ASBESTOS FIBER REENTRAINMENT DURING
DRY VACUUMING AND WET CLEANING OF
ASBESTOS-CONTAMINATED CARPET
John R. Kominsky
Ronald W. Freyberg
PEI Associates, Inc.
Cincinnati, Ohio 45246
EPA Contract No. 68-03-4006
Technical Project Monitor
William C. Cain
Project Officer
Thomas J. Powers
Water and Hazardous Waste Treatment Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISC LA I fri ER
The information in this document has been funded wholly or in part by
the U.S. Environmental Protection Agency under Contract 68-03-4006 to PEI
Associates, Inc. It has been subjected to the Agency’s peer and administra-
tive review, and it has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
ii
-------
FOREWORD
Today’s rapidly developing and changing technologies and industrial
products and practices frequently carry with them the increased generation of
materials that, if improperly dealt with, can threaten both public health and
the environment. The U.S. Environmental Protection Agency (EPA) is charged
by Congress with protecting the Nation’s land, air, and water resources.
Under a mandate of national environmental laws, the Agency strives to formu-
late and implement actions leading to a compatible balance between human
activities and the ability of natural systems to support and nurture life.
These laws direct the EPA to perform research to define our environmental
problems, measure the impacts, and search for solutions.
The Risk Reduction Laboratory is responsible for planning, implementing,
and managing research, de velopment, and demonstration programs to provide an
authoritative, defensible, engineering basis in support of the policies,
programs, and regulations of the EPA with respect to drinking water, waste-
water, pesticides, toxic substances, solid and hazardous wastes, and Super-
fund—related activities. This publication is one of the products of that
research and provides a vital communication link between the researcher and
the user community.
This report provides information on airborne asbestos fiber reentrain-
merit during dry vacuuming and wet cleaning of asbestos-contaminated carpet
under experimental conditions. Airborne asbestos concentrations were deter-
mined before and during carpet cleaning. Overall, airborne asbestos con-
centrations were two to four times greater during the carpet cleaning activ-
ity. The level of asbestos contamination and the type of cleaning method
used had no statistically significant effect on the relative increase of
airborne asbestos concentrations during carpet cleaning.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
111
-------
ABSTRACT
A study was conducted to evaluate the potential for asbestos fiber
reentrainment during cleaning of carpet contaminated with asbestos. Two
types of carpet cleaning equipment were evaluated at two carpet contamination
levels. Airborne asbestos concentrations were determined before and during
carpet cleaning. Overall, airborne asbestos concentrations were two to four
times greater during the carpet cleaning activity. The level of asbestos
contamination and the type of cleaning method used had no statistically sig-
nificant effect on the relative increase of airborne asbestos concentrations
during carpet cleaning.
This document was submitted in fulfillment of Contract No. 68—03-4006 by
PEI Associates, Inc., for the U.S. Environmental Protection Agency’s Office
of Research and Development, Risk Reduction Engineering Laboratory. This
report covers a period of January 1988 to July 1989, and work was completed
as of July 31, 1989.
iv
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vi
Acknowledgments vii
1. Introduction 1
Background 1
Objectives 1
2. Conclusions and Recommendations 2
Conclusions 2
Recommendations 2
3. Study Design 3
Test facility 3
Experimental c esign 5
Sampling strategy 6
4. Materials and Methods 8
Selection of carpet 8
Selection of carpet cleaning equipment 8
Sampling methodology 9
Analytical me1 hodology 9
Statistical analysis 9
5. Experimental Procedures 11
Prestudy air monitoring 11
Carpet contamination 11
Disposal of asbestos-containing material 17
Site cleanup 17
Poststudy air monitoring 18
6. Quality Assurance 19
Sample chain of custody 19
Quality assurance sample analyses 19
Spray—application technique 21
7. Results and Discussion 25
References 32
Appendix 1k — Chrysotile Fiber Size Distribution in the High- and
Low-Concentration Arupules 33
Appendix B - Total Airbor e Asbestos Structure Concentrations Before
and During Carpet Cleaning for Samples Analyzed by
Transmission Electron Microscopy 35
Appendix C — Structure Length Distributions of Airborne Asbestos
Before and During Carpet Cleaning 40
V
-------
FIGURES
Number Page
1 Layout of test facility 4
2 Distribution of chrysotile fiber lengths in the low and
high concentration aqueous asbestos suspensions 16
3 Fiber size distributions from preliminary study of
asbestos dispersion by spraying 24
4 Average airborne asbestos concentrations before and
during carpet cleaning 26
5 Comparative plot of cumulative percentages of airborne
asbestos fibers during dry vacuuming and wet cleaning
of carpet with asbestos fibers in the low and high
concentratior? suspensions 29
6 Airborne asbestos concentrations for varying fiber
lengths for samples collected during dry vacuuming
and wet cleaning of carpet 30
TABLES
Number Page
1 Experimental Design 5
2 Probability of Detecting a Statistically Significant
Difference at the 5 Percent Level of Significance
Between Two Groups of Airborne Asbestos Measurements 6
3 Summary of Prestudy Airborne Asbestos Concentrations in
Test Facility 11
4 Summary of Results of Transmission Electron Microscopy
Analyses for’ Low and High Concentration Ampules 15
5 Summary of Field and Laboratory Blank Analyses 20
6 Results of Replicate and Duplicate Sample Analyses 21
7 Results From Preliminary Study of Asbestos Dispersion
by Spraying--Fibers and Fiber Bundles 22
8 Fiber Length Distributions From the Preliminary Study of
Asbestos Dispersion by Spraying 22
9 Summary Statistics for Airborne Asbestos Concentrations
Before and During Carpet Cleaning 27
10 Summary of ANOVA Results for Airborne Asbestos Concen-
trations Measured Before and During Carpet Cleaning 27
11 Structure Morphology Distribution for Air Samples
Collected Be fore and During Carpet Cleaning 28
12 Comparison of TEM and PCM Analyses 0 f Selected Air Samples 31
vi
-------
AC KNO WL EDGMENTS
This document was prepared for EPP’s Office of Research and Development,
Risk Reduction Engineering Laboratory, in fulfillment of Contract No. 68—03-
4006. Mr. Thomas J. Powers, P.E., served as the EPA Project Officer. Mr.
Powers also offered the invaluable suggestion of contaminating the carpet
using an aqueous suspension of asbestos. Mr. William C. Cain served as the
Technical Project Monitor for this project. The administrative efforts and
support given by Mr. Roger Wilmoth of EPA ’s Office of Research and Develop-
ment is greatly appreciated.
The technical assistance provided by Dr. Eric Chatfield of Chatfield
Technical Consulting, Limited, in the preparation and characterization of the
aqueous asbestos suspensions used to contaminate the carpet is gratefully
acknowledged. Colonel Stephen F. Kollar, Commander, USAF, authorized the use
of a building at Wright Patterson Air Force Base to conduct this research
study. Administrative support from Behram Shroff, Douglas Post, and Suzette
Smith of the U.S. Air Force is also acknowledged. Review comments arid sug-
gestions provided by William Burch, P.E., Larry Longanecker, Kin Wong, Ph.D.,
Elizabeth Dutrow, and Joseph Breen, Ph.D., of EPA’s Office of Toxic Substanc-
es; and William McCarthy and Michael Beard of EPA’s Office of Research and
Development, is also appreciated. Jean Chesson, Ph.D., Chesson Consulting.
provided statistical consultation arid peer review. Christopher Frebis of
Computer Sciences Corporation also provided a statistical review of this
report.
John R. Kominsky, C.I.H., and Ronald W. Freyberg of PET Associates,
Inc., were the principal authors. Mr. Robert S. 4mick, P.E., of PET Associ-
ates, Inc., served as senior reviewer. Marty Phillips and Jerry Day of PET
Associates, Inc., performed the technical edit and copy edit, respectively.
vii
-------
SECTION 1
INTRODUCTION
BACKGROUND
Buildings that contain friable asbestos-containing materials (ACM) may
present unique exposure problems for custodial workers. Under certain condi-
tions, asbestos fibers can be released from fireproofing, acoustical plaster,
and other surfacing material. The release of asbestos by aging and deterio-
rating ACM is known to be episodic and to relate to a myriad of factors, such
as the condition and amount of asbestos present, the accessibility of the ma-
terial, activity within the area, vibration, temperature, humidity, airflow,
use patterns, etc. A major concern is the extent to which carpet and fur-
nishings may be serving as reservoirs of asbestos fibers and what happens to
these fibers during normal custodial cleaning operations.
OBJECTIVES
The U.S. Environmental Protection Agency (EPA) performed a series of
controlled experiments in an unoccupied buildinq to evaluate the effective-
ness of a high—efficiency particulate air (HEPA)—filtered vacuum cleaner and
a HEPA-filtered hot-water extraction cleaner in the removal of ashestos from
carpet, and to evaluate the potential for reentrainment of asbestos fibers
during carpet—cleaning activities. The study was designed to compare carpet
asbestos concentrations before and after cleaning with each cleaning method
at two known contamination levels. Work area airborne asbestos concentra-
tions before and during carpet cleaning were also compared.
