Attachment A
                   AIRBORNE ASBESTOS
                     A Technical Report by

             Arthur R.  Wlckman, Principal Investigator
                Daryl W. Roberts, Project Manager
           Terry L. Hopper, Industrial Hygiene Supervisor
                  Missouri  Department of Health
              Bureau of Environmental Epidemiology
               U.S. Environmental Protection Agency
             Office of Pesticides and Toxic Substances
                 EPA  Project No.  J1007468-01-0
              EPA Grant Administrator:  David Treece
exposure of Custodial Employees                     Page 1 of 22

    The  objective of this study  was to evaluate the  occupational
exposure of custodians to airborne  asbestos  during  routine activities.
The sample population included eight custodians working in six public
buildings in Missouri.   Forty-seven personal  samples and 91  area
samples  were collected and  analyzed  by transmission electron
microscopy using direct preparation  techniques.  Personal samples
resulted  in an arithmetic mean  of  0.0009  structures/cubic  centimeter
(s/cc) and area samples resulted in  an arithmetic mean  of  0.0003 s/cc.
The highest personal sample concentration was 0.0255 s/cc, or  26% of
the OSHA action level of 0.1f/cc.
 Exposure of Custodial Employees                       Page 2 of 22


      The goal of this study was to assess the exposure of custodians
to airborne  asbestos fibers as  they  conduct routine custodial  activities
in buildings where asbestos containing materials (ACM) are present.
The  study was designed to provide quantitative exposure data in the
range of the current OSHA action level, 0.1  fiber/cubic centimeter
(f/cc).  Exposures did not  involve the intentional  disruption  of ACM, as
might occur with maintenance personnel, plumbers, electricians,  and
asbestos abatement  workers.  As studied here,  custodial exposure was
incidental and unintentional, often involving the redispersal  of
asbestos dust which had accumulated on surfaces  after previously
being released from its  original source  material.  The  study  related
exposure levels to the type of custodial activity  involved,  such as
vacuuming,  dust  mopping, and stripping  of vinyl asbestos tiles.


      STUDY  DESIGN.  Samples were collected for three consecutive
days on  each of two visits to six.selected sites.  Forty-seven personal
samples  were collected  from  the  eight custodians  involved in the study.
These samples represented full  shift  exposures, and were computed as
an eight  hour  time weighted average  (8  hr. TWA).  In addition to the
personal samples, area samples were taken when a custodian was
involved  in  a specific  task.   Six  categories of custodial activities were
monitored during area sampling:  stripping vinyl asbestos  tile (VAT),
buffing VAT, vacuuming,  dust  mopping, dry broom sweeping, and hand
dusting.  These area samples provided information  on  the  concentration
of asbestos rendered airborne by the particular  custodial  activity.  The
area samples also gave information regarding the tasks during the day
which contributed to the  8 hr. TWA exposure  recorded on the
custodian's  personal sample.   Custodians were instructed to perform
their work  routinely  but  in as much as  possible  to  schedule  their work
so that the six relevant  tasks would  be performed at some time during
the  three day sampling  periods.

      SITE  SELECTION.  A list of  potential sites  for  this  study  was
generated from the  records of the Missouri Department of Health.  The
Department maintains records of asbestos inspections for
approximately 2900  public buildings  in Missouri,  inspected since 1988
under state law 701.122, RSMo 1991.  The criteria used to determine
potential building sites  was the  presence of accessible, friable,  and

Exposure of Custodial Employees                       Page 3 of 22

extensive ACM which was damaged to some degree.  Managers of these
buildings were contacted to  determine the nature of  custodial services
in the building.  Buildings were  identified in  which  custodians worked
routinely in the presence of ACM.  Visits were made to each of the
sites, with the result that  six sites and eight custodians  meeting the
exposure criteria of the study were chosen.   These sites  are  listed

Student  Union. Missouri  Western  State College,  St. Joseph (MWSC).  This
site  had 10,000 square feet of asbestos sprayed onto the underside of
roof and metal trusses on two floors.   Custodial and maintenance areas
of the building were directly  exposed  to  this  insulation. The space
between the suspended ceiling and the metal trusses  served as the
return air plenum on both floors.   Some damaged areas were  visible in
the  sprayed on insulation.  This  material was friable  and subjected to
airstream  abrasion.  Two  janitors  worked here full time.

