Asbestos Release during Building Demolition Activities
IT Corp., Cincinnati, OH

Prepared for:
Environmental Protection Agency,  Cincinnati, OH
1993

-------
                                   TECHNICAL REPORT DATA
                                                                   PB93-199537
4. TITLE *NO 3UITITk.£

   Asbestos Release  During  Building Demolition
     Activities
                                                           A, KirOMT OAT I
              t, Pf MFOMMINO OHOANHATION COOC
7. AUTHOfttSI     1                 ">              \
R.  C. WJlmoth1,  K.  A.  Bracketr,  P.  0.  Clarr,  B. A.
Hollett*, T.  J.  Powers1,  and  J.  R. Millette
                                                           kP£A»OAMIMG
                                                                                  MCPOAT MO
». HHfOHMINO ORGANIZATION NAME AND AOOftCSS
 EPA, RR£L, WHHTRO,  Toxics  Control  Branch,  Cincinnati, OH
45268;   IT  Corporation,  11499 Chester Road,  Cincinnati,
OH   45246;  and Gillette Vander  Wood and Associates,
5500 Oakbrook Parkway, Norcross,  GA  30093
                                                           1C. rROSMAM EltMENTNO.
                , CDNTMACT/CnANT NO.
                    68-C9-0036
 13. SPOMSOMIHC AGENCY MAMl ANO ADOKESS
  Risk Reduction Engineering  Laboratory—Cincinnati, Ohio
  Office of Research  and Development
  U.S. Environmental  Protection Agency
  Cincinnati, Ohio  45268
              11 TYPf Of MEPOflT AMO PEMIOO
                   Journal  Article
              U. S^OMSOMINC ACEMCV COOt
                  EPA/600/14
 is,su»»i.eM«MTA»rNOTES   Project Officer = l"atricic  Clark
  Choices - Technical Supplement", Volume 1(2}:6-11
                b69-/bol;  "tnvironmcntal
1C. ABSTRACT
      The Risk Reduction Engineering Laboratory has monitored  block-wide building
 demolition  and debris disposal activities at Santa Cruz and Uatsonsville, California
 following the earthquake,  an implosion demolition of a 26-story building  in
 Cincinnati,  Ohio,  and the  demolition of two school buildings  in Fairbanks, Alaska to
 evaluate if the demolition activities and their associated dust control practices
 were able to prevent downwind elevations of asbestos concentrations.

      The analyses  of the air samples were performed using the Transmission Electron
 Microscope  for ambient air levels.   Personnel monitoring at the Santa Cruz landfill
 and  the  Fairbanks  landfill  during disposal  activities were analyzed by both TEH  and
 phase contrast microscopy.

      The paper documents the conditions of the buildings,  demolition practices,  dust
 control  practices, and up'  and downwind asbestos concentrations during demolition.
                               KEY WORDS AND DCCWMf MT ANALVIIS
                 DESCRIPTORS
                                                                        c. COSATI
 Asbestos
 Exposure
  Demolition
  Fibers
«. DISTHIIUTION STATEMENT
RELEASE TO PUBLIC
It. ItCUplTY
  UNCLASSIFIED
31. NO. O» ?AG<9
     8
                                                                        aa.'nice
  UNCLASSIFIED

