HVAC Systems as Emission Sources Affecting Indoor Air Quality a Critical Review


EPA- 600 / R- 9 5- 014
February 1995

HVAC SYSTEMS AS EMISSION SOURCES
AFFECTING INDOOR AIR QUALITY:
A CRITICAL REVIEW

By

Stuart Batterman

University of Michigan
Ann Arbor, MI 48109

and

Harriet Burge

Harvard University
Boston, MA 02115

EPA Cooperative Agreement CR815391-01-0
(American Society of Heating, Refrigerating and Air-Conditioning Engineers)

EPA Project Officer: Russell N. Kulp

Air and Energy Engineering Research laboratory
Research Triangle Park, NC 27711

Prepared for:

U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460

-------
EPA REVIEW NOTICE

This report has been reviewed by the U.S. Environmental Protection Agency, and
IpprovL (or publication. Approval does not signify thai the contents necessarily

reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document Is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

ABSTRACT

This study evaluates heating, ventilating and air conditioning (HVAC) systems as contaminant
emission sources that affect indoor air quality (IAQ). Various literature sources and methods
for characterizing HVAC emission sources are reviewed. Available methods include in situ
tests, longitudinal and cross-sectional studies, and laboratory studies. A critique of the
literature reveals that few studies are well-controlled, comprehensive and quantitative.
Significant gaps in the data are highlighted and procedures are suggested to improve the
characterization of bioaerosol and volatile organic compound (VOC) emission sources. Based
on the available literature, several HVAC components are cited fairly frequently as emission
sources, and there is broad agreement regarding their significance. The components include
biological growth and bioaerosol generation in the presence of moisture provided by air
washers and other recirculating water systems, poor control of humidity, poorly designed
humidifying systems, poorly maintained cooling coils and drip pans. IAQ problems appear to
be exacerbated by dust accumulation and by the presence of fibrous insulation. Other
problems include entrainment, migration, and infiltration of indoor and outdoor contaminants
that are distributed to indoor spaces by the HVAC system. The importance of good design and
operation of HVAC systems, including the appropriate placement and maintenance of air
intakes, building pressurization, and local exhaust in source areas, is also well accepted. More
limited data implicate dust (resulting from inadequate filtration and maintenance of filters) as a
sink and secondary source for VOCs.

-------
CONTENTS

Page

ABSTRACT...................................					....ii

TABLES....			 			................			iv

INTRODUCTION

Data Published in the Peer -reviewed Journal literature			2

Data Published in Books and Proeeedinp					„„3

Reviews

APPROACHES TO IDENTIFYING AND CHARACTERIZING HVAC SOURCES......4

In Situ Component Studies...........									5

Longitudinal Studies

Cross-secLional S^tud^es...	................a..........................."1 "1.

Laboratory Tests								11

Summary					—........		—		12

HVAC SYSTEMS AND IAQ							13

Intrinsic Emission Sources					13

Seals, caulks, etc.	.....................—				13

Fibers								13

Metal degradation products										 15

Lubricating oils, etc							15

Ozone......							15

Emission Sources Resulting from Contamination			15

Dust							...15

Other organic debris..............	.......							18

Growth of microorganisms.....							................ 18

Cooling coils						19

Drain pans, drains, traps and sumps							. 19

Filters			—							......19

Insulation.		20

Duct surfaces.................	.......							21

Flemi i lis		21

Humidifiers and evaporative coolers..............					21

Cooling towers.............					...21

VOC sinks	22

Cleaning compounds and biocides	..............			22

Boiler steam							22

Design/Operational Effects on IAQ			22

Entrainment and re-entrainment	...			22

Rotary heat exchangers										23

Building pressurization.................	....					23

Transport...												24

Climate control						24

Ventilation and air exchange.........							24

Cleaning										25

SENSORY STUDIES												....25

DISCUSSION								26

Important Gaps in the Available Data					27

Interpretation of Available Studies									28

Suggested Strategies									30

iii

-------
CONTENTS (Cont.)

Page

CONCLUSION.....					33

ACKNOWLEDGEMENTS 						33

REFERENCES										 33

APPENDIX A. PRIMARY LITERATURE SOURCES								 A-1

APPENDIX B. ANNOTATED BIBLIOGRAPHY							B-1

TABLES

No. Title	Page

I Emission sources and problems identified in HVAC systems	14

iv

-------
INTRODUCTION

The view that closing buildings and recirculating air can significantly affect indoor air quality
(IAQ) is not new. Woods (1983) cited Leeds (1968) quoting Benjamin Franklin's
correspondence to a physician:

"I considered [fresh air] an enemy, and closed with extreme care every crevice in rooms
I inhabited. Experience has convinced me of my error. I am persuaded that no common
air from without is so unwholesome as the air within a closed room that has been often
breathed and not changes. ... You physicians have of late discovered, after a contrary
opinion had prevailed some ages, that a fresh and cool air does good to persons in the
small-pox and other fevers. It is to be hoped, that in another century or two we may
find out that it is not bad even for people in health,"

We are operating, however, on the premise that heating, ventilating and air conditioning
(HVAC) systems are essential to modern life, and that, when properly designed, installed,
operated and maintained, HVAC systems do provide healthy, comfortable indoor
environments. However, it has recently been suggested that sick building syndrome and
occupant complaints are related primarily to mechanical ventilation. In a questionnaire-based
investigation of 43 British office buildings, Burge et al. (1987) found that complaints occurred
more frequently in buildings where HVAC systems provided cooling and humidification. In
the US, NIOSH blamed inadequate ventilation for complaints in 52% of 484 buildings
(Crandall 1987). Health and Welfare Canada found the same percentage in 1362 Canadian
buildings (Kirkbride et al. 1990). In apportioning emission sources affecting IAQ, Fanger et
al. (1988), Pejtersen et al. (1991) and Molhave and Thorsen (1991) attributed a large fraction
of perceived air degradation to pollutant sources within HVAC systems.

-------
As emission sources, HVAC systems are poorly characterized in comparison to other
sources of air pollutants in buildings, such as building materials, furnishings, cleaning
products, and personal-use products. This has occurred for several reasons. HVAC systems
are usually considered to improve IAQ by diluting, filtering and removing contaminants, and
until recently have not been considered as sources. HVAC systems can have complex and
time-varying effects on contaminant emissions and transport, sometimes acting as sinks, and
other times as sources, resulting in several practical difficulties when characterizing emissions.
Finally, HVAC components and systems are large and complex, and rarely amenable to
chamber studies, composition analyses or other tests that provide the desired degree of
experimental control and relevance to operating conditions.

Over the last decade, IAQ concerns have prompted many studies of pollutant emission
sources, transport processes, health effects, mitigation strategies, and other aspects. This
review focuses on the HVAC system as a source of pollutants, either intrinsically from off-
gassing or the release of particles from components, or because of contamination.

SOURCES OF INFORMATION

The published literature can be divided into four categories:

1. Data in peer-reviewed journals

2. Data in books and proceedings

3. Reviews in peer-reviewed journals

4. Reviews in books or proceedings.

Data Published In The Peer-Reviewed Journal Literature

Peer-reviewed journals restrict publication to those papers that are deemed acceptable to
a panel of experts in the field. Journals have a wide range of standards for publication, so that
the presence of a study in a peer-reviewed journal is not a guarantee of quality, and all papers

-------
need to be read critically. Data published in peer-reviewed journals should (but does not
always) provide conclusive evidence for a particular hypothesis, and should include sufficient
information so that the reader can repeat the experiments or judge the appropriateness of the
methods, results and conclusions.

Little of this kind of literature exists in the area of HVAC sources and effects pertaining
to IAQ. Most that is available deals primarily with health effects, and provides a clear
indication that, under some conditions, HVAC systems adversely affect human health and
comfort. Several studies have been published in the peer-reviewed journal literature that deal
directly with HVAC systems as pollutant sources, either in situ, or in laboratory situations.
These studies have emphasized the discontinuous nature of bioaerosol sources in HVAC
systems, have begun to document source strengths for total VOCs implicating HVAC dust as a
reservoir, and have begun to approach basic source characterization, especially for bioaerosols.
Many journal articles address pollutant sampling and analysis methods. While few of these
specifically address problems encountered in HVAC systems, many do present applicable data.

Data Published In Books And Proceedings

Most of the literature on IAQ that is related to HVAC systems as pollutant sources is
published in the proceedings literature. These papers are usually peer-reviewed, but with less
stringent requirements than the journal literature. Papers representing data that are too limited
to support conclusions are often accepted because they present interesting new ideas or
preliminary clues that might stimulate further study. Care must be taken in interpreting the
proceedings literature, which ranges from reports of nearly complete studies to speculations
based on extremely small data sets derived from very limited study protocols. Especially
important is careful consideration of the data presented as it relates to the conclusions (if any).

Reviews

Reviews can provide, as we hope this one does, an overview of the subject. However,
their relevance depend on the reviewer's accurate interpretation of the literature, and it is

-------
important to go back to original literature to support critical decisions. Reviews often provide
access to the extensive experience of the reviewer. Likewise, the value of commentary in
reviews is dependent on the expertise of the reviewer. A review written for a peer-reviewed
journal is more likely to be carefully researched than one published in the proceedings
literature.

Good general reviews of IAQ problems specifically related to HVAC systems have been
published in the journal literature by Woods (1983), Hughes and O'Brien (1986), Morey and
Shattuck (1989), Moseley (1990) and others. Reviews emphasizing microorganisms in HVAC
systems are provided by McCunney (1987), Breif and Bernath (1988), and Burrell (1991).
Particle filtration issues are discussed by Ottney (1993), and duct cleaning is surveyed by
Luoma et al. (1993).

Summary

This review focuses on the peer-reviewed journal literature, and on papers in the

proceedings literature that represent complete or nearly complete well-designed studies of
comparable quality.

APPROACHES TO IDENTIFYING AND CHARACTERIZING HVAC SOURCES

HVAC emission sources can be investigated using several approaches:

1.	Case reports

2.	In situ component studies

3.	Longitudinal studies of HVAC and building systems

4.	Cross-sectional studies of HVAC and building systems

5.	Chamber or other laboratory tests of HVAC system components

4

-------
Case Reports

Nearly all of the literature relevant to HVAC emission sources consists of anecdotal case
reports of investigations of individual buildings or small groups of buildings. Typically, case
reports describe the motivation and/or purpose of the IAQ investigation, essential features of
the building and HVAC system, and summaries of results of visual inspection, contaminant
monitoring, and HVAC or source analyses. Reports sometimes include occupants' symptoms,
suggested mitigation strategies, and results of mitigation efforts. Case reports generally focus
on problem buildings and thus help to identify many potential HVAC emission sources.
However, case reports do not reveal a systematic and representative picture of emission
sources in HVAC systems. Quantification of results is rarely emphasized or even possible
based on the data presented. Good case reports published in the journal literature include
Banaszak et al. (1970), Bernstein (1983), Fink et al. 1976, Flaherty et al. 1984, Hodgson et
al. (1987), McJilton et al. (1990), all dealing with microbiological contamination, and Burton
(1990), a study of re-entrainment.

In Situ Component Studies

In situ component studies can be used to identify and characterize emission sources from
HVAC system components. By monitoring the concentration increment AC; (mg/m3) across
a component for contaminant i, i.e., the difference between upstream and downstream
concentrations, the emission rate E; (mg/hr) for the component can be estimated as:

Ei = ^Cj Q (1)

where Q is the flow rate (m3/hr). For surfaces, area emission rates (mg/m2-hr) may be
derived by dividing by the source's surface area. This approach is potentially applicable for
vapors and aerosols released from duct surfaces, filters, heating/cooling coils, fans, silencers,
etc.

-------
The estimation of in situ emission rates requires a number of assumptions with respect to
conditions in the system, source characteristics, and sampling.

Good mixing is assumed as normally expected in HVAC systems (Molhave and Thorsen
1991), In cases of poor mixing, as possibly in mixing boxes, a velocity-weighted
concentration increment may be needed, as obtained using a number of transects with
simultaneous measurements of air velocity and concentration. Since emissions might be
affected by temperature, pressure, flow rates and possibly other factors, these parameters
should remain constant over the measurement period. HVAC systems that cycle on and off
may require monitoring over a number of cycles to reflect intermittent conditions.
Alternatively, if the effect of these parameters on the emission rate is known, one might
carefully measure these parameters and calculate the role of each in the measured emission
rate. Air leakage into the system near the component being tested, either present initially or as
a result of the monitoring, must be avoided. Leakage is likely when the component is under
negative pressure.

Emission rates are assumed to be constant over the measurement period. For particles,
desorbing VOCs and any other pollutant with a finite reservoir, this criterion is unlikely to be
met. For example, fungus spores resulting from active growth on surfaces are produced in
batches over periods of time and in concentrations that depend on ambient conditions.
Amplification of biological agents in HVAC reservoirs may occur while the system is
inoperative (i.e., during shut-down periods), with bursts of emission occurring as the system is
turned on. This means that multiple measurements must be taken over short periods to
characterize these emissions. For vapors, local equilibrium must be assumed, thus, desorbing
(source) or absorbing (sink) processes are assumed to be balanced. Alternately, non-
equilibrium assumptions may be invoked to quantify emissions during desorbing periods.

Up- and downstream measurements should be made simultaneously. Monitoring must
avoid any disturbances to the HVAC system to prevent measurement artifacts. Because £>Cj

-------
is normally small, especially in comparison to ambient or in-duct levels, concentration
measurements must be precise to detect small increments, and representative to avoid
anomalies. In situ gaseous measurements are straightforward. Accurate particulate
measurements, however, require isokinetic sampling for particles with diameters exceeding
about 5 /Xm, which may be difficult or impossible without accessible and favorable duct
configurations. The accuracy of aerosol sampling may also be compromised in variable air
volume systems where air velocities are changing. Particulate sampling also must consider:
the large range of particle diameters of interest (e.g., 0.01 to perhaps 100 jum dia.); the need
to maintain viability of biological agents for culturing and identification; and the large sample
sizes needed to quantify mass. These constraints greatly restrict the utility of most types of
particulate measurements, and essentially prohibit the use of commercially available culture
plate bioaerosol samplers.

An alternative to emission-rate determination is to evaluate HVAC components in situ
for the potential release of pollutants. This, in fact, is the procedure usually used to document
the presence of specific pollutants in most case studies (Hugenholtz and Fuerst 1992).

Usually, the system is carefully inspected, and sampling sites are chosen on the basis of visible
contamination, or because known pollutant reservoirs are present. Bulk samples of material
from each site are usually analyzed to verify the nature and strength of the potential source,
and air samples are collected under as representative conditions as is possible. For biological
agents, the system is usually shut down, and air samples collected immediately in the vicinity
of the suspected source, and in other areas of the system. Sometimes, deliberate disturbances
of the site may be used to create a "worst case" situation. Ambient (room) air and outdoor air
are collected to determine the contribution of specific pollutants from the suspected site of
contamination.

Another type of in situ test uses side-by-side comparisons of alternate materials or
possibly systems in the same building, air handler, duct, etc. This approach has been used to

-------
test resistance to microbial infestation in colonization studies (e.g., Ahearn et al. 1992), and

appears adaptable to other studies, e.g., dust accumulation on variously coated

surfaces.

Longitudinal Studies

The role of HVAC systems in IAQ can be characterized using longitudinal studies in
which HVAC system components, ventilation rates, occupancy, and other factors are altered in
(usually one) building during interventions that typically last periods from hours to weeks.
IAQ impacts of the interventions are determined by comparison to the earlier (baseline)
condition. Variations of this approach have been used to estimate IIVAC systems as emission
sources and to evaluate effects on IAQ of different ventilation and air exchange rates .