Dallas County Courthouse,  Buffalo (Dallas  CC).   This building had
approximately 20,000 square feet  of  ceiling  area throughout  that
contained asbestos.  About 90%  of this ceiling had  been painted with
latex paint rendering it non-friable. The remaining 10% of ceiling area
was  not painted and was  friable.  Isolated areas of ceiling were
damaged to the point  of releasing fibers.  Hallways in the building
served  as the return air plenum, which created a continuous  flow of air
over this ACM.  Floors  throughout the building were  vinyl asbestos tile.
Two half-time janitors were employed  here.

Independence  Power and  Light Generating Station, Independence (IP&L).  An
extensive amount  of TS) was present on the  steam pipes of this coal
fired electrical  power plant, some in  significantly damaged condition.
A definitive quantity of TSI  was not determined,  but  amount  was
estimated as hundreds of linear  feet.   A program of asbestos removal
and/or  abatement, by  outside  contractors occurred periodically  at  the
plant, but not during the period  of this study.  Approximately  75% of
asbestos had been removed from the  plant.   Remaining TSI is subjected
to  strong air currents and vibration from operations  of the plant.
Floors  in office  areas contained vinyl  asbestos floor  tiles.  Two
janitors  worked here -spending 'time' both' in the generating plant and in
the office  areas.

Franklin  County  Courthouse. Union (Franklin  CC).  This  building  had  2700
square  feet of hidden  spline ceiling tile throughout which contained
asbestos.  Approximately 5% of  the ceiling  was  damaged and friable.

 Exposure of Custodial Employees                        Page 4 of 22

Five hundred linear feet of hot water pipes  in the  basement were
wrapped with TSI, some of which was in damaged condition.  Two  full-
time and  two part-time custodians worked  in the building.

Western Missouri  Medical  Center.  Warrensburg  (WMMC).   This building
had  36,000  square feet of ceiling containing asbestos.  The  ceiling  had
been painted and was not friable.  Approximately 25% of ceiling area
had  been cut to  allow access for equipment, lights,  mirrors, etc,
Isolated spots in the ceiling showed  damage. In some damaged areas,
there was expected  fiber release from the many openings into the
ceiling.   Approximately  7,200  square feet of flooring  was VAT.   The
building also had approximately  150 thermal system joints  packed with
asbestos, of which 15  were damaged.  Four full-time janitors were

Ha//  of  Waters  (City  Hall), Excelsior Springs (Ex Spr).  This building
served as the  city's municipal building.  Prior to the 1960's  it had also
been used as a center for therapeutic mineral bath  treatments.   The
building had an extensive system of hot water pipes insulated with
asbestos throughout.  The total amount of TSI could not be  determined
accurately, but was estimated at several thousand  linear  feet.   Varying
degrees  of damage  existed In this material.  Air chases in certain areas
of the building were built around these insulated pipes, which  created
the  possibility  of widespread  dispersal of asbestos fibers.  One full
time custodian worked  here.

      MATERIALS   Materials were selected in accordance  with the
NIOSH  Method 7402 protocols for asbestos fiber collection.   All
samples were  collected -on Millipore 0.045  micron  mixed  cellulose
ester filters using open faced 25 mm preloaded cassettes.  Personal
sampling pumps were SKC Model 224-PCXR3 which  were  calibrated on
site  to a flow range between  2.0 and 2.5 liters  per  minute.  Area
samples were  collected using Gilian Aircon 520 high volume samplers,
Calibrated to 10  liters per minute.  Calibration  of  the air samplers was
performed using  a Gilian Gilibrator  electronic  flow  meter with  a
representative  filter  in  line.   Pumps were  calibrated at the beginning
and end of  each  sample collected.