-------
                   Asbestos  Release
                     During  Building
             Demolition Activities
                        Roger C. WamothA, Kim A. Brackclt8, Patrick J. Clark*,
                      Brace A. Holl«ttA, Thomas J. Powers*, and James R. Millttt«c
lite U.S. Environmental Protection
Agency's (IPA) Risk Reduction Engi-
neering ynrotory (RREl) nmdored
block-wide bundng iemofton and
debris disposal activities at Santa Cna
and Watsoflsvie, California following
the 1989 earthquake; an najosm
detaoktion of a 26-story WAng n
unamwn, Onia; the deoNwhoii of eight
wooden barracks at Foil Kss, Texas;
and tKe demoKtion of two school
buldmgs in Fairoanlu, Alaska f o
evaluate if the denofition achvitws and
their associated dust control practices
wore able to prevent downwind
elevations of asbestos concentrations
and to measure the tonconritant worker
exposure levels.
     For the study detailed here,
     air sample"! for tran.vmisxion elec-
     tron microscopy (TEM) were col-
lected on short-cowled ihree-picce sam-
pling cassettes containing a 0.45 pm pore
mixed cellulose ester (MCE) membrane
fil ler hacked by a 5 (am pore MCE diffusing
filter and support pud.  Analysis of the
ambient air samples was performed in the
RREL Electron Microscopy Facility using
the TEM following al I rhe standard QA/QC
procedures and methodology required by
die Asbestos Hazardous Emergency Re-
sponse Act1 (AHERA) protocol. Person-
nel monitoring at the Santa Cruz landfill
and (he Fairbanks landfill during disposal
were collected and analyzed in accordance
with the National Institute for Occupa-
tional Safety and Health (NfOSH) 7400
phase contrast microscopy (PCM) proto-
col2 and by TEM according to a modified
AHERA protocol, where all length and
width measurements am recorded. Sample
volumes during these investigation* varied
according to the duration of the particular
task and ranged from about 130 liters to 545
liters, yielding limits of detection of about
0.01 fibers per cubic centimeter (f/cc)
maximum to 0.003 f/cc minimum. Person-
nel monitoring at the Fort Bliss siic was by
PCM alone,  Flo* rates  for personal
samples typically were between 1.5 to 2
liters per minute. For siaiistical purposes.
TEM samples in which no asbestos struc-
tures were counted were reported as zero
-structures per cubic centimeter (N/CC).

(oifonua Earthquake

   After the 1989 California earthquake
along the San  Andreas fault,  damaged
buildings were condemned and rapidly
demolished.  The Monterey Bay Unified
Air Pollution Control Di.strkt  and EPA
Region IX asked RREL to evaluate the
effectiveness of emission control practices
used during demolition in preventing sig-
nificant airborne asbestos release in ihc
communities involved RREL monitored
two separate demolition activities, one in
the Pacific Garden Mall in Sana Cruz and
one in downtown WatsonsvilJe. In bodi
locations, building construction was simi-
lar, being mostly two-story brick buildings
with common walls to adjacent buildings.
The existence of asbestos in the structures
couid not be confirmed, because access 10
the interior of the buildings for observa-
tions and bulk sampling wa.s prohibited far
safety reasons. Since asbestos existed in
similar nearby undamaged buildings, it
was presumed to wist in the demolition
areas a* well. Typical asbestos-containing
materials  (ACM) In adjacent buildings
consisted of vinyl asbestos tile and ihentul
system insulation on pipes and boilers.
Control practices consisted of spnying the
demolition site with water from fire boxes
while demolition bulldozers, cndloudcr*.
anft trucks were operating.  In Santa Cruz
(Table 1). the demolition activity released
minimal asbestos. In Watsons«illc. higlwr
levels (Table I) measured downwind of the
demolition were sutuatcully significani (y
= O.OU2) and were probably caatad by ihe
collapse of u three-story building during
the monitoring period.
   At Santa Cruz, the cJemnJition ilcbris

-------
             ToUe h- Asbestos RtiMrw Owing DemofrkM and Wsposd, Santa Cruz and Watsonvie, CaWofnia, 1989
bomtnag xn
Pacific Garden
Mai, Santa Cm





Munkiad LandR
Santa Cruz





Worstmvllle

Loamon

ladqpmvnd
Downwind
BadtfrMMd
V^^AiUUMMJL4^H^M
INIHHMnHn
•odarmnd



•wicnjrwnid
Dawnwind
MdezcrCah
Opnroto/


Iwfcflrouiid
Downwind
Daft
•MnnjilM

11/03/89
11/03/89
11/04/89
11/04/89
11/06/89
11/06/89

11/03/89
11/03/89
11/04/89
11/04/89


11/07/89
11/07/89
*$/« = Slroctares per eubk cMiftaetef
Ninaoef BI
Samples

9
10
10
10
10
10

4
5
6
2


4
5
Al means
mean AJTLWKI
Volume (L)

