United States
Environmental Protection
Agency
Critical Assessment of
Building Air Cleaner Technologies
FINAL REPORT
-------
-------
EPA/600/R-08/053 December 2008 www.epa.gov/ord
FINAL REPORT ON
Critical Assessment of
Building Air Cleaner Technologies
Contract No. GS-10F-0275K
Task Order 1105
to
Joseph Wood and Les Sparks
Project Officers
U.S. ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, NC
Prepared by
Kent C. Hofacre
Rick T Hecker
Anbo Wang
Mary C. Shell
Sara J. Lawhon
(614) 424-5639
(614) 424-7955
(614)424-7144
(614) 424-7219
(614) 424-7376
"WARNING - This document may contain technical data whose export is restricted
by U.S. law. Violators of export control laws may be subject to severe legal penalties.
Do not disseminate this document outside the United States or disclose its contents
to non-U.S. persons except in accordance with applicable laws and regulations and
after obtaining any required authorizations."
BATTELLE COLUMBUS OPERATIONS
505 King Avenue
Columbus, Ohio 43201-2693
Office of Research and Development
National Homeland Security Research Center, Decontamination and Consequence Management Division
-------
This report is prepared for the United States Government by Battelle. References
to specific commercial products, processes, or services by trade name, trademark,
manufacturer, or otherwise, does not constitute or imply endorsement, recommendation,
or favoring by Battelle or the United States Government.
-------
Table of Contents
List of Acronyms x
Executive Summary xiii
1.0 Introduction 1
2.0 Objective 1
3.0 Approach 3
3.1 Search and Selection of Literature 3
3.2 Technology Selection 3
3.3 Technology Reviews 4
3.4 HVAC Particle Removal System Background 4
3.4.1 Particle Size Considerations 4
3.4.2 ASHRAE 52.2 and MERVRatings 5
4.0 Critical Assessment of Mechanical Filtration 7
4.1 Technology Description 7
4.2 Theory of Mechanical Filtration 10
4.3 Summary of Relevant Studies 12
4.3.1 Performance and Variables That Affect Performance 14
4.3.2 Assessment in an HVAC System 14
4.3.3 Additional Factors 19
4.4 Critical Assessment 22
4.4.1 Technology Assessment 22
4.4.2 Impact on HVAC System 22
4.4.3 Cost Analysis 23
5.0 Critical Assessment of Electrostatically Enhanced Filtration 25
5.1 Technology Description 25
5.2 Theory of Electrostatically Enhanced Filtration 25
5.3 Summary of Relevant Studies 26
5.3.1 Performance and Variables That Affect Performance 27
5.3.2 Assessment in an HVAC System 29
5.3.3 Additional Factors 30
5.4 Critical Assessment 31
5.4.1 Technology Assessment 31
5.4.2 Impact on HVAC System 31
5.4.3 Cost Analysis 31
6.0 Critical Assessment of Electret Media 33
6.1 Technology Description 33
6.2 Theory of Electret Media 33
6.3 Summary of Relevant Studies 33
6.3.1 Performance and Variables That Affect Performance 33
6.3.2 Assessment in an HVAC System 38
6.3.3 Additional Factors 38
6.4 Critical Assessment 38
6.4.1 Technical Assessment 38
6.4.2 Impact on HVAC System 39
6.4.3 Cost Analysis 39
-------
Table of Contents
7.0 Critical Assessment of Electrostatic Precipitation 41
7.1 Technology Description 41
7.2 Theory of Electrostatic Precipitation 41
7.3 Summary of Relevant Studies 42
7.3.1 Performance and Variables That Affect Performance 42
7.3.2 Assessment in an HVAC System 51
7.3.3 Additional Factors 52
7.4 Critical Assessment 53
7.4.1 Technology Assessment 53
7.4.2 Impact on HVAC System 53
7.4.3 Cost Analysis 53
8.0 Critical Assessment of Ultraviolet Germicidal Irradiation 55
8.1 Technology Description 55
8.2 Theory of UVGI 55
8.3 Summary of Relevant Studies 55
8.3.1 Performance and Variables That Affect Performance 56
8.3.2 Assessment in an HVAC System 57
8.3.3 Additional Factors 57
8.4 Critical Assessment 57
8.4.1 Technology Assessment 57
8.4.2 Impact on HVAC System 58
8.4.3 Cost Analysis 58
9.0 Cost Analysis 59
9.1 Approach 59
9.2 Model Estimation for Mechanical Filtration 61
9.2.1 Initial Purchase Cost (Cp) 61
9.2.2 Installation Cost (Q) 61
9.2.3 Service Cost (Cs) 61
9.2.4 Retrofit Cost (Cr) 61
9.2.5 Operating Cost (C0) 61
9.2.6 Maintenance Cost (Cm) 62
9.2.7 Mechanical Filtration Cost Summary 62
9.3 Model Estimation for Electrostatically Enhanced Filtration (EEF) 62
9.3.1 Initial Purchase Cost (Cp) 62
9.3.2 Installation Cost (Q) 62
9.3.3 Service Cost (Cs) 62
9.3.4 Retrofit Cost (Cr) 63
9.3.5 Operating Cost (C0) 63
9.3.6 Maintenance Cost (Cm) 63
9.3.7 EEF Filtration Cost Summary 63
9.4 Model Estimation for Electret Media Filtration (EMF) 63
9.4.1 Initial Purchase Cost (Cp) 63
9.4.2 Installation Cost (Q) 63
9.4.3 Service Cost (Cs) 63
9.4.4 Retrofit Cost (Cr) 63
9.4.5 Operating Cost (C0) 64
9.4.6 Maintenance Cost (Cm) 64
9.4.7 Electret Media Filtration Cost Summary 64
-------
Table of Contents
9.5 Model Estimation for Electrostatic Precipitation 64
9.5.1 Initial Purchase Cost (Cp) 64
9.5.2 Installation Cost (Q) 64
9.5.3 Service Cost (Cs) 64
9.5.4 Retrofit Cost (Cr) 64
9.5.5 Operating Cost (C0) 65
9.5.6 Maintenance Cost (Cm) 65
9.5.7 Electrostatic Precipitation Cost Summary 65
9.6 Model Estimation for UVGI 65
9.6.1 Initial Purchase Cost (Cp) 65
9.6.2 Installation Cost (Q) 65
9.6.3 Service Cost (Cs) 65
9.6.4 Retrofit Cost (Cr) 65
9.6.5 Operating Cost (C0) 66
9.6.6 Maintenance Cost (Cm) 66
9.6.7 UVGI Filtration Cost Summary 66
10.0 Conclusions 67
11.0 References 69
Appendix A Databases and Conferences Searched for Relevant Information A-l
-------
List of Tables
Table 1. Minimum Efficiency Reporting Value (MERV) Parameters 6
Table 2. Minimum Efficiency Reporting Value (MERV) Parameters With Minimum Final Resistance 10
Table 3. Influence of Filter Parameters on Filtration Efficiency and Airflow Resistance 11
Table 4. Summary of Mechanical Filter Studies 13
TableS. Comparison of ASHRAE Standards 52.1 and 52.2 13
Table 6. Filter Results vs. Manufacturers' Claims 15
Table?. Mean Reduction of Indoor PM10 Levels Below "No-Filter" Case 15
Table 8. Percent Reduction of Outdoor Particles Penetrating Indoors 16
Table 9. Percent Reductions in Particle Concentrations Due to a Central Fan, MECH Filter, and ESP 17
Table 10. Filter Types and Efficiency Classes 18
Table 11. Summary of Percent Reductions of Particles 18
Table 12. Average Percent Reductions for MERV Ratings in High Class 19
Table 13. Summary of Recent Studies on Electrically Enhanced Devices 26
Table 14. Summary of Electret Studies 34
Table 15. Degradation of Electret Media Measured by Barrett and Rousseau (1998) 35
Table 16. Summary of Recent EAC/ESP Studies 43
Table 17. Summary of UVGI Studies 56
-------
List of Figures
Figure 1. Typical Size Ranges of Commonly Occurring Aerosols 4
Figure 2. Typical Panel Filters 7
Figure 3. Unusual Panel Filter Shapes 8
Figure 4. Four Mechanisms of Particle Capture 10
Figure 5. Primary Mechanisms of Capture for Various Particle Diameters 11
Figure 6. ASHRAE Standard 52.2 Test Data for a MERV 9 Filter Showing How Collection Efficiency
Increases as the Filter Loads 12
Figure 7. Comparison of Particle Deposition Rates for Different HVAC Configurations 16
Figure 8. Deposition Rates by Particle Diameter for Each Test Configuration 17
Figure 9. Illustration of Particle Charging Upstream of Fibrous Filter by lonization Array 25
Figure 10. Illustration of Fiber Polarization by Oppositely Charged Electrodes 26
Figure 11. Penetration as a Function of Particle Diameter at Several Voltages (Thorpe and Brown, 2003) 28
Figure 12. Penetration of Monodisperse Particles as a Function of Face Velocity at Several Voltages 28
Figure 13. Penetration of Monodisperse Particles as a Function of Voltage for Positive and
Negative Electrical Fields 29
Figure 14. Aerosol Penetration of Monodispersed Particles Through an Electrically Enhanced Filter and
a Permanently Charged Filter as a Function of Operation Time 30
Figure 15. Example of Indoor Particle Measurements for a 24-hour Period With an
Electrostatic Precipitation Device 31
Figure 16. Effect of Diesel Fume Aerosol on the Removal Efficiency of an Electret Filter 35
Figure 17. Effect of Cigarette Smoke on the Removal Efficiency of an Electret Filter 36
Figure 18. Effect of Arizona Road Dust on the Removal Efficiency of an Electret Filter 36
Figure 19. Efficiency Reduction of the Rigid-Cell Electret Filter With Aerosol Loading 37
Figure 20. Efficiency Reduction of the Residential Pleated Panel Electret Filter With Aerosol Loading 38
Figure 21. Schematic of ESP Process 41
Figure 22. Collection Efficiency as a Function of Particle Diameter After HVAC Use 43
Figure 23. Aerosol Penetration as a Function of Particle Diameter at Various Applied Voltages 44
Figure 25. Collection Efficiency as a Function of Particle Diameter for Various Face Velocities 45
Figure 24. Collection Efficiency as a Function of Particle Diameter for Several Face Velocities 45
Figure 26. Aerosol Penetration as a Function of Particle Diameter for Several Face Velocities 46
Figure 27. Aerosol Penetration as a Function of Particle Diameter for Several Face Velocities 46
Figure 28. Aerosol Penetration as a Function of Particle Diameter at Various Applied Voltages 47
Figure 29. Collection Efficiency as a Function of Particle Diameter for Several Applied Voltages 48
Figure 30. Aerosol Penetration as a Function of Particle Diameter for Several Positive and
Negative Corona Voltages 49
Figure 31. Pressure Drop as a Function of Face Velocity for Various Filter Configurations 49
Figure 34. Ozone Production as a Function of Current for Positive and Negative Corona 52
Figure 35. Microbial Populations Before and After Filters andaUVGI System 56
-------
List of Acronyms
ABO aerosols of biological origin
ADSP Atmospheric Dust Spot Discoloration Method
AFS American Filtration and Separations
ANSI American National Standards Institute
APS aerodynamic particle sizer
ASHRAE American Society of Heating, Refrigerating, and Air-Conditioning Engineers
ASME American Society of Mechanical Engineers
BCG Bacillus Calmette-Guerin
Bg Bacillus globigii
BLCC Building Life-Cycle Cost Program
BTU British thermal unit
BW biological warfare
CADR Clean Air Delivery Rate
CBIAC Chemical and Biological Defense Information Analysis Center
CDC Centers for Disease Control and Prevention
CD-ROM Compact Disc read-only memory
cfm cubic feet per minute
cm centimeter
CPI Consumer Price Index
CSEPP Chemical Stockpile Emergency Preparedness Program
dp particle diameter
DHHS U.S. Department of Health and Human Services
DOE U.S. Department of Energy
DOL U.S. Department of Labor
OOP di-w-octyl phthalate
DTIC Defense Technical Information Center
EAC electronic air cleaner
ECBC Edgewood Chemical Biological Center
EE electrostatically enhanced
EEF electrostatically enhanced filtration
EIA Energy Information Administration
EMF electret media filtration
EPA U.S. Environmental Protection Agency
ERDEC Edgewood Research, Development and Engineering Center
ESP electrostatic precipitator
ETV Environmental Technology Verification
fpm foot (feet) per minute
h hour
-------
HAC heating and air conditioning
HEPA high efficiency paniculate air
HP horsepower
HPAC heating, piping and air conditioning
HVAC heating, ventilation and air conditioning
IAQ indoor air quality
IEEE Institute of Electrical and Electronics Engineers
IEST Institute of Environmental Sciences and Technology
in. inch
INTC International Nonwovens Technical Conference
JAPCA Journal of the Air Pollution Control Association
kV kilovolt
KWH kilowatt-hour
L liter
m meter
MECH fibrous mechanical filter
MERV minimum efficiency reporting value
min minute
MLGW Memphis Light Gas and Water
mm millimeter
MPPS most penetrating particle size
NBC nuclear, biological and chemical
NIOSH National Institute for Occupational Safety and Health
NIST National Institute of Science and Technology
nm nanometer
NSF National Science Foundation
O&M operations and maintenance
OSHA Occupational Safety and Health Administration
Pa pascal
PM1 paniculate matter with a diameter less than 1 micrometers
PM5 paniculate matter with a diameter less than 5 micrometers
PM10 paniculate matter with a diameter less than 10 micrometers
PSL polystyrene latex
RDECOM Research Development and Engineering Command
RH relative humidity
RPM revolutions per minute
RS Research Summary(ies)
SBCCOM Soldier Biological and Chemical Command
sec, s second
SIP Structural Insulated Panel
SMPS scanning mobility particle sizer
-------
TB tuberculosis
TFP turbulent flow precipitator
TR Technical Report
TWA time-weighted average
ULPA Ultra Low Penetrating Air
um micrometer, micron
UV ultraviolet
UVGI ultraviolet germicidal irradiation
V volt
VAV variable air volume
w.g. water gauge
w/o without
r) efficiency
-------
Executive Summary
Recent events have shown that buildings and other
infrastructure are vulnerable to terrorist attacks with
biological agents. This report provides a review of the
literature on technologies that could be used in heating,
ventilation, and air conditioning (HVAC) systems to reduce
contamination of a building following such an attack.
There are, however, no identified "safe" levels of exposure
to biological threat agents, thus it is not known the extent
reduction of these particles required to provide protection
to building occupants from illness or death resulting from
exposure to these threat agents. This report was designed
therefore to provide an evaluation of the removal efficiencies
of the technology, and also to address space, power
requirements, and cost factors.
The five technologies selected for critical review were
deemed the most appropriate to reduce or inactivate
biologically active paniculate matter. Each review provides
a description of the technology, a summary of the available
literature, and a critical assessment that addresses technology
performance, trends that affect performance, the impact of
the technology on an HVAC system, and a cost analysis.
The five technologies selected for review are as follows: (1)
mechanical filtration, (2) electrostatically enhanced filtration,
(3) electret filters, (4) electrostatic precipitation, and (5)
ultraviolet germicidal irradiation (UVGI).
In general, the performance of particle removal technologies
depends on the particle size. Thus, to aid the discussion of
the technology reviews and critical assessments, background
information regarding typical particle sizes for various types
of aerosols, and particle removal concepts, are presented.
In general, particles of biological origin range in diameter
from less than 0.1 micron up to about 50 microns. Bacillus
anthracis spores are between 1 and 2 microns in diameter.
This report also provides background information on the
American Society of Heating, Refrigerating, and Air-
Conditioning Engineers (ASHRAE) standard test method
52.2 and the minimum efficiency reporting value (MERV)
ratings. Air filtration technologies used in HVAC applications
are tested with this method. ASHRAE 52.2 allows filters to
be tested in a consistent fashion and performance ratings to
be assigned. The MERV performance ratings for the various
types of filters assessed are provided.
In general, the typical HVAC filtration system in a building is
a relatively low efficiency mechanical filter that is intended
to remove particles to keep the remainder of the HVAC
system clean and to remove nuisance dust for the occupants.
However, mature technologies exist to enhance particle
removal without requiring extensive retrofits or significant
duct modifications. Operating costs may increase because the
technology is more expensive to maintain or operate.
Mechanical filters are by far the most widely used type of
air cleaner for residential and commercial building HVAC
systems. In general, the advantages of mechanical filters
are their low cost and wide availability in a variety of sizes,
types, and performance ranges. The key disadvantage
of mechanical filters is that their pressure drop increases
with use, thus requiring an increase in the power needed
to maintain airflow and requiring replacement with a
frequency that is proportional to their efficiency. Fibrous
media are commonly used in mechanical filters because
they can provide good filtration efficiency at a low pressure
drop. The fibers are either woven into the filter frame or
randomly oriented and thermally or chemically bonded
to each other and to the filter frame. Particle capture
using mechanical filters occurs through four primary
mechanisms: (1) inertial impaction, (2) interception,
(3) diffusion, and (4) electrostatic attraction.
Electrostatically enhanced filtration technology improves the
performance over standard fibrous filters that rely solely on
mechanical means for aerosol collection. The principle of
operation is to ionize the incoming airstream and particles so
that a surface charge is achieved on the incoming particles
upstream of the filter. Fibrous filter media are located
between a negatively charged electrode upstream and a
positively charged electrode downstream. When power is
applied to the electrodes, an electrical field is generated and
the fibrous filter media are polarized, i.e., the fibers of the
media form areas of negative and positive charge. In this
manner, electrostatically enhanced filtration is similar to
electret filtration (discussed next), except the fibers are not
permanently charged as with electret filters.
The performance of electrostatically enhanced filtration
technologies depends on several factors. Because mechanical
filters are used, the performance depends on the fiber
diameter and the number of fiber layers. The addition of the
electrical field over the filter creates a dependence on the
voltage, and performance also depends on the particle size
and the face velocity through the filter.
