United States
Environmental Protection
Agency
Office of Toxic Substances
Office of
Toxic Substances
Washington, DC 20460
EPA 560 13-80-017A
December 1980
Asbestos - Containing Materials
in School Buildings
Guidance for Asbestos
Analytical Programs
I
nmosite
chrysotile
crocidolite
-------
Cover slides supplied by:
McCrone Research Institute
Chicago, Illinois
-------
EPA 560/13-80-017A
ASBESTOS-CONTAINING MATERIALS
IN SCHOOL BUILDINGS
Guidance For Asbestos Analytical Programs
December 1980
by
D. Lucas
T. Hartwell
A. V - Rao
Research Triangle Institute
Research Triangle Park,
North Carolina 27709
EPA Contract Number 68-01-5848
EPA Task Manager: Cindy Stroup
EPA Project Officer: Joe Carra
Design and Development Branch
Exposure Evaluation Division
Office of Toxic Substances
Washington, D. C. 20460
-------
DISCLAIMER NOTICE
This report was prepared under contract to an agency of
the United States Government. Neither the United States
Government nor any of its employees, contractors, subcon-
tractors, or their employees, makes any warranty, expressed
or implied, nor assumes any legal liability or responsibility
for any third party's use or the results of such use of any
information, apparatus, product or process disclosed in this
report, nor represents that its use by such third party
would not infringe privately-owned rights.
Publication of the data in this document does not
signify that the contents necessarily reflect the joint or
separate views and policies of each sponsoring agency.
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
11
-------
PREFACE
This document is one in a series prepared in support of
the EPA Asbestos-In-Schools Program. It was developed to
provide guidance to local school officials and their staffs
in determining the presence or absence of asbestos in school
buildings. Data and information generated during the EPA
Technical Assistance Program have been used to design a
rigorous sampling and analysis scheme for bulk materials.
Implementation of the enclosed sampling protocol will reli-
ably document the presence or absence of asbestos in the
bulk materials and provide an interval estimate of the
asbestos content.
EPA has prepared rules which, when final, would require
the examination of public school buildings for asbestos. The
EPA Asbestos-In-Schools Identification and Notification Rule
was proposed in September 1980 and is planned to be final in
early 1981.
111
-------
ACKNOWLEDGEMENTS
The authors greatly appreciate the helpful suggestions
of Larry Longanecker and Joe Breen of the U.S. Environmental
Protection Agency and Steve Williams and Martin Rosenzweig
of Research Triangle Institute. Many thanks are also given
to Carol Mitchell for the outstanding formatting and typing
job and to Lynne Srba for the cover design and printing.
IV
-------
TABLE OF CONTENTS
Page
DISCLAIMER ii
PREFACE iii
ACKNOWLEDGEMENTS iv
LIST OF TABLES vii
LIST OF FIGURES viii
1. INTRODUCTION 1
Asbestos Analytical Program Coordinator 2
Sampling of Materials Suspected to Contain
Asbestos 2
Laboratory Analytical Technique 3
Laboratory Selection 4
Specific Information to be Reported by
Laboratory 6
Quality Assurance Measures 6
Recordkeeping 7
CHECKLIST FOR AN ASBESTOS ANALYTICAL PROGRAM.. 9
2. SAMPLING FRIABLE MATERIAL 11
Sampling Procedure 13
Establishing an Asbestos Analytical Program
File 14
Inspecting for Friable Material 15
Establishing Sampling Areas 15
v
-------
Table of Contents (continued)
Page
Diagram Preparation 16
Number of Samples to Take 18
Selection of Sample Locations 21
SELECTION OF SAMPLE LOCATIONS WORKSHEET 23
Sample Collection 27
Precautions to be Taken During Sampling 29
An Illustration of the Sampling Procedure 30
3. LABORATORY QUALITY ASSURANCE 37
Split-Sample Techniques for Quality Assurance.. 38
Continuing Quality Assurance Program 40
4. LABORATORY ANALYSIS AND STATISTICS 49
Forwarding Samples to Laboratory 50
LABORATORY DATA SHEET 53
Statistical Analysis of Laboratory Results 55
INSTRUCTIONS FOR STATISTICS COMPUTATION
WORKSHEET 57
STATISTICS COMPUTATION WORKSHEET 59
REFERENCES 63
APPENDIX A: EPA-Sponsored Analytical Proficiency
Program For Asbestos Bulk Sample
Analysis 65
APPENDIX B: Quality Assurance Program for Initial
Laboratory Evaluation 77
APPENDIX C: How to Use A TABLE OF RANDOM DIGITS 83
APPENDIX D: EPA Regional Asbestos Coordinators 91
APPENDIX E: Toil-Free Information Number 95
VI
-------
LIST OF TABLES
Number Title Paqe
2.1 TABLE OF RANDOM DIGITS Used For The
Example 34
VI1
-------
LIST OF FIGURES
Number Title Page
2.1 Example Sampling Area Diagram 19
2.2 Example Sampling Area Diagram 20
2. 3 WORKSHEET 23
2.4 Example Sampling Area Diagram With Sample
Locations Marked 31
2.5 WORKSHEET EXAMPLE 32
3.1 Case 1: School Systems For Which Fewer
Than 25 Samples Will be Analyzed 42
3.2 Case 2: School Systems With Expected Num-
ber of Samples Over 25 But Not Over 100.... 44
3.3 Case 3: School Systems With Expected Num-
ber of Samples Over 100 46
A.I Example of Reports to Laboratories 70
B.I Case 1: School Systems For Which Fewer
Than 25 Samples Will be Analyzed 80
B.2 Case 2: School Systems With Expected Num-
ber of Samples Over 25 81
VIXI
-------
CHAPTER 1: INTRODUCTION
The objective of this document is to provide guidance
to local school officials and their staffs for the effective
implementation of an asbestos analytical program that will
generate adequate information for decision-making and yet
not be too costly in terms of dollars and human resources.
Participants in the school asbestos program should be
aware of some of the pitfalls associated with carrying out
an asbestos analytical program. The proper implementation
of a valid program to characterize suspected asbestos-
containing materials requires an appreciation of the inter-
dependence of the various elements of the overall process.
The importance of random sampling, appropriate chemical
analytical techniques, selection of a laboratory to do the
bulk sample analyses, and an effective laboratory monitoring
program are emphasized throughout this document.
The following paragraphs outline seven elements that
are necessary in an asbestos program. Chapters 2, 3, and 4
then discuss in detail sampling procedures, laboratory
quality assurance, laboratory analysis, and statistical
analysis.
-------
ASBESTOS ANALYTICAL PROGRAM COORDINATOR
The first element in the program is to identify an
asbestos analytical program coordinator to be responsible
for overseeing the entire asbestos program. In particular,
the coordinator is responsible for supervising the sampling
of suspected asbestos-containing materials, selecting labo-
ratories to analyze the bulk samples for asbestos content,
monitoring the laboratories' performance throughout the
analysis period, and preparing a summary report. If possi-
ble, someone with a technical background, such as mathema-
tics or science, should be designated coordinator.
SAMPLING OF MATERIALS SUSPECTED TO CONTAIN ASBESTOS
The second element in the asbestos analytical program,
sampling of the suspect material, is considered the single
most important step in the process. Proper sampling, that
is, random sampling, is the basis upon which the validity of
the subsequent laboratory analysis program and decision-
making processes rest. If the suspect material is impro-
perly sampled, the analyses that follow will be compromised.
Chapter 2 describes inspection for suspect material,
identification of Sampling Areas, and the recommended sam-
pling procedure. A simple random sampling procedure is
employed to ensure the reliability of the results. To aid
in this process, a SELECTION OF SAMPLE LOCATIONS WORKSHEET
is provided in Chapter 2. The number of samples recommended
-2-
-------
provides a high chance of detecting asbestos in bulk mate-
rials, if present. For example, if a Sampling Area has 5%
or more asbestos content, taking at least three samples
would give greater than a 90% chance of detecting the pre-
sence of asbestos. (The assumptions underlying this state-
ment are described in detail in a Statistical Background
Document, [2]. These assumptions are based on data made
available to EPA by the Bureau of Mines and the Battelle-
Columbus Laboratories.) To achieve greater than 90% assured-
ness, the number of samples taken in each sampling area
would have to be increased.
LABORATORY ANALYTICAL TECHNIQUE
The third element in the asbestos program is the appro-
priate choice of a laboratory technique to analyze the
suspect materials. In view of the health and economic
implications, accurate determination of the presence or
absence of asbestos is critical.
The method of choice for the determination of asbestos
in suspect materials is polarized light microscopy (PLM)
with or without dispersion staining (DS), and with X-ray
diffraction CXRD) as necessary to supplement the PLM ana-
lysis [1]. PLM is the only method that depends on the
unique optical crystallographic properties of the sample.
These properties uniquely identify the individual asbestos
types: chrysotile, actinolite, amosite, anthophyllite,
-3-
-------
crocidolite, and tremolite. These crystal aspects coupled
with the fiber shape will uniquely identify the asbestos
present in the material being analyzed and will also char-
acterize non-asbestos fibers present such as fiberglass and
cellulose.
Another analytical technique used in asbestos analysis
is phase contrast microscopy. This method was developed by
the National Institute of Occupational Safety and Health
(NIOSH) for use in occupational settings when a significant
asbestos insult is known to exist. This technique is used
to count fibers based solely on their shape and size and
does not distinguish between asbestos fibers and non-asbestos
fibers such as cellulose, hair, and fiberglass. Consequently,
analysis of bulk samples for the determination of asbestos
content by this laboratory technique is unacceptable.
A detailed analytical protocol for the bulk analysis of
asbestos-containing insulation and sprayed-on materials is
being prepared and tested [3], This protocol should serve
as an authoritative guide to any bulk sample analysis pro-
gram using PLM and XRD as analytical tools.
LABORATORY SELECTION
The fourth element of the asbestos analytical program
is the selection of a competent and reliable laboratory.
The identification of asbestos in bulk samples involves
expertise in optical crystallography and is not a routine
-4-
-------
laboratory procedure. Only laboratories actively engaged in
using polarized light microscopy for the analysis of bulk
samples for asbestos materials should be considered for this
service.
EPA is sponsoring an analytical proficiency program for
bulk sample analysis. Presently, 52 commercial and 23 non-
commercial laboratories are participating on a voluntary
basis. A brief description of this program, the results of
round one, sample reporting, and a list of participating
laboratories are provided in Appendix A. This is not a
laboratory certification process; however, these laboratories
have demonstrated proficiency in analyzing bulk samples
using polarized light microscopy.
It is recommended that laboratories from this list be
selected for school asbestos programs. If it is not possi-
ble to select a laboratory from this list, a procedure for
evaluating the performance of an unknown laboratory is
provided in Appendix B.
A laboratory proficient in NIOSH asbestos fiber-count-
ing methodology using phase contrast microscopy may lack
both the equipment and expertise for PLM identification of
asbestos in bulk samples. As stated above, phase contrast
microscopy is inappropriate for the differentiation of
asbestos from non-asbestos fiber materials.
-5-
-------
SPECIFIC INFORMATION TO BE REPORTED BY LABORATORY
The fifth element in the asbestos analytical program is
to specify the information to be reported by the laboratory
for each sample submitted for analysis. It is important
that complete reporting of the analytical results be obtain-
ed from the laboratory.
The laboratory report should include: school's "blind"
sample ID numbers, laboratory sample ID numbers (assigned by
laboratory), analytical method, sample appearance, sample
treatment, amount of material examined, type and percent of
asbestos present, type and percent of non-asbestos fibrous
and nonfibrous materials present, method of quantitation,
laboratory quality control program, analyst's name and
address, and the school system's return address. A LABORA-
TORY DATA SHEET incorporating this information is provided
in Chapter 4. Send this form to the laboratory with every
set of samples.
QUALITY ASSURANCE MEASURES
Quality assurance is a term used to describe measures
for determining and maintaining laboratory reliability. The
selection of a competent laboratory for the analysis of bulk
samples suspected of containing asbestos is an important
step in the implementation of a successful asbestos program,
and such a selection must be made prudently.