This report presents only air monitoring results from dry vacuuming and
wet cleaning of asbestos-contaminated carpet to evaluate the potential for
fiber reentrainment during cleaning. The results of the carpet sample
analyses and the effectiveness of two cleaning methods in the removal of
asbestos fibers from contaminated carpet are presented in a separate report.
1
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
CONCLUSIONS
Both dry vacuuming and wet cleaning of carpet artificially contaminated
with asbestos fibers resulted in a statistically significant increase in
airborne asbestos concentrations. The increase did not vary significantly
with the type of cleaning method (wet or dry) or with the two levels of
asbestos contamination applied to the carpet. While this research observed
significant increases in airborne asbestos concentrations during cleaning
activities in a controlled study under artificial, simulated conditions, it
is not known if such increases occur in real-world custodial operations.
Obviously, this possibility is a concern.
RECOMMENDAT IONS
This research suggests that normal custodial cleaning of asbestos-con-
taminated carpet may result in elevated airborne asbestos concentrations.
Further research is needed to determine actual exposure risk to custodial
workers performing these activities in buildings containing friable asbestos—
containing materials.
2
-------
SECTION 3
STUDY DESIGN
TEST FACILITY
This study was conducted in an unoccupied building at Wright-Patterson
Air Force Base in Dayton, Ohio. Two rooms, each containing approximately 500
square feet of floor space, were constructed in a larger bay of the building.
Figure 1 presents the layout of the test facility. The rooms were
constructed of 2-in. x 4-in, lumber with studs spaced on 24-in, centers and
3/4-in, plywood floors. The ceiling, floor, and walls were double-covered
with 6-niil polyethylene sheeting. (The interior layer of polyethylene sheet-
ing was encapsulated and replaced after each experiment.) Where the joining
of separate sheets of polyethylene was necessary, the sheets were overlapped
at least 12 in. and joined with an unbroken line of adhesive to prohibit air
movement. Three-inch-wide tape was then used to seal the joint further on
both the inside and outside of the plastic sheeting.
Entry from one room to another was through a double-curtained doorway
consisting of two overlapping sheets of 6-mu polyethylene placed over a
framed doorway; each sheet was secured along the top of the doorway, and the
vertical edge of one sheet was secured along one vertical side of the doorway
and the vertical edge of the other sheet, along the opposite vertical side of
the doorway.
Room size (approximately 29 ft x 17 ft x 7.5 ft) was determined based on
the minimum amount of time required to vacuum or wet-clean the room and to
attain an adequate volume of sample air to achieve a specified analytical
sensitivity. A 52—inch, ceiling-mounted, axial-flow, propeller fan was
installed in each room to facilitate air movement and to minimize temperature
stratification.
Separate decontamination facilities for workers and waste materials were
connected to the experimental areas. The worker decontamination facility
consisted of three totally enclosed chambers as follows:
1) An equipment-change room with double curtained doorways, one to the
work area and one to the shower room.
2) A shower room with double—curtained doorways, one to the equipment
change room and one to the clean change room. The one shower
installed in this room was constructed so that all water was col-
lected and pumped through a three-stage filtration system. The
3
-------
HEPA-FILTERED
NEGATIVE AIR
UNIT
FRESH-AIR
INTAKE
SITE
OFFICE
PERSONNEL
DECON
• = AIR SAMPLE LOCATION
Figure 1. Layout of test facility.
-------
three-stage filtration system consisted of a 400-micrometer, nylon-
mesh, filter-bag prefilter; a 50-micrometer, filter-bag second-
stage filter; and a 5—micrometer final-stage filter. Filtrate was
disposed of as asbestos-contaminated waste. Water was drained from
the filtration system exit into a sanitary sewerage system.
3) A clean change room with double-curtained doorways, one to the
shower room and one to the noncontaminated areas of the building.
Air Filtration
High-efficiency particulate air filtration systems were used to reduce
the airborne asbestos concentrations to background levels after each experi-
rnent. These units were operated during both preparation and decontamination
of the test rooms. The air filtration units were not intended to be operated
during the carpet clearing phase of each experiment.
One HEPA filtration system was dedicated to each test room (Figure 1).
Each unit provided approximately 8 air changes per every 15—minute period.
The negative pressure inside the test rooms ranged from -0.08 to -0.06 in. of
water. All exhaust air passed through a HEPA filter and was discharged to
the outdoors (i.e., outside the building). All makeup air was obtained from
outside the building through a window located on the opposite side of the
building from the exhaust for the HEPA filtration systems.
EXPERIMENTAL DESIGN
Two carpet clearing methods, dry vacuuming with a HEPA—filtered vacuum
and wet cleaning with a HEPA-filtered hot—water extraction cleaner, were
evaluated on carpet artificially contaminated with approximately 100 million
and I billion asbestos structures per square foot (sift 2 ). Each combination
of cleaning method and contamination level was replicated four times. Four
different (same model) HEPA-filtered vacuums and four different (same model)
HEPA-filtered hot-water extraction units were used in this study so the
results would not be influenced by the peculiarities of a single unit. Each
machine was usea only once per combination of cleaning method and contamina-
tion level. This experimental design, which yielded a total of 16 experi-
ments, is summarized in Table 1.
TABLE 1. EXPERIMENTAL DESIGN
Approximate
contami nation
level, s/ft 2
Cleaning
method
Wet cleaning
Dry
vacuuming
100 million
Experiments 1,
4, 5, 8
2,
3, 6, 7
1 billion
9, 12, 13, 16
10,
11, 14, 15
5
-------
Two experiments were conducted each day of the study. Each combination
of cleaning method and contamination level was tested twice in each test
room. A single experiment consisted of contaminating a new piece of carpet
(approximately 500 square feet) with asbestos fibers, collecting work-area
air samples, dry vacuuming or wet cleaning the carpet while concurrently col-
lecting a second set of work area air samples, removing the carpet, and
decontaminating the test room. Each test room was decontaminated by encapsu-
lating the polyethylene sheeting on the ceiling and walls and the carpet
prior to their removal. These materials were replaced after each experiment.
SAMPLING STRATEGY
The number of samples collected was based, in part, on power calcula-
tions made during the design phase of the study. Statistical power is de-
fined as the probability of detecting a difference between two sets of mea-
surements (e.g., before and during cleaning) when a true difference actually
exists. The probability of detecting a difference depends on the absolute
magnitude of the airborne asbestos concentrations, their variability, and
their statistical distribution. For planning purposes, it was assumed that
individual airborne asbestos measurements would follow a negative binomial
distribution with a coefficient of variation of approximately 100 percent.
Table 2 shows the relationship between the number of samples and the proba-
bility of obtaining a statistically significant result at the 5 percent
level, assuming a t-test will be used to compare two groups of measurements.
Twelve samples per group are needed to detect a five-fold difference, with
high probability (greater than 0.85).
TABLE 2. PROBABILITY OF DETECTING A STATISTICALLY SIGNIFICANT
DIFFERENCE AT THE 5 PERCENT LEVEL OF SIGNIFICANCE BETWEEN
TWO GROUPS OF AIRBORNE ASBESTOS MEASUREMENTS
Assumed
airborne
asbestos
level
for
group 1 =
0.005 s/cm 3
Actual difference
between
groups
Number
of samples
per group
4
8
12
Twofold
Fivefold
Tenfold
0.11
0.34
0.69
0.23
0.73
0.95
0.26
0.88
1.0
Assumed
airborne
asbestos
level
for
group 1 =
0.02 s/cm 3
Actual difference
between
groups
Number
of samples
per group
4
8
12
Twofold
Fivefold
Tenfold
0.21
0.45
0.78
0.25
0.76
0.97
0.39
0.91
1.0
6
-------
The study was designed to achieve at least this power by having four
replicates of each experiment and three samples per replicate. The actual
power is expected to be greater than that indicated in Table 2 because the
design permits comparisons involving more than two sets of measurements
(i.e., analyses of variance rather than individual t—tests).
7
-------
SECTION 4
MATERIALS AND METHODS
A survey was made of 14 General Service Administration (GSA) field
offices in 11 States distributed across the United States to determine the
most prevalent types of carpet, HEPA-filtered vacuum cleaner unit, and HEPA-
filtered hot—water extraction unit to use in this study. Building managers
were asked to identify 1) the specific type and manufacturer of carpet used
in GSA buildings, 2) the manufacturer and model of HEPA-filtered vacuum
cleaner commonly used, and 3) the manufacturer and model of HEPA-filtered
hot-water extraction equipment routinely used in their buildings.
None of the GSA offices routinely wet-cleaned their carpet. When wet-
cleaning was necessary, contractors were hired to perform the work. There-
fore, six trade associations (the American Institute of Maintenance, the
Building Service Contractors Association, the International Maintenance
Institute, the Environmental Management Association, the International Sani-
tary Supply Association, and the Vacuum Cleaner Manufacturers Association)
were surveyed to obtain their recommendations on a HEPA-filtered hot—water
extraction cleaner.
