EPA 600-D-91-104
PB91-196048
Bake-Out o£ a Portion of a New High-Rise Office Building

(U.S.) Environmental Protection Agency, Research Triangle Park, NC
1991

-------
TECHNICAL REPORT DATA
(Mtest rttdInttrutiiont on ihr recent btfoft romplfl
\. *ero*T no. 2 j
EPA/600/D-91/104
> P By 1- 60 Id
«. TlJLE AHO SU9TITLC
Bike-Out of a Portion of a New
April 22, 1991
High-Riae Office Building
». ^EKFOKMIWC OKGANilAHOKi CODE
7. *.yTHO«isi
Redrew 6, Lmdstrom, Richard K. Taft (U.S. EPA Region 13
Larry C. Michael, Maurice C. Oberg (addressee below J
(,^£«F0AMtNG OPlCANIlAI ION AEFORT NO.
)
e *lAJO*MING ORGANIZATION NAM.C AWD AOOKtSS
U.S. Environmental Protection Agency
Atmospheric Research and Exposure Assesament Laboratory
Exposure Assessment Research Division
Research Triangle Park, NC 27711
10. r*0G*AM I Ut ME NT NO.
11 COnTHACVgRANI HO
)? i^QMSOpmC AC£WCV NAi/l A.NG AOOBES5.
0.5. Environmental Protection Agency
Atmospheric Research and Exposure Assessment Laboratory
Exposure Assessment Research Division
Research Triangle Park, NC 2,J71l
iZ.lYH Of T AND HRlOD CO vlNtD
r-i rj ]
M. SPONSORING ACtNC* coot
,\^lcutuff,r'*t "" 8"h 'n"U*1 of th' s
Association, Vancouver, B.C. June 16-21, 1991
i
RTP/ HC 2")709'i anri u>	* Triangle -.ns-itutej 304C Cornwall is Rd.
C' °b"S «=•««'- ""If Services, 1,„ Tulate
"Li*din"^"s^1F',°ci^rtc"?°ft'*rfOC"*d " °'S'	»e«io" ™ Headquarters'
r.duce Ldoor ^ir	i?,J"ly' 15'°' **'	«• bake-out ... tb
compounds (VOCo) found in a variety o^cm^Io ^"Tirt^8"3 With r*sldlJal vclatilc organic
Four nearly identical fi^« f 1 L ,f	building materials and finishing agents.
for 34, 38, 54 and 66 houra ^ h 6 &tory office building were heated above 30 «C
pollutant ;=";«r«Ln^ »c .L a*^	*"*" °'	du"tl" P=«-"«>«
after the bake-out. The results LnA-r*^ ® -amPliRS w®b conducted before, during, and
source strengths were reduced durinn .h \ v***	volatile organic compound (TVOCJ
targeted Wc	V **-Mt b* « " « »• si.llar reductions L
periods. While the most abundant i*"v! j served from tha Pre- to post-bake monitoring
approximately 65t dur"o	V " ! ""f""	increased
conc.ntr.tio!:.
,-j IC£V WORDS AND DOCUMENT ANALYSIS
», OC*CHI»TORE
b.tpENlimtrtWlN ENDED TERMS
t. COSATi Fteid/Croup
"Bake-Out
Volatile Organic Compounds
Indoor Air Pollution

-
tk.lDl«'T*l»UTl6N ITATIMtNT
Release Unlimited
19, CtCUftlTV CUAtt ftfcu tttparo
21.>0. Of »AGES
14 Pages
to, ttcumrv ci\tt
». miti
fw» Jiifl-H**" 4-77) PACVIOUI tOJTlON It C»»0vtTI

-------
Pii 91 — 19 60 ^ 6
EPA/600/D-91/104
91-62.3
Bake-Out of a Portion of a New High-Rise Office Building
Andrew B. Lindstrom
U.S. Environmental Protection Agency
Atmospheric Research and Exposure Assessment Laboratory
MD-56
Research Triangle Park, NC 27711
Richard M. Taft
Health and Safety Office
EPA Region IX, M/S P-l
75 Hawthorne Street
San Francisco, CA 94105
Larry C. Michael
Research Triangle Institute
Dreyfus Laboratory
3040 Comwallis Rd.
Research Triangle Park, NC 27709-2194
Maurice C, Oberg
Certified Health Services
1772 Tulare Avtnue
Richmond, CA 94805
DISCLAIMER
This paper has been reviewed in accordance with the U.S. Environmental Protection Agency'* peer and
administrative review policies and approved for presentation and publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
1

