United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-89/010 Aug. 1989
Project Summary
Evaluation of Asbestos Levels in
Two Schools Before and After
Asbestos Removal
Mark A. Karaffa, Jean Chesson, James Russell, and Thomas J. Powers
This report presents a statistical
evaluation of airborne asbestos data
collected at two schools before and
after removal of asbestos-containing
material (ACM). Although the moni-
toring data are not totally consistent
with new Asbestos Hazard Emer-
gency Response Act (AHERA) re-
quirements and recent EPA guide-
lines, this study evaluates these
historical data by standard statistical
methods to determine if abated work
areas met clearance criteria.
This Project Summary was devel-
oped by EPA's Risk Reduction Engi-
neering Laboratory, Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering Information at
back).
Introduction
The U.S. Environmental Protection
Agency (EPA) has undertaken a variety
of technical assistance and regulatory
activities designed to control ACM in
buildings and .to minimize inhalation of
asbestos fibers. In 1979, the EPA began
publishing state-of-the-art guidance to
help school administrators and building
owners identify and control asbestos
hazards in buildings. An important
objective of these guidance documents
has been to address the question of how
to determine when an asbestos-abate-
ment worksite has been successfully
cleaned and is acceptable for reoccu-
pancy.
The EPA's second asbestos guidance
document, "Guidance for Controlling
Friable Asbestos-Containing Materials in
Buildings" (Blue Book) (EPA-560/5-83-
002), describes for the first time a two-
part process for determining when an
abatement project is complete and the
contractor can be released. The two
steps are (1) a visual test to determine if
the ACM has been property abated and
the worksite is free of debris and dust,
and (2) an air test to determine if residual
asbestos fibers generated during removal
have been reduced below a predeter-
mined level. The recommended method
of sampling and analysis of air samples
presented in the 1983 EPA guidance
document was the National Institute for
Occupational Safety and Health (NIOSH)
method based on phase-contrast micros-
copy (PCM). The sampling and analysis
specifications suggested for air moni-
toring after project completion included
no mention of aggressive sampling and
recommended the PCM detection limit as
the criterion level for clearance. The
limitations of PCM analysis and static
sampling techniques for post-abatement
clearance testing are now well known and
have led to the use of more sophisticated
and accurate methods of asbestos identi-
fication.
The EPA guidance document, "Guid-
ance for Controlling Asbestos-Containing
Materials in Buildings" (EPA-56075-85-
024) was published in June 1985. Later in
1985, EPA published "Measuring Air-
borne Asbestos Following An Abatement
Action" (EPA-600/4-85-049), which dis-
-------
cusses the subject in more detail. These
documents recommended and presented
supportive arguments for the use of
aggressive sampling and transmission
electron microscopic (TEM) analysis of
air samples. In addition, these more
recent guidelines contain a recommend-
ed protocol for aggressive sampling, a
sampling strategy for post-abatement
clearance monitoring, and a statistical
method for evaluating the TEM results
and the adequacy of the contractor's
cleanup.
Sampling, analytical, and statistical
protocols for clearance testing of an
abatement site have undergone further
revisions as a result of the Asbestos
Hazard Emergency Response Act of
1986 (AHERA). The final rule (40 CFR
Part 763, Subpart E), which was pub-
lished October 30, 1987, specifies a
detailed aggressive sampling protocol
that incorporates the use of a leaf blower
and fans, the collection and TEM analysis
of a representative and statistically defen-
sible number of air samples from inside
and outside the work area, a statistical
method (Z-test) for evaluating the TEM
results, and numerous mandatory quality
control and quality assurance proce-
dures.
This report compares historical TEM air
monitoring data collected in two schools
(under static and aggressive conditions)
before and after asbestos removal. To the
extent that the data allow, this study
attempts to evaluate these data by
applying standard statistical methods
designed to determine whether the inside
airborne asbestos concentrations are
significantly higher than the outside
asbestos concentrations.
Study Design, Experimental
Methods, and Site Descriptions
During the summer of 1985, air
samples were collected on 0.8-u,m pore
size, 37-mm mixed diameter cellulose
ester filters at two schools before and
after ACM. During each sampling period,
samples were first taken without any
deliberate attempt to disturb the air (static
sampling). A second set of samples was
then taken after leaf blowers and fans had
been used to resuspend any settled as-
bestos fibers (aggressive sampling). Each
abatement work area at both schools had
an aggressive clearance concentration of
less than 0.01 fiber/cm3 (by PCM) of air
before the work area was cleared by
PCM and before subsequent TEM air
analyses were performed.