-6-
-------
It is not, however, sufficient to carefully select a
laboratory and then presume that all will run smoothly
throughout the course of the asbestos program. The experi-
ences of several state and local efforts in dealing with
asbestos analysis strongly suggest that additional measures
are not only recommended but even necessary if the program
is to be successful. Thus, the sixth element in an asbestos
program is laboratory quality assurance. Recommendations
for a program of laboratory quality assurance are detailed
in Chapter 3. Flowcharts are provided for three different
situations depending on the number of samples taken.
RECORDKEEPING
The seventh and final element of the asbestos analyti-
cal program relates to proper recordkeeping of the analyti-
cal data collected during the program. Close attention must
be paid to the accurate recording of the -sampling process
and the final disposition of the laboratory reports. The
laboratory analytical reports should be inserted into the
permanent file of the asbestos program. Reports of results
from school surveys should be forwarded to the school dis-
trict office. Additional recordkeeping details are present-
ed in Chapters 2, 3 and 4.
-7-
-------
The preceeding paragraphs have given an overview of a
school asbestos program. Chapter 2 describes the recommen-
ded sampling procedure. A SELECTION OF SAMPLE LOCATIONS
WORKSHEET is provided. Chapter 3 presents recommendations
for a laboratory quality assurance program. Flowcharts
outline the procedures to be followed. In Chapter 4, labo-
ratory reporting and statistical analysis are discussed. A
LABORATORY DATA SHEET and a STATISTICS COMPUTATION WORKSHEET
are provided.
The following CHECKLIST FOR AN ASBESTOS ANALYTICAL PRO-
GRAM provides a chronological list of events that normally
comprise a thorough asbestos analytical program. This list
is provided as a convenient reference for the program coor-
dinator.
-8-
-------
CHECKLIST FOR AN ASBESTOS ANALYTICAL PROGRAM
1. Appoint an Asbestos Analytical
Program Coordinator
2. Establish Program File
3. Inspect For Friable Materials
4. Follow Sampling Protocol {Use
SELECTION OF SAMPLE LOCATIONS
WORKSHEETS]
5. Follow quality assurance protocol
[Use flow charts]
6. Send samples to laboratories [Use
LABORATORY DATA SHEET]
7. Interpret laboratory results [Use
STATISTICS COMPUTATION WORKSHEET]
8. Enter all information in program
file
9. Report to district office
Date Completed
-9-
-------
-10-
-------
CHAPTER 2: SAMPLING FRIABLE MATERIALS
Friable material is material that can be easily crumbled,
pulverized, or reduced to powder in the hand. It may be an
asbestos-containing material, or it may be a material that
contains other fibers, such as cellulose and fiberglass.
Since friable materials crumble easily, it is believed they
have the potential to release fibers readily- For that
reason, it is imperative to determine whether friable mate-
rials contain asbestos fibers and to take corrective action
where necessary.
Friable material may be found on the ceilings of class-
rooms, corridors, auditoriums, cafeterias, machinery rooms,
storage rooms, indoor pools, and gymnasiums. It may also be
found on steel support beams and columns and, occasionally,
on walls and pipes. Neither visual inspection of friable
material nor checking building records can determine the
presence or absence of asbestos. Such a determination must
be made through proper sampling and analysis.
The sampling procedure outlined in this chapter is a
refinement of the methodology presented in Chapter 5 of
Asbestos-Containing Materials in School Buildings; A Guid-
-11-
-------
ance Document, Part 1 [1]. It was developed using recently-
available data and based on standard statistical theory. If
a school sampling program has been completed prior to the
release of this revised sampling procedure, EPA will not
require additional sampling unless there still remains some
question as to the presence or absence of asbestos in the
friable material.
The recommended sampling procedure should be carefully
followed. Improper sampling could result in incorrect
decisions, even when the accompanying laboratory analysis
and quality assurance programs are excellent. Incorrect
decisions would lead either to costly, time-consuming, and
unnecessary corrective action or to no action for poten-
tially hazardous situations. Especially critical to a valid
asbestos analytical program is the use of a random sampling
technique. The importance of this aspect cannot be over-
emphasized.
Since it is clearly not reasonable to remove all the
material from a ceiling to examine for the presence of
asbestos, a few small specimens, a sample of the ceiling
material, is taken. The basis for extending the results of
the sample to the entire ceiling is statistical theory which
assumes random sampling. This is easily explained by an
example.
Suppose a handful of marbles are blindly withdrawn from
a jar full of marbles. If the handful of marbles withdrawn
-12-
-------
is half white and half blue, it would be believed that the
jar contains only white and blue marbles, and that the
composition is approximately half of each color. Since the
selection is random, the composition of the handful of
marbles would be expected to reflect the contents of the
jar. That is, if the jar contained mostly blue marbles,
then selecting marbles purely by chance ought to produce a
mostly blue sample.
On the other hand, if for convenience just the top
layer of marbles are selected and 6 white and 6 blue marbles
are found, not much can be said about the contents of the
jar. It contained at least 6 blue marbles and 6 white
marbles, but that is all that can be said. Thus, this
purposively selected (chosen on purpose) or convenience
sample, does not provide much information about the nature
of the jar's contents. These ideas underlie the concept of
statistical inference.
Given the wide variation in asbestos content observed
in some ceilings, a similar judgmental or convenience sam-
pling method has led to incorrect characterization of the
material. In some cases, the asbestos was entirely missed.
In other cases, it was significantly over-estimated.
SAMPLING PROCEDURE
The recommended sampling procedure includes the fol-
lowing steps:
-13-
-------
- establish an asbestos analytical program file.
- locate all friable materials in the buildings of
concern.
- identify and establish homogeneous Sampling Areas of
friable material.
- diagram each homogeneous Sampling Area reasonably to
scale on graph paper.
- clearly indicate all inaccessible areas and water-
damaged areas in Sampling Areas on diagrams.
- determine the appropriate number of bulk samples to
be taken.
- using a random selection process, select the loca-
tions within all Sampling Areas where bulk samples
will be taken.
- collect the samples, using proper precautions.
- enter all pertinent data in the program file.
The rest of this chapter presents this process in
detail. Worksheets are provided to assist in the somewhat
complicated steps necessary to ensure reliable results. A
detailed example of the sampling process is provided at the
end of this chapter.
ESTABLISHING AN ASBESTOS ANALYTICAL PROGRAM FILE
This step, though apparent, is listed here for empha-
sis. Maintain all worksheets and data forms in a permanent
central file for future reference.
-14-
-------
INSPECTING FOR FRIABLE MATERIAL
The next step in determining the presence or absence of
asbestos is an inspection for friable material. Visually
inspect all areas of the school building including student,
administrative, maintenance, and custodial areas for friable
material. Follow the guidelines for inspection given in
Chapter 4 of Asbestos-Containing Materials in School Build-
ings; A Guidance Document, Part 1 [1]. If friable material
is located during inspection, collect samples of the mate-
rial for laboratory analysis according to the sampling
procedure outlined below.
ESTABLISHING SAMPLING AREAS
Following the inspection for friable material, estab-
lish Sampling Areas. A Sampling Area is defined as a homo-
geneous area of friable material—that is, all friable
material in ct single Sampling Area is of the same type and
was applied during the same time period. A decision as to
the presence or absence of asbestos in the friable material
is necessary for each Sampling Area.
The procedure for establishing Sampling Areas is des-
cribed below. Their proper establishment is extremely impor-
tant as incorrectly established Sampling Areas will yield
results that do not accurately reflect the asbestos content
of the friable material in the school building. This in
turn may lead to very costly and unnecessary corrective
-15-
-------
action, or to no corrective action at all when it is need-
ed.
Partition the total friable material area of the school
building into Sampling Areas. The partitioning will be
based upon visual inspection, knowledge of the school build-
ing's history, and building records, if available.
The following example should clarify the method of
partitioning.
Example: Suppose that friable material is found on the
ceiling of a school's library and on the ceilings of
first floor classrooms of an annex constructed six
years after library construction. The friable material
on the library ceiling appears to all be of one type,
and the friable material on the ceilings of the annex
classrooms appears to all be of a second type. In this
situation, two Sampling Areas are required: (1) library
ceiling and (2) ceilings of first floor classrooms of
the annex. An estimate of the percentage of asbestos
present will be obtained for each of these Sampling
Areas, and separate decisions as to the necessity of
corrective action will be made. If an unacceptably
high percentage of asbestos is found only in Sampling
Area (2), then corrective action needs to be considered
only for that Sampling Area.
DIAGRAM PREPARATION
For each Sampling Area, prepare a diagram showing all
friable materials in the Sampling Area. The diagram should
be constructed on graph paper as follows:
(.1) Clearly indicate the approximate dimensions of all
rooms, corridors, or other school building areas
included in the diagram. If these measurements
are not readily available, rooms will need to be
-16-
-------
measured using a tape measure or by pacing.
Prepare the diagram to scale.
02) Distinguish between friable material areas of the
Sampling Area and areas in the diagram that are
not contained in the Sampling Area.
C3) Draw on the diagram (to scale) any of the follow-
ing features that are found within the Sampling
Area.
(a) Damage caused by water or high humidity.
(b) Damage due to vandalism, rough use, or other
factors.
(c) Patched or repaired material.
(d) Areas that are inaccessible for the purpose
of sampling the friable material.
The reason for noting (a) water damage is that it is appro-
priate to take corrective action for all these areas regard-
less of asbestos content. Information noted in (b) may be
useful in assessing the appropriate corrective action to be
taken if asbestos is found to be present. Inaccessible
areas (d) are marked so that no sample locations will be
selected in these areas.
If one Sampling Area contains friable material areas
that are not adjacent (.for example, areas on different
floors of the school building where the material is the
same), sketch each separate area according to the above
instructions. Place all sketches on the same graph, as
-17-
-------
close together as possible. The Sampling Area may contain
areas that are not in the same plane (for example, a ceiling
and a wall with the same type of friable material). In this
case, sketch each flat surface according to the above instruc-
tions and place these sketches on the same graph, as close
together as possible.
On each Sampling Area diagram, record the following
information:
(1) Sampling Area identification (ID) number. (A
number assigned by the school official to the
Sampling Area that distinguishes the Sampling Area
from all others of the school building.)
(2) Brief description of the Sampling Area.
(3) Area dimensions and scale.
(4) Name and address of the school.
(5) Name and telephone number of the asbestos analyti-
cal program coordinator of the school.
(.6) Name of inspector and date of inspection.
(7) Name of person preparing the diagram and date
prepared.
Include these diagrams in the program file. (Example Samp-
ling Area diagrams are displayed in Figures 2.1 and 2.2.)
NUMBER OF SAMPLES TO TAKE
The number of samples to be collected will be based on
the overall size of the Sampling Area. From the dimensions
-18-
-------
H
reo-
CeiliMj
S dl^Od I
ss
e
: r J r M
M
ft -o r
be.f
o
f
D , a_q r P«>-> P re par e d fc> V
Figure 2.1: Example Sampling Area Diagram
-19-
-------
ft £«* X D
Descripf ion '.
Classroom ANN ex (c0Ns4ru.c4eJ in
ritttle Ceiling |Of\R.Wi «J of F»rs4 Floor Classrooms
Gray "fex-lurcJ spr*y f»Nisl\
S-ttLccaed IN fVppenrAiJce.
ft II ceiltMj OJTCjaus ske.4«Uie,J be/ou)
RreA. f3») ; w/'ff* one
132.'.
1
3
>
/
3
\
4
3
$
Room 101
0'
^ c-^S
°' Room 103
f
^
O ftvivn I/J^T ,
poon\ i
-------
recorded on the Sampling Area diagram, compute the total
square feet in the Sampling Area. Then from the table
below, determine the number of samples to be collected.
If the size Csquare Then the number
feet) of the Sampling of samples to
Area is be collected is
Less than 1,000 3
Between 1,000 and 5,000 5
Greater than 5,000 7
SELECTION OF SAMPLE LOCATIONS
After preparing the diagram(s) and determining the
number of samples to be collected in each Sampling Area,
determine the approximate location of each sample. The
method for selecting sample locations described below uti-
lizes a TABLE OF RANDOM DIGITS. This is designed to elimi-
nate any inadvertent bias which would jeopardize the cor-
rectness of the final decision as to whether or not asbestos
is present. Unfortunately, this method involves a certain
amount of numerical work. No other method of site select-
ion, though, can guarantee unbiased results. Following this
step-by-step procedure carefully will give reliable, unbiased
sample site selections.