-------
91-62.3
INTRODUCTION
Building materials and finishes commonly used in renovation and construction can emit appreciable
concentrations of residual volatile organic compounds (VOCs) into the indoor environment.I,: While the
concentrations of many of these compounds tend to decrease over time,3 they remain a concern because many
are thought to be responsible for the sensory irritation and fatigue associated with Sick Building Syndrome.4
Building bake-out, a process which involves the sustained heating of a new or recently renovated structure (e.g.,
to approximately 30 - 35 eC) to accelerate off-gassing of VOCs, has been suggested as an effective method for
reducing the concentrations of these potentially problematic compounds.3 While many uncertainties concerning
bake-out duration, temperature requirements, and optimal ventilation rates remain unresolved, the available data
indicate that appreciable reductions in indoor VOC concentrations result from an increase in a building's interior
temperature to 32 CC or higher for several days/
Three main objectives were identified for the present study: (1) to characterize the concentrations of
selected VOCs and aldehydes in a recently constructed high-rise structure (2) to examine the influence of bake-
out duration on the reduction in VOC source strengths observed within a building and (3) to assess the
effectiveness of implementing a partial bake-out procedure (i.e., a bake-out involving only a portion of a
particular building).
MATERIALS AND METHODS
This investigation was conducted on the mid-level floors of a 21-story office building in downtown San
Francisco, California. At the time of the study, floors 1-7 had been occupied for approximately one year by a
variety of clerical offices, floors 8-12 had just been completed, and floors 13-21 were still under
construction. Prior to occupancy, the new tenant's management scheduled a bake-out for floors 9 - 12 of the
building. If the results indicated a substantial reduction in contaminant concentrations, and the procedure was
determined to be manageable by the building staff, bake-out of other areas within the structure was to be
considered.
Two separate variable volume air handling systems are located in the study area; one on the 10th floor
(serving floors 9 & 10), and the other on the 12th floor (serving floors 11 &. 12). Each system consists of a fan
room with a filter bank, blowers, and an outdoor air intake, along with the supply and return ducting. The
systems are designed to deliver air at 800,900 1/min on the lower floors (9 & 10) and 708,900 Wmin on the
upper floors (11 & 12). Both systems deliver a minimum of 1096 outdoor air; fresh air intakes are located on
the eastern side of the building, 10 and 12 stories above street level, respectively. Each floor also has its own
heating system consisting of three fan coil units and an independent hot air supply duct located in the plenum.
There is approximately 2,950 m3 of floor space on the lower floors and 2,560 m5 on the upper floors.
Most of this space is covered with wall-to-wall carpeting, except the northwest quadrant of the 11th floor and
several small storage rooms on each floor, which are tiled. Most of the final construction activities, including
painting and installation of carpeting, were completed three to six weeks prior to pre-bake testing. Some minor
activities (e.g., installation of the cove moldings) continued up until the day before pre-bake monitoring. In
addition to a single set of freight elevators, the building has a staggered set of passenger elevators. One bank
serves floors 1*11, while the other serves floors 11 - 21.
Bake-out
Previous studies have demonstrated that bake-out duration is approximately proportional to the degree
of contaminant reduction achieved.* In this study two different heating periods were employed in the same
2

-------
91-62.3
building: the upper floors (11 & 12) were scheduled for a 108-h heating period, and the lower floors (9 & 10)
were scheduled for 73 h of heating.
Since recent bake-out studies have demonstrated lhat heating systems in moderate climates can have
some difficulty generating and sustaining the high temperatures required for a bake-out,® and since these studies
have also indicated that elevated air change rates during a bake-out may be less important to VOC reduction
than sustained high interior air temperatures, it was decided to '¦'ose the test floors' fresh air supply dampers to
reduce infiltration of cool outdoor air during the procedure. Both air handling systems' fresh air dampers were
adjusted to provide the normal minimum flow of fresh air (10%), and then further reduction of fresh air input
was achieved by partially covering the intake (approximately 50%) with a sheet metal plate. After initiating the
bake-out, however, it became apparent that interior temperatures would not reach target levels unless auxiliary
heating measures were instituted. To this end, boiler water temperature (to the fan coil heating units) was
increased from 49 "C to 54 °C, and both air handling systems' fresh air intakes were completely covered with
the sheet metal plates. This procedure ensured an absolute minimum infiltration rate for the remainder of the
bake-out. At the end of the bake-out period, the fresh air dampers in the test floors' air handling systems were
set to introduce fresh air at the maximum flow rate for approximately 60 h to accelerate removal of outgassed
contaminants and to reduce building temperatures. Twelve hours prior to post-bake sampling, the air handling
systems were readjusted to their original pre-bake fresh air flow rates.
Sampling
All integrated and grab VOC samples were collected in precleaned, evacuated Summan,-polished
stainless steel canisters following established indoor air measurement protocols.1 Aldehyde samples were
collected on 2,4-dinitrophenylhydrazine (DNPH)-coated silica gel cartridges and were identified and quantified
by high-performance liquid chromatography.* Temperature and humidity data were collected at each sample site
with recording hygrothermographs and recording thermometers.
Air samples were collected on three different occasions during the study: pre-bake sampling on July 26,
starting ten hours before the bake-out began; mid-bake sampling on July 29, approximately 60 hours into the
bake-out; and post-bake sampling on August 3, about 72 hours after the bake-out was completed. Sampling
generally consisted of one 8-h hour integrated VOC and aldehyde sample collected between 8:00 a.m. and 4:00
p.m. in the same south-central locations on each of the test floors. Ambient air samples were also collected on
the roof adjacent to the upper floors' (13-21) air intake vent during the pre- and post-bake phases. In addition
to the integrated VOC samples, a single VOC grab sample was collected at the primary sampling site on each
floor at approximately 11:00 a.m. during the pre- and post-bake phases of the study. The mid-bake sample
collection consisted of 8-h integrated VOC and aldehyde samples on floors 10, 11, and 12, approximately 24 h
after the temperature on all the test floors reached 30 °C (- 60 h into the bake-out). Air change rates were
determined during the pre- and post-bake phases of the study by American Society for Testing and Materials
sulfur hexafluoride tracer gas methods.* Table I is a summary of the sampling conducted during the study.
Compound source strengths' (5 in ftg/ml h) were calculated from the measured concentrations by using:
S - (C,-CJ*V* ACH I FS
where C, is the indoor compound concentration (/ig/m3), Ct is the outdoor concentration, V is the volume of the
floor (m5), ACH is the number of air changes per hour, and FS is the floor space in mJ.
While every effort was made to prevent unauthorized access to the test floors during the experiment,
construction-related activities were discovered between the mid-bake and post-bake sampling periods. On
7/30/90 (the day after the mid-bake monitoring period), workers removed and replaced small tactions of tile that1
covered electric cable access points on the 1 Ith floor and tiled a small storage room on the 12th floor. Both
3

