Successful Mold Growth Remediation in HVAC Systems
- By Paul Buckmaster, M.S., CIH, CIEC
- Jan 04, 2008
The summer of 2003 was one of the
wettest in Maryland history. Later
that fall, an indoor environmental
quality (IEQ) investigation in a leased
facility revealed mold growth in the air
handing units (AHUs) and main supply
duct of all 12 of the building’s HVAC systems.
During the next two months, we
worked with occupants, our leasing agent,
the building owner, and his contractor to
successfully remediate the mold growth.
We decided to assess mechanical hygiene
in 15 other buildings.
This article focuses on one building and
describes the strategies and lessons learned
by a team of facility, engineering, and
industrial hygiene staff during the remediation
project completed in May 2004.
HVAC System Description
The building is a four-story concrete slab
and column structure built in the mid-
1950s with a total area of 1.5 million
square feet. Each floor is served by eight
HVAC systems. Renovation has been
ongoing since 1997, and 17 of the 32 systems
have been replaced. The original
design consisted of constant-volume
AHUs with electric steam generators to
provide space humidification. These were
abandoned in the 1980s because of maintenance
problems. Fiberglass insulation
(FGI) was used extensively inside AHU
cabinets and on air stream surfaces in both
supply and return ductwork.
TABLE 1. MYCOMETER-TEST CATEGORIES
The level of mold is not above normal background levels.
25 < MV < 450
The level of mold is above normal background levels. This may
be due to high concentrations of spores in dust deposits or
may in some cases indicate the presence of old mold growth.
The level of mold is high above normal background level due
to active mold growth.
During the renovation, existing systems
are replaced with variable air volume
AHUs. Duct static pressure is monitored
and controls a variable frequency drive on
the supply fan. Humidification is provided
by a steam-to-steam heat exchanger
located between the cooling coil and fan
inlet. Controls include a humidistat
located in the return air duct and a high
limit sensor in the supply discharge. New
systems were designed with several features
to improve mechanical hygiene.
Condensate pans were constructed of
stainless steel and sloped to improve
drainage. Particulate filters with a Minimum
Efficiency Reporting Value
(MERV) of 12 were installed to keep coils
and ductwork cleaner and improve energy
efficiency. AHU cabinets have doublewall
construction to eliminate exposed
FGI; however, engineers required FGI to
be installed in the initial 30 feet of supply
and return ductwork and downstream of
fan-powered VAV boxes to reduce noise.
Mold spores are ubiquitous, and one
should expect to find them in the indoor
air and on surfaces inside the building
envelope. In well-maintained buildings
with effective particulate filtration, we
typically find that:
¦ indoor levels are less than 5 percent
of outdoor levels,
¦ rank order of outdoor and indoor
taxa are similar, and
¦ concentrations of indicator species
are absent or very low.
Air sampling has resulted in false negatives
in several building investigations
when there was extensive mold growth in
the HVAC system. Because of this, we did
not use air sampling as part of the methodology.
We adopted a standard of care that
stipulated mold growth was not acceptable
and used visual inspection and surface
sampling to determine conformance.
Historically, we used vacuum pumps
and MCEFs to collect surface samples and
used a contract laboratory to culture samples
to determine total fungal counts and
taxa. This worked well during the investigation,
but the cycle time was too long for
post-remediation verification. After the
first project, we began to look for other
methods to characterize mold growth. We
learned about the Mycometer®-Test from
the conference proceedings1 and contacted
the distributor for a demonstration. After
reviewing the literature and talking to
other users, we purchased the instrument.
The Mycometer®-Test is a patented
method that allows the quantification of
surface mold in less than one hour at a
cost of about $25 per sample. The test
uses the enzyme activity found in the
spores and mycelium of all mold. Activity
is measured using an enzyme substrate
that releases a fluorescent compound
when excited by UV light. Fluorescence is
measured by a portable fluorometer that
displays a unitless mycometer value (MV),
which is proportionate to the amount of
biomass on the surface. Later research2
showed a correlation between this enzyme
and ergosterol, the gold standard for characterizing
fungal biomass on surfaces.
Establishing categories that defined
fungal biomass on a surface was an important
part of method development. To do
this, 102 samples were collected from clean
surfaces in well-maintained buildings.
These were found to have an MV < 25.
This was defined as Category A and represents
normal fungal ecology. Next, 127
samples were collected from visibly dusty
surfaces with moderate to heavy spore
deposition. About 96 percent of these samples
had an MV < 450. Category B was
defined as an MV > 25 but < 450. Samples
collected from surfaces with active mold
growth had MVs > 450 and were defined as
Category C. Categories and interpretive
guidelines are found in Table 1.
Industrial hygienists and facility technicians
visually inspected all 32 HVAC systems.
IHs collected swab samples from
cooling coils, condensate pans, and supply
ductwork. There was no visible growth in
any location, and samples from coils and
pans were found to be Category A. However,
samples from ductwork revealed
problems. Results and descriptive statistics
for these are presented in Table 2.