SELECTION OF CARPET
Eight of the fourteen GSA offices indicated a preference for the same
manufacturer and type of carpet. The selected carpet was first—grade, 100
percent nylon, with 0.25-inch cut pile, 28 ounces of yarn per square foot,
and dual vinyl backing. The carpet was manufactured in roll sizes of 4.5 by
90 ft.
SELECTION OF CARPET CLEANING EQUIPMENT
HEPA-Filtered Vacuum
The HEPA—filtered vacuum selected for this study was the model most
frequently mentioned in the GSA survey. The unit had an airflow capacity of
87 cubic feet per minute and a suction power of 200 watts. The standard fil-
tration system consisted of a main cotton filter that permits a steady even
airflow and has a high retention efficiency and an exhaust diffuser that
insures a low exhaust velocity and additional air filtration. A HEPA exhaust
filter was added to this standard filtration system to trap small particles
and keep them from escaping into the air. The HEPA-filter had a retention
8
-------
efficiency rating of 99.97 percent for particles larger than 0.3 micrometer.
This unit was also equipped with a motor—driven carpet nozzle with a rotating
brush.
Hot—Water Extraction Cleaner
Three of the trade associations surveyed recommended the same hot-water
extraction unit. The selected cleaner was equipped with a HEPA-filtered
power head and a moisture—proof, continuous-duty, 2-horsepower vacuum motor
that develops a 100-inch waterlift. This unit was also equipped with an
extractor tool that uses a motor-driven 4-inch-diameter by 14-inch-long
cylindrical nylon-bristle brush to agitate and scrub the carpet during the
extraction process.
SAMPLING METHODOLOGY
Air samples were collected on open-face, 25-mm-diameter, 0.45-pm pore-
size, mixed cellulose ester membrane filters with a 5—pm pore-size, mixed
cellulose ester backup diffusing filter and cellulose ester support pad
contained in a three-piece cassette. The filter cassettes were positioned
approximately 5 feet above the floor with the filter face at approximately a
45-degree angle toward the floor. The filter assembly was attached to an
electric-powered vacuum pump operating at a flow rate of approximately
10 liters per minute. In each test room, the air samplers were positioned in
a triangular pattern (Figure 1). Air samples were collected for a minimum of
65 minutes before and during carpet cleaning to achieve a minimum air volume
of approximately 650 liters. The sampling pumps were calibrated both before
and after sampling with a precision rotameter.
ANALYTICAL METHODOLOGY
The mixed cellulose ester filters were analyzed by transmission electron
microscopy (TEM). These filters were prepared and analyzed in accordance
with the nonmandatory TEM method as described in the Asbestos Hazard Emer-
gency Response Act (AHERA) final rule (52 CFR 41821). Because no OSHA per-
missible exposure limits or NIOSH recommended exposure limits have been
established for airborne asbestos measured by TEM, a subset of filters was
selected for additional analysis by phase contrast microscopy (PCM) in ac-
cordance with NIOSH Method 7400. Battelle Laboratories, Columbus Division,
performed the TEM and PCM analyses on the field samples under separate con-
tract with EPA’s Risk Reduction Engineering Laboratory (RREL) in Cincinnati,
Ohio.
STATISTICAL ANALYSIS
Airborne asbestos concentrations were determined before and during
carpet cleaning to study the effect of the cleaning method and contamination
loading on fiber reentrainment during carpet cleaning. Three work-area sam-
ples were collected before and during the carpet cleaning for each experi-
ment. A single estimate of the airborne asbestos concentrations before and
9
-------
during cleaning was then determined by averaging the three respective work-
area samples. As a measure of relative change in airborne asbestos concen-
tration, the ratio of the concentration during cleaning to the concentration
prior to cleaning was computed. The natural log of this ratio was then ana-
lyzed by using a two-factor analysis of variance (ANOVA)’ with the cleaning
method and contamination level as the main factors. The two-factor interac-
tion term was also included in the model. This analysis is equivalent to
assuming a lognornial distribution for airborne asbestos measurements and
analyzing the log-transformed data for differences between . irborne asbestos
concentrations before and during cleaning. The lognormal distribution is
commonly assumed for measurements of asbestos and other air pollutants.
Sumary statistics (arithmetic mean and standard deviation) were calculated
according to cleaning method and contamination level.
10
-------
SECTION 5
EXPERIMENTAL PROCEDURES
PRESTUDY AIR MONITORING
Before construction of the contamination enclosure system, air samples
were collected to determine a baseline airborne asbestos concentration inside
the test facility. Seven interior air samples and two field blanks were
collected in accordance with sampling procedures described in Section 4. The
air samples were collected for a period of approximately 200 minutes to
achieve a minimum air volume of 1260 liters for each sample. These samples
were analyzed in accordance with the nonmandatory TEM method as described in
the AHERA Final Rule.
The average airborne asbestos concentration for the seven samples col-
lected was 0.0031 s/cm 3 . The TEM analysis of the seven samples yielded a
total of 6 asbestos structures (4 chrysotile and 2 amphibole). One chryso-
tile fiber was detected on each field blank. Table 3 summarizes these re-
sults.
TABLE 3. SUMMARY OF PRESTUDY AIRBORNE ASBESTOS
CONCENTRATIONS IN TEST FACILITY
Number of
structures Concentration,
Sample observed s/cm 3
001
1
0.0028
002
0
<0.0039
003
2
0.0077
004
0
<0.0038
005
1
0.0039
006
1
0.0039
007
1
0.0038
Field
blank
1
Field
blank
1
CARPET CONTAMINATION
Selected levels of carpet contamination for this study were based on
field data reported by Wilmoth et al. 2 Asbestos concentrations in
contaminated carpet ranging from approximately 8000 s/ft 2 to 2 billion
s/ft 2 were detected by use of a microvac technique. Bulk sample sonication
11
-------
of the samples revealed levels ranging from 30 million to 4 billion s/ft 2 .
Based on these preliminary results, the target experimental asbestos
contamination levels of approximately 100 million and 1 billion s/ft 2 were
thought to represent carpet contamination likely to be present in buildings
where asbestos—containing materials are present.
The carpet was contaminated with a spray—applied dispersion of Union
International Centre le Centre Calidria chrysotile asbestos in distilled
water. The asbestos was dispersed uniformly on the carpet by use of a manual
pesticide sprayer equipped with a stainless steel container.
Preparation of Concentrated Aqueous Suspensions of Chrysotile
Aqueous suspensions of chrysotile are not stable for long periods unless
they are specially prepared. 3 Even small amounts of high-molecular-weight
organic materials, such as those generated by bacteria, result in the desta-
bilization of chrysotile suspensions and the attachment of fibers to the
walls of the container. This process can be reversed only by carrying out
oxidation of the organic materials with ozone and ultraviolet light treat-
ment. 3 If precautions are taken to exclude all organic materials and to
prevent bacterial growth, however, chrysotile suspensions can be prepared
that remain stable for several years. This can be achieved by sterilizing
all containers used in the preparation, using freshly distilled water for the
dispersion process and storing the preparation in flame—sealed glass arnpules
that are autoclaved immediately after sealina.
For this project, the decision was made to prepare sealed ampules of
fiber dispersions so that the contents of one ampule dispersed in 6 liters of
freshly distilled water would provide the concentration of suspension re-
quired for artificial contamination of one 500-ft 2 sample of carpet. Calcu-
lations of the amount of chrysotile required were based on the assumption
that all of the fibers needed to contaminate one carpet sample would be
contained in a volume of 50 ml sealed in one ampule.
For the higher of the two concentrations used, the fiber concentration
required in each ampule was calculated as follows:
Higher contamination level required = 1O fibers/ft 2
Number of fibers required to contaminate
500 ft 2 = 6.5 x 1011 fibers
Fiber concentration required for this
number of fibers to be in a volume of 50 ml = 1.3 x iO’ fibers/liter
The lower of the two concentrations used was a factor of 10 lower than
this. To ensure an exact factor of 10 ratio between the two concentrations,
the lower—concentration dispersion was prepared by diluting an aliquot of the
high—concentration dispersion.
Because the original suspension was to be prepared by dispersing a known
weight of chrysotile in water, knowledge of what numerical concentration of
12
-------
fibers would result from this dispersion was required. Previous work on
preparation of ampules indicated that a suspension of purified Calidria
chrysotile in water, with a mass concentration of 1 pg/liter yielded a numer-
ical fiber concentration of approximately 200 million fibers per liter.
Based on this conversion, the weight of chrysotile is calculated as follows:
Weight required = 1.3 x i0 x 106/(2 x 108) g/liter
= 65 mg/liter
Therefore, the preparation of 1.5 liters of a suspension with this concentra-
tion requires 97.5 mg of chrysotile.