-------
91-62.3
activities involved the use of tile adhesives and a white gasoline torch. On 7/31/90 (three days before the post-
bake sampling period) a single door in the southeast section of the 9th floor was painted with an oil-based
enamel paint.
VOC and TVOC Analysis
Canister samples were analyzed for selected target compounds by using an LKB 2091 magnetic sector
GC/MS/COMP system operated in the full-scan mode. Separation of compounds was achieved on a 30-m x
0.32-mm i.d. DB-624 column, temperature programmed from -20 to 200 °C at 5 "C/min.
Total volatile organic compound (TVOC) estimates were obtained by integrating the total
chromatographic areas encompassed in the retention time window from perfluorobenzene through dodecane.
The cumulative area in vacb chromatogram for mass to charge ratio (iq/j) 55 added to that for jn/g 57
represented the aliphatic hydrocarbons. Similarly, the cumulative areas for ffl/j 91 and jjj/i 105 were combined
to estimate total aromatic hydrocarbons. Surrogate aliphatic and aromatic relative response factors (RRF) were
calculated from peak areas in calibration standard runs for Q}/2 55 plus mtj. 57 relative to d«-benzene and
similarly for n]/j 91 plus jn!% 105. Analyte concentrations incorporated into these calculations are sums of the
amounts of aliphatic hydrocarbons (octane, decane) and aromatic hydrocarbons (toluene, ethylbenzne, m,p-
xylene, styrene, 1,3,5-trimethylbenzene, 1,2,4-trimethyIbenzene), respectively, present in the calibration
mixture. As with target compound quantitation, these RRFs were subsequently used to estimate aliphatic and
aromatic hydrocarbon levels in the samples. In all cases, the areas of individual chromatographic peaks were
summed manually across all appropriate retention times to yield total areas in each channel10. This
approach includes only compounds that fragment into the specific mass ions indicated. Benzene, which has
neither jq/j 91 nor 105 ions, is not included, but was measured as a target compound.
RESULTS
The temperature of the upper floors reached 30 °C approximately 20 h after the start of the bake-out;
the lower floors, with their larger volume and surface area, required an additional 20 h to reach this level.
Temperatures were held at or above 30 °C for 34 h on the 9th floor, 54 h on the 10th floor, 86 h on the 11th
floor, and 38 h on the 12th floor. Maximum temperatures for floors 9 through 12 were 32.2, 33.8, 33.9, and
32.0 °C, respectively. Figure 1 represents the temperature profiles of each of the test floors during the bake-
out. Degree-hours (#C hours) of heating (the product of the air temperature minus 27 *C and the number of
hours at that temperature)6 were also calculated. With this method, the lower floors (9 and 10) had a combined
heating degree-hour value of 223 aC hours, and the upper floors (11 and 12) had a combined heating degree
value of 366 °C hours.
Air change rates were fairly constant duruig the pre- and post-bake monitoring periods, ranging from
1.11 to 1.26 air changes per hour (ACH) on the upper floors and from 0.74 to 0.90 ACH on the lower floors
(Table II). As noted above, the fresh air intake vents were partially, and then completely, blocked during the
bake-out period to maximize heating within the structure. (Measurements made with the fully blocked air
handling system on the 12th floor after the study was completed demonstrated an air change rate of
approximately 0.63 ACH.)
Although the extremely short planning period available for this study prevented the preparation and
shipment of field spikes and blanks before the sampling was scheduled to begin, performance evaluation samples
were prepared and analyzed after the field sampling was completed. The results of these performance
evaluation samples (Table III) indicate, with two exceptions (vinyl chloride and styrene), percent bias was
generally on the order of 5 - 15 %.
4

