600/A-98-047
Effectiveness of Vacuum Cleaning on Fungally
Contaminated Duct Materials
Karin K. Foarde and Douglas W. VanOsdell
Center for Engineering and Environmental Sciences
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709-2194
J.C.S. Chang
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ABSTRACT
Ventilation system materials are of particular signincance as potential micruuuu coiuammauuu
sources because of their potential to rapidly spread contamination throughout a building. Portions of
ventilation systems near cooling coils and drain pans are known to be exposed to high moisture levels for
extended periods, and fibrous duct insulation materials are known to have become sources of microbial
contamination in some buildings.
Cleaning has been suggested as a possible strategy for the prevention of microbial contamination,
as well as for the remediation of already contaminated materials. However, the efficacy of cleaning has
not been determined. Recommendations that materials appearing to be wet or moldy should be discarded
are not always followed.
The objectives of this research program were to: determine, under constant temperature, relative
humidity, and air flow test conditions, whether fungal spores levels on HVAC (heating, ventilating, and
air-conditioning) duct material surfaces could be substantially reduced by thorough vacuum cleaning, and
evaluate whether subsequent fungal growth would be limited or contained by a single mechanical cleaning
treatment.
Three fiberglass duct materials were tested. Ail were artificially soiled. The results showed that
notable amounts of surface dust were removed, and surface spore levels could be reduced in the short
term on all materials by vacuuming. However, regrowth occurred within 6-12 weeks.
INTRODUCTION
Because of their location and consequent potential to rapidly spread contamination throughout a
building, ventilation system materials are of particular significance as potential microbial contamination
sources. Portions of ventilation systems near cooling coils and drain pans are known to be exposed to
high moisture levels for extended periods, and fibrous duct insulation materials are known to have
become sources of microbial contamination in some buildings1. The cause is invariably the presence of
adequate moisture and nutrients in those materials for extended periods of time.
Microbial growth can be limited by controlling the amount of moisture and nutrients in or on
building materials2,3,4'5. And in general, the availability of nutrients may be influenced by: good cleaning
and maintenance practices, building pressurization and outdoor air filtration, and selecting materials that
provide limited encouragement to microbial growth. Many of these practices are currently a routine part
of good construction practices and good operation and maintenance (O&M). Unfortunately, either
through failure to follow what are generally considered good practices or some other combination of

-------
circumstances, building materials do at times become contaminated and serve as sources of microbial
contamination in buildings. When this situation arises, it can severely impact the building's IAQ (indoor
air quality). Exposure to microbial spores can cause severe allergic and/or toxic responses among the
occupants.
To help address this issue in duct materials, both the National Air Duct Cleaners Association
(NADCA) and the North American Insulation Manufacturers Association (NAIMA) have published
guidelines: "Understanding Microbial Contamination in HVAC Systems"6; and "Cleaning Fibrous Glass
Insulated Air Duct Systems"7. In them the authors address a number of important issues associated with
the cleaning of fibrous duct liner, including a discussion of when and if insulated ducts should be cleaned.
The NAIMA document states that fibrous glass insulation that appears to be wet or moldy should be
discarded. Unfortunately, that advice is not always followed. Instead, cleaning is being used in the field
as a method for the prevention of microbial growth, as well as for the remediation of fibrous glass duct
liner that is already contaminated with microbial growth. However, the efficacy of cleaning as either a
prevention or remediation strategy for microbial contamination has not been determined.
As summarized in NAIMA's recommended practice, three types of cleaning techniques are most
commonly used for air duct cleaning: contact vacuuming, air washing, and power brushing. Of the three
cleaning methods, contact vacuuming has the most potential for cleaning fiingally contaminated fibrous
liner, because fungal growth is not confined to the surface of the fiberglass materials. Air washing and
power brushing are largely surface techniques. Contact vacuuming is hand-operated to ensure direct
contact between the brush on the vacuum nozzle and the interior surfaces of the ducts to dislodge and
remove dirt and debris. Operating parameters such as contact time, number of passes, and vacuum level
can also be controlled in the laboratory. Therefore, contact vacuum was selected as the cleaning method
to be evaluated during this research to provide the most rigorous test of duct cleaning's potential
effectiveness.
The objectives of this research program were to: determine, under dynamic test conditions,
whether fungal spores levels on HVAC duct material surfaces could be substantially reduced by thorough
vacuum cleaning; and evaluate whether subsequent fungal growth was limited or contained by that clean-
ing. The constant high relative humidity (RH) environmental condition to which the test materials were
exposed during this study was chosen because it can be achieved downstream of a continuously operating
air-conditioning coil, albeit at a lower temperature. (We have previously shown that lower temperatures
slow but do not stop growth3.) There was no liquid water carryover in the high RH air, however, as
might occur in an operating HVAC system. Operated this way, the miniducts were intended to provide a
serious but realistic challenge for duct materials.
METHODS AND MATERIALS
Duct Materials Tested
Three new duct materials were tested: two brands of fiberglass duct liner and one brand of fiber-
glass ductboard. The materials were purchased from local commercial vendors. The compositions of the
new fiberglass materials compiled from the Material Safety Data Sheets are summarized in Table 1.
FDL-B contained a permanent (bound) antimicrobial in the coating of the airstream surface. Both
FDL-A and FDL-B were nominally 2.5 cm thick, and were classed as 24.0 kg/m3 (1.5 lb/ft3) in density.
In appearance, these duct liners were very similar, with an uncoated surface intended to be attached to a
rigid duct material and a polymer coated surface intended to be in contact with the moving air in the duct
The fiberglass ductboard material was classed as a 72.0 kg/m3 (4.5 lb/ft3) material, with a reinforced foil
outer coating and a dense but uncoated duct interior surface.
2

