United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-93/155 September 1993
Project Summary
Evaluation of Three Cleaning
Methods for Removing
Asbestos from Carpet:
Determination of Airborne
Asbestos Concentrations
Associated with Each Method
John R. Kominsky, Ronald W. Freyberg, and Kim A. Bracket!
A study was conducted to compare
the effectiveness of three cleaning
methods for removal of asbestos from
contaminated cprpet and to determine
the airborne asbestos concentrations
associated with each. Baseline mea-
surements before cleaning showed an
average concentration of 1.6 billion
asbestos structures per square foot
(s/ft2) of carpet. The effectiveness of
dry vacuuming using cleaners with and
without a high-efficiency particulate air
filter was compared with that of wet
cleaning with a hot-water extraction
cleaner. The wet cleaning method re-
duced the level of asbestos contamina-
tion in the carpet by approximately 60%,
whereas neither dry cleaning method
had any notable effect on the asbestos
level. The type of cleaner used had
little effect on the difference between
the airborne asbestos concentration
before and during cleaning.
This Project Summary was developed
by EPA's Risk Reduction Engineering
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
Asbestos-containing materials (ACM)
may release asbestos fibers into the build-
ing air as a result of disturbance, damage,
or deterioration over time. A concern ex-
ists about the extent to which carpet and
furnishings may be reservoirs of asbestos
fibers and about the behavior of these
fibers during normal custodial cleaning op-
erations.
A 1988 study by the Risk Reduction
Engineering Laboratory (RREL) of the U.S.
Environmental Protection Agency (EPA)
compared how effectively dry vacuuming
and wet cleaning removed asbestos fi-
bers from artificially contaminated carpet.
Airborne asbestos concentrations also
were measured during the carpet-clean-
ing activities. Artificially contaminating the
carpet with known levels of asbestos re-
sulted in a carefully controlled experiment
with sufficient replication to demonstrate
that the wet cleaning method removed
significantly more asbestos material from
the carpet than did the dry cleaning
method. Both methods increased airbornfe
asbestos concentrations significantly.
As a follow-up to this study, EPA's RREL
conducted a "real-world" study to deter-
mine whether the experimental results ob-
tained with artificially contaminated carpet
would also apply to carpet naturally con-
taminated with asbestos fibers released
from in-place ACM. The carpet on which
these methods were tested was naturally
contaminated over a period of 15 to 20 yr
as a result of asbestos-containing ceiling
material and spray-applied fireproofing
above the ceiling. The effectiveness of dry
vacuuming using vacuum cleaners with
and without a high-efficiency particulate
Printed on Recycled Papei
-------
air (HEPA) filter was compared with that
of wet cleaning with a hot-water extraction
cleaner.
The primary objectives of this study were
(1) to determine the ability of three clean-
ing methods to remove asbestos struc-
tures from carpet, (2) to determine air-
borne asbestos levels during carpet clean-
ing by each method, and (3) to compare
fiber concentrations measured by phase
contrast microscopy (PCM) during each
cleaning method with the Occupational
Safety and Health Administration (OSHA)
action level of 0.1 fiber per cubic centime-
ter (f/cm3).
Study Design and Methods
Test Site
This study was conducted in an unoc-
cupied cafeteria area of the East High
Rise Building of the Social Security Ad-
ministration, Baltimore, MD, where the re-
moval of asbestos-containing ceiling ma-
terial and spray-applied fireproofing above
the ceiling was planned. The acoustical
ceiling material contained 1% to 5%
chrysotile, and the fireproofing contained
35% to 40% amosite.
Approximately 3700 ft2 (58 ft x 64 ft) of
the carpeted dining area was isolated as
the test area. Within this area, nine equally
dimensioned areas (19 ft 4 in. by 21 ft 4
in.), each with approximately 400 ft2 of
carpet, were defined as experimental test
cells. Each test cell was covered by a
floor panel 19 ft 4 in. by 21 ft 4 in., which
served as a protective barrier against
cross-contamination during each experi-
ment. The floor panel was removed for
each experiment and replaced when the
experiment was complete. The floor panel
frame was constructed of 2- by 4-in. lum-
ber, and 6-mil-thick plastic sheeting was
stretched across the top surface. A 24- by
27-ft office enclosure was constructed ad-
jacent to the test area. The test area was
entered from the office area through a 5-
by 13-ft decontamination facility. The de-
contamination enclosure consisted of three
equally dimensioned chambers: an equip-
ment-change room, a shower room, and a
clean room.