-------
91-62.3
Formaldehyde, aceUldehyde, acetone, and hexanaldehyde consistently accounted for about 93% of the
mass measured in the indoor aldehyde and carbonyl samples (range 92 - 94%). Concentrations of these four
compounds increased between 11 and 110% (mean = 65%) from the pre- to mid-bake measurement period, but
their calculated source strengths (which adjust for changes in air exchange rates) changed between -25 and
+ 47% (mean = -2%) from the pre to post-bake sample periods. Average formaldehyde source strengths
decreased by 7% during the bake-out. Acetone source strengths increased on the lower floors by approximately
42%, possibly as a result of the unauthorized painting on the 9th floor described above.
Of the 28 targeted VOCs monitored during this study, 1,1,1 - trichloroethane, benzene, toluene, n-
octane, ethylbenzene, m,p-xylene, o-xylene, 1,3,5-trimethylbenzene, n-decane, and 1,2,4-trimethylbenzene
consistently accounted for more than 95% of the mass of all the targeted compounds in the integrated samples.
Together, the mass of these 10 compounds ranged from 6 to 53% (mean = 17%) of the TVOC mass in any
single integrated sample. During any given phase of the study, TVOC and individual target VOC
concentrations and source strengths sharing a common air handling system generally remained similar (Tables
IV - VII).
As illustrated in Table VIII (selected compound concentrations on the 11th floor), both TVOC and
individual VOC levels were generally found in highest concentration during the pre-bake measurement period,
concentrations decreased during the mid-bake period, and the lowest levels were measured during the post-bake
monitoring period. While this general decrease in compound concentration was true for all floors, it was more
pronounced on the upper floors (11 and 12), possibly because these floors had much higher initial contaminant
levels than floors 9 and 10. Average pre-bake TVOC source strengths on the 11th and 12th floors
(«• 8,378 jig/m5 h) (Tables IV & V) were more than three times higher than the average TVOC source
strengths on floors 9 and 10 (¦» 2,502 /xg/m* h) (Tables VI &. VII). On the upper floors, 64% of the initial
TVOC level was aromatic, while only 16% of the lower floors' TVOC level was attributable to aromatic
compounds. Individual target VOC source strengths were typic/ "y higher on the upper floors.
n-Octane and n-decane were the only target compounds that generally increased in concentration from
the pre- to mid-bake measurement periods. n-Decane increased on all floors between 59 and 120%, and n-
octane increased on the upper floors by 26 - 37%. Both compounds also demonstrated large decreases in source
strength in the post-bake monitoring phase (decreases of between 39 and 93% of the pre-bake level). The
aliphatic portion of the TVOC concentration also generally increased during the mid-bake monitoring period
(2% on floor 12, 64% on floor 11, and 33% on floor 10), while its source strengths decreased in the post-bake
measurements (between 25 and 58%). In contrast, the aromatic portion of the TVOC level decreased from the
pre- to mid-bake phase by 11% on the 10th floor, 74% on the 11th floor, and by 85% on the 12th floor (Table
IX).
After the bake-out, average post-bake TVOC source strengths on the 11th and 12th floors
(» 2407 fig/m1 h) were still more than 89% higher than TVOC source strengths on floors 9 and 10
(«¦ 1272 fig/mJ h). While individual target VOC source strengths remained similar in the shared air handling
zones, the upper floors' individual source strengths were again generally higher than the lower floors'. Specific
compounds that consistently showed major reductions in source strength on all the test floors from the pre- to
post-bake monitoring periods include n-decane (39 - 93%), n-octane (41 - 91 %), ethylbenzene (45 - 95%), m,p-
xylene (48 - 95%), and o-xylene (41 - 94%). The TVOC source strengths were reduced by 45 - 76% of their
pre-bake levels (Tables IV - VII).
The data from the grab samples are consistent with the integrated sample analysis: Average TVOC
source strengths were reduced by 67% on the upper floors and by 38% on the lower floors during the bake-out.
(Because no pre-bake grab sample was collected on the 9th floor, data from that floor's integrated sample was
used for this calculation.) The TVOC levels from the grab samples were on average within +/- 15% of the
5