Six sites (rooms) were sampled at
School 1, and four sites were sampled at
School 2. Three air samples, each con-
sisting of approximately 1,450 L, were
collected at each site. The flow rate was
approximately 3 L/min. Outdoor samples
were collected on the roof of each
school16 at School 1 (6 before removal
and 10 after removal) and 10 at School 2
(4 before removal and 6 after removal).
Field blanks (filters handled that are
subject to the same as standard samples,
but through which no air is drawn) were
used at both schools (13 at School 1 and
11 at School 2) to check for sources of
asbestos contamination other than the air
being sampled.
The samples were prepared in accord-
ance with the NIOSH 7402 technique,
which is a direct-preparation technique
for cellulose ester membrane filters.
Samples were analyzed by TEM in
accordance with the procedures in
Yamate et al. The results are expressed
in asbestos structures per cubic centi-
meter (s/cm3) for ajr samples and
asbestos structures per square millimeter
(s/mm2) for blanks. When more than one
analysis was done on a single filter, the
average of the multiple readings was
used in the statistical analysis. Results
are not available for two samples from
School 1, and one sample from School 2
was not analyzed because there was no
filter in the cassette when it reached the
laboratory. Statistical analyses are based
on 99 TEM results from School 1 and 68
from School 2.
Methods of Statistical Analysis
Asbestos concentrations on the blank
filters were examined first to ensure that
a contamination problem did not exist
before proceeding with the analysis of
the field samples. Samples with an
asbestos count of zero were assigned an
estimated airborne asbestos concentra-
tion of zero s/cm3. A concentration of
zero was used in all subsequent calcu-
lations and analyses with the exception of
the Z-test.
Some researchers set the airborne
asbestos concentration to the analytical
sensitivity when the structure count is
zero. (The analytical sensitivity, also re-
ferred to as the detection limit, is the
estimated airborne asbestos concentra-
tion calculated when a single fiber is
counted in a sample.) In this research
study, which involved statistical analysis
of multiple samples, using the estimate of
zero is preferable to substituting the
analytical sensitivity because the latt
approach introduces a positive bias th
may obscure trends of interest.
Objectives
The objectives of this evaluation wer
(1) To compare airborne asbestos leve
after removal with outdoor levels, (2) 1
compare airborne asbestos levels befoi
and after removal, (3) To compare tw
methods of sampling airborne asbesto
static sampling and aggressive sampling
Statistical analyses were designed "
address each of the three research objei
tives. Summary statistics (mean an
standard deviation) were generated fi
each sampling period, sampling metho<
and site.
Comparison of Airborne
Asbestos Levels After Removal
With Outdoor Levels
The Kruskal-Wallis one-way analysis <
variance and the Z-test were used to te
for differences between indoor and ou
door sites after asbestos removal. Th
Kruskal-Wallis analysis is a nonpar;
metric test that uses ranks rather the
actual data values. Although it is not i
definitive as the standard analysis <
variance, it does not require assumptior
about the distribution of the data.
The Z-test is a standard comparison <
means for data that are normal!
distributed. Because it is based on a Ic
transformation of the data, the particuli
form of the Z-test required under th
AHERA EPA Asbestos-Containing Mat<
rials in Schools, Final Rule and Notici
specifies that zero concentrations are I
be replaced by the analytical sensitivil
before calculating the Z statistic, ft
noted earlier, this is the only situation i
which the analytical sensitivity was use
in place of zero. The studies wer
conducted before the AHERA clearanc
criteria were developed, and they do n<
meet all the sampling and analysi
requirements. In particular, in this stud1
fewer location samples were taken p<
site (three instead of five) and th
analytical sensitivity was less (0.00
instead of 0.007 s/cm3). This means th<
the Z-test (required under AHERA) is les
likely to detect differences betwee
inside and outside concentrations i
these studies than would be detected i
studies in which the requirements at
met. Nevertheless, the Z-test was applie
to each site to gauge its performanc
under nonideal circumstances.
-------
Comparison of Airborne
isbestos Levels Before and
After Removal
The average airborne asbestos concen-
tration at each site before removal was
subtracted from the average concen-
tration after removal to give a measure of
the effect of removal. A t-test was used to
test whether this measure, which is
approximately normally distributed, is
significantly different from zero.
Comparison of Static and
Aggressive Sampling
For each site and sampling period, the
average airborne asbestos concentration
obtained by aggressive sampling was
plotted against the corresponding con-
centration obtained by static sampling
and a correlation coefficient was calcu-
lated.