Select sample locations according to instructions (1)
through (9) below. It is very important to properly use the
random number procedure to select sample locations. Refer
to the SELECTION OF SAMPLE LOCATIONS WORKSHEET in Figure 2.3
for the following steps.
-21-
-------
-22-
-------
SELECTION OF SAMPLE LOCATIONS WORKSHEET
Sampling Area ID No. (a)
Split-Sample ID No.(s).
Dimensions of rectangle covering the Sampling Area:
Base (b)
Height (c)
RANDOM NUMBERS
First
Second
_feet (Choose first random number
between 0 and no. in (b)}
_feet (Choose second random number
between 0 and no. in (o)}
LOCATION FALLS
WITHIN SA
Yes No
SAMPLE
LOCATION
ID Numbers
Unique
Sample ID #
(For Lab]
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
School Name:
District:
State:
Asbestos Analytical Program Coordinator:
Date:
Figure 2.3: WORKSHEET
-23-
-------
-24-
-------
(1) Record the Sampling Area ID Number on the SELEC-
TION OF SAMPLE LOCATIONS WORKSHEET (a).
(2) Construct on the Sampling Area diagram an imagi-
nary rectangle enclosing the entire Sampling Area.
Record the dimensions of this imaginary rectangle
on the WORKSHEET: first the number of feet along
the rectangle base (b) and then the number of feet
along the side or height (c).
(3) From the TABLE OF RANDOM DIGITS, choose a pair of
random numbers. Record the random numbers on the
WORKSHEET. A TABLE OF RANDOM DIGITS and instruc-
tions for its use are provided in Appendix C. The
first random number of the pair should be between
0 and the number of feet along the rectangle base
Cb). The second random number of the pair should
be between 0 and the number of feet along the side
or height (c) of the rectangle.
(4) The random number pair describes a location within
the rectangle. The first number of the pair
specifies the number of feet from the left side of
the rectangle, and the second number specifies the
number of feet from the bottom of the rectangle.
The point should be plotted on the Sampling Area
diagram.
C5) If the point described by the random number pair
is within the Sampling Area and not within any
-25-
-------
area designated on the diagram as inaccessible for
the purpose of sampling, then that point is a
sample location. Otherwise, the point is not a
sample location. This elimination of inappropri-
ate random number pairs does not adversely affect
the random selection process so long as the pairs
are chosen in continuous sequence. (If the random
number selection process is done off-site, select
some extra pairs of random numbers in case one or
more are later found to be inaccessible. The
SELECTION OF SAMPLE LOCATIONS WORKSHEET provides
room to select 12 pairs of random numbers.)
(6) Continue using the above random number pair proce-
dure until at least the required number (3, 5 or
7) of appropriate sample locations have been
selected.
(7) All random number pairs should be recorded on the
WORKSHEET. Beside each random number pair, indi-
cate by a check if the location the pair describes
is within the Sampling Area (and not within any
area designated as inaccessible for the purpose of
sampling) and is thus a sample location.
C8) Assign a sample location number to each of the
sample locations. Any system of numbers that
assigns a unique number to each sample location is
satisfactory. Record these location numbers on
-26-
-------
the WORKSHEET, and on the Sampling Area diagram.
(9) At the same time, assign a non-systematic but
unique sample ID number to each sample location.
This ID number, not the sample location number,
will be on the sampling container when it goes to
the laboratory for analysis. Using unique non-
systematic numbers will prevent the laboratories
from knowing which samples come from the same
Sampling Areas or the same buildings. This "blind"
procedure helps prevent bias on the part of the
analyst. Record these unique sample ID numbers
(for laboratories) on the WORKSHEET. Choosing
numbers from the TABLE OF RANDOM DIGITS is a quick
and easy technique for assigning sample ID num-
bers.
SAMPLE COLLECTION
Sampling containers should be small, sealable tin,
metal or plastic containers. Suggested sampling containers
are plastic 35 mm film canisters or small wide-mouthed
aspirin bottles. Prior to sampling, thoroughly clean a
sufficient number of sampling containers.
Collect the bulk samples, i.e., samples taken from the
friable material by penetrating the depth of the friable
material, at the specified locations according to the fol-
lowing guidelines:
-27-
-------
(1) Gently twist the open end of the sampling con-
tainer into the material. A core of the material
should fall into the container. A sample can also
be taken by using a clean knife to cut out or
scrape off a small piece of the material and then
placing it in the container. Be sure to penetrate
any paint or protective coating and all the layers
of the material. If the sampling container cannot
penetrate the material, consider whether the
material is really--friable, or not.
(2) Tightly close the sampling container; wipe its
exterior with a damp cloth to remove any material
which may have adhered to it during sampling.
(3) Tape the sampling container cap to prevent the
accidental opening of the container during ship-
ment or handling. In addition, it is recommended
that each container be placed in a sealed plastic
bag because film canister caps, even when taped,
may come off in transport.
(4) Record the unique sample ID number chosen in (8)
above on a label and tape the label to the cor-
responding sampling container. Be sure to record
the unique sample ID numbers on the WORKSHEET as
part of the asbestos analytical program file.
(5) See Chapter 3 for laboratory quality assurance
procedures.
-28-
-------
Collect samples at Cor as close as possible to) the
selected locations and collect all samples. Exact measure-
ments (i.e., by ruler) are not necessary for finding the
sample locations. Quicker, easier techniques such as pacing
may be employed.
PRECAUTIONS TO BE TAKEN DURING SAMPLING
To avoid causing unnecessary exposure to asbestos
fibers, take the following precautions while sampling fri-
able materials [1].
Cl) Sample the material when the area is not in use.
(.2) Have only those persons needed for the sampling
present.
C3) Hold the sampling container away from the face
during actual sampling.
(4) Do not disturb the material any more than neces-
sary.
C5) Spray the material with a light mist of water to
reduce fiber release during sampling.
C6) If a large number of samples are taken, NIOSH
recommends that the sampler wear an approved
respirator. Contact a NIOSH Regional Office for
information on approved respirators [1J.
(7) Wear a respirator if moving ceiling tiles or in
any other way disturbing possible fallen asbestos
or its debris.
-29-
-------
(8) If pieces of material break off during sampling,
wet mop the areas where they have fallen.
AN ILLUSTRATION OF THE SAMPLING PROCEDURE
The sampling procedure is illustrated by this example.
A school was visually inspected for friable materials. Five
classrooms in an annex were found to contain suspect ceiling
materials. All the materials were believed to be the same
and thus comprise one Sampling Area (SA).
Approximate room dimensions were obtained by pacing and
diagrammed as shown in Figure 2.4. Pertinent information
such as the location of damaged, non-friable, and inaccessi-
ble materials was diagrammed and labelled.
The total area of friable materials in the five class-
rooms determines the number of samples to be taken from the
SA. The example SA is 10,080 square feet, as calculated by
Area = [60' x 90'] + [121 x 90'] + 160' x 60']
= 10,080 square feet.
Since this area is greater than 5,000 square feet, seven
samples are required.
The SA ID number and the base and height of the SA were
recorded on the SELECTION OF SAMPLE LOCATIONS WORKSHEET.
The completed WORKSHEET is shown in Figure 2.5. The example
SA's base is 132' (Note: 3-digit number) and height is 90'
-30-
-------
C I ASS room rtNNex CCON
Fricttle CeiliM^ nlfvltri
G^rftV ^ex-Kcred
_ . f I •
ft,
->„ l-.,^ l-terA I D -# ^
•^Jflmpl'N^ r\K.^Pt ' ^ ^
Ass room rtNNex CcoNs-fru.c4eJ in I ^2.)
Ceiling |T1 frier i«J of Firs+ Floor Classrooms
Grriw "f ex-lured spr*y -fiNis^
$-hu.cCdea IN f^ppeitr ntJc£.
r^
of
1
1
3
>
/
2
^
s
3
8Z
^ J
*N 1 1
^ Room 101
0/*- *e
3?
iCSiOl
0/ Room 103
1 x*-°3
f
1
,O PA/I»W I/J^T
pwOni |V^
«K ^X 1
L
»
•
L
»
•
r
J
•
<
%
w
^^
[ 1
^
1
L 1
IB^VH
••
1
^
^
*
L
Room 102. #*
^
r
[
^^Room 104-
•
r
^t ,n/
Sj|10 5r
V -JC
1 C~f
-------
SELECTION OF SAMPLE LOCATIONS WORKSHEET
Sampling Area ID No. (a)
Split-Sample ID No.(s)
imensions of rectangle covering the Sampling Area:
Base (b) 133* feet (Choose first random number
between 0 and no . in (b) }
Height (.c) \ O feet (Choose second random number
between 0 and no. in (c))
LOCATION FALLS SAMPLE Unique
RANDOM NUMBERS WITHIN SA LOCATION Sample ID #
(1)
First
£ /
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
11
10$
10
11^
32
)<+
£7
Second
S.I
^^
12
41
7G
(*(o
7/
/2
res
/
/
^X
y
^
^"
I/
^Jo
V
ID Numbers
3,01
AOZ
3(03
3iO
-------
(Note: 2-digit number). The seven sample locations were
chosen using the TABLE OF RANDOM DIGITS in Table 2.1. The
first random number of each pair had to be between 0 and
132, and the second random number of each pair had to be
between 0 and 90. Beginning in the upper left-hand corner
of the random number table, the following 3-digit random
numbers were crossed off as they are greater than 132: 632,
715, 998, 671, 744, and 511. Then the digits 021 were
circled and used in the first random number pair since the
number 21 is between 0 and 132.
To find the second random number of this pair, 2-digit
numbers were considered. The digits 51 were circled since
51 is between 0 and 90. The first random number pair (21,51)
was recorded on the SELECTION OF SAMPLE LOCATIONS WORKSHEET.
This procedure was repeated until seven pairs of random
numbers were selected.
After seven pairs of random numbers were selected and
recorded on the SELECTION OF SAMPLE LOCATIONS WORKSHEET,
these numbers were plotted on the SA diagram which was drawn
to scale on graph paper (see Figure 2.4.). In the example
SA, the first random number pair is (21,51). This desig-
nates the point 21 feet from the left side of the rectangle
and 51 feet from the bottom of the rectangle. Since that
point (21,51) is within the SA, it is a valid sample loca-
tion and was marked on the diagram. If a random number pair
designates a point within the rectangle that is not within
-33-
-------
TABLE OF RANDOM DIGITS
*#&# '.
55363
69393
13186
17726
36520
81628
84649
63291
70502
06426
20711
41990
72452
37042
53766
90585
32001
62606
10078
91561
13091
73864
66668
84745
48068
54310
14877
78295
67524
58268
97158
04230
94879
71446
32886
62048
84534
84707
19409
57978
57295
£?<£3--
07449
92785
29431
28652
64465
36100
48968
11618
53225
24771
556.09
70538
36618
40318
52875
58955
96293
64324
28073
46145
98112
83014
25467
41042
26805
96175
33095
23179
02865
57219
28672
16831
56606
15232
05644
33711
42351
15885
40868
48015
98298
94044 83785
30014 25879
07265 69563
84404 88642
21778 02085
^«c;
34835
49902
88190
56836
05550
39254
75215
12613
03655
59935
29430
77191
76298
57099
15987
53122
37203
46354
85389
24177
53959
72457
48894
29493
94595
97594
10924
02771
39593
68124
50685
69085
30401
66715
79316
25290
21628
84710
64220
25973
11199
93388
71763
64268
30263
27762
3&2te
15290
58447
04588
78351
30157
56835
75498
75055
05915
49801
70165
25860
26678
10528
46962
16025
64516
72157
50324
15294
79607
22682
51043
01836
47907
88616
58013
43464
54278
73455
01181
30802
02602
26385
09819
21526
53669
35866
80861
66777
96510
07833
96679
88802
80310
46097
45573
76616
42048
38733
47327
82242
37636
49539
43915
37140
11082
45406
55204
89334
09925
67342
84299
51530
67248
14500
10061
52244
03033
02365
09044
13357
42035
61439
59061
04237
83236
24262
65559
57658
91518
00813
02223
81352
06446
13860
45924
75228
38216
90603
72264
11522
43324
84358
67191
30378
81290
18518
29520
02421
74240
26488
57051
66762
78484
73417
33938
89773
77592
53310
37069
20135
15562
98124
63303
61714
91726
51926
38412
38093
21882
71411
92441
08710
19427
09205
70091
70566
88407
75947
95152
86311
68493
56144
41600
31413
99396
66540
57810
34354
*• — ^ ~
21625^
12777
87618
89541
92222
69753
98063
03466
41116
48393
94477
31639
83920
95567
41335
57651
67380
40261
49804
64165
75732
10413
88173
09365
43630
33318
36745
42059
05697
26602
04284
52106
71829
54986
02888
17461
66466
08107
32648
52908
24742
47192
70555
74557
01782
27627
09369
16999
21861
26933
70290
55201
72602
89641
49292
64531
91322
02494
52009
69468
29380
96244
95508
84249
61374
09226
06125
00815
63839
90835
63167
63470
26098
56702
24177
67194
63835
55005
34308
06498
41394
79941
73925
06232
98814
88141
26374
96702
43267
03023
74224
08396
78376
14966
13385
68689
40640
40113
27340
23756
64953
36401
56827
25553
88215
18873
74972
75906
29002
80033
25348
05815
64419
71353
83452
74762
00634
95264
76508
82782
40644
58739
30495
38032
84171
73685
85650
60437
39684
53037
10913
72743
73902
63297
88200
35973
54147
18211
19251
36240
10158
PV" ~~tr-
22782
03263
16281
08243
10493
54935
99337
45525
30825
06543
27191
96927
38712
91807
46453
69828
04332
06714
29457
77669
97355
50289
85169
45643
14194
42851
83514
60170
21157
94770
42596
74246
38707
03195
54315
91904
75336
12849
69981
45052
66162
23152
06647
91637
83613
48952
76089
Table 2.1: TABLE OF RANDOM DIGITS Used For the Example
-34-
-------
the SA, it is not used as a sample location and the next
random number pair is used. In the example SA, the third
random number pair selected (108,13) designated a point
outside of the SA and was disregarded.