-------
Table 1. Composition of Fiberglass Duct Material Tested
Material
Composition
Fiberglass duct liners A
(FDL-A)
> 44 - 98% fiberglass, 1 -18% urea polymer of phenol and formaldehyde or
urea-extended phenol-melamine- formaldehyde resin,
<0.1% formaldehyde
Fiberglass duct liner B
(FDL-B)
82 - 98% fiberglass, 2-18% urea-extended phenol-formaldehyde resin
(cured) or urea-extended phenol-melamine-formaldehyde resin (cured), < 1%
non-woven, Foil-Skrim-Kraft or vinyl facings or vinyl or latex coatings
Fiberglass ductboard
(FGD)
85 - 96% fiberglass wool, 4 - 15% cured binder, < 1% formaldehyde
Artificially Soiling of Duct Materials
All the materials included in the test were artificially soiled. This was accomplished in an aerosol
deposition chamber. Sieved (250/zm) duct dust obtained from a local duct cleaner was used to soil the
samples. Duct material samples were placed around the periphery of the deposition chamber floor, duct
dust was injected using an air injector, mixed in the chamber, and allowed to settle on the test material
samples.
Prior to the study, the uniformity of duct deposition in the chamber was evaluated by comparing
the mass deposited on small test samples placed at different locations on the chamber floor. Acceptable
deposition uniformity was achieved except at the center of the deposition chamber, which was not used
for the studies.
Table 2 presents the precleaning dust loadings for the artificially soiled test materials. The top
row lists the targeted amount of dust for deposition in milligrams per 100 cm2. The amount used was
considered moderately soiled (approximately 100 mg dust /100 cm2). This level was selected with
reference to the 1.0 mg dust / 100 cm2 definition of cleanliness given in NADCA Standard 92-018,
NADCA 92-01 states that a surface may be verified as clean only if the surface is visibly clean and if the
weight of the debris collected by the NADCA vacuum test does not exceed 1.0 mg dust /100 cm2. This
standard is intended only for non-porous surfaces, but its definition of the amount of soil that may remain
in a "cleaned" duct provides the only quantitative benchmark available. (In this study the NADCA
vacuum test was not used to determine dust loading.) Therefore, moderately soiled, as targeted in this
research, was about 100-fold higher than the standard.
The bottom row shows the actual amount of HVAC dust deposited on the surface of the
artificially soiled materials. Glass microscope slides were used as coupons during the artificial soiling
process. The slides were weighed, placed in the chamber, artificially soiled simultaneously with the test
materials, and weighed. The actual dust loading correlated well with the targeted amounts.
Table 2. Amounts of HVAC Dust Deposited on the Test Materials in mg dust /100 cm2