The calculation for determining the mass of chrysotile reQuired is based
on data from very dilute suspensions. Initial experiments indicated that
some difficulty could arise in obtaining complete dispersal of the chrysotile
at the high concentrations in this program; if some aggregation were to
occur, the numerical structure count would be somewhat lower than that re-
quired. For this reason, the suspensions were prepared to have a higher mass
concentration than that indicated in the preceding calculation.
Before preparation of the fiber suspensions, the 50-mi ampuies were
thoroughly cleaned. Each ampule was filled to the top with freshly distilled
water and placed in an ultrasonic bath for a period of 15 minutes; the water
was then removed by suction. This process was repeated twice, and the
ampules were then considered ready for filling.
The higher-concentration chrysotile suspension was prepared first. All
water used for preparation of these dispersions was freshly distilled (within
8 hours of preparation). A weight of 409.5 mg of purified Calidria chryso•-
tile was placed in an agate mortar and lightly ground with a small volume of
water by use of a pestle. More freshly distilled water was added gradually
until a creamy liquid was obtained. Up to 400 mL of this liquid was made up
in a disposable polypropylene beaker, and the beaker was placed in an ultra-
sonic bath for approximately 30 minutes. Up to 1500 ml of the chrysotile
suspension was then made up with water in a 1-gallon polyethylene bottle.
The bottle was placed in an ultrasonic bath for approximately 30 minutes.
During this time, the bottle was removed several times and shaken vigorously.
For the lower-concentration suspension, a volume of 150 ml, up to 1500 ml of
this suspension was made up with water in another 1-gallon polyethylene
bottle. The two suspensions had concentrations of 273 and 27.3 mg/liter,
respectively.
A disposable polyethylene funnel was used to place a volume of 50 mL of
suspension in each of the ampules. This left adequate space in the ampule to
permit efficient shaking of the contents. The filled anipules were flame-
sealed immediately and then autoclaved for 30 minutes at a temperature of
121°C to sterilize the contents. After the ampules cooled, they were labeled
in the order of their filling.
13
-------
Preparation of Asbestos Dispersion
The following steps were followed precisely in the preparation of the
asbestos dispersions used to contaminate the carpet:
1. All water used for dilution of the anipules of chrysotile suspension
was freshly distilled from a glass still.
2. Before the ampule was opened, it was shaken vigorously for 1 minute
and then placed in an ultrasonic bath for 30 minutes. During the
ultrasonic treatment, the ampule was removed every 5 minutes and
again shaken vigorously for 1 minute.
3. A new 32-ounce glass bottle was washed with several changes of
freshly distilled water. The ampule was then opened, and the
entire contents were emptied into 450 ml of freshly distilled water
in the glass bottle. For the high—concentration ampules only, the
pH was adjusted to approximately 4.0 by adding approximately 300 to
400 pl of glacial acetic acid. The bottle was capped, shaken
vigorously, and then placed in an ultrasonic bath for 15 minutes.
No surface active agents were added.
4. The pesticide sprayer was sterilized and cleaned by rinsing it with
a 10 to 15 percent solution of Clorox for approximately 15 minutes.
The sprayer, including the interior of the outlet pipe, was then
thoroughly washed with several changes of freshly distilled water.
5. The sprayer was filled with 5.5 liters of freshly distilled water,
and the contents of the bottle were added. The sprayer was then
shaken before the carpet was sprayed.
The sprayer was not allowed to dry before it was washed after each
experiment because chrysotile is much more difficult to remove from the
interior surfaces when it has dried.
To ensure no bacterial growth had occurred in the sprayer between uses,
the inside of the sprayer and the outlet pipe were treated with a 10 to 15
percent solution of Clorox to remove any bacteria and their byproducts. Any
bacterial growth would scavenge fibers from the suspension and cause fibers
to become attached to the wall of the container. The container and outlet
pipe were then rinsed with isopropyl alcohol.
Concentrations of Suspensions
Several of the ampules were used to make precise measurements of the
fiber concentrations andalso to determine the fiber size distributions. In
order to measure these very high fiber concentrations, a total dilution fac-
tor of 1 in 25,000 was necessary for the low-concentration ampules, and 1 in
250,000 for the high—concentration arnpules. This was achieved by successive
dilutions in freshly distilled water. For the low-concentration ampules, the
contents of one ampule were first dispersed in 500 nil. In the second dilu-
tion, 10 ml were dilutedto 500 ml, and 10 nil of the second dilution were
14
-------
diluted to 500 nil. Three filters were prepared from this final suspension,
using the EPA Analytical Method for Determination of Asbestos Fibers in
Water. For the high—concentration ampules, the final suspension was diluted
by a further factor of 10 before preparation of the filters.
The dilution factors and the volumes of suspension filtered were select-
ed to yield fiber counts of approximately 40 per grid opening. One fiber
count incorporating approximately 600 asbestos structures was made for each
of the two concentrations.
It was found that the high—concentration anipules yielded asbestos struc-
ture counts which were significantly lower than those obtained during the
initial tests on the suspension at the time the ampules were prepared. This
effect was investigated, and found to be a consequence of a rise in pH of the
suspension after packing and autoclaving. The increase in the pH was probab-
ly due to some leaching of the chrysotile during the autoclave treatment,
giving rise to destabilization of the dispersion, and aggregation of the
fibers into bundles and clusters. The effect was found to be reversible by
adjusting the pH of the dispersion to approximately 4.0 with acetic acid at
the time of the first dilution. The measurements on the high-concentration
ampules were repeated using another ampule and adjusting the pH during prepa-
ration of the first dilution. The aggregation effect did not occur in the
low-concentration ampules, and therefore no pH adjustment was required when
these ampules were used.
Table 4 shows the results of the fiber concentration measurements made
on the low- and high—concentration ampules. The analysis of the laboratory
dilution was continued for approximately 600 chrysotile structures to provide
a precise concentration value and to provide a size distribution with a
sufficient number of structures in each size classification. Appendix A
contains the size distributions for the measurements made on the low- and
high-concentration ampules. Figure 2 illustrates the fiber size distribution
in the low— and high—concentration ampules.
TABLE 4. SUMMARY OF RESULTS OF TRANSMISSION ELECTRON MICROSCOPY
ANALYSES FOR LOW AND HIGH CONCENTRATION AMPULES
Sample
description
Fiber type
Structure concentration, 1012 structures/liter
Mean
95% con-
fidence
interval
Analytical
sensitivity
Equivalent
volume
sampled,
pl
No. of
struc-
tures
counted
Low-concentra-
tion ampule
Chrysotile
2.2
2.0-2.5
0.0036
0.400
619
High-concentra-
tion ampule
Chrysotile
25
22-27
0.0409
0.040
601
15
-------
I
I Low Concentration I
—I Suspension L........
Total Fibers 619 J
100.0
160-
140-
120.
100
80-
60.
40
20
0
160
140
120
100
80
60
40
20
0
Fiber Length, micrometers
Figure 2. Distribution of chrysotile fiber lengths in the low and high concentration aqueous asbestos
suspensions.
I
I
1.0 10.0
Fiber Length, micrometers
1.0 10.0
100.0
16
-------
Application of Dispersion to Carpet
A meticulously cleaned hand-pumped garden sprayer was used to apply the
asbestos dispersion to the carpet. A fixed number of pumps was used for each
batch to provide consistent spray pressure. The desired controlled spray was
experimentally determined by trial and error before the tests with asbestos
began. The pressure was kept within the desired range by adding a fixed
number of pump strokes after each fixed area was sprayed in a predetermined
pattern by following a grid work of string placed over the carpet before the
beginning of each experiment. The tank was periodically agitated to help
keep the asbestos fibers suspended. Dehumidifiers were placed in the room
overnight to aid in drying the carpet. The following day a 200—pound steel
lawn roller was rolled over the carpet surfaces to simulate the effects of
normal foot traffic in working the asbestos into the carpet.
Carpet Cleaning Technique
The carpet was vacuumed or wet-cleaned for a period of approximately 65
minutes to allow the collection of a sufficient volume of air samples to
obtain an analytical sensitivity of 0.005 s/cm 3 of air. The carpet was
cleaned in two directions, the second direction at a 90-degree angle to the
first.
DISPOSAL OF ASBESTOS-CONTAINING MATERIAL
Asbestos—contaminated materials, including carpeting, polyethylene, pro-
tective clothing, etc., were placed in disposable 6-mil polyethylene bags and
labeled according to EPA regulations. When filled, the disposal bags were
sealed, sponged clean, and removed from the test room to the primary waste-
loadout work area (Figure 1). The disposal bag was then sponged a second
time, taken through the equipment-change area, and placed in the shower
chamber for a thorough washing. The clean disposal bag was taken into the
clean chamber, loaded into a fiberboard drum, labeled with an EPA-approved
asbestos warning label, and transported to a disposal site approved by the
Ohio Environmental Protection Agency.
SITE CLEANUP
Prior to removal of the primary polyethylene barrier (i.e., the first
barrier installed to isolate the work area, including test rooms), the sur-
face was thoroughly wet-wiped with amended water. The HEPA filtration system
continued to operate during site cleanup.
