United States
 Environmental Protection
 Agency
Risk Reduction
Engineering Laboratory
Cincinnati OH 45268
 Research and Development
 EPA/600/S2-89/041  Feb. 1990
Project Summary
Evaluation  of Airborne Asbestos
Concentrations Before and
During  an  O&M  Activity:
A  Case Study
John Kominsky and Ronald Freyburg
  The current  lack of information
regarding the effect of Operation and
Maintenance (O&M) activities on the
potential  for asbestos exposure to
building staff and occupants prompt-
ed this study. This report presents a
statistical  evaluation of airborne
asbestos  data collected before and
during an  O&M  activity  involving
removal of thermal surface insulation
from a feedwater preheated tank in a
boiler room. Transmission electron
microscopy (TEM)  analysis on  0<4*ii-"
pore-size polycarbonate (PC) is
compared with TEM analysis on  0.8-p-
pore-size mixed  cellulose  ester
(MCE) membrane filters.
  This Project Summary was devel-
oped by EPA's  Risk Reduction  Engi-
neering Laboratory, Cincinnati, OH, to
announce key findings of the research
project that Is fully documented In a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
  The U.S.  Environmental Protection
Agency (EPA) has issued guidance docu-
ments to assist building owners in devel-
oping an O&M  program  to  control
exposure to  asbestos  when asbestos-
containing material (ACM) is present in
the building. ACM includes surface mate-
rial,  thermal  surface insulation, or other
materials such as ceiling and floor tiles.
In theory,  an O&M program should be
designed and implemented  to protect
building staff and occupants from the
release of asbestos  fibers and to warn
them of potential hazards created by the
disturbance of ACM under uncontrolled
conditions. The  O&M  program should
continue as long as the ACM remains in
the building.
  The overall goal of an  asbestos O&M
program is to maintain the building
environment free of asbestos  contamina-
tion. The specific program objectives are
(1) to remove asbestos fibers that  may
have been released from the ACM, and
(2) to minimize future release and distri-
bution of fibers by controlling activities
that might disturb the ACM.
  The O&M  program  focuses on the
activities of custodial and maintenance
workers and service contractors. Special
procedures for routine cleaning by custo-
dial  workers are designed to achieve the
first program objective, the collection of
previously released  asbestos fibers.
Special work practices and procedures
are designed for maintenance workers to
achieve the second  program  objective,
minimization of ACM disturbance.  The
work practices and procedures are tailor-
ed to three types of projects: (1) those
that are unlikely to involve any contact
with ACM; (2) those that may cause
accidental  disturbance of ACM; and (3)
those that involve small-scale manipu-
lation or removal of ACM.  The  O&M
program also  specifies  emergency re-
sponse actions for asbestos-fiber-release
episodes.

-------
Objectives
  The following were the primary objec-
tives of the study:
•  To document a one-site case study on
   the effectiveness  of  work practices
   and procedures in  preventing  the
   release of asbestos fibers outside the
   work  area  during  an O&M activity
   involving  removal  of thermal ACM
   surface insulation.
•  To determine if 0.4-ii-pore-size poly-
   carbonate and 0.8-n-pore-size mixed
   cellulose ester membrane filters pro-
   duce  equivalent estimates of  airborne
   asbestos concentrations.