-------
91-62.3
corresponding integrated sample value. Much of this variation is attributable to the pre-bake sample on the 10th
floor, which was 43lower than the corresponding integrated TVOC level.
DISCUSSION
Despite exemplary building owner and tenant cooperation, at least two factors made this building less
than an ideal choice for a study of this type: (1) Although the floors scheduled for the bake-out were essentially
completed at the time of the study, predetermined moving schedules prevented any of the floor's furnishings
(e.g., desks, chairs, computers, etc.) from being present; (2) because the test floors were scheduled for
occupation shortly after their completion, the bake-out and monitoring of these floors had to be conducted while
construction work was still in progress on the upper floors. However, because this represented a rare
opportunity to gather data on a new building commissioning procedure, the study was performed despite these
compromises. Another potential problem associated with interpretation of these data is that the ambient air
samples collected on the roof (data in Tables VIII and IX) during the ore-bake phase appeared to have been
contaminated by a building exhaust vent. Roof top concentrations of 1,1,1-trichloroethane, toluene, and the
xylene isomers were from 2-5 times higher than some of the levels measured indoors. This being the case, all
of the source strengths were calculated by assuming that the pre- and post-bake ambient pollutant levels were
equivalent or that any differences in ambient concentrations were negligible. Meteorological data from San
Francisco International Airport suggest this assumption is valid, as the pre- and post-bake sampling dates'
weather conditions were very similar. In any case, the great reductions in TVOC and individual VOC source
strengths documented in this study suggest more significant changes than those attributable to ambient pollutant
levels alone.
This study shows that large reductions in TVOC and target VOC source strengths can be achieved
during a bake-out conducted with minimal air exchange rates. It is more difficult to conclude, however, that
differences in the intensity or duration of the heating period are responsible for the apparent differences in
source strength reduction achieved. The percentage reduction of the aromatic portion of TVOC source strength
from the pre- to post-bake period on the upper floors (90%) is approximately three times greater than the
aromatic reduction on the lower floors (28%). This large reduction is probably related to the initial aromatic
levels in the two air handling rones: Pre-bake aromatic source strengths averaged 5347 jig/m3 h on the upper
floors, but only 406 /ig/m5 h on the lower floors. This 13-fold greater aromatic burden on the upper floors
gives the potential for much greater reductions during the bake-out. Average post-bake aromatic source
strengths on the upper floors were 522 //g/ml h while on the lower floors they averaged only 2-2 /ig/m3 h.
Average aliphatic concentrations decreased on the upper floors by 38% (3031 to 1884 /ig/mJ h) and on the
lower floors by 53% (2096 to 979 fig/m* h) as a result of the bake-out. Taken together, these data do not
indicate any obvious advantage to the extended heating period on the upper floors.
One recent bake-out study has shown mid-bake increases in TVOC and target VOC concentrations
ranging from 50 - 400% higher than pre-bake levels3 despite a mid-bake air change rate of 1.59 ACH. This
large mid-bake elevation in concentration is generally taken as evidence that the high temperatures are
accelerating the volatilization and removal of the contaminants from the structure. The mid-bake sample
collection period in that study, however, apparently occurred only 5 - 10 h after heating began within the
structure. In contrast, the present study's mid-bake sample period began some 60 h after the bake-out began,
approximately 24 h after all the test floors reached 30 °C. This relatively long heating period before mid-bake
sampling probably accounts for the Finding that most of the target VOCs and the aromatic portion of the TVOC
measurement showod large decreases in concentration in the mid-bake sampling period. This suggests that the
most volatile compounds were driven off before the mid-bake sampling took place. Future bake-out studies
might include continuous VOC monitoring before, during, and after the bake-out so an estimation of the
volatilization rates during each critical phase of the study can be made.
6

-------
91-62.3
Although flo^. s 9 & 10 and 11 & 12 are on independent air handling systems, which are also separate
from the rest of the building, some cross contamination from upper and lower floors by way of the elevator
shafts was likely. As noted above, the building's elevators are staggered, and the upper set serves only floors
11 -21. Cross contamination from upper floors still under construction may help explain why floors 11 & 12
typically had higher contaminant concentrations than the lower floors during this study. Another explanation of
the higher contaminant levels on the upper floors, and the more dramatic reductions achieved there, could
involve the application of solvent-laden building materials just prior to pre-bake testing. If significant
construction activities took place on the upper floors before pre-bake testing began, one would expect initial
pollutant concentrations to be higher, and the compound reductions achieved greater, simply because of higher
starting concentrations. Unfortunately, records of the construction activities prior to bake-out are incomplete,
and, because of the day-to-day variability of the work practices observed during construction, would likely be of
little help in resolving this matter, even if they were available.
Previous bake-out studies6 have shown more modest reductions (20 - 30%) in individual target VOCs,
but similar reductions in TVOC levels. The large reduction in individual VOCs noted in this study (Tables IV -
VII) (particularly toluene, xylene isomers, elhylbenzene) might be related to the high initial pollutant
concentrations and the brief period of time between final construction activities and the start of the pre-bake
sample period. Few (if any) bake-out studies have indicated initial pollutant levels as high as some of those
noted in the present study (Table VIII). Comparably elevated pollutant concentrations inside recently completed
buildings have been reported, however, Sheldon el al.,J in an investigation that included recently completed
office buildings (monitored within one month of construction), reported elevated levels of some of the same
compounds identified in the present study: straight chain aliphatic hydrocarbons (n-decane, n-undecane, n-
dodecane) and aromatic hydrocarbons (xylene isomers, ethylbenzene, ethyltoluene isomers, and
tri methyl benzene isomers). Indoor concentrations of these compounds generally ranged between 10 and 400
fig/m3 one month after construction. Interestingly, several months later, without any bake-out, the
concentrations of these compounds generally ranged between 3 and 30 ^g/mJ. Additional studies which
compare the natural rate of contaminant depletion with the accelerated rate of depletion resulting from bake-out
should help to further characterize the effectiveness of the bake-out process.
CONCLUSIONS
This bake-out study shows reductions in TVOC and individual targeted VOC source strengths as great
as, or greater than similar recent studies. Although some compounds increased in concentration during the mid-
bake phase, this study did not demonstrate a large increase in target compound concentrations during the
building bake-out. The long period of sustained heating before mid-bake measurements were made (* 60 h)
probably allowed the concentrations of the most volatile VOCs to be reduced before the mid-bake samples were
taken. Post-bake sampling shows that TVOC source strengths were reduced from 45 to 76% of their pre-bake
levels. Differences in initial contaminant concentrations in the two air handling zones make it difficult to
determine if lengthened heating periods correspond with increased reductions in contaminant levels. Many
targeted VOC were reduced from 45 to 95 % as a result of the bake-out. Although the most abundant aldehyde
species concentrations increased approximately 65 % during the mid-bake monitoring period, post-bake aldehyde
concentrations remained similar to pre-bake levels. We conclude that a partial building bake-out, conducted on
only a few floors of a new high-rise office building, can be an effective technique in the reducing human
exposure to potentially irritating VOCs.
7