Air Filtration
Five HEPA filtration units were used to
reduce the airborne asbestos concentra-
tions to background levels after each ex-
periment. These units were operated dur-
ing the preparation phase of the experi-
ment but not during the carpet-cleaning
phase. Four of the units cleaned and
recirculated the air; the fifth unit cleaned
and discharged the air to the outdoors
via flexible ducting. Makeup air was
brought into the test area from outdoors
via the door at the adjacent decontami-
nation facility.
Experimental Design
Three methods of carpet cleaning were
evaluated: (1) dry vacuuming with a HEPA-
filtered vacuum cleaner, (2) dry vacuum-
ing with a conventional vacuum cleaner
(i.e., without HEPA filtration), and (3) wet
cleaning with a hot-water extraction
cleaner. Each method was tested three
times (with different cleaners of the same
model) to yield a total of nine experiments.
The carpeted area was divided into nine
equal 400-ft2 areas. To allow for possible
spatial trends in the contamination level
across the carpet, the three cleaning meth-
ods were applied according to a 3 x 3
Latin square design. The entire carpet was
divided by a grid of three rows and three
columns, and each cleaning method was
applied once in each row and each col-
umn. This provided three tests of each
method.
A single experiment consisted of col-
lecting six baseline work-area air samples
and six bulk carpet baseline samples; dry
vacuuming or wet cleaning the carpet for
60 min; concurrently collecting a second
set of six work-area air samples and three
personal breathing zone samples; collect-
ing a set of six postcleaning bulk carpet
samples; dry vacuuming or wet cleaning
the carpet a second time for 60 min; col-
lecting a second set of three personal
breathing zone samples; collecting a sec-
ond set of postcleaning bulk carpet
samples; covering the carpet with a pro-
tective floor panel; and ventilating the area
with five HEPA-filtration units for 4 hr.
Although six air samples were collected
before and during cleaning and six carpet
samples were collected before and after
cleaning, three randomly selected samples
from each set of six were analyzed. Sta-
tistical significance was achieved with the
reduced set, and the remaining samples
were archived.
Materials and Methods
Fourteen General Service Administra-
tion (GSA) field offices in 11 states across
the country were surveyed to identify the
commonly used conventional vacuum
cleaner. In the 1988 EPA study, a similar
survey was made of 14 GSA offices and 6
trade associations to select the HEPA-
firtered dry vacuum cleaner and hot-water
extraction cleaner. The same model HEPA-
filtered dry vacuum cleaner was used in
this study. Because the HEPA-filtered hot-
water extraction cleaner used in the 1988
study is no longer manufactured, a hot-
water extraction cleaner without HEPA fil-
tration (but manufactured by the same
company) was selected. The conventional
dry vacuum cleaner selected was the
model most frequently mentioned during
the GSA survey.
Carpet Cleaning Equipment
The HEPA-filtered dry vacuum had an
airflow capacity of 87 ftVmin and a 75-in.
static water lift and was equipped with a
16-in. carpet nozzle with a rotating brush.
The hot-water extraction cleaner had an
airflow capacity of 95 ft3/min and a 117-in.
static water lift and was equipped with a
3-in.-diameter by 14-in.-long motorized
agitator brush. The conventional vacuum
cleaner was an upright unit with an airflow
capacity of 110 ft3/min and a 10-in. static
water lift and was equipped with a belt-
driven agitator brush.
Carpet Cleaning Technique
The carpet in each experiment was me-
thodically vacuumed or wet-cleaned for
approximately 60 min to collect enough
air volume to obtain an analytical sensitiv-
ity of 0.005 s/cm3. Each of the two clean-
ing periods consisted of three passes over
the carpet with each cleaner. Each pass
was at a 90° angle to the previous pass.
Sampling Methodology
Carpet Samples
Bulk carpet samples were collected be-
fore and after cleaning with a 10-cm (4-
in.) square template and a utility razor
knife. Each sample was cut in half to
provide a duplicate sample for archiving.