All debris and waste resulting from the experiments were removed from
the building. All the drummed waste was removed from the site and disposed
of in an Ohio EPA—approved landfill.
17
-------
POSTSTUDY AIR MONITORING
After removal of the polyethylene sheeting from the floor, ceiling, and
walls, air samples were collected to determine the airborne asbestos concen-
trations inside the building. Four interior air samples were collected in
accordance with the sampling procedures described in Section 5. These sam-
ples were collected for a period of approximately 180 minutes to achieve a
minimum air volume of approximately 1800 liters for each sample. These
samples were analyzed in accordance with the nonrnandatory TEM method as
described in the AHERA Final Rule. No asbestos was detected in any of these
samples.
18
-------
SECTION 6
QUALITY ASSURANCE
The Quality Assurance Project Plan (QAPP) contains complete details of
the quality assurance procedures followed during this research project. The
procedures used for this study are summarized in the following subsections.
SAMPLE CHAIN OF CUSTODY
Sample chain-of-custody procedures were an integral part of both sam-
pling and analytical activities during this study. They were followed for
all air samples collected. The applied field custody procedures documented
each sample from the time of its collection until its receipt by the ana-
lytical laboratory. Internal laboratory records then documented the custody
of the sample through its final disposition.
Standard sample custody (traceability) procedures were used. Each
sample was labeled with a unique project identification number, which was
recorded in the field log book along with other information specified by the
QAPP.
QUALITY ASSURANCE SAMPLE ANALYSES
Specific quality assurance procedures for ensuring the accuracy and pre-
cision of the TEM analyses of air samples included the use of lot, laborato-
ry, and field blanks and replicate and duplicate analyses.
Lot Blanks
Filter lot blanks consist of unused filters selected at random and sub-
mitted for prescreening analysis for background asbestos contamination before
the start of field work to determine the integrity of the entire lot of
filters purchased for EPA research studies. One hundred lot blanks were
submitted for TEM analysis. No asbestos structures were detected in the
1000 grid openings analyzed. The lot of filters was subsequently considered
acceptable for use.
Field and Laboratory Blanks
During the setup of the air sampling pumps, preloaded filter cassettes
were labeled and handled in a manner similar to that for the actual sample
filters, but they were never attached to the pump. One field blank was
19
-------
collected for each of the 16 experiments. Two of the 16 filters each contained
1 asbestos structure. Also, prior to each of the 16 experiments, one sample
cassette was selected from the filter inventory to be used as a laboratory
blank. These samples were sealed and submitted for use by the analytical
laboratory to ensure against any blank interference during the analytical
procedures. Two of the 16 sealed blanks each contained 2 asbestos struc-
tures. Analysis of the field and laboratory blanks demonstrated that filter
contamination was comparable to background levels of asbestos air filters
(defined as 70 s/mm 2 in AHERA). Table 5 summarizes the results of the field
and laboratory blanks.
TABLE 5. SUMMARY OF FIELD AND LABORATORY BLANK ANALYSES
Asbestos
conc
entration,
Experiment
s/mm 2
Field blank
Laboratory
blank
1
<14
<14
2
14
<14
3
14
28
4
<14
<14
5
<14
28
6
<14
<14
7
<14
<14
8
<14
<14
9
<14
<14
10
<14
<14
11
<14
<14
12
<14
<14
13
<14
<14
14
<14
<14
15
<14
<14
16
<14
<14
Duplicate and Replicate Sample Analyses
Duplicate sample analysis provides a means of quantifying iritralaborato-
ry precision and refers to the analysis of the same grid preparation by a
second microscopist. Five samples were randomly selected for duplicate anal-
ysis. Replicate sample analysis provides a means of quantifying any
analytical variability introduced by the filter preparation procedure and
refers to the analysis of a second grid preparation from the original filter.
Five samples were randomly selected for replicate analysis.
The coefficient of variations for the duplicate and replicate analyses
were estimated by assuming a lognormal distribution for the data on the
original scale and estimating the variance on the log scale. The variance
was estimated by the mean square error obtained from a one-way ANOVA of the
log-transformed data with saniple ID as the experimental factor. The co-
efficient of variations associated with the duplicate and replicate sample
20
-------
analyses were 22 and 32 percent, respectively. Since the replicate
analyses used different filter preparations, a higher coefficient of
variation is not unexpected. Table 6 presents the results of the duplicate
and replicate analyses.
TABLE 6. RESULTS OF REPLICATE AND DUPLICATE SAMPLE ANALYSES
Sample
Original
Duplicate
Rep
licate
N
s/cm 3
N
s/cm 3
N
s/cm 3
O1-A444B
47
0.1810
33
0.1271
-
-
04-A464D
37
0.1242
39
0.1309
-
—
07-A482D
57
O.275
53
0.2565
-
-
14-A525D
53
0.3368
50
0.2174
-
-
16-A533B
8
0.0306
12
0.0459
-
-
02-A451D
2
0.0070
-
-
2
0.0070
05-A467B
6
0.0220
-
-
4
0.0147
1O-A500D
51
0.4891
-
-
51
0.3113
13 —A516B
19
0.0719
-
-
10
0.0378
15—A529D
41
0.1482
-
-
26
0.0940
SPRAY-APPLICATION TECHNIQUE
To confirm the validity of the spraying technique, an additional experi-
ment was conducted using a pesticide sprayer identical to those used to apply
the chrysotile to the carpet samples. An ampule of low-concentration suspen-
sion was diluted to 500 ml, and then further diluted to 6 liters in the
pesticide sprayer, using freshly distilled water. The sprayer was thoroughly
shaken, and the contents, were sprayed out into several containers. Three
500-nil samples of the spray were collected, one at the beginning of spraying,
one when approximately 50 percent of the contents had been discharged, and
one just before the end of spraying. These three samples were analyzed to
determine if the concentration and size distribution of the fibers changed
during the period of spraying. Structure concentrations for the three sam-
ples are presented in Table 7. These results indicate no significant loss of
fibers during the transfer of the diluted liquid suspension through the
sprayer’s hose and nozzle.
The size distributions for these three samples are shown in Table 8 and
illustrated in Figure 3. Since the distributions all approximate logarith-
mic-normal, the size range intervals for calculation of the distribution must
be spaced logarithmically. Another characteristic required for the choice of
size intervals is that they allow for a sufficient number of size classes,
while still retaining a statistically—valid number of fibers in each class.
Interpretation is also facilitated if each size class repeats at decade
21
-------
TABLE 7. RESULTS FROM PRELIMINARY STUDY OF ASBESTOS DISPERSION BY
SPRAYING--FIBERS AND FIBER BUNDLES
Volume in sprayer
at time of sample
collection, liters
Structure concentration,
1012 structures/liter
95% con- Number of
fidence Analytical structures
Fiber type Mean interval sensit.ivity counted
6
(Beginning of spray)
4
(50% point of spray)
Chrysotile 2.33 1.87-2.79 0.0118 198
Chrysotile 2.18 1.54-2.82 0.0118 185
2
(End of spray)
Chrysotile 2.38 1.90-2.85 0.0118 202
TABLE 8. FIBER
LENGTH DISTRIBUTIONS FROM THE PRELIMINARY STUDY OF
ASBESTOS DISPERSION BY SPRAYING
Particle
size range, lim
Number of fibers, fiber bundles (cumulative percentage)
Beginning of spray 50% point of spray End of spray
0.23-0.34
0 (0) 0 (0) 0 (0)
0.34-0.50
0 (0) 0 (0) 0 (0)
0.50-0.73
28 (14.14) 33 (17.84) 24 (11.88)
0.73-1.08
48 (38.38) 55 (47.57) 43 (33.17)
1.08-1.58
34 (55.56) 28 (62.70) 45 (55.45)
1.58-2.32
30 (70.71) 20 (73.51) 28 (69.31)
2.32-3.41
34 (87.88) 17 (82.70) 22 (80.20)
3.41-5.00
18 (96.97) 14 (90.27) 19 (89.60)
5.00-7.34
4 (98.99) 10 (95.68) 13 (96.04)
7.34-10.77
1 (99.49) 5 (98.38) 5 (98.51)
10.77-15.81
1 (100.00) 3 (100.00) 1 (99.01)
15.81-23.21
0 (100.00) 0 (100.00) 1 (99.50)
23.21-34.06
0 (100.00) 0 (100.00) 0 (99.50)
34.06-50.00
0 (100.00) 0 (100.00) 1 (100.00)
22
-------
Number of Fibers Observed
0.5-0.7 0.7- 1.1 1.1-1.6 1.6-2.3 2.3-3.4 3.4-5.0 5.0-7.3 7.3- 10.8 10.8- 15.8 15.8-23.2
Particle Size Range, micrometers
Figure 3. Fiber size distributions from preliminary study
35
30
25
20
15
10
5
0
F )
(. .)
BEGINNING OF SPRAY
50% POINT OF SPRAY
LIII END OF SPRAY
of asbestos dispersion by spraying.