-------
91-62.3
Acknowledgment
We thank V. Ross Highsmith and Winona Victory for their help in planning and arranging this research
opportunity and Mark Johnson, Barbara Gross, and Pierre Belanger for their assistance in sample collection and
study coordination. Special thanks are also extended to Kheay Loke and Steve Wilkerson for their invaluable
on-site assistance,
REFERENCES
1.	L.A. Wallace, E. Pellizzari, B. Leaderer, H. Zelon and L. Sheldon, "Emissions of volatile organic
compounds from building materials and consumer products," Atmos. Environ. 21 (2): 385 (1987).
2.	J.R. Girrnan, A.T. Hodgson, A.S. Newton and A.W. Winkes, "Emissions of volatile organic compounds
from adhesives with indoor applications," Environ. Int. 12: 317 (1986).
3.	L. Sheldon, H. Zelon, J. Sickles, C. Eaton, T. Hartwel! and L. Wallace, Indoor Air Quality in Public
Buildiny: volume II. EPA-600/S6-88/009b U.S. Environmental Protection Agency, Research Triangle Park,
1988.
4.	L. Molhave, B. Bach and O. Petersen, "Human reactions to low concentrations of volatile organic
compounds," Environ. Int. 12; 167 (1986).
5.	J. Girman, L. Alevantis, G. Kulasingam, M. Petreas and L. Webber, "The bake-out of an office building: a
case study," Environ. Int. 15: 449 (1989).
6.	J.R. Ginnan, L.E. Alevantis, M.X. Petreas and L.M. Webber, "Building bake -out studies," in Proceedings
of the 5th International Conference on Indoor Air Quality and Climate:Indoor Air '90. Vol 3. International
Conference on Indoor Air Quality and Climate, Ottawa, 1990, pp 349-354.
7.	W.T. Winberry, L. Forehand, N.T. Murphy, A. Ceroli, B. Phinney and A. Evans, Compendium of
Methods for the Determination of Air Pollutants in Indoor Air. EPA-600/S4-90/010, U.S. Environmental
Protection Agency, Research Triangle Park, 1990.
8.	S.B. Tejada, "Evaluation of silica gel cartridges coated in-situ with acidified 2,4-dinitropJ "nylhydnuune for
sampling aldehydes and ketones in air," Int. J. Environ. Anal. Chem. 26: 167 (1986).
9-	Standard Test Method for Determining Air Leakage Rate bv Tracer Dilution. Designation E 741, The
American Society for Testing and Materials, Philadelphia, PA, 1983.
10-	Pilot Study to Calculate Total Levels of Orcanics bv Chemical Classes in Air Samples. EPA Final Report,
Contract No. 68-02-4544, U.S. Environmental Protection Agency, 1990.
DISCLAIMER
This paper has been reviewed in accordance with the U.S. Environmental Protection Agency's peer and
administrative review policies and approved for presentation and publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
8

-------
7/31/90
7/27/90
7/29/90
7/26/90
7/28/90
7/30/90
Q> 30
Pre-Bake
Monitoring
Mid-Bake
Monitoring
		
I	l	I
ooooooooooooooooooooooo
ooooooooooooooooooooooo
<0-•
i
cr>
ho

-------
91-62.3

TABLE 1.
Sample collection schedule.

Floor I




Date
integrated voc
GRAB VOC
CARBONYl AIR
EXCHANGE RATE
12th
7/26
7/26
7/26
7/26
Floor
7/29
...
7/29
—

8/03
8/03
8/03
8/03
11th
7/26
7/26
7/26
7/26
Floor
7/29
- - •
7/29
—

8/03
8/03
8/03
8/03
10th
7/26
7/26
7/26
7/26
F loor
7/29
...
7/29
...