      SAMPLE ANALY^ig   Analysis of the samples was  performed by
transmission electron microscopy  using  direct transfer methods.
Analysis was conducted  according to the non-mandatory  protocols of
the  Asbestos Hazard  Emergency Response Act (AHERA) method,  as
described in 40 CFR 763, Appendix A, Subpart E. QuanTEM  Laboratory,

Exposure of Custodiat Employees                       Page 5 of 22

Oklahoma  City, Oklahoma, served as the principal analytical laboratory.
QuanTEM's role Involved the analysis of 156 samples and the
repreparation  of 7 samples.  Included in the laboratory report were the
following factors:   structure type  (free  fibers, bundles,  clusters,  or
matrices),  species of asbestos  (chrysotile/amphibole), the count  of
fibers less  than or greater than  5  microns, and the length and diameter
of all structures  identified  as  asbestos.   Quality control  analysis was
performed  by  Professional Service Industries (PSI) at their Pittsburg,
Pennsylvania  facility.  All analysis was  performed at or below a  level
of sensitivity  of  0.005  s/cc.

the background contamination of the filters in the production  lots,  two
lot blanks were selected at random from each of the 2 production lots.
They  were submitted  for analysis without opening  the  cassettes.   One
field blank  was selected  for each day of sampling.  Field blanks were
handled in the same  fashion as field samples, including  removal of  the
cover in the field  and connection to the  pump  tubing.  This was a
control for contamination occurring in the field but not  resulting from
air sampling.   Nineteen duplicate area samples were  collected  side  by
side.  The duplicate  samples were sent to an external laboratory,  PSI,
for  analysis.   Results from the  external  laboratory were compared to
results from  the  primary analytical laboratory.  The degree of
agreement between the two was used to define the overall precision of
the sampling  and analytical techniques.   In order to  test intra-
laboratory  analyst reliability,  ^preparations were  done on seven (7)
samples  at the primary  laboratory.
Exposure of Custodial Employees                        Page 6 of 22


    A total of 138 samples were collected, of which 47 (34%) were
personal samples and 91 (66%) were area samples.  Collection volumes
averaged  1309 liters (SD +/- 302 L).  Average  analytical sensitivity
was 0.0044 s/cc (SD +/  0.0007s/cc).  Seventeen samples (12.3%)
were rejected as too dirty to count (TDTC) using  the AHERA  counting
rules for TEM analysis by direct preparation.  The remaining  121
samples were divided between 38 personal samples and 83 area
samples.   No asbestos structures were counted in 99 (81.8%) of the
samples.   Detectable asbestos structures were found in 22 (18.2%) of
samples,  with total concentrations ranging  from  0.0029 s/cc to 0.1247
s/cc.  When  a level of interest  was established  at total concentrations
of 0.005  s/cc or greater, detectable asbestos structures were  reported
in 12 (9.9%) samples.

	ARITHMETIC MEAN VALUES  Table 1 gives the arithmetic mean
values  for personal and area samples.   These are reported first for all
samples, and then for samples from each of the sites.  The mean 8 hr-
TWA for all personal samples was 0.0009 s/cc (SD W-  0.0043).   The
mean  total concentration value for all area samples was 0.0033 s/cc
(SD +/- 0.0147).  Standard deviations were calculated  based on total
concentration for area samples, and on  8 hr-TWA's for personal
Exposure of Custodial Employees                      Page 7 of 22

Table  1:   Arithmetic  means,  as structures/cubic centimeter, of personal and  area
                 Total Cane.  flHr  TWA
< B u
            >S a  Sample  Sign   SO
All Site*
Dallas CC
Exc. Spr.



Franklin CC


0.0064 0.0009

0.0472 0.0099

Personal 0.0202

Area 0.0011

Personal 0
Area 0.0009

0 0
0 0

0 0
0 0

Personal O







0 13





















6 0



a o

               1. The  category of total concentration Is composed of structures <5ji and

         2.   Standard  deviations (SO) were calculated from total  concentrations for area

                      samplesand  from 8Hr-TWA concentrations for personal  samples.
 Exposure of Custodial Employees
                     Page 8 of 22

      As  indicated in Table 1, sample results are skewed based on the
building in which sampling  occurred.  Three buildings accounted for all
reported asbestos structures.  Two  buildings, Dallas  County Courthouse
and  Excelsior Springs City  Hall,  accounted for all values reported at a
total concentration level of 0.005 s/cc or greater.  Figure  1  graphically
depicts this  skewed  distribution with  respect to  the  personal  samples.
Eight hour TWA's in Figure 1 were based on asbestos structures greater
than 5u in length.
   Figure 1: Arithmetic  mean  concentrations  of  fiber  lengtha<5|t, 5u, and  of
           BHr-TWA's for personal samples grouped by building  collection site.
            1.  Calculations of 8 Hr-TWA's  were  baaed  on  the  concentration
                  of  asbestos  structures
Exposure of Custodial Employees                        Page 9 of 22