Study and Design Methods

Site Description
  The selected site, a boiler room of a
commercial research laboratory, met the
following criteria.
1,   No abatement of ACM had occurred
    inside  the  building  site within  the
    preceding 6 months.
2.   The  asbestos was being abated  as
    part of a  scheduled  maintenance
    operation.
3.   The  building owner agreed  to  co-
    operate with the EPA and to allow air
    to be  monitored during the  main-
    tenance activity.
  The O&M activity involved removing of
approximately  57 ft2 of  thermal surface
insulation  from  a 6-ft-high feedwater
preheated  vessel with  a  2.5 ft base. The
feedwater  vessel was  to be replaced as
part of scheduled maintenance of boiler
room equipment. The vessel was located
on the operating platform on the second
deck of the boiler room,  which measured
75  ft  x  150 ft x 20  ft. An abatement
contractor was  removing the  ACM
removal.
O&M Activity
  The  glove-bag technique  was  used
during  removal of the thermal surface
insulation. This containment method  in-
volved  enclosing the vessel  with one
layer of 6-mil polyethylene plastic sheet-
ing and splicing several sleeve and glove
assemblies (standard glove bags) into the
enclosure. The containment  did  not
include  a negative-pressure  air filtration
system.
  Two  workers wearing  full  protective
clothing  and  powered  air-purifying
respirators removed the  surface insula-
tion. All work was performed from outside
the enclosure as the workers  insert their
hands  and  arms through the built-in
sleeves of the glove bags. A nozzle to
wet the insulation with an  amended water
solution  was introduced  through  an
auxiliary  sleeve.  During removal, asbes-
tos-containing debris fell to the bottom of
the bag. Some debris  escaped through
seams  in the bag and fell approximately
7 ft to  the uncovered  floor of the boiler
room. Analysis of this  material showed
that it  contained approximately 42 per-
cent  chrysotile, 33 percent amosite, and
10 percent crocidolite.
  Subsequent to the  removal of the
insulation, all  substrate  surfaces  were
repeatedly wire-brushed  and  wet-wiped
to removal  all residual  material.  The
asbestos-containing  debris  and  the
plastic  glove  bags were  then  placed in
double 6-mil polyethylene bags and dis-
posed of at an approved sanitary landfill.

Sampling Strategy
  Table 1 presents  an  air-sampling ma-
trix. Area air  samples  were  collected
before and during removal of the thermal
surface insulation from  the  feedwater
vessel.  Both the before-  and during-
removal  samples were collected in the
perimeter area of the  feedwater vessel.
                         Outdoor area air samples were also
                         collected before removal.
                           Two side-by-side air samples were
                         lected at each sampling location, or
                         a 25-mm,  0.4-n-pore'size PC filter
                         the other one a 25-mm, 0.8-ii-por<
                         MCE filter. When this study was de
                         ed and implemented, EPA recomme
                         using either of these membrane filt<
                         collect airborne asbestos fibers. Ci
                         Asbestos  Hazardous  Emergency
                         sponse Act (AHERA) guidance spe
                         the same filter types but a different
                         size (0.45-n) for the MCE filter.


                         Sampling Methods
                           Two side-by-side area  air  sar
                         were collected at each sampling loc
                         Each pair of samples consisted of
                         mm,  0.4-n-pore-size  PC filter and
                         mm, 0.8-n-pore-size MCE filter. Eac
                         mm filter was  mounted  on a 5-ji
                         size,  MCE  backup diffusing  filte
                         cellulose support pad and was con
                         in a three-piece cassette with a 5
                         conductive cowl and face cap. The
                         and cowl  sections  of the cassettes
                         sealed with vinyl  adhesive  tai
                         prevent air filtration through the sej
                         the cassettes during sampling. Thi
                         cassettes were  4 to 5 ft above th<
                         and arranged  in  a horizontal  lii
                         clipping them to a  sturdy stand. Th
                         cassettes were approximately 5 err
                         and oriented in  the same direction
                         filter  face,  angled slightly dowr
                         During sampling,  the face  caf
                         removed to expose the full face
                         filter to the air stream.
                           The filter assembly was attachec
                         electric-powered vacuum pump. Ai
                         calibrated precision rotameter was
                         to regulate  the airflow  through th
                         assembly at 9.3 to  11.3 L/min.
                           The air  samples were generall
                         lected for 6 to 7 hr to achieve a mi
                        Table 1.    Air Sampling Matrix
                                                         Location and Number of Samples
                                                   Perimeter
                          Outdoors
                            Samples Collected
        PC'
MCE"
PC
MCE
                         Total samples
         19
                          Field Blank
PC
MCE
Before asbestos removal
During asbestos removal
10
9
10
flc
5
0
5
0
1
1
1
1
                        "Polycarbonate filter.
                        bMixed cellulose ester filter.
                        CA ninth sample could not be analyzed.