TABLE 2. MYCOMETER RESULTS FOR SUPPLY DUCTWORK
7 ( 41%)
In existing systems, MVs ranged from
9 to 138 with a mean of 32. Fifty-nine percent
of the samples were Category A, and
41 percent were Category B. In contrast,
MVs from new systems (less than six years
old) ranged from 3 to 12,000 with a mean
of 1,228. Sixty percent of the systems were
Category A, 13 percent were Category B,
and 27 percent were Category C. Active
growth was found in ductwork in four systems,
where MVs ranged from 491 to
12,000 with a mean of 4,467. In each,
growth was found on FGI within 20 feet
of the fan discharge.
Risk communication is a critical part of
any mold remediation project. We crafted
a good strategy for the first project and
carried that forward in this project. Initial
notification began with an e-mail to
affected supervisors stating we’d found
mold growth in the HVAC system serving
their space and would:
¦ clean that system and return it to a
normal fungal ecology,
¦ clean the office space to remove
mold spores and other particulate,
¦ provide an occupational medicine
consultation to any occupant with
symptoms from mold exposure, and
¦ present a series of town meetings
to inform occupants about mold, its
health effects, and our plans to clean the
systems and affected rooms.
After 24 hours, a similar notification
was e-mailed to affected employees. Town
meetings began within a week, with 14
held over the next five days. We also developed
a Web site with streaming video
about mold and its health effects, Power-
Point presentations from the town meetings,
copies of supervisor and occupant
notices, and the proposed cleaning
schedule for the HVAC systems and associated
rooms. The strategy helped to control
outrage and rumors so well that more
facility and health professionals than occupants
attended the last two town meetings.
Scope of Work
Facilities and contacting specialists
teamed with the industrial hygienist to
develop a scope of work (SOW) for the
project. The National Air Duct Cleaners
Association (NADCA) published two documents
dealing with the cleaning and
restoration of HVAC systems3,4 that were
used as the basis for an SOW that would
cover multiple projects in multiple buildings
over a three-year period. We were
concerned about the knowledge and skills
of firms bidding on the contract proposal,
so in addition to the usual requirements
for bonding, insurance, and previous work
experience, we required a “competent
person” on the job site at all times work
was being performed. This person needed
at least one of the following credentials
from NADCA or the American Indoor
Air Quality Council (AmIAQC):
¦ Certified Air Systems Cleaning
¦ Certified Ventilation System Mold
¦ Certified Microbial Remediation
¦ Certified Microbial Remediator
Five companies responded to a
Request for Proposal, and the three
judged to be most qualified were
selected to work on our campus. Bidders’
past safety performance was considered
in the selection.
Prior to start-up, we met the contractor to
discuss its safety plan and the project
scope, which included:
1. Remediating mold growth on FGI
in the initial 20 feet of supply duct in four
2. Cleaning the coils, condensate pan,
and interior surfaces of four AHU cabinets.
3. Cleaning 200,000 square feet of
office space, including HEPA-vacuuming
carpet and chairs and damp wiping nonporous
We discussed security issues, cleaning
methods, containment strategies, differential
pressure requirements, post-remediation
verification, the role of the competent
person, and how the government
would monitor contractor performance.
Work was scheduled after hours and
weekends during April and May 2004.
Facilities Services managed the project,
and a Certified Industrial Hygienist served
as the safety and health consultant. Containment
was designed to isolate the AHU
and affected supply duct from the rest of
the HVAC system and occupied space. To
do this, the contractor installed a HEPAfiltered
negative air machine at one of the
AHU’s access doors. The return air
damper was closed and the outdoor air
damper was opened to provide make-up
air. A 24-inch by 24-inch access panel was
installed in the supply duct downstream of
the mold growth, and foam pillows were
used to zone off. A floor-to-ceiling plastic
enclosure was erected and used to enter
the supply duct for cleaning. A second
access door was used to enter and clean the
AHU and components.
The negative air machine was operated
to maintain a differential pressure of at least
-0.02 inches water gauge. The contractor
continuously monitored this requirement.
The industrial hygienist inspected the containment
each day and issued a notice to
proceed. A particle counter was used to determine the effectiveness of the HEPAfiltered
negative air unit and to monitor
background levels of particulate.
There was some concern that fiberglass
insulation could not be adequately
cleaned. One study5 showed it was possible
to remove approximately 90 percent of the
mold using mechanical methods. Another6
showed cleaned liners sealed with an EPA
registered antifungal protective coating
(AFPC) remained free of mold growth
after 10 years. Using this data, we developed
the following decision logic:
¦ Eroded FGI would be removed,
and the metal duct would be cleaned
and treated with an AFPC to lock
down residual FGI.
¦ FGI with an MV > 1000 would be
removed, and the metal duct would be
cleaned and treated with an AFPC to
lock down residual FGI.
¦ FGI with an MV < 1000 would be
cleaned and coated with an AFPC.