Electrostatically enhanced filtration offers benefits
over electret filters because the electrical field in the
filter is less apt to degrade. Filtration efficiencies
for electret and electrostatically enhanced filters are
similar, although electret filters are more frequently
used because they do not require electrical power.
However, electrostatically enhanced filtration devices are
relatively new to the market and relatively little research
regarding their performance is available compared to
established air cleaning technologies such as fibrous
filters or electrostatic precipitators. The impact of using
electrostatically enhanced technologies on an HVAC
system is minimal. The pressure drops are not significantly
different from what they are with other fibrous filters.
Electret filters use electrically charged media to attract
particles. In contrast to the electrostatically enhanced filter,
electret media are permanently charged in the course of
manufacturing. Therefore, electret media do not need an
-------
electrode system to charge filter media or an ionizer to charge
incoming particles during operation. Another advantage of
electret media is their relatively high collection efficiency at
very low pressure drops.
Electret filters collect particles using a combination of
mechanical and electrostatic mechanisms. The efficiency of
electret media depends on parameters such as charges on
particles, charge density of fibers, and chemical compositions
of particles and fibers. Efficiency also depends on mechanical
factors, such as fiber diameter and packing density of the
fibrous materials.
In the HVAC filtration market, electret filters are becoming
increasingly popular. The electret filters used for commercial
HVAC filtration generally have MERV ratings ranging from
8 to 16. The main concern with using electret filters is the
effect of aerosol loading on collection efficiency. However,
in spite of the collection efficiency degradation over time
(primarily due to dust loading), the efficiency of an electret
filter always exceeds that of an uncharged filter with the
identical mechanical structure.
Electret filters have a lower pressure drop than conventional
uncharged fiber filters; therefore, they can be installed
into an existing HVAC filtration system without extensive
modification such as the addition of an extra fan. These types
of filters might require a new access door to be added to the
existing unit and installation of new pressure gauges. Initial
and installation costs would be inexpensive since there is no
need for electric service.
Typical operating and maintenance costs are low for
electrets. Maintenance includes yearly changing of not
only the electret filters, but of the prefilters as well,
thus increasing the maintenance costs of the typical
office building somewhat. In general, however, electret
filters are usually less expensive than glass fiber
(mechanical) filters with the same MERV rating.
Electrostatic precipitators (ESPs) utilize particle collection
technology that has been used for decades in industrial and
combustion applications. Commercial and residential devices
employing this technology are now widely available. With
this technology, an electrical charge is imparted to incoming
dust particles as they pass through an electrical field in the
ionizing section. The charged particles are collected on plates
of an opposite charge in the collection section. ESPs offer
several advantages over traditional fibrous filters, such as
high collection efficiencies at relatively low pressure drops,
and with infrequent replacement.
In general, although the type of particle to be collected does
not impact the collection efficiency of an ESP, particle size
is a strong determinant. Also, the higher the voltage used
to ionize and collect particles, the greater the collection
efficiency will be. But as with traditional high-efficiency
filtration, the performance of ESP devices can degrade over
time. Performance degradation with ESP devices can be due
to several effects, such as dust loading. However, as long as
the ESP is cleaned regularly, performance can be maintained
at a relatively high level.
While ESPs remain an effective technology for air cleaning,
there are some negative effects, such as the additional
electrical power requirement, and the potential for these
devices to produce ozone. ESPs, because of their design,
typically will not fit into an existing air handler or existing
ductwork without major modifications. These filters would
also require new electric service. For these reasons, the
installation and initial purchase costs of these filters are
very high. However, ESPs have relatively low operating and
maintenance costs.
Ultraviolet germicidal irradiation (UVGI) can be used as an
in-duct air disinfection system, as a recirculation system used
to treat the air in a room, and as an "upper air" disinfection
system. Unlike the other filter technologies discussed in the
report, UVGI is used to inactivate the biocontaminant, rather
than remove it from the air stream. While it is possible that
these systems could be used in commercial and residential
buildings, their application is not very common.
UVGI in wavelengths of 225 to 302 nm is frequently used for
microbial disinfection, and DNA absorption of UV radiation
is maximal at 254 nm. Lethality of the UVGI system
depends on the dose of radiation that the microorganism
receives. Environmental and design variables also affect the
performance of UVGI systems and include relative humidity,
temperature, air velocity and air mixing, lamp selection,
the use of reflectors, and the combination of UVGI with
filtration. Unfortunately, few experimental data are available
for HVAC applications of UVGI.
The combination of UVGI and mechanical filtration appears
to be the most likely use of UVGI due mainly to the fact
that UVGI systems would probably be added to an existing
HVAC system that already employs some type of mechanical
filtration. This approach is advantageous since UVGI is
most effective against biocontaminants in the particle size
range where mechanical filtration is less efficient (lum
and smaller). Retrofit UVGI systems would have to be
installed downstream of the original mechanical filtration
system to aid in maintaining the fully developed light field.
Periodic cleaning of the lamps will also be important in
establishing the light field, which is critical in maintaining
the effectiveness of the UVGI system.
A UVGI system can be installed in-duct in existing
ventilation systems, although modifications are required.
Moreover, because of the more complex design of
these systems, their initial purchase cost is extremely
high—much higher than for any of the other filters
analyzed in this report. Maintenance and operating
costs for these systems are also high as they include
cleaning and changing the bulbs periodically, and
because they use a large amount of electricity.
-------
1.0
Introduction
Recent events have shown that buildings and other
infrastructure are vulnerable to terrorist attacks with
biological agents. The most serious effects of such an
attack are on the health of the occupants of the buildings.
Building occupants may suffer health effects ranging from
irritation, to severe sickness, to death. The attack may
also have long-term economic and other impacts due to
contamination of the building. Although guidelines exist
on how to prevent and/or mitigate a terrorist attack on a
building, a thorough examination of all of the available
scientific data has not yet been made to determine the
optimum course of action. This report provides a review
of the literature on technologies that could be used in
heating, ventilation, and air conditioning (HVAC) systems
to reduce the amount of biological threat agent in the indoor
environment. There are, however, no identified "safe"
levels of exposure to biological threat agents, thus it is
not known the extent reduction of these particles required
to provide protection to building occupants from illness
or death resulting from exposure to these threat agents.
2.0
Objective
The objective of this report is to provide a critical assessment
of technologies that could be used to reduce contamination
of a building following an attack with a biological agent. The
assessment was designed to provide not only an evaluation of
the scientific merit of the technology, but also consider space,
power requirements, and cost factors. In addition, although
the focus of this report is primarily on technologies for
protecting buildings from biological agents, the majority of
this report deals with air filtration technologies, which would
be applicable to any threat agent in aerosol form.
-------
-------
3.0
Approach
The critical assessment of air cleaner technologies began
with a search for relevant technical literature, conference
proceedings, and manufacturers' literature, with an
emphasis on literature written within the past five years.
Technologies were categorized under aerosol filtration or
gas-phase filtration. A cursory evaluation of each technology
was performed, and it was decided to focus the critical
assessment primarily on aerosol filtration technologies
(the exception is that UV germicidal technology is
assessed) and include more mature technologies. Five
technologies were selected for critical review as those
most appropriate to protect a building against a biological
terrorist attack. The five reviews are detailed in separate
sections of this report. Each review provides a description
of the technology, a summary of the available literature,
and a critical assessment that addresses technology
performance, trends that affect performance, the impact of
the technology on an HVAC system, and a cost analysis.
Note that the purpose of this report is not to specify which
technology is best, more effective, or beneficial. These
questions cannot be answered through an assessment of air
cleaner technologies alone; factors specific to each building
under investigation must be considered. General trends
regarding the importance of filtration and its effectiveness
have been addressed in a related report (Wang and Hofacre,
2007). Another related report addresses which parameters
are most important in mitigating a hazardous release and
how accurately these parameters need to be measured and
controlled (Hawkins and Hofacre, 2007).
3.1 Search and Selection of Literature
Information regarding air cleaner technologies was obtained
by searching military technical databases, commercial
technical databases, peer-reviewed journals, and relevant
Web sites. A full list of those searches is provided in
Appendix A. A summary of the search strategy and results
from the primary search are also given in Appendix A.
Searches were performed using the following keywords:
• air
• aerosol
• building
• electrostatic or electronic
•HEPA
• clean, cleaner, or cleaning
• particle, or paniculate
•HVAC
• germicidal irradiation
1 filter, filtering, or filtration
1 indoor
1 electret
•UVGI
Subsequent to the initial primary search, additional
articles were obtained through an EPA database that
was searched using similar keywords. Although some
overlap with the previous searches existed, this database
provided useful papers for the major technologies
investigated. Reference sections of relevant articles were
also used to identify further citations of use and interest.
Although the searches focused on literature published in
the past five years, important and representative studies
from earlier years were included when appropriate.
3.2 Technology Selection
The literature search generated hundreds of citations. Titles
and abstracts were reviewed for relevance. Prior to initiating
the primary search, air cleaning technology surveys and
searches previously conducted by Battelle were reviewed.
As a result, the following categories of gas-phase and aerosol
technologies were established:
1 Mechanical filtration
1 Electrostatically
enhanced filtration
1 Electret media
1 Electrostatic
precipitation
1 Ultraviolet germicidal
irradiation (UVGI)
• Reactive fibers/
membranes
• Inertia! separation
• Aerosol membranes
• Scrubbers
• Cold plasma
A cursory evaluation of each technology category was
performed based on previous experience in the field to focus
the critical assessment primarily on aerosol filtration and
more mature technologies. Five technologies were selected
for critical review as those that could be used to reduce
contaminant levels in buildings: (1) mechanical filtration,
(2) electrostatically enhanced filtration, (3) electret filters,
(4) electrostatic precipitation, and (5) ultraviolet germicidal
irradiation (UVGI).
The above five technology categories were selected
because they represented distinguishable particle removal
technologies (or in the case of UVGI, a technology for
inactivation of biological threat agents) that were considered
practical for consideration in a building HVAC (or collective
protection) application. Technologies such as wet scrubbers,
which can be used to remove aerosolized particles from
airstreams, were not considered reasonable to consider
in the assessment. Articles, conference proceedings, and
manufacturers' data relevant to the performance of each
technology, variables that affect the performance of each
technology, the impact of the technology on an HVAC
system, and cost parameters were sought.
-------
3.3 Technology Reviews
Each review provides a description of the technology, a
summary of the available literature, and a critical assessment
that addresses technology performance, trends that affect
performance, the impact of the technology once installed in
an HVAC system, and cost analysis. Section 10 of this report
details areas in which data were lacking that would have
been helpful in further assessing the building air cleaning
technologies.
3.4 HVAC Particle Removal System Background
To aid the discussion of the technology reviews and critical
assessments, background regarding particle removal is
presented below. First, a discussion of particle size of interest
and consideration is given. This is followed by a discussion
of the American Society of Heating, Refrigerating, and Air-
Conditioning Engineers (ASHRAE) standard test method
52.2 and the minimum efficiency reporting value (MERV)
ratings. The performance of all air cleaning technologies
that remove particles from an air stream, regardless of
application (i.e., regardless of whether to improve air quality
or to remove a hazardous aerosol from a terrorist incident),
depends on the particle size (diameter). Thus, the particle
size of interest is important when comparing and assessing
air cleaning technologies. Air filtration technologies used
in HVAC applications are tested against American National
Standards Institute (ANSI)/ASHRAE 52.2. ASHRAE
52.2 allows filters to be tested in a consistent fashion and
performance ratings to be assigned. Summaries of that
method and the performance ratings are provided.
3.4.1 Particle Size Considerations
The typical size ranges of commonly occurring aerosols
are shown in Figure 1. In general, particles of biological
origin range in diameter from less than 0.1 micron up to
about 50 um. Bacillus anthracis spores are between 1 and 2
microns in diameter (Carrera, 2006).
cojnmoji
gas mok-
vin
con
oil co
.tobacco
jcules
ises
ijjustioiu
nbustion
smoke ..
ir
-^
luclei
clouds
secticide
bacteria
milled
.fog
dust
pollens
_Jiair_
flour
£each S)
jnd
1E-4 1E-3 0.01 0.1 1 10 100 1000 10000
aerosol particle diameter, microns
Figure 1. Typical Size Ranges of Commonly Occurring Aerosols (Owen et al., 1992)
-------
Different air cleaning devices rely on different mechanisms
to capture particles of varying size, as discussed in
more detail under mechanical filtration in Section 5.4.1.
To summarize, relatively large particles, larger than
approximately 0.5 um in diameter, are collected mostly
by inertial effects and sedimentation, complemented by
interception at the boundaries of the collection elements,
while collection of smaller particles (<0.1 um) is mostly
due to diffusion. In the range from approximately 0.1 um
to 0.5 um, none of these mechanisms dominates, which
often results in a point of minimum collection efficiency
(n) in the air cleaner performance curve. Air cleaning
mechanisms that rely, at least in part, on the electrostatic
force acting on charged particles in an electric field
can affect the most penetrating particle size associated
with filters that rely solely on mechanical filtration
mechanisms. A larger charge-to-mass ratio is usually
achieved for smaller particles, which, therefore, increases
the collection efficiency of these smaller particles.
3.4.2 ASHRAE 52.2 and MERV Ratings
In the United States, ANSI/ASHRAE Standard 52.2-1999
(ASHRAE, 1999) is the standard method that is used to
evaluate and rate HVAC filters. Although the ASHRAE
52.2 standard was designed for assessing mechanical
filtration, the basic concept can be used to assess all types
of air filtration devices. This standard describes the test
procedures used to evaluate filters with capacities between
472 and 3,000 cfm (13 and 85 mVmin). Potassium chloride
particles in water are generated for the challenge aerosol.
The concentrations upstream and downstream of the filter
are measured to determine collection efficiency. These
penetration measurements are taken with the new filter at
four regular intervals as the filter is loaded with a specific
test dust until its pressure drop has doubled. The measured
collection efficiencies are then averaged over three particle
size ranges (0.3 to 1.0 um, 1.0 to 3.0 um, and 3.0 to 10.0 um)
to classify the filter into one of 16 classifications referred to
as minimum efficiency reporting values (MERVs). Table 1
lists the definitions of the various MERV ratings. Four MERV
categories (17-20) are included in this list to demonstrate
where High Efficiency Paniculate Air (HEPA)TUltra Low
Penetrating Air (ULPA) filter performance rates, however,
ASHRAE 52.2-1999 is not intended to be used to evaluate
HEPA filter performance. Therefore, as indicated in Table 1,
a different standard (IEST RP-CC001.3, 1993) is required
to properly classify HEPA filters. Also, as noted in Table
1, filters with a MERV rating of less than 5 must be tested
per ANSI/ASHRAE 52.1-1992 in order to determine their
MERV rating. For filters with MERV ratings of less than
5, the average arrestance of the filter must be measured.
As described in ANSI/ASHRAE 52.1-1992 (ASHRAE,
1992), the average arrestance of the filter is measured by
exposing the filter to a relatively coarse dust (composed of
72% standardized air cleaner fine test dust, 23% powdered
carbon, and 5% number 7 cotton linters) that has an average
particle diameter and concentration significantly higher
than for typical atmospheric dusts. The total mass fed and
the downstream mass that penetrates the filter are measured
at four regular intervals as the filter is loaded with the test
dust. The arrestance is determined from the ratio between
the total dust collected downstream of the filter and the
total dust fed. The average arrestance is then determined
by weighting the individual arrestances by the amount of
dust fed to the filter between successive measurements.
The average arrestance is then used to determine the
MERV rating of the filter, as shown in Table 1. Due to the
relatively new acceptance of ASHRAE 52.2-1999, the
performance of some HVAC filters is still reported by some
manufacturers using ASHRAE 52.1-1992. A recent study
(Burroughs, 2004) has shown that filters with MERV ratings
of less than 7 do not provide sufficient particle reduction to
avoid particle accumulation in the duct system. Examples
of typical applications for filters of various ranges of
MERV ratings are listed below (Spengler et al., 2000).
MERV 1-4 Residential: pollen, dust mites
MERV 5-8 Industrial: dust, molds, spores
MERV 9-12 Industrial: Legionella, dust
MERV 13-16 Hospitals: smoke removal, bacteria
MERV 17-20 Clean Rooms: surgery, chem-bio, viruses
-------
Table 1. Minimum Efficiency Reporting Value (MERV) Parameters (ASHRAE, 1999)
Standard 52.2
Minimum Efficiency
Reporting Value
(MERV)
Composite Average Particle Size Efficiency
(%) in Size Range (urn)
0.3- 1.0 (E)
1.0-3.0 (E2)
3.0-10.0 (EJ
Comments
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Ej<75
7599.97% for 0.3 urn
particles, IEST Type A
>99.99% for 0.3 urn
particles, IEST Type C
>99.999% for 0.3 urn
particles, IEST Type D
>99. 999% for 0.1-
0.2 urn particles, lESTType
F
-------
4.0
Critical Assessment of Mechanical Filtration
4.1 Technology Description
Mechanical niters are by far the most commonly used air
cleaning devices in both residential and commercial HVAC
systems. Mechanical niters come in a variety of shapes, sizes,
compositions, and forms; however, panel niters are the most
common. A rectangular frame containing a sheet of filter
medium, as shown in Figure 2, comprises a panel filter. Panel
filters come in an extremely wide variety of compositions,
sizes, capacities, and efficiencies. The filter medium used in
panel filters is usually constructed of woven or nonwoven
fibers and is composed of a wide variety of materials,
including glass, metal, synthetic (polymeric) materials (such
as polypropylene), paper, or woven fabrics (such as cotton
or nylon). The filters shown in Figure 2 have a 20 x 20 inch
(51x51 cm) cross-section and a 2-inch (5.1-cm) depth.