MS-FDL-A
MS-FDL-B
MS-fgd
Target amount
100
100
100
Actual amount, mean ± SD
95 ± 10
95 ± 10
115 ± 10
Note: MS- Moderately Soiled
SD- Standard Deviation
3

-------
All of the artificially soiled material pieces were autoclaved after soiling (but before inoculation
with P. chrysogenum) on a short cycle for better adhesion of the HVAC dust Because the short
autoclave cycle was not sufficient for sterilization, the artificially soiled materials were considered
naturally inoculated as well as inoculated with the test organism.
Selection of Test Organisms and Inoculation
PeniciIlium chrysogenum was selected as the inoculated test organism for these studies. It has
been reported as one of the most frequently isolated molds from the air, dust, and surfaces of indoor
environments9. It has been proposed as a causative agent of allergic alveolitis10. In addition, this
organism has also been isolated from a number of air-conditioning systems in environments where
patients were suffering from allergic disease. Skin challenge testing against P. chrysogenum isolated
from these systems yielded more positives than any of the other organisms isolated11.
The particular P. chrysogenum strain selected for these studies was isolated from a contaminated
building material by RTI and cultivated for use in the laboratory. The culture is being maintained in the
University of Texas Medical Branch Fungus Culture Collection as UTMB3491.
The P. chrysogenum was prepared for inoculation onto the test materials as previously
described3,4. The suspension was nebulized into the aerosol deposition chamber utilizing a six-jet BGI-
Collison nebulizer at 138 kPa (20 psi) for 2 hours and allowed to settle on the duct material pieces.
Surface Sample Collection
To quantify the growth as a function of exposure time, each material was periodically sampled. A
closed-faced filter cassette sampler with a pipette nozzle was used to sample the airstream surface of the
duct material. The sample was obtained from a 10 cm2 surface area as determined by a template. Dust
mass was determined for each sample gravimetrically. Each of the filters was weighed before sampling
and again after sampling and the weight change computed. After the second weighing, the membrane
filters were analyzed using routine plating/counting techniques to determine the colony forming units
(CFUs) per 10 cm2.
Experimental Apparatus and Procedure for Dynamic Growth Experiments
The growth experiments were conducted in the Dynamic Microbial Test Chamber (DMTC)12,
The DMTC is a room-sized test facility designed and constructed to conduct studies on the conditions
and factors that influence biocontaminant emissions and dissemination. The chamber, a cube with inside
dimensions of 2.44 m, was constructed with stainless steel walls and floor and an acrylic drop-in ceiling.
Temperature (18 - 32°C) and relative humidity control (55 - 95% RH) are provided through an air
handler, conventional ductwork, and ceiling diffusers with an air circulation rate between 1.4 and 4.8
m3/min
Of the conditions that must be reproduced in a duct-like test, temperature, humidity, dust
(nutrient) loading, and air velocity over the duct material, temperature has been shown to be of secondary
importance. Lowering the temperature slowed the beginning of a fungal growth response but did not
change the eventual amount of measured growth3. The bulk of the previous tests were conducted at
ambient temperature, as were the dynamic tests. Humidity and dust loading are both known to be
important, while air flow rate has not been investigated.
To provide these conditions in the DMTC, the chamber was adapted to contain eight miniducts.
An artist's rendition of the DMTC containing the miniduct apparatus is shown in Figure 1. The blower
forces the conditioned DMTC air into a High Efficiency Particulate Air (HEPA) filter, from which the air
for the eight miniducts is obtained. The miniduct channel design was chosen to limit the total amount of
air required for a single test, allowing multiple tests to be run simultaneously.
4

