point of
release from the ACM to other areas in the building. Air
plenums, air shafts and elevator shafts represent different types
of air flow.
Air monitoring: the process of collecting air samples in a
building.
Asbestos: a group of naturally occurring minerals that separate
into fibers. There are six asbestos minerals used commercially
(chrysotile, amosite, crocidolite, anthophyllite, tremolite, and
actinolite).
Bulk sample: a portion of friable material collected in order to
measure the asbestos content of the material.
Categories of buildings:
Category 1a building in which no friable asbestos-
containing surfacing materials or TSI were noted in the GSA
records and none was found during the building inspection.
Category 2a building in which all or most of the areas
with friable asbestos-containing surfacing materials or TSI
were in good condition allowing for a limited number of
areas of moderate damage.
Category 3a building which had at least one significantly
damaged area of friable asbestos-containing surfacing
material or TSI, or there were numerous moderately damaged
areas.
Condition: See Appendix C for definitions of ACM condition.
Disturbance: (classifications revised March 19, 1987)
High potential for disturbanceACM which has two or more of
the three factors (accessibility, vibration, air erosion)
rated "high," or one factor sufficiently high that the
material is almost certainly going to be disturbed.
Examples are (1) acoustic plaster on a low ceiling in a high
school band room; (2) thermal system insulation on air ducts
connected to ventilation fans and readily accessible to
workers conducting maintenance on the ventilation system;
and (3) fireproof ing on low beams in a work room located
just downstream from an air vent.
183
-------
Moderate potential for disturbanceACM which is accessible,
subject to vibration, or subject to air erosion, but has no
more than one factor rated as high. ACM on corridor walls,
on a ceiling underneath a gymnasium, or in an elevator shaft
are examples of material with a moderate potential for
disturbance.
Low potential for disturbanceACM which has low
accessibility, is not subject to vibration, and is not
subject to air erosion.
External analysis: an analysis in which a sample is analyzed a
second time by another analytical laboratory. This type of
analysis is performed as a QC check on the performance of the
method by the primary laboratory. The degree of agreement of the
original analysis with the external analysis indicates the
consistency of the method performance.
Field blank: a filter taken into the field, handled in the same
manner as exposed air sample filters, and analyzed for
contamination which might occur in the field but not as a result
of air sampling.
Friable: capable of being crumbled, pulverized, or reduced to
powder by hand pressure.
Production lot blank: a filter chosen prior to field work and
analyzed by the laboratory to check for filter contamination.
PLM: polarized light microscopy.
Replicate analysis: an analysis in which a sample is analyzed a
second time by the same analytical laboratory. The degree of
agreement of the original analysis with the replicate analysis
indicates the level of precision in the laboratory analysis
procedures.
Side-by-side duplicate: a sample collected in the immediate area
of the original sample but handled separately. The degree of
agreement of the analyses of the original sample with its
duplicate indicates the level of precision in the sample
collection and field handling procedures.
Structure: An asbestos fiber, bundle, cluster, or matrix.
Surfacing: ACM sprayed or troweled on surfaces, such as
acoustical plaster on ceilings and fireproofing material on
structural members.
TEM: transmission electron microscopy.
184
-------
Thermal systems insulation: ACM applied to pipes, boilers,
tanks, ducts, etc. to prevent heat loss or gain or water
condensation.
w
condensation.
TSI: thermal systems insulation.
185
-------
50272-101
REPORT DOCUMENTATION
_J>AGE
4. Title and Subtitle
1. REPORT NO.
EPA 560/5-88-002
2.
3. Recipient's Accession No.
Assessing Asbestos Exposure in Public Buildings
-.j
5. Report Date
May. 1988
7. Author**)
Hatfield, J. et al
9. Performing Organization Name and Address ' "
Battelle Columbus Division, Washington Operations, 2P30 M St. NW
Washington, D.C. 20036; Price Associates,Inc., 1825 K St, NW,
Wash, DC, 20006; Alliance Technologies Corp, 213 Burlington Rd,
Bedford, MA 07130; R. J. Lee Group, Inc., 350 Hochberg Rd, Mon-
roeville, PA 15146; Midwest Research Inst., Kansas City, MO 641K>(G)
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
Office of Toxic Substances
Exposure Evaluation Division (TS-798)
401 M St., S.W. Washington, D.C. 20460
15. Supplementary Notes
8. Performing Organization Rept. No.
10. Project/Task/Work Unit No.
"* B^-O^-^?^""^' N°'
(C) 68-02-3997
'68-03-3406,68-02-4252
13. Type of Report & Period Covered
Peer-reviewed report
14.
is. Abstract (Limit: 200 words) Airborne asFes"fos levels were measured "by "dl'recrfc l/ransmissio
electron microscopy in 49 public buildings from three categories: (1)
buildings without asbestos-containing material (ACM); (2) buildings with all
or most of-the ACM in good' condition allowing for a limited number of areas
of moderate damage; and (3) buildings which had at least one area of
significantly damaged- ACM or numerous areas of moderate damage. Approximatel
seven areas were monitored inside and one area outside each building.
Although the absolute airborne asbestos levels were very low, Category (3)
had the highest median levels followed by Category (2), Category (1), and
outdoors. Category (3) levels were significantly higher than Category (1);
Another objective was to field test an assessment method for ACM developed t
facilitate abatement decision making in the context of an asbestos managemen
program. Material condition, potential for disturbance, and air flow were
assessed by trained raters in 257 areas in 60 public buildings. Using rater(
consistency as an evaluation criterion, the three factors showed promise as j
assessment tools for use- in the field. Each factor showed statistically .
significant consistency among raters. A further observation associates |
disagreement among raters with imprecision in definitions and the absence ofi
proper training.
17. Document Analysis m. Descriptors
Asbestos, asbestos exposure, asbestos air monitoring, asbestos assessment, TEM,
transmission electron microscopy
b. Identifiers/Open.Ended Terms
c. COSATI Field/Group
18. Availability Statement
19. Security Class (This Report)
unclassified
20. Security Class (This Page)
unclassified
21. No. of Pages
202
22. Price
(See ANSI-Z39.18)
See Instructions on Reverse
OPTIONAL FORM 272 (4-77
(Formerly NTIS-35)
C ,:-,erce
------- |