-------
intervals. A ratio of 1.468 from one class to the next satisfies all of
these requirements. The other constraint is that the length distribution
should include the minimum fiber length of 0.5 irn at the first interval
point. The decade repeat automatically ensures that the other significant
fiber length of 5 i.im occurs as an interval point.
No significant change in the fiber size distribution was evident during
the transfer of the diluted liquid suspension.
24
-------
SECTION 7
RESULTS AND DISCUSSION
Figure 4 presents the average airborne asbestos concentrations measured
before and during cleaning for each cleaning method and carpet contamination
loading. The samples collected before cleaning were obtained after the
carpet was contaminated to determine the baseline concentration in the test
room. Table 9 presents the summary statistics (arithmetic average and stan-
dard deviation). Individual air sampling results analyzed by TEM are listed
in Appendix B.
Air sampling results from 2 of the 16 experiments showed that the aver-
age airborne asbestos concentrations decreased during both wet cleaning and
dry vacuuming of the carpet. The explanation for this anomaly is that the
HEPA filtration system used to ventilate the test rooms was inadvertently
operating during the carpet cleaning phase of these two experiments. There-
fore, these results were omitted from the statistical analysis of the data.
Results from the two-factor ANOVA are summarized in Table 10. There was
no statistically significant interaction between cleaning method and con-
tamination level (p = 0.8901). That is, the effect of cleaning method on
airborne asbestos did not vary significantly with contamination level. No
statistically significant difference was evident between cleaning methods
with respect to fiber reentrainment (p = 0.5847); that is, the mean relative
increase in airborne asbestos concentration during carpet cleaning with a dry
vacuum was not significantly different from that found during wet cleaning.
Similarly, no statistically significant difference was evident between carpet
contamination loadings with respect to fiber reentraininent (p 0.0857); that
is, the mean relative increase in airborne asbestos concentrations during
carpet cleaning when the carpet contamination level was 100 million s/ft 2 was
not significantly different from that found when the carpet contamination
loading was 1 billion s/ft 2 . The ANOVA results do, however, indicate that,
overall, the mean airborne asbestos concentration was significantly higher
during carpet cleaning than just prior to cleaning (p = 0.0001). Specifi-
cally, a 95 percent confidence interval for the mean airborne asbestos con-
centration during carpet cleaning as a proportion of the airborne concentra-
tion before cleaning showed that the mean airborne asbestos concentration was
between two and four times greater during carpet cleaning.
Airborne Asbestos Fiber Distribution
The TEM analysis of the 95 work-area samples before and during cleaning
yielded a total of 2839 structures. Of these, 2757 (97.1%) were chrysotile,
25
-------
0.3
0.25
0.2
0.15
0.1
0.05
0
Figure 4.
Hot Water Dry Hot Water Dry
Extraction Vacuum Extraction Vacuum
Average airborne asbestos concentrations before and during
carpet cleaning.
Average
Airborne
Asbestos /7”
Concentration //
(s/cm 3 )
Low Contamination
High Contamination
r )
, öjöi
During
Before
-------
TABLE 9. SUMMARY STATISTICS FOR AIRBORNE ASBESTOS CONCENTRATIONS
BEFORE AND DURING CARPET CLEANING
Airborne asbestos
Approximate
contamination HEPA-
loading, filtered Number of
s/ft 2 cleaner data 01 t 5 a
concentration, s/Cm 3
Standard
Average deviation
Before cleaning
100 million Hot-water 3
0.0673 0.0874
extraction
Dry-vacuum 3
0.0571 0.0315
During cleaning
Hot-water 3
0.1639 0.0911
extracti on
Dry-vacuum 3
0.2531 0.1655
Before cleaning
1 billion Hot-water 4
0.0761 0.0471
extraction
Dry-vacuum 4
0.1424 0.1235
During cleaning
Hot-water 4
0.1577 0.0690
extracti on
Dry-vacuum 4
0.2248 0.1499
a Each data point is the average of three work-area
samples.
TABLE 10. SUMMARY OF ANOVA RESULTS FOR AIRBORNE
ASBESTOS CONCENTRATIONS
MEASURED BEFORE AND DURING CARPET
CLEANING
Degrees of Sum of
Source of variation freedom squares
F value P value
Contamination level 1 1.5326
3.63 0.0857
Cleaning method 1 0.1345
0.32 0.5847
Interaction 1 0.0U85
0.02 0.8901
Average 1 15.5827
36.94 0.0001
Error 10 4.2179
27
-------
8 (0.03%) were amphibole, and 74 (2.6%) were ambiguous. The structure nior-
phology distribution is surm arized in Table 11.
TABLE 11. STRUCTURE MORPHOLOGY DISTRIBUTION FOP AIR SAMPLES
COLLECTED BEFORE AND DURING CARPET CLEANING
Structure
type
Number
bundi
of
es
Number of
clusters
Number of
fibers
Number of
matrices
Total
Chrysotile
30
7
2661
59
2757
Amphibole
0
2
5
1
8
Ambiguous
2
0
70
2
74
Total
32
9
2736
62
2839
These data indicate that the original chrysotile fibers used to prepare
the diluted asbestos suspension remained intact as fibers. There appeared to
be no significant tendency for the fibers to clump together as a result of
the suspension preparation, the carpet contamination, or the cleaning tech-
nique.
The presence of amphibole asbestos fibers in the air was probably due to
conditions existing prior to the experiment. Prestudy air monitoring identi-
fied two amphibole asbestos fibers in seven air samples collected.
Appendix C presents the structure-length distributions of asbestos par-
ticles found in the air before and during carpet cleaning. Eighty—four
percent of the chrysotile structures identified were 1 micrometer or less in
length. Only nine particles were identified with lengths greater than 5
micrometers. Figure 5 compares the fiber sizes of airborne asbestos during
carpet cleaning with fibers in the low- and high-concentration asbestos
suspensions. For example, approximately 60 percent f the asbestos fibers
used to contaminate the carpet with 100 million s/ft were greater than
1.1 pm. Less than 15 percent of the fiber observed in the air during carpet
cleaning were greater than 1.1 pm. These data suggest that the larger as-
bestos particles either remained in the carpet or were prevented from escap-
ing into the air by the carpet cleaning activity.
Figure 6 presents average airborne asbestos concentrations based on
particles greater than or equal to a given length. These “cumulative 0
concentrations illustrate that for both dry vacuuming and wet cleaning, the
overall airborne asbestos concentrations observed in this study were based
primarily on asbestos structures less than 1.5 pm in length.
Samples Analyzed by PCM
Twelve samples were selected to be analyzed by phase contrast microscopy
(PCM) based on their respective high asbestos concentrations determined by
28
-------
% Of Fibers
>.7 >1.1
Low Carpet Contamination,
100 million s/ft 2
Asbestos Suspension
Dry Vacuuming
Wet Cleaning
jiLtF
I I I
>1.6 >2.3 >3.4 >5.0 >7.3 10.8
Fiber Length, micrometers
High Carpet Contamination,
1 billion s/ft 2
Asbestos Suspension
Dry Vacuuming
Wet Cleaning
Fiber Length, micrometers
Figure 5. Comparative plot of cumulative percentages of airborne asbestos fibers
during dry vacuuming and wet cleaning of carpet with asbestos fibers in
the low and high concentration suspensions.
100
90
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10-
0-
I -— —
% of Fibers
100
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10
0
>.5 ‘.7 >1.1
>1.6 >2.3 >3.4 >5.0 >7.3 10.8
29
-------
Average Asbestos Concentration, s/Cm
Low Carpet Contamination,
100 million s/ft 2
— Dry Vacuuming
Wet Cleaning
I I
‘0 >0.5 >1.0 ‘1.5 ‘2 0 ‘2 5 ‘3.0 >3.5 ‘4.0 ‘4 5 ‘5 0 ‘5 5
Fiber Length, micrometers
Average Asbestos Concentration, s/cm
0.25
0 225
02
0 175
0 15
o 125
01
0 075
0 05
0.02 5
0
‘0 0 ‘0 5 ‘1.0 ‘1.5 ‘2.0 ‘2 5 ‘3.0 >3.5 >4.0 ‘4.5 >5.0 ‘5 5
Fiber Length, micrometers
Figure 6. Airborne asbestos concentrations for varying fiber lengths for samples
collected during dry vacuuming and wet cleaning of carpet.
30
03
0.275
0.25
o 225
02
0 175
0 15
0.125
01•
o 075
0.05-
0.025 -
0-
High Carpet Contamination,
1 billion s/ft 2
-------
TEM. Results from both TEM and PCM analyses are compared in Table 12. As
expected, airborne fiber concentrations determined by PCM were significantly
lower than the corresponding asbestos concentrations determined by TEM. This
difference is presumably due to the limitation of PCM to detect small fibers.
Furthermore, the majority of asbestos fibers applied (Figure 2) did not meet
the dimensional criteria (length >5 m) of NIOSH Method 7400 and hence were
not counted.