As illustrated in Figure 3, there are several varieties of
panel filters that have specific names and shapes. The filters
depicted typically have a cross-section of approximately
61 x 61 cm and 30 cm depth. Cube filters have five filtration
surfaces and can be force fit into an airstream duct without
requiring clips, latches, or any type of frame. Shown in
Figure 3 is the "pocket" filter or "multi-bag" filter. These
are specific types of panel filters that use filter pockets or
bags instead of a flat filter surface. The air flows through the
pocket walls while the particles are collected inside. These
filters are generally claimed to have higher dust loading
capacities than standard panel filters because of their depth-
loading nature. Also shown in Figure 3 is the "extended
surface" filter. These filters generally use pleated paper
media with aluminum separators. They generally have high
collection efficiencies (up to HEPA) and higher initial and
final pressure drops than other panel filters.
Figure 2. Typical Panel Filters
-------
"Pocket" Filters
"Cube" Filters
fc:
Double header
model
Single deader
model
Rigid or Extended Surface Area Filters
Figures. Unusual Panel Filter Shapes
The size of a panel filter depends on the airflow rate it is
intended to handle; however, panel filters generally remain
less than a square yard in cross-sectional area and handle
flow capacities of less than 2,500 cfm (71 mVmin). Banks of
panel filters are used in high airflow filtration applications.
Pressure drops of clean panel filters range between 0.05 and
1 in. w.g. (12 and 249 Pa), and (except for high efficiency
filters) the filters are generally replaced when the pressure
drop reaches between 3 and 10 times the original pressure
drop. In applications where high efficiency filtration is
required, a sequence of increasingly efficient panel filters is
generally employed to lengthen the service life of the more
expensive high efficiency filters.
Fibrous media are commonly used in air filtration because
they can provide good filtration efficiency at a low pressure
drop. The fibers are either woven into the filter frame
(woven) or randomly oriented and thermally or chemically
bonded to each other and to the filter frame (nonwoven). The
collection efficiency of fibrous media is directly related to
the average fiber diameter, the media density, and the media
depth or thickness. The smaller the fiber diameter, the smaller
the particles that can be collected. Increasing the density
and thickness increases the collection efficiency but also
increases the pressure drop. Fiber diameters in fibrous media
generally range between submicron and hundreds of microns.
Nonwoven fibrous media can be produced by the wet-laid
process, which is a modification of the normal paper-
making process. In this process, a slurry of fibers and water
is introduced to a porous base. The water drains, while
the fibers are collected and dried. The fibers tend to lie in
the same plane but are randomly oriented. These fibers,
when dried, form a continuous sheet of filter media. Glass,
cellulose, and other materials can be used to make filter
media using this method.
-------
Spunbond nonwoven filter media are formed from continuous
filaments that are extruded, drawn, laid into a filter web,
bonded together, and collected in a roll goods form on a
single process line. They are made from a wide range of
polymers, including polyethylene, polypropylene, polyester,
nylon, or combinations thereof. They are bonded together
by thermal bonding, chemical bonding, or needlepunch
fiber entanglement bonding. Because spunbond media are
composed of large diameter (20 to 250 um) fibers, they do
not collect small particles effectively and are generally used
as support structures for more efficient media.
Carding is another method of nonwoven fibrous filter
media production. Carding is a process in which fibers are
repeatedly combed with metal hooks to disentangle fiber
clumps. Carded fibers, which are roughly aligned, can be
compressed into a filter medium. Wool, cotton, and synthetic
fiber media can be produced in this way. Carded filter media
generally have fiber diameters larger than 15 um and are
generally very weak in the fiber plane direction but very
strong perpendicular to the fiber orientation. Carded filters
are generally weak and are not used in sheet form. Carded
filters can be "felted," a process in which the irregular
surfaces of the fibers are used to hold the fibers together by
the application of heat, humidity, and pressure. Synthetic
carded filters cannot be felted but can be "needled" together
with barbed needles. These processes improve the filter
media strength but often lower collection efficiency.
Metallic media are generally produced by weaving or
sintering. Because of the strength of these media and the lack
of any chemical sealant or bonding materials, metallic filters
are very corrosion resistant. In the meltblown production
process, polymer pellets are fed into an extruder where they
melt, pass through a metered pump, pass through an array of
thin "capillary" tubes, and then into a high velocity hot air jet
and are deposited on a collecting drum. The fibers thermally
bond, generally have thicknesses between 1 and 15 um, and
form high porosity webs. They possess clean, unused (virgin)
surfaces, which makes them ideally suited for filtration.
The fiber diameter and web thickness can be controlled by
altering the pump flow rate, the size of the capillary tubes, the
velocity of the air jet, and the speed of the collection step.
In the United States, ANSI/ASHRAE Standard 52.2-1999
(ASHRAE, 1999) is the standard method used to evaluate
and rate HVAC filters. This fairly recent standard describes
the test procedures used to evaluate filters with capacities
between 472 and 3,000 cfm (13 and 85 mVmin). Table 2
lists the definitions of the various MERV ratings, along with
the minimum final resistance of each rating. An additional
discussion regarding ASHRAE 52.2 and MERV rating is
provided in Section 3.4.2.
In the residential market, the most inexpensive filters
dominate. These include fiberglass, disposable polyester/
cotton blends, and pleated air filters. The lowest MERV-
rated filter identified for residential use was 4, and the
highest rated filter available in the residential market was 12
(manufactured by 3M). Electrostatic filters were found to
be dominant for the medium to higher efficiency residential
filters. In fact, it was quite difficult to identify a residential
filter with a MERV rating of 11 or greater that did not possess
electrostatic media.
A much larger range of filters was identified in the
commercial market. The most popular design in commercial
applications is the pleated air filter. In the commercial
market, it was not difficult to identify filters with MERV
ratings between 1 and 15. MERV 16 filters were more
difficult to identify than the other classes of filters, but some
were identified.
-------
Table 2. Minimum Efficiency Reporting Value (MERV) Parameters With Minimum Final Resistance (ASHRAE, 1999)
Standard 52.2
Minimum
Efficiency
Reporting Value
(MERV)(a>
Composite Average Particle Size Efficiency (%)
in Diameter Range (urn)
0.30-1.0 (EJ 1.0-3.0 (E2) 3.0-10.0 (E3)
Minimum Final Resistance™
in. of water
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
El<75
75
-------
Particle fractional penetration (P), expressed in terms of
fractional efficiency (E) by P=1-E, is dependent on the
diameter of the aerosol particles, filtration velocity (based
on flow rate and available filter surface area), and filter
parameters, including the thickness, fiber diameter, and
solidity or fiber packing density (i.e., the ratio of solid fiber
volume to gross filter volume). The general trends regarding
the impact of these parameters on the filtration efficiency
and airflow resistance are summarized in Table 3. As shown
in Figure 5, the collection efficiency of filter media depends
strongly on particle size. The most penetrating particle size
(MPPS) for high efficiency filters is typically in the range of
0.1 to 0.3 urn.
The fiber diameter and packing fraction affect the filtration
performance and the airflow resistance of the media. In
general, the efficiency and resistance both increase as the
fiber diameter decreases or packing fraction increases.
High efficiency filters typically contain blends of fibers of
varying diameter to satisfy the filtration requirements and to
provide the physical strength. For example, High Efficiency
Paniculate Air (HEPA) filter media are typically less than
0.5 mm thick, have packing densities ranging from 0.03 to
0.05, and contain fibers of diameters ranging from 0.55 um
to approximately 6.5 um but possess a nominal mean fiber
diameter of approximately 0.65 to 0.70 um.
The efficiency of a filter is typically inversely proportional
to the filtration velocity (air velocity through the filtration
media) for particles of all diameters. The only exception
would be those particles that are collected predominantly by
diffusional mechanism, typically those less than 0.1 um. As
the velocity increases, the diffusion mechanism becomes less
effective and the interception and impaction mechanisms are
enhanced. In any case, filters should always be used at the
filtration velocity recommended by the manufacturer.
In general, particle penetration of a filter decreases
(sometimes by as much as an order of magnitude) as particles
are collected on the filter surface. The collected particles
form an additional layer (termed a filter cake or dust cake)
on the filter surface, which contributes significantly to the
collection efficiency of the filter. The filter cake does not
increase the filter's collection efficiency indefinitely, but
collection efficiency increases quickly as particles are first
collected and then levels off. Figure 6 illustrates the effect
of particle loading on the collection efficiency of a MERV 9
filter. Of course, the airflow resistance across the filter also
increases with the loading of particles on the filter as portions
of the filter surface area become clogged.
Table 3. Influence of Filter Parameters on Filtration Efficiency and Airflow Resistance
Parameter Penetration Airflow Resistance
Fiber Diameter
Thickness
Solidity
Surface area
Decreases with decreased fiber diameter
Decreases with increased thickness
Decreases with increased solidity
Decreases with increased surface area
Increases with decreased fiber diameter
Increases with increased fiber thickness
Increases with increased solidity
Decreases with increased surface area
Inertia!
impaction
and
interception
regime
Diffusion and
interception
regime
0.01 0.1 1.0
Particle diameter (microns)
Figure 5. Primary Mechanisms of Capture for Various Particle Diameters (CDC, 2005)
-------
100
90
80
—, 70
•sP
Jeo
| 50
o
IE 40
LJJ
30
20
10
0
0.1
T
i—i i i i 111 r
1.0
i i I i i i
Geometric particle size (micrometers)
10
Figure 6. ASHRAE Standard 52.2 Test Data for a MERV 9 Filter Showing How
Collection Efficiency Increases as the Filter Loads (CDC, 2005)
Excessive heat can affect the performance of niters by
causing degradation of the materials that bond the filter
fibers, the materials used to bond the media to the frame, or
even the filter media itself. In general HVAC applications,
excessive heat is unlikely to be a significant concern, but care
should be taken to ensure that the filters are suitable for the
likely operating temperatures.
Relative humidity also can affect the performance of
mechanical filters. Under unusual conditions, excessively
high relative humidity can affect the collection efficiency and
airflow resistance of mechanical filters as water can condense
onto the filter fibers and collected particles, causing a rapid
increase in pressure drop that could damage the filter, or even
cause the filter to burst. Therefore, care must be taken to
ensure proper design of HVAC systems to prevent exposure
of mechanical filters to saturated or supersaturated airstreams.
4.3 Summary of Relevant Studies
Recent studies in mechanical media are summarized in
Table 4. Note that Table 4 focuses primarily on studies
conducted since 1995. Only important and representative
studies conducted before 1995 are included in Table 4.
The residential furnace filtration market has seen a large
increase in the number and variety of available filters (Fugler
et al., 2000). Consumers now have the option to purchase
anything from a traditional mechanical filter made of
recycled material to a HEPA filter for their HVAC system.
Several studies were identified that provided background
information on both filter performance and contaminants
found in the air (Brown, 2001; Miller, 2002; Kowalski and
Bahnfleth, 2002) to help homeowners select the correct
balance of efficiency versus cost for their particular needs.
Filtration demands in commercial environments vary with the
type of environment that uses the filtered air. For example,
ASHRAE recommends a 90% average dust spot efficiency
filter preceded by a 25% dust spot efficiency filter for general
areas of hospitals, a 25% dust spot efficiency filter for the
administrative areas, and a 25% dust spot efficiency filter,
90% dust spot efficiency filter, and HEPA filter in series
for an operating room (Kowalski and Bahnfleth, 2002).
The average commercial building will not meet these strict
guidelines, and certain overviews (Miller, 2002; Kowalski
and Bahnfleth, 2003) were identified to assist building
owners in selecting an adequate filtration system.
The dust spot efficiency does not directly convert to a MERV
rating due to the differences in the standard test methods. The
dust spot efficiency represents an overall efficiency while
the MERV is based on particle-size dependent efficiencies.
For ASHRAE 52.2, potassium chloride particles ranging
in size from 0.3um to lOum are used to test the filter, with
penetration determined in 12 distinct size bins between
0.3um and lOum. For the previous standard, ASHRAE
52.1, collection efficiency was measured by two methods:
arrestance and dust spot efficiency. Arrestance is measure
by weighing the fraction of a synthetic test dust that passes
through the filter. Dust spot efficiency is measured by
comparing opacity meters upstream and downstream of the
filter. Since collection efficiency is strongly related to particle
size and the particle sizes of the materials used in the three
tests are quite different, the tests are not directly comparable.
However, Table 5 can be used as a general guideline. For
example, a filter with a dust spot efficiency of 30% roughly
corresponds to a MERV 8 filter.
-------
Table 4. Summary of Mechanical Filter Studies
Basic Scope Content/Conclusion Reference
Overview of Filtration, Filters, or
Contaminants
Performance Data of Mechanical
Filters
Effect of Parameters on Performance
Innovations in Mechanical Filters
Impact in Actual Residential or
Commercial Environments
Designing an Air Filter System
Inert versus Bioaerosol Filtration
These papers provide an overview of mechanical filtration,
including standards, capabilities of current filters, types of
media, and potential contaminants.
These studies provide experimental data on filtration
performance for a wide variety of residential/commercial
filters, tested in the laboratory according to ASHRAE
standards.
In these studies, different parameters of a filter setup were
adjusted to test the effect on performance, including the
orientation of different filters in series, variable housing
geometry of a filter, and filter media.
These studies and patents explore promising new
developments, techniques, or equipment:
Filter immersed in a liquid
New hygroscopic filter media
Indoor air purification system (patent)
NBC-Building protection system and method
(patent)
These studies provide data on the impact of mechanical
filters in actual home or commercial environments.
These papers discuss the concept of Clean Air Delivery Rate
(CADR) to help optimize air filter needs.
These studies compare collection efficiency of bioaerosols
compared to inert aerosols of comparable particle size.
Brown, 2001; Miller, 2002;
Kowalski and Bahnfleth, 2002;
Kowalski and Bahnfleth, 2003
Rivers and Murphy, 2000; Owen
etal., 2003
Chambers et al., 2001; Peters
etal., 2001; Letts etal., 2003
Agranovski et al., 2001
Kemp etal., 2001
Homeyer et al., 2001
Fuchsetal., 2004
Howard-Reed etal., 2003;
Wallace et al., 2004; Chimack
and Sellers, 2000; Fugler et al.,
2000
Rudnick, 2004; Ward etal.,
2003
Brosseau et al., 1994; Willeke
et al., 1996; McCullough et al.,
1997; Hofacreetal., 1996
Table 5. Comparison of ASHRAE Standards 52.1 and 52.2
(a)
ASHRAE 52.2 ASHRAE 52.1 Particle
Particle size range Test size range, Applications
MERV 3 to 10 Mm 1 to 3 urn .3 to 1 urn Arrestance Dust spot u™
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
<20%
<20%
<20%
<20%
20-35%
35-50%
50-70%
>70%
>85%
>85%
>85%
>90%
>90%
>90%
>90%
>95%
-
-
-
-
-
-
-
-
<50%
50-65%
65-80%
>80%
>90%
>90%
>90%
>95%
-
-
-
-
-
-
-
-
-
-
-
-
<75%
75-85%
85-95%
>95%
<65%
65-70%
70-75%
>75%
80-85%
>90%
>90%
>95%
>95%
>95%
>98%
>98%
>98%
>98%
>98%
>98%
<20%
<20%
<20%
<20%
<20%
<20%
20-25%
25-30%
40^15%
50-55%
60-65%
70-75%
80-90%
90-95%
-95%
>95%
>10
3.0-10
1.0-3.0
0.3-1.0
Residential
light pollen
dust mites
Industrial
dust
molds spores
Industrial
Legionella
dust
Hospitals
smoke removal
bacteria
'a'www.cdc.gov/niosh/docs/2003-136/2003-136c.html
-------
4.3.1 Performance and Variables That Affect
Performance
Filter performance information is currently provided by
either an ASHRAE rating or a manufacturer's claim. Recent
literature (Rivers and Murphy, 2000; Owen et al, 2003) has
helped make public significant amounts of data on residential
and commercial filter efficiency test results using ASHRAE
52.2-1999. Owen et al. (2003) tested 26 different filters,
both residential and commercial, in the laboratory according
to ASHRAE 52.2-1999. MERV ratings were determined
from the test data, based on the assumption that an initial
efficiency curve would represent the lowest efficiency values
of each filter in an ASHRAE 52.2-1999 test (Owen et al.,
2003). Particle diameters in the test ranged from 0.03 to
10 urn, as compared to the standard ASHRAE 52.2-1999 test
range of 0.3 to 10 um. The lowest efficiencies for all filters
were located in the 0.1 to ~0.5 um diameter range, with the
higher-rated filters showing an increase in efficiency as the
particle diameter increased in the particle size range (Owen
et al., 2003). The filters with lower MERV ratings showed
a definite increase in efficiency when the particle diameter
increased but exhibited an inconsistent performance as
the diameter decreased (Owen et al., 2003). The data were
inconsistent in that both increases and decreases in collection
efficiency have been reported with particle diameters of less
than 1 um.
Rivers and Murphy (2000) conducted a series of laboratory
tests on 31 different air filters with a variety of ASHRAE dust
spot efficiencies. The tests included the ASHRAE Standard
52.1 tests, filtration efficiency evaluations for different
particle diameters, and reentrainment tests to determine
whether the filter media contributed to the downstream
concentration of particles. The filters were evaluated under
both constant airflow and variable air volume (VAV) flow.
The goal of the study was to be able to predict air filter
resistance and efficiency in VAV systems. As a result, much
of the paper is devoted to generating that model as opposed
to presenting laboratory test results. Rivers and Murphy
(2000) concluded that there was no significant drop in filter
performance under VAV test conditions, with the exception
of the lowest efficiency filters showing a greater loss of
collected dust. These filters, however, showed high dust
losses under both operating conditions and could not even
withstand all of the tests (Rivers and Murphy, 2000).