The plotting procedure was continued until a total of
seven valid sample locations were specified on the diagram.
Each sampling location was assigned a sample location num-
ber. (Any simple numbering system can be used so long as
each sample location within a SA receives a unique number.)
In the example SA, sample locations were numbered from 201
to 207. The unique non-systematic sample ID numbers for
laboratory submission were randomly selected from a TABLE OF
RANDOM DIGITS.
-35-
-------
-36-
-------
CHAPTER 3: LABORATORY QUALITY ASSURANCE
As previously stated, a rigorous quality assurance (QA)
program is important to ensure the reliability of results
from laboratory analyses. Once a laboratory is chosen from
the list in Appendix A, or from subsequent laboratory lists
provided by EPA, the implementation of split-sample tech-
niques outlined in this chapter is recommended to monitor
laboratory output on a regular basis. If it is not possible
to choose a laboratory from this list, Appendix B contains a
procedure for evaluating the performance of an unknown
laboratory.
A quality assurance program need not be pursued on an
individual school basis. In fact, measures described here
would be much more effective and useful when applied by a
school district or other collection of schools. By utiliz-
ing a coordinated quality assurance program, officials from
many schools or districts could greatly improve their
chances of receiving reliable and consistent laboratory
results.
-37-
-------
jjlUT-SAMPLE TECHNIQUES FOR QUALITY ASSURANCE
To carry out a successful quality assurance program
requires that a certain proportion of samples of suspect
materials be split. Therefore, take each bulk sample in
large enough quantity to allow it to be divided into two
equivalent portions. The two portions are referred to
collectively as a split-sample; that is, a split-sample is
one sample that has been split into two equivalent parts.
The two portions of a split-sample must not differ from
each other in any substantial way, and the amount of mate-
rial taken for each part of the split-sample should be the
same as for regular samples in the overall program. One
simple method used for split sampling is to take two samples
immediately adjacent to each other and use each as a portion
of the split-sample. However, it is crucial that the two
—i^plits" be equivalent material.
Pack the two parts of the split-sample separately in a
manner similar to other samples. Affix a unique sample ID
number to each sampling container so the laboratory will not
know which samples are split-samples. Record the unique
sample ID numbers so results can be accurately compared.
The two portions of a split-sample must be analyzed
independently of each other, and the results from the two
analyses should then be compared. The two samples might be
analyzed by two different laboratories. One would be the
laboratory that is to carry out the overall program and the
-38-
-------
second would be a qualified laboratory, either a commercial
facility (perhaps a back-up laboratory in the event the
other falters and proves unreliable) or a noncommercial
facility that is willing to do a small number of samples on
a limited basis but unable to handle a large volume.
Alternatively, both parts of the split-sample may be sub-
mitted to the same lab.
If the two parts of a split-sample are not equivalent,
then the results reported by the laboratories may correctly
differ significantly one from the other and pose difficult
interpretation problems for the school official. The fail-
ure of the laboratories to agree in their analyses of the
split-samples may require that all samples (split-samples
and non split-samples) be reanalyzed.
The laboratory reports for the split-samples will state
whether asbestos is present or absent in that sample. It is
this point of the analytical report - the presence or absence
of asbestos - which should serve as the first focal point
for the school official in deciding whether the laboratory
performance is acceptable or not for the program.
To aid in this decision, the following procedure is
recommended: The laboratory is performing satisfactorily if
the number of disagrggmjgnjbs (as to presence or absence of
asbestos) between the results of the two portions of the
split-samples is less than a predetermined critical number.
This critical number depends on the number of split-samples
-39-
-------
analyzed. The following table, based on certain statistical
considerations [2], gives the critical number of disagree-
ments for different numbers of split-samples.
Critical Number of Split-
Number of Split-Samples Sample Disagreements
5 2
6 to 8 3
9 to 14 4
15 to 20 5
21 to 25 6
According to this table, if two or more disagreements are
observed in five split-samples, or three or more disagree-
ments are observed for six to eight split-samples, etc.,
then the laboratory's performance is suspect. On the other
hand, if the number of disagreements is less than the crit-
ical number, the laboratory's performance is satisfactory -
However, even if the laboratory ijs satisfactory, all split-
sample results which do not agree must be resolved.
CONTINUING QUALITY ASSURANCE PROGRAM
As previously stated, the procedure for monitoring
laboratory performance on an on-going basis includes a
certain proportion of split-samples for the purpose of
quality assurance. The number of split-samples with dif-
ferences between the two parts of the split-sample will be
monitored as described to determine whether laboratory
performance is acceptable.
The number of split-samples that the school official
will take will depend on the expected number of samples from
-40-
-------
the school system. If an unacceptable number of disagree-
ments is noted, then there is reason to suspect the labora-
tory's performance. In such a situation, it is suggested
that the school official ask the laboratory to investigate
and correct the cause(s) for these discrepancies. All
samples analyzed during the period the laboratory's proce-
dure is suspect must be reanalyzed. Again, even if the
laboratory is considered satisfactory, all split-sample
results which do not agree must be resolved.
The split-samples in a school should be divided among
the sampling areas. That is, all splits in a school should
not come from the same Sampling Area, assuming that there is
more than one Sampling Area in a school.
The following flowcharts summarize the split-sample
procedures under situations with different total numbers of
samples. Figure 3.1 (Case 1) is for school programs with
less than 25 samples; Figure 3.2 (Case 2) is for school
programs with between 25 and 100 samples; and Figure 3.3
(Case 3) is for school programs with greater than 100 sam-
ples.
Note, it should be clear in examining the figures that
there are several advantages if there is one statewide
authority for the purpose of quality assurance monitoring.
The effort needed in monitoring will be considerably re-
duced.
-41-
-------
Figure 3.1
Case 1:
Split-Sample Procedures For School Systems For
Which Fewer Than 25 Samples Will Be Analyzed
Take a minimum of 5 split-samples for the purpose of
monitoring. If the number of split-sample disagreements is
2 or more, the laboratory's performance is suspect. All the
samples analyzed during the period when the laboratory's
procedure was suspect should be reanalyzed after the labora-
tory resolves the problems.
If the number of split-sample disagreements is less
than two, then the lab performance is considered satisfac-
tory. However, all split-sample disagreements should be
resolved. No further monitoring is required for that labo-
ratory.
-42-
-------
QA
Monitoring
Have 5
Split-Samples
Analyzed
Is the
Number of
Disagreements
Less Than
2?
No
Yes
Have Lab
Resolve the
Problems
Have All
Samples
Reanalyzed
No Further
Monitoring
Required
Laboratory Is
Satisfactory
Resolve All
Split-Sample
Disagreements
No Further
Monitoring
Required
Figure 3.1
Case 1: School Systems For Which Fewer
Than 25 Samples Will Be Analyzed
-43-
-------
Figure 3.2
Case 2:
Split-Sample Procedures For School Systems With Expected
Number of Samples Over 25 But Not Over 100
One out of every 5 samples should be split. The re-
sults of each consecutive set of 5 split-samples should be
monitored. If 2 or more disagreements are noted, the labo-
ratory must correct the problems and all samples analyzed
during the period when the lab procedure was suspect should
be reanalyzed.
Figure 3.2 assumes sets of 5 split-samples. If the
number of split-samples in a given set is not a multiple of
5, then use the appropriate critical number in the table to
determine the unacceptable number of disagreements.
-44-
-------
QA
Monitoring
Let One of 5
Samples Be
Split-Sample
Monitor a Set
of 5 Split-
Samples
Resolve All
Split-Sample
Disagreements
^
r
Continue
Monitoring
If More
Samples
Remain
Number of
Split-Samples
in Set
5
6-3
9-14
15-20
21-25
Critical
Number
2
3
4
5
6
Is The
Number of
Disagreements
Less Than
2?
Have Lab
Resolve
The Problems
Have All
Samples
Reanalyzed
Figure 3.2 Case 2:
School Systems With Expected Number of
Samples Over 25 But Not Over 100
-45-
-------
Figure 3.3
Case 3:
Split-Sample Procedures For School Systems
With Expected Number of Samples Over 100
Among the first 100 samples, one out of every 5 samples
should be a split-sample. For these split-samples, the re-
sults of each consecutive set of 5 split-samples should be
monitored. If 2 or more disagreements are noted, the lab
procedure is suspect. All the samples analyzed during the
period when the lab procedure was suspect should be rean-
alyzed after the laboratory resolves the problems.
After the first 100 samples, the split-sample rate may
be reduced to 1 in every 10 samples. Then the results of
each consecutive set of 20 split-samples should be monitor-
ed. If 5 or more disagreements are noted in the results of
any set, the lab procedure may be suspect. If less than 20
split-samples comprise a set, use the appropriate critical
number in the table to determine the unacceptable number of
disagreements. All the samples analyzed during the period
when the lab procedure was suspect should be reanalyzed
after the laboratory resolves the problems.
Number of
Split-samples
In Set
5
6-8
9-14
15-20
21-25
Critical
Number
2
3
4
5
6
-46-
-------
QA
Monitoring
Let 1 of 10
Samples Be
Split-Sample
Let One of
5 Samples Be
Split-Sample
(tfirst 100)
Monitor a Set
of 20 Split-
Samples (If
Possible)
Monitor a
Set of 5
Split-Samples
Is The
Number of
Disagreements
Less Than 5? (Or
Appropriate
Critical
Number)
Have Lab
Resolve the
Problems
Is The
Number of
Disagreements
Less Than
2?
Have Lab
Resolve The
Problems
Have
Samples
Reanalyzed
Have Samples
Reanalyzed
Resolve All
Split-Sample
Disagreements
Resolve All
Split-Sample
Disagreements
Continue
Monitoring
If More
Samples
Remain
Is the
Number of
Samples
Already
Monitored
at
Least 100
Figure 3.3 Case 3:
School Systems With Expected Number of
Samples Over 100
-47-
-------
-48-
-------
CHAPTER 4: LABORATORY ANALYSIS AND STATISTICAL ANALYSIS
Now that the asbestos analytical program is established
and samples have been collected, the "blind" samples are
sent to a qualified laboratory for analysis. To ensure com-
pleteness and consistency in the data received from labora-
tories, this chapter provides a LABORATORY DATA SHEET which
should accompany all samples.
When the analytical results are received, the asbestos
analytical program coordinator will need to interpret them.
A simple statistical analysis technique is provided to
facilitate this process.
The calculation of 90% confidence intervals as outlined
below provides interval estimates of the asbestos content in
each Sampling Area. The meaning of this estimate can be
described as follows. If sampling is repeated 100 times and
a confidence interval is calculated each time, then 90 of
the 100 confidence intervals for the Sampling Area will con-
tain the true average concentration of asbestos in that
Sampling Area.