Each piece of carpet was placed in a
separate, labeled, wide-mouth polyethyl-
ene jar with a polypropylene screw cap.
The template and utility razor were thor-
oughly cleaned between each sample col-
lection to avoid cross-sample contamina-
tion.
Area Air Samples
Air samples were collected on open-
face, 25-mm-diameter, 0.45-u.m-pore-size,
mixed cellulose ester (MCE) filters with a
5-u,m-pore-size cellulose support pad con-
tained in a three-piece cassette. The filter
cassettes were positioned approximately
5 ft above the floor with the filter face at a
45° angle toward the floor. The filter as-
sembly was attached to an electric-pow-
ered vacuum pump operating at a flow
rate of approximately 9 L/min. Air volumes
ranged from 487 to 705 L. The sampling
pumps were calibrated both before and
after sampling with a precision rotameter.
-------
Personal Breathing Zone Air
Samples
The person cleaning the carpet during
each experiment wore a personal sam-
pling pump with the filter assembly posi-
tioned in his/her breathing zone. The
samples were collected on open-face, 25-
mm-diameter, 0.8-u,m-pore-size MCE
membrane filters and cellulose support pad
contained in a three-piece cassette with a
50-mm conductive cowl. The filter assem-
bly was attached to a constant-flow, bat-
tery-powered vacuum pump operating at
a flow rate of approximately 2 L/min. The
sampling assembly was worn for the du-
ration of each carpet-cleaning activity. Air
volumes ranged from 110 to 192 L The
sampling pumps were calibrated both be-
fore and after sampling with an electronic
mass flowmeter.
Analytical Methodology
Carpet Samples
A sonication procedure was used to
extract asbestos structures from the bulk
carpet samples for subsequent analysis
by transmission electron microscopy
(JEM).
Area Air Samples
The MCE filters were prepared and ana-
lyzed in accordance with a modified
nonmandatory TEM protocol, as described
in the Asbestos Hazard Emergency Re-
sponse Act final rule (40 CFR Part 763, p.
41870).
Personal Breathing Zone Air
Samples
The 0.8-|im-pore-size MCE filters used
to collect the personal breathing zone
samples were analyzed in accordance with
NIOSH Method 7400 by using PCM at the
EPA TEM laboratory. The analytical sen-
sitivity was approximately 0.01 f/cm3. A
subset of these samples was also ana-
lyzed by TEM in accordance with the pro-
tocol described for the area air samples.
Statistical Analysis
Carpet Samples
A single estimated concentration for
each cleaning method and replicate com-
bination was obtained before and after
cleaning by calculating the arithmetic mean
of the three individual estimates. This
yielded nine pairs of concentrations, three
for each cleaning method. The relative
change in asbestos concentration was
measured by the ratio of the concentra-
tion after cleaning to the concentration
before cleaning. These ratios were com-
pared by taking the natural logarithm and
comparing the averages by standard analy-
sis of variance (ANOVA) techniques.
Area Air Samples
The statistical analysis of the area air
concentrations was similar to that for the
carpet samples. A single estimated con-
centration for each cleaning method and
replicate combination was obtained be-
fore and during cleaning by calculating
the arithmetic mean of the three individual
estimates. The relative change in asbes-
tos concentration was measured by the
ratio of the concentration during cleaning
to the concentration before cleaning. These
ratios were compared by taking the natu-
ral logarithm and comparing the averages
by standard ANOVA techniques.
Personal Breathing Zone
Samples
The three personal breathing zone
samples collected during both cleaning
stages in an experiment yielded a total of
54 personal samples. For each experi-
ment, a single estimated concentration was
then obtained during the first and second
cleanings by taking the arithmetic mean
of the three individual estimates. This
yielded nine pairs of concentrations, one
for each experiment. The relative change
in asbestos concentrations was measured
by comparing the ratio of the concentra-
tion during the first cleaning with the con-
centration during the second cleaning.
These ratios were compared by taking the
natural logarithm and comparing averages
by standard ANOVA techniques.
Quality Assurance
Specific quality assurance procedures
for ensuring the accuracy and precision of
the TEM analyses of air samples included
the use of lot, laboratory, and field blanks
and replicate and duplicate analyses.