Several studies have used an approach suggested by Fanger et al. (1988) to estimate
emissions from three broad source classes:

1.	The HVAC system

2.	Occupant and occupant-related activities,

3.	Building materials, furnishings, and other interior constituents.

These estimates are obtained using three sets of IAQ measurements and a simple IAQ model:

^indoor,t — ^"outdoor,! ^total^Qexehange,t

(2)

¦^total — ^11 VAC ^occupant ^building	®

where Cindo()r and Coutdoor are indoor and outdoor IAQ measurements, e.g., concentrations of
specific pollutants (mg/m3), and Qexchange *s exchange rate (m3/hr), usually measured with
a tracer gas. Subscript t denotes the time dependence of parameters. Etotal is the source
strength of all sources, including those related to the HVAC system (Ehvac)' occupancy and
occupant activities (Eoccupant), and building materials, furnishings, etc. (Ebuilding). The

8

-------
emission term in cq. (2) is solved by monitoring or estimating levels of specific pollutants
indoors and out, and by measuring exchange rates. To separate the emissions terms in eq. (3),
measurements of CjndtK)r and Qcxchange are taken under three conditions:

1. Occupancy with the HVAC system functioning, giving
Eryac ^occupant ^building

2. Occupancy with the HVAC system off or blocked from the space under study,
giving

F + F

occupant building

3. No occupancy, with the HVAC system operating, giving
%VAC + Ebuilding

The IAQ measurements can include physical-chemical or biological characterizations as well as
results from sensory panels. Results are assumed to represent steady-state conditions, and only
the aggregate HVAC impact, rather than those from specific components, is identified.

An alternate longitudinal approach has been used by Menzies et al. (1993) in which
ventilation and exchange rates were varied. Both physical-chemical measurements and sensory
and symptom data were collected. Unfortunately, this study did not achieve a large or
representative range of ventilation and exchange rates. As in Fanger et al. 's (1988) approach,
only the aggregate HVAC impact can be determined from this kind of study.

A combination of longitudinal and in situ studies can be used to isolate specific sources.
The activities of specific emission sources are determined and then a multiple regression
analysis is used to estimate time-varying emissions from the sources (Franke and Wadden
1987). For HVAC sources, this regression resembles:

Ct = (fi0 + Ei=1 p Bj Xi t)/Qexchange>t (4)

-------
where Ct is the measured concentration of a specific pollutant at time t; B, (mg/aetivity) are
least-squares estimates of the emission factor for source i; Xi t is the activity or intensity of
source i of p sources; and Qexchange(t (m3,'hr) is the exchange or air flow rate. This approach
requires that source activities are measurable and varying in time, and that all confounding
variables, e.g., sources that emit the same pollutant, are controlled or measured. This
approach appears applicable to several HVAC components. For example, emissions of
anticorrosive agents from steam humidifiers might be estimated using humidifier steam
consumption as dependent variable Xi t. Results for HVAC components using this approach
have not been published.

Case-control comparisons are a variant of longitudinal studies that include some aspects
of cross-sectional studies (described below). They may offer one of the more sensitive and
reliable methods of investigating effects because controls are incorporated that adjust for
variations in occupancy, weather, environmental factors (such as ambient levels of the
pollutants of interest), and other potentially confounding variables that may be encountered
over the study period. This approach permits the determination of odds-ratios and other
indicators showing the impact of the intervention with respect to the control case. Case-
control studies appear especially useful for tests that require long study periods.

Longitudinal studies have several limitations. Often, the ability to alter HVAC system
operation is limited. While HVAC system schedules can be altered, filters can be replaced,
and ventilation and exchange rates modified within narrow ranges, many components cannot
be tested. Second, longitudinal studies are time consuming. Interventions may require several
weeks or possibly years for a full evaluation. Third, in case-control studies, two comparable
buildings or spaces that can be individually controlled and monitored must be studied
simultaneously. This approximately doubles the level of effort and expense.

-------
Cross-Sectional Studies

Cross-sectional studies of HVAC systems involve collecting data on a single occasion in
many buildings. HVAC system emissions can be identified using cross-sectional studies in
buildings with different system types and components. For example, Burge et al. (1987)
studied complaint rates as a measure of IAQ in groups of buildings with natural ventilation,
mechanical ventilation without cooling, and mechanical ventilation with cooling. Such
comparisons require that confounding factors be controlled or corrected. Thus, it may be
necessary to use building types, occupant densities, smoking policies, exchange rates, interior
equipment, geographic regions, etc., that are similar in all study buildings. Statistical
adjustment for dissimilarities may be possible. In some cases, however, this is difficult given
the strength and diversity of confounding factors, and a large sample size is needed to obtain
statistically meaningful results. The likelihood of unknown, uncertain, unmeasured, or
difficult to quantify factors decreases the statistical power to distinguish and evaluate HVAC
emission sources in cross-sectional studies. Furthermore, surveys may not provide sufficient
detail to evaluate specific HVAC sources and source interactions. Cross-sectional analyses of
HVAC emission sources have not been published.

Laboratory Tests

Laboratory tests include collecting material from buildings and performing laboratory
analyses, and testing of new materials for performance or release of pollutants. Laboratory
component studies have been used for many years to determine filter efficiencies in tests that
compare up- and downstream contaminant levels. Similar tests have also been used to
determine filter shedding, and to evaluate HVAC components for biological contamination or
the potential for such contamination. Laboratory tests are essential to characterize complex
sources, e.g., biological growth (Pasanen et al. 1991b).

Environmental test chambers are routinely used to measure VOC emissions under
controlled temperatures, humidities, air flows, and loading ratios designed to reflect interior
conditions and environments. U.S. EPA has developed standardized protocols for small

-------
chamber tests (Tichenor 1989; ASTM 1991) that provide good precision and accuracy.

Certain HVAC components appear amenable to such tests, including materials that potentially
outgas (e.g., caulks, sealants, paints, etc.). Such tests have been performed for odors and
VOCs from filters (Hujanen et al. 1991; Rivers et al., 1992) (although EPA protocols were
not used in these cases). Also, modifications of the chamber setup permit the evaluation of
sorption-desorption processes (Rothenburg et al. 1989).

Although chambers provide a controlled environment so that individual characteristics of
HVAC components can be tested, laboratory conditions can never truly duplicate dynamic
building conditions; thus the quantification of emission rates may not be relevant for some
HVAC components. This is unlikely to be an issue for VOCs that are rapidly outgassed, e.g.,
"wet" products like caulks and sealants. Chamber measurements may not accurately represent
VOC emissions that depend on the previous time history of concentrations, humidities, etc.
Chambers may not realistically simulate high flows, turbulence, fan cycling, etc., and thus
may not account for entrainment and aerosolization, critical processes for some particulate
emissions.

Summary

A wide range of methods is available to identify and estimate emissions from HVAC
sources. While all methods have limitations, many are complementary. The most reliable
results will be obtained using a combination of methods. For example, case reports may
identify a suspected problem; laboratory component or chamber studies may identify and
confirm specific emissions, or may allow identification of specific pollutant reservoirs; in situ
tests may help quantify emissions or document the potential for emissions; and longitudinal
and/or cross-sectional studies may confirm and extend results. At this point, no HVAC
component has progressed through all of these stages for any pollutant.

-------
HVAC SYSTEMS AND IAQ

It is clear from the literature that HVAC systems can be sources for some pollutants,
although the extent and human health impact of the problem remains unknown. Table I
provides an overview of HVAC emission sources and problems that are discussed below.
Understanding the kinds of pollutants that can be generated in HVAC systems should help
to direct efforts at control so that impacts can be minimized.

Intrinsic Emission Sources

Seals, caulks, etc. Based on laboratory component tests, adhesives, sealants and
caulks used in HVAC systems can contain latex acrylic, styrene, butadiene rubber,
neoprene rubber, butyl rubber, vinyl, silicone, and urethane (Leovic et al. 1993). These
and other compounds may be released in the curing and aging of these products and, if they
are placed in contact with HVAC air streams, then emissions may enter occupied spaces.
No studies have been found that estimate IAQ impacts of these sources. However,
standardized chamber tests can be used to provide accurate measurements of VOC
composition and source strength for these materials, from which impacts may be estimated.
Like other "wet" products, these materials are likely to exhibit emissions which rapidly
decrease or decay after application (Tichenor and Mason 1988).

Fibers. Fibers from fiberglass linings damaged during installation or worn and
disintegrating may be entrained and discharged to occupied spaces (Morey and Shattuck
1989). Spray-on fibrous fireproofing and exposed fibrous insulation in plenums may
become dislodged and entrained (Morey and Shattuck 1989). Shumate and Wilhelm
(1991) used laboratory tests to investigate fiber shedding for various types of filters. Only
minimal amounts of fiber were shed in short and long term tests.

-------
Table I. Emission sources and problems identified in HVAC systems.
SOURCES AND PROBLEMS	TYPICAL EXAMPLES

A. Intrinsic emission sources

1.

Seals, caulks, adhesives

outgassing of VOCs, deterioration

2.

Fibers

asbestos, fiber shedding

3.

Metal degradation products

deterioration and entrainment of coatings,
platings, metal surfaces

4.

Lubricating oils, etc.

fans, motors in the air stream

5.

Ozone

release by electrostatic air cleaners

B. Emission sources resulting from contamination

1.	Dust	construction material, skin cells, etc., with

accumulation possibly leading to microbial
contamination, VOC sorption-desorption,
and low flows

2.	Other organic debris	leaves, bird droppings

3.	Growth of microorganisms	growth and aerosolization of bioaerosols and

VOCs from microorganisms at sites
including: cooling coils; drain pans,drains,
traps and sumps, filters, insulation, duct
surfaces, plenums, humidifiers and
evaporative coolers, cooling towers

4.	VOC sinks	filters, sound absorbers, insulation materials,

deposited dust

5.	Cleaning compounds and biocides biocides, disinfectants, deodorizers

6.	Boiler steam	anticorrosives, biocides, slimicides, oxygen-

scavenging or filming chemicals, anti-
corrosives, pH control neutralizers

C. Design/operational effects on IAQ

1.

Entrainment and re-entrainment

leaks, polluted outside air, building exhaust

2.

Rotary heat exchangers

sorption-desorption of VOCs

3.

Building pressurization

intake of polluted outside air

4.

Transport

odor, VOC and particle migration

5.

Climate control

high humidity

6.

Ventilation and air exchange

inadequate dilution of internal sources,
inadequate outside air

7.

Cleaning procedures

inadequate filter maintenance, clogged
condensate drains and traps, open traps, poor
access to AHUs

14

-------
Price and Crump (1992) review releases of asbestos containing materials found in older
HVAC systems. These particles may be released during building and HYAC maintenance and
repair operations. Asbestos contamination and mitigation measures are well established and
beyond the scope of this review.

Metal degradation products. Deterioration of platings on metal surfaces and subsequent
entrainment may release particles containing toxic metals (e.g., Zn and Cd). One study did
find that a small subset of dust particles contained high concentrations of Fe, Cr and Mn
(Rothenberg et al. 1989); however, the source of these particles is unknown. Also poorly
studied, microbial agents metabolize metals, and can produce metal-containing gases and
aerosols.

Lubricating oils, etc. Levin and Moschandreas (1990) mention lubricating oils in fans
and motors as potential VOC sources. Morey (1990a) recommends that motors should be
located outside of ventilation air streams, possibly for the same reason. No study
substantiating this source has been identified.

Ozone. Electrostatic air cleaners produce ozone (03) as a consequence of the
electrostatic field used to ionize and collect particles. This phenomenon is well-known (Viner
et al. 1992). Most units produce 03 increments below 50 ppb, the EPA guideline, and well
below the 120 ppb ceiling established as a National Ambient Air Quality Standard (NAAQS).
The significance of this source increases, however, in areas where ambient or building air
approaches or exceeds the 03 standard.

Emission Sources Resulting from Contamination

Dust. Dust is an accumulation of particles entrained from outside air and from air
recirculated from the ventilation system and the occupied space. Dust accumulates on filters
and on surfaces within the HVAC system. Outdoor dust contains (among other things) silica,

-------
combustion products, rubber, fungus spores, bacteria, and other whole and fragmented
organisms. Laatikainen et al. (1991) found that dust loadings were related to the height of the
air intake and the filter type used. Pasanen et al, (1992a) investigated surface density and
accumulation rates of dusts and pollen. Dust accumulation depended on filter efficiency
(including leakage between filters and frames).

Dust particles released from ventilation system components include fibers, rubber from
fan belts, metal degradation products (see above), and particles from microbial sources. Dust
from occupied spaces consists primarily of human skin scales, along with fibers, combustion
products, and microorganisms. It should be noted that return ducts typically do not employ
filtration, and low velocity returns and ceiling plenums can accumulate significant levels of
dust and debris.

In an in situ study, Krzyanowsksi (1992) observed particulate "puffs" downstream of
filters when HVAC system fans were turned on during fan cycling. Based on optical particle
counter measurements, puffs consisted largely of small (0.5-1 jum dia.) particles, possibly
previously settled material that was reentrained, or a burst of particles jarred from the filter
bank.

Laatikainen et al. (1991) analyzed 17 samples of duct dust, measuring deposition rates,
fungal spores, bacteria, pollen and total protein concentrations. Inorganic residues comprised
the bulk (58 to 91 %) of the dust, but concentrations of fungal spores and bacteria were high
and highly correlated. Nyman and Sandstrom (1991) found high levels of culturable spores
and bacteria in supply ducts. Levels decreased along the duct, suggesting removal by
deposition.

No effect on air flow rates was seen as a function of dust loading by Pasanen et al.
(1992a). However, all loading rates were low. In contrast, studies by Wall in (in Luorria et al.
1993) found a 20-30% increase in air flows after duct cleaning, although case study buildings

-------
had small and very dirty ducts. Luoma et al. (1993) suggest that the main reason for
preventive duct cleaning is to minimize the possibility of microbial growth that would be likely
in dirty ducts should water leaks or high humidity conditions occur, rather than to prevent
particle resuspension or maintain air flows.

Dust can act as a direct source of VOC emissions, especially for odiferous particles like
cigarette smoke and biological VOCs. Reemission of VOCs from extremely dirty ducts has
been implicated as an indirect source of VOC emissions in a chemical laboratory environment
(Downing and Bayer 1991). Molhave and Thorsen (1991) quantify VOC emissions from dust
and debris from a 16 year old office building containing smokers, a small kitchen and
cafeteria. Based on in-duct concentration differences, the HVAC duct was estimated to emit
161 mg/hr of TVOCs, an emission rate four times higher than the HVAC filters (42 mg/hr),
or the internal sources in the cafeteria (44 mg/hr). Thus, VOCs from dust are implicated as a
major VOC source. This study, which has not been repeated, does not include replicates, has
fairly high detection limits of 50 ng/m3, and provided only preliminary speciation of the
VOCs. (Heptane, hexane and 3 other unidentified alkanes in both room air and ducts were
identified). VOC speciation was not used to apportion sources. Using a sensory panel,
Hujanen et al. (1991) evaluated odors from dirty filters removed from AHUs in office
buildings. Odor strength was associated with filter age, buildings in polluted areas, and
possibly several unmeasured variables that included outdoor particulate concentrations and the
chemical composition of the filter loadings. Given the building space under study, filter
emissions probably resulted from cigarette smoke (Morey 1990a) and food odors.