- PERSONAL  SAMPLES  Detectable asbestos structures were found
in 6 (15.8%) of the analyzed personal samples.  Listed in Table 2 are the
personal samples with total concentrations 2 0.005  s/cc.   Eight hour
TWA's were calculated using  asbestos structures greater than 5u.,
following the  OSHA definition of  fiber length.   Sample #79, from  Dallas
County Courthouse, was the only personal sample which met the criteria
for both concentration 0>0.005sfcc) and fiber  length
Table 2: Results  of  personal  samples with  total concentrations a).005s/cc.
  ID #       Site

   79        Dallas CC0.0598

   120       Ex Spr 0.0098

   74        Dallas CC0.0495

   176       Ex Spr 0.0450
   112       Ex Spr 0.0414

0.0276        0.0874

0.0049        0.0147

0.0000        0.0495




Sample  #79.  The composition  of custodial work during this personal
sample was  as follows:   Sweeping (7%), dust  mopping (7%), dusting
(7%), miscellaneous activities (45%), and stripping VAT (34%).  The
miscellaneous category included activities such  as  wet mopping,
emptying trash, and glass cleaning.  The custodian was also involved in
sweeping and dusting in a small store  room that contained
approximately  12 square feet of a friable asbestos  ceiling.  Although
this  sweeping and  dusting were of short duration (five minutes),  the
activity was  very dusty.  Area  sampling (sample #81) in  this room
during the cleanup  did not confirm the  presence of airborne 'asbestos.
The  custodian also stripped wax from  approximately 300 square  feet of
vinyl asbestos floor tiles for a period of one hour using a rotary
stripper  and stripping solution.   Area  sampling of this operation
recorded levels  of  airborne asbestos concentrations of 0.0043 s/cc
(structure _sjze 25u).  From observations of custodial  work during the
day, the elevated, concentration..of this, personal sample  is.most  like/ly
attributable to the dusting  and sweeping  of the  store  room, despite  the
negative  results of sample #81.

Sample  # 120   The composition of custodial work during this personal
sample was  as follows:   Sweeping (11%), dust mopping (22%),
Exposure of Custodial Employees
                          Page 10 of 22

miscellaneous (60%), vacuuming (7%).  The custodian cleaned a seldom
used storage room during this  sampling  period.  Asbestos containing
materials  in  this area included significantly damaged thermal  system
insulation  located  above the storage area.  Approximately 15 minutes
was spent sweeping the area with a dry broom.  This area had a large
accumulation of dust and was seldom cleaned.  The majority of other
work on this day  involved little generation of dust.

      DISTRIBUTION  OF  CUSTODIAL  TIME.   Custodial  activities were
monitored  during each  personal sampling period.  Time spent on each
activity has  been averaged for  all personal samples.  Figure  2 gives the
distribution of time, by  percentage, spent on each of the listed job

Figure  2:  Average  percentage distribution of an 8 hour workday  on  each
          task  during  personal  sampling.
    The distribution of work time, indicated  in Figure 2 gives an
approximation of the actual job exposure of custodians on an  average day.
A job such as dry broom sweeping, "sweep," which is high in  exposure
potential, occupies only about 4% of the work day.  Thus the contribution
of dry broom sweeping to daily exposure is  low.

	AREA  SAMPLES^  Area  samples were collected in seven categories
of custodial work.   The numbers of samples distributed among these
categories  were as follows:  Sweeping (5); buffing (8); dust mopping  (32);
dusting (8); stripping vinyl asbestos floor  tile (5); vacuuming (12); and
miscellaneous (13).   Detectable asbestos  structures were found  in 16
Exposure of Custodial Employees                       Page 11 of 22

(19.3%) of the analyzed area samples.  Figure 3 gives the mean
concentrations of area  samples categorized by building  location.