The Environmental Technology Verification (ETV) Program,
established by the U.S. Environmental Protection Agency,
conducted laboratory tests on the performance of air filters
in building HVAC systems. The results for fourteen of
these evaluations, focusing primarily on pressure drop and
filtration efficiency of bioaerosols and inerts, can be found
on the EPA National Homeland Security Research Center
Web site (US EPA, 2004). The filters were assigned MERV
ratings based on the ASHRAE Standard 52.2 test from 0.3 to
10 um diameter particles. The results showed an increase in
filtration efficiency of bioaerosols with MERV rating, with
higher efficiencies for dust-loaded filters. There was also a
general increase in pressure drop with MERV rating. As with
the previous references, the ETV data are laboratory data and
do not represent actual HVAC impact data.
4.3.2 Assessment in an HVAC System
As noted in Table 4, a number of papers were identified that
examined the impact of mechanical filters on the air quality
in actual use environments (Howard-Reed et al., 2003;
Wallace et al., 2004; Burroughs, 2004; Chimack and Sellers,
2000; Fugler et al., 2000). These papers are discussed in turn
in the following sections.
Residential Environment. Filter efficiency in an actual
residential environment was evaluated by Fugler et al.
(2000) by comparing the performance of various furnace
filters in six different test houses. Ten filters were used in an
initial test house, and then five of these filters were chosen
to be evaluated in the remaining five houses. Table 6 lists
the efficiencies for these five filters, four of which were
mechanical, after being tested in six houses.
The efficiencies in Table 6 were calculated for each particle
diameter, based on upstream and downstream conditions.
Experimental efficiencies were also found at PM5 (mass
of particles below 5 um) but were not provided because
all diameters showed similar results to all tests (Fugler et
al., 2000). The calculated efficiencies were meant to be
comparable to ASHRAE dust spot efficiencies.
For the most part, the experimental data provided efficiencies
that were similar to the filters' rated efficiencies. The 25-mm
high quality media filter exhibited a range of 29 to 45%. As
seen in Table 6, it was claimed to be 20 times better than an
ordinary filter and 7 times better than an ordinary pleated
filter. The results compare well to the ordinary furnace filter
that showed a negative efficiency in the test house and a
regular 25-mm pleated media filter that showed an efficiency
of around 5% in the test house (Fugler et al., 2000). It
is unclear how the electronic charged pad experimental
efficiency data relate to the manufacturer's claimed
performance. Both the 100-mm pleated and ESP filters
performed at the level claimed by the manufacturer. The
HEPA and TFP filters did not reach their expected efficiencies
because they were used as bypass units, filtering only about
30% of the system air (Fugler et al., 2000). These filters did,
however, operate at their expected efficiencies for the fraction
of air that they handled (Fugler et al., 2000).
Fugler et al. (2000) also evaluated filter impact in
a residential setting by measuring the reduction in
concentration of respirable particles in the indoor
environment. Table 7 lists the reduction of these particles
based on percent improvement compared to a no-filter
condition, with the experimental efficiency also provided for
reference (Fugler et al., 2000).
-------
Table 6. Filter Results vs. Manufacturers' Claims (Fugler et al., 2000)
Test Results, Upstream/Downstream Efficiency'3'
Filter Description Manufacturer Claimed Performance Eo/0 PM1(b) E%PM10(b>
25-mm high quality
pleated media
Electronic charged pad
100-mm pleated media
Electronic plate and wire
type (ESP)
HEPAorTFP(c)
20 times better than ordinary filters, 7 times
better than ordinary pleated filters
Efficiency at 0.3-0.5 urn: 33-75%;
0.5-1.0 urn: 75-95%
32% average dust spot efficiency
75% average dust spot efficiency
99.97% DOP(H EPA)
84% - 99%, based on particle diameter (TFP)
29%
17%
21%
84%
27%
45%
20%
36%
90%
30%
(a) The upstream/downstream efficiency technique is based on the mean particle concentrations at the upstream and downstream sampling points
over the duration of the data period. It is meant to be comparable to dust spot efficiency, an ASHRAE evaluation method.
(b) PM1 and PM10 represent mass of particles below 1 and 10 pm diameter, respectively.
(c) HEPA or TFP (turbulent flow precipitator) filters were both used as high efficiency filters in the study but did not reach their expected
efficiencies because they were used as bypass units, filtering only about 30% of the system air.
Table 7. Mean Reduction of Indoor PM10 Levels (a) Below "No-Filter" Case (Fugler et al., 2000)
. Experimental Percent Improvement
Filter Description r«. .
Efficiency Active Nonactive
100-mm pleated media
Electronic charged pad
Bypass filter (HEPA or TFP)(b)
25-mm pleated media high quality
Electronic plate and wire type (ESP)
36%
20%
30%
45%
90%
9%
9%
23%
21%
31%
13%
29%
38%
57%
71%
(a) Although this table is for PM10 only, Fugler et al. (2000) reported similar results for all diameter ranges.
(b) The HEPA and TFP filters did not reach their expected efficiencies because they were used as bypass units, filtering only about 30% of the
system air.
The percent improvement is a measure of how the
concentration of PM10 particles in rooms of the house
was lowered when a filter was used in the home HVAC
system. Each filter's improvement is relative to a no-filter
condition, in which the fan runs but no filter is in the duct.
Active data refer only to periods of known activity in the
house, causing the resuspension or generation of particles
on real-time data charts. Nonactive data include periods
when there is no activity in the house, such as when the
occupants are sleeping or the house is unoccupied. As seen
in Table 7, percent improvements during active periods were
consistently much lower than the efficiencies. The exception
is the bypass filter, but the 23% reduction was the result of
handling only 30% of the system air. As would be expected,
reductions in concentration were greater during periods
of inactivity than activity (Fugler et al., 2000). Two of the
filters showed a reduction lower than the filter efficiency,
while the remaining three filters showed a reduction greater
than the filter efficiency. The filters analyzed by Fugler et al.
(2000) performed well when compared to their ASHRAE or
manufacturer claimed efficiencies. Their impact in reducing
the amount of indoor particles was much less significant than
their single-pass efficiency during active periods. Inactive
periods showed more comparable results, as some percent
reductions were higher than single-pass efficiencies.
Howard-Reed et al. (2003) investigated the impact of central
heating and air conditioning (HAC) forced-air fans and in-
duct filters on the deposition rates and reduction of particle
concentrations in a residential environment. Deposition rate
includes removal of particles by deposition to room surfaces,
removal by operation of the HAC fan (when the fan is on),
and removal by a mechanical or electrical air cleaner (when
a filter is used). The study took place in an occupied three-
story townhouse over several years. Three different sources
(cooking with a gas stove, burning a citronella candle, and
pouring kitty litter) were used to introduce particles of
varying shape and composition into the home. The deposition
rates and particle reductions were calculated for four HAC
configurations: (1) fan off, no filter; (2) fan on, no filter; (3)
fan on, typical furnace filter; and (4) fan on, electrostatic
precipitator (ESP) unit.
The results of the experiment showed the deposition rates
for each particle diameter range were not influenced by the
three different sources (Howard-Reed et al., 2003). The
particle deposition rates did, however, vary with both particle
diameter and HAC configurations, as seen in Figure 7
(Howard-Reed et al., 2003).
-------
^ 9-
e 8-
<3
"^ 7 -
c 6 •
o
1 5-
o
Q
4> 3-
« 2-
"^ 1-
0-
,
2
T
1
1 Fan Off
2 Fan On w/o Filter
3 Fan On w/Furnace Filter
4 FanOnw/ESP
3
f
1
t R h
,.
. rf
,t
3
T
..
4
f
ft
2
f
3
••
4
rr
1
I
2
1
1
•
2
0.3-0.5 0.5-1.0 1.0-2.5 2.5-5.0 5.0-10 >10
Particle Diameter (\im)
Figure 7. Comparison of Particle Deposition Rates for Different HVAC
Configurations
Error bars are ± one standard deviation. The standard niters
and ESP were used only in the cooking and candle burning
source events, resulting in a lack of measurable decay rates
for the larger diameter particles. (Howard-Reed et al, 2003)
Deposition rate clearly increased as particle diameter
increased. The fan on, no filter configuration showed a
marked increase over the fan off condition. The addition
of a standard furnace filter demonstrated very little
improvement over the fan on, no filter configuration.
The addition of an ESP proved to have a very significant
influence on particle deposition rates. The standard
filters and ESP were used only in the cooking and
candle burning source events, resulting in a lack of
measurable decay rates for the larger diameter particles.
A similar experiment in a test house was conducted as part
of the study to investigate the effects of room surface area
and furnishings. An OFF!® citronella candle was burned in
an unfurnished room with a similar floor surface area and
slightly higher volumetric flow rate through the fan system.
The results provided no significant deviations from the
furnished townhouse (Howard-Reed et al., 2003).
The percent reduction of particles among the test conditions
was estimated based on the mean deposition rates (Howard-
Reed et al., 2003). Table 8 lists the percent reductions in
particle levels for both a "tight" house (air change rate of
0.2 h'1) and a "drafty" house (air change rate of 1.0 fr1).
Simply turning on the fan in the central heating and air
conditioning unit had a significant effect in reducing indoor
particles. The addition of an ESP filter had an even greater
effect. The effect of the standard filter was too insignificant to
be included in the table (Howard-Reed et al., 2003).
Table 8. Percent Reduction of Outdoor Particles Penetrating Indoors (Howard-Reed et al., 2003)
Ventilation/Filtration Particle Diameter Range (urn)
Setting'3' 0.3-0.5 0.5-1 1-2.5 2.5-5 5-10 > 10
Tight house
Fan on
Fan on, ESP
Drafty house
Fan on
Fan on, ESP
48%
84%
27%
67%
47%
85%
28%
70%
42%
77%
29%
64%
53%
67%
43%
57%
42%
-
36%
-
26%
-
23%
-
(a) The ESP filter was used only for cooking and candle burning (smaller particles) so there is a lack of measurable decay rate for larger diameter
particles (Howard-Reed et al., 2003).
-------
10
0.1
SMPS APS ^ „"-«-._ ^
1 «
-J.^ ..'.
* ESP filter
• Mech filter
• Fan on, no filter
• Fan off
0.01
10
0.1 1
Particle diameter ((Jim)
Figure 8. Deposition Rates by Particle Diameter for Each Test Configuration (Wallace et al., 2004)
Wallace et al. (2004) performed an extension of the Howard-
Reed et al. (2003) study in the same townhouse to include
particles of a much smaller diameter and also replaced
the standard filter with a higher quality mechanical filter
(Wallace et al., 2004). The fibrous mechanical filter, or
MECH, had an extended surface area and an ASHRAE
Standard 52.1 average arrestance of 93%. The same four
HAC configurations were tested. Figure 8 illustrates how the
deposition rates vary with particle diameter for the four test
configurations.
Two particle sizing instruments, a Scanning Mobility Particle
Sizer (SMPS) and an Aerodynamic Particle Sizer (APS).
were required to gather the data over the proposed particle
diameter ranges. Therefore, the graph is split into two halves.
one for each instrument. As in the Howard-Reed et al. (2003)
study, use of the central forced-air fan led to the reduction
of indoor particles, and the addition of a filter reduced them
further. The MECH filter did show an improvement over
the no-filter condition, especially at the smallest particle
diameters. The ESP filter exhibited the best performance
but required cleaning (at intervals between 500 and 2000
hours) to maintain the high level of performance (Wallace
et al., 2004). The graphs have minimum deposition values
at a range of 0.11 -0.13 um, which agrees with theoretical
predictions (Wallace et al., 2004).
As in the Howard-Reed et al. (2003) study, the percent
reductions of particles among the test configurations were
estimated. The values, presented in Table 9, are based on the
mean deposition rates.
The air change rates for the tight, typical, and drafty houses
were 0.2 fr1, 0.64 fr1, and 1.2 fr1, respectively. The typical
air change rate was determined by the average rate in the
house over the course of a year. The table shows that simply
running the fan with no filter has a significant reduction effect
on indoor particles. The MECH filter produces an additional
reduction effect, as does the ESP filter.
Commercial Environment. Few studies were found that
included data from actual commercial environments. One
such study by Chimack and Sellers (2000) compared the
Table 9. Percent Reductions in Particle Concentrations
Due to a Central Fan, MECH Filter, and ESP
(Wallace et al., 2004)
Tight house
Fan on
Fan on, MECH
Fan on, ESP
Typical house
Fan on
Fan on, MECH
Fan on, ESP
Drafty house
Fan on
Fan on, MECH
Fan on, ESP
Percent Reduction
18%
28%
59%
14%
23%
51%
11%
20%
44%
-------
use of an office building's current standard bag-type filters
to a premium bag-type filter with extended surface area and
a lower pressure drop. The results of the 40-week study
show that the premium filter was able to maintain the lower
pressure drop and would have a longer service life (Chimack
and Sellers, 2000).
A more in-depth study by Burroughs (2004) evaluated several
different filters, from MERV 5 to MERV 16, in a variety of
commercial environments for a period of one year. The study
included 55 different sites in 5 different cities, incorporating
a variety of filtration needs. The sites were grouped into
low (L), medium (M), and high (H) efficiency classes. Each
class had its own MERV-rated filters, as shown in Table 10
(Burroughs, 2004).
Paniculate loading in the air stream was identified and
compared in upstream and downstream conditions,
with data being acquired primarily at the filter bank.
The impact of the filters was presented as percent
reduction of particles for each diameter. The three
filter efficiency results are listed in Table 11.
The data gathered in the study produced a very large range
for all particle diameters and all efficiency classes, as there
are high standard deviations for the numbers in the table
(Burroughs, 2004). As shown in Table 11, higher MERV-
rated filters generally remove more particles from the filtered
space and produce higher quality air than lower MERV-rated
filters, validating the MERV ratings as adequate predictors
of field performance (Burroughs, 2004). Low-level filters
showed very slight percent reductions at small diameters
and moderate reductions at larger diameters. Other factors,
including particle counts and on-site buildup, demonstrate
that filters below MERV 7 do not adequately prevent particle
accumulation in the system (Burroughs, 2004). Medium-level
class filters provided a respectable improvement over the
low class levels. High-level MERV filters proved superior in
performance over the other two levels. They also, however,
exhibited some variations within their class for certain
particle diameters, as shown in Table 12 (Burroughs, 2004).
The detailed analysis of individual MERV ratings
shows definite variations in the high class performance,
in percent reduction by both particle diameter and
MERV rating. Despite these variations, all high
class MERV-rated filters maintained a superior
performance overall (Burroughs, 2004).
The literature review indicates that the use of mechanical
filters does, in fact, lead to the reduction of particles.
In general, filters with higher MERV ratings (or higher
expected efficiencies if not assigned a MERV rating)
performed better than filters with lower MERV ratings
when impacts in actual HVAC environments were
evaluated. Note that there is not a significant amount
of actual HVAC environment data available.
Table 10. Filter Types and Efficiency Classes (Burroughs, 2004)
Efficiency Filter Type and Description ASHRAE52.1 ASHRAE 52.2
Class ADSP % MERV
Low
Medium
High
Link-Panel - 1" using synthetic blanket media that may or may not be
enhanced
Conventional Pleated Filter - 2" or 4" pleat depth, low capacity
pleating
Pleated Filter - 2" or 4" pleat depth (An upgraded pleat using
enhanced electret media)
Pleated Filter - 2" or 4" pleat depth (An upgraded pleat using
enhanced electret media)
Medium Efficiency Extended Media Bag Filter - with pockets made of
synthetic media
High Efficiency Cartridges - minipleat type with high capacity pleat
High Efficiency Cartridges - minipleat type with high capacity pleat
High Efficiency Cartridges - minipleat type with high capacity pleat
High Efficiency Cartridges - with pockets made of synthetic media
High Efficiency Cartridges - minipleat type with high capacity pleat
<20%
<30%
30-35%
50-65%
40-55%
70-75%
80-85%
90-95%
90-95%
98%
MERV 5
MERV 6
MERV 8
MERV 11
MERV 9
MERV 12
MERV 13
MERV 14
MERV 14
MERV 16
ADSP - Atmospheric Dust Spot Discoloration Method (ASH RAE 52.1-1992)
Table 11. Summary of Percent Reductions of Particles (Burroughs, 2004)
Level Particulate Diameter (|jm)
0.3 0.5 1 3 5 10(a>
Low
Medium
High
2.3%
13.8%
41%
10.7%
23.6%
53.8%
26.4%
41.2%
67.7%
54%
67.7%
87.3%
65.7%
77.6%
92.3%
83.9%
88.3%
94.8%
"'Actual counts of these particles were too low to derive any statistical conclusions from the data (Burroughs, 2004).
-------
Table 12. Average Percent Reductions for MERV Ratings in High Class (Burroughs, 2004)
Particulate Diameter (urn)
MERV value
0.3 0.5 1 3 5 10
MERV 12
MERV 13
MERV 14
MERV 16
23.3%
50.5%
32.3%
84.5%
33.7%
59.2%
44.5%
93.5%
51.3%
76.7%
62.3%
96.1%
82.9%
90.5%
84.7%
97.1%
89.5%
92.9%
91.8%
97.3%
96.6%
93.1%
94.7%
98.6%
'"'Actual counts of these particles were too low to derive any statistical conclusions from the data (Burroughs, 2004).
4.3.3 Additional Factors
Filtration of Inert vs. Bioaerosols. Several studies have
compared measured penetrations of inert and biological
aerosols through masks and respirator filters (Brosseau
et al., 1994; Chen et al, 1994; Willeke et al., 1996;
McCullough et al., 1997; Wake et al., 1997; Qian et al.,
1998). Consistently, no significant difference in filter
penetration was found between spherical inert and spherical
bioaerosol particles of similar aerodynamic diameter,
suggesting that inert particles of the same size may be
used to predict bioaerosol efficiency. In some cases, the
penetration of rod-shaped bacteria was lower than that of
spherical organisms of the same aerodynamic diameter,
indicating that in addition to particle size, particle shape
also may affect penetration through respirator filters.