Results and Discussion
School 1
No asbestos fibers were found on any
of the 13 blank filters, which indicates
that contamination from sources other
than the air being sampled is not an
important factor. Figure 1 presents the
mean airborne asbestos level at each site
for each sampling period and method.
Arithmetic, rather than geometric, means
are reported because of the large number
of zero measurements.
Indoor Airborne Asbestos Levels
Before and After Removal
Compared With Outdoor
Asbestos Levels
Average indoor airborne asbestos
levels after removal were higher than
were outdoor levels. The Kruskal-Wallis
test indicates significant differences
among sites for both static (p = 0.001)
and aggressive samples (p = 0.002).
Significant differences also existed
among the indoor sites (p = 0.02 for both
static and aggressive samples), which
indicates that levels after removal can
differ from room to room in the same
school.
Results of the Z-test show that five of
the six sites failed the test under static
sampling, and four failed under
aggressive sampling.
Past experiences with final clearance
criteria suggest that sufficient air
exchange in abatement areas following
final cleaning (via negative air systems) is
a major factor in passing the test. Also,
meticulous, repetitive wet cleaning and
HEPA vacuuming of all surfaces are
necessary to remove reentrainable
asbestos fibers.
Comparison of Airborne
Asbestos Levels Before and
After Removal
Average airborne asbestos levels at
each site before and after removal are
presented graphically in Figure 1, which
illustrates both static and aggressive
sampling results. Levels were higher after
removal at all sites, although the dif-
ference was not statistically significant
(t = 2.01, p = 0.l, for static samples;
t = 1.17, p = 0.3, for aggressive samples)
because of the large variability from site
to site. The Kruskal-Wallis test detected
no significant differences between indoor
and outdoor sites before removal
(p = 0.99 for static samples, p = 0.81 for
aggressive samples), but detected signif-
icant differences between indoor and
outdoor sites after removal. This confirms
that the situation before removal differed
from that after removal. Final air quality
following asbestos removal appears to be
related directly to the adequacy of the
final cleaning and to the degree of air
exchange occurring in the work area as a
/O1
W
10'
Static Sampling
6
.0/3
I
.000^ .000
1
.141
.1
.000
032
I
.(
.002
r
334
1
*V>
OV
000
I
.000
I
Sampling Sites
/O1
10°
10'
JO'
.008
00 1
Aggressive Sampling
.000
.003
.004
234
Sampling Sites
.323
.005
Legend
I I Before Removal
E2ZZ1 After Removal
Figure 1. Average JEM airborne asbestos levels measured by static and aggressive sampling at School 1 before and after asbestos removal
at each site. Sites are identified by number.
-------
Z-Test Statistic
Static
3 4
Location Number
Aggressive
Pass-Fail Line
Figure 2. Z-test for School 1 post abatement static and aggressive sampling.
result of engineering control systems
(e.g., negative air). Figure 2 presents the
Z-test results for School 1.
School 2
No asbestos fibers were found on any
of the 11 blank filters, which indicates
that contamination from sources other
than the air being sampled is not an
important factor. The mean airborne
asbestos level at each site for each
sampling period and method are pre-
sented in the full report. Arithmetic, rather
than geometric, means are reported
because of the large number of zero
measurements.
Indoor Airborne Asbestos Level
Before and After Removal
Compared with Outdoor
Asbestos Levels
Average airborne asbestos levels after
removal were low. No asbestos fibers
were detected outdoors. The Kruskal-
Wallis test detected no significant dif-
ferences between indoor and outdoor
sites for both static (p = 0.46) and ag-
gressive samples (p = 0.44).
Results of the Z-test show that none of
the four sites failed the test under static
sampling; one site failed the test under
aggressive sampling.
Comparison of Airborne
Asbestos Levels Before and
After Removal
Average airborne asbestos levels at
each site before and after removal are
plotted in Figure 3, which illustrates both
static and aggressive sampling results.
Levels were lower after removal at all
sites except site 4 under static sampling
conditions. The reduction was not statis-
tically significant (t = -1.5, p = 0.23, for
static samples; t = 1.4, p = 0.26, for
aggressive samples). The Kruskal-Wallis
test detected no significant differences
between indoor and outdoor sites before
removal for static samples (p = 0.25), but
it did detect a significant difference for
aggressive samples (p = 0.01). No sii
nificant differences were detected aft
removal, which indicates that the situatic
before removal differed from the situatic
after removal.