Figure  3:   Arithmetic mean  <<5n, 5|i,  and  total) concentrations  of area
          samples  listed by collection site.
      The means for area samples given here  follow the trend established
for personal samples in Figure 1. That is,  Dallas County Courthouse shows
the  highest airborne concentration and the length  of fibers is primarily
 Exposure of Custodial Employees                       Page 12 of 22

    Figure * gives  the  arithmetic mean of area samples categorized by job
class.  Three means are given for each job class (<5|i,  5u,, and total
structure count).  Of these three means, the mean for  structures &5u, is
the most relevant for comparison to occupational standards.  As  seen in
previous figures,  airborne  asbestos fibers are primarily <5jj. in length.

Figure 4: Arithmetic  mean concentrations for area samples,  categorized by
          job class.  Calculations are made for  means of  structures  <5\i,
                and for total  structure counts.
 Exposure of Custodial Employees                        Page 13 of 22

    Listed in Table  3 are the 7 area samples with concentrations  at or
above 0.005 s/cc.  Because these samples are Intended  to represent
ambient concentrations, as compared to personal exposures,  TWA's have
not been calculated. - Six of the seven  of these samples were collected at
the Dallas County Courthouse.

Table 3.  Results  of  area samples with total concentrations  at or above 0.005
 ID  #


  2 7    0.0240



Total Cone.    Site       Job note
0.1247      Dallas CC  Sweeping In fumcace room with
                     In room.

0.0301      Dallas CC  Stripping VAT down to exposed

.0.0300      Dallas CC  Vacuum beneath paint encapsulated
                     ACM surfaced ceiling.

0.0288      Dallas CC  Dry buffing on VAT with poor wax
  2 5    0.0096

surfaced ceiling.
poor wax
           0.0000    0.0096      Dallas CC  Dust mop beneath paint
                                                  encapsulated ACM

           0.0096    0.0000      0.0096    Dallas CC   Dustmop on VAT with

           0.0050    0.0000      0.0050    Ex Spr     45 minutes vacuuming.

                                         minutes background.
Sample ft 78.  Area sample #78 was collected during the cleaning of a
mechanical and air  handling room.  The concrete ceiling of this  room was
identified as  having a surface coating  of  friable asbestos in  the
Department of  Health's 1988 inspection.  This surfacing  had been removed
at the time this study was conducted.  Remnants of the original surface
have been covered  with paint.  Debris  on the floor and oth'eV horizontal
surfaces  of the room appeared to contain remnants of asbestos  surfacing.
The custodian  worked in this  room for approximately 90  minutes.  Area
sample #78  was  collected concurrently with personal sample #74.
Personal  sample #74 which does not indicate a-high personal exposure,
had an 8 hr-TWA of 0.0000 s/cc.
Sample # 90.  Area sample #80 was collected during the stripping of
approximately 300 square feet of VAT  floor.   Stripping was done by
applying  a chemical stripper  solution to  the  old floor wax, followed  by
removal  using a  rotary  stripper/buffer machine.   The  stripper/buffer
 Exposure of Custodial Employees
                                                Page 14 of 22

machine was operated for one hour, resulting in the complete removal of
wax and some  direct abrasion of the VAT.  This sample was collected as
an area sample associated with the personal  sample #79, which had an
8 hr-TWA of 0.0255 s/cc.

Sample #  27.   Area sample #27 was collected during the dry buffing of
1000 square feet of VAT.  This floor was not frequently  buffed and  was
poorly  coated with wax at the  time of sampling.  The  procedure  for buffing
the floor consisted of  an initial cleaning with an oil impregnated dust
mop.   This was followed by rotary  buffing using a nylon pad.  The operation
required one  hour.  The area sample was collected in  association with
personal sample #24,  which was  rejected from analysis  as too dirty to
Exposure of Custodial Employees                       Page 15 of 22

     REJECTED  SAMPLES.  Samples that were too dirty to count
represented 12.3% of the samples collected (19. 1% of personal samples
and 8.8% of area samples).  These samples are listed in Table 4.