Relevant studies are highlighted below. Qian et al. (1998)
found the filtration efficiencies of N95 respirators were
equivalent when challenged with inert NaCl particles and
polystyrene latex (PSL) spheres, and two bioaerosols,
Bacillus subtilis (0.8 um) and B. megatherium (1.2 urn),
of the same mean aerodynamic diameter. These results
were consistent with Brosseau et al. (1994), who found
no significant difference in penetration though flatsheet
fiberglass filter media when challenged with PSL spheres
andMycobacterium chelonae (0.78 um) of the same
aerodynamic size. Chen et al. (1994) report this same
conclusion for experiments under equivalent test conditions,
with a surgical mask and several disposable dust-mist-fume
and HEPA respirators. Based on their results, Brosseau et
al. suggest that an inert aerosol with aerodynamic particle
size similar to a bioaerosol of interest is an appropriate
test aerosol to predict bioaerosol filter collection.
In a recent report by RTI International, Hanley and
Foarde (2003) assessed the filtration efficiency provided
by the C2A1 canister against inert KC1 particles and a
bioaerosol of B. globigii (Bg) spores. In nearly all test
cases, no Bg spores were detected downstream of the C2A1
canister. Results similarly demonstrated that inert aerosol
penetration in the 0.7 to 1.0 um range were consistent
with those measured using the Bg spore challenge.
Research efforts on filter efficiency have largely focused on
bacterial aerosols, thus limited information on viral aerosols
is currently available. Hofacre et al. (1996) measured the
penetration of PSL spheres (0.173 um) and a viral aerosol
of MS2 phage through a HEPA filter used for collective
protection. Aerosol penetration measured using the light
scattering technique was statistically equivalent for both the
MS2 phage and PSL particles, while the penetration of MS2
phage measured using the bioassay method was consistently
lower by a factor of approximately four. The study concluded
that inert aerosols of similar size provided a conservative
indication of HEPA filter performance against a bioaerosol.
Previous reports on filter performance against particles of
various sizes, shapes, and aspect ratios (a measurement
comparing the length and diameter of nonspherical
particles) contain conflicting results. Willeke et al. (1996)
compared penetrations through surgical masks and
dust-mist respirators of several rod-shaped bacteria (B.
megatherium, Pseudomonasfluorescens, B. alcalophilus)
of varying aspect ratios with that of spherical Streptococcus
salivarius and inert corn oil particles. Results indicate
the penetration of spherical S. salivarius bacteria were
approximately the same as spherical corn oil particles in the
aerodynamic size range from 0.9 to 1.7 um. The penetration
of rod-shaped bacteria was lower than S. salivarius and
decreased with increases in the aspect ratio, showing that
filter penetration of bacteria is a strong function of their
shape. Willeke et al. postulated that due to greater surface
area of nonspherical particles compared to spherical ones,
interception and electrostatic attraction cause greater removal
(i.e., less penetration) for rod-shaped bacteria compared
to spherical particles of the same aerodynamic size.
In comparison, McCullough et al. (1997) evaluated the
penetration of inert PSL spheres (0.55 um) andM. abscessus
(0.69 um), B. subtilis (0.88 um), and Staphylococcus
epidermidis (0.87 um) bioaerosols through a variety of
dust-mist-fume filters and surgical masks. In all cases,
penetration of the inert aerosol was greater than any of the
biological aerosols due to its smaller aerodynamic diameter.
However, results showed that B. subtilis (a rod) was more
penetrating than S. epidermidis (a sphere) at approximately
the same aerodynamic diameter. Contrary to Willeke et al.
(1996), the authors suggest that the aerodynamic diameter
of the bacteria may not be an accurate predictor of aerosol
penetration for nonspherical particles in these filters,
particularly when electrostatic forces are dominant.
Reaerosolization. Limited information is
available on the reaerosolization of particles from
filter materials. The most relevant studies were
conducted by Qianetal. (1997a, 1997b).
Qian et al. (1997a) evaluated the reaerosolization of inert
and biological aerosols from three different models of N95
-------
half-mask respirators at velocities of up to 300 cm/sec,
intended to represent violent sneezing or coughing. The
filters were loaded with aerosols of either sodium chloride
(NaCl) in water, PSL spheres, or the bacterium Bacillus
subtilis or B. megatherium. Reaerosolization was found to
be insignificant for particles of less than 1.0 um, with less
than 0.025% becoming reentrained. For the condition of
violent sneezing or coughing, only larger particles were
reaerosolized in significant amounts: about 1% of 3 um and
6% of 5 um PSL particles. This is of importance as single
bacteria may aggregate or attach to inert particles to form
larger clusters, increasing their risk of reaerosolization. In
addition, results indicate that bacteria become reentrained
at high air velocities more easily than the NaCl particles of
similar size, which the authors reason may signify a weaker
bond to the filter fibers or more surface area exposed to
the airflow. Finally, no reaerosolization of particles was
observed when the relative humidity was increased to 35%,
which the authors attribute to liquid bridging between
particles and filter fibers that increases the adhesion force.
In a related study, Qian et al. (1997b) assessed the effects of
particle size, inert particle type, filter type, and reentrainment
velocity on reaerosolization of inert aerosols including NaCl
in water, PSL spheres, corn oil, and dust. Test methods
similar to the previous study were employed. Flat sheets
of three types of fibrous filter media used in half-mask
respirators were evaluated. Reaerosolization trends were in
agreement with the previous study as reentrainment of 0.6
to 5.1 um particles increased approximately with the square
of particle size and the square of reentrainment velocity
and decreased with relative humidity. Reaerosolization
was also found to be a function of particle type as dust was
found to have the highest reentrainment, while corn oil
was not reentrained under any of the test conditions. This
difference in reentrainment was attributed to differences in
interaction with the filter fibers of oily particles compared
to solid, irregularly shaped particles. Consistent with
previous studies, electrostatic charges on the filter fibers
significantly increased the collection of submicrometer
particles; however, particle reaerosolization was only slightly
impeded by the embedded charges. Finally, filter properties
were found to significantly affect particle reaerosolization.
The number of reaerosolized particles decreased slightly
with filter thickness, an observation that supports the concept
that most of the particles are reentrained from the front
layer of the filter. Essentially no particle reentrainment
was observed from charged felt, compared with up to 5 to
15% from glass fiber HEPA and polypropylene filters.
Reentrainment of bioaerosols from air ventilation filters
has also recently been studied. Jankowska et al. (2000)
compared the collection efficiency and reentrainment rate of
the fungal spores Penicillium brevicompactum and P. melinii
against that of inert potassium chloride (KC1) particles,
using a medium prefilter and a higher efficiency fine filter.
Reaerosolization increased with reentrainment velocity
for all test particles. When the reentrainment velocity
was the same as the loading velocity, the reaerosolization
was less than 0.4%. When the reentrainment velocity was
increased to 3.0 m/s, the reaerosolization of fungal spores
was higher than that of KC1 particles, ranging from 2 to
6% for P. brevicompactum, 5 to 12% for P. melinii, and
0.2 to 0.6% for KC1 particles. The higher reentrainment of
fungal spores was attributed to the presence of aggregated
spores, where the reentrainment velocity may become
sufficient to break up the aggregates and reentrain the
spores. The differences between fungal spores were
attributed to surface structure: P. melinii spores have a spiny
surface, imparting weaker contact with the filter fibers.
In summary, particle size, shape and composition, air
velocity, humidity, and filter properties were all factors found
to affect particle reentrainment from filter materials. Large
particles of approximately 5.0 um diameter and greater
were found to be most susceptible to reaerosolization.
Biological particles were found to become reentrained
more easily than inert particles of comparable size, possibly
because of weaker contact with filter fibers due to irregular
shapes and surface characteristics. In addition, aggregation
of bacteria and fungal spores to form larger clusters was
determined to also increase potential reaerosolization.
Pulsed Flow. A limited number of studies assessed the
effect of pulsed or variable flow on filter performance.
Two studies compared the aerosol penetration through
respirator filters under constant and cyclic flow conditions
(Stafford et al., 1973; Brosseau et al., 1990).
Stafford et al. (1973) measured the penetration of
monodisperse PSL (0.176 to 2.02 um) and dioctyl phthalate
(DOP) (0.3 um) aerosols through respirator filter cartridges
at three cyclic flows with mean flow rates of 30, 35, and
53 L/min. These flows were selected to correspond to a
range of work rates from moderate to heavy. Tests were also
conducted at a constant flow rate of 16 L/min (equivalent
to 32 L/min through a pair of cartridges). Under steady-
flow conditions, the maximum penetration occurred at a
particle size of 0.3 um, which is consistent with theoretical
single-fiber predictions. However, during cyclic flow the
particle size that produced maximum penetration varied
between filters, in one case less than 0.3 um and in the other
approximately 0.5 um. Also, the maximum penetration
was considerably higher than corresponding steady-flow
values, suggesting that tests conducted under steady-
flow conditions may overestimate filter performance.
Brosseau et al. (1990) compared the collection of silica and
asbestos aerosols by dust/mist respirators under breathing
and constant flows. The cyclic-flow was sinusoidal with a
minute volume of 32 L/min, mean flow of 76 L/min, and a
peak flow of 100 L/min. The constant flow rate was 32 L/
min. In general, the silica penetration under cyclic flow
conditions was about one and a half times as great as that
measured under steady-flow conditions, which was consistent
with the results of Stafford et al. (1973). The asbestos
results were inconclusive as the results varied by filter.
Nanoparticle Filtration. Nanoparticles or ultrafine
particles are generally defined as less than 0.1 um (100
nm) in diameter. Based on the theory described in Section
4.2, the filtration efficiency is expected to improve as the
-------
particle size decreases below 100 nm due to the increased
efficiency in capture due to Brownian diffusion. However,
at some point, the particles will behave more like a vapor
than a particle, and, thus, collection efficiency is expected
to decline due to thermal rebound. Diffusion is the primary
collection mechanism in the nanoparticle size range. The
impact velocity is referred to as the thermal velocity.
Thermal rebound occurs when the thermal velocity exceeds
a critical value that results in particle bounce from a fiber
surface following impaction. Wang and Kasper (1991)
have modeled the filtration efficiency of nanometer-sized
aerosol particles through fibrous filters and predicted that
thermal rebound is unlikely to cause enough degradation
in performance to be detected experimentally down to at
least 2 to 5 nm. Several studies have recently assessed the
performance of fibrous filtration media against nanometer-
sized aerosols to determine the size at which thermal
rebound becomes evident (Kim et al., 2006; Heim et al.,
2005; Chen et al., 2006; Balazy et al., 2004; Balazy et al.,
2005; Kim et al., 2007). Results to date have suggested
that filtration theory remains valid down to particle sizes
of at least 3 nm. Relevant studies are summarized below.
Kim et al. (2006) measured the filtration efficiency of two
glass fiber filters against a sodium chloride aerosol over
the range of 1 to 100 nm. The specific application for
the filters was not provided. The effects of particle size,
relative humidity, and particle charge were assessed. The
measured efficiencies were independent of humidity over
the range tested (dry to 92% at ambient temperature).
The measured collection efficiencies were highest against
the charged aerosol. The magnitude of the difference
decreased as the particle size decreased. This observation
was attributed to the fact that collection by diffusion
becomes more efficient with decreasing particle size. The
data suggest that thermal rebound may begin to occur at
particle sizes below 2 nm. This conclusion is supported
by the work of Heim et al. (2005) who characterized the
performance of three low-efficiency filters/meshes over the
range of 2.5 to 20 nm and concluded that thermal rebound
was not detected in the size range down to 2.5 nm.
Chen et al. (2006) assessed the penetration of a salt
aerosol over the range of 4.5 nm to 10 um through two
filtering facepiece respirators. No evidence of thermal
rebound was observed down to a particle size of 4.5 nm.
Tests were also performed after dipping the respirator
filters in isopropanol to remove their electrostatic charge.
Removal of the charge led to a shift in the most penetrating
particle size from 50 nm to 200 nm. The collection
efficiency of particles less than approximately 20 nm
was unaffected, demonstrating that diffusion is the most
important collection mechanism in this size range.
Japuntich et al. (2006) evaluated various test methods to
measure nanoparticle penetration through a fibrous filter.
The particle size range evaluated was 10 to 400 nm. The
penetration appeared to decrease with decreasing particle
size down to 10 nm as predicted based on the single fiber
theory. Kim et al. (2007) characterized the performance of
several fibrous filters against nanoparticles ranging from
3 to 20 nm. Penetration was observed to continuously
decrease with decreasing particle size over the range tested.
In a companion study, Wang et al. (2007) demonstrated
that the measured collection efficiencies were in good
agreement with classical filtration theory down to 3 nm.
Balazy et al. (2005) assessed the performance of two
commercially available N95 filtering facepiece respirators
against a nanoaerosol challenge over the range 10 to 600
nm. The effects of particle size, flow rate, and particle
charge were assessed. The most penetrating particle size was
between 40 and 70 nm for the respirator filters evaluated
and penetrations exceeded 5% at the higher flow rate. The
most significant increases in penetration were observed for
particles of less than 100 nm when the flow rate increased
from 30 to 85 L/min. Collection efficiencies of particles in the
range of 10 to 20 nm were near 100%, indicating that thermal
rebound was not an issue. As expected, charge neutralized
particles were more penetrating than charged particles.
Fewer studies were identified in the literature specific
to HVAC filters. Roth et al. (1999) assess the filtration
efficiency of three fibrous filters used in HVAC systems.
Two filters were reported as electrostatic filters, and one
was reported as a HEPA filter. The challenge aerosol was
ambient air and a differential mobility analyzer was used
to measure efficiencies over the range of 10 to 200 nm.
The airborne concentration of particles larger than 200
nm was not sufficient for measurement of efficiencies.
The HEPA filter showed a minimum collection efficiency
at about 60 nm. The two electrostatic filters behaved
differently. One, a commercially available product, showed
poor performance against particles ranging from 10 to
30 nm (with efficiency less than 20%) but a general trend
of increased efficiency with increased particle size. The
efficiency of the other electrostatic filter was fairly constant
and over 90% for particles in the 10 to 100 nm size range.
Balazy et al. (2004) assessed the efficiency of two
commercial fibrous filters, classified as F5 and G4 per
EN 779, over the range of particle diameters from 10
to 500 nm. The challenge aerosol was oil and a wide-
range spectrometer (WPS, Model 1000, MSP Corp.) was
used to measure the challenge and downstream aerosol
concentrations. A local maximum in collection efficiency
was observed at about 20 nm, below which collection
efficiency decreased with decreasing particle size. The
authors suggested that this decrease was attributed to
thermal rebound. Others have attributed this decrease to
a low challenge concentration for particles less than 20
nm (< 100/cm3) and errors in the downstream sampling
approach (Harrington, 2005; Heim et al., 2005).
Owen et al. (2003) characterized the filtration efficiency
of 26 residential and industrial filters for HVAC systems.
The filters were selected to cover a range of MERV ratings
(~5 to 16). The effect of particle size was assessed over
the range of 30 nm to 10 um. The challenge aerosol was
potassium chloride. As expected, efficiency tended to
increase with increased MERV rating. This trend was
-------
generally true for all particle sizes evaluated. The lowest
efficiencies were generally measured in the 0.1 to 0.5 particle
size range. The higher efficiency filters showed increased
efficiency as the particle size increased or decreased away
from the most penetrating. The lower efficiency filters
performed very poorly against particles less than 100 nm.
For example, collection efficiencies of two MERV 6 filters
was less than 20% over the range from 30 to 100 nm.
In summary, high efficiency fibrous filters, especially
those used in respirator applications, have been shown to
perform as predicted based on single fiber filtration theory
down to particle diameters of at least 2.5 nm. In general,
collection efficiency has been observed to increase with
decreased particle size due to the enhanced collection
by diffusion. Studies completed with charge neutralized
media have shown that diffusion is the dominating capture
mechanism below 20 nm. Fewer studies were identified
that assessed the performance of HVAC filters. However,
assuming that fibrous filters were used, they would be
expected to perform as predicted based on filtration theory.
Owen et al. (2003) showed that collection efficiency of
nanoaerosols tended to increase with increasing MERV
rating. Several of the lower rated filters performed poorly
against nanoparticles with penetrations nearly 100%.
4.4 Critical Assessment
4.4.1 Technology Assessment
Air cleaning methods that rely on mechanical filtration
for particle removal are well established, reliable,
understood, and described (predictable). [In general, HVAC
applications such as residential or commercial buildings
with no need for air cleaning other than to control nuisance
dust and protection of mechanicals typically rely on
relatively low (< 90%) efficiency mechanical filters.]
Filters that rely solely on mechanical particle capture
mechanisms, especially HEPA filters, have been
extensively studied and characterized. HEPA filters
were originally developed to control emissions in the
nuclear energy field and have since been the principal
means of particle removal for individual and collective
protection applications that require high efficiency.
In general, the literature review did not identify conflicting
or controversial data. The literature is in good agreement
regarding the controlling filter and particle characteristics
that affect collection efficiency—hence, the well-
established theory to predict aerosol penetration.
One area of interest within the past decade is the performance
of mechanical filters with respect to aerosols of biological
origin (ABO), or bioaerosols. Of specific interest is whether
the capture efficiency of bioaerosols is comparable to that
of inert aerosols of similar aerodynamic size, especially
in applications for individual (respiratory) and collective
protection systems, notably for military applications but
also for healthcare workers. Battelle is currently supporting
research for NIOSH to further study the penetration of inert
and biological aerosols for respirator filters. In an assessment
of literature on biological versus inert aerosol filtration, the
general consensus is that mechanical filters are comparably
effective at removing particles, independent of the particle
type, as long as they are of comparable aerodynamic size.
Related to the bioaerosol versus inert aerosol collection
efficiency is the concern regarding grow-through of
organisms and reentrainment (particle shedding). There have
been few studies of organism grow-through, which is the
concern that collected organisms can grow on/in the filter,
reaching the back side, and then shed to become an inhalation
hazard. In summary, for sustained growth, moisture and a
nutrient source are required. Although conditions can be
achieved to support growth of organisms on filters, it does
not yet appear to be an established hazard or problem. Proper
filter maintenance can avoid these potential problems.