-49-
-------
FORWARDING SAMPLES TO LABORATORY
It is very important to obtain complete and accurate
information from the analytical laboratories. To ensure
that goal, a sample laboratory reporting form is provided.
Pertinent requested information includes:
(1) The Sample Identification Number
The analyst should not know whether he or she is
running a routine sample or a sample which has
been selected for split-sample analysis. To
assure that all samples are run "blind", label the
sampling containers with only the unique sample
identification number assigned by the asbestos
analytical program coordinator. Do not use sample
location numbers.
(2) The Analytical Method(s) Used In the Analysis
The method of choice is polarized light microscopy
with or without dispersion staining and X-ray
diffraction as appropriate (see [3] and Appendix H
of [1]).
(3) Sample Appearance
A comment should be made on the homogeneity of the
sample and the steps taken to assure that proper
analytical sampling was employed. If the analyst
selects only fibrous-looking particles from the
sample submitted for analysis, he or she is apt to
miss small pea-like coated aggregates of asbestos
-50-
-------
which were formed during the spray-on procedure at
the time of application. Improper sampling by the
analyst may result in a false negative, i.e.,
reporting no asbestos present when it is in the
sample. Several slides may be required for accu-
rate analysis of the sample.
(4) Sample Treatment
Some analysts prepare the sample by grinding or
washing prior to microscopic analysis. This
process should be briefly described.
(5) Amount of Material Examined
The analyst should estimate the total number of
milligrams of material examined.
(6) Type and Percent of Asbestos Present
The analyst should identify all asbestos fibrous
materials and estimate the percent of each type
present and specify associated precision.
(7) Percent Total Asbestos Present in Sample
The analyst should record the percent of total
asbestos present Call types) in the sample.
(8) Type(s) and Amount(s) of Other Fibrous Materials
The non-asbestos fibrous materials should be iden-
tified by the analyst as to type(s) with an esti-
mate of the amount of each type present. The pre-
cision to be associated with the percentage reported
should also be specified. The basis for that
-51-
-------
judgment and characterization should be provided.
Such verification may help to minimize the report-
ing of false negatives (i.e., reporting asbestos
as cellulose) or false positives (i.e., reporting
fiberglass as chrysotile).
(9) Type(s) and Amount(s) of Nonfibrous Material
Present
The nonfibrous materials should be identified by
the analyst with an estimate of the amount of each
present.
(10) Description of Method of Quantitation
The analyst should briefly describe the quantita-
tion technique employed. ,.--'' ~^-\^
^
(11) Description of Laboratory's Quality Control •
Program v--.___ ../. -
The laboratory should give appropriate comments on
their in-house "good laboratory practices" that
provide quality control for their PLM analyses in-
cluding the number of slides per sample and the
number of splits per set.
-52-
-------
LABORATORY DATA SHEET
Sample ID //
Laboratory Sample ID #
Analytical Method 1. PLM
(enter number) 2. PLM + dispersion staining
3. X-ray diffraction
Gross Sample 1. Homogeneous, fibrous
Appearance 2. Homogeneous, nonfibrous
(enter number; 3. Heterogeneous, fibrous
note color) 4. Heterogeneous, nonfibrous
5. Heterogeneous, mixed
Sample Treatment 1. Homogenized
(enter number) 2. Untreated
3. Other, specify
Total Amount of Material Examined (mg)
Asbestos Present 1. Amosite
(enter number and 2. Chrysotile
percent) 3. Crocidolite
4. Other, specify
Percent Total Asbestos Present in Sample
Other Fibrous 1. Fiberglass
Materials Present 2. Mineral Wool
(enter number and 3. Cellulose
percent) 4. Other, specify
Nonfibrous Materials Present (enter description
below by number and percent in columns)
1.
2.
3.
4.
5.
(Jl
(Continued: Please provide requested information on reverse side of this form.)
-------
Description of Method of Quantitation
Description of Quality Control Program (e.g., # siides/sample, # splits/set)
Comments
Analyst's Name:
Analyst's Telephone Number:
Confirmation By:
Report Reviewed By:
Address Correction, Please:
Return Laboratory Forms to:
-54-
-------
STATISTICAL ANALYSIS OF LABORATORY RESULTS
Because of the imperfection of the analytical tech-
niques currently available for bulk sample analysis and the
heterogeneous nature of these friable materials, it is not
unusual to get a range of results reported by the laboratory
on samples taken from within a single Sampling Area. It
would be unusual for the asbestos concentrations for all
three, five, or seven samples reported to be within a few
percent of each other.
For that reason, a statistical analysis must be per-
formed to compute a confidence interval for the average
percentage of asbestos within a Sampling Area. The proce-
dure outlined below was developed to be simple and easy-to-
follow and is accompanied by a brief explanation.
Use the confidence interval to reach a conclusion as to
the presence or absence of asbestos in the Sampling Area.
If the entire confidence interval is below 1%, then conclude
that asbestos is not present. If the entire confidence
interval is above 1%, then conclude that asbestos is present.
If the confidence interval contains the value 1%, then there
is still some question as to the presence or absence of
asbestos in the Sampling Area.
Note that the confidence interval's upper bound can be
interpreted as a maximum probable value. To narrow the
confidence interval, it would be necessary to take a second
series of samples. In most situations, obtaining a more
-55-
-------
precise estimate when the maximum probable value is small
would not normally warrant the additional expense involved
in a second series of samples.
If it is concluded that asbestos is present in the
Sampling Area, then the potential exposure of users of the
building should be evaluated and a decision on corrective
action should be made according to the guidelines presented
in Asbestos-Containing Materials in School Buildings: A
Guidance Document, Part 1 [1J.
A STATISTICS COMPUTATION WORKSHEET and instructions are
provided in this section. A completed WORKSHEET is also
shown. In these calculations, use the value for percent
asbestos from the line on the LABORATORY DATA SHEET titled
"Percent Total Asbestos Present in Sample". Compute one
confidence interval for each Sampling Area.
-56-
-------
INSTRUCTIONS FOR STATISTICS COMPUTATION WORKSHEET AND EXAMPLE
I. List the results of the laboratory analysis of the
friable material for the Sampling Area. There should
be 3, 5 or 7 observations to enter in column (A). The
order of entry is unimportant. In column (B) enter
the square of the number in column (A). Sum both
columns.
For Example;
(A) (B)
No. Percent Asbestos Squares of Column (A)
1
2
3
4 SS 625*
5 £& I^ZS
60 6
1
(A)
15* %
(B)
II. Enter the number of observations. Should be 3, 5, or 7
For Example; From above N = 7
III. The mean is the sum of the observations divided by the
number of observations.
(A) I(A) ISA 7*\
For Example; Mean =
N N = H
IV. The variance (V) is a measure of precision and is cal-
culated according to the formula below.
1 (A} ^
For Example; V = ~r [ (B) - ^' ]
7 *
- 3300.S-7]
= 733.^.3
V = /3<3.5-7
-57-
-------
V. The standard deviation (SD) is the square root of the
sample variance.
For Example: SD =-Vv = \Jl3o .SI
VI. The half-range (HR) depends on the number of observa-
tions, and is the product of the standard deviation
(SD) and one of the following constants.
If N = then HR = (SD) times
3 1.69
5 0.95
7 0.73
For Example; HR = CSD) /|.*f3 % x o.73 (N = 7)
VII. Confidence bounds are the upper and lower limits of the
confidence interval and are found by subtracting the
half-range from the mean to obtain the lower confidence
bound, and adding the half -range to the mean to get the
upper confidence bound.
For Example ;
Lower Confidence Bound = (Mean) SLI.1I % - (HR) 8.3*/ %
= (LCB) J3.37 %
Upper Confidence Bound = (Mean) 3/.7/ % + (HR) g»3V %
= (UCB) 36.05* %
VIII. The 90% Confidence Interval (CI) consists of all the
numbers between the upper and lower confidence bounds .
For Example: 90% CI = ( (LCB) j_3137_% , (UCB) 30.0^ % )
NOTE: Negative values can be achieved through this
statistical analysis process. Since there obviously
cannot be a negative concentration of asbestos in a
ceiling, all negative values should be considered
equivalent to zero.
IX. If the 90% CI Then Conclude
is below 1% asbestos absent
is above 1% asbestos present
contains 1% uncertain
For Example: The 90% CI (13.37%, 30.05%) is completely
above 1%, conclude that asbestos is present,
X. The maximum probable value equals the upper confidence
bound (UCB), in this case, ^o-OS" %.
-58-
-------
STATISTICS COMPUTATION WORKSHEET
SAMPLING AREA ID #
No.
1
2
3
4
5
6
7
Percent Asbestos
(B)
Squares of Column (A)
(.A)
(B)
Total
II. No. of observations (3, 5, or 7), N =
Sum of Column (A)
III.
IV.
Mean = =
N Total No. Observations
Variance (.V) =
N-l
N
(B)
(A)
V.
Standard Deviation (SD) =
V
VI.
VII
The half-range (HR)
If N = then HR = (SD) times
3
5
7
HR = (SD)
1.69
0.95
0.73
% x
Confidence Bounds
Lower Confidence Bound (LCB)
Upper Confidence Bound (UCB) =
-59-
Mean - HR
Mean
(LCB)
- HR
Mean + Half-range
Mean % + HR
(UCB) %
(continued)
-------
VIII. The 90% Confidence Interval
90% CI = [ (LCB) , (UCB) ] = [ %, % ]
IX. The 90% CI Q is below 1%. Conclude asbestos absent.
|| is above 1%. Conclude asbestos present.
|| contains 1%. Uncertain.
X. Maximum Probable Value = (UCB) = %.
-60-
-------
EXAMPLE STATISTICS COMPUTATION WORKSHEET
SAMPLING AREA ID #
I.
No.
1
2
3
4
5
6
7
(A)
Percent Asbestos
(B)
Squares of Column (A)
30
LAS'
(A)
(B)
Total j
II. No. of observations (.3, 5, or 7), N = "J_
III.
V.
VI.
VII
Sum of Column (A)
M = (A) _
N Total No. Observations
% = ;*/.?/%
IV. Variance (.V) = ^=- [ (B) -
N
]
= 130.57
Standard Deviation (SD) =
13
-------
VIII. The 90% Confidence Interval
90% CI = [ (LCB) , (UCB) ] = [ ;3.37 %, 2>Q.QS~ % )
IX. The 90% CI [~~| is below 1%. Conclude asbestos absent.
^>
-------
REFERENCES
[1] Asbestos-Containing Materials in School Buildings: A
Guidance Document, Part 1. Office of Pesticides and
Toxic Substances, United States Environmental Protec-
tion Agency, March 1979.
[2] Lucas, D., A. V. Rao and T. D. Hartwell, Asbestos-
Containing Materials in School Buildings: Guidance for
Asbestos Analytical Programs. Background Document.
Research Triangle Institute, Research Triangle Park,
North Carolina, December 1980.
[3] Interim Method for the Determination of Asbestiform
Minerals in Bulk Insulation Samples. Environmental
Monitoring Systems Laboratory and Office of Pesticides
and Toxic Substances, United States Environmental Pro-
tection Agency, June 1980.
[4] Brantly, E.P. Jr., and D.E. Lentzen, Asbestos Contain-
ing Material in School Buildings: Bulk Sample Analysis
Quality Assurance Program. EPA-560/13-80-23. Office
of Pesticides and Toxic Substances, United States En-
vironmental Protection Agency, August 1980.
-63-
-------
-64-
-------
APPENDIX A: EPA-SpoNSORED ANALYTICAL PROFICIENCY PROGRAM
FOR ASBESTOS BULK SAMPLE ANALYSIS
-65-
-------
APPENDIX A
EPA-Sponsored Analytical Proficiency Program
For Asbestos Bulk Sample Analysis
Growing public concern with the effects of exposure to
asbestos fibers has resulted in a greatly increased demand
for laboratory analysis to determine the content of bulk
insulation samples. In the course of the Environmental
Protection Agency school asbestos program, many differences
have been noted in analytical services contracted for by
public school systems. Discrepancies among laboratories may
be attributed to variations in analytical methods, lack of
appropriate reference standards, and inadequate reporting of
analytical results.