Significant differences also existe
among the indoor sites prior to remov;
which indicates that levels can differ fro
room to room in the same school. Figu
4 presents the Z-test results for School;
Post Abatement Clearance
Concentrations
The cumulative concentrations of ai
borne fibers for School 1 and School
were analyzed. The data for School
included 18 aggressive, 18 static, and 1
ambient samples. Figure 5 {School
illustrates the cumulative average coi
centrations for the specific type <
sampling employed. The data for Scho
2 are depicted in Figure 6. The po
abatement clearance data for School
includes 12 aggressive, 11 static, and
ambient samples.
-------
Static Sampling
Aggressive Sampling
701
10°
to
70-
70"
.020
.003
.000
.013
.012
015
.002
.000
701
70°
70-
70"
.513
.110
.016
.019
.007
.002
.002
.053
2 3
Sampling Sites
2 3
Sampling Sites
Legend
\ I Before Removal
After Removal
Figure 3. Average TEM airborne asbestos levels measured by static and aggressive sampling at School 2 before and after asbestos removal
at each site. Sites are identified by number.
Z-Test Statistic
-1.
T
2
~T
3
Static
Location Number
Aggressive
Pass-Fail Line
Figure 4. Z-test for School 2 post abatement static and post aggressive sampling.
-------
Results and Conclusions
Results
The main results are summarized by
school as follows:
School 1
Indoor airborne asbestos levels prior to
removal were not statistically signif-
icantly different from outdoor levels.
The indoor levels (using both static and
aggressive sampling techniques) in-
creased after removal and were signifi-
cantly higher than outdoor levels at all
but one of the sites. The levels differed
significantly at different sites.
School 2
Indoor airborne asbestos levels meas-
ured by aggressive sampling were
significantly higher than outdoor levels
removal. Differences among indoor
sites were also significant. Levels
declined after removal, and with the
exception of one site under aggressive
sampling, they were not significantly
different from outdoor levels.
Conclusions
The preceding results led to the
following conclusions;
(1) At the schools surveyed, the effect of
asbestos removal was unpredictable.
Measured airborne asbestos levels
may reflect reductions after removal
in some cases and increases in
others.
(2) The variability among different sites
within the same building, even when
the abatement operation is carried
out in accordance with presumably
uniform specifications, argues for the
need to treat different sites as
separate areas for the purpose of
clearance.
(3) Aggressive sampling is appropriate
for clearance testing. It tends to
capture more asbestos (i.e., meas-
urements by aggressive sampling
generally produce larger values th
do measurements by static sa
pling), which lowers the chance
declaring a worksite clean wh
entrainable asbestos is still present
Recommendations
Based on the findings of this stuc
three recommendations can be made:
(1) An immediate research objecti
should be to identify abateme
projects in which work acceptan
was achieved (per the current El
clearance criteria) and to descril
the final cleaning methods ai
engineering control strategies usi
to achieve acceptance.
(2) The long-term effectiveness of ci
rently recommended abateme
methods should be investigah
further. The use of aggressive sar
pling and TEM Methods to Monit
previously abated building are;
(including those that were clear*
Cumulative fibers/cm3
0.26
Number of samples
Q 18 Aggressive
A- 18 Static
0 12 Ambient
25 3 35
Length of fiber (micrometers)
4.5
5.5
Figure 5. Post abatement clearance data for School 1
-------
Cumulative fibers/cm3
0016
0014-
0012-
05
-r
2 2.5 3 35
Length of fiber (micrometers}
0.016
-0.014
-0.012
Number of samples
0 12 Aggressive
-A- / / Static
0 8 Ambient
Figure 6. Post abatement clearance data for School 2
only by PCM methods) could help
identify trends in indoor asbestos
fiber concentrations over time.
(3) Future asbestos abatement research
studies should focus on AHERA-Rule
requirements.
The full report was submitted in ful-
fillment of Contract No. 68-03-4006 by
PEI Associates, Inc., under the sponsor-
ship of the U.S. Environmental Protection
Agency.
-------
Mark A. Karaffa is with PEI Associates, Inc., Cincinnati, OH 45246; Jean Chesson
and James Russell are with Price Associates, Inc., Washington, DC 20037; and
the EPA author, Thomas J. Powers (also the EPA Project Officer, see below), is
with the Risk Reduction Engineering Laboratory, Cincinnati, OH 45268.
The complete report, entitled "Evaluation of Asbestos Levels in Two Schools
Before and After Asbestos Removal," (Order No. PB 89-765 922/AS; Cos*:
$13.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAN
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-89/010
230 5 051BB08N
IL 6D6S"
------- |