VOC sorption and dcsorption on dust in HVAC filters and ducts may help explain results
obtained by Fanger et al. (1988), Downing and Bayer (1991), Molhave and Thorsen (1991),
Pejtersen et al. (1993) and others. Dust can contain hundreds of VOCs and semivolatile
compounds. Laboratory studies examining the surface area, adsorption, and desorption of dust
have been completed by Rothenberg et al. (1989), Kjaer and Nielsen (1993), and others. The

-------
VOC composition of dusts and fibers has been measured by Wilkins et al. (1993) and has been
shown to demonstrate some explanatory power with respect to mucous membrane irritancy and
concentration difficulty in a small cross-sectional study. However, these results have not yet
been directly linked to HVAC system performance.

Pejtersen et al. (1992) and others (Anom. 1991) have noted that odors from ducts
decreased substantially after duct cleaning.

Other organic debris. Absence of bird screens permits roosting inside air intakes,
accumulation of bird droppings, and the risk of exposure to the infectious fungi Histoplasma
and Cryptococcus, as well as other fungi and bacteria (Burrell 1991).

Growth of microorganisms. Microorganisms growing in HVAC systems produce VOCs
and aerosols that have well-known human health effects. Usually such growth is associated
with water, either as standing pools, condensation on surfaces, or absorbed in hygroscopic
materials. Virtually any part of a HVAC system can support active microbial growth if
sufficient water is present. Standing water and very wet surfaces tend to support bacterial
growth. Fungal growth predominates on dryer surfaces (Hugenholts and Fuerst 1992).

Bacteria may survive dry environments and reappear after the reintroduction of water.
Laboratory work on wallpaper substrates under varying moisture conditions indicates that
fungal microcolonies can develop within a week on occasionally wet surfaces (Pasanen et al.
1992b). Similar growth would be expected for HVAC components.

Aerosoli?ation of microorganisms from HVAC reservoirs can occur via air movement
and turbulence, mechanical disturbance (e.g., duct cleaning devices, high pressure water or
steam sprays), movement of the component (e.g., fans), droplet splash (standing water), and
by discharge mechanisms that are intrinsic to many microorganisms, especially the fungi.

-------
Cooling coils. High concentrations of bioaerosols (spores of Penicillium, Cladosporium,
Aspergillus, etc.) have been found downstream of cooling coils in many problem buildings,
especially after agitation of the coils dislodged spores (Morey 1988, 1992). Hugenholtz and
Fuerst (1992) present a scanning electron micrograph of a bacterial biofilm on a cooling coil
surface. Since cooling coils are in a rapidly moving air stream, a mechanism to suspend
aerosols exists. Inadequate maintenance (e.g., cleaning) and poor filtration (allowing high
concentrations of outdoor-source organic material to accumulate on the coils) contribute to this
source.

Drain pans, drains, traps and sumps. Standing water due to clogged condensate drains
and traps in AHUs and other mechanical spaces is frequently noted in inspections as a potential
IAQ problem (e.g., Ager and Tickner 1983, Downing and Bayer 1991, Trent 1992), and has
been identified as potential source of microbial contamination. Although aerosolization of the
contaminated water has not been identified or confirmed, droplet splash mechanisms and
intrinsic discharge mechanisms in the yeasts commonly found in these environments (i.e.,
Sporobolomyces, Itersonilia) are sufficient to assume that aerosols are produced. Water in
normally operating traps and sumps can be ejected and aerosolized under a high vacuum "pull-
through" system as air rushes in the open drain line (Trent 1992). Contaminated condensate
water can produce odors without aerosolization. McJilton et al. (1990), for example, found
odors that were apparently caused by bacterial growth in condensate water in three buildings.
Two VOCs (2-methyl propionic acid and l-butoxy-2-propanol) were associated with the
bacteria.

Filters. Case reports show that microbial contamination and amplification occur on
filters in the presence of sufficient moisture. Being in the air stream, spores and other
bioaerosols may be released. Microbial growth can deteriorate the filter media, decrease
filtration efficiency, cause clogging, and decrease the filter's useful life. Bernstein (1983), for
example, describes fungal contamination (primarily Penicillium) on filters and surrounding

-------
areas. Problems due to poor filter maintenance are described in other case reports (Acierno et
al. 1985, Morey et al. 1987, Pasanen et al. 1991a). Rivers et al. (1992) measured VOCs
emitted by pure bacterial strains isolated from organisms recovered from residential air filters,
VOCs emitted by many of the cultures included ethanoh methyl mercaptan and
dimethyldisulfide; other VOCs included methanol, trimethylamine, ethanol, acetone, methyl
ethyl ketone, dimethyl disulfide, dimethyltrisulfide, indole, cresol and phenol. The VOC
composition and emission rates depended on the bacterial strains, the metabolic activity, and
possibly the growth medium. Pasanen et al. (1991b) found that high (96%) relative humidity
rapidly stimulated fungal growth on filters removed from office buildings.

Insulation. Some materials used to line HVAC components for thermal and acoustic
purposes can support microbial growth in the presence of sufficient nutrients and moisture.
Morey (1988; 1990b; 1992) and Morey and Williams (1991) identify porous fiber linings in
air-handling units, ductwork and air terminal boxes as reservoirs and amplifiers of
microorganisms. Both insulated and non-insulated surfaces near these water reservoirs may
also become contaminated. Insulated surfaces may be less likely to allow condensation than
metal surfaces. However, if the insulated surface is hygroscopic, or becomes dirty,
contamination becomes likely at lower ambient water levels than is required for metal surfaces.
Microbial problems are more likely in supply ducts during the cooling season, with relative
humidities over 70-80%, when filters are missing, and with malfunctioning humidifiers.
Removal and cleaning of contaminated insulation may aerosolize microbes (Morey 1992).
Insulation may be resistant to sterilization, possibly due to embedded fungal material and
adherent bacteria that recolonize the insulation (Morey 1988). New linings with impermeable
surfaces may reduce the potential for colonization without compromising acoustical or
insulating effectiveness. For example, Ahearn et al. (1992) show that rigid compressed
fiberglass with a foil facing supported little microbial growth, while plastic-faced insulation
was colonized by xerophilic fungi.

-------
Duct surfaces. Duct surfaces provide a reservoir and amplifier (if moistened) for fungi
and bacteria. Duct cleaning and removal of contaminated materials (e.g., porous insulation)
may also aerosolize biological contaminants (Morey 1992).

Plenums. Ceiling tiles, fireproofing materials, and fibrous insulation in plenums that
become wet due to roof leaks, inadequate humidity control, etc., may support microbial
growth (Morey 1988; Morey and Shattuck 1989; Morey 1990a).

Humidifiers and evaporative coolers. Microbial contamination and amplification may
occur in water reservoirs and sumps of humidifiers and air washers that use recirculated water,
and in standing reservoirs and stagnant water of cold water humidification systems (Liebert et
al. 1983; Flaherty et al. 1984; Breif and Bernath 1988; Morey and Shattuck 1989; Morey
1992). Aerosolization is likely with cold water spray humidifiers and air washers. These
systems have been associated with humidifier fever and hypersensitivity pneumonitis (Arnow
et al. 1978; Fink et al. 1976; Rylandcr and Haglind 1984; Acicrno et al. 1985; Hodgson et al.
1987). Aerosols may be transported past demisters or baffle plates (Ager and Tickner 1983;
Flaherty et al. 1984). If excessive moisture is emitted from humidifiers due to leaks or other
malfunctions, microbial growth may occur at downstream components, such as heat
exchangers and duct linings. Banaszak et al. (1970) and Morey (1988) suggest respiratory and
systemic symptoms have been caused by thermophilic actinomycetes in evaporative coolers
using a cold water spray and city water. Ager and Tickner (1983) review problems associated
with systems that store and recirculate water.

Cooling towers. As is well known due to incidents of Legionnaires' Disease and Pontiac
fever (e.g., Winn 1985), cooling tower water may provide a site for microbial amplification
and bioaerosol generation. Aerosols may escape demisters or baffle plates and enter HVAC
systems with intakes nearby or downwind (Ager and Tickner 1983; Breif and Bernath 1988).

21

-------
VOC sinks. Potential sinks for VOCs and semivolatile compounds include filters, sound
absorbers, and insulation materials (in addition to dust as discussed above) (Levin and
Moschandreas 1990). The few studies examining sink-source effects have been mention above
under dust. The potential sinks that are intrinsic to HVAC systems have not been studied.

Cleaning compounds and biocides. Detergent and chemical cleaning, and sterilization of
HVAC surfaces can produce aerosols and VOCs, including toxic chlorine-containing and
antimicrobial compounds (e.g., bleach, copper-8-quinolinolate, alcohols, phenols, aldehydes,
and iodides (ACGIH 1989; Luoma et al. 1993). The U.S. EPA does not list any biocides as
approved for use in ducts or humidification systems. A number of biocides are registered for
use in cooling towers. However, cooling tower effluent is assumed not to enter the HVAC
system and the occupied space. Deodorizers release VOCs when used on filters to disguise
odors (Downing and Bayer 1991).

Boiler steam. Direct steam injection humidifiers routinely emit into the air stream
hazardous steam conditioning anticorrosive agents (e.g., soluble nitrosated amines like
morpholine) (National Research Council 1983; Morey and Shattuck 1989). Other chemicals
used in these systems, but not normally during humidifier operation, include biocides,
slimicides, oxygen-scavenging or filming chemicals, anti-corrosives, and pH control
neutralizers (Halas 1991a; 1991b).

Design/Operational Effects On IAQ

Entrainment and re-entrainment. Inadequate filtration may allow high levels of outdoor
aerosols (fungi and bacteria) to enter the ventilation airstream. Excessively dirty or clogged
filters reduce ventilation and thus increase concentrations of contaminants emitted by occupants
and building materials. If filters are missing, improperly installed (e.g., with gaps and air
leakage between filters and the filter housing, or have low or unrated efficiency against fine
particles, the airstream may contain high particulate concentrations (Ottney 1993). These

-------
particles can deposit in the HVAC system leading to problems discussed above, or may be
transported directly to the occupied space.

Leaves, soil, vegetable matter, stagnant water, etc., near or in air intakes may allow the
growth of fungi and bacteria that may subsequently enter the HVAC system (ACGIH, 1989).

While some re-entrainment of exhaust air almost always occurs in any building with
intakes and exhausts, special concerns include air intakes placed near or downwind from
exhaust ducts, cooling towers, sanitary vents, idling vehicles, laboratory hood exhausts, and
other emission sources (Ager and Tickner 1983; Godish 1986; Hughes and O'Brien 1986;
Morey and Shattuck 1989; Burton 1990; Hodgson et al. 1991). These situations can be
avoided given acceptable dilution factors, avoidance of the recirculation cavity, and sufficient
stack exit velocities.

Rotary heat exchangers. Rotary heat exchangers may transfer VOCs from exhaust
(relief) air to supply air. Several studies show significant VOC contributions (Hughes and
O'Brien 1986; Ekberg 1991). Unfortunately, these studies did not specify the design and
materials of the exchanger. Conversely, no pollutant transfer was found at a heat recovery
wheel coated with a molecular sieve desiccant (Bayer and Downing 1991).

Building pressurization, Insufficient building pressurization may increase the infiltration
of contaminated air (Morey and Shattuck 1989; Morey 1990a). Such situations include
infiltration of vehicular exhaust in parking decks below offices (Godish 1986; Hodgson et al.
1991), street-level carbon monoxide from traffic (Collett et al. 1991), and exhaust air from
building combustion devices (e.g., furnaces). Building depressurization can also increase the
transport of radon containing soil gas into a structure. Relationships between building
depressurization and radon levels are reasonably well understood (e.g., Nazaroff et al. 1987;
Mosely 1992), and mitigation measures have been extensively studied (Henschel 1988).

Radon contamination and mitigation are beyond the scope of this review.

-------
Transport. The conveyance of contaminated air in HVAC systems will occur in any
building that has any degree of recirculation. Contaminant migration will also occur within
and between rooms and floors if pressure imbalances exists (Hughes and O'Brien 1986). For
example, Bloch (1985) shows the transport of measles in a physician's office; and Chang and
Guo (1991) show transport of contaminant gases in a residence. Aerosols and VQCs may also
be transported. Reynolds et al. (1990) show transport of fungi via the HVAC system or
possibly by human activity in several homes and office buildings. Morey (1990a) mentions the
need to avoid movement of contaminants between apartment by proper zoning or provision of
separate ventilation systems. These problems can be avoided by providing local exhaust to
source areas, e.g., reproduction areas, cafeterias, smoking areas, and by minimizing pressure
imbalances.

Climate control. In especially hot and humid climates, inadequate humidity control may
result in microbial contamination on surfaces of building materials and possibly building
contents. One case report indicates that high humidities softened and impaired the curing of
carpet adhesives that then appeared to emit high levels of VOCs (Bayer and Downing 1992).
While design temperatures were achieved in the hot and humid climate, the HVAC system was
unable to remove excessive moisture from the air.

Ventilation and air exchange. Inadequate provision of outside air to interior spaces leads
to unacceptable IAQ (e.g., Morey and Shattuck 1989; Wolter 1991; Persily 1993).

Contributing factors may include purposely closed, inoperable and/or uncalibrated outside air
dampers; insufficient air flow, especially during "pinch down" of VAV systems; inadequate
air distribution; low ventilation efficiency; en train men t and contaminant migration (as
discussed above); and other circumstances. Air exchange rates have been found to be
inversely proportional to carbon dioxide (C02) and some other pollutant levels, as expected
(Persily 1993). However, several longitudinal studies in which exchange rates were altered

-------
have not provided the expected results with respect to odor, irritant and health outcomes or
perceived IAQ, For example, in a 19 story office building at low outside air ventilation rates,
higher ventilation rates increased the infiltration of outside air containing carbon monoxide
from street level sources, indicating that building pressurization was insufficient (Collett et al.
1991). Menzies et al. (1993) found that increases in outside air ventilation over a narrow,
high range (up to 50 cfm/person) did not decrease the incidence of sick building syndrome
symptoms and complaints reported by office workers. Secondary factors that may have
confounded results in both studies include deficiencies in the monitoring instrumentation and
monitoring approaches, the inability to characterize microenvironments, and unaccounted
emission sources.

Cleaning. As mentioned earlier, duct cleaning may aerosolize biological contaminants
found in ducts due to rapid air movement and turbulence, mechanical disturbance, and
movement of components.

Control of microbial contamination involves limiting access to water, and rigorous

maintenance of components that are necessarily wet. Access to small AHUs installed and
sealed within building walls (Morey and Shattuck 1989), and the lack of access doors in large
AHUs are common problems that prevent necessary maintenance.

SENSORY STUDIES

Several studies have apportioned contaminants to major classes of emission sources, including
the HVAC system, using sensory measures. Using a sensory panel, Fanger et al. (1988)
investigated 20 spaces in Copenhagen with occupants present, occupants absent, and with the
occupants absent and the HVAC system turned off. Using the approach described by eqs. (2-
3), the contribution of the HVAC system to the total perceived IAQ degradation varied from

-------
space to space but averaged 40%. The sensory measurements were not correlated with CO,
C02» air exchange rate, or TVOC levels measured in the buildings. Using the same
longitudinal technique, Pejtersen et al. (1991) studied 10 kindergarten classrooms, obtaining
comparable results. However the ventilation systems produced a higher contribution (in
olf/m2 floor area) than found in offices and assembly halls. Based on C02 measurements,
several schools were poorly ventilated. In a variation of this approach, Pejtersen et al. (1993)
used a sensory panel, CO, and C02 measurements to apportion the perceived IAQ degradation
to three source categories: occupants; smoking; and building materials and the ventilation
system (combined). CO served as a surrogate for cigarette smoking, and C02 for bioeffluents.
The non-cigarette and non-bioeffluent component is attributed to the materials and ventilation
system. In 9 buildings, 62% of the olf load is attributable to the building and ventilation
system.