Table  4:   Samples too  dirty to count (TDTC) by AHERA  counting rules for TEM, by
                 probable  source of overexposure.
 ID  a
T vn e
Probable Source of Qverexooaure
Dallas CC
Dallas CC
Dallas CC
Dallas CC
Dallas CC
Ex Spr
Franklin CC
Franklin CC
Franklin CC
Franklin CC
Franklin CC
No Identified probable source of overexposure
Very dusty mechanlcal/furance room cleanup
Very dusty mechanical/furnace room cleanup
Very dusty mechanical/furnace room cleanup
Very dusty mechanical/furnace room cleanup
Very dusty store room, with TS1 on pipes above area
Very dusty basement hallway cleanup
Very dusty basement hallway cleanup
Includes dusting Venetian blinds
Includes non-routine sweeping back stairs and dusting
Vicinity of old ACM ceiling tile, seldom cleaned
Extremely dusty, fly ash at coal power plant
Extremely dusty, fly ash at coal power plant
Dusting rails in area of coal lired boiler
Mostly dustmopplng, ongoing construction on boilers
Includes dustmopping In construction area
replace VAT and dustmop In construction area
measurements of 115 structures  identified as asbestos were made.
Arithmetic mean value for  length of all structures was 2.45u, (SD W-1.94
s/cc).  The range of length data was 0,5n to  9.5u.  Nine structures were
measured  at 5u, length.  Arithmetic mean width was 0.14|i (SD +/- 0.18
s/cc).  The range of width data was 0.1u to 1.75u.  Lengths and widths
were compared for asbestos structures collected at Dallas County
Courthouse and  Excelsior Springs City Hall,  the two primary sites at
which recordable asbestos structures were collected.   No significant
differences were found  in  length or width of structures coming from  these
two sites  (p=0.05).  Asbestos  structures were proportionately  divided in
the following  categories: Bundles, 7%; fibers, 15%; matrices, 78%.

      QUALITY CONTROL SAMPLES.  Nineteen samples were collected
side-by-side for  duplicate analysis  at PSI,   No significant difference  was
found between the  original samples and the  duplicates (p-0.05).
 Repreparations were done  on seven samples.  No  significant difference
was found between the original sample and the reprepared sample
(p-0,05).  One asbestos structure was found on a field blank, which gave a
total concentration  of 0.0045 s/cc.   No explanation for this  occurrence
 Exposure of Custodial Employees
                                              Page 16 of 22

was made.  No other asbestos structures were found on field blanks or
production lot  blanks.


    As an occupational group, custodial exposure to asbestos has not
received  the same attention which has been focused on persons whose
work directly disturbs ACM  in buildings.  This  paucity of research data
has recently been noted in  the comprehensive review of asbestos  exposure
research undertaken  by the Health  Effects Institute - Asbestos  Research
(HEI-AR.1991).   From  the  limited data currently available,  estimates  of
the occupational exposure of custodians to asbestos  vary greatly.   Sawyer
(1977) measured levels of airborne asbestos resulting from custodial
dusting of  book shelves beneath a friable   20%  chrysotile ceiling in a Yale
University  building.   Concentration  levels, as counted by polarized  light
microscopy, ranged from  1.6 fibers/cubic  centimeter  (f/cc) lor dry
sweeping to 4.0 f/cc for dry  dusting.  These concentrations would indicate
a  significant health  hazard" if they  were representative of normal
custodial exposures.  By contrast, one  of the  mean values of ambient
concentration  recorded in EPA's Public Buildings Study (1988)  was
0.00073  structures/cubic centimeter (s/cc).  The samples  that
contributed to  this mean were collected in areas of public buildings with
significantly damaged ACM.   Samples were subsequently analyzed by
transmission electron microscopy (TEM).   If predictions of custodial
exposure were- made based on the deposition and resuspension of asbestos
dust,  as measured in the Public  Buildings Study, it would be.reasonable  to
estimate that  exposures would be quite low,  well below the OSHA action

    Concern has been expressed that  routine custodial exposure may be a
significant health hazard (HEI-AR,  1991;   Indoor Air Review,  1991;
Bricher,  1990).  Catherine  Oliver's  review  (1991) of  the pulmonary
radiographs and lung functions of custodians has focused attention on this
issue.  In Oliver's study, pulmonary radiographs and spirometric
measurements were  reviewed from  a cohort  of custodial  employees who
had been  occupation ally exposed to asbestos in the Boston Public Schools.
The prevalence  of pleural plaques  and pulmonary restrictions,  in excess  of
background rates,..was. attributed to custodial occupational exposure, to
asbestos.  Pleural plaques are  considered diagnostic traits of  asbestos
exposure.  Advocates of custodial employees have called for more
stringent controls of ACM  in buildings to ensure occupational  health
 Exposure of Custodial Employees                        Page 17 of 22