¦ Metal duct would be cleaned only.
HEPA-filtered vacuums were used to
remove particulate from surfaces inside
the AHU and supply ductwork. Heat
exchange coils and condensate pans were
cleaned using detergent and high-pressure
water because there was no exposed FGI.
After cleaning, the industrial hygienist
completed a visual inspection and collected
samples to determine the residual
fungal biomass. Surfaces with an MV < 25
met the post-remediation verification criteria
and were judged to be clean.
Causal Factor Analysis
No remediation is successful until the
moisture source is identified and corrected.
Despite manufacturers’ claims,
many studies have shown that FGI supports
mold growth.7 The material is
hygroscopic and retains water. Over time,
enough mold spores and nutrients are
deposited on the surface that growth will
occur when adequate moisture is present.
The problem is that moisture exists in
any HVAC system. During the cooling
season, chilled water coils produce condensation
and high-humidity air, and face
velocities greater than 400 feet per minute
can cause water stripping. The new AHUs
had an additional moisture source. Steam
humidification was condensing at the fan
inlet and wetting the FGI surface.
The engineer’s investigation revealed
several other moisture problems. AHUs
were installed directly on the floor slab,
and there was not enough clearance to
install a trap with the proper stem height.
As a result, condensate pans did not drain
properly and were filled with 30 to 40 gallons
of water during the cooling season.
This water aerosolized and deposited
downstream of the fan onto the ductwork.
Installation of AHUs also was problematic.
Cabinets were designed in three
sections that were supposed to be assembled
using a come-along. This was never
done, and sections leaked excessively.
Mechanical rooms were not air-conditioned,
so warm, humid air leaked into the
cabinet and condensed on metal surfaces.
The project team initiated the following
actions to address causal factors and improve mechanical hygiene:
1. Design standards were modified to
prohibit FGI from air stream surfaces in
the first 20 feet of supply and return ductwork.
Insulation is permitted downstream
of fan-powered VAV boxes because
Mycometer sampling has shown these
surfaces remained free of mold growth
after four years in service.
2. Humidifiers were taken out of service
immediately. Our health care providers
developed a fact sheet that discussed steps
to take during the winter months to deal
with low humidity at work and home.
3. Condensate traps were redesigned.
Contractors penetrated the floor slab to
allow additional clearance to increase
stem height. Later AHUs were installed
on a 14-inch-high housekeeping pad to
provide adequate clearance.
4. AHU construction was tightened
up. Contractors installed additional gasket
material between sections and torqued
corner bolts to manufacturers’ specifications.
Supply terminals were installed in
the main duct to introduce conditioned
air into the mechanical room.
5. Construction IAQ management
plans are now required for all major renovation
projects. Contractors are required
to protect ductwork and building materials
from moisture and dirt.
6. Construction standards were revised
to include provisions for duct access openings
to permit inspection and cleaning.
7. Mechanical hygiene assessments are
conducted on all new HVAC systems.
Surfaces with MVs > 25 units are required
to be cleaned by the contractor prior to
8. Operation and maintenance standards
were modified to conform with
ASHRAE Standard 62.1. We are currently
looking adopting ASHRAE Standard
180P, Standard Practice for the Inspection
and Maintenance of Commercial Building
HVAC Systems, as the standard of care.
9. Particulate filtration has been
upgraded to MERV 13 to conform to the
Leadership in Energy and Environmental
Design (LEED) criteria.
We continue using the Mycometer to
characterize fungal biomass in new HVAC
systems. Ninety percent of the surfaces
remain free of mold growth (i.e., MVs <
100) after three years.
1. Reeslev, M. and Miller, M.: The
Mycometer Test: A New Rapid Method
for Detection and Quantification of
Mould in Buildings, Healthy Buildings
2. Reeslev, M., Miller, M., and Nielson,
K.F.: Quantifying Mold Biomass on
Gypsum Board: Comparison of Ergosterol
as Mold Biomass Parameters, Applied and
Environmental Microbiology, July 2003,
3. General Specification for the Cleaning
of Commercial HVAC Systems, National
Air Duct Cleaners Association, 2001.
4. Assessment, Cleaning and Restoration
of HVAC Systems, National Air Duct
Cleaners Association, 2003.
5. Krause, J.D., and Hammad, Y.Y.: Measuring the Efficacy of Mold Remediation on Contaminated Ductwork, Indoor Air 2002 Proceedings.
6. Yang, C.S., and Ellringer, P.J.: Evaluation of Treating and Coating HVAC Fibrous Glass Liners for Controlling Fungal Colonization and Amplification, ASHRAE 1996 Indoor Air Quality Conference Proceedings.
7. Van Loo, J.M., Robbins, C., and Swenson, L.: Growth of Mold on Fiberglass Insulation Building Materials, Journal of Occupational and Environmental Hygiene, June 2004, pp. 349-354.
This article originally appeared in the January 2008 issue of Occupational Health & Safety.