8/03
8/03
8/03
8/03
9th
7/26

7/26
7/26
F loor


...


8/03
8/03
8/03
8/03
Roof
7/26
7/26
7/26
...

8/03
8/03
8/03
...
* 7/26 «
Pre-lake, 7/29
* Kid-bake,
8/03 » Post-bake

* Not
Measured



TABLE II. Average air change rates
FLOOR P RE-BAKE	MID-BAXE	POST-BAXE
(7/26/90)	(7/29/90)	(8/03/90)
12 1.11	0.63	1.26
11 1.20	--	1.21
10 0.81	--	0.84
9 0.74	—	0.90
Measured air exchange rate under identical
conditions (i.e., fresh air intake fully
blocked) made on 11/01/90.
—¦ Hot Measured
10

-------
TABLE III. Performance evaluation samples ).
91-62.3
COMPOUND
Vinyl Chloride
Vin'ylidene Chloride
Hethylene Ch\or^de
1.1-D	ichloroethane
Ch Wof orm
1,1,1 -Tr ichloroethane
Carbon Tetrachloride
Beruene
Toluene
TeVrach^oroethylene
1.2-D	ibromoethane
Chlorobenzene
Ethylbenzene
o-Xylene
Styrene
NOMINAL
SPIKE 106
REPORTED
X BIAS NOMINAL
SPIKE 089
REPORTED
X BIAS
BLANK 004
REPORTED
8.8
0.8
90.9
6.2
0.6
90.3
ND
13.6
13.9
-2.1
9.6
10.3
- 7.2
ND
23.9
23.2
2.7
17.2
15.9
7.5
ND
14.7
12.9
12.3
10.6
9.2
12.4
ND
16.8
15.7
6.3
11.8
11.8
0.0
ND
18.7
16.7
10.9
13.2
12.6
4.8
NO
20.3
20.8
-2.4
16.1
15.7
2.5
NO
22.3
20.5
7.9
16.2
15.0
7.0
ND
13.7
12.6
8.0
9.9
9.3
6.0
T
24.7
22.2
9.9
17.8
16.0
10.1 .
NO
27.9
25.4
9.1
20.2
18.0
10.8
NO
33.5
30.8
8.0
23.7
23.3
1.8
ND
32.5
29.7
6.S
23.-3
22.0
5.4
HO
36.9
27.8
24.6
26.3
21.1
19.8
ND
15.1
ND
98.6
10.8
NO
99.0
NO
T * Trace
ND « Not Detected
TABLE IV. Selected VOC source strengths (^g/m2 h)
from integrated samples on the 12th floor.
X REDUCTION*
COMPOUND	PRE-BAKE POST-BAKE PRE TO POST
formaIdehyde
Acetaldehyde
Acetone
Hexanaldehyde
1,1,1-Trlchloroethane
Benzene
Toluene
n-Octane
Ethylbenzene
m,p-Xylene
o-Xylene
1,3,5-Trimethylbenzene
n-Oecane
1,2,4-Trlmethylbenzene
Aliphatic Portion
Aromatic Portion
Total TVOC
* A negative value Indicates an Increase In source strength
T • Trace
NC • Not Calculated, Chromatographic Interference
98.4
23.0
49.6
10.4
44.6
16.0
522.9
16.1
229.4
872.7
295.5
27.3
56.3
41.7
3153.4
5423.7
8577.1
99.1
22.7
52.0
8.9
57.5
17.8
67.0
T
12.0
44.1
16.6
7.0
NC
22.2
1583.8
SOB. 1
2091.9
-0.6
1.3
-4.8
13.9
-28.4
-10.8
87.2
40.8
94.8
94.9
94.4
74.4
46.7
49.8
90.6
75.6
11