-------
Figure 2 shows an expanded view of the miniduct apparatus. The upper part of the figure shows
the blower, the duct leading to the HEPA filter unit, and an individual duct delivering conditioned air to a
miniduct. The bottom part of the figure shows an expanded view of one of the miniducts.
Recommended air velocities in ventilation system ducts are in the range of 2.5 - 4.6 m/s (500 -
900 ft/min)13. A velocity of 2.5 m/s (500 ft/min) was chosen for the present study as a reasonable
velocity that could be provided to the eight miniducts by a small fan. Higher velocities might also lead to
unrealistically high surface velocities and turbulence in the narrow channels. Achieving 2.5 m/s in a
miniduct, whose flow channel is 2.5 cm high and 40 cm wide, required a volumetric flow rate of 0.025
m3/s (53 ft3/min). The dampers shown in Figure 2 were used to obtain the desired flow rate to each
miniduct. To reduce flow development and edge effects within the area of the duct material samples, the
material samples were placed in a recess in the middle of the miniduct flow area with the upper surface of
the material 2.5 cm below the channel top. A 10 cm flow development buffer space was provided
upstream, and 5 cm buffers were provided on the sides and downstream.
Miniduct flow rates were routinely measured downstream of the samples at the channel center
point. Flow distribution at the upstream end of the channel was enhanced with screens to achieve a
horizontally uniform velocity profile. The inoculated duct material samples were placed in the miniducts
with the upper surface flush with the bottom of the flow channel, allowing the filtered and conditioned air
to flow across the material just as it would when in a duct.
After artificial soiling and inoculation, the 30.5 x 91.4 cm (1 x 3 ft) pieces of test material were
placed in the miniducts. For all experiments the temperature and relative humidity throughout the mini-
ducts were at 23 .5°C and 94% RH The air velocity through the miniducts was 2.5 m/s (500 ft/min).
Duplicate surface samples were collected on each of the test days. For the FDL-A and FDL-B
experiments, duplicate surface samples were collected the first, second, third, fourth, sixth, and eighth
weeks prior to cleaning. The materials were then contact vacuumed and postcleaning surface samples
were collected the first through fourth and sixth weeks. In the FGD experiment that schedule was
amended slightly. Duplicate surface samples were collected the first through fourth weeks prior to
cleaning. The materials were then contact vacuumed and postcleaning surface samples were collected
the first through sixth weeks, and again the twelfth and thirteenth weeks.
Duct Material Cleaning
Once mature growth was reached and quantitatively evaluated, the duct material was cleaned in
place using a Minuteman Model C82906-03 HEPA vacuum cleaner operating at 2.7 m3/min (95 ft3/min).
Prior to cleaning, the bag was weighed. The combination floor tool was used. The tool measures 14x4
cm and one side houses a single-edge brush. A thorough cleaning procedure was developed and followed
rigorously for all pieces. The cleaning pattern consisted of four passes over each surface crosswise and
four passes lengthwise. The first pass was across the width of a piece from left to right. The second pass
was back across the same width from right to left. This was repeated for passes three and four. The
brush was then moved down the piece by a brush-width, and this pattern was repeated down the length of
the whole piece of the test material. Once the entire piece had been cleaned crosswise, the procedure was
repeated for the length of the material. During cleaning, the material was inspected and any light spots
cleaned. Relative to field duct cleaning observed by the authors, this constitutes extremely thorough
cleaning. Care was taken not to abrade the surface of the materials. The combination floor tool was
decontaminated with 70% ethanol between vacuuming the different materials.
RESULTS
Precleaning
Table 3 presents the levels of growth attained on the three materials before the contact vacuuming
protocol was performed. The data are expressed as the log increase between the initial levels measured
on week 1 and the levels measured on weeks 4 and 8.
5