These studies indicate the potential importance of the HVAC system as a contaminant
source. Unfortunately, culpable HVAC components are not identified. Additionally, it is
unclear whether results reflect intrinsic emission sources or those resulting from
contamination, and study designs may have determined whether sorption or desorption
processes were emphasized. Finally, the relationship of the sensory responses to physical-
chemical measures is unclear. In particular, nonlinearity and thresholds of sensory responses
should be taken into account when apportioning sources quantitatively.

DISCUSSION

It is important to realize that this paper has focused on contaminant emission sources and has
highlighted sources and problems in HVAC systems. Few studies focus on the acceptable IAQ
that is provided by HVAC systems in many or most buildings. It should also be noted that all
of the longitudinal or cross-sectional health outcome studies have focused on sick building

-------
syndrome. It is probably true that the environment in most mechanically ventilated buildings
is protective for people who are sensitive to the common outdoor allergens (e.g., pollens and
many kinds of fungus spores). Ambient fungus spores levels in mechanically ventilated
buildings appear to be considerably lower than those in outdoor air, or in naturally ventilated
homes.

Based on the available literature, however, many HVAC components can act as direct or
indirect sources of particles and/or VOCs which may affect IAQ under some conditions.
Several components appear fairly frequently in the literature, and there is broad agreement
regarding their significance. Prominent among these is biological growth and bioaerosol
generation in the presence of moisture provided by air washers and other recirculating water
systems, poor control of humidity, poorly designed humidifying systems, poorly maintained
cooling coils and drip pans, etc. These problems appear to be exacerbated by dust,
accumulation, and by the presence of fibrous insulation. A number of studies describe
entrainment, migration, and infiltration of indoor and outdoor contaminants that are distributed
to indoor spaces by the HVAC system. The importance of good design and operation of
HVAC systems, including the appropriate placement and maintenance of air intakes, building
pressurization, and local exhaust in source areas, is also well accepted. More limited data
implicate dust (resulting from inadequate filtration and maintenance of filters) as a sink and
secondary source for VOCs. Evidence is inconclusive or inconsistent for the role of adhesives,
coatings, sealants, drain pans, sumps, rotary heat recovery wheels, and plenums as primary
sources for air pollutants.

Important Gaps Jii The Available Data

We have found no single study (or collection of studies) of HVAC emission sources that
was comprehensive, used robust physical-chemical measurements, and examined and isolated
pollutant contributions from major HVAC components. This review has relied on case reports
undertaken in problem buildings. These limitations have led to a number of important gaps in
our knowledge:

-------
The frequency with which various HVAC-related problems occur can not be estimated.

With the exception of a few studies examining VOCs, pollutant emission rates have not
been quantified, and thus the contribution of HVAC systems and components to overall
IAQ has not been established.

The temporal and spatial variability of emissions has not been demonstrated.

The relationship between bulk samples of water, dust, or solid materials and the quality
of delivered air is uncertain.

No studies have evaluated the efficacy of preventative measures regarding microbial
colonization or VOC sink-source relationships.

The relationship between laboratory tests and building studies has not been verified.

In general, the uncertainty of results has not been characterized, either internally (e.g.,
using replicates within a study) or externally (e.g., using intercomparisons across
studies), thus it is difficult to generalize many findings.

The tradeoffs between preventative maintenance actions (e.g., minimization of biological
growth and dust accumulation) and the impacts of cleaning, sterilization, and other
remediation procedures are unknown.

Interpretation Of Available Studies

As discussed above, most of the literature is qualitative and consists of case reports.

28

-------
Such studies are useful as they identify potential sources. However, several factors must be
addressed before generalizations can be made from this literature.

A key issue concerns the comparability, completeness, and relevance of the contaminant
measurements obtained in the studies. For airborne particles, contaminant measurements
include total and size-specific mass concentrations, gross number concentrations (using optical
particle counters), speciated concentrations (e.g., counts of culturable microorganisms), and
dust accumulation rates. Often, bulk sampling from reservoirs is used in lieu of air sampling,
especially for microbial agents, with no evidence presented as to the representativeness of
these methods. For VOCs, contaminant measurements include total and speciated VOCs
recovered and analyzed by a variety of methods which may not be comparable. Sensory
panels have provided information of yet a different nature. In most cases, alternate
measurement approaches have not been evaluated or cross-validated, nor have standardized
approaches been used.

A second factor is the timing and duration of measurements. The short monitoring
periods that are typically mandated by the case nature of nearly all studies may not provide
representative results due to effects of changing ambient conditions (e.g., humidity, air
velocities, building loads, and other dynamic factors) that may alter emissions. For example,
spore release rates depend on humidity and air velocity (Fasanen et al. 1991b) as well as the
(finite) number of spores available for dispersal. To accurately evaluate emission rates for a
specific case of fungus contamination requires the collection of (probably) hundreds of short
term (minutes) samples collected over a long period (months) of time. Interpretation of the
" snapshot" data provided in virtually all of the literature is problematic. For VOCs, the timing
of measurements may reflect sorption or desorption periods with very different results.

Results also may differ by season and HVAC operation mode (e.g., heating versus cooling,
system start-up versus constant operation, etc.).

-------
A third factor that makes interpretation of the existing literature difficult is the absence
of controls for outdoor pollutants, and for smoking. Most reported studies have been done in
urban areas, yet few studies examined outdoor pollutant levels. Smoking appears to have
influenced outcomes in several studies. The current trend to limit smoking in indoor
environments makes it mandatory that any study of HVAC system components as sources
control for smoking, or that such studies be done in non-smoking environments.

Finally, although it is clear that HVAC components can release potential air pollutants,
either intrinsically or as the result of contamination, it has not been clearly demonstrated that
these sources contribute significantly to pollutant levels in indoor air. For example, the
presence of a biologically-contaminated site or reservoir (e.g., a water sump or drain pan) may
not constitute a significant HVAC emission source for bioaerosols. Likewise, high levels of
dusts (including biological particles) found deposited in ducts, air handling units, etc., do not
constitute a particulate source unless the particles are reentrained and transported to the
occupied space.

Suggested Strategies

In any characterization study, decisions regarding overall study goals and data quality
objectives must be made prior to the development of the sampling and analysis strategy. The
following strategies are formulated with the primary goal of accurate quantification of emission
sources in typical existing HVAC systems. Cost-effectiveness is a secondary goal.

For VOCs, a combination of in situ system and component studies, in conjunction with
laboratory chamber tests, seems effective and comprehensive. The chamber studies confirm
the VOC composition and provide qualitative estimates of emission rates; the in situ studies
provide more representative and hopefully quantitative information under operating conditions.
In situ component studies may indicate emission rates from specific sources, while system
studies show the cumulative impact. In most buildings, such sampling appears feasible,

-------
although it will be practical to test only a fraction of the supply and return ducts. Obtaining
the necessary analytic precision and the temporal and spatial representativeness in the in situ
tests is the major challenge. Actions that increase concentration increments across HVAC
components will improve accuracy. This might include, for example, economizer mode
(assuming that concentrations in outside air are low), reduction or removal of internal building
sources (e.g., monitor during unoccupied periods), and sampling during periods of low
outdoor concentrations of the pollutants of interest. VOC sampling and analysis strategies
should attempt to separate sorption and desorption processes that arc antipatcd to significantly
affect apparent VOC emissions. Thus, monitoring should be performed both before and after
the onset of high concentrations produced by the sum of building and outdoor emission
sources. Continuous (or very frequent) monitoring may be needed to determine VOC patterns
and to identify representative monitoring periods. Detailed test plans, including QA/QC,
should be developed prior to sampling activity.

For particles (including bioaerosols), the sampling strategy is greatly influenced by
restrictions in available measurement technology. The ability to collect representative and
accurate samples is limited by changing duct velocities (especially in variable air volume
systems), the range of particle sizes encountered, and typical HVAC configurations. No
commercially available sampler will remain isokinetic in the constantly changing velocities and
turbulence of an operating ventilation system. The errors that are introduced are particle size-
specific and cannot be reliably predicted in any routine way. The use of optical particle
counting and sizing methods may circumvent many sampling problems. Unfortunately, these
methods do not allow chemical analysis of particles. With the technology that is currently
practical for use in building investigations, the accuracy of most in situ particulate
measurements will be compromised, leading to largely qualitative results.

A further problem has been mentioned with respect to evaluating emission rates of
biological particles from active growth. Fungus spores, in particular, are produced in batches
under conditions that do not necessarily favor particle release. Spores are often released in

-------
mass under specific conditions, and over very short periods of time. Capturing the peak of
release is statistically unlikely given the relatively short sampling times mandated by the
commercially available culture plate sampling devices. Thus, false negatives (or falsely low
emission rates) are likely for these particles.

Until devices are available that allow isokinetic, time-discriminated measures of particles
in operating HVAC systems, we suggest that bulk samples be collected to identify potential
contamination sites, and that air samples be collected in occupied spaces to document that
identified agents actually are released from the system. Such studies should be carefully
designed to include a range of HVAC operating conditions (e.g., immediately after a weekend
shut-down, immediately before start-up, immediately after start-up, periodically during a week
of building operation, etc.). They should include controls for any other potential sources of
the agent, including outdoor air. In addition, air samples can be collected immediately
downstream from suspected reservoirs within seconds of turning the HVAC system off. Small
particles should remain entrained under these circumstances, and the turbulence that might
occur at this instant might create a useful "worst case" situation. Such samples can be
collected before and during mechanical agitation of the reservoir, although these data are
difficult to interpret with respect to actual HVAC operating conditions.

For a few HVAC components, case-control techniques could reduce some of the
uncertainties arising from measurement errors and environmental factors. These techniques
are applicable to particulate and VOC emissions for those HVAC components that can be
easily altered in controlled interventions. With filters, for example, different filter types or
clean versus dirty filters could be compared in adjacent, similar buildings, or in a single
building with more than one ventilation system. Emissions could be monitored at both sites
under both conditions.

-------
CONCLUSION

This review has identified numerous sources and problems related to indoor air contaminants
and HVAC systems. These have been characterized as intrinsic emission sources, emission
sources resulting from contamination, and HVAC design and operational impacts on IAQ,
Many of the sources and HVAC system deficiencies have the potential to critically affect IAQ.
However, limitations in the available studies do not permit quantitative estimates of the
frequency or the intensity of specific sources. We have also also evaluated approaches to
characterizing HVAC emission sources. The various measurement approaches provide
different and complementary information. The use of several approaches is recommended to
provide the most reliable information.

ACKNOWLEDGEMENTS

This work was performed with the financial support of the American Society of Heating,
Refrigeration and Air-Conditioning Engineers, under the sponsorship of the Environmental
Health Subcommittee (662-TRP). We appreciate comments by M. Luoma, R. Kulp, L,
Sparks, B. Tichenor, and W. Whelan.

REFERENCES

ACGIH. 1989. Guidelines for the assessment of bioaerosols in the indoor environment.

Cincinnati, OH: American Conference of Governmental Industrial Hygienists.

Acierno, L.J., J.S. Lytle, M.H. Sweeny. 1985. Acute hypersensitivity pnuemonitis related to
forced air systems ~ a review of selected literature and a commentary on recognition and
prevention, J, Env, Health (48) 3:138-141.

33

-------
Ager, B., J.A. Tickner. 1983. The control of microbiological hazards associated with air-
conditioning and ventilation systems. Ann. Occup. Hyg. (27) 4:341-358.

Ahearn, D.G., D.L. Price, R.B. Simmons, S.A. Crow. 1992. Colonization studies of various
HVAC insulation materials. Proceedings of IAQ "92 Environments for People, ASHRAE,
Atlanta, GA, p. 101-105.

Anom., 1991. Duct cleaning -- troubleshooting foul air. Air Cond., Heating & Refrigeration
News p. 32. Oct. 7

Amow, P.M., J.N. Fink, D.P. Schluete. 1978. Early detection of hypersensisitivity
pneumonitis in office workers. Am. J. Med. 64:236-243.

ASTM. 1991. Standard guide for small-scale environmental chamber determinations of organic
emissions from indoor materials/products. Philadelphia, PA: American Society for Testing
and Materials 5116-90.

Banaszak, E.F., W.H. Thiede, J.N. Fink. 1970. Hypersensitivity pneumonitis due to
contamination of an air conditioner. New Eng. J. Med. (283) 6:271-6.

Bayer, C.W., C.C. Downing. 1992. Indoor conditions in schools with insufficient humidity
control. Proceedings of IAQ '92 Environments for People. ASHRAE, Atlanta, GA, p. 115-
118.

Bayer, C.W., C.C. Downing. 1991. Does a total energy recovery system provide a healthier
indoor environment? Proceedings of IAQ '91 Healthy Buildings, ASHRAE, Atlanta, GA,
p. 74-76, Sept. 4-8.

Bernstein, R.S. 1983. Exposures to respirable, airborne peniciIlium from a contaminated
ventilation system: clinical, environmental and epidemiological aspects. Am. Ind. Hyg.
Assoc. J. (44) 3:161-169.

Bloch, A.B. 1985. Measles outbreak in a pediatric practice: airborne transmission in an office
setting. Pediatrics (75) 4:676-683.

Breif, R.S., T. Bernath. 1988. Indoor pollution: Guidelines for prevention and control of
microbiological respiratory hazards associated with air conditioning and ventilation systems.
Appl. Ind. Hyg. Assoc. (3) 1:5-10.

34

-------
Burge S.» A. Hedge, S. Wilson, J.H. Bass, A. Robertson. 1987. Sick building syndrome: A

study of 4373 office workers. Attn. Occup. Hyg. (31) 493-504.

Burrell, R. 1991. Microbiological agents as health risks in indoor air. Env. Health Persp.
(95) 29-34.

Burton, D.J. 1990. Re-entrainment of building's exhaust air creates complex engineering

problem. Occup. Health Safety, 36-38.

Chang, J.C.S., Z. Guo. 1991. The effects of building features on indoor air and pollutant

movements. Proceedings of IAQ '91, ASHRAE, Washington, DC, p. 67-73.

Collett, C.W., J.A. Vcntresca, S. Turner. 1991. The impact of increased ventilation on
indoor air quality. Proceedings of IAQ '91 Healthy Buildings, ASHRAE. Atlanta, GA,
Sept. 4-8, p. 97-100.

Crandall, M.S. 1987. NIOSH indoor air quality investigations: 1971 through 1987.

Proceedings, Air Pollution Control Association Annual Meeting, Dallas, TX.

Downing, C.C., C.W. Bayer. 1991. Operation and maintenance for quality indoor air.
Proceedings of IAQ '91 Healthy Buildings, ASHRAE, Atlanta, GA, p. 372-374, Sept. 4-
8.

Ekberg, L.E. 1991. Indoor air quality in a new office building. Proceedings of IAQ '91

Healthy Buildings, ASHRAE, Atlanta, GA, p. 125-127, Sept. 4-8.

Fanger, P.O., Lauridsen, J., Bluyssen, P., Clausen. 1988. Pollution sources in offices and

assembly halls quantified by the olf unit, Energy and Buildings, 12:7-19.

Fink, J.N, E.F. Banaszak, JJ. Baroriak. 1976. Lung disease from forced air systems. Clin.

Notes on Respiratory Dis. (84) 4:406-413.