Results and Discussion
Area Air Samples
Table 1 presents summary statistics for
airborne asbestos concentrations mea-
sured before and during the first cleaning
stage. The three fixed-station area samples
collected before and during the first clean-
ing stage in each experiment yielded a
total of 54 area air samples. For each
experiment, a single estimated concentra-
tion was then obtained before and during
cleaning by taking the arithmetic mean of
the three individual estimates. This yielded
nine pairs of concentrations, one for each
experiment.
Figure 1 shows the average airborne
asbestos concentrations measured before
and during the carpet-cleaning activity with
each of the three cleaners. Average air-
borne asbestos concentrations increased
during carpet cleaning with each of the
three cleaners. Results from the one-fac-
tor ANOVA indicated that the type of clean-
ing method had no statistically significant
effect on the difference between airborne
asbestos concentrations before and dur-
ing cleaning (p=0.3127); i.e., the mean
relative increase in the airborne asbestos
concentration during carpet cleaning did
not vary significantly with the type of
cleaner.
The increase in airborne asbestos con-
centration during the carpet-cleaning ac-
tivity was statistically significant (p=0.004).
Specifically, a 95% confidence interval for
the mean airborne asbestos concent'a-
tion during carpet cleaning as a propor-
tion of the baseline concentration before
cleaning showed that the overall mean
airborne asbestos concentration was be-
tween 1.3 and 2 times greater during car-
pet cleaning.
Personal Breathing Zone
Samples
All 54 individual samples showed PCM
concentrations below the OSHA (TWA*)
action level of 0.1 f/cm3. The maximum
personal breathing zone concentration was
0.333 f/cm3. (=1-hr sample)
Results of the one-factor ANOVA
showed that the type of cleaning method
had no statistically significant effect on
the difference between personal breath-
ing zone concentrations during the t^rst
and second cleanings (p = 0.5716).
Thirteen of the 54 personal breathing
zone samples, selected to represent those
with the highest concentration measured
by PCM, were also analyzed by TEM.
Overall, the concentrations determined by
TEM were consistently higher, which was
not unexpected because PCM is unable
to detect fibers less than 5 (im in length
and less than 0.25 jim in width. Most of
the structures measured by TEM were
less than 2 u,m in length. The Pearson
correlation coefficient associated with these
measurements (r = 0.03) indicates no sig-
nificant linear relationship between TEM
and PCM concentrations.
Effectiveness of Cleaning
Methods
Figure 2 shows the average (geometric
mean) concentrations of asbestos struc-
tures in the carpet before and after clean-
ing. The 95% confidence intervals for the
* Time Weighted Average
-------
Table 1.
Cleaning Method
Summary Statistics for Airborne Asbestos Concentrations Before and During First
Cleaning
Number of Asbestos Concentration, s/cm3
Data Points *•*
Mean
Minimum
Maximum
Before cleaning
Conventional dry vacuum
HEPA-filtered dry vacuum
Hot-water extraction
2
3
3
0.034
0.079
0.046
0.053
0.025
0.040
0.015
0.163
0.056
During cleaning
Conventional dry vacuum
HEPA-filtered dry vacuum
Hot-water extraction
0.047
0.094
0.093
0.030
0.043
0.066
0.065
0.168
0.109
* Each data point represents the average of three work-area samples.
* Results from Experiment 1 are not included because they were an apparent anomaly.
0.1
0.08
0.06
0.04
0.02 I
Conventional
HEPA-filtered
Vacuum Type
Hot water extraction
Sample Phase
Before Cleaning pyj During Cleaning
Figure 1. Airborne asbestos concentrations (arithmetic mean) before and during carpet
cleaning. (Samples were analyzed by transmission electron microscopy.)
geometric mean concentrations are pre-
sented in Table 2. For each experiment, a
single estimated concentration was ob-
tained before cleaning, after the first clean-
ing, and after the second cleaning by tak-
ing the arithmetic average of the three
individual estimates. This yielded nine trip-
licates of concentrations, one for each ex-
periment.