1379
1527
1029
1023
895
1030

975
1301
210
176


3111
804
•re QntineiK.
wrraiKNi
s/cc* (Mean)

0.005
0.003
0.006
0.008
O.OOS
0.004

0.006
0.005
0.009
0,060 (
1
i
0.006
0.051

was taken 10 the local municipal landfill.
causing interest in the potential for worker
exposure. At the request of the Monterey
Bay Unified Air Pollution Control District,
two days of activity at the dump  were
monitored, concurrent with the sampling of
the demolition. Area samples were taken
on  the first  day (for TEM analyses) and
personal monitoring samples (for  both
TEM and PCM analyses) were taken on die
second day. No significant increase was
seen (p = 0.82S4) when the upwind and
downwind asbestos levels were compared
using TEM analysis; however, TEM analy-
sis  of personal samples taken on the  bull-
dozer operator revealed elevated levels
(Table I).   The PCM counts on these
samples averaged 0.32 f/cc. The high PCM
counts prompted an inspection of those
filters by TEM at low magnification to
determine if the fibers seen by PCM were
asbestos — none were.
    The size distribution of the .structures
counted by TEM for the Pacific Garden
Mall and Watsonvjlle appear in Figure 1.
Figure 2 presents similar data for the area
and personal samples taken at the Santa
Cruz landfill. In all cases, the majority of
the structures seen was leas than 2 MJTI in
length: however, Acre was an apparent
increase in .structures above 2 urn on the
samples from  rhe pump mounted in the
bulldozer cab and in those from the breath-
           Figure 1— Size Range (Area Sonifies)

                  Pacific Gordea Mail
Background
Downwind
0-1        1-2        2-3        3-4        4-5
               Parrifl* Length |««§* (Micr«M*t*ri)
                 VL'afionville,
    >5
            OMCK TcdnM!

-------
Table II — Asbestos Refease During Implosion Demolition
U5
ing rone of the operator, relative K> the area      Visible emissions were observed dur-
simples.  All asbestos siruaures observed  ing the structural collapse of buildings but
in the area and personal samples contained  were generally not apparent during loading
chrysoiile.                              operations when firehoses were used to wet
                      the debris. _ There
                      were, however, in-
                      stances of statisti-
                      cally  significant
                      elevation  (p  <
                      0,03)  of  airborne
                      asbcsios  levels
                      above background
                      during the  han-
                      dling of debris de-
                      spite the lack of
                      visible emissions.
                      These limited data
                      support      the
                      premise in the Na-
                      tional  Emission
                      Standards    for
Hazardous Air Pollutants (NESHAP) (Pro-
posed Rules, January 10,1989; 40CFR Part
61, page 925) lhat the  absence of visible
emission is not sufficient evidence to as-
sume no fugitive p&nicuJate emission oc-
curs.

Implosion

    RREL was able to  monitor an implo-
sion-type demolition of a 26-story building
from which all known asbestos (other than
vinyl  asbestos rile [VAT])  had been re-
moved in full accordance widi the provi-
sions of the applkable asbestos NESHAP
Ak sampling siics  were placed at  four
quadrants centered about the building at a
distance of approximately 500 to 600 feet,
Samples were collected  downwind prior to
che implosion to serve as baseline data. The
downwind sampling protocol consisted of
one set of samples to be started immedi-
ately after implosion which were tn run for
1 _5 hours. Additionally, two sets of three
sequential samples  of  0.5-hour duration
were collected.  One set was started at ihe
rime of the detonation, with the second set
being started ten minutes later. The initial
1.5-hour samples  were overloaded with
debris within minute*, so a second set was
started approximately 20 minutes after the
implosion.  Tho first sample  of the rwo
sequential sets was  also found to be too
heavily loaded to analyze by direct prepa-
ration methods. The results from die re-
maining samples are presented in Table H,
with the size range of panicles at the vari-
ous time intervals given in Figure 3.  The
data showed initially elevated airborne as-
bestos levels tlownwjnd of the site, which
rapidly decreased in concentration and stzc