-------
As discussed previously, although the materials were autoclaved before inoculation with P.
chrysogenum, only short-exposure autoclaving was performed to promote HVAC dust adhesion to the
artificially soiled samples. Killing spores by autoclaving (sterilization) requires longer exposure times and
higher temperatures; therefore, a certain natural inoculation from the spores dormant on the surface was
anticipated. The dominant organism that grew from this natural contamination was A. versicolor, whose
growth on fiberglass materials is well documented14. No other organism resulted from the natural
contamination in detectable levels. A. versicolor is of particular concern because it is a toxigenic fungus,
and its growth was documented as part of the experiment. The data for the two test organisms, P.
chrysogenum and A. versicolor, are reported separately.
Table 3. Log Difference in Levels of CFUs for P. chrysogenum and A. versicolor between Week 4
or Week 8 and Week 1
Material
Week 4
Week 8

P. chrysogenum
A. versicolor
P. chrysogenum
A. versicolor
MS-FDL-A
2.3 ±0.1
1.2 ± 0.1
2.9 ±0.2
1.9 ±0.3
MS-FDL-B
2.4 ± 0.2
3.5 ±0.2
2.8 ±0.1
4.5 ±0.2
MS-FGD
3.1 ±0.2
2.8 ±0.7
ND
ND
ND = Not Done
As can be seen from Table 3, after 4 weeks the two artificially soiled fiberglass duct liners (MS-
FDL-A and MS-FDL-B) had sustained at least a 1 log increase for both organisms. The levels of P.
chrysogenum were similar for both duct liners, but the levels of A. versicolor were considerably higher on
the MS-FDL-B than the MS-FDL-A. Levels of both test organisms increased approximately 3 logs on
the fiberglass duct (FGD).
As discussed previously, cleaning by contact vacuuming was performed on MS-FDL-A and MS-
FDL-B on the eighth week of the study. The levels of both of the organisms continued to increase from
the fourth to the eight week; although, with the exception of A. versicolor on FDL-B, the increases were
small. Previous experiments using static chambers have demonstrated similar or slightly lower increases
for similar time spans and conditions3,15. That is, the addition of air flow in the dynamic experiments may
have yielded slightly higher levels of growth than were found in static experiments. Because growth
between weeks 4 and 8 was small, the FGD experiment did not go beyond week 4 before cleaning.
Impact of Contact Vacuuming on Dust Mass
The total weight of the dust removed by contact vacuuming from each of the materials is shown
in Table 4. The weights were obtained by determining the dust mass in the vacuum cleaner bag after
cleaning
Table 4. Dust Mass From Dust Removed by Contact Vacuuming in mg dust /100 cm2

MS-FDL-A
MS-FDL-B
MS-FGD
Amount removed by contact vacuuming
75
89
127
For the moderately soiled FDL-A and -B, the total mass removed was 75 and 89 mg dust /100
cm2, respectively. Based on the 100 mg / 100 cm2 targeted amount, this was 75 and 89% of the amount
6

-------
initially deposited. However, the moderately soiled FGD yielded 127 mg dust /100 cm2; that is,
approximately 25% more was removed than was added initially.
Table 5 shows a comparison of the amount of dust measured on the test materials both before and
after contact vacuuming. These values are the means and standard deviations of the replicate dust mass
measurements collected by surface sampling. Of particular interest is that the difference between the
precleaning and postcleaning numbers is not equal to the total mass removed by contact vacuuming as
shown in Table 4. Visual examination of the surface samples showed that a notable proportion of the
sample content was fibers removed from the surface during the surface sample collection process.
Examination of the contact vacuuming dust in the vacuum cleaner bag also showed similar fiber content.
The results demonstrate that: 1) evaluating total dust on a duct surface is a difficult measurement, and 2)
there appears to be continued fiber loss from the surface of the fiberglass materials with repeated
vacuuming.
Table 5, Dust Levels Determined by Surface Sampling in mg dust /100 cm2