-------
TABLE V. Selected VOC source strengths 0*g/m: h)
from integrated samples on the 11th floor.
X REDUCTION*
COMPOUND	PRE-BAKE POST-BAKE PRE TO POST
Formaldehyde
142,2
126.3
11.2
AcetaIdehyde
27.0
24.8
8.2
Acetone
57.0
54.3
4.8
HexanaIdehyde
15.0
11.8
21.4
1,1, l-Trich1oroethane
61.2
39.1
36.1
Benzene
19.2
16.3
15.2
Toluene
547 .3
69.1
87.4
n-Octane
30.2
12.3
59.3
Ethy lbenzene
218.7
12.5
94.3
m,p-Xylene
830.8
48.3
94.2
o-Xylene
273.2
16.8
93.8
1,3,5-Trimethylbenzene
25.3
ND
98.3
n-Decane
78.5
ND
93.0
1,2,4-TrImethylbenzene
44.0
25.4
42.4
Aliphatic Portion
2907.9
2184.8
24.9
Aromatic Portion
5270.9
536.3
89.8
Total TVOC
8178.8
2721.2
66.7
* A negative value indicates an increase in source strength
ND ¦= Not Detected
TABLE VI. Selected VOC source strengths (jiglm7 h)
from integrated samples on the 10th floor.
% REDUCTION*
COMPOUND	PRE-BAKE POST-BAKE PRE TO POST
Formaldehyde
111.5
103.4
7.3
AcetaIdehyde
17.5
20.2
-15.4
Acetone
27.5
37.7
-36.9
HexanaIdehyde
14.4
12.7
11.4
1,1,1-Trichloroethane
14.8
12.5
15.5
Benzene
12.4
11.0
10.9
Toluene
46.2
44.7
3.1
n-Octane
47.5
1
90.8
Ethylbenzene
11.8
5.7
51.7
m,p-Xylene
46.2
20.6
55.3
o-Xylene
15.3
7.2
53.2
1,3,5 - Tr tme'chy 1 benzene
4.1
3.6
11.9
n-Oecane
32.8
NO
88.4
1,2,4-Tr imethylbenzene
13.6
11.9
12.9
Aliphatic Portion
2138.6
906.1
57.6
Aromatic Portion
408.9
277.6
32.1
Total TVOC
2547.8
1183.B
53.5
* A negative value Indicates an Increase in source strength
T » Trace
ND ¦ Hot Detected
12

-------
91-62.3
TABLE VII. Selected VOC source strengths (jiglm7 h)
from integrated samples on the 9th floor.
% REDUCTION*
COMPOUND	PRE-BAKE POST-BAKE PRE TO POST
Formaldehyde
114.0
103.3
9.4
Acetaldehyde
14.5
16.3
-12.9
Acetone
24.8
36.5
-46.9
Hexanaldehyde
13.7
10.3
24.6
1,1.1-Trichloroethane
1069.4
13.3
98.8
Benzene
9.9
12.3
-24.4
Toluene
46.3
47.6
-2.9
n-Octane
49.0
T
90.4
Ethylbenzene
12.3
6.8
44.8
m,p-Xylene
44.8
23.4
47.8
o-Xylene
14.6
8.6
41.4
1,3,5-Trimethylbenzene
4.2
3.9
6.0
n-Decane
32.6
T
38,6
1,2,4-Trimethylbenzene
12.5
12.6
-0.3
Aliphatic Portion
2053.6
1052.6
48.7
Aromatic Portion
403.2
307.3
23.8
Total TVOC
2456.8
1359.9
44.6
* A negative value indicates an Increase In source strength
T « Trace
TABLE VIII, Selected VOC concentrations Qig/m}) from
integrated samples on the roof and the 11th floor.
% REDUCTION ROOF	ROOF
COMPOUND	PRE-BAICE MIG-BAKE* POST-BAKE PRE TO POST PRE-BAKE POST-BAKE
Formaldehyde
32.6
48.7
28.7
11.2
7.2
6.9
Aceta ldehyde
6.2
11.1
5.6
8.2
11,1
6.0
Acetone
13.1
16.3
12.3
4.8
8.2
B.O
1,1.1-Trichloroethane
14.0
7.0
8.9
36.1
28.5
4.1
Benzene
4.4
1.6
3.7
15.2
3.1
2.5
Toluene
125.5
49.2
15.7
87.4
27.8
7.5
n-Octane
6.9
8.7
2.8
59.3
NO
T
Ethylbenzene
50.2
7.3
2.9
94.3
7.8
1.3
m,p-Xylene
190.6
30.6
11.0
94.2
31.3
4.9
o-Xylene
62.7
12.5
3.8
93.8
10.8
1.9
1,3,5-Trimethylbenzene
5.8
2.4
NO
98.3
1.9
T
n-Decane
18.0
39.1
NO
93.0
T
T
1,2,4-Trlmethylbenzene
10.1
7.7
5.8
42,4
4.3
2.4
Aliphatic Portion
667.0
1097.0
497.0
24.9
324.0
181.0
Aromatic Portion
1209.0
309.0
122.0
89.8
231.0
55.0
Total TVOC
1876.0
1406.0
619.0
66.7
555.0
236.0
* Mld-bake air change rates not measured. Pre- and Post-Bake air change rate ¦ 1.2.
NO • Not Detected
T » Trace
13

-------
TABLE IX. Hydrocarbon concentrations from integrated samples (^g/m').*
COMPOUND or
CLASS
FLOOR
PRE
HID
POST
Aliphatic
9
764

322
Aromatic
9
ISO
...
94
Total
9
914
—
416
Aliphatic
10
727
969
297
Aromatic
10
139
124
91
Total
10
666
1093
368
Aliphatic
11
667
1097
497
AromatIc
11
1209
309
122
Total
'.1
'*876
1406
619
Aliphat fc
12
782
797
346
Aromat1c
12
1345
201
111
Total
12
2127
996
457
Aliphatic
Roof
324
—
181
Aromat1c
Roof
231
...
55
Tota 1
Roof
555
—
236
* Concentrations not adjusted for changes In air
exchange rate
14

-------