Polarized light microscopy (PLM) is the EPA method of
choice for detecting asbestos in bulk insulation samples
[1]. EPA is sponsoring an analytical proficiency program
directed at qualifying, to a limited extent, the services
provided by commercial laboratories claiming capability in
PLM analysis. Commercial and noncommercial laboratories
were invited to participate in the program. Accepting
laboratories were provided with four characterized samples
and their analytical reports were compared with reference
analyses. This was not an accreditation program and did not
seek to certify or endorse participating laboratories. A
-66-
-------
performance rating based on a fairly lenient criterion was
determined for each laboratory.
Laboratories had been notified at the start of the
project that such a rating would be made. Participation in
the program was required for laboratories to be included on
the published listing.
Four bulk samples were sent to each laboratory. Two
contained asbestos fibers, anthophyllite and chrysotile, and
two were non-asbestos fiber materials, mineral wool and fi-
berglass, commonly found in insulations. The samples were
doublebagged, coded, and packaged with a reporting form and
instructions for analysis. Sample packages were mailed on
December 28, 1979, to all laboratories on the source list-
ing.
Seventy-one percent of the laboratories contacted
reported results including 52 of 72 commercial labs and 23
of 34 noncommercial labs. Results included were received on
or before January 25, 1980. For the 300 (75 x 4) samples
analyzed, no false negatives and only two false positives
were reported. Mineral wool CSample 1) was incorrectly
identified by one laboratory as crocidolite and by another
laboratory as amosite. The other 73 laboratories correctly
identified Sample 1 as either mineral wool, fiberglass or
glass wool.
Anthophyllite-asbestos was frequently misidentified as
either amosite (.15 labs) or tremolite (10 labs) . This was
-67-
-------
most likely due to unfamiliarity with anthophyllite-asbestos
because no standard reference samples exist and it is not
commonly found in insulation materials. Fiberglass was
identified as fiberglass, mineral wool, or glass wool by all
laboratories. Chrysotile was properly identified by all
laboratories. Chrysotile is the most common asbestos fiber
found in insulation materials.
The laboratories estimated the relative amounts of
sample constituents. These estimates were averaged for each
sample lot, disregarding errors in fiber identification.
Means and standard deviations were included on reports to
the laboratories. The distribution of quantitative esti-
mates were recorded on histograms in 5 percent intervals.
The histograms were included on individual reports to allow
laboratories to place themselves within the distribution.
Because of the lack of an accepted quantitation procedure,
values reported were not used in rating laboratory perform-
ance.
Reports were issued to individual laboratories on March
25, 1980 (commercial), and April 3, 1980 (noncommercial).
Reports included the results of reference analyses, data
reported by the individual laboratory, and summary data on
quantitative estimates. An example of the reports to labo-
ratories is shown in Figure A-l.
A listing of laboratories which participated in the
quality assurance program is included in this appendix.
-68-
-------
Updated lists of participants in subsequent rounds may be
obtained by calling the EPA toll-free number for technical
assistance, 1-800-334-8571.
-69-
-------
ASBESTOS BULK SAMPLE ANALYSIS PROGRAM
RESULTS OF ROUND 1
Sample I.D. #:
Asbestos Present (%)
Laboratory report
Reference report
Other Fibrous Material (%)
Laboratory report
Reference report
220
801
273
854
0
0
100 mineral wool
98 mineral wool
75 anthophyjlite
53 anthophyllite
0
0
0
0
95 fiberglass
98 fiberglass
95 chrysotile
95 chrysotile
0
0
Summary of Laboratories Reporting:
Mean % (Standard deviation)
Asbestos present
Other fibrous material
Distribution of Asbestos Quantitation
0(0)
96.1 (5.4)
53.0(19.3)
1.4(7.1)
0(0)
97.7 (4.0)
84.5 (17.4)
1.2(3.1)
a
1
1
e
(4
UJ
S
O
F—
£
O
Sample I.D. #: 801 *
O
c
vu
O
s
z
13 -
12
11
10
9
8
7
5
4
3
2
1
-
—
-
-
23 IS \2
-------
LABORATORIES PARTICIPATING IN THE ANALYTICAL PROFICIENCY
PROGRAM
# Correct/
4 Samples
American Can Company 4/4
Safety & Industrial Hygiene Laboratory
U.S. Highway 22
Union, New Jersey 07083
American Microscopy Laboratory 4/4
D. 3410 12th Avenue E.
Tuscaloosa, Alabama 35405
Analytical Center, Inc. 4/4
P. 0. Box 15635
Houston, Texas 77020
Boeing Technology Services 4/4
9R-25
P. 0. Box 3707
Seattle, Washington, 98124
Brewer Analytical Laboratories 4/4
311 Pacific Street
Honolulu, Hawaii 96810
C.E.D., Inc. 4/4
Environmental Microscopy International
135 West Cutting Boulevard
Richmond, California 94804
Casalina Associates, Inc. 4/4
47-345 Mahakea Road
Kaneohe, Hawaii 96744
Certified Testing Laboratories, Inc. 4/4
2905 East Century Boulevard
South Gate, California 90280
Clayton Environmental Counsultants, Inc. 4/4
25711 Southfield Road
Southfield, Michigan 48075
Colorado School of Mines 4/4
Research Institute
P. 0. Box 112
Golden, Colorado 80401
Fay Goldblatt 4/4
407 N. Butrick Street
Waukegan, Illinois 60085
-71-
-------
Continental Technical Services
Environmental Health Division
9742 Skillman
Dallas, Texas 75243
Department of Chemistry
New Jersey Institute of Technology
323 High Street
Newark, New Jersey 07102
Department of Geological Sciences
SUNY, New Paltz
New Paltz, New York 12562
Department of Geology
Illinois State University
Normal, Illinois 61761
Eastern Analytical Laboratories
One "A" Street
Burlington, Massachusetts 01803
EMS Laboratories
12563 Crenshaw Boulevard
Hawthorne, California 90250
EMV Associates, Inc.
Microanalysis Laboratory
15825 Shady Grove Road
Rockville, Maryland 20850
Environment/One Corporation
2773 Balltown Road
Schenectady, New York 12301
Environmental Consulting & Testing Services
P. 0. Box 3521
Cherry Hill, New Jersey 08034
Environmental Health Services, Inc.
5206 Lindbergh Boulevard
W. Carollton, Ohio 45449
Erie Testing Laboratories
2401 W. 26th Street
Erie, Pennsylvania 16506
Erlin, Hime Associates
811 Skokie Boulevard
Northbrook, Illinois 60062
# Correct/
4 Samples
4/4
4/4
4/4
4/4
4/4
4/4
4/4
4/4
3/4
4/4
4/4
4/4
-72-
-------
# Correct/
4 Samples
GCA Corporation 4/4
Technology Division
Burlington Road
Bedford, Massachusetts 01730
Geoscience Consultants, Inc. 4/4
P. 0. Box 341366
Coral Gables, Florida 33134
Hager Laboratories 4/4
12000 E. 47th Avenue
Denver, Colorado 80239
Health Science Associates 4/4
Suite B/C
10941 Bloomfield Street
Los Alamitos, California 90720
Herron Testing Laboratories 4/4
5405 Schaaf Road
Cleveland, Ohio 44131
IIT Research Institute 4/4
10 West 35th Street
Chicago, Illinois 60616
Industrial Analytical Laboratory 4/4
1523 Kalakaua Avenue
Suite 101
Honolulu, Hawaii 96826
Industrial Hygienics, Inc. 4/4
755 New York Avenue
Huntington, New York 11743
Industrial Testing Laboratories, Inc. 4/4
2350 Seventh Boulevard
St. Louis, Missouri 63104
Inter-City Testing & Consulting Corporation 4/4
P. 0. Drawer "0"
609 Middle Neck Road
Great Neck, New York 11023
Interscience Research 4/4
2614 Wyoming Avenue
Norfolk, Virginia 23513
-73-
-------
# Correct/
4 Samples
Jesse H. Bidanset & Associates, Inc. 4/4
P. 0. Drawer "O"
609 Middle Neck Road
Great Neck, New York 11023
Law Engineering Testing Company 4/4
3301 Winton Road
Raleigh, North Carolina 27619
LFE Corporation 4/4
Environmental Analysis Lab Division
2030 Wright Avenue
Richmond, California 94804
Maryland Mineral Analysis Laboratory 4/4
Department of Geology
University of Maryland
College Park, Maryland 20740
MJH Associates 4/4
Mineralogical Consultants
13345 Foliage Avenue
Apply Valley, Minnesota 55124
Northrop Services, Inc. 4/4
P. 0. Box 12313
Research Triangle Park, North Carolina 27709
PEDCo Environmental, Inc. 4/4
11499 Chester Road
Cincinnati, Ohio 45246
Princeton Testing Laboratory 3/4
P. 0. Box 3108
Princeton, New Jersey 08540
R. J. Kuryvial & Associates 4/4
Mineralogy/Microscopy Consultants
12185 W. 29th Place
Lakewood, Colorado 80215
Southwestern Laboratories 4/4
P. 0. Box 10687
Dallas, Texas 75207
St. Paul Fire & Marine 4/4
Environmental Services Analytical Laboratory
494 Metro Square Building
7th and Robert Streets
St. Paul, Minnesota 55101
-74-
-------
# Correct/
4 Samples
Sunbelt Associates, Inc. 4/4
6961 Mayo Road
New Orleans, Louisiana 70126
Thomas A. Kubic & Associates 4/4
8 Pine Hill Court
Northport, New York 11768
Tri-State Laboratories, Inc. 4/4
54 Westchester Drive
Austintown, Ohio 44515
Truesdail Laboratories, Inc. 4/4
4101 N. Figueroa Street
Los Angeles, California 90065
United States Testing Company, Inc. 4/4
1415 Park Avenue
Hoboken, New Jersey 07030
Utah Biomedical Test Laboratory 4/4
520 Wakara Way
Salt Lake City, Utah 84108
Walter McCrone Associates, Inc. 4/4
2820 S. Michigan Avenue
Chicago, Illinois 60616
Wausau Insurance Companies 4/4
Environmental Health Laboratory
2000 Westwood Drive
Wausau, Wisconsin 54401
-75-
-------
-76-
-------
APPENDIX B: QUALITY ASSURANCE PROGRAM FOR INITIAL
LABORATORY EVALUATION
-77-
-------
APPENDIX B
Quality Assurance Program For
Initial Laboratory Evaluation
A laboratory must document its experience with PLM and
XRD prior to being used in a school's asbestos analytical
program. All the laboratories on the list in Appendix A
have demonstrated proficiency with these analytical tech-
niques and should be utilized by school systems.
In the event that a state or school district selects a
laboratory not on the list which does not have documented
evidence of successful bulk sample analysis using PLM/XRD,
the following laboratory evaluation procedure is suggested
as a model for initial quality assurance.
Following the split-sample guidance offered in Chapter
3, use a minimum of twenty-five split-samples for judging
the acceptability of a laboratory's performance. If the
number of disagreements is more than 6, then the labora-
tory's performance is suspect. There may be situations in
which it is not possible to have twenty-five split-samples,
but the school official would like to have independent
evidence of the performance of the laboratory. In such
situations, analyze a minimum of five split-samples. If two
or more disagreements are observed in these five split-
samples, the laboratory's performance will be considered
suspect.
-78-
-------
The school official has two options. If the results of
the split-sample analyses are unsatisfactory, reject the
laboratory ties) involved and identify other possible candi-
date laboratories. The second option is, encourage the
laboratory to resolve the analytical problems surrounding
the disagreements. When the laboratory has identified and
corrected the problem, re-submit a new set of split-samples
and again compare the analytical results in order to ensure
that the problem has been resolved. Of course, if the
official is still uneasy after sending several split-samples
to a laboratory, he or she may opt to split every sample
that they send to the laboratory.
In either case — identifying new laboratories or re-
solving the problems at hand with previously selected lab-
oratories — the school official should not forward large
numbers of samples for analysis until a laboratory has been
determined to be acceptable in its analytical performance.
Figures B.I and B.2 summarize the requirements for
initial quality assurance under two different situations;
namely, school systems with less than 25 samples and school
systems with greater than 25 samples.