    The OSHA asbestos standard was used  in this study for comparison to
the obtained analytical results, with the acknowledgment that OSHA uses
phase contrast microscopy (PCM) for analysis rather than TEM.  The
OSHA/PCM  standard was used in the absence of a comparable TEM
standard.  In  this study the arithmetic  mean for  all personal samples with
fibers longer than 5 microns, is  0.0009  s/cc, 8 hour TWA.  This mean is
below the OSHA action level of  0.1 f/cc (95% Upper Confidence Level of
0.029 s/cc).  In the case  of the highest exposure building, Dallas County
Courthouse, the building mean for personal exposure was 0.0099  s/cc,
from Table 1.   The 95% Upper Confidence Level for this mean is 0.032
s/cc, or  32% of the action  level.  The personal sample with the highest
measured concentration, also from Dallas County Courthouse,  was 0.0255
s/cc, Table  2, or 26% of the OSHA action level.

     Mean values for all area samples  was 0.0033  s/cc, with the mean  for
the area samples from Dallas County Courthouse at 0.0176 s/cc.   These
means  included fibers of  all lengths.  Among individual area sample
results, the highest concentration, 0.1247 s/cc, came from a  sample
collected during the dry broom sweeping of a furnace room at  the Dallas
County Courthouse.  The  ceiling of this room was  originally coated with a
friable asbestos  surfacing material.  This surfacing  had been  scraped off
within  the last five  years  and replaced with decorative paint.  Debris and
dust in the  room were evident and cleanup of the room was rarely done,
occurring less  than  once  per year.  Rve additional  area samples at Dallas
County Courthouse also  registered concentrations  exceeding 0.005 s/cc.
The custodial  activities conducted during each of these  sampling  periods
are described in  the Job Notes, Table 3.

     Included  in the miscellaneous category of Figure 5 are two  instances
of small scale replacement of VAT flooring  and  the replacement  of three
damaged suspended ceiling tiles beneath a sprayed-on asbestos
fireproofing.  The  miscellaneous category also included a high proportion
.of activities, such  as wet.mopping  and moving equipment, when  custodial
exposure could not reasonably be expected to be higher than that of
general building occupants.
 Exposure of Custodial Employees                       Page 18 of 22

     REJECTED  SAMPLES.  The use of direct preparation TEM analysis
presented  difficulties  in the assessment of custodial exposure to
asbestos.  In the custodial occupational environment, small concentrations
of asbestos fibers were suspended in higher background concentrations of
non-asbestos building  dusts.  In order to detect low concentrations of
asbestos efficiently,  it was  necessary to draw a minimum volume
(approximately  1000L) of air through  the filter.  In dusty environments,
this volume of air caused .the excessive buildup of  dust on the filter.   When
25% of the filter's grid area was  obscured, the concentration of asbestos
was considered too dirty to  count  (TOTC) by AHERA counting rules.

    This problem  was apparent in the attempts made in  this study to
collect peak custodial exposures.  An  example  of this was the cleanup of
the mechanical room at Dallas  County Courthouse.  The  cleanup of this
room, which  reportedly had  not been cleaned in the last three years, was
not part  of the custodian's routine duty. However, in an effort  to define a
"worst case scenario" for custodial  exposure, the custodian was requested
to clean  this room during trie study.  A total of 7 samples (4 area, 3
personal) were collected.  Of these 7 samples,  2 personal samples and 2
area samples were rejected as too dirty to  count.  The highest
concentration for the samples  which could  be  counted  was 0.0874 s/cc,
8  hr-TWA.