-------
ir volume of 3200 L for each sample. At
le end of  the  sampling period, filters
«re turned upright before being disconn-
ected from the  vacuum pump and were
 ored in this position.

 nalytical Methods
  The MCE membrane filters and the PC
lembrane filters were analyzed by TEM.
'he TEM analytical protocols  used for
le PC  membrane filters are those found
i WERL  SOP 87-1, Revision.  MCE
lembrane filters were prepared accord-
ig to protocols found in the  Asbestos
 azard  Emergency Response Act in
 inal Rule;  counting  rules  and  other
nalytical protocols were consistent with
 lose used for the PC membrane filters.
  ITT Research Institute performed the
 EM Level  II  analyses on  the  field
•amples under  separate contract  with
 PA's Risk Reduction Engineering Labo-
 tory (RREL) in  Cincinnati, Ohio.

 duality Assurance
  The Quality Assurance Project  Plan
 1APP) contains the complete  details of
 e quality assurance procedures follow-
 1 during this  research project. These
 rocedures  are  summarized in  the
  (lowing subsections.

  ample Chain-of-Custody
  Sample chain-of-custody procedures,
 n integral  part of  both sampling  and
 nalytical  activities, and were  imple-
  ented for  all  air and bulk  samples
 ollected. The applied  field custody pro-
 edures documented each sample  from
  e time of  its collection until its receipt
 y the analytical laboratory. Internal  labo-
atory  records  then documented  the
custody of  the  sample through its final
disposition.
  Standard  sample custody  (traceability)
procedures were used during this project.
Each sample was labeled with a unique
sroject  identification number, which  was
•ecorded in the  field log  book along  with
jther information specified by the QAPP.
Quality Assurance Sample
Analyses
  Specific quality assurance procedures
for ensuring the accuracy and precision
of the  TEM analyses  of  air  samples
included the use of lot  and field  blanks
and replicate and duplicate analyses.

Lot Blanks —
  Lot blanks are filters chosen before the
start of field  work.  These blanks  are
analyzed by the analytical  laboratory to
check  for filter contamination.  Two
laboratories  analyzed  5 percent  of  the
total  number of PC filters and  5 percent
of the total number of MCE filters used in
the 1987 field studies by TEM  Level II in
accordance with the  Yamate procedure.
The PC filters were all from the same lot.
The filters were considered "acceptable"
for use  if the average asbestos structure
count per 10 grid openings was less than
3. If the average asbestos structure count
for the group exceeded 3 asbestos struc-
tures  per 10 grid openings, the entire lot
of filters was  considered  contaminated.
The  TEM analysis of  the  PC  filter lot
blanks showed background  filter contam-
ination of 1.8 asbestos structures  per 10
grid openings (or 180 asbestos structures
in 1000 grid  squares examined). The
TEM analysis of the MCE filter lot blanks
showed background  filter contamination
of 0.12 asbestos structures per 10 grid
squares (or 12  asbestos structures in
1000  grid openings  examined).  There-
fore,  the  analysis of  the lot  blanks
showed that  the background  asbestos
filter contamination  was  within specific
limits.

Field Blanks —
  Field  blanks  are filters taken into the
field and handled in the  same manner as
exposed air sample  filters  to  check for
contamination  that might not be a result
of air sampling. During  setup  of  the air
sampling  pump, preloaded filter  cas-
settes were  selected as  field  blanks.
These filters were labeled and handled in
a manner similar to that  used for the
sample filters, but they were not attached
to the sampling pump. Field blanks were
collected and TEM-analyzed for both PC
and MCE filters (Table 2). The analysis  of
the  field blanks showed that asbestos
filter contamination was within the  guide-
line  of an  average  of  3  asbestos
structures per 10 grid openings.

Replicate  and Duplicate
Analysis —
  Replicate  sample analysis  provides a
means of assessing analyst precision and
refers to a second analysis of the same
grid  preparation by  the same analyst.
Three samples collected on PC and three
samples collected  on  MCE  membrane
filters were  randomly selected for repli-
cate analysis.
  Duplicate  sample analysis  provides a
means of assessing  analytical variability
introduced by the  filter preparation pro-
cedure  and  refers to the  analysis of a
second  grid  preparation from  the original
filter. Three samples collected  on PC
filters and three samples on  MCE mem-
brane filters were randomly selected for
duplicate analysis.