Flaherty, D.K., F.H. Deck, J. Cooper. 1984. Bacterial endotoxin isolated from a water spray
air humidifieation system as a putative agent of occupation-related lung disease.

Infection and Immunity (43) 1:206-212.

Franke, J.E., R.A. Wadden. 1987. Indoor contaminant emission rates characterized by

source activity factors. Environ. Sci. Technol 21:41-45.

Godish, T. 1986. Indoor air pollution in offices and other non-residential buildings, /.
Environ. Health (48) 4:190-195.

35

-------
Halas, J.J. 1991a. Reflections on steam-humidified room air, part 1. Engin. Systems, Jan/Feb.
p. 98-100.

Halas, J.J. 1991b. Reflections on steam-humidified room air, part 2. Engin. Systems, March,
p. 72-75.

Henschel, D.B. 1988. Radon reduction techniques for detached houses. Technical guidance
(second edition), Air and Energy Engineering Research Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, NC, EPA/625/5-87/019 (NTIS PB88-184908).

Hodgson, M.J., P.R. Morey, J.S. Simon. 1987. An outbreak of recurrent acute and chronic
hypersensitivity pneumonitis in office workers. Am. J. Epidemiol. (125) 4:631-638.

Hodgson, A.T., J.M. Daisey, R.A. Grot. 1991. Sources and source strengths of volatile
organic compounds in a new office building. J. Air and Waste Manage. Assoc. 41:1461-
1468.

Hugenholtz, P., J. A. Fuerst. 1992. Heterotrophic bacteria in an air handling system. Appl.
Environ. Microbio. (58) 12:3914-3920.

Hughes, R.T., D.M. O'Brien. 1986. Evaluation of building ventilation systems. Am. Ind.
Hyg. J. (47) 4:207-213.

Hujanen, M.„ O. Seppanen, P. Pasanen. 1991. Odor emissions from the used filters of air
handling units. Proceedings of IAQ '91 Healthy Buildings, ASHRAE, Atlanta GA, Sept. 4-
8, p. 329-333.

Kirkbride, J., H.K. Lee, C. Moore. 1990. Health and Welfare: Canada's experience in indoor
air quality investigations. Proceedings of Indoor Air '90. Toronto, Canada.

Kjaer, U.D., P. A. Nielsen. 1993. Adsorption studies on dust samples from the indoor
environment. Proceedings of Indoor Air '93, Helsinki, Finland, 2:579-584.

Krzyanowski, M.E. 1992. Use of airborne particle counting to evaluate indoor air quality for
remediation and control. IAQ '92 Environments for People, ASHRAE, Atlanta, GA, p.
289-299.

Laatikainen, T., P. Pasanen, L. Korhonen, A. Nevalainen, J. Ruuskanen. 1991. Methods for
evaluating dust accumulation in ventilation ducts. Proceedings of IAQ '91 Healthy
Buildings, ASHRAE, Atlanta, GA, Sept. 4-8, p. 379-382.

-------
Leeds, L.W. 1968. Lectures on ventilation at Franklin Institute, 1866-67, New York, New
York: John Wiley and Sons, p.8-9.

Leo vie, K.W., J.B. White, C. Sarsony. 1993. EPA's indoor air/pollution prevention

workshop. Presented at the 86th Annual Air and Waste Manage. Assoc. Meeting, Denver,
CO, June 13-18.

Levin, H., D. Moschandreas. 1990. Source assessment. Proceedings of IAQ '90 1990,
ASHRAE, Ottawa, Canada, p. 461-462.

Liebert, C.A., M. A. Hood, P.A. Winter, F.L. Singleton. 1983. Observations on biofilm
formation in industrial air-cooling units. Devel. Ind. Microbio. 24:508-517.

Luoma, M., A.L. Pasanen, P. Pasanen, Y. Fan. 1993. Duct cleaning - a literature survey,
AIVC Air Infiltration Rev. (14) 4:1-5.

McCunney, R.J. 1987. The role of building construction and ventilation in indoor air
pollution. N. Y. State. J. Med. (87) 4:203-209.

McJilton, C.E., S.J. Reynolds, S.K. Streifel, R.L. Pearson. 1990. Bacteria and indoor odor
problems -three case studies. Am. Ind. Hyg. Assoc. 51:545-549.

Menzies, R. et al. 1993. The effect of varying levels of outdoor-air supply on the symptoms of
sick building syndrome. New Eng. J. Med. 12:821-827.

Molhave, L., M. Thorsen 1991. A model for investigations of ventilation systems as sources
for volatile organic compounds in indoor climates. Atmos. Environ. (25A) 2:241-249.

Morey, P.R. 1988. Microorganisms in buildings and HVAC systems: a summary of 21
environmental studies. Proceedings of IAQ '88 Engineering solutions to indoor air
problems, ASHRAE, Atlanta, GA, p. 10-24.

Morey, P.R. 1990a. Internal HVAC pollution, pilot study on indoor air quality. Report on
Meeting Held in Sainte-Aldele, Quebec, Canada, Aug. 6-8.

Morey, P.R. 1990b. Porous insulation in buildings — a potential source of microorganisms.
IAQ '90, ASHRAE, Ottawa, Canada.

Morey, P.R. 1992. Microbiological contamination in buildings: precautions during

remediation activities. IAQ '92 Environments for People, ASHRAE, Atlanta, GA, p. 94-
100.

37

-------
Morey, P.R., W.G. Jones, J.L. Clere, W.G. Sorenson. 1987. Studies of sources of airborne
microorganisms and on indoor air quality in a large office building. IAQ '86, Managing
Indoor Air for Health and Energy Conservation, ASHRAE, Atlanta, GA, April 20-23.

Morey, P.R., D.E. Shattuck. 1989. Role of ventilation in the causation of building-associated
illness. Occup. Med. State of Art Revs. (4) 4:625-642.

Morey, P.R., C.M. Williams. 1991. Is porous insulation inside an HVAC system compatible
with a healthy building? IAQ 91 Healthy Buildings, ASHRAE, Atlanta, GA. p. 128-135,

Moseley, C. 1990. Indoor air quality problems, a proactive approach for new or renovated
buildings. J. Environ. Health (53) 3:19-23.

Mosely, R.B. 1992. A simple model for describing radon migration and entry into houses.
Indoor radon and lung cancer: reality or myth ? F.T. Cross, ed. Batelle Press, Columbus,
OH, 339-356.

National Research Council. 1983. An assessment of the health risks ofmorpholine and
dimethylaminoethunol. Washington, DC: National Academy Press.

Nazaroff, W.W., S.R. Lewis, S.M. Doyle, B.A. Moed, A.V. Nero. 1987. Experiments on
pollutant transport from soil into residential basements by pressure-driven airflow. Environ.
Sci. Tech. 21:459-466.

Nyman, E., N. Sandstrom. 1991. Mikroorgnismer i ventilationssytem, Sweden,
Byggforskningsrdet, 38 pp. R45 (in Swedish).

Ottney, T.C. 1993. Particle management for HVAC systems. ASHRAE Journal (35) 7: 26-34.

Pasanen, P., M. Hejanen, A.L. Pasanen, A. Nevalainen, J. Ruuskanen. 1991a. Criteria for
changing ventilation filters. Proceedings of IAQ '91 Healthy Buildings, ASHRAE, Atlanta,
GA, Sept. 4-8, p. 383-385.

Pasanen, A.L., P. Pasanen, M.J, Jantunen, P. Kalliokoski. 1991b. Significance of air
humidity and air velocity for fungal spore release into the air. Atmos. Environ, (25A)
2:459-462.

Pasanen, P., A. Nevalainen, J. Ruuskanen, P. Kalliokoski. 1992a. The composition and
location of dust settled in supply air ducts. 13th AIVC Conference on Ventilation for
Energy Efficiency and Optimum Air Quality, Nice, France, Sept. 15-18.

38

-------
Pasanen, A., H. Heinonen-Tanksi, P. Kalliokoski, M.J. Jantunen. 1992b. Fungal

microcolonies on indoor surfaces - an explanation for the base-level fungal spore counts in
indoor air. Atmos. Environ. (26B) 1:117-120.

Pcjtersen, J. et al. 1991. Air pollution sources in kindergartens. Proceedings of IAQ '91
Healthy Buildings, ASHRAE, Atlanta, GA, Sept. 4-8, p. 221-4.

Pejtersen, J., G. Clausen, O. Fanger. 1992. Olf och energi - Fase 2: Olf-vaerdier for og efter
rensning af ventilationsanlaeg. Laboratory of Heating and Air Conditioning, Technical
University of Denmark, 43 pp. Draft, (in Danish).

Pejtersen, J., L. Oie, S. Skar, G. Clausen, P.O. Fanger. 1993. A Simple Method to
Determine the Olf Load in a Building. Proceedings of Indoor Air, Toronto, Canada.

Persily, A.K. 1993. Ventilation, carbon dioxide, and ASHRAE standard 62-1989. ASHRAE
Journal (35) 7:40-44.

Price, B., K.S. Crump. 1992. Exposure inferences from airborne asbestos measurements in
buildings. IAQ '92 Environments for People, ASHRAE, Atlanta, GA, p. 63-68.

Reynolds, S.J., A.J. Streifel, C.E. McJilton. 1990. Elevated airborne concentrations of fungi
in residential and office environments. Am. Irid. Hyg. Assoc. J. (51) 11:601-604.

Rivers, J.C., J.D. Pleil, R.W. Wiener. 1992. Detection and characterization of volatile
organic compounds produced by indoor air bacteria. J. Exp. Anal. Env. Epid. Suppl.
1:177-188.

Rothenberg, S.J., P.A. Nagy, J.A. Pickrell, C.H. Hobbs. 1989. Surface area, adsorption, and
desorption studies on indoor dust samples, Am. Ind. Hyg. Assoc. J. (50) 1:15-23.

Rylander, R., P. Haglind. 1984. Airborne Endotoxins and Humidifier Diseases. Clin. Allergy
14:109-112.

Shumate, M.W., J.E. Wilhelm. 1991. Air filtration media—evaluations of fiber shedding
characteristics under laboratory conditions and in commercial installations. Proceedings of
IAQ '91 Healthy Buildings, ASHRAE, p. 337-341.

Tichenor, B.A., M.A. Mason. 1988. Organic emissions from consumer products and building
materials to the indoor environment. J. Air Pol. Control. Assoc. 38:264-268.

39

-------
Tichenor, B.A. 1989, indoor air sources: using small environmental test chambers to
characterize organic emissions from indoor materials and products. EPA-600/8-89-074
(NTIS PB90-110131), Air and Energy Engineering Research Laboratory, Research Triangle
Park, NC, August.

Trent, W. 1992. IAQ and HVAC — some major problems and a practical remedy. Air Cond.
Heating & Refrigeration News, Oct. 19. p. 2-5.

Viner, A.S., P.A. Lawless, D.S. Ensor, L.E. Sparks. 1992. Ozone generation in DC-
energized electrostatic precipitators. IEEE Trans. Indus. Appl. (28) 3:504-512.

Wilkins, C.K., P. Wolkoff, F. Gyntelberg, P. Skov, O.I. Valbjorn. 1993. Characterization of
office dust by VOC and TVOC release — identification of potential VOCs by principal least
squares analysis. Proceedings of Indoor Air '93. Helsinki, Finland. 4:37-42.

Winn, W. 1985. Legionella and Legionnaires' disease: a review with emphasis on
environmental studies and laboratory diagnosis. C'RC Crit. Rev. Clin. Lab. Sci. (21)
40:323-338.

Wolter, R. 1991. Proper ventilation improves indoor air quality. Safety and Health, June, 36-
39.

Woods, J.E., 1983. Sources of indoor air contaminants. ASHRAE Trans. 2:462-497.

40

-------
Appendix A
Primary Literature Sources

Conferences

Indoor Air '93, Indoor Air Quality and Comfort, Helsinki, Finland, July, 1993.
Indoor Air '90, Indoor Air Quality and Comfort, Toronto, Canada, July, 1990,
Indoor Air '87, Indoor Air Quality and Comfort, Berlin, Germany, July, 1987.
IAQ '92, Environments for People, San Francisco, CA, October, 1992.
IAQ '91, The Human Equation: Health and Comfort, San Diego, CA, April, 1991.
IAQ '90, Healthy Buildings, Ottawa, Canada, April, 1990.

IAQ '88, Engineering Solutions to Indoor Air Problems, Atlanta, GA, April, 1988.
IAQ '86, Managing Indoor Air for Health and Energy Conservation, Atlanta, GA,
April, 1986.

American Industrial Hygiene Conference, Salt Lake City, UT, May, 1991.
13th AIVC Conference Proceedings, Nice, France, September, 1992.

9th AIVC Conference Proceedings, Gent, Belgium, September, 1988.

Journals and Trade Magazines

American Industrial Hygiene Association
American Review of Respiratory Disease
Applied and Environmental Microbiology
ASHRAE Journal
Atmospheric Environment
Energy and Buildings
Engineered Systems
Environmental Health Perspectives
Indoor Air Journal

Journal Air and Waste Management Association
Journal Allergy Clinical Immunology
Journal of Environmental Health

Journal of Exposure Analysis and Environmental Epidemiology
New England Journal of Medicine
Occupational Health and Safety

Miscellaneous

States: Dept. of General Adm., Washington State, December, 1989.

Federal: EPA, NIOSH
Personal collections
Literature reviews

A-l

-------
Appendix B
Annotated Bibliography

(1)	Ah earn, D.G., D.L. Price, R.B. Simmons, S.A. Crow, "Colonization Studies of
Various HVAC Insulation Materials," Proceedings of IAQ '92 Environments for People,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta,
GA, 101-105, 1992.

Keywords: insulation, microbial, fungi
Study type: in situ component study

Various insulating materials were tested in a residential HVAC system after two years of
operation. No evidence of excessive dust or moisture was found. New materials were
also examined. Results show that rigid compressed fiberglass with a foil facing
supported little microbial growth. However, plastic faced insulation was colonized by
xerophillic fungi, e.g., Eurotium and Cladosporium. Penetration into the fiber mat of
the 2 year old insulation was not observed. Fungal growth is observed in some, well-
maintained HVAC systems. Suggestions are made to use insulation media with low
available water content.

(2)	Acierno, L.J., J.S. Lytle, M.H. Sweeny, "Acute Hypersensitivity Pneumonitis Related
to Forced Air Systems — A Review of Selected Literature and a Commentary on
Recognition and Prevention," J. Env. Health, 48, 3, 138-141, 1985.

Keywords: microbial, humidifiers, filters.

Study type: literature review, case report

While primarily providing a discussion of causes and diagnoses of HP, a brief
description of prevention and control techniques for HVAC systems is provided.
Improved filtration, use of steam humidifiers, regulation of water temperatures,
bactericidal agents, ultraviolet light and maintenance are suggested to reduce bioaerosol
levels.

(3)	Ager, B.P., J.A. Tickner, "The Control of Microbiological Hazards Associated with
Air-Conditioning and Ventilation Systems," Ann. Occup. Hyg., 27, 4, 341-358, 1983.

B-l

-------
Keywords: microbial, humidifiers, filters, entrainment.

Study type: literature review, case report

A review of diseases, including "humidifier fever" and legionnaires' disease, that are
related to HVAC systems is presented. Water in humidification systems is the site of
contamination; microbial contamination is most likely to occur in systems incorporating
storage and recirculation of water; spray humidifiers produce aerosols containing
biologic matter. Cooling towers are a second microbial source. In both cases, aerosols
may be generated and transported, past baffle plates, to other environments. Turbulent
eddies outside a building are of the scale of the building, and thus contaminants from
cooling towers can be transported to air inlets in the same building or other buildings
some distance away. Microbial controls and maintenance practices are recommended for
humidifiers and cooling towers.