After First Cleaning
Results of the one-factor ANOVA indi-
cated that the type of cleaning method
had a statistically significant effect on the
difference between asbestos concentra-
tions in the carpet before and after the
first cleaning (p=0.0164); i.e., the mean
relative change in asbestos concentration
in the carpet after cleaning varied signifi-
cantly with the type of cleaner. The esti-
mated asbestos concentration in the car-
pet after cleaning as a proportion of the
asbestos concentration before cleaning for
each cleaning method and the correspond-
ing 95% confidence interval are presented
in Table 3.
The asbestos concentration after wet
cleaning was approximately 0.4 of the as-
bestos concentration before cleaning (i.e.,
a 60% reduction in the concentration). The
upper 95% confidence limit for this pro-
portion (Table 3) is less than 1, which
indicates this is a statistically significant
reduction.
The asbestos concentration in the car-
pet after dry vacuuming with a conven-
tional and a HEPA-filtered dry vacuum
cleaner was 1.3 and 1.2 times the con-
centration before cleaning, respectively.
The 95% confidence intervals for both es-
timates include the number 1, which indi-
cates the data do not provide statistically
significant evidence of either an increase
or a decrease in asbestos concentration
after dry vacuuming with either a conven-
tional or a HEPA-filtered vacuum cleaner.
These results are consistent with the
findings from the 1988 EPA controlled re-
search study, which evaluated the effi-
cacy of HEPA-filtered dry vacuum and
HEPA-filtered hot-water extraction clean-
ers on carpet that was artificially contami-
nated with asbestos. The controlled study
also showed that the efficacy of wet clean-
ing was significantly greater than that of
dry vacuuming. That study showed an ap-
proximately 70% reduction in carpet con-
tamination levels after wet cleaning, com-
pared with an approximately 60% reduc-
tion in this study. The 1988 study also did
not show statistically significant evidence
of either an increase or a decrease in
asbestos concentration after dry vacuum-
ing.
-------
2,500
2,000
=1 7,500 -
0 1,000 -\
500
Conventional
HEPA-filtered
Vacuum Type
Hot water extraction
Sample Phase
Baseline \^\ After 1st Cleaning [s^j After 2nd Cleaning
Figure 2. Asbestos concentrations in the carpet before and after cleaning.
After Second Cleaning
The carpet was dry-vacuumed or wet-
cleaned a second time to determine the
effect of repeat vacuuming or cleaning.
The type of cleaning method used had no
statistically significant effect on the differ-
ence between asbestos concentrations in
the carpet after the first and second
cleanings (p=0.5314). The estimated as-
bestos concentration in the carpet after
the second cleaning as a proportion of the
asbestos concentration after the first clean-
ing is given in Table 4 for each cleaning
method, together with a 95% confidence.
The 95% confidence intervals for these
estimates include the number 1, which
indicates the data do not provide statisti-
cally significant evidence of either an in-
crease or a decrease in asbestos concen-
tration after cleaning the carpet a second
time.
Comparison With 1988
Controlled Carpet Study
In the controlled carpet-cleaning study
performed in 1988, new carpet was
sprayed with an aerosol containing known
concentrations of chrysolite asbestos sus-
pended in water. After the carpet dried, it
was rolled with a 200-lb steel roller to
simulate the effects of normal foot traffic
in working the asbestos into the carpet.
The results of the present study, which
represent a real-world carpet (with un-
known contaminants, similar asbestos con-
tamination levels [1.6 billion s/ft2 average],
and wear characteristics) are quite com-
parable with the results of the high-con-
centration (1 billion s/ft2) controlled experi-
ment in terms of the reentrainment of as-
bestos during cleaning procedures; i.e.,
the airborne asbestos concentrations mea-
sured in this study were 1.3 to 2 times
greater during carpet cleaning versus 2 to
4 times greater in the 1988 study. The
results of the present study are also com-
parable regarding the effectiveness of the
cleaning methods to remove asbestos
structures from carpet; i.e., the present
study showed a 60% reduction in asbes-
tos concentrations in the carpet after wet-
cleaning compared with a 70% reduction
in the 1988 study. Both studies showed
that dry vacuuming did not significantly
change the asbestos concentration in the
carpet.
Conclusions
Wet cleaning reduced the asbestos con-
centration in the carpet by approximately
60%, whereas no significant evidence of
an increase or decrease was found in
asbestos concentrations after dry vacu-
uming.