-------
 with time. No elevation above background
 was observed in the other three quadrant
 samples.   Bulk samples of the airborne
 debns which settled on surfaces at the same
 distance downwind as the air samples were
 analyzed  by Polarized Light Microscopy
 and cantainedTrace amounts of chrysoole.
     There is no way to assure that all the
 asbestos originally present in the building
 was accessible and/or totally removed (ex-
 cept for the VAT, of course,  which re-
 mained in place).
     Our  conclusion was that the forces
 involved  in the sudden collapse of a  26-
 story building provide sufficient energy to
 make non-friable materials (such is VAT)
 friable and this may also have contributed
 to the  observed asbestos concentration at
 the  time  of the demolition. No control
 options (such is wetting) were utilized dur-
 ing this demolition; it is difficult, however,
 to envision control technologies that would
 be  100%  effective in preventing asbestos
 release considering the massive forces in-
 volved in  this  demolition mode.

 U,S. Army — Foil tfos, B Paso, Ttxos

     RREL assisted the U.S. Army Corps
 of Engineers, Tulsa District, in evaluating
 asbestos release during demolition of sev-
 eral jwo-story wooden barracks. The only
 known asbestos remaining in the building
 was VAT that contained as much as 20%
 chrysotile asbestos over mastic that con-
 tained  15% asbestos and caulking corn-
 pound. The buildings were demolished by
 a bulldozer and a backhoe with a demoli-
 tion claw  bucket  No wetting was used
 during tine demolition process. This work
 was done  under the auspices of a Federal
 agency asbestos workgroup known as the
 Asbestos  Development and Demonstra-
 tion Initiatives Group (ADDIG), which is
 comprised of the EPA, components of the
 Department of Defense (DoD) (including
 the Air Force, Navy, Army, and DoD De-
 pendents Schools)  and several other fed-
 eral agencies.
    Air samplers were placed at varying
height";  and distances downwind of the
demolition site.  Dust collection samplers
 were placed at four locations near the up-
wind and  selected downwind  sampling
sites and were open for Ihc duration of fee
corresponding air collection period (ap-
proximately  seven  hours.)  The  dujt
samples were analyzed at MVA, Inc. using
Table Ml— Asbestos Release During Fort ttss Dewofitwo Actmfy
Sampfttf
Site
UpwM
Pcwtrwiod
"s/a = Srmetwes
• • *•]

Ne.of MeaiAir
Sairfles VtfcftM (U
4 1264
19 9700
per arfnc ctBtteeter
WM€TlC«
CoKefltratmi
0.0-0.009
.0-0.041

CoKMtralien
s/a {M«M)
0.003
0.014

                             10

                             M

                             40

                             20
                                       Figure 3—Asbestos Pvtkk ln§tfi inge (Unptavoa Demotion)
                                                                        m
                              =Jo«
                                                           SM«MM^au  yQjs^Miy^M
                                                                        m s«^Mty «T« hr 01
                                                           StMnri«40lw  ^^
      1-2
P«rtJ«U
                                                                2-3       3-4
                                                               «§ * (MJ«ra««)
    4-5
>S
                                                 Figure 4—Asbestos Pertide Length Range
                                                          (fort Kst larrafcs)
                                    0-1
    1-2       2-3       3-4
P«rtuf« Length R»§e (Microns)
4-5
                           *n experimental indirect preparation proto-
                           col.  A large number of air and personal
                           samples were collected and sent to an inde-
                           pendent laboratory for analysis and will be
                           reported by the Corps or' Engineers, Sub-
                      sets of these samples were analyzed by the
                      EPA-RRELTEM Laboratory. The results
                      of the EPA air samples are presented in
                      Table ID, with the accompanying size dis-
                      tributions in Figure 4.  The asbestos »truc-
ONNces Ttcfaicrf
                                   • MorA/Apd 1993