Shedding is a concern because post collection of a
hazardous aerosol could result in reaerosolization. Very
little has been published regarding reaerosolization,
and the topic is currently being researched by Battelle
for NIOSH. Although mechanical force can be used to
dislodge heavily loaded dusts, whether lightly loaded
filters with a bioaerosol challenge will dislodge is
unknown. According to inert particle collection theory,
once a particle has collected on a fiber, it is retained.
The main area of research regarding fibrous filters is
apparently the development of nanofibers and surface
treatments. Filtration media producers such as Donaldson
and Freudenberg are developing "nanofibers" incorporated
into filters. According to filtration theory, the smaller the
fiber, the higher the collection efficiency. These media
are in consideration for the advanced filter prototype to
be developed under a different subtask of this contract.
Treatment of fibers is also being explored forbiocidal activity
and for enhancing collection efficiency. For example, anionic
surfactants have been applied to fibrous filters with marginal
decreases (10 to 20% of -0.3 um particles) in penetration,
without change in filter physical or mechanical properties.
The concept is to enhance the electrostatic collection
properties by establishing a surface change on the fibers.
4.4.2 Impact on HVAC System
Mechanical filters are by far the most widely used type of
air cleaner. Mechanical filters are the standard against which
other types of air cleaners and air cleaning technologies
must be compared. Future developments, therefore, are
not likely to include significant improvements in their
performance. In general, the advantages of mechanical
filters are their low cost, wide availability in a variety of
sizes and performance ranges, and well-known performance.
The collection efficiency of mechanical filters generally
increases as they age, so their efficiency is generally
lowest at installation. The key disadvantage of mechanical
filters is that their pressure drop increases with use,
requiring an increase in the power needed to maintain
airflow, and requiring replacement with a frequency that
is proportional to their efficiency (higher efficiency filters
require more frequent replacement). In addition, high
-------
humidity air can cause rapid increases in pressure drop in
some circumstances, as well as mold growth on the filter.
If filters selected for use in an HVAC system have higher
pressure drops than in the original design, it will adversely
impact the HVAC system. Additional booster fans may have
to be used, or the speed of the existing blower may have to
be modified to overcome the increase in pressure drop.
4.4.3 Cost Analysis
An in-depth cost analysis of mechanical filtration is provided
in Section 9.2. If new mechanical filtration were added to
an existing air handler in a typical office building, no major
air handler modifications would be required because the
filters would typically fit into an existing air handler and
their pressure drop would be low. Mechanical filtration has
the least expensive initial purchase and installation cost
of all types of air cleaners analyzed in this report. Adding
these types of filters might require installation of a new
access door in the existing unit and new pressure gauges.
The operating and maintenance cost increase would
be very small. Mechanical filters require only yearly
changing (filters are easy to change), which may amount
to a maintenance cost increase of 19% for a typical office
building. The static pressure added to the air handler from
the mechanical filtration would not be great; however, the
fan speed would need to be adjusted, which would increase
the operating cost of typical office building by 6%.
-------
-------
5.0
Critical Assessment of
Electrostatically Enhanced Filtration
5.1 Technology Description
Electrostatically enhanced nitration technology improves the
potential performance over standard fibrous filters that rely
solely on mechanical means of aerosol collection. Aerosol
nitration by fibrous niters has been described previously
in the mechanical nitration technology description in
Sections 4.1 and 4.2. Likewise, the enhancement that can
be achieved by increasing the electrostatic capture forces
will be discussed in the technology description for electret
filtration media in Section 6.1. The electret media make use
of polarization of the fibers to enhance particle collection by
electrostatic forces, which supplement the forces of particle
collection by mechanical mechanisms of interception,
impaction, and diffusion. The electrostatically enhanced
media are treated as a separate technology category because
of the physical differences from standard fibrous media or
electret media.
5.2 Theory of Electrostatically Enhanced Filtration
Technological advancements have been recently made and
products such as StrionAir® GC Filter are now commercially
available. The principle of operation is to ionize the incoming
airstream and particles so that a surface charge is achieved
on the incoming particles upstream of the filter. (See Figure
9 for an illustration of this concept.) Charging of these
particles increases their electrical mobility and also the
attractive force to oppositely charged surfaces. Fibrous filter
media are located between a negatively charged electrode
upstream and a positively charged electrode downstream.
When power is applied to the electrodes, an electrical field
is generated, and the fibrous filter media are polarized, i.e.,
the fibers of the media form areas of negative and positive
charge. (See Figure 10 for an illustration of this concept). In
this manner, electrostatically enhanced filtration is similar to
electret media. In the case of the electrostatically enhanced
filter, the fibers are not permanently charged like electrets,
but rather are charged only in the presence of the electrical
field. Particle collection thus occurs predominantly due
to the electrostatic forces. Because particle collection is
predominantly associated with electrostatic force, larger
fiber diameters can be used for the fibrous filter (it is the
small diameter fibers that are prevalent in particle capture
in mechanical filters). All other parameters of a filter are
constant: the larger the fiber diameter, the lower the airflow
resistance. Rather than increase the collection efficiency
of a fibrous filter by reducing the fiber diameter and thus
increasing the pressure drop, the collection efficiency is
enhanced by the charging of the particles and polarization of
the fibers.
lonization Array Upstream Field Electrode
Figure 9. Illustration of Particle Charging Upstream of Fibrous Filter by
lonization Array (From StrionAir, Inc. Web site, www.strionair.com)
-------
Upstream Field Electrode Filter Media Downstream Electrode
Figure 10. Illustration of Fiber Polarization by Oppositely Charged Electrodes
(From StrionAir, Inc. Web site, www.strionair.com)
5.3 Summary of Relevant Studies
Many authors of the reviewed papers noted the improved
performance of electrostatically enhanced media, regardless
of how the electrical enhancement was achieved (Brown,
2001; Carlsson, 2001; Drouin, 2000; Emmerich and
Nabinger, 2001; Fugler et al., 2000; Thorpe and Brown,
2003; Wang, 2001). However, several papers identified in this
search actually used electret media, which are also described
by some to be electrostatically enhanced (Brown, 2001;
Carlsson, 2001; Drouin, 2000; Raynor and Chae, 2003).
Electret media are discussed as a separate technology in
Section 6.0 of this report.
Several recent studies of electrostatically enhanced devices
are summarized in Table 13. One study performed a detailed
examination of the variables affecting collection efficiency
for an electrostatically enhanced filter, though incoming
particles were not ionized before collection as described
above (Thorpe and Brown, 2003). Other studies measured
the in-duct efficiency in a house and the expected reduction
in indoor particle concentrations with the use of filtration
(Emmerich and Nabinger, 2001; Fugler et al., 2000). Both of
these studies examined a range of devices and compared the
performance of different technologies.
Table 13. Summary of Recent Studies on Electrically Enhanced Devices
Performance Data
Measured aerosol penetration for a range of fiber diameters, face
velocities, applied voltages, and particle sizes; performance was as
expected (increasing with decreased face velocity and fiber size and
increased voltage and particle diameter; applied electrical fields offer
improvements over mechanical filtration similar to electret media.
Thorpe and Brown, 2003:
Agranovski et al., 2006.
Effectiveness Data
Measured in duct efficiency and indoor particle concentrations for a
range of devices for comparison purposes; electrostatically enhanced
filters performed better than mechanical filters, though not nearly as
well as ESP filters.
Emmerich and Nabinger, 2001:
Fugler et al., 2000.
Models of Electrostatic
Collection Forces
Present equations describe electrostatic attraction forces for various
scenarios involving charges on particles or particles in electrical fields;
mathematical descriptions of performance are lacking for electrostatic
filter forces, particularly for filters with significant loading.
Thorpe and Brown, 2003: Wang,
2001.
-------
5.3.1 Performance and Variables That Affect
Performance
The performance of electrostatically enhanced filtration
technologies depends on several factors. Because
the filters used are mechanical or fibrous filters, the
performance depends on the fiber diameter and the number
of fiber layers. The addition of the electrical field over
the filter creates a dependence on the voltage used, as
is the case with electrostatic precipitation technology.
Also, the performance depends on the particle size and
the face velocity through the filter, as is the case with
both mechanical and electrostatic precipitation.
Three of the studies listed in Table 13 — Thorpe
and Brown (2003), Emmerich and Nabinger (2001),
and Fugler et al. (2000) — measured the collection
efficiency of an electrostatically enhanced filter,
with the latter two studies measuring the in-duct
efficiency with the filter installed in a house.
All three studies examined the effect of particle size on
collection efficiency. Emmerich and Nabinger (2001)
measured in-duct efficiencies of about 7% for particle
diameters of 0.3 to 0.5 urn, 10% for 0.5 to 1.0 urn, and 13%
for 1.0 to 5.0 um in an uninhabited test house. In Fugler et al.
(2000), a collection efficiency of 20% for PM10 particles and
17% for PM1 particles was reported based on measurements
taken in six different inhabited test houses. Both studies
were short in duration and did not study performance over
an extended period of time. Thorpe and Brown (2003)
performed an experimental study of an electrostatically
enhanced filter in a laboratory setting. Figure 11 shows the
variation in aerosol penetration with particle diameter for
several applied voltages, ranging from no field to 600 kV/m.
Analogous to the results above, greater efficiency (or lower
penetration) is seen at larger particle diameters. In fact,
collection efficiencies are shown to increase by an order of
magnitude for all particle sizes. Collection efficiency is also
shown to increase with increased voltage, especially for the
larger particles.
Of the above studies, Fugler et al. did not measure the in-
duct velocity. Emmerich and Nabinger (2001) measured an
average duct velocity of about 5 m/s with the filter in place
and did not study other velocities. Figure 12 shows the
results of Thorpe and Brown (2003) for filtration velocity for
several voltages, again ranging from no field to 600 kV/m.
As can be seen in the figure, the effect of face velocity is less
significant with low electric field voltages. At higher electric
field voltages (300 and 600 kV/m), aerosol penetration is
greater at higher velocities. Emmerich and Nabinger reported
a relatively high penetration (around 90% for all particles) at
a much higher velocity, though it is unclear what electrical
field voltage was tested.
As is clear from Figures 11 and 12, collection efficiency
increases (and aerosol penetration decreases) with increasing
electric field voltage (Thorpe and Brown, 2003). Emmerich
and Nabinger (2001) and Fugler et al. (2000) did not report
the operating voltages of the particular filters tested, so it is
not possible to compare the performance measured in those
studies with the study of Thorpe and Brown (2003).
Thorpe and Brown (2003) demonstrated little significance
with regard to the polarity of the electrical field, as seen in
Figure 13. This figure is another example of the improved
performance at higher voltages, demonstrating decreasing
penetration with increasing voltage.
Pressure drop was measured by both Thorpe and Brown
(2003) and Emmerich and Nabinger (2001). However,
pressure measurement results were not directly reported.
Thorpe and Brown do report a 13% increase in pressure drop
with a high loading test, as discussed in the next section.
Thorpe and Brown (2003) demonstrated only minor
changes to aerosol penetration with time of operation
when loading a filter with sodium chloride particles. To
maintain performance for a reasonable test time, the charged
electrode was exchanged with the grounded electrode in the
downstream position, resulting in about 303 minutes of test
time versus 10 minutes with the charged electrode upstream.
Regardless of the position of the charged electrode, the filter
failed by short circuiting due to the accumulation of sodium
chloride particles in contact with the electrode generating
the field. Without the electrical field, the performance of the
filter was reduced, as shown above. The authors performed
similar measurements with a permanently charged filter
(probably electret) and noted significant increases in the
aerosol penetration, presumably due to the shielding of the
electric charge by the sodium chloride particles. The results
can be seen in Figure 14. The other studies did not examine
performance degradation with time.
-------
100-r
10--
0)
c
0)
Q_
0.1 ••
0.01-•
0.001
01234567
Particle diameter (|jm)
Figure 11. Penetration as a Function of Particle Diameter at Several Voltages
Thorpe and Brown, 2003)
100-r
10
0.1
0.01
0.001
H
0.05
0.1
0.15
0.2
Filtration velocity (m s~1)
Figure 12. Penetration (with 37 urn diameter fibers) of Mo nod is perse Particles (of
4.8 urn diameter) as a Function of Face Velocity at Several Voltages
(Thorpe and Brown, 2003)
-------
100-r
c
o
Q)
c
Q)
Q_
0.1 --
0.01
n
n
o
200
400
600
800
1,000 1,200
Electric Field (kV nv1)
Figure 13. Penetration (with 37 urn diameter fibers) of Monodisperse Particles
(of 3 pen diameter) as a Function of Voltage for Positive and Negative
Electrical Fields (Thorpe and Brown, 2003)
Agranovski et al. (2006) assessed the impact of unipolar
ionization on the collection efficiency of two low efficiency
HVAC filters. The effects of particle size (0.5 to 1.5 urn)
and distance between the emission source and filter surface
were assessed. The face velocity was held constant at 1.1
m/sec resulting in pressure drops of 82 and 68 Pa across
the two filters, respectively. Efficiencies of both filters
were less than 20% for all particle sizes tested without
the ionization source. The collection efficiency increased
for all particle sizes with the use of the ion emitter. For
example, for 1-um particles, the collection efficiency
jumped from 5 -15% to 40-90%. There was not a significant
difference between the measured collection efficiencies
with the ion emitter placed 5 and 10 cm upstream of
the filter. The enhancement was less pronounced as the
emitter was placed 25 cm from the filter. This concept was
previously applied to respirator filters, and it was shown
that particle penetrations were reduced by an order of
magnitude (Lee et al., 2004b). In both studies, the authors
attribute the improved performance to unipolar ions being
deposited on the particles and filter media fibers resulting
in a repelling force that shields a fraction of particles
from the filter. The effect of loading was not assessed.
5.3.2 Assessment in an HVAC System
In studies by Emmerich and Nabinger (2001) and Fugler et
al. (2000), collection efficiency was measured for a number
of different devices installed in a house. In both studies, the
electrostatically enhanced filter did not perform exceptionally
well. In the study performed by Emmerich and Nabinger
(2001), the electrostatically enhanced filter performed better
than several other fibrous filters, but not as well as an electret
filter, and much more poorly than an electrostatic precipitator
device. The electrostatically enhanced filter performed poorly
in Fugler et al. (2003), with a lower collection efficiency
than several fibrous filters and an electrostatic precipitator
(which again easily exceeded the performance of all the other
devices tested).
-------
100 T-
10 --
c
.g
U—»
CO
-i—»
CD
C
CD
0.
0.1
0.01 --
0.001
o
o
o
O Electrically enhanced
Q Permanently charged
0 50 100 150 200 250 300 350
Exposure time (mins)
Figure 14. Aerosol Penetration of Monodispersed Particles (3 urn diameter) Through
an Electrically Enhanced Filter and a Permanently Charged Filter as a
Function of Operation Time (Thorpe and Brown, 2003)
5.3.3 Additional Factors
Reduction of Particles. Device effectiveness is often
quantified by the reduction in particles. Both Emmerich
and Nabinger (2001) and Fugler et al. (2000) measured
indoor particle concentrations, but Emmerich and Nabinger
(2001) do not directly report the reduction in indoor
particles with the operation of the filter technology. Fugler
et al. (2000) observed a 9% reduction in indoor PM10
levels during "active" periods (times when there is human
activity) and a 29% reduction in indoor PM10 levels during
"nonactive" periods (times when no one is home, everyone
is sleeping, etc.). With these reductions, the authors note
that the electrostatic precipitator, which showed an active
reduction of 31% and nonactive reduction of 71% in particle
concentration, did not significantly reduce occupants'
exposure to indoor particulates. Figure 15 shows an example
of the PM10 and PM1 concentrations measured within an
occupied house for a 24-hour period while an electrostatic
precipitation device was operating.
-------
PM10 House Avg.
PM1 House Avg.
Time
Figure 15. Example of Indoor Particle Measurements for a 24-hour Period With an
Electrostatic Precipitation Device (Fugler et al., 2000)
Electrostatically enhanced nitration offers benefits over
mechanical filtration since actively charging the filter region
can significantly increase filtration efficiency, as is seen with
passively charged electret filters. Electrostatically enhanced
filters have the capability of using an adjustable electrical
field strength, so increasing the electrical field voltage can
improve filtration. Also, the electrical field in the filter is less
apt to degrade due to the collection of charged particles or
deterioration of the media, as can happen with electret filters.
However, some electrical power is required to produce the
electrical field, unlike electret. Filtration benefits are similar
for electret and electrostatically enhanced filters, but electret
filters are more frequently used because they do not have an
additional electrical power requirement.
5.4 Critical Assessment
5.4.1 Technology Assessment
Electrostatically enhanced filtration devices are relatively
new to the market and relatively little research regarding
their performance is available compared to established air
cleaning technologies such as fibrous filters or electrostatic
precipitators. The principle of operation is not new, however.
Electrostatically enhanced filtration is effectively the addition
of electrostatic force to a variation of fibrous filtration in an
effort to improve collection efficiency.
Consistent with electrostatic attraction theory (discussed in
Section 4.2), collection efficiency increases with an increase
in the applied voltage—an increase in the fiber charge.
Depending on particle size, collection efficiency gains of
one to three orders of magnitude were achieved, with the
greatest increase experienced by the largest particles. The
enhancement in collection efficiency is most significant as the
filtration velocity decreases. Due to an increased residence
time—longer time for the electrostatic forces to act on the
particles—and reduction in inertial effects. Polarity of the
field did not appear to matter.
Research conducted by Battelle (Kogan et al., 2007)
regarding in-facility testing of units by measuring
reduction in room aerosol concentration provides no
useful information about the unit performance. There
was only a 15% reduction in aerosol concentration
found in a room equipped with a unit.