(4)	Banaszak, E.F., W.H. Thiede, J.N. Fink, "Hypersensitivity Pneumonitis Due to
Contamination of an Air Conditioner," New Eng. J. Med., 283, 6, 271-6, 1970.

Keywords: evaporative cooling
Study type: case report

An IAQ study was undertaken in a small office when 4 of 27 workers had respiratory
and systemic symptoms. Thermophilic actinomycetes of a micropolyspora species were
suspected in an evaporative cooler using a cold water spray and city water. Later, the
system was replaced with electric refrigeration and the system was steam cleaned. No
symptoms returned.

(5)	Batterman, S., "Sampling and Analysis of Biological Volatile Organic Compounds," in
Burge, H.A., M.L., Muilenberg, eds., Monographs in indoor air quality: State of the
Art Review of Biological Aerosols. Lewis Publishers, Chelsea, MI, 1994.

Keywords: VOCs, bacteria, fungi
Study type: literature review

A literature review and discussion is presented that describes VOCs emitted from
bacteria and fungi. Many studies are found in the food-related literature. VOCs are
emitted from many organisms with composition and emission rates that appear to depend
on the species, substrate, and evolution (phase). The polar VOCs are responsible for the

B 2

-------
various musty odors. While difficult to measure, due both to the low levels encountered
and their polarity, these VOCs may provide clues to contamination sources in HVAC
systems. Approximately 35 references are given that detail sampling and analysis
strategies, and laboratory and field studies.

(6)	Bayer, C.W., C.C. Downing, "Indoor Conditions in Schools with Insufficient Humidity
Control," Proceedings of IAQ '92 Environments for People, Atlanta, GA, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, 115 118, 1992.

Keywords: humidity, fungi, microbial, VOC

Study type: case report

In hot and humid climates, some HVAC systems arc unable to remove excessive
moisture from the air, although design temperatures are achieved. Investigations in three
school buildings indicate microbial contamination was present and occupants were
complaining of recurrent respiratory illness. Buildings were selected on the basis of
complaints. The lack of humidity controls contributed to favorable conditions for visible
microbial growth on various surfaces. Additionally, high humidities appeared to soften
and impair proper curing of carpet adhesives that continued to emit high levels of VOCs.

(7)	Bayer, C.W., C.C. Downing, "Does a Total Energy Recovery System Provide a
Healthier Indoor Environment?", Proceedings of IAQ *91 Healthy Buildings, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, 74-76,
Sept. 4-8, 1991.

Keywords: VOC, heat recovery, desiccant wheel.

Study type: in situ component test

A newly constructed 27 story office building used a molecular sieve desiccant coating on
a heat recovery wheel to recover energy in the bathroom and janitor closet exhaust air.
No pollutant transfer was found.

(8)	Bernstein, R.S., "Exposures to Respirable, Airborne Penicillium from a Contaminated
Ventilation System: Clinical, Environmental and Epidemiological Aspects," Am. lnd.
Hyg, Assoc. J. 44, 3, 161-169, 1983.

B 3

-------
Keywords: filter, fungi, symptoms
Study type: case report

Improperly maintained forced air heater/cooler units were found to be contaminated
with fungi, primarily Penicillium, on the filter and surrounding area. Describes
sampling and symptoms in depth.

(9)	Black, M.S., C.W. Bayer, H.L. Brackett, "An Office Building IAQ Problem Involving
Volatile Organic Compounds," Proceedings of Indoor Air '87—Practical Control of
Indoor Air Problems, Arlington, VA, 72-85, May, 1987.

Keywords: filters, maintenance, rcentrainment.

Study type: case report

A series of measurements were made over 5 months in a large, 5-story office building in
which occupants had experienced a number of adverse health symptoms. C02 averaged
940 ppm indoors and 325 ppm outside; TVOC averaged 364 ug/m3 indoors and 29
Uglm3 outside. A number of problems concerning ventilation and maintenance practices
in the HVAC system were identified, including missing, improperly installed and
excessively dirty filters, a large amount of dust and dirt, blocked diffusors, and
improperly placed inlets. These problems were suspected as decreasing ventilation and
increasing particulate levels in the building.

(10)	Bloch, A.B., "Measles Outbreak in a Pediatric Practice: Airborne Transmission in an
Office Setting," Pediatrics, 75, 4, 676-683, 1985.

Keywords: migration, infectious bioaerosols
Study type: case report

An outbreak of measles in patients widely distributed throughout a building appeared to
be caused by airborne virus as an aerosol from the source (a frequently coughing
patient). Chicken pox may also be transmitted by the airborne route. Airflow and tracer
studies using aerosols indicated transmission and dispersion throughout the office.

(11)	Breif, R.S., T. Bernath, "Indoor Pollution: Guidelines for Prevention and Control of
Microbiological Respiratory Hazards Associated with Air Conditioning and Ventilation
Systems," Appl. lnd. Hyg. Assoc. 3, 1, 5-10, 1988.

B-4

-------
Keywords: fungi, bacteria, management, mitigation
Study type: literature review, case report

Review of microbiological organisms, effects, sampling methods, and mitigation
approaches for biologicals found in HVAC systems, especially cooling towers and
humidifiers, and evaporative condensers. Provides 39 references, approximately half are
case studies.

(12) Burrell, R., "Microbiological Agents as Health Risks in Indoor Air," En v. Health Persp.
95, 29-34, 1991.

Keywords: filtration, biologic
Study type: literature review, case report

Study describes the enhancement and maintenance of IAQ related to bioaerosols. Bird
roosting sites near intakes that may spread histoplasmosis and cryptococcosis should be
eliminated; HVAC equipment should be maintained, including the prevention of leaks
and chlorination of water in humidification systems; special attention should be given to
any system or process that uses recirculating liquid and can produce an aerosol; water
should be prevented from contacting a suitable organic substrate. Air purification, once
contamination is present, may not be satisfactory. Additionally, operating costs may be
high due to high pressure drops for low porosity filters. UV lamps are cosmetic in
actual practice.

(13) Burton, D.J., "Re-entrainment of Building's Exhaust Air Creates Complex Engineering
Problem," Occup. Health Safety, 36-38, 1990.

Keywords: re-entrainment

Study type: case report, design guidance

Some re-entrainment of exhaust air almost always occurs on any building with intakes
and exhausts. Particular concerns include laboratory hood exhausts near building air
intakes. Guidelines are given to (1) achieve acceptable dilution factors, (2) avoid the
recirculation cavity, and (3) provide sufficient stack velocities.

B-5

-------
(14) Chang, J.C.S., Z. Guo, "The Effects of Building Features on Indoor Air and Pollutant
Movements," Proceedings of IAQ '91, Washingon, DC, American Society of Heating,
Refrigerating and Air-Conditioning Engineers, Atlanta, GA, 67-73, 1991.

Keywords: transport, aerosols, VOC
Study type: system study

Using full scale tests on a residential house, activated HVAC systems transported
pollutants from local sources, e.g. aerosol cans. Local exhaust is suggested in rooms
with significant emission sources. CO was used as a tracer gas, although conclusions are
stated to apply to aerosols and VOCs.

(15) Collett, C.W., J.A. Ventresca, S. Turner, "The Impact of Increased Ventilation on
Indoor Air Quality," Proceedings of IAQ '91 Healthy Buildings, American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, 97-100, Sept. 4-8,
1991.

Keywords: pressurization, infiltration, CO
Study type: longitudinal system study

CO levels increased to about 2.3 ppm in a 19 story office building at high outside air
ventilation rates, indicating that infiltration of outside air, in the presence of street-level
CO sources, was occurring, and that building pressurization was insufficient.

(16) Downing, C.C., C.W. Bayer, "Operation and Maintenance for Quality Indoor Air,"
Proceedings of IAQ '91 Healthy Buildings, American Society of Heating, Refrigerating
and Air-Conditioning Engineers, Atlanta, GA, 372-374, Sept. 4-8, 1991.

Keywords: ducts, dust, VOC
Study type: case resports

Results from 35 air quality investigations show the importance of building operation and
maintenance. Specific examples of HVAC system components acting as pollution
sources in these studies include: (1) use of a vanilla-scented deodorizer on HVAC filters
to disguise odors emanating from air intakes near six sewer roof vents; in addition,
standing water in the mechanical room and clogged condensate drains were noted. (2) A

B-6

-------
university building with laboratory fume hoods used fiberglass lined ducts. Few of the
fume hoods were operating properly, chemicals were stored improperly, and
housekeeping was poor. The extremely dirty ducts were said to be secondary sources of
VOC emissions due to the continued exposure of chemicals during building use.

(17) Ekberg, L.E., "Indoor Air Quality in a New Office Building," Proceedings of IAQ '91
Healthy Buildings, American Society of Heating, Refrigerating and Air-Conditioning
Engineers, Atlanta, GA, 125-127, Sept. 4-8, 1991.

Keywords: heat exchanger, VOC, ducts
Study type; in situ component study

Testing of a new office building in Sweden determined that a rotary heat exchanger
transferred VOCs from exhaust to supply air. GC/MS measurements using sorbent tube
samples indicated that VOC concentrations increased from 47 ^g/m3 to 104 ^g/m3 at the
exchanger. The exhaust concentration during this measurement was 182 fig/m3. A
purge section of the rotary heat exchanger was included. Further details on the type and
operation of this unit are not provided. VOCs found in the building included toluene,
Texanol, alcohols and formaldehyde, all materials found in construction materials.

(18) Fanger, P.O., J. Lauridsen, P. Bluyssen, P. Clausen, "Pollution Sources in Offices and
Assembly Halls Quantified by the Olf Unit," Energy and Buildings, 12, 7-19, 1988.

Keywords: VOC, C02, sensory

Study type: longitudinal sensory system study

A study of 20 spaces in Copenhagen was undertaken to identify sources of pollutants as
measured by sensory measures (olfs) using a panel of judges. Additional measurements
included CO, C02, air exchange, and TVOCs. Little correlation was found between
pollutant measures and olfs. By comparing perceptions in three cases (with occupants
present, without occupants, and with HVAC systems off), the relative contributions of
materials in the space (0.02-0.45 olf/m2), the ventilation system (0.02-0.59 olf/m2) and
the total (0.19-0.54 olf/m2) were found. In summary, the study found for each
occupant, 6-7 olfs from other pollutant sources, 1-2 olfs from materials, 3 olfs from
ventilation, and 2 olfs from smoking. It should be noted that C02 levels were very low,
and that only a subset of spaces in the buildings were studied. This study indicates the
potential importance of the HVAC system as a source, although its specific role is not
identified.

B-7

-------
(19) Fink, J.N, E.F. Banaszak, J.J. Baroriak, "Lung Disease from Forced Air Systems,"
Clin. Notes on Respiratory Dis., 84, 4, 406 413, 1976.

Keywords: humidifiers
Study type; case report

This case report describes acute HP found in four individuals. The disease was
associated with office air conditioners or home furnace humidifiers.

(20) Flaherty, D.K., F.H. Deck, J. Cooper, "Bacterial Endotoxin Isolated from a Water
Spray Air Humidification System as a Putative Agent of Occupation-Related Lung
Disease," Infection and Immunity, 43, 1, 206-212, 1984,

Keywords: humidifiers
Study type: case report

The study cites a number of studies that found biological growth in chilled-water spray
humidifiers, particularly on demister vanes, associated with lung disease. This report is
aimed at isolating specific endotoxins.

(21) Godish, T.. "Indoor Air Pollution in Offices and Other Non-Residential Buildings," J.
Environ. Health, 48, 4, 190-195, 1986.

Keywords: fungi, cross-contamination, entrapment, air washers
Study type: literature review

General review of problems is presented, including results of NIQSH surveys to 1983.
HVAC related sources include cross-contamination, especially likely where buildings
serve multiple purposes. A second problem is entrainment, due to local exhaust ducts
being placed upwind of outside air ducts, idling vehicles near intakes, parking decks
below offices, and buildings under negative pressure. These problems can be mitigated
using local exhausts, isolating contaminant generating areas, preventing negative
pressure, balancing HVAC systems, and raising or relocating exhaust stacks. AIIUs are
mentioned as they provide a medium for the growth of microorganisms. Locations
include condenser pans, air washers (baffle plates, sumps, aerosolization), deposits of
organic dust, and contaminated filters. HVAC systems are said to spread contaminants
throughout the building.

B-8

-------
(22) Halas, J J., "Reflections on Steam-Humidified Room Air, Part 1, Engin. Systems, 98-
100, Jan./Feb., 1991.

(23) Halas, J.J., "Reflections on Steam-Humidified Room Air, Part 2, Engin. Systems, 72-
75, Mar., 1991.

Keywords: humidifiers, steam, biocides
Study type: literature review, guidance

Steam injection humidifiers are virtually universal in most institutional and industrial
projects, and in most cases carry some hazardous chemicaJs into the workplace. Water
soluble volatile amines, used for steam conditioning as anticorrosive agents, are
recognized as hazardous. ACGIH states amine levels of 5 ppm can occur in rooms. 1
ppm of exposure leads to 270 mg/year of absorption by employees. High temperatures
and oxidation processes can nitrosate the chemicals, leading to nitrosamines that are
carcinogens, mutagens and teratogens. This can occur where open flames, heat sources
or smoking occurs. Alternatives include clean-steam. and ultrasonic fogging systems.
Guidelines for humidification equipment are given. Biocides, slimicides, oxygen-
scavenging or filming chemicals, anti-corrosives, or pH control neutralizers should not
be used during the operating cycle of the humidifier; direct steam should be used only if
there is no alternative.

(24) Hugenholtz, P., J. A. Fuerst, "Heterotrophic Bacteria in an Air Handling System,"
Appl. Environ. Microbio., 58, 12, 3914-3920, 1992.

Keywords: bacteria, cooling coils
Study type: case report

Bacteria are identified and quantified in an air handling system in a "healthy" building, a
5 story library in Australia. Despite no visible signs, the air handling system harbored
significant reservoirs of bacteria, primarily as a biofilm on cooling coils. The
concentration on the supply-side cooling coils was 105-106 CFU/cm2. (CFU -colony
forming unit.) Consisting primarily of Blastobacter, bacteria were found over the course
of the year, without change in composition, including after the annual cleaning of the
coils. Bacteria were also found in drain pan water (105-107 CFU/ml), and in sump
water of the evaporative condenser (105 CFU/ml).

B-9

-------
(25) Hughes, R.T., D.M. O'Brien, "Evaluation of Building Ventilation Systems," Am. lnd.
Hyg. J. 47, 4, 207-213, 1986.

Keywords: heat exchanger, odors, re-entrainment, ventilation
Study type: review of case reports

A review of problems NIOSH has encountered in office environments is presented.
With respect to HVAC systems, these include (1) migration of odors (including the
effects of pressure imbalance within a single ventilation system); (2) re-entrainment of
building exhaust; (3) re entrainment through heat recovery wheels; (4) poor odor
control due to insufficient ventilation; (5) microorganisms as contaminants.

(26) Hujanen, M., O. Seppanen, P. Pasanen, "Odor Emissions from the Used Filters of Air
Handling Units, Proceedings of IAQ '91 Healthy Buildings, American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, 329-333, Sept. 4-
8, 1991.

Keywords: odors, filters

Study type: laboratory component study

Dirty filters from AHUs in 10 office buildings were removed and installed in a
laboratory setup. Odors were measured by a panel using the ASTM standard STP 440
butanol scale. Filters with a high odor had long been in use or were from a building
located in a polluted area. Differences may also be related to outdoor particulate
concentrations and the chemical composition of the filter loadings.