TABLE 12. COMPARISON OF TEM AND PCM ANALYSES OF SELECTED AIR SAMPLES
Sample
number
PCM fiber
concentration,
f/cm 3
TEM asbestos
concentration,
s/cm 3
03-A457D
0.0035
0.5507
03-A458D
0.0023
0.3658
03-A459 0
0.0081
0.3464
1O-A496B
0.0026
0.3656
1O-A497B
0.0078
0.2909
1O-A498B
0.0068
0.3375
10-A499D
0.0116
0.3871
1O-A500D
0.0109
0.4891
1O-A5O1D
0.0000
0.0070
14—A523D
0.0061
0.3177
14-A524D
0.0138
0.3779
14-A525D
0.0138
0.3368
31
-------
REFERENCES
1. Neter, J., W. Wasserman, and M. H. Kutner. Applied Linear Statistical
Models. 2nd Ed. Richard D. Irwin, Inc., Homewood, Illinois. 1985.
2. Wilmoth, R., 1. J. Powers, and J. R. Millette. Observations in Studies
Useful to Asbestos 0&M Activities. Presented at the National Asbestos
Council Conference in Atlanta, Georgia, February 1988.
3. Chatfield, E. J., and N. J. Dillon. Analytical Method for Determination
of Asbestos Fibers in Water. PB 83-260-471. U.S. Environmental Re-
search Laboratory, Athens, Georgia. Contract 68-03—2717. National
Technical Information Service, Springfield, Virginia. 1983.
4. Chatfield, E. J., M. J. Dillon, and W. R. Stott. Development of Im-
proved Analytical Techniques for Determination of Asbestos in Water Sam-
ples. PB 83-261-471. U.S. Environmental Research Laboratory, Athens,
Georgia. 1983.
32
-------
APPENDIX A
CHRYSOTILE FIBER SIZE DISTRIBUTION
IN THE HIGH- AND LOW-CONCENTRATION AMPULES
TABLE A-i. FIBER LENGTH DISTRIBUTION IN THE
LOW CONCENTRATION AMPULE
Number
Particle
size range, pm
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23 - 0.34
0
0
0.00
0.00
0.34 - 0.54
0
0
0.00
0.00
0.50 - 0.73
107
107
17.29
17.29
0.73 - 1.08
147
254
23.75
41.03
1.08 - 1.58
106
360
17.12
58.16
1.58 - 2.32
90
450
14.54
72.70
2.32 - 3.41
69
519
11.15
83.84
3.41 - 5.00
57
576
9.21
93.05
5.00 - 7.34
26
602
4.20
97.25
7.34 - 10.77
11
613
1.78
99.03
10.77 - 15.81
5
618
0.81
99.84
15.81 - 23.21
0
618
0.00
99.84
23.21 - 34.06
1
619
0.16
100.00
34.06 - 50.00
0
619
0.00
100.00
50.00 - 73.40
0
619
0.00
100.00
73.40 - 107.70
‘ 0
619
0.00
100.00
107.70 - 158.10
0
619
0.00
100.00
158.10 - 232.10
0
619
0.00
100.00
232.10 - 340.60
0
619
0.00
100.00
33
-------
TABLE A-2. FIBER LENGTH DISTRIBUTION IN THE
HIGH CONCENTRATION AMPULE
Number
Particle
size range, pm
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23 - 0.34
0
0
0.00
0.00
0.34 - 0.54
0
0
0.00
0.00
0.50 - 0.73
101
101
16.81
16.81
0.73 - 1.08
135
236
22.46
39.27
1.08 - 1.58
119
355
19.80
59.07
1.58 - 2.32
85
440
14.14
73.21
2.32 - 3.41
82
522
13.64
86.86
3.41 - 5.00
40
562
6.66
93.51
5.00 - 7.34
20
582
3.33
96.84
7.34 - 10.77
16
598
2.66
99.50
10.77 - 15.81
3
601
0.50
100.00
15.81 - 23.21
0
601
0.00
100.00
23.21 - 34.06
0
601
0.00
100.00
34.06 - 50.00
0
601
0.00
100.00
50.00 - 73.40
0
601
0.00
100.00
73.40 - 107.70
0
601
0.00
100.00
107.70 - 158.10
I o
601
0.00
100.00
158.10 - 232.10
0
601
0.00
100.00
232.10 - 340.60
0
601
0.00
100.00
34
-------
APPENDIX B
TOTAL AIRBORNE ASBESTOS STRUCTURE
CONCENTRATIONS BEFORE AND DURING
CARPET CLEANING FOR SAMPLES ANALYZED
BY TRANSMISSION ELECTRON MICROSCOPY
NOTE: Sample numbers ending with “B” indicate that the sample was
taken before the experiment; those ending with “D” indicate
that the sample was taken during the experiment.
35
-------
Sa mple Number of Asbestos Concentration,
Number Asbestos Str. S/Cm 3 s/nun 2
EXPERIMENT 2. - WET CLEAN
01—A442B 21 0.0809 170
01—A443B 25 0.0963 203
01—A444B 47 0.1810 381
01—A445D 25 0.0996 181
01—A446D 15 0.0597 109
01—A447D 19 0.0757 138
EXPERIMENT 2 - DRY VACUUM
02—A448B 6 0.0234 49
02—A449B 57 1.2596 2617
02—A450B 12 0.0468 97
02—A451D 2 0.0070 12
02—A452D 6 0.0209 36
EXPERIMENT 3 - DRY VACUUM
03—A454B 9 0.0349 73
03—A455B 7 0.0271 57
03—A456B 22 0.0853 178
03—A457D 53 0.5507 913
03—A458D 44 0.3658 606
03—A459D 50 0.3464 574
EXPERIMENT 4 - WET CLEAN
04—A460B 4 0.0154 32
04—A461B 2 0.0078 16
04—A462B 5 0.0194 41
04—A463D 39 0.1309 234
04—A464D 37 0.1242 222
04—A465D 44 0.1477 264
36
-------
Sample Number of Asbestos Concentration,
Number Asbestos Str. s/cin s/nun 2
EXPERIMENT 5 - WET CLEAN
05—A466B 2 0.0073 15
05—A467B 6 0.0220 •44
05—A468B 1 0.0037 7
05—A469D 28 0.1004 276
05—A470D 28 0.1004 276
05—A471D 11 0.0392 108
EXPERIMENT 6 - DRY VACUUM
06—A472B 6 0.0212 44
06—A473B 13 0.0465 94
06—A474B 8 0.0286 58
06—A475D 15 0.0523 90
06—A476D 15 0.0511 90
06—A477D 37 0.1235 222
EXPERIMENT 7 - DRY VACUUM
07—A478B 26 0.1008 211
07—A479B 20 0.0770 162
07—A4BOB 24 0.0924 195
07—A481D 48 0.1828 315
07—A482D 57 0.2758 491
07—A483D 51 0.3291 586
EXPERIMENT 8 - WET CLEAN
08—A484B 37 0.1399 300
08—A485B 38 0.1446 308
08—A486B 49 0.2453 519
08—A487D 53 0.3046 664
08—A488D 51 0.2703 586
08—A489D 48 0.2575 551
37
-------
Sample Number of Asbestos Concentration,
Number Asbestos Str. s/cm’ s/mm 2
EXPERIMENT 9 -WET CLEAN
09—A490B 32 0.1265 315
09—A491B 23 0.0854 211
09—A492B 41 0.1523 377
09—A493D 51 0.2211 502
09—A494D 53 0.2171 487
09—A495D 51 0.2063 468
EXPERIMENT 10 - DRY VACUUM
10—A496B 52 0.3656 895
10—A497B 52 0.2909 716
10—A498B 54 0.3375 827
1O—A499D 57 0.3871 785
1O—A500D 51 0.4891 1004
10—A5O1D 2 0.0071 15
EXPERIMENT 11 - DRY VACUUM
11—A502B 6 0.0217 55
11—A503B 9 0.0326 83
11—A504B 21 0.0752 193
11—A505D 27 0.0981 219
11—A506D 47 0.1687 381
11—A507D 25 0.0898 203
EXPERIMENT 12 - WET CLEAN
12—A508B 8 0.0288 74
12—A509B 5 0.0179 46
12—A51OB 4 0.0143 37
12—A511D 17 0.0608 123
12—A512D 23 0.0823 167
12—A513D 23 0.0823 167
38
-------
Sample Number of Asbestos Concentration,
Number Asbestos Str. s/cin S/m m 2
EXPERIMENT 13 - WET CLEAN
13—A514B 22 0.0832 178
13—A515B 16 0.0601 130
13—A516B 19 0.0719 154
13—A517D 23 0.0804 186
13—A518D 51 0.2507 586
13—A519D 49 0.2251 519
EXPERIMENT 14 - DRY VACUUM
14—A520B 42 0.1562 340
14—A521B 32 0.1190 259
14—A5228 42 0.1562 340
14—A523D 50 0.3177 530
14—A524D 50 0.3779 626
14—A525D 53 0.3368 562
EXPERIMENT 15 - DRY VACUUM
15—A526B 20 0.0715 145
15—A527B 14 0.0500 102
15—A528B 9 0.0322 65
15—A529D 41 0.1482 246
15—A530D 33 0.1200 198
15—A531D 43 0.1571 258