No comparison of performance of an electrostatically
enhanced (EE) filter to an electret or fibrous filter was found.
Excluding the cost of operation, the EE filters should be
compared to other filters using the figure-of-merit metric
given by:
FOM = -In (Pen)
AP
where Pen is the fractional penetration and AP is the pressure
drop at a specified flow. In this manner, the increase in
efficiency with reduction (or no increase) in pressure drop
can be quantified for consistent comparisons.
5.4.2 Impact on HVAC System
The impact on an HVAC system of using electrostatically
enhanced technologies is minimal. Because traditional
fibrous filtration material is used, the pressure drops are
not significantly different from what they are with other
fibrous filters. Modifications to the HVAC system would
not necessarily need to be made due to the pressure drop
of an added electrostatically enhanced technology. As with
any air cleaning device that employs an ionizing source,
generation of ozone is a concern. There were no reported
studies regarding ozone concentration measurements, but that
does not mean that there was no ozone production. Ozone
generation should be measured in future research.
5.4.3 Cost Analysis
Electrostatically enhanced filters can typically fit into an
existing air handler without major modifications. Most
filters that have been electrostatically enhanced do not
increase the static pressure of the system. Units that charge
particles upstream of fibrous filters to enhance filtration
(e.g., using an ionization array) present an additional, finite,
resistance on the system. The retrofit of an HVAC system
with electrostatically enhanced filters or arrays might require
a new access door to be added to the existing unit and new
pressure gauges. Although these filters do not use a great deal
of electricity during normal operation, installation would
-------
require new electric service. Due to the complexity of the static pressure added to the air handler from the mechanical
filter itself, initial and installation costs for this type of filter filtration would not be great. However, the fan speed would
are very high. need to be adjusted, which would increase the operating cost
The operating and maintenance cost increase would be small. for a ^P^1 office buildin§ b^ 6%-
These filters require only annual changing (only the pads An in-depth cost analysis on electrostatically enhanced
need to be changed), which may amount to a maintenance filtration is provided in Section 9.3.
cost increase of 24% for a typical office building. The
-------
6.0
Critical Assessment of Electret Media
6.1 Technology Description
Many air niters in the market are currently manufactured
using electrically charged media to attract particles. Filters
that use this technology are commonly referred to as
"electrostatic," "electrically charged," or "electret" media.
In this report, the filters are referred to as electret media. The
advantage of electret media is their relatively high collection
efficiency at very low pressure drops.
Electret media are made of dielectric materials that have
a significant microscopic bipolar charge on the fibers and
a very low net macroscopic charge. Different from the
electrostatically enhanced filter described in Section 5.0,
electret media are permanently charged in the course of
manufacturing. Therefore, electret media do not need an
electrode system to charge filter media or an ionizer to charge
incoming particles during operation.
There are many types of electret media, due to the variety
of fiber-forming technologies (i.e., meltblown, split fiber,
bi-component spunbond, needlefelt, etc.) and the variety
of electrostatic treating technologies (i.e., corona charged,
triboelectric charged, induction charged, etc.). A recent study
demonstrated that electret filter media can also be generated
by applying anionic surfactants on some polypropylene
fibrous filters (Yang and Lee, 2005).
The composition of electret media varies from polycarbonate,
polypropylene, and polyolefin, to a binary mixture of
polypropylene and chlorinated acrylic fiber. The media
manufactured by different technologies and different
polymers could demonstrate a significant difference in
filtration performance and degradation behavior (Barrett and
Rousseau, 1998 andRomay etal., 1998b).
6.2 Theory of Electret Media
Electret filters collect particles using a combination of the
conventional mechanical mechanisms (i.e., impaction,
interception, and diffusion) and the electrostatic mechanisms
(i.e., Coulombic attraction and dielectrophoretic capture).
Charged particles are attracted to oppositely charged fibers
by the Coulombic force. For singly charged particles, the
attraction increases as particle size decreases. Neutral
particles that are unaffected by Coulombic force are collected
by dielectrophoretic force — the polarization force induced
by local electrical fields within the filter media. Charged
particles are also collected by dielectrophoretic capture. The
efficiency of the dielectrophoretic capture increases with
particle size.
The efficiency of electret media depends on parameters
such as charges on particles, charge density of fibers, and
chemical compositions of particles and fibers; efficiency also
depends on factors that affect the efficiency of conventional
uncharged filters, such as fiber diameter and packing density
of the fibrous materials. Several theoretical models are
available to predict the capture efficiency of electret filters
(Pich, 1978; Pich et al., 1987; Brown, 1981; Lathrache and
Fissan, 1987; and Otani et al., 1993). These models relate the
electret collection efficiency to parameters, such as average
charge density of fibers, number of charges on particles, fiber
packing density, and fiber and aerosol diameters. The models
provide a good qualitative description of the behavior of
clean electret filters and show a generally good agreement
with experimental results (Romay et al., 1998b; and Wang,
2001). For example, the empirical power law expressions for
single-fiber efficiencies obtained by Romay et al. (1998b)
when testing commercial electret filters were in good
agreement with those predicted by Brown (1981).
Electret media capture particles by the same mechanisms
as fibrous filters do, as described in Section 4.2. It is
the enhancement of particle capture via the electrostatic
mechanism that distinguishes electret media from fibrous
media, and therefore a separate technology category is
considered in this analysis. The local (particle-fiber regime)
electrostatic force increases capture efficiency without the
need for increasing thickness, increasing fiber packing
density, or reducing fiber diameter. Thus, the overall particle
collection efficiency is increased—all other parameters of
the filter being equal—while the pressure drop (airflow
resistance) is maintained or reduced.
6.3 Summary of Relevant Studies
Recent studies in electret media are summarized in Table 14.
Note that Table 14 focuses primarily on studies conducted
since 1995. Only important and representative studies
conducted before 1995 are included.
6.3.1 Performance and Variables That Affect
Performance
In the HVAC filtration market, electret filters are becoming
increasingly popular (Myers and Arnold, 2003; Homonoff,
2004). Task 2 of the current overall project to select the
representative filters to test found that most of the medium to
higher efficiency filters were electrostatic filters. Nearly all
high efficiency (MERV 11 or higher) residential filters were
composed of electret material as well.
The electret filters available for residential HVAC
filtration generally have MERV ratings ranging from 8
to 12. 3M is the leading company to produce high-end
electret filters for residential application. 3M's Filtrete™
Ultra Allergen filter, Filtrete™ Micro Allergen filter, and
Filtrete™ Dust & Pollen filter are rated as MERV 12,
11, and 8, respectively. The typical pressure drop for
residential pleated electret filters ranges from 0.13 to
-------
Table 14. Summary of Electret Studies
Basic Scope
Review of Electret Media
Content/Conclusion
An overview of current electret media types,
charging techniques, and the fundamental
impact of environmental factors on filter
performance.
Reference
Myers and Arnold, 2003.
Collection Efficiency Model
• Models
• Models validation
• Models review
Major models that are available for
predicting the collection efficiency of
electret; these theoretical models provide a
good qualitative description of the behavior
of clean electret filter and are in general
agreement with experimental results.
Rich, 1978; Rich et al., 1987; Brown,
1981; Lathrache and Fissan, 1987; Otani
et al., 1993; Romay et al., 1998a; Wang,
2001; Leeetal., 2002.
Methods for Making Electret
Recent studies and developments in
making electret.
Rousseau, 1998; Nifuku, 2001; Drouin,
2002; and Tsai, 2002; Yang, 2005.
Degradation With Aerosol Loading
Electret media degradation studies:
generally, the degradation depends on the
type of aerosol, but filter properties can
also affect the degradation; the degradation
by oil aerosols such as DOR (dioctyl
phthalate), diesel soot, and cigarette smoke
is particularly severe in many electret
filters.
Brown et al., 1988; Tennal et al., 1991;
Lehtimaki and Heinonen, 1994; Lehtimaki,
1996; Walsh and Stenhouse, 1997; Walsh
and Stenhouse,1998; Lifshutz, 1997; Liu
and Romay, 1997; Barrett and Rousseau,
1998; Hofacreetal., 1999; Hanleyetal.,
1999; Hanley and Owen, 2003; Arnold
and Myers, 2002; Janssen et al., 2003a;
Janssen et al., 2003b; Ji et al., 2003;
Raynor and Chae, 2002; Raynor and Chae,
2003; Romay et al., 1998a.
Performance Data for Electret Media
Experimental data on the performance of
commercially available electret media;
test data available include efficiency as a
function of particle size, and pressure drop
vs. face velocity curves at different filter
basis weights.
Carpin et al., 1997; Liu and Romay, 1997;
Romay etal., 1997; Romay etal., 1998b.
0.35 in. w.g. (32 to 87 Pa) at 300 fjpm (1.52 m/s) of
face velocity (3M Brochure, Improve Indoor Air).
The electret filters used for commercial HVAC filtration
generally have MERV ratings ranging from 8 to 16.
Freudenberg is the leading producer of high-end electret
filters for commercial HVAC applications. Freudenberg's
pleated electret filter Viledon® MV95 is rated as MERV
15 and has a pressure drop of only 0.35 in. w.g. (87 Pa) at
500 fpm (2.54 m/s) face velocity. Its pocket electret filter
Viledon® MF95 is rated as MERV 16 with a pressure drop of
0.5 in. w.g. (125 Pa) at 500 fpm.
Fiber charge density affects the collection efficiency.
Manufacturers and researchers have tried to increase
the electrical charge density of electret fibers in order
to improve collection performance (Nifuku et al.,
2001; Lee et al., 2002). The effects of relative humidity
(RH) (up to 100%) and temperature (up to 70°C),
however, are almost negligible in most electret filter
media available today (Liu and Romay, 1997; Arnold
and Myers, 2002; and Myers and Arnold, 2003).
A major concern with using electret filters is the effect of
aerosol loading on collection efficiency. A number of studies
were conducted previously to determine the efficiency
degradation of electret filters during aerosol loading (Brown
et al., 1988; Tennal et al., 1991; Lehtimaki and Heinonen,
1994; Lehtimaki, 1996; Walsh and Stenhouse, 1997; Walsh
and Stenhouse, 1998; Lifshutz, 1997; Barrett and Rousseau,
1998; Romay et al., 1998b; Arnold and Myers, 2002; Hanley
et al., 1999; Hanley and Owen, 2003, Raynor and Chae,
2002 and 2003; Janssen et al., 2003a and 2003b; and Ji et al.,
2003). Generally, degradation depends on the type of aerosol,
but filter properties can also affect it (Barrett and Rousseau,
1998; Romay et al., 1998b).
Degradation by oil aerosols such as dioctyl phthalate
(DOP), diesel fumes, diesel soot, and cigarette smoke is
particularly severe in many electret filters (Lehtimaki, 1994;
Lifshutz, 1997; and Ji et al., 2003). There appear to be two
mechanisms for electret degradation by oil aerosols. The first
involves wetting and coating of the fiber by the oil aerosol
(Pierce and Lifshutz, 1997; Barrett and Rousseau, 1998).
This wetting and coating process shields the electret charge
on the surface. The second mechanism involves fine particles
carrying a Boltzman distribution of electrostatic charges.
As the oil aerosol wets and coats the fibers, these individual
charges can migrate to and neutralize some fixed charges on
the surface of the fibers. This process cannot occur with solid
aerosols, since the charges on a solid particle are not mobile
(Pierce and Lifshutz, 1997).
Due to the potential degradation of electret filters by oil
aerosols, NIOSH's "Testing and Certification Standard
for Paniculate Respirators and Filters - 42 CFR 84"
requires a loading test with DOP aerosol (Sigma-
Aldrich, St. Louis, MO). Oil-resistant electret filters,
which have much higher resistance to degradation by
-------
oily aerosols than the conventional electret niters, were
developed and used in particle respirators (Barrett
and Rousseau, 1998; Hofacre et al., 1999; Janssen
et al., 2003a and 2003b; Romay et al., 1998a).
In a study conducted by Hofacre et al. (1999), the electret
media fabricated by a leading media manufacturer
were tested for aerosol penetration when being
loaded with oil aerosol. The result demonstrated that
improvement in electret media resistance to oil aerosol
degradation can be achieved. An experimental electret
media sample from the manufacturer exhibited no
measurable increase in aerosol penetration when loaded
with up to 6 mg/cm2 of fog oil aerosol, whereas an
earlier version of electret was adversely affected.
Barrett and Rousseau (1998) reported that in their study, DOP
loading tests were conducted with four different types of
electret filters, including tribocharged polypropylene/acrylic,
corona-charged polypropylene, fibrillated electret film, and
new advanced electret media at a face velocity of 7.8 cm/s.
The measured initial and final efficiencies at DOP loading of
0.82 mg/cm2 are summarized in Table 15. As shown there,
the initial efficiency and the efficiency degradation with DOP
loading varied with the types of media. After being loaded
with 0.82 mg/cm2 of DOP, the efficiency decreased 0.1 to
8 % for the electret media tested. Among them, the P-type
advanced electret filter demonstrated the best oil resistance,
with efficiency decreased only 0.1% at DOP loading of
0.82 mg/cm2. This P-type filter is an oil resistance medium
developed by the manufacturer for the application of P-series
respirators. The oil-resistant electret filters were not found in
applications of HVAC filtration probably because oil aerosol
is not the major component of ambient or indoor aerosols.
Figures 16, 17, and 18 are examples of degradation of electret
by aerosol loading of diesel fumes, cigarette smoke, and
Arizona road dust, which were measured by Lehtimaki and
Heinonen (1994).
Table 15. Degradation of Electret Media Measured by Barrett and Rousseau (1998)
Electret Media DOP Aerosol
Initial m;% Final m;% Change
Tribocharged Filter Media (media weight: 300 g/m2)
Fibrillated Electret Film (media weight: 200 g/m2)
Corona-charged Polypropylene (media weight: 120 g/m2)
R-type
P-type
98
98
97
99.9
97.0
90
92
91
99.2
96.9
-8%
-6%
-6%
-0.7%
-0.1%
u
z
W
—
u
-
$
o
FILTER
EXPOSURE &
PRESSURE
(9/m2) (Pa)
O 0.00 - 76
o 1.39-77
• 2.61-78
X3.80-78
4.94 - 79
A 5.97-79
^9.36-80
PARTICLE SIZE (juim)
Figure 16. Effect of Diesel Fume Aerosol on the Removal Efficiency of an Electret Filter
(Lehtimaki and Heinonen, 1994)
-------
100
CLEAN FILTER
AFTER 20 CIGR.
PARTICLE SIZE (juim)
Figure 17. Effect of Cigarette Smoke on the Removal Efficiency of an Electret Filter
(Lehtimaki and Heinonen, 1994)
U
z
-
£
o
100
80
60
40
20
0
0.1
10
PARTICLE SIZE (jmm)
FILTER
EXPOSURE &
PRESSURE
(9/m2) (Pa)
Figure 18. Effect of Arizona Road Dust on the Removal Efficiency of an Electret Filter
(Lehtimaki and Heinonen, 1994)
As shown in Figures 16 and 18, the aerosol loading with
Arizona road dust was significantly higher, compared to
diesel fume loading. The removal efficiency, however,
decreased only slightly with the dust loading, compared to
the significant efficiency decrease when the filter was loaded
with diesel fume aerosol.
Because solid aerosols are the major components of ambient
and indoor aerosols, the study on the potential degradation
of electret media used for HVAC filtration applications
should focus on the effect of loading with solid particles. The
collection efficiency of an electret filter for solid particles
decreases with operation time in its early stage of collection
as the fibers are coated and shielded. Then the collection
efficiency becomes relatively constant but increases with
time because of the mechanical collection mechanism for the
filter media loaded with the solid particles.
Arizona road dust is the ASHRAE test dust that is currently
used for the conditioning step in the ASHRAE Standard
52.2. Several studies (Lehtimaki 1996; Hanley et al., 1999;
Raynor and Chae, 2002 and 2003) revealed, however, that
the degradation of an electret filter when loaded with the
ASHRAE dust is less significant than when the filter was
exposed to a real ambient condition. Lehtimaki (1996)
conducted tests to compare ASHRAE 52.2 test results
-------
for electret filters to field test results. Two commercially
available EU7 electret filters were tested. The collection
efficiency of both filters was reduced significantly in the field
tests (after up to four-month exposure), with a reduction of
the efficiency of up to 4 times for 0.3 um particles and up to
2 times for 1 um particles. When the filters were loaded with
the ASHRAE dust, however, the collection efficiency was
reduced only slightly (less than 10%) for one test filter and
the efficiency even increased for the other test filter.
A series of tests conducted with the support of EPA's
Environmental Technology Verification Program (ETV)
compared the efficiency reduction for electret filters under
real-life exposures and laboratory test conditions (Hanley
et al., 1999; Hanley and Owen, 2003). The electret filters
included a rigid-cell filter charged via an electrodynamic
spinning process and a residential filter charged via a split-
fiber process. Exposures consisted of outdoor ambient air,
in-home air, ASHRAE dust, ASHRAE dust without carbon
black, and a sub-micron KC1 aerosol.
The results, as shown in Figures 19 and 20, indicated
that the efficiency of the electret filters exposed to
outdoor ambient air decreased significantly. The
laboratory test used the ASHRAE dust, however, and
did not reproduce these reductions. Similarly to the
tests conducted by Lehtimaki (1996), the ASHRAE
dust tests showed either significantly less reduction in
efficiency with loading than the ambient exposure test
(the residential electret filter) or even an increase in
efficiency with loading (the rigid-cell electret filter).
In addition, as shown in Figures 19 and 20, there were no
significant differences in the effects of the loading with the
ASHRAE dust and the ASHRAE dust without carbon black.
The sub-micron KC1 aerosol demonstrated an effect closer
to the ambient aerosol exposure than the ASHRAE dust,
although the magnitude of the efficiency decrease was still
underestimated.