Both wet cleaning and dry vacuuming
of carpet resulted in a statistically signifi-
cant increase in airborne concentrations
in the work area. Concentrations were 1.3
to 2 times greater during carpet-cleaning
activities than before.
Although the personal breathing zone
samples analyzed by PCM were all below
the OSHA action level of 0.1 f/cm3 of air,
considerably higher exposures were indi-
cated by the samples analyzed by TEM
because PCM does not detect the smaller
fibers (<5 u,m in length and <0.25 urn in
width) measured by TEM. The structures
measured by TEM analyses were pre-
dominantly <5 u,m in length, i.e., 99.6%
and 97.1% during dry and wet carpet-
cleaning activities, respectively.
The results of this study involving car-
pet with natural asbestos contamination
and wear characteristics are comparable
with those obtained in a controlled study
under artificial, simulated conditions in both
efficacy of the carpet cleaning methods
and reentrainment of asbestos structures
during cleaning activities.
Recommendations
In buildings containing friable ACM,
vacuuming of carpets during routine cus-
todial activities should be performed with
HEPA-filtered dry vacuum cleaners. Car-
pets should be cleaned periodically by a
-------
Table 2. Summary Statistics for Asbestos Concentrations in Carpet Before and After Cleaning
Cleaning Method
Number of
Data Points *
Asbestos Concentration, Billion s/ft*
Geometric mean 95% Cl *
Baseline
Conventional dry vacuum 3
HEPA-filtered dry vacuum 3
Hot-water extraction 3
After 1st Cleaning
Conventional dry vacuum 3
HEPA-filtered dry vacuum 3
Hot-water extraction 3
1.6
1.1
2.0
2.1
1.3
0.85
(0.85, 3.1)
(0.28, 4.0)
(1.1, 3.5)
(1.2, 3.7)
(0.39, 4.3)
(0.32, 2.3)
After 2nd Cleaning
Conventional dry vacuum
HEPA-filtered dry vacuum
Hot-water extraction
3
3
3
1.3
1.4
0.88
(0.23, 7.3)
(0.82, 2.4)
(0.24, 3.3)
Each data point represents the average of three work-area samples.
95% confidence interval for the geometric mean.
wet-cleaning method (e.g., a hot-water
extraction cleaner). If ACM has been re-
leased onto a carpeted area during an
operation and maintenance activity or as
a result of fallen surfacing material, the
gross debris should be removed by a
HEPA-filtered dry vacuum cleaner, fol-
lowed by wet cleaning of the carpet.
The full report was submitted in fulfill-
ment of Contract No. 68-CO-0016 by In-
ternational Technology Corporation under
the sponsorship of the U.S. Environmen-
tal Protection Agency.
Table 3. Estimated Asbestos Concentration in Carpet After First Cleaning as a Proportion of the
Concentration Before Cleaning (P)
Cleaning Method
95% Confidence Interval
Conventional dry vacuum 1.3
HEPA-filtered dry vacuum 1.2
Hot water extraction cleaner 0.43
(0.75, 2.1)
(0.74, 2.0)
(0.26, 0.72)
Table 4. Estimated Asbestos Concentration in Carpet After Second Cleaning as a Proportion
of the Concentration Before Cleaning (P)
Cleaning Method
95% Confidence Interval
Conventional dry vacuum 0.63
HEPA-filtered dry vacuum 1.1
Hot water extraction cleaner 1.0
(0.26, 1.5)
(0.45, 2.6)
(0.43, 2.5)
•&U.S. GOVERNMENT PRINTING OFFICE: 1993 - 750-071/80053
-------
-------
J.R Kominsky and R. W. Freyberg are with Environmental Quality Management
Inc., Cincinnati, OH 45240 (formerly with International Technology Corpora-
tion). K.A. Brackett is with International Technology Corporation, Cincinnati,
OH 45246.
W.C. Cain is the EPA Project Monitor, and M. Lehmkuhl is the EPA Project
Officer (see below).
The complete report, entitled "Evaluation of Three Cleaning Methods for
Removing Asbestos from Carpet: Determination of Airborne Asbestos
Concentrations Associated with Each Method," (Order No. PB93-218568;
Cost: $19.50, 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 Monitor can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-93/155
------- |