MS-FDL-A (n=8)
MS-FDL-B (n=8)
MS-FGD (n=4)
Before contact vacuuming
73 ±31
100 ± 28
196 ± 17
After contact vacuuming
42 ± 16
62 ±35
133 ±34
Impact of Contact Vacuuming on Microbial Load
Tables 6 and 7 show the results of cleaning by contact vacuuming as the percent reduction on the
fungal load for the three different test materials. The tables show the comparison of the precleaning and
immediate postcleaning levels for both of the test organisms. Table 6 presents the results for P,
chrysogenum, which was inoculated onto the surface of the materials at the beginning of the experiments;
and Table 7 shows the results for A. versicolor, the natural inoculum.
As can be seen in the tables, there were large decreases in both P. chrysogenum and A. versicolor
levels immediately postcleaning. This was probably due to a high percentage of the fungal growth being
in or on the dust that was removed by contact vacuuming, MS-FDL-A, MS-FDL-B, and MS-FGD all
decreased by at least 95%, except less of a decrease was seen for the P. chrysogenum on MS-FGD.
Table 6. Percent Reduction of P. chrysogenum on the Surface of the Test Material Immediately after
Contact Vacuuming	

MS-FDL-A
MS-FDL-B
MS-FGD
Before contact vacuuming,CFU/cm2
46,000
61,000
410,000
After contact vacuuming, CFU/cm2
680
200
48,000
% Reduction
98.5
99.7
88.3
Table 7. Percent Reduction of A. versicolor on the Surface of the Test Material Immediately after
Contact Vacuuming	

MS-FDL-A
MS-FDL-B
MS-FGD
Before contact vacuuming,CFU/cm2
25,000
1,800,000
58,000
After contact vacuuming,CFU/cm2
<25
1,300
<2,500
% Reduction
99.9
99.9
95.7
7

-------
Impact of Cleaning on Fungal Growth	*
Figures 3 and 4 show the levels of P. chrysogenum and A, versicolor, respectively, immediately
post cleaning, 6 weeks post cleaning, and 12 weeks postcleaning. All have been normalized to the initial
precleaning (first week) levels for that particular test material. The black bars represent the immediately
postcleaning levels. The gray bars show levels at the 6 weeks postcleaning, and the white the 12 week
levels. The figures show that on all materials regrowth occurred for both organisms.
The levels of organisms isolated from the immediately postcleaning samples varied greatly
between materials and organisms. P. chrysogenum was isolated from all the samples, but on FDL-A A.
versi-color levels were below the detection limit. As seen in Figure 3, by the 6 weeks postcleaning
measurement, the level of P. chrysogenum had rebounded from the immediate postcleaning levels for all
samples. FDL-A and FDL-B showed increases of at least 2 logs. FGD showed a smaller increase
probably because the levels on that material had not been reduced as much by cleaning. The 12 weeks
postcleaning levels showed continued increases in CFUs on most materials. As discussed earlier, the
experiments with FDL-A and -B were discontinued at 6 weeks so no data are available at the twelfth
week.
Figure 4 shows the comparison of the immediate and 6 and 12 weeks (where available)
postcleaning levels of A. versicolor normalized to the initial preclean (first week) levels for each of the
test materials. As discussed earlier, A. versicolor was not inoculated onto the test materials but was
considered a natural inoculum. Generally, the overall results were similar to those seen and discussed
above for P. chrysogenum.
DISCUSSION AND CONCLUSIONS
Surface cleaning by contact vacuuming was able to remove notable amounts of dust from the
surface of the artificially soiled duct materials. Large amounts of fibers were removed from the fiberglass
materials. Fiber shedding was evident with the new as well as the used duct liners. Although the amount
of shedding was not quantified in this study, visual inspection suggested that a fair percentage of a surface
sample was composed of fibers or fragments of fibers.
Cleaning artificially soiled materials caused noticeable reductions for both of the organisms in the
immediate postcleaning period. The surface dirt was readily removed from the artificially soiled materials
along with the organisms colonizing the dirt.
In the longer term, the levels of both A. versicolor and P. chrysogenum recovered to precleaning
levels within 6 weeks and, where studied, growth continued for both organisms over the entire 3 months.
Therefore, mechanical cleaning by contact vacuuming, at best, was able to only temporarily reduce the
surface fungal load. The current guideline to discard contaminated or potentially contaminated materials
should be followed. This may have been because cleaning is a surface treatment, and the organisms were
growing both on the surface dirt and deep in the material.
These data showed that, at least for the conditions used in this study, a bound antimicrobial had
no effect on fungal growth on soiled and cleaned contaminated fiberglass duct liner. There was no
difference between regrowth on the fiberglass duct liner with or without the bound antimicrobial. This
finding supports the results of our previous static chamber studies3. Additional work is needed to
determine the benefit of using a biocide or antimicrobial as part of the duct cleaning process for
prevention and remediation of fungal growth on the various duct materials.
ACKNOWLEDGMENTS
The authors greatly appreciate the technical assistance of Eric A. Myers, Environmental Biologist
II, in collecting the data. Additional thanks for technical assistance go to Tricia D. Webber, Environmen-
tal Biologist I. Both are from the Research Triangle Institute.
8