-------
            Table IV— Asbestos Release Dfriq DonofitiM end ftspasd
                            Aurora Bamfltvy Sd»oJ
Sampling
 Site

 Upwind
 Downwind
Damp Downwind
No. of
Samples

 24
 32
  6
                              McanAir
   KRSOHAl SAMPtES
   Bockhc*
   Twk Driver
   Dump Cot
S
6
1
                              2112
                              2295
                              1906
                           JU
                           425
                           390
                       0-0-0,004
                       0.04.028
                       0,04.002
                       0.004-0.176
                       O.Or 0.05) difference
 between the upwind and downwind asbes-
 tos levels. A single spike of asbestos re-
 lease (0.096 s/cc)  was  observed  in one
 sample as detected by TEM.  All structures
 found contained chrysolite <3 (im in length
 with the exception of one amphibole fiber
 (actmoJittj. There were no significant re-
 leases of asbestos observed by TEM analy-
 sis  in die downwind samples at the dump
 site. TEM levels from the Fort Wainwrighi
 school were also negligible (Table V). The
 downwind samples from the two days of
 demolition activity had a total of rwelve
 observed chrysolite asbestos structures, all
 < 1.5 jam in length.
    The worker exposure levels as deter-
 mined by PCM were all below ihc OSHA
 action level of 0.1 f/cc, TEM examination
 of the personal samples was conducted to
 identify and quanticate the actual concen-
 tration of asbestos fibers  lunger than 0.5
 \an. The results appear in Table IV and
 Figure 5. The operations performed by the
backhoe and truck drivers caused the high-
est exposure concentrations  (as opposed
to   umc-weighced-averagcs,  including a
                                                                          d dMius Tedmol S«HI|»HI* • tort/April 1993

-------
                                   Figure 5—4sb«tos Particle Ungth Range
                                                  Aurora Dfs»o»l
                                                                lackhoc Operator

                                                                Truck Driver
                                                                      Cat Operator
           0-1
1-2
2-3
        3-4            4-5
Unf th laiig* (MUroni)
>5
number of chrysotile fibers over 5 um long
and several long amphibole fibers includ-
ing amosite, tremolite and  actmolite.
Ticrnolite and actinolite may  have been
present in the floor tile as they often occur
as "contaminants" of the chrysotile used.
Amosite was the most common amphibole
and suggests the presence of an unidenti-
fied source.

Sommory

    These four  studies include three sites
in which  the friable asbestos had been re-
moved prior to demolition (Fort Bliss,
Fairbanks, and die implosion site) and two
sites in which no pre-demoliuon removal
had   been  done (Santa Cruz  and
Watsonsville).
    While these few number of sites can-
not be considered  as representative of all
demolition activities,  the sites  where the
friable asbestos had been removed prior to
demolition had no statistically  significant
(where p values were < 0.05) increase in the
downwind asbestos concentration as a re-
sult of the demolition activity, except in die
case of the implosion technique. The sites
where no pre-removal was done experi-
enced several instances of brief, statisti-
cally  significant elevations of  downwind
asbestos concentrations. None  of the per-
sonal  Camples exceeded the OSHA action
            level when measured by the NIOSH 7400
            PCM method; however, the presence of
            chrysotile and amphibole fibers on these
            samples as revealed by TEM would indi-
            cate thai these activities should be closely
            monitored.
                     ACKNOWLEDGMENTS

               The authors wish to thank the follow-
           ing for their contributions TO the research
           aummmrued in this paper Asbestos Dem-
           onstration and Development  Group, and
           particularly  Gary  Jacks  of die U.S. Air
           Force and Dick Caldwell of NAWAC for
           partial funding; Diana R edmond of RREL:
           Burl Ragland  and Richard Smith of the
           U.S. Army Corps  of Engineers; Eugenia
           Shirom, Cory DeMaris, and Susan Seitz of
           IT Corporation for TEM analytical support
           at RRJEL; Barb Meyer 
-------