(27) Krzyanowski, M.E., "Use of Airborne Particle Counting to Evaluate Indoor Air Quality
for Remediation and Control," Proceedings of IAQ '92 Environments for People,
Atlanta, GA, American Society of Heating, Refrigerating and Air-Conditioning
Engineers, 289-299, 1992.

Keywords: filter, fan, particles, operation
Study type: in situ component study

White light and laser optical particle counters were used to measure 5 ranges of particle
diameters (all under 5 ^m) in offices. Particulate "puffs" are produced downstream of
the filters when HVAC system fans are turned on during fan cycling. This puff may
represent previously settled dust that is reentrained, or a burst of particles jarred from the

B-10

-------
filter bank. Puffs are seen more with 0.5 - 1 fxm particles than larger particles. The
author suggests that intermittent fan cycling should not be employed during occupancy
for variable occupancy buildings, but continuous fan operation should be used to protect
dust sensitive equipment.

(28)	Laatikainen, T., P. Pasanen, L. Korhonen, A. Nevalainen, J. Ruuskanen, "Methods for
Evaluating Dust Accumulation in Ventilation Ducts," Proceedings of IAQ *91 Healthy
Buildings, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Atlanta, GA, 379-382, Sept. 4-8, 1991.

Keywords: dust, ducts, bacteria, fungi
Study type: in situ component study

Seventeen samples of settled dust in ventilation ducts were collected in three types of
buildings (an apartment house, a school, and four office buildings). Dust mass was
determined gravimetrically, and fungal spores, bacteria, pollen and total protein
concentrations were also measured. Surface density of dust varied from 3.6 to 140.8
g/m2. Repeated measurements (after vacuuming) were used to determine dust
accumulation rates. These ranged from 0.5 to 13 g/m2-year. The buildings ranged in
age from 5 to 11 years; none of the HVAC systems had been cleaned. Inorganic
residues comprised from 58 to 91 % of total dust (average of 80%). Deposition of fungal
spores ranged from 200-22,500 CFU/g (average of 6,100 CFU/g); bacteria counts
ranged from 490-35,000 CFU/g (average of 7,000 CFU/g). Loadings were related to
the height of the air intake and the filter type used. Bacteria and fungal counts were
highly correlated (r—0.875).

(29)	Levin, H., D. Moschandreas, "Source Assessment," Proceedings of IAQ '90, Ottawa,
Canada, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Atlanta, GA, 461-2, 1990.

Keywords: dusts, VOC, filters
Study type: panel discussion

A panel discussion included HVAC systems as sources and sinks and named filters,
sound absorbers, and insulation as potential sources. In addition, dusts on filters were
mentioned as a sink and secondary source for volatile (especially semi volatile)
compounds. Oils in fans were also mentioned as sources.

B-ll

-------
(30) Leovic, K.W., J.B. White, C. Sarsony, "EPA's Indoor Air/Pollution Prevention
Workshop," Presented at the 86th Annual Air and Waste Manage. Assoc. Meeting,
Denver, CO, June 13-18, 1993.

Keywords: pollution prevention, duct liners, VOCs, adhesives, sealants.

Study type: discussion

Discussions of workgroups on pollution prevention activities related to HVAC systems,
included the following: (1) Adhesives, sealants, and caulks used in HVAC systems (and
other indoor locations). These materials may contain latex acrylic, styrene, butadiene
rubber, neoprene rubber, butyl rubber, polymers, vinyl, silicone, and urethane.
Pollution prevention options included material substitution, material elimination, new
technology and process substitution. Technology transfer aspects, emission assessments,
source modeling, and realtime monitoring were also suggested. (2) Ductwork surfaces.
Here, use of safe biocides in air stream, improved humidity control, external duct liners,
biocide impregnated panels, low dust accumulation technology, and sealing were
discussed. Research needs here included assessments of emission mechanisms, toxicity
of biocides. The workshop could not identify the major IAQ problems.

(31) Liebert, C.A., M.A. Hood, P.A. Winter, F.L. Singleton, "Observations on Biofilm
Formation in Industrial Air-Cooling Units," Devel Industrial Microbiology, 24, 508-
517, 1983.

Keywords: air washers, microbial
Study type: case report

Four air cooling/air washing units were studied over a 60 day period, and counts of
bacteria in water and slime were made using several approaches. After a few days,
water counts remained fairly constant, while concentrations in the slime increased over
time.

(32) McCunney, R.J., "The Role of Building Construction and Ventilation in Indoor Air

Pollution," N.Y. State. J. of Med., 87, 4, 203-209, 1987.

B-12

-------
Keywords: humidifiers
Study type: literature review

Review of causes and diagnosis of hypersensitivity pneumonitis, humidifier fever,
infections, nonspecific reactions, and psychogenic illness. Also provided is a brief
review of sources, NIOSH building studies, and suggestions for investigating IAQ
problems.

(33) McJilton, C.E., S.J. Reynolds, S.K. Streifel, R.L. Pearson, "Bacteria and Indoor Odor
Problems-Three Case Studies," Am. hid. Hyg. Assoc. 51, 545-549, 1990.

Keywords: bacteria, odors, VOCs, water
Study type: case report

Odors associated with air conditioning and condensate water were apparently caused by
"red" bacteria in the HVAC system. HVAC systems utilized a heat pump. These odors
were dispersed throughout three buildings and appeared related to odor and symptom
complaints of occupants. Analysis of the air in the building indicated 2-methyl propionic
acid and 1 -butoxy-2-propanol, the latter was also found in the head space of a pure
culture. The bacteria apparently survive dry environments and reappear after water is
reintroduced. pH and other controls were being investigated as control strategies.

(34) Molhave, L., M. Thorsen, "A Model for Investigations of Ventilation Systems as
Sources for Volatile Organic Compounds in Indoor Climates," Atmos. Environ., 25A, 2,
241-249, 1991.

Keywords: dust, VOCs, source estimation, modeling
Study type: in situ component study

A 16 year old office building in Denmark that included a small kitchen and cafeteria was
studied. Smoking was permitted; about 10 smokers were observed in the cafeteria.
Exchange rates were determined after sampling using tracer techniques. Total VOCs
were collected in rooms, room air inlets and exhausts, and within the HVAC system
using a carbon sorbent. The estimated VOC source strength of the cafeteria was 44
mg/hr; the HVAC filter contributed 42 mg/hr; other sources in the HVAC supply duct
contributed 161 mg/hr, all as total VOCs in toluene equivalents. The duct contribution

B-13

-------
was measured by concentration differences between the mixing box, downstream of the
filter, and at the terminal diffuser. Duct concentration differences were 278 and 27
/ug/m3. Filter dust was estimated as 1 kg; assuming VOC emissions are proportional to
dust, 4 kg of dust in HVAC ducts is estimated. This would result in a thin (0.02 mm)
layer throughout the ductwork. Concentrations increased by 15 times when fans were
turned back on. Room concentrations ranged from 0.269 to 0.842 jug/m3. In sum,
VOCs from dust are implied as a VOC source; the dust may have absorbed VOCs
during high pollution periods. The study does not include (or mention) replicates, the
detection limits are 50 fig/m3, fairly large given the levels observed. Finally, speciation
of VOCs was preliminary, indicating heptane, hexane, and 3 other unidentified alkanes
in both room air and ducts. In sum, accumulated dust and dirt within the HVAC system
may be a source for IAQ pollutants, including VOCs and dust. Also mentioned is a
possible reduction of air flows due to the accumulation of dirt in ventilation systems in
older builders ,

(35) Morey, P.R.. "Porous Insulation in Buildings. A Potential Source of Microorganisms."

Proceedings of IAQ *90, Ottawa, Canada, American Society of Heating, Refrigerating
and Air-Conditioning Engineers, Atlanta, GA, 1990.

Keywords: microorganisms, ducts, liners
Study type: case report review

Mineral fiber linings in air-handling units, ductwork and air terminal boxes are identified
as harbors and amplifiers of microorganisms, especially in supply ducts with relative
humidities over 70% during the cooling season. New linings offered with
nonpermeable, nonflammable surfaces may replace these without compromising
acoustical or insulating effectiveness.

(36) Morey, P.R. "Microbiological Contamination in Buildings: Precautions During
Remediation Activities," Proceedings of IAQ '92 Environments for People, Atlanta,
GA, American Society of Heating, Refrigerating and Air-Conditioning Engineers, 94-
100, 1992.

Keywords: drain pans, insulation, microbial, fungi, maintenance, removal
Study type: case report review, guidance

B-14

-------
Removal of porous materials such as insulation that are grossly contaminated by
microorganisms may result in aerosolization of microbes that can cause allergic
respiratory illness. Cleaning of HVAC systems may also produce aerosolization.
Specific HVAC components mentioned include drain pans, cooling coils, humidifier
sumps, insulation, and air handling units. Case studies of several buildings, including 2
hospitals, are detailed. Microbial contamination was severe where filters were missing,
humidifiers were operating improperly, and where humidity was above 80%. Steps to
protect workers and occupants are given, including negative pressurization of
remediation zone, isolation of remediation zone, protective clothing, and source
sampling.

(37)	Morey, P.R., "Microorganisms in Buildings and HVAC Systems: A Summary of 21
Environmental Studies," Proceedings of IAQ '88 Engineering Solutions to Indoor Air
Problems, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Atlanta, GA, 10-24, 1988.

Keywords: fungi, drain pans, insulation, AHU, maintenance
Study type: case report review

Microbial contamination was found in 21 problem buildings investigated. Spore trap
samples taken 3-6 feet downstream of cooling coils indicated mold contamination
(Penicillium, Cladosporium, Aspergillus); agitating of the heat exchanger increased
levels by a factor of 10,000. Stagnant water in drain pans was contaminated with slime
(yeasts). Soundliner using porous insulation in AHUs was found contaminated with
molds (Penicillium, Cladosporium) that was resistant to sterilization, possibly due to
embedded fungal spores that recolonized the insulation. Thermophilic actinomycetes
may be found in water spray components of HVAC systems, and some evidence exists
for their amplification in AHUs. Recommendations include using better filtration,
maintenance and materials to prevent dirt and debris on surfaces and stagnant water.

(38)	Morey, P.R., D.E. Shattuck, "Role of Ventilation in the Causation of Building-
Associated Illness, Occup. Med, State of the Art Revs. 4, 4, 625-642, 1989.

Keywords: insulation, humidifiers, fibers, fungi, ceiling cavities, maintenance
Study type: literature review, guidance

B-15

-------
The authors present a comprehensive review of ventilation systems and its relationship to
IAQ, including description of system functioning. Microbial sources and fibers (rather
than VOCs) are emphasized. Sources mentioned include inappropriate outdoor air
intakes; inadequate system design leading to negative pressurization and infiltration of
outdoor air contaminants; fiberglass linings that are damaged during maintenance, then
entrained and discharged to occupied spaces; above-ceiling cavities used as return air
plenums containing spray-on fibrous fireproofing, exposed fibrous insulation, or ceiling
tiles that become wet and support microbial growth; leaks in natural gas piping systems
in above-ceiling cavities; poorly maintained fan-coil or induction units that may contain
microbial slime, construction debris, leaves, dead mice, insects, pest control bait,
cigarette butts, pens, pencils, damaged insulation that allow microbial amplification;
humidification systems that wet duct linings, and stagnant water in cold water
humidifiers; volatile amines (e.g., morpholine) that are nitrosated and potentially toxic;
and finally inadequate rates of outdoor air exchange. Examples of common HVAC
deficiencies are given. Deficiencies are summarized as (1) system design related, e.g.,
outdoor air intakes located near emission sources; (2) construction related, e.g., leaks;
(3) building-operation related, e.g., inadequate fan cycling; (4) building maintenance
related, e.g., filter replacement, cleaning; (5) building renovation related, e.g., adding
local exhaust without supplying additional makeup air; (6) occupant related, e.g.,
exceeding design capacity of the HVAC system.

(39) Morey, P.R., C.M. Williams, "Is Porous Insulation Inside an HVAC System
Compatible with a Healthy Building?" Proceedings of IAQ '91 Healthy Buildings,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta,
GA, 128-135, 1991.

Keywords: insulation, microbial,fungi
Study type: case reports

Porous materials in air handling units are installed for acoustic and thermal insulation
purposes at the inside surface of fan coil housings, induction units, and air ventilators.
This insulation can become a secondary emission source of outdoor air contaminants.
Dirty, wet insulation can almost always become a microbial amplification site.
Mitigative approaches include upgrading the efficiency of upstream filtration, and
discouraging the use of porous insulation inside HVAC components. In building A,
Penicillium was found to grow on fiberglass in AHUs. Building B had Penicillium,

B-16

-------
Cladosporium and Epicoccum in insulation in induction units. HVAC systems in
buildings A and B were poorly maintained with inadequate filtration and humidifi cation.
Building C had Cladosporium deposited in fiberglass insulation in AHUs. Building D
had a moist, mycelial-like layer on rigid fiberboard insulation and a soft porous
fiberglass insulation lining the air supply plenum where humidities were approximately
90%. In building E a malfunctioning steam humidifier produced a sporulating fungal
growth on the porous insulation in a small AHU.

(40) Morey, P.R., W.G. Jones, J.L. Clere, W.G. Sorenson, "Studies of Sources of Airborne
Microorganisms and on Indoor Air Quality in a Large Office Building," Proceedings of
IAQ '86, Managing Indoor Air for Health and Energy Conservation, American Society
of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, April 20-23,
1986.

Study type: case report

Keywords: fungi, bacteria, drain pans, insulation, maintenance, filters, humidity

In response to complaints identified as building-related-illness at a large building
complex in Washington, DC, a number of IAQ studies were initiated. The building
contained seven large AHUs serving 15, 7 and 4 story building units, and hundreds of
small AHUs and FCUs. Penicillium, Cladosporium, Aspergillus niger and other
Aspergillus species were the dominant fungi collected by air sampling; levels exceeded
background concentrations by three orders of magnitude. Also, comfort levels were
low, and C02 concentrations exceeded 1000 ppm in many spaces. The study showed a
complex mixture of IAQ problems that include: (1) Water leaks from roof leaks and
overflowed AHU and FCU drain pains. Stagnant water in the drain pans contained 107
viable bacteria/ml. Microbial slimes were found on metal surfaces of drain pans and
nearby cooling coils. (2) Suspended ceiling tiles were also wet and sometimes covered
with mold. Space above the ceiling served as a pressurized plenum. (4) Missing ceiling
tiles and other problems led to inadequate distribution of air within the building. (5)
Summer relative humidities often exceeded 70% and further promoted growth of fungi
on building materials and furnishings. (6) Filter maintenance was inadequate. Filters
were seldom, if ever replaced. (7) Access to small AHUs was precluded as these units
were installed and sealed within building walls. (8) Porous insulation inside AHUs was
suspected of acting as a reservoir or source of bioaerosols.

B-17

-------
Moseley, C., "Indoor Air Quality Problems, A Proactive Approach for New or
Renovated Buildings," J. Environ. Health, 53, 3, 19-23, 1990.

Keywords: building commissioning, duct, dust
Study type: review and guidance

A general review of prevention-oriented strategies is provided, including discussion of
ventilation requirements, balancing, purging (prior to use), humidification, and other
aspects. Construction dust left in new HVAC system should be determined, with
purging or cleaning mentioned as a mitigation method.