    Other samples that were TDTC are described in Table 4.  From the
right column of Table 4, it can  be seen that overexposed samples
generally occurred in association  with the  clean up of  areas in buildings
described as having  a large amount of background  dust.  With the exception
of Independence Power and Light, these areas of the buildings  were
excessively dusty  because they were seldom, if ever, cleaned.  These
exposures  have been  considered  episodic,  or non-routine, rather than
routine.  Independence Power and Light was an exception to this, because
excessive dust in  the  plant  was due to  the operation of the coal fired
boilers, rather  than to long periods between custodial  services.  This is a
type of  exposure which is  routine"," but 'applicable to only lirtiifed industrial
 Exposure of Custodial Employees                       Page 19 of 22

      LIMITATIONS  OF  DATA  in discussing the results of this  study, a
distinction should  be made between the two classes of samples collected.
Personal samples, which are taken in  the  breathing zone of the custodian
for an 8 hour work  shift, represented  actual occupational exposure.  Area
samples represented a potential  exposure  to building occupants who are in
the vicinity  of custodial activity for a time duration of  approximately  two
hours.  Area samples do not represent occupational exposure and can  not
be related to  occupational  standards.  They did however, give an indication
of the concentrations of airborne asbestos fibers generated during
custodial  work.

    The present data can  be used to  assessed  custodial  exposure during
tasks which were routinely performed on  a daily,  weekly, or monthly
cycle.  The  data can not be used to make  a valid assessment of tasks which
fall  under the general duties of custodians, but which are done on a non-
routine  basis.  Routine exposures to custodians were reported to fall  at or
below the values obtained  at  Dallas County Courthouse, 0.0099 s/cc.  The
evaluation of  custodial exposure levels for non-routine tasks  can not be
made from  this data.

      STUDY BIAS.  A  bias was established in this study which
emphasized the occupational exposure of custodians to  ACM.  Work
regimes were scheduled so that air sampling would predominately include
activities which had  the potential to  dislodge or resuspend  asbestos
fibers.  The buildings themselves were  selected with a bias  toward those
buildings with damaged, friable ACM.  The study did not attempt to define
the "average" exposure of custodians.  The rationale for  the study's bias
was to focus attention  on those routine  custodial activities  which
directly involved ACM and to  document the resultant level of custodial


    This  study determined that custodians who performed routine
activities  in  buildings which  contained  friable, asbestos materials were
not exposed to  levels of airborne asbestos which approached the OSHA
action level of 0.1 f/cc.  The  arithmetic mean value for 38 personal
samples, analyzed by TCM,- was-0.0009 s/cc, 8 hr-TWA  for structure
lengths 5u,.  The study attempted to  bias sampling in order  to maximize
the number of occurrences in which custodians worked with  asbestos
containing  materials.  Even with this  bias, resulting exposure levels  were
well below the OSHA action  level.
Exposure of Custodial Employees                       Page 20 of 22

    A proportion of  samples (12.3%) were rejected  from  analysis as too
dirt/ to  count.   These  samples predominantly reflect  collections made
under non-routine conditions.  Hypothetically,  the rejected samples
contain  higher asbestos content than the analyzed samples.  The
conclusion remains  firm, however, that during routine activities  the
exposure of custodians is very low and does not pose a significant risk for
the development of  asbestos related diseases.
 Exposure of Custodial Employees                       Page 21 of 22


Bricher,  Julie Larson.  1990. The Third Wave Risk.  Asbestos Issues.
     August, pp. 8-17.

EPA. 1988.  Assessing Asbestos Exposure in Public Buildings.  EPA 560/5-

Health  Effects Institute-Asbestos Research (HEI-AR).  1991.  Asbestos in
     Public and Commercial Buildings:  A literature review and synthesis
        of current knowledge.  HEI-AR, Cambridge,  Mass.

Indoor Air  Review* 1991.  Victims Group Says  Asbestos  Report 'Short-
     Changes' Public.  Vol. 1, #8, p.29.

Oliver, C. L;  Sprinca,  N. L'u.and  Greene, R. 1991. Asbestos-Related
	Disease in Public School Custodians.  American Journal of
            industrial  Medicine.  Vol 19, pp. 303-316.

Sawyer, R. 1977.  Asbestos Exposure in a Yale  Building: Analysis and
     Resolution.  Environmental  Research.  Vol 13:1, pp. 146-168.
 Exposure of Custodial Employees                       Page 22 of 22