-79-
-------
Initial QA
Have At Least 5
Split-Samples
Analyzed
Number of
Split-Samples
5
6-8
9-14
15-20
21-25
Critical**
Number
2
3
4
5
6
Determine
The Critical
Number **
Is the
Number of
Disagreements
Less Than The
Critical
Number?
All samples must be split-samples. (A
minimum of 5 split-samples should be
analyzed.) If the number of disagree-
ments is equal to or greather than the
critical number of disagreements, the
lab is suspect. All samples must be
reanalyzed either by the lab in ques-
tion (after the problems are resolved)
or by another lab.
No
Have Lab
Resolve The
Problems
Yes
Resolve All
Split-Sample
Disagreements
Have All
Samples
Reanalyzed
Laboratory is
Satisfactory
Figure B.I Case 1;
School Systems For Which Fewer
Than 25 Samples Will Be Analyzed
-80-
-------
Twenty-five samples should be
split-samples. If 6 or more
disagreements are observed in
the results of these 25 split-
samples, the laboratory's per-
formance is suspect. In this
case, reanalyze the samples
in that lab after the problems
are resolved, or in another
lab until the differences are
resolved.
Initial
QA
Have 25
Split-Samples
Analyzed
Have Lab
Resolve the
Problems
Have All
Samples
Reanalyzed
Is the
Number of
Disagreements
Less Than
6?
Resolve All
Split-Sample
Disagreements
Laboratory is
Satisfactory
Figure B.2 Case 2:
School Systems With Expected Number
of Samples Over 25
-31-
-------
-82-
-------
APPENDIX C: How To USE THE TABLE OF RANDOM DIGITS
-83-
-------
APPENDIX C
Table of Random Digits
The Table of Random Digits
(.1) Begin on the first line in the upper left-hand
corner of the table.
(2) Proceed horizontally line by line as if reading a
book.
(3) Cross off or circle each digit as it is used.
This prevents using the same digit more than once
and also is a marker that tells where the last
digit used was located and that the following
digit is the next digit to use.
(4) The first number of a random number pair is to be
between 0 and the rectangle length, where the rec-
tangle length is the number of feet along the base
of the rectangle.
(5) If the length is less than or equal to 9, follow
instruction (5)a. If the length is less than or
equal to 99 but greater than 9, follow instruction
C5)b. If the length is less than or equal to 999
but greater than 99, follow instruction (5)c. If
the length is less than or equal to 9,999 but
greater than 999, follow instruction (5)d.
-84-
-------
a. Select a one-digit random number from the
table. If the number is less than or equal
to the length, then circle the digit and use
the selected number. If the number is great-
er than the length, then cross the digit off
the table and repeat instruction (5)a.
b. Select a two-digit random number from the
table. If the number is less than or equal
to the length, then circle the digits and use
the selected number. If the number is
greater than the length, then cross the
digits off the table and repeat instruction
(5)b.
c. Select a three-digit random number from the
table. If the number is less than or equal
to the length, then circle the digits and use
the selected number. If the number is
greater than the length, then cross the
digits off the table and repeat instruction
(5)c.
d. Proceed as in instruction (5)c, but select
four-digit random numbers instead of three-
digit random numbers from the table.
(6) The second number of a random number pair is to be
between 0 and the rectangle height, where the rec-
tangle height is the number of feet along the side
of the rectangle.
C7) To select a random number between 0 and the rec-
tangle height, follow instruction (5), using
height wherever instruction (5) specifies length.
C8) Repeat instructions (4), (5), (6), and (7) until
all sample locations have been selected.
-85-
-------
TABLE OF RANDOM DIGITS
44582 46158
11784 55778
99467 43708
90320 11450
06734 40248
82283 16274
51551 55040
30133 39596
92677 88177
20634 56942
08361 25792
13302 02709
18222 53745
65099 27691
64133 99840
07765 11446
86522 81825
97368 79881
32443 94023
33031 64521
74697 72451
78918 04504
36213 47336
04034 34973
06738 25056
29556 67044
81832 00308
11699 40232
78536 25859
45357 95941
91745 99309
63150 41097
78706 82406
59656 00816
25122 39001
90048 70633
02444 92190
71285 96559
54442 26299
91806 60730
47900 04052
06884 46248
63232 28448
82633 53703
01007 38553
87049 98691
47465 74463
05479 75460
70796 68225
99193 15667
74093 59121
96965 77810
60959 43228
73504 86473
77182 05454
09896 94822
81801 37118
45464 86248
70879 86262
49147 06248
30352 54229
30614 30335
08060 08712
45346 50864
50925 39609
36839 00489
13175 64851
10561 68625
28821 42701
86988 73789
05019 10565
88179 33101
07554 52634
42422 99407
07063 41729
13303 41677
94554 40814
81608 95101
81910 42731
13175 59939
09543 38282
45876 87587
75403 52466
87648 64644
12244 93763
62174 06914
62427 85667
04405 68298
19385 79929
94109 13507
27184 67140
52972 63320
16119 73968
05848 89692
00993 33323
32243 81284
08384 91756
63621 05928
16339 64992
64280 75582
03692 64813
28022 82425
56188 98074
51064 47355
20024 26053
62222 66638
71410 00794
93148 44411
94259 94425
76430 05003
53236 94426
40006 04093
56392 87580
20147 42495
08126 56368
88258 32257
08704 89514
60526 67411
58581 19811
28505 42204
29076 09914
23620 53718
72398 42061
80696 99841
82225 35912
12570 05242
58271 51033
07853 48490
38146 27317
39556 86303
14598 28129
15919 53571
25138 21409
22866 98689
26217 68717
82768 62243
40884 17930
03070 87508
96552 86418
51182 31990
66333 21785
06351 46409
50564 32583
55194 28290
94835 76620
95851 50037
04492 36274
43283 65080
96378 02442
56897 81094
56617 72993
97392 39476
89262 81629
62408 78119
13507 27746
68755 71100
71918 36927
74090 79115
29510 64693
72932 14219
77004 05433
07178 41372
23034 55944
47582 03439
72383 28310
35044 28220
96470 08114
09184 24437
42181 61941
02517 77293
28512 75211
55562 98261
77639 49208
11789 08289
99622 80506
46668 48845
28225 28064
40069 82796
23664 93340
06350 46941
31073 87258
41519 82838
90490 92092
46717 58949
16188 02145
36682 93485
45671 28713
97855 21833
21539 95381
71202 32365
73333 01556
58128 21791
24189 15900
22382 00717
63754 19570
83905 65371
05635 85528
80680 11215
10670 80181
38630 74347
66590 38058
95151 72894
51108 67418
07511 78476
72340 77341
92155 42435
07989 49983
65126 47529
71494 76666
13357 07189
00344 78295
72597 22166
90350 53690
32451 93211
10877 67532
16318 63172
36113 48110
45333 55607
08487 31468
98604 94733
92115 43363
31001 29718
81819 00403
34200 92468
90053 17421
09351 20785
16880 15286
68081 34319
79871 44193
78486 26193
56462 44062
89215 59770
54212 59940
90949 64380
97902 50149
63136 70143
53460 76216
06881 32352
37457 50425
74616 99592
53675 21605
70982 71557
80466 82194
68939 49539
64557 42351
78121 12498
36472 40136
88705 28273
11517 81744
61703 65114
-86-
-------
TABLE OF RANDOM DIGITS
56627
05810
89821
74795
03084
10592
55104
57702
12705
71542
22915
55282
21194
28581
06581
24169
94242
22064
35483
38693
87129
00355
13755
64530
77438
06630
04218
54035
37007
98120
28539
12625
75932
54119
92236
36086
50860
65044
33940
01300
35302
55164
49834
73457
55376
91208
14646
77715
30356
36176
89443
99759
05097
19802
10365
20389
18126
23085
74357
08402
70587
64458
21729
02659
16639
20306
11143
94868
99930
19511
74119
87686
93590
08357
71193
66615
20098
23575
00509
14199
31469
35935
54154
15877
75184
82355
76490
53596
00785
32373
51965
63088
24686
50641
37603
37322
20965
85702
56178
89348
89943
57021
12239
44149
69855
24714
71404
31275
83461
60006
46557
33971
74204
86808
01698
21504
98837
20349
24221
95501
47871
30589
98426
53375
80776
54245
09332
77351
09406
94106
95760
03679
55152
61714
44456
75887
13586
24769
77009
80475
48059
34248
09548
77156
68343
13423
34955
51857
04291
94997
45868
21715
70006
58632
02042
05413
76804
80363
51295
25569
01760
80916
42410
58957
33379
33688
71924
33047
12262
43606
92109
90363
02615
64358
58326
47822
16093
27694
26693
17951
25687
96762
05819
55847
06416
37395
61583
51227
12605
37957
72212
67068
74846
30478
33754
24895
74655
40755
30972
40154
04832
69080
03807
04157
30270
14972
59839
88216
02179
55566
11615
75969
40657
43708
43667
49543
04353
12763
50769
36911
92866
99904
31694
22149
77773
19499
49780
40111
16013
30106
76515
19802
83772
69913
95893
05622
53330
44605
05755
60550
13114
99651
33276
95911
44885
61120
04284
47604
25587
34109
64523
72643
45385
78258
49892
43781
57359
37966
08139
89610
99346
18812
80693
52585
23579
77966
57665
77184
32705
63597
15117
87015
68531
36431
18577
76229
39373
70949
10380
99747
83987
76493
70113
28748
95836
39763
67945
42879
26706
95824
97556
51387
74292
85434
70128
47571
05374
66135
31805
42126
14934
10142
68770
14536
71914
43550
84468
32228
62600
71594
92400
70378
42406
79657
81702
75911
36146
42874
81402
76769
77445
33950
82169
14200
23726
45945
67266
52630
65039
59047
72554
35939
28816
31796
76425
48572
07524
30712
68063
38518
81660
21851
38857
45677
04680
12832
74116
07167
96334
34050
60455
92591
33824
94714
20928
99496
07251
04721
08093
23893
29472
09583
13049
33885
72391
21129
25962
91559
46572
03700
79735
76535
96693
45329
55381
48395
34104
82284
86530
32848
62189
16105
75194
30689
90192
38715
71677
58219
27324
59896
30668
35714
93825
87215
50994
79322
26931
15892
39109
02873
05710
83450
00538
53840
71598
21971
32749
68771
89871
15786
60589
40489
49410
05617
26900
41678
45509
40501
63047
44851
44122
09430
48118
75795
63641
18342
05649
75828
43662
75046
05647
45021
65653
76748
42402
22043
11122
22696
20719
10317
65910
15457
67127
13636
45426
53611
44458
27494
23244
05943
45757
61043
78897
25035
94787
84332
88488
71920
13790
12227
84236
88956
71809
24783
78616
77126
30276
80582
24256
64002
49203
52527
69197
11705
82235
28652
77246
73896
03513
90435
63969
17328
08365
30283
8940.8
13062
74120
18305
43279
58221
90048
08299
00611
66906
70685
84159
10575
82112
50481
82698
-87-
-------
TABLE OF RANDOM DIGITS
96779 94885
06973 61333
22366 71653
37197 91054
15234 35530
75554 64074
47230 79000
30-159 83599
28979 73275
65855 05534
95348 50091
41774 64236
03354 96795
88886 09883
48189 54316
29323 88380
57944 15793
26473 35895
90941 14121
15200 48466
03704 21488
06976 19232
58784 61149
92687 63644
68635 28907
25136 53356
10939 52366
98361 61960
34201 75389
94946 95350
92459 46807
01990 61688
56357 03811
36783 05002
88822 11796
03478 89017
15272 84614
29596 47534
71904 81693
05201 51312
16510 95406
83816 94852
19962 86326
66852 52392
84161 37020
58837 30960
12971 62671
21036 13175
34152 24555
50434 17800
33674 52860