-------
REFERENCES
1.	Morey, P. R. and C. M. Williams. 1991. is Porous Insulation Inside an HVAC System Compati-
ble with a Healthy Building? Healthy Buildings: Proceedings of IAQ '91. ASHRAE, Atlanta,
GA, pp. 128-135.
2.	Chang, J.C.S., K.K. Foarde, and D.W. VanOsdell. 1996. Assessment of Fungal (Penicillium
chrysogenum) Growth on Three HVAC Duct Materials. Envir. Inter., 22 (4):425-431.
3.	Chang, J.C.S., K.K. Foarde, and D.W. VanOsdell. 1995 Growth Evaluation of Fungi
(Penicillium and Aspergillus spp.) On Ceiling Tiles. Atmospheric Environment. 29:2331-2337.
4.	Foarde, K.K., D.W. VanOsdell, and J.C.S. Chang. 1996. Evaluation of Fungal Growth on
Fiberglass Duct Materials for Various Moisture, Soil, Use, and Temperature Conditions. Indoor
Air. 6:83-92.
5.	Foarde, K., D. VanOsdell, and J. Chang. 1996. Static Chamber Method for Evaluating the
Ability of Indoor Materials to Support Microbial Growth. ASTM STP 128 6 Methods for
Characterizing Indoor Sources and Sinks, pp. 87-97.
6.	NADCA. 1996. Understanding Microbial Contamination in HVAC Systems. National Air Duct
Cleaners Association, Washington, DC.
7.	N AIMA. 1993. Cleaning Fibrous Glass Insulated A ir Duct Systems, Recommended Practice.
North American Insulation Manufacturers Association, Alexandria, VA.
8.	NADCA 1992-01. 1992. Mechanical Cleaning of Non-porous Air Conveyance System Compo-
nents. National Air Duct Cleaners Association, Washington, DC, p. 11.
9.	Hunter, C.A. and R.G. Lea. 1995. The airborne ftingal population of representative British
homes. Health Implications of Fungi in Indoor Environments, R.A. Samson et al., eds. Air
Quality Monographs, 2:141-153.
10.	Fergusson, R J., L.R.J. Milne, and G.K. Crompton. 1984. Penicillium allergic alveolitis: faulty
installation of central heating. Thorax. 39:294-298.
11.	Schata, M., W. Jorde, J.H. Elixman, and H.F. Linskens. 1989. Allergies to molds caused by fungal
spores in air conditioning equipment. Environment International, 15:177-179.
12.	VanOsdell, D., K. Foarde, and J. Chang. 1996. Design and Operation of a Dynamic Test
Chamber for Measurement of Biocontaminant Pollutant Emission and Control. ASTM STP 1286
Methods for Characterizing Indoor Sources and Sinks, pp. 44-57.
13.	ASHRAE. 1992. Air Distribution Design for Small Heating and Cooling Systems. HVAC
Systems and Equipment, ASHRAE, Atlanta, GA, pp. 9.4 - 9.5.
14.	Ezeonu, I.M., J.A. Noble, R.B. Simmons, D.L. Price, S.A. Crow, and D.G. Ahearn. 1994. Effect
of relative humidity on fungal colonization of fiberglass insulation, Applied and Environmental
Microbiology. 60(6):2149-2151.
15.	Foarde, K. K„ D. W. VanOsdell, and J. C. S. Chang. 1995. Susceptibility of Fiberglass Duct
Liner to Fungal (Penicillium chrysogenum) Growth. Proceedings of Engineering Solutions to
Indoor Air Quality Problems, EPA and Air and Waste Management Association, Research
Triangle Park, NC, July.