Pasanen, P., M. Hejanen. A.L. Pasanen, A. Nevalainen, J. Ruuskanen, "Criteria for
Changing Ventilation Filters," Proceedings of IAQ '91 Healthy Buildings, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, 383-
385, Sept. 4-8, 1991.

Keywords: fungi, filter

Study type: laboratory component study

Filter samples from 19 office buildings were tested in chamber studies at room
temperature and 75 and 96% relative humidity. Fungal spores and C02 emissions were
used to indicate microbial activity. After 5 to 20 days incubation, 96% relative humidity
stimulated fungal growth. The main genus after incubation was Penicillium found at
levels from 7 to 38 million CFU/g of filter. Aspergillus, Aureobasidium, and
Cladosporium were found initially.

Pasanen, A.L., P. Pasanen, M.J. Jantunen, P. Kalliokoski, "Significance of Air
Humidity and Air Velocity for Fungal Spore Release into the Air," Atmos. Environ.,
25A, 2, 459-462, 1991.

Keywords: fungi, air velocity, humidity
Study type: laboratory study

Spore release from fungi is tested and found to be a function of species, humidity, and
air velocity. For the species studied, spore release was increased with air velocities
above 0.5 to 1.0 m/s and with lower (<42%) relative humidity.

B-18

-------
(44) Pasanen, P., A. Nevalainen, J. Ruuskanen, P. Kalliokoski, "The Composition and
Location of Dust Settled in Supply Air Ducts," poster presented at the 13th AIVC
Conference on Ventilation for Energy Efficiency and Optimum Air Quality, Nice,
France, Sept. 15-18, 1992.

Keywords: filters, dust, leakage
Study type: in situ component study

Supply ducts in six mechanically ventilated buildings 4 to 31 years old were examined.
Surface density and accumulation rates were determined, including pollens. The average
surface density was 10.6 g/m2 (range 1.2-58.3 g/m2); the average accumulation rate
was 3.5 g/m2 (1.2-8.3 g/m2). 82% of dust was inorganic material; 71 mg/g was found
to be pollen, mostly from coniferous trees; low levels (average of 990 CFU/g) of fungal
spores were found. Higher filter efficiencies had a significant effect on dust
accumulation. The dust accumulation was low with no significant effect on air flow
rates; similar indoor and outdoor compositions were found; the pollen indicated
significant leakage between filters and frames.

(45) Pejtersen, J., et al., "Air Pollution Sources in Kindergartens," Proceedings of TAQ '91
Healthy Buildings, American Society of Heating, Refrigerating and Air-Conditioning
Engineers, Atlanta, GA, 221-224, Sept. 4-8, 1991.

Keywords: sensory, HVAC

Study type: longitudinal sensory system study

Using the methods of Fanger (1988), 10 kindergarten classrooms were studied to classify
pollutant sources. On average, ventilation systems were found to provide 0.32 olf/m2
floor, compared to 0.07 olf/m2 from materials and 0.38 olf/m2 from occupants. The
absolute HVAC system contribution is higher than in other schools (0.20 olf/m2), offices
(0.25 olf/m2) and assembly halls (0.28 olf/m2), although the relative load is smaller
given the stronger occupant sources. In 3 rooms, C02 increments exceeded 1000 ppm
(+ background of 460) indicating poor ventilation.

(46) Price, B., K.S. Crump, "Exposure Inferences from Airborne Asbestos Measurements in
Buildings," Proceedings of IAQ '92 Environments for People, Atlanta, GA, American
Society of Heating, Refrigerating and Air-Conditioning Engineers, 63-68, 1992.

B-19

-------
Keywords: fibers, asbestos, repair, maintenance
Study type: case report review, guidance

A review and statistical model is presented for asbestos containing materials (ACMs) and
related human exposures in buildings. Numerous measurements have been completed in
several hundred buildings indicating that fiber release episodes may occur during
building maintenance and repair operations. Such operations may disturb ACMs,
causing falling or dislodging. Reference is made to EPA and other studies.

(47)	Reynolds, S.J., A.J. Strcifel, C.E. McJilton, "Elevated Airborne Concentrations of
Fungi in Residential and Office Environments," Am. Ind. Hyg. Assoc., 51, 11, 601-604,
1990.

Keywords: fungi, humidity, transport
Study type: case report

Sampling of 6 homes and office buildings conducted in response to health complaints
showed indoor/outdoor ratios of airborne fungi from 9 to 364. Most studies indicated
fungal contamination on building materials and furnishings due to water leaks. Fungi
were found in the affected room and nearby rooms, apparently aerosolized from
contaminated materials either via air movements in HVAC systems or by human activity
(children playing). Recommendations for remedial measures include routine
maintenance of HVAC systems, HEP A vacuuming of contaminated areas, disinfectants
(bleach, copper-8-quinolinolate), or replacement.

(48)	Rivers, J.C., J.D. Pleil, R.W. Wiener, "Detection and Characterization of Volatile
Organic Compounds Produced By Indoor Air Bacteria," J. of Exp. Anal. Env. Epid.
Suppl. 1, 177-188, 1992.

Keywords: bacteria, VOCs, odor
Study type: laboratory study

This study used pure bacteria strains isolated from cultures found on residential air
filters. These isolates were grown in a laboratory culture medium (tryticase soy broth)
for several days. VOCs found to be emitted from the cultures included ethanol, methyl
mercaptan, and dimethyl disulfide. In addition, methanol, trimethylamine, acetone,

B-20

-------
methyl ethyl ketone, dimethyltrisulfide, indole, cresol and phenol were identified.
Several VOCs depended on individual strains; the composition and emission rate
depended on the metabolic activity. Some indication was found that the growth medium
influences the emissions, as it affects both the growth rate and the VOC composition.
The study suggests that the use of VOC analysis for the detection of microorganisms in
indoor air must account for several physiological factors that alter emissions and
compositions of biologically derived VOCs.

(49)	Rothenberg, S.J., P.A. Nagy, J.A. Piekrell, C.H. Hobbs, "Surface Area, Adsorption,
and Desorption Studies on Indoor Dust Samples, Am. Ind. Hyg. Assoc. J., 50, 1, 15-23,
1989.

Keywords: dust

Study type: laboratory study

Dust samples collected from homes, libraries, and offices were analyzed for
morphology, size, composition, volatilizable and combustible material, and sorption-
desorption of formaldehyde vapor. Dusts were very heterogeneous. Dusts were
comprised largely of Si, Al, Ca, Mg, typical soil elements, but some particles contained
high concentrations of Fe, Cr, Mn, suggesting that they were produced by abrasion or
combustion of alloys containing these materials. Specific surface areas for the dusts
were small, about 1.2 m2/g, and may have reflected the dominance of the arid location
(Albuquerque). Formaldehyde absorption was estimated about 10 ng/mg dust at 1 ppm
formaldehyde concentration.

(50)	Rylander, R., P. Haglind, "Airborne Endotoxins and Humidifier Diseases," Clin.
Allergy, 14, 109-112, 1984.

Keywords: humidifiers
Study type: case report

The presence of humidifier disease in a printing factory indicated that water in the
reservoir of a humidifier was contaminated with Pseudomonas. Measurements of air in
the building indicated airborne endoxins concentrations of 0.13 - 0.39 /ig/m3, airborne
viable bacteria of 700-3100/m3, and Gram-negative bacteria of 100/m3. Twenty of 50
employees reported symptoms when the humidifiers were operating.

B-21

-------
Shumate, M.W., J.E. Wilhelm, "Air Filtration Media-Evaluations of Fiber Shedding
Characteristics Under Laboratory Conditions and in Commercial Installations,"
Proceedings of IAQ '91 Healthy Buildings, American Society of Heating, Refrigerating
and Air-Conditioning Engineers, Atlanta, GA, 337-341, 1991.

Keywords: odors, filters, fibers
Study type: laboratory component tests

Methods developed and tested to measure fiber shedding of various types of filtration
media. One method used short-term laboratory tests. The second method used long
term monitoring at a commercial facility. Similar results were obtained using either test.
Results indicate that filters shed a minimal amount of fiber with some fibers being
respirable. Concentrations were 10 to 100 times smaller than fibers found in ambient
air.

Skaret, E., "Ventilation Criteria, Effectiveness, Measurement," IAQ '90, Ottawa,
Canada, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Atlanta, GA, 430-435, 1990.

Keywords: ventilation, cleanliness
Study type: literature review

A general review of ventilation requirements is presented. The review states that HVAC
sources can be a major source of pollutions. HVAC systems should be clean, cleanable,
employ filtration, and should consist of non-polluting materials that are designed to
prevent the accumulation and condensation of water.

Trent, W. "IAQ and HVAC: Some Maj or Problems and a Practical Remedy, Air
Conditioning, Heating and Refrigeration News, 2-5, Oct. 19, 1992.

Keywords: condensate, traps
Study type: literature review

Discussion of condensate problems encountered in draw-through HVAC systems that can
impede removal of condensate and allow entry of sewer fumes or other undesirable air.

B-22

-------
Traps that can prevent these problems are difficult to maintain or poorly maintained in
practice. In some cases (e.g., high negative pressure), the trap may generate an aerosol
mist by a "geysering" effect. Also, dry traps found during the heating season are not
functional.

(54) Wells, M.W., W.A. Holla, "Ventilation in the Flow of Measles and Chicken Pox
Through a Community," J. Am. Med. Assoc., 142, 1337-1344, 1950.

Keywords: ventilation
Study type: case report

This study compared the incidence of measles and chicken pox in a town using UV
radiation in HVAC systems in schools and other locations to a control town. Differences
were seen with low humidities during the heating season, indicating that the control
strategy was effective in disinfecting the indoor air.

(55) Weiss, N.S., Y. Soleymani, "Hypersensitivity Lung Disease Caused by Contamination
of an Air-Conditioning System," Annals Allergy, 29, 154-156, 1971.

Study type: case report
Keywords: construction, humidifier

This report describes a case of lung disease apparently caused by construction debris that
contaminated a forced hot air, water cooled and poorly filtered HVAC system in an
office building. Thermophilic aetinomycetes were suspected agents. The building was
cleaned but symptoms remained. The installation of a new central air conditioning
system (air-cooled) eliminated the problems.

(56) Wilkins, C.K., P. Wolkoff, F. Gyntelberg, P. Skov, O.I. Valbjorn, "Characterization of
Office Dust by VOCs and TVOCs Release-Identification of Potential VOCs by Principal
Least Squares Analysis," Proceedings of IAQ '93, Vol. 4, 37-42, Helsinki, Finland,
1993.

Study type: cross-sectional study, laboratory analysis
Keywords: dust, VOCs

B-23

-------
Dust collected from nine town halls in Denmark was analyzed for VOCs and TVOCs.
Using GC-MS, 188 VOC and SVOC compounds were identified from thermal
desorption of the dust, most of which were similar to those found in household dust.
Saturated aldehydes (C4-11), carboxylic acids (C2-14), saturated hydrocarbons (C6-21)
and phthalate esters were the dominant compound groups. The aldehydes and carboxylic
acids could be a result of microbial degradation of lipids. Also preseent in relatively
high concentrations in some buildings were nicotine, 2-pentylfuran and 2-methylpyrrole,
contained in sidestream cigarette smoke. Collected fibers contained 121-238 fig TVOC
per gram of fiber; collected particles contained 51-260 fig TVOC per gram of dust.
VOC concentrations did not have simple correlations to mucous membrane irritation
indices; however, the VOCs with the greatest explanatory power included 2-
methypropanol, hexanoic acid, 2-alkanone, 3-methylbutanaJ, octane, pentanoic acid,
heptanoic acid, 2-undecanone, 5 methyl-3-methylene-5-hexene-2-on. Statistical analysis
of VOC levels with respect to concentration difficulty identified three dominant VOCs:
pentanoic acid, hexanoic acid, hexanal.

(57) Wolter, Rich, "Proper Ventilation Improves Indoor Air Quality," Safety and Health, 36-
39, June, 1991.

Keywords: migration, re-entrainment, ventilation, microbial
Study type: case report

Four basic aspects of ventilation must be considered: (1) sufficient air must be
delivered; (2) improper air changes must be avoided to minimize the migration of
pollutants; (3) exterior air intakes must be checked for obvious contamination sources
(e.g.. plumbing vents or kitchen exhaust fans); (4) ventilation units must be in good
working order, free of dust and debris. General suggestions for assessment and control
or correction of IAQ problems are provided.

B-24

-------
TECHNICAL REPORT DATA

(Please read Instructions on the reverse before eomplet ||| |||| || |||||| ||||||[| || III III

1, REPORT NO. 2,

EPA-6QG/R-95-014

3. Ill IIIIII Illil llfllllll IIHI III

PB95-178596

4. TtTLE AND SUBTITLE

HVAC Systems as Emission Sources Affecting Indoor
Air Quality: A Critical Review

5. REPORT DATE

February 1995

6. PERFORMING organization code

7, AUTHOR(S)

Stuart Batterman (UM) and Harriet Burge (HU)*

8. PERFORMING ORGANIZATION REPORT NO.

ASHRAE 662-TRP

9. PERFORMING ORGANIZATION NAME AND ADDRESS

University of Michigan
Environmental and Industrial Health
Ann Arbor, Michigan 48109-2029

10, PROGRAM ELEMENT NO,

11. CONTRACT/GRANT NO.

CR815391-01-0 (ASHRAE)

12. SPONSORING AGENCY NAME AND ADDRESS

EPA, Office of Research and Development

Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711

13. TYPE OF REPORT AND PERIOD COVERED

Final report; 9/93 - 6/94

14. SPONSORING AGENCY CODE

EPA/600/13

15, supplementary notes AEERL project officer is Russell N. Kulp, Mail Drop 54, 919 /541-
7980. (*) Harvard University, Environmental Health, Boston, MA 02115.

is,abstractytufjy evaluates heating, ventilating, and air-conditioning (HVAC) sys-
tems as contaminant emission sources that affect indoor air quality (IAQ). Various
literature sources and methods for characterizing HVAC emission sources are re-
viewed. Available methods include in situ tests, longitudinal and cross-sectional stu-
dies, and laboratory studies. A critique of the literature reveals that few studies are
well-controlled, comprehensive, and quantitative. Significant gaps in the data are
highlighted, and procedures are suggested to improve the characterization bioaerosol
and volatile organic compound (VOC) emission sources. Based on available litera-
ture, several HVAC components are cited fairly frequently as emission sources, and
there is broad agreement regarding their significance. The components include bio-
logical growth and bioaerosol generation in the presence of moisture provided by air
washers and other recirculating water systems, poor control of humidity, poorly de-
signed humidifying systems, and inadequately maintained cooling coils and drip pans.
IAQ problems appear to be exacerbated by dust accumulation and by the presence of
fibrous insulation. Other problems include entrainment, migration, and infiltration
of indoor and outdoor contaminants that are distributed to indoor spaces by the HVAC
system.

17. KEY WORDS AND DOCUMENT ANALYSIS

a. DESCRIPTORS

b. IDENTIFIERS/OPEN ENDED TERMS

c. cosati Field/Group

Pollution Maintenance

Emission Organic Compounds

Heating Volatility

Ventilation

Air Conditioning

Biological Aerosols

Pollution Control
Stationary Sources
Indoor Air Quality
Operation

Volatile Organic Com-
pounds (VOCs)

13B 15E
14G 07 C
13H, 13A 20M

15B

13. DISTRIBUTION STATEMENT

Release to Public

19. SECURITY CLASS (This Report)

Unclassified

21. NO. OF PAGES

70

20, SECURITY CLASS (This page)

Unclassified

22. PRICE

EPA Form 2220-1 (9-73)

-------