00465 70079
64852 69137
45316 64212
10147 65273
37544 34863
08569 74977
72906 07861
87178 48764
44208 08903
44611 49700
05346 57370
86666 35232
77679 07972
64441 32520
34403 29290
46141 77291
03768 48263
32494 52627
68764 30111
23373 27179
77725 26152
89620 88225
39013 63475
63317 16301
21610 96745
77537 80180
02082 44879
40418 63925
19640 24501
00742 98068
21317 58136
04824 53455
71761 35852
28561 27091
30466 54463
27404 33686
89805 95170
94887 45573
78986 27330
39078 31468
73159 76123
99855 14146
32115 75977
79694 35717
84272 38937
87151 80924
77916 31978
54366 40704
99805 32819
39750 47056
02538 83123
36552 25495
63635 68992
07553 78481
36478 79281
06680 99658
13625 35611
58960 40528
19491 82126
54373 80200
74027 46196
38206 24653
20542 81125
06350 71271
29057 74103
54098 37292
09733 22819
65420 12249
29052 75579
78622 98536
82770 07884
38005 81411
45033 98679
35291 27832
14276 83374
98287 14191
33803 64194
01612 60875
58261 86334
05715 91914
81372 32479
88755 30122
40640 62630
93013 64939
32998 45826
51283 72980
89816 58314
76874 74548
63194 98096
43577 67990
05010 08393
28341 93570
80723 96562
73417 15617
27926 95403
08413 22879
78898 69869
33111 00490
71033 83674
59836 10552
86995 05706
85845 71503
02608 93110
62311 36134
58549 44237
07458 17435
03043 69904
14378 03612
66860 32840
76787 16563
05323 43858
39718 80864
54583 70123
93086 52857
18949 37051
71554 16467
43269 63159
66149 47064
92279 88993
85425 92276
32089 25244
29645 40186
44963 28862
49665 26975
38793 27121
09983 42701
41519 20487
27928 54277
12535 12853
30368 76830
89450 54188
02839 71763
26769 02587
94299 98240
92196 84866
53589 61318
03649 64285
36851 48630
93212 74891
11287 27068
62827 13728
34163 59623
19388 64446
93437 46981
61816 32202
51701 84303
22225 13043
53198 52317
84640 67470
26093 40520
71111 40435
31631 58633
21593 56327
89043 56110
19801 31240
08308 11027
55051 74144
90075 96905
54979 22213
18303 66995
84458 81397
28193 86369
13780 74558
63361 98260
93231 73949
07860 47556
38560 13548
51607 98475
69782 27641
97238 28716
20896 06246
35101 89938
51162 71792
36918 71635
02809 18908
69101 73946
22554 69494
23320 23997
97546 80748
01471 31879
15032 52447
49139 06246
44623 95577
57450 18672
90728 60701
78649 06703
14682 12486
77916 78922
55099 02679
37874 61734
34709 39578
14103 63367
73949 83823
94838 12418
11343 99925
65556 20152
49858 81615
77478 38052
60922 25920
-88-
-------
TABLE OF RANDOM DIGITS
49514
92631
40278
81803
06725
03003
96786
13867
60153
03723
70071
85798
03645
36129
67883
35303
65451
•97984
98435
52684
71328
06873
61478
20195
82781
76507
73673
82662
59057
30927
17377
03973
62945
74341
76481
87994
76542
67803
61450
24626
33885
87145
62761
20334
24130
04179
85691
34157
87159
83231
56977
91973
24410
01934
70141
02041
77447
16828
27028
43680
65642
61647
69342
12616
39077
54831
36814
69925
99382
47602
66786
45241
04063
95679
38482
43573
01651
09745
47915
49665
75935
99123
41517
11492
59719
93471
43827
33971
33708
08923
02906
85850
59668
07868
28154
24070
65031
75648
03307
85739
61091
05484
94768
67431
56283
43212
36503
74299
68048
40520
85632
84450
80292
76291
21592
27109
24793
21930
72279
92545
05495
68658
32137
18798
38817
73800
36947
20817
45427
55809
06637
73738
32909
37721
09182
23310
76797
40771
98687
56849
61019
45170
24486
42065
13623
19024
42942
34808
93587
30743
92612
19712
72614
06570
94081
10780
85678
53165
88563
88434
14420
78632
51763
44000
26234
85724
55271
43520
81923
85307
43089
15507
97949
09786
84125
68958
50655
86570
02391
72006
44475
23889
38415
82149
68138
60321
38054
56798
68345
98557
47791
98396
32044
07963
81346
08587
15172
89094
93706
47635
26282
21723
30133
68573
68772
87488
29760
91721
30737
76364
76911
92907
21843
80406
40105
80252
22937
42903
92711
27889
30332
61812
28894
95489
14227
25504
18842
01312
14171
26053
30935
45586
08517
56743
38681
94807
26127
62945
65812
55227
40655
11448
76952
98165
26267
93957
70876
64196
61341
23682
71434
75125
24250
37985
77224
68377
78551
17584
57731
37846
25034
49621
96110
83704
38901
68075
76975
07952
22859
22652
98484
84899
18848
95477
43948
31547
02009
56188
44789
02678
58790
52624
18177
24237
71424
28031
86957
35850
57718
38891
52065
18418
55372
92197
91707
13056
05431
03966
36517
86210
09611
33490
19139
63397
71218
59637
57885
03773
62797
60306
62113
00396
69921
74393
68564
56866
23783
22909
73579
94664
57912
24073
43781
52307
54543
36068
10041
06850
60728
65196
60159
68847
74068
88066
35332
87369
31770
97353
95158
86332
28583
93095
51677
33502
23440
37274
66260
90926
68751
52789
38514
95926
14774
88439
45373
41731
76095
75051
92327
56808
70957
42732
63435
54442
95491
69977
72291
20514
64006
56700
37416
91467
53423
66635
73141
37290
59211
24880
39158
08531
05190
03264
17193
47337
93887
14031
06380
45876
87176
85219
52772
80923
48772
57407
54639
78170
18839
63590
97776
32573
22110
01748
19686
29490
42207
97081
46889
83129
86703
68513
59171
31003
77284
56252
48295
15694
37108
22268
32401
09358
05057
37399
68867
13626
85779
58145
76968
50141
18481
29780
37755
33298
32115
23687
35799
64574
48235
71431
56817
77847
58993
65885
26826
55643
27918
35724
33916
19298
66750
86446
33105
92453
49597
71102
88661
88146
10514
80955
45231
05431
79161
40448
62427
38487
35200
56340
44979
00017
38988
41365
64606
66597
86084
53577
05960
97945
15850
32037
82991
50992
06495
45310
17989
84734
73919
31007
93607
12206
86537
02664
22900
00582
64561
36305
89509
50626
95212
75077
75672
07199
-89-
-------
TABLE OF RANDOM DIGITS
02743
74802
06933
40345
70055
34552
45253
71558
95474
34619
44546
22917
33043
99357
01072
90838
35914
87047
93727
37439
98892
95398
28982
31303
08457
59982
59354
78651
80092
98685
76373
86947
21692
76468
91898
75524
96024
31433
54593
31679
50179
39441
77284
46613
50362
53633
77381
24589
70209
10085
30698x80818
27142
36775
02560
36744
66482
76375
95772
59013
52392
08027
27284
20513
36076
60679
49416
65757
17379
00757
68276
64716
83695
58275
58005
99993
41186
63628
51679
66697
04302
41539
72925
81632
04440
07629
39416
38581
12821
43862
58370
39149
77731
13129
79035
91696
11496
66797
84170
80083
92806
91213
45636
50587
10244
40928
42417
84077
12019
28499
68535
04784
47833
21688
80961
42064
90149
12753
48045
44171
33909
21912
88896
42174
35741
90073
52273
70856
79600
08331
29770
65940
19454
85000
45628
04339
57313
82309
68723
43675
63738
11780
65133
09648
78273
45448
57066
35380
29999
08810
62853
26293
77509
18535
11760
93696
28778
17814
04274
00279
77434
05809
75234
64216
34029
62987
67957
45644
49685
18495
81674
24873
31137
10757
79416
78320
81087
43164
23297
50201
46201
57820
63712
39180
34976
77570
03508
69951
37934
03653
87515
92494
44979
07644
83412
92281
48153
41155
23631
07244
39755
18112
28610
19001
21952
97711
14936
33316
01893
35351
18543
52788
74539
85938
56463
13072
16955
47843
28385
57370
91956
26372
87512
98531
72457
83675
67396
88426
36434
56303
04588
29283
21401
99975
92012
47155
71443
82658
62818
21060
39290
41335
90939
81689
97328
73854
74754
44389
93032
42067
42550
93831
34307
82179
73985
15818
94099
49494
23930
87736
15479
83577
38529
51742
09594
84227
39960
55781
37200
77691
84531
12044
33216
35725
59502
78361
16795
51415
17174
09171
42575
94564
96665
73253
16596
77128
42543
60958
64157
81096
87656
35835
29184
68088
81003
67452
56383
94860
41592
76669
31081
01473
68045
12572
68903
31004
90775
31817
88771
83909
58850
02269
57893
12898
11939
24060
26142
06672
98473
28006
60422
43234
29665
68208
46986
49929
98542
37514
00109
55995
99318
96598
77750
46534
28596
24498
73335
18290
54590
71332
36130
69882
76634
34882
65938
52145
09253
42074
55688
19686
38860
10798
15107
77589
66934
03263
42001
90127
31142
43736
73802
68632
45997
09737
51777
59438
45600
57548
56808
55318
55574
08503
26014
61239
09051
70495
83820
24870
02731
60232
84406
00619
21218
59083
15493
67455
68620
60534
98263
28930
23490
56431
58007
04971
85510
41582
65456
31178
78599
64064
35712
18229
62935
33459
42442
05626
86675
39485
65859
19258
10636
23310
71749
61504
64796
05695
75486
84706
86774
39723
31670
86716
02919
26705
70963
92247
48765
55665
05909
31305
83405
60468
02990
60243
80754
06320
87559
30741
70109
44207
30963
08091
34873
27565
78367
81549
32438
27967
59902
67141
78632
18642
34149
35130
07608
07460
62770
92302
28948
95068
82735
4626.1
28666
18946
52452
38136
74103
25453
07305
25299
61059
08095
17639
43916
19759
04318
24164
05311
58959
30689
21186
97375
50796
27664
48760
82004
77927
89647
06755
08354
57350
01981
83816
66577
29316
82434
96357
12666
21568
80520
-90-
-------
APPENDIX D: EPA REGIONAL ASBESTOS COORDINATORS
-91-
-------
Region 1
Mr. Paul Heffernan, Asbestos Coordinator
Air & Hazardous Materials Division
Pesticides & Toxic Substances Branch
EPA Region 1
JFK Federal Building
Boston, Massachusetts 02203
(617) 223-0585
Region 2
Mr. Peter Flynn, Asbestos Coordinator
EPA Region II
Room 1015, 26 Federal Plaza
New York, New York 10278
(212) 264-4479
Region 3
Ms. Pauline Levin, Asbestos Coordinator
EPA Region III
Curtis Building
Sixth & Walnut Streets
Philadelphia, Pennsylvania 19106
(215) 597-9859
Region 4
Mr. Dwight Brown, Asbestos Coordinator
EPA Region IV
345 Courtland Street
Altanta, Georgia 30308
(404) 881-3864
Region 5
Mr. Anthony Restaino, Asbestos Coordinator
EPA Region V
230 S. Dearborn Street
Chicago, Illinois 60604
(312) 886-6003
Region 6
Mr. Larry Thomas, Asbestos Coordinator
EPA Region VI
First International Building
1201 Elm Street
Dallas, Texas 75270
(214) 767-2723
-92-
-------
Region 7
Mr. Wolfgang Brandner, Asbestos Coordinator
EPA Region VII
324 East llth Street
Room 1500
Kansas City, Missouri 64106
(816) 374-6538
Region 8
Mr. Steve Farrell, Asbestos Coordinator
Region VIII
1860 Lincoln Street
Denver, Colorado 80295
(303) 837-3926
Region 9
Mr. Kirby Narcisse, Asbestos Coordinator
EPA Region IX
215 Fremont Street
San Francisco, California 94105
(415) 556-3352
Region 10
Ms. Margo Partridge, Asbestos Coordinator
EPA Region X
1200 Sixth Avenue
Seattle, Washington 98101
(206) 442-5560
REGIONAL OFFICES
-93-
-------
-94-
-------
APPENDIX E: TOLL-FREE INFORMATION NUMBER
-95-
-------
APPENDIX E
Toll-Free Information Number
ENVIRONMENTAL PROTECTION AGENCY
The following number is to be used for general infor-
mation on the EPA school asbestos program and to request
copies of the guidance manuals and new documents:
800—424-9065
(554-1404 in Washington, D.C.)
This report is also available from:
National Technical Information Service
U.S. Department of Commerce
5285 Port Royal Road
Springfield, Virginia 22161
-96-
------- |