Statistical Analyses Methods
  The data  were grouped according  to
time of  sampling (before and   during
removal activities),  location  of sample
(outside the work area and  outdoors), and
filter type (MCE and  PC). The data were
then log-transferred (base  e)  and  tested
for normality by use of the Shapiro-Wilk
procedure to determine an  appropriate
statistical analytic approach for the com-
parisons.  The  transformed  data sug-
gested   reasonable  normality;   hence,
parametric  statistical procedures were
chosen for further analysis of the data.
  Samples with a structure count of zero
were assigned  an estimated airborne
asbestos concentration  of zero structures
per cubic centimeter (s/cm3). A concen-
tration  of zero  s/cm3  was used  in  all
summary statistic  calculations and plots.
                       Table 2.    TEM Level II Analyses of Field Blanks
                                                                  Asbestos Structures
Type of Filter
Polycarbonate
Cellulose ester
Number of
Field
Blanks
2
2
Total
Number
4
1
Average Number
per 10 Grid
Openings
2
0.5
Range per 10
Grid Openings
0-4
0-1
                        Blank Guideline
                             3.0

-------
Because the  data  were log-transformed
for  comparisons,  zero s/cm3 concen-
trations  were  replaced  by the sample's
analytical sensitivity  before the  data
transformation. (The analytical sensitivity
for TEM, also  referred to as the detection
limit, is  the estimated airborne asbestos
structure concentration calculated when a
single structure is counted in  a sample).
This approach is recommended  by the
AHERA  Final  Rule for  the clearance z-
test.

Results and Discussion

Airborne Asbestos
Concentration Data
  Both   0.8-ii  MCE  and   0.4-p PC
membrane filters were  used  to measure
the airborne  asbestos concentrations
before and during removal of the thermal
surface  insulation.  Summary  statistics
(arithmetic mean, standard deviation, and
                  coefficient of variation) for the  asbestos
                  concentrations determined  by  TEM
                  analysis of the cellulose  ester and  PC
                  filters are presented in Tables  3  and 4,
                  respectively. The  results  are presented
                  by sampling phase (before and  during
                  removal) and  sampling  location (per-
                  imeter area and outdoors). Figures 1, 2,
                  and 3 present  the mean  asbestos con-
                  centrations for  MCE and PC filter types
                  jointly.
                  Comparison of Airborne
                  Asbestos Concentrations on the
                  Two Filter Types
                    The paired  sample t-test  revealed no
                  statistically significant difference in mean
                  airborne  asbestos  concentrations  be-
                  tween the two filter types (t =  1.10,  p  =
                  0.2842).  The asbestos  concentrations
                  measured on 0.8-ii-pore-sized MCE filters
                  are  plotted against  the corresponding
                  measurements made on 4-ii-pore-size PC
                                         filters (Figure 4).  Neither filter typ<
                                         sistently demonstrated a  tenden
                                         show higher or lower airborne ast
                                         concentrations.
                                           Because of  this  lack  of statis
                                         significant different asbestos  co
                                         trations between the two  filter type
                                         because  PC filter  blanks indica
                                         marginal (but  not  significant)  co
                                         (nation  problem, data from the
                                         membrane filters  were  used fc
                                         remaining comparison.


                                         Comparison of Airborne
                                         Asbestos Concentrations in
                                         Perimeter Area Before and
                                         During O&M Activity
                                           The mean airborne asbestos o
                                         tration (0.0600  s/cm3)  in the  per
                                         area  during ACM removal  was c
                                         than  the mean  concentration ((
                                         s/cm3) before  the O&M  activity
                                         samples collected on MCE  filters
 Table 3.   Summary Statistics of TEM Analyses on Mixed Cellulose Ester Filters

                                        Airborne Asbestos Concentration, s/cm3
      Location
Sample Size
Mean
Standard
Deviation
Coefficient  of
   Variation
                               Before Removal
 Outdoors
 Perimeter
     5
    10
0.0004
0.0246
 0.0004

 0.0194
   0.9170

   0.7868
                               During Removal
 Outdoors
 Perimeter
     0"
     8
                                        0.0600
                             0.0295
                         0.4919
 "No samples were collected.
 Table 4.   Summary Statistics of TEM Analyses on Polycarbonate Filters

                                        Airborne Asbestos Concentration, s/cm3
Location
Sample Size Mean
Standard
Deviation
Coefficient of
Variation
Before Removal
Outdoors
Perimeter
5 0.0009
10 0.0329
0.0008
0.0237
0.8810
0.7226
                                During Removal
  Outdoors
  Perimeter
     Qa

     9
0.0606
 0.0467
    0.7710
 *No samples were collected.