EXPERIMENT 16 - WET CLEAN
16—A532B 33 0.1271 267
16—A533B 8 0.0306 65
16—A534B 30 0.1156 243
16—A535D 52 0.1900 421
16—A536D 33 0.1199 267
16—A537D 43 0.1562 349
39
-------
APPENDIX C
STRUCTURE LENGTH DISTRIBUTIONS OF
AIRBORNE ASBESTOS BEFORE AND
DURING CARPET CLEANING
TABLE C-i. FIBER LENGTH DISTRIBUTION OBSERVED IN
AIR SAMPLES COLLECTED BEFORE CARPET CLEANING
Number
Particle
size range, m
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23 - 0.34
0
0
0
0
0.34 - 0.54
0
0
0
0
0.50 - 0.73
666
666
64.3
64.3
0.73 - 1.08
239
965
23.1
87.4
1.08 - 1.58
82
987
7.9
95.3
1.58 — 2.32
33
1020
3.2
98.5
2.32 - 3.41
9
1029
0.9
99.4
3.41 - 5.00
4
1033
0.4
99.8
5.00 - 7.34
1
1034
0.1
99.9
7.34 - 10.77
0
1034
0
99.9
10.77 - 15.81
1
1035
0.1
100
15.81 — 23.21
0
1035
0
100
23.21 - 34.06
0
1035
0
100
34.06 - 50.00
0
1035
0
100
50.00 - 73.40
0
1035
0
100
73.40 - 107.70
0
1035
0
100
107.70 - 158.10
0
1035
0
100
158.10 — 232.10
0
1035
0
100
232.10 - 340.60
0
1035
0
100
40
-------
TABLE C-2. FIBER LENGTH DISTRIBUTION OBSERVED IN AIR SAMPLES
COLLECTED DURING DRY VACUUMING OF CARPET CONTAMINATED WITH THE
LOW—CONCENTRATION DISPERISON
Number
Particle
size range, 1Jm
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23- 0.34
0
0
0
0
0.34- 0.54
0
0
0
0
0.50 - 0.73
238
238
63
63
0.73 — 1.08
104
342
27.5
90.5
1.08 - 1.58
24
366
6.3
96.8
1.58 - 2.32
5
371
1.3
98.1
2.32 - 3.41
4
375
1.1
99.2
3.41 - 5.00
2
377
0.5
99.7
5.00 - 7.34
1
378
0.3
100
7.34 - 10.77
0
378
0
100
10.77 - 15.81
0
378
0
100
15.81 — 23.21
0
378
0
100
23.21 - 34.06
0
378
0
100
34.06 - 50.00
0
378
0
100
50.00 - 73.40
0
378
0
100
73.40 - 107.70
0
378
0
100
107.70 — 158.10
0
378
0
100
158.10 - 232.10
0
378
0
100
232.10 - 340.60
0
378
0
100
41
-------
TABLE C-3. FIBER LENGTH DISTRIBUTION OBSERVED IN AIR SAMPLES
COLLECTED DURING WET CLEANING OF CARPET CONTAMINATED WITH
THE LOW-CONCENTRATION DISPERSION
Number
Particle
size range, .zm
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23 - 0.34
0
0
0
0
0.34- 0.54
0
0
0
0
0.50 - 0.73
238
238
60.1
60.1
0.73 - 1.08
101
339
25.5
85.6
1.08 - 1.58
47
386
11.9
97.5
1.58 - 2.32
7
393
1.8
99.2
2.32 - 3.41
1
394
0.3
99.5
3.41 - 5.00
1
395
0.3
99.7
5.00 - 7.34
1
396
0.3
100
7.34 - 10.77
0
396
0
100
10.77 - 15.81
0
396
0
100
15.81 - 23.21
0
396
0
100
23.21 - 34.06
0
396
0
100
34.06 - 50.00
0
396
0
100
50.00 - 73.40
0
396
0
100
73.40 - 107.70
0
396
0
100
107.70 - 158.10
0
396
0
100
158.10 - 232.10
0
396
0
100
232.10 - 340.60
0
396
0
100
42
-------
TABLE C-4. FIBER LENGTH DISTRIBUTION OBSERVED IN AIR SAMPLES
COLLECTED DURING DRY VACUUMING OF CARPET CONTAMINATED WITH
THE HIGH-CONCENTRATION DISPERSION
Number
Particle
size range, im
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23- 0.34
0
0
0
0
0.34 - 0.54
0
0
0
0
0.50 - 0.73
326
326
68.1
68.1
0.73 - 1.08
102
428
21.3
89.4
1.08 - 1.58
41
469
8.6
97.9
1.58 - 2.32
5
474
1.0
99.0
2.32 - 3.41
2
476
0.4
99.4
3.41 - 5.00
1
477
0.2
99.6
5.00 - 7.34
2
479
0.4
100
7.34 - 10.77
0
479
0
100
10.77 - 15.81
0
479
0
100
15.81 - 23.21
0
479
0
100
23.21 - 34.06
0
479
0
100
34.06 - 50.00
0
479
0
100
50.00 - 73.40
0
479
0
100
73.40 - 107.70
0
479
0
100
107.70 - 158.10
0
479
0
100
158.10 - 232.10
0
479
0
100
232.10 - 340.60
0
479
0
100
43
-------
TABLE C-S. FIBER LENGTH DISTRIBUTION OBSERVED IN AIR SAMPLES
COLLECTED DURING WET CLEANING OF CARPET CONTAMINATED WITH
THE HIGH-CONCENTRATION DISPERSION
Number
Particle
size range, j m
of fibers
counted
Cumulative
fiber count
Percent
of total
Cumulative
percent
0.23- 0.34
0
0
0
0
0.34- 0.54
0
0
0
0
0.50 - 0.73
319
319
68
68
0.73 - 1.08
82
401
17.5
85.5
1.08 - 1.58
44
445
9.4
94.9
1.58 - 2.32
11
456
2.3
97.2
2.32 - 3.41
6
462
1.3
98.5
3.41 - 5.00
4
466
0.9
99.4
5.00 - 7.34
2
468
0.4
99.8
7.34 - 10.77
1
469
0.2
100
10.77 - 15.81
1
469
0
100
15.81 - 23.21
0
469
0
100
23.21 - 34.06
0
469
0
100
34.06 - 50.00
0
469
0
100
50.00 - 73.40
0
469
0
100
73.40 - 107.70
0
469
0
100
107.70 - 158.10
0
469
0
100
158.10 - 232.10
0
469
0
100
232.10 - 340.60
0
469
0
100
44
-------
TECHNICAL REPORT DATA
(Please read Instructions on the rcicrsc before compicnng)
1 REPORT NO 2
3 RECIPIENTS ACCESSION NO.
4 TITLE AND SUBTITLE
Asbestos Fiber Reentrainment During Dry Vacuuming and
.
Wet Cleaning of Asbestos—Contaminated Carpet
5 REPORT DATE
7/31/89
6 PERFORMING ORGANIZATION CODE
ORGANIZATION REPORT NO
7 AUTHOR(S)
John R. Kominsky, flonald U. Freyberg
8 PERFORMING
ELEMENT NO
9 PERFORMING ORGANIZATION NAME AND ADDRESS
PEI Associates, Inc.
11499 Chester Road
Cincinnati, OH 45246
10 PROGRAM
CONTRACT/GRANT NO
11
68-03-4006
12 SPONSORING AGENCY NAME AND ADDRESS
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
1/88 - 7/89
14.SpONSORINGAGENCYC006
15 SUPPLEMENTARY NOTES —-
Project Officer: Thomas J Powers FTS: 684-7550
COMM: 569-7550
16 ABSTRACTA study was conducted to evaluate the potential for asbestos fiber reentrain-
ment during clean .ng of carpet contaminated with asbestos. Two types of carpet
cleaning equipment were evaluated at two carpet contamination levels. Airborne
asbestos concentrations were determined before and during carpet cleaning. Overall,
airborne asbestos concentrations were two to four times greater during the carpet
cleaning activity. The level of asbestos contamination and the type of cleaning
method used had no statistically significant effect on the relative increase of
airborne asbestos concentrations during carpet cleaning.
This document was submitted in fulfillment of Contract No. 68-03-4006 by
PEI Associates, Inc., for the U.S. Environmental Protection Agency’s Office of
Research and Development, Risk Reduction Engineering Laboratory. This report
covers a period of January 1988 to July 1989, and work was completed as of July
31, 1989.
17 KEY WORDS AND DOCUMENT ANALYSIS
a DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
C COSATI Field/Group
18 DISTRIBUTION STATEMENT
19 SECURITY CLASS (misReport)
21 NO OF PAGES
PRICE
20 SECURITY CLASS (This pagel
EPA Form 2220—1 (Rev. 4—77)
PREVIOUS EDITION IS OBSOLETE
------- |