-------
Figure 1. Drawing of Dynamic Chamber with "Miniduet" Apparatus.

-------
Dynamic Test Chamber
Flow Control
Damper 		
5 £ _v_ i'.Y J, _v_ j-.
Mixed Air
y Temperature and
\^7 Humidity Controlled
To other
Miniducts

1 of 8 Miniducts
Entrance
Transition
Window
Gaskets to Seal^
Flow Channel
Duct Material Sample,
30.5 x 91.4 cm
0.6 - 2.4 m3/min
Volumetric Flow
I -4 m/s Velocity
in Flow Channel
. ^
2.5 cm Flow
Channel
Figure 2.Diagram of Miniduct Apparatus.
11

-------
6,00
immediately postcleaning
~ 6 weeks postcleaning
012 weeks postcleaning
if
ilhi
u 1.00
0.00
moderately soiled FDL-A
moderately soiled FDL-B
moderately soiled FGD
Figure 3. Immediately postcleaning, 6 weeks postcleaning, and 12 weeks postcleaning levels
of P. chrysogenum.
QJ
£
D
PL,
O
00
o
X
8
>
D
u*
U
m
o
j ¦ immediately postcleaning
! E2 6 weeks postcleaning
l_l l A WCCNb pUMUCCtmilg
moderately soiled FDL-A moderately soiled FDL-B moderately soiled FGD
! Figure 4. Immediately postcleaning, 6 weeks postcleaning, and 12 weeks postcleaning levels
of A. versicolor.
i	* Below Detection Limit
12

-------
NRMRL-RTP- P-314 (Please recd^^^^^lecornpU
1, REPORT NO. 2.
EPA 600/A-98-047
3
4. TITLE AND SUBTITLE
Effectiveness of Vacuum Cleaning on Fung ally
Contaminated Duct Materials
S. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHORts) K. Foarde and D. VanOsdell (RTI); and
J. Chang (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME ANO AODRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR822642
12. SPONSORING AGENCY NAME ANO ADDRESS
EPA, Office of Research and Development
Mr Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 3-7/97
14. SPONSORING AGENCY CODE
EPA/600/13
16.supplementarynotesAppcd project 0fficer is John C.S. Chang, Mail Drop 54, 919/
541-3747. Presented at EPA / AW MA Conference, Engineering Solutions to 1AQ Pro-
blems. RTP. NC. 7/21-23/97.
is. abstract pap6r gives results of research to: determine, under constant temper-
ature, relative humidity, and air flow test conditions, whether fungal spore levels
on heating, ventilating, and air-conditioning (HVAC) duct material surfaces could be
reduced substantially by thorough vacuum cleaning; and evaluate whether subsequent
fungal growth would be limited or contained by a single mechanical cleaning treat-
ment. Three fiberglass duct materials were tested. All were soiled artificially. The
results showed that notable amounts of surface dust were removed, and surface
spore levels could be reduced in the short term on all materials by vacuuming. How-
ever, regrowth occurred within 6-12 weeks. Ventilation system materials are of
particular significance as potential microbial contamination sources because of their
potential to rapidly spread contamination throughout a building. Portions of ventila-
tion systems near cooling coils and drain pans are known to be exposed to high mois-
ture levels for extended periods, and fibrous duct insulation materials are known to
have become sources of microbial contamination in some buildings. Cleaning has
been suggested as a possible strategy for preventing microbial contamination, as
well as for remediating already contaminated materials. However, the efficacy of
cleaning has not been determined.
17. key words and document analysis
a. DESCRIPTORS
b.tOENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Ducts
Vacuum Cleaners Heating Equipment
Cleaning Ventilation
Fungi Air Conditioning
Spores Dust
Fiberglass Reinforced Plastics
Pollution Control
Stationary Sources
13 B 13K
13 G 13 A
13 H
06 C
11G
11D, 11F
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage/
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)

-------