-------
      Mean
   Asbestos
 Concentration
    (•/cm*)
              Perimeter
Outdoors
Figure 1.   Mean airborne asbestos concentrations on cellulose ester filters before and during removal of thermal
          surface insulation.
      Mean
    Asbestos
 Concentration
     (s/cm»)
                                                                               During
              Perimeter
Outdoors
Figure 2.   Mean airborne asbestos concentrations on polycarbonate filters before and during removal of thermal
          surface insulation.

-------
     Mean
   Asbestos
Concentration
    (a/cm*)
                                                                                     ilycarbonato
               Before
During
Outdoors
Figure 3.   Mean airborne asbestos concentrations on two filter types before and during removal of thermal surface
           insulation.
  0.16
         Polycarbonate Concentration. s/c«
                                                            (Thl« HIM r«pra*«nt« • 1:1
                                                            corr»«pond«fte* b*t«««n I liter*)
                    0.02
   0.04        0.06        0.08
   Caltaloa* Ecltr Concentration. a/CM9
                          0.12
 Figure 4.  Relationship between airborne asbestos concentrations measured on 0.8-n-pore-size mixed cellulose ester
           and 0.4-ii-pore-sizo polycarbonate membrane filters.

-------
 ). Student's t-test showed this increase
o be statistically significant (t  = 2.54, p
=  0.0216).  The  increase in  airborne
isbestos concentration was most  likely
tttributable to the  asbestos-containing
jebris that escaped the plastic enclosure
md fell to the boiler room floor.
Conclusions
  The principal conclusions reached dur-
ng this study of O&M procedures are:
 .  Inadequate sealing  of  a glove-bag
   type enclosure resulted in  a statis-
   tically significant increase in airborne
   asbestos concentrations in the perim-
   eter area during removal of  the ther-
   mal surface insulation.
I.  The  TEM analysis of paired O.S-ii-
   pore-size  MCE and  0.4-ii-pore-size
    PC filters produced statistically equiv-
    alent estimates of  airborne asbestos
    concentrations.  (At the time  of  this
    study,  EPA  sampling  guidance
    recommended the  use of 0.8-ii-pore-
    size cellulose ester or 0.4-p-pore-size
    PC filters to collect airborne asbestos
    fibers, whereas AHERA specifies the
    same  filter  type but  a pore  size of
    0.45-n for cellulose  ester filters.)

Recommendations
  This  research study  illustrates  that
O&M  activities  involving  the  removal of
thermal  system ACM  may result in
elevated asbestos fiber concentrations in
adjacent areas  of the building. Research
is  needed to develop or identify effective
minicontainment  systems for controlling
the release of asbestos fibers into adja-
cent building areas.
  The  full report was submitted in  ful-
fillment of Contract Number  68-03-4006
by PEI Associates, Inc., under the spon-
sorship of the U.S. Environmental Protec-
tion Agency.

-------
John Kominsky and Ronald Freyburg are with PEI Associates, Inc., Cincinnati, OH
  45246.
Thomas J. Powers is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Airbornen Asbestos Concentrations
  Before and During an O&M Activity:  A  Case Study","  (Order No. PB 89-224
  463/AS; Cost: $13.95, subject to change) will be available only from:
        National Technical Information Service
        5285 Port Royal Road
        Springfield, VA 22161
        Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
        Risk Reduction Engineering Laboratory
        U.S. Environmental Protection Agency
        Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
      BULK RATE
POSTAGE & FEES F
         EPA
   PERMIT No. G-3!
Official Business
Penalty for Private Use $300

EPA/600/S2-89/041
          000085836   TOXICSPEST*"  "
          USEPft   REGION  V  LIBRARY
          2JO  3  DSAR8QHS  ST
          Rn 1670
          CHICAGO               IL  60604

-------