The decreasing efficiency over the 3.0 to 10 um range, as
shown in Figure 20 for the "6 weeks ambient test" data, was
attributed to particle bounce (Hanley et al., 1999) The authors
do not further explain the reason or definition of particle
bounce in this context.
Figure 3. Rigid-Cell Filter Conditioning Tests
Initial
With carbon
W/0 carbon
eWksAmb -
KCI
0.1
1
Particle Diameter (pm)
10
Figure 19. Efficiency Reduction of the Rigid-Cell Electret Filter With Aerosol
Loading (Hanley et al., 1999)
-------
0.1
Initial
With carbon
W/Ocartwn
6 Wks amb
KCI
1
Panicle Diameter (pm)
10
Figure 20. Efficiency Reduction of the Residential Pleated Panel Electret Filter
With Aerosol Loading (Hanley et al., 1999)
6.3.2 Assessment in an HVAC System
In the two studies conducted by Raynor and Chae (2002
and 2003), a series of tests was conducted to investigate the
degradation of electret niters in a real HVAC system. Electret
filters made from polyolefin fibers were used continuously in
the system for more than 19 weeks. Collection efficiencies of
the electret filters were found to decline substantially during
the test. The efficiencies forthe 0.337, 0.626, and 1.1 urn
particles were reduced to 2, 2.2, and 1.6 times, respectively.
In these studies, the same types of electret filters were also
tested according to ASHRAE Standard 52.2-1999. Unlike
the results observed in the real HVAC tests, accelerated dust
loading tests run according to the ASHRAE Standard 52.2
did not show any efficiency decrease.
The efficiency differences when the filters were loaded with
the ASHRAE dust and the ambient dust may be caused by
the difference in the particle sizes. In Raynor and Chae's
study (2003), most of the mass of the particles collected
by the filters in the real HVAC system was contributed by
particles smaller than 1 um, while most of the mass in the
ASHRAE dust was contributed by particles with diameters
larger than 1 um. Most of the atmospheric particles are
small in diameter and the prefilters in the real HVAC
systems collected almost all particles larger than 3 um in
diameter. Several studies regarding aerosol loading and
filter performance clearly showed that smaller particles
cause a more rapid degradation in efficiency of electret
filters than larger particles (Walsh and Stenhouse, 1997
and 1998; Ji et al., 2003) probably because the smaller
particles may be more capable than larger particles of
masking or screening the charges on electret filters.
The studies described above revealed that the ASHRAE 52.2
dust loading procedure does not adequately reproduce the
reduction in filtration efficiency that electret filter undergoes
in actual HVAC systems. The ASHRAE Standard 52.2,
which was developed to determine the minimal efficiencies
of a filter over its lifetime, may actually provide an artificially
higher MERV rating for an electret filter.
Realizing that the ASHRAE Standard 52.2 tends to show an
artificially higher MERV rating for electret filters, ASHRAE
supported a research project conducted by Research Triangle
Institute (RTI) to develop a loading dust for a new loading
test method that will more nearly represent the minimum
efficiency points of an electret filter in a real-world
application (Hanley and Foarde, 2003). In this project, the
new test method was developed to replace the first dust
loading step (or the conditioning step) of ASHRAE 52.2,
using nano-sized solid-phase KCI aerosol (with number mean
diameter of 0.035 um) as the conditioning aerosol. The new
method provided a means of accelerating the drop-off in
efficiency that electret filters undergo in real-life applications.
A draft addendum (Addendum C) to ASHRAE Standard
52.2 was prepared in the project; a detailed protocol for
conditioning electret filters using nano-sized KCI aerosols to
mask (or screen) the charges on electret filters was included.
Addendum C is currently undergoing public review.
6.3.3 Additional Factors
No additional factors were identified for electret media.
6.4 Critical Assessment
6.4.1 Technical Assessment
Electret media are found in a variety of filter products,
primarily respirator filters and HVAC filters. Electret
filters have gained significant market share and acceptance
in HVAC filtration applications over the past few years
(Arnold and Myers, 2002; Homonoff, 2004) despite
the potential efficiency degradation of electret media
with use. As discussed above, filters made of electret
media offer the advantage of a lower airflow resistance
for an equivalent efficiency, or a higher efficiency for
an equivalent airflow resistance. Also, electret filters
are usually less expensive than mechanical filters (glass
fiber filters) with the same MERV rating. In addition,
in spite of the collection efficiency degradation, the
efficiency of an electret filter always exceeds that of an
uncharged filter with the identical mechanical structure.
-------
When selecting an electret filter for an HVAC application,
it is important to evaluate filter performance data for the
particular application conditions. If the real-life performance
data are not available, a laboratory loading test, which is able
to more nearly represent the minimum efficiency points of an
electret filter in a real-world application, should be conducted
to ensure that the selected electret filter can meet the design
goals of a particular HVAC application.
Although the advantage of improved filtration efficiency
compared to that of standard fibrous media is well
documented, the single greatest concern regarding electret
filters is degradation. Degradation is associated with aerosol
loading, aging, and environmental effects. The studies
discussed above clearly show the effect is real and a valid
concern. The loading effects have been considered of
such importance that both the respirator industry (through
filter certification standards for NIOSH) and the HVAC
industry (through specific filter conditioning specifications
in ASHRAE 52.2) have attempted to address this issue
by requiring filters to meet standards with filter loading
requirements.
The use of ambient aerosol was found to be more degrading
than the original loading dust (Arizona road dust) for
ASHRAE 52.2 testing—hence, the exploration of a KC1
nano-sized loading aerosol. Similarly, filters for respirator
applications are loaded with DOP aerosol to demonstrate
oil resistance and meet performance requirements, since
the particle size more closely matches ambient conditions.
In the respirator industry, there has been criticism that the
DOP aerosol is not representative of the use conditions; the
concentrations and loadings are much greater than those
that would be experienced in use. Likewise, the selection
of KC1 as a conditioning aerosol for ASHRAE filter testing
can be questioned. Previous research has demonstrated
that different aerosols can cause different extents of
degradation. The difference appears to be associated with
the aerosol composition and not necessarily the particle size.
Nonetheless, the selection of KC1 as the conditioning aerosol
for electret filters, as well as the loading concentration, merit
further consideration.
Another concern with selection of a specific aerosol and
loading concentration for ASHRAE 52.2 testing is that media
manufacturers will produce media specifically to meet or
pass performance requirements. For example, respirator
manufacturers design media such that their filters will meet
certification requirements for P100 filters. Passing a standard
test will not necessarily ensure resistance to degradation.
Filters of all media types should be tested in the same
manner, including conditioning aerosols.
Other degradation concerns with electret media are
operating temperature and redistribution of charge
where local (on the fiber) charge separation no longer
exists or has been reduced. Temperature may affect
charge redistribution. Also, the polymeric fibers are
not suitable for relatively high temperature (> 120°C)
operation because of the melting point of the polymer.
Within the context of HVAC applications, operating
temperatures should not cause the polymer fiber to melt.
Due to the limitations listed above in tests such as the
ASHRAE 52.2 conditioning step, standardized tests may
not provide a completely reliable measure of electret
performance degradation and thus not provide a reliable
MERV rating. Because electret filters are used in high
efficiency applications and because electret filters show
significant reduction in efficiency for many applications, care
should be used in selecting electret media. Where possible,
electret filters should be tested to determine degradation
under the conditions where they will be used.
Despite the concerns regarding electret media degradation,
electret filters merit use as HVAC filters. Strategies to
reduce the effects of degradation are possible and are
being explored. One approach is to mix fibers (specifically
nanofibers) to provide additional mechanical filtration
capability. Also, as respirator filter manufacturers have
demonstrated, improvements in resistance to degradation can
be achieved.
6.4.2 Impact on HVAC System
Electret filters have a relatively lower pressure drop than
conventional uncharged fiber filters; therefore, they can be
installed into an existing HVAC filtration system without
extensive modification such as the addition of an extra fan.
The performance degradation of electret filters with service
(or loading) may cause the actual efficiency of an electret
filter to be significantly lower than the design efficiency
over the entire service life. This reduction in filter efficiency,
however, should not have any impact on the performance of
the HVAC system.
6.4.3 Cost Analysis
Electret media filters can typically fit into an existing air
handler without major modifications, and the static pressure
increase in the system would be small. These types of filters
might require a new access door to be added to the existing
unit and installation of new pressure gauges. Initial and
installation costs are very inexpensive since there is no need
for electric service.
Typical operating and maintenance costs are low.
Maintenance includes yearly changing of not only the electret
filters, but of the prefilters as well, increasing the maintenance
costs of the typical office building by 14%. The operating
cost increase would also be small. One could expect a 9%
increase in the operating cost of a typical office building.
An in-depth cost analysis of electret media is provided in
Section 9.4.
-------
-------
7.0
Critical Assessment of
Electrostatic Precipitation
7.1 Technology Description
Electrostatic precipitation (ESP) is a technology that has
been used for some time in industrial applications to reduce
the amount of paniculate matter resulting from combustion
and other processes. Commercial and residential devices
employing ESP, often called electronic air cleaners (EACs),
are now widely available. Figure 21 shows a schematic of
the ESP process. In general, an electrical charge is imparted
to incoming dust particles as they pass through an electrical
field in the ionizing section. The charged particles are
collected on plates of an opposite charge in the collection
section. Additional filters may be used to reduce the number
of large particles before the ESP stages (i.e., a prefilter
as shown in Figure 21), collect agglomerated particles
dislodged from the collection plates (i.e., the after-filter), or
to remove odors. ESPs and EACs offer several advantages
over traditional fibrous filters. By employing electrical
forces, ESPs may achieve high collection efficiencies
at relatively low pressure drops. EACs also require less
frequent replacement. EACs require regular cleaning, with
lower operating costs. General ESP theory is well explained
in several sources, including Oglesby and Nichols (1970),
White (1963), and Rose and Wood (1956).
7.2 Theory of Electrostatic Precipitation
In the ionizing section, particles acquire a charge from ions
generated within the electrical field. These ions are created by
corona generation. Corona is the phenomenon of ionization
of gas molecules in regions of high electrical field strength.
A relatively high voltage is applied to the ionizing electrode,
resulting in a high electrical field near the electrode.
Electrons present in the field are accelerated and impact gas
molecules, releasing more electrons and creating positive
ions. Depending on the polarity of the electrical field, the
electrons and positively charged ions move in different
directions. A negative ionizing electrode creates a negative
corona where the positive ions are attracted to the electrode,
producing more free electrons when they collide with the
electrode. The electrons are attracted to the positively
charged collection plate(s), impacting gas molecules to create
negative ions as the strength of the ionizing electrical field
diminishes. A positive ionizing electrode (positive corona)
produces an opposite effect. Both positive and negative
coronas are used in ESP, though negative coronas are used
less often for cleaning air in occupied spaces because
greater amounts of ozone are generated (Huang and Chen,
2001). Negative coronas do offer advantageous electrical
performance, resulting in greater efficiency for the same
operating conditions. The corona is affected by electrode
geometry and gas composition and conditions.
The ions created by corona generation impact particles in
two ways. Larger particles tend to travel along electrical field
lines and directly impact particles in a process called field-
dependent charging. As a particle becomes saturated with
charge, it diverts the electrical field lines so that other ions do
not impact the particle. The saturation charge of the particle
is related to the magnitude of the electrical field responsible
for charging the particle, the size of the particle, and the
dielectric constant of the particle.
Smaller particles (0.2 um and less) receive less charge
through field-dependent charging and more charge from
direct collisions between the ions and particles due to thermal
motion, or diffusion charging. As with field-dependent
charging, as charge is accumulated on the particle, the
probability of impact with additional ions is decreased.
However, since there is no upper limit to diffusive motion,
there is no saturation limit for diffusion-charged particles. In
either case, the higher the charge on the particle, the greater
the electrical force between the particle and the collection
electrode. Greater residence times within the electrical field
will result in higher charges when the saturation charge has
not been reached.
The collection of the charged particles is controlled by the
forces on the particle, which include electrical, gravitational,
inertial, and aerodynamic forces. Electrical and aerodynamic
Dirty Air
Ionizing
Pre-Filter Section
I I
Collection
Plates
After-Filter Clean Air
I
Figure 21. Schematic of ESP Process
-------
forces are the most significant in this case. The electrical
force will be acting on the particle to move it toward the
collection electrode while the aerodynamic drag force will
oppose the forward motion of the particle. When these
two forces are balanced, the particle will have reached a
sort of terminal velocity called the migration velocity. The
migration velocity will be dependent on the charge of the
particle, the size of the particle, the strength of the electrical
field, and the viscosity of the gas around the particles. The
following is a general equation for the migration velocity:
w =
6-tf-a-fl
where w is the migration velocity, q is the charge on the
particle, E is the precipitating electrical field strength, a
is the particle radius, ft is the gas viscosity, and n is the
irrational number pi.
The collection performance for ESP can be predicted by
a range of models (Park et al., 2004). One of the simplest
relationships to predict the collection efficiency is the
Deutsch-Anderson equation:
vi-A
r\ =\-e Q
where r\ is the collection efficiency, A is the total surface
available for collection, w is the migration velocity, and Q
is the flow rate of air through the EAC device. As migration
velocity (w) or collection area (A) increases, the collection
efficiency increases. As the flow rate (Q) increases, the
collection efficiency decreases. This relationship makes a
number of assumptions:
• Particles are charged instantaneously.
• Turbulent and diffusive transport causes particles to be
uniformly distributed through the device.
• Gas velocity does not affect migration velocity.
• Viscous drag follows Stokes' law.
• Particles always move at migration velocity and are
identically sized.
• Mutual repulsion between dust particles can be neglected
because they are sufficiently separated.
• Collisions between ions and neutral gas molecules are
neglected.
• Unusual effects such as uneven gas flow, backward
corona, erosion or particle reentrainment are neglected.
7.3 Summary of Relevant Studies
Recent studies of ESP/EAC devices are summarized in
Table 16. Note that Table 16 focuses on recent studies
(conducted since 1995). In the reviewed literature, the
purposes of the studies varied. Several studies focused on
measuring the effectiveness of ESP devices in reducing
indoor air paniculate concentrations. Other studies measured
the performance of ESP devices by quantifying the collection
efficiency as a function of particle size. Some studies also
examined other variables, including face velocity, ionization
voltage, and even corona polarity.
The devices in the reviewed literature also varied
significantly. Several papers reviewed stand-alone ESP/
EAC devices that were not connected to a building's HVAC
system but instead were self-contained devices that sample
the air and filter it. Other papers reviewed devices that were
designed to operate within an HVAC system. One study
appears to discuss an ESP device used with a window-
mounted room air conditioner (Park et al., 2002), and one
other study examines an industrial ESP system (Zukeran et
al., 1999). It is not clear exactly what types of devices are
used in some studies.
A number of studies directly address the performance, or
collection efficiency, of ESP/EAC devices. ESP performance
is impacted by the size of the particle to be collected, the
velocity of the air and particulates through the device, and
the magnitude of the electrical charge applied to the particles.
Some studies consider additional variables based on the
particular device studied. Part of the device performance
is the pressure drop associated with the device, though
because ESP devices typically have lower pressure drops
than traditional HVAC filters, pressure drop was not often
measured in the reviewed studies. Several studies also
addressed collection of biological particles specifically.
7.3.1 Performance and Variables That Affect
Performance
In general, the type of particle to be collected does not
impact the collection efficiency of an ESP device. Studies
by Morawska et al. (2002), Howard-Reed et al. (2003), and
Mainelis et al. (1999) comparing several types of particulates
show that neither the type of particle nor the particle shape
has a significant effect on particle collection. However,
particle size is a strong determinant of collection efficiency.
As with traditional types of high efficiency filters, most data
sets demonstrate a particular particle size with a minimum
collection efficiency, often referred to as the most penetrating
particle size (MPPS). The value of the most penetrating size
will depend on the operating conditions of the ESP device,
but values of the MPPS with ESP are similar to values for
traditional filters (particle diameters of around 0.3 um, Huang
and Chen, 2001). However, Huang and Chen (2002) note that
collection efficiency of ESP devices has also been observed
to decrease with decreasing particle size for particles smaller
than tens of nanometers.
Particle collection efficiency increases with increasing
particle size over the range of 0.3 to 10 um, as demonstrated
in Figure 22. Figure 22 also shows the effect of loading on
the performance of the ESP as the collection efficiency is
reduced with time. Figure 23 shows aerosol penetration as a
function of particle size and applied voltage. As shown, the
penetration decreases across all particle sizes as the applied
voltage increases. The MPPS is also observed to shift toward
smaller particles as the applied voltage is decreased. For
comparison, the MPPS is 0.4 to 0.5 um at 8 kV and 0.2 to
0.3 um at 4.5 kV
-------
Table 16. Summary of Recent EAC/ESP Studies
Performance Data
• Free-standing units
• Duct-mounted units
• Room AC filter
Recent studies of the effectiveness of ESP filtration;
studies typically varied one or more variables,
including flow rate or face velocity, ionization voltage,
and corona polarization; performance generally
increases with increasing voltage and particle size,
and decreasing face velocity.
Huang and Chen (2001, 2002); Mainelis et
al. (1999); Morawska et al. (2002); Park et
al. (2002)
Effectiveness Data
• Free-standing units
• Duct-mounted units
Studies using ESP for reduction of airborne and/
or surface dust; ESP devices can offer significant
reduction (values reported from 20 to 85%) of
indoor particulate concentrations as tested in both
residential and commercial use.
Croxford et al. (2000); Richardson et
al. (2001); Howard-Reed et al. (2003);
Wallace et al. (2004); Emmerich and
Nabinger (2001); Fugler et al. (2000)
Degradation With Duration of Use
Describes method for appropriately simulating filter
degradation to quantify reduction in performance
with time. Approach appears reasonable, though did
not produce anticipated results for all devices tested.
Hanleyetal. (2002)
Models of Performance or
Effectiveness
Detailed experimental studies linked with a detailed
indoor model, employing a network of well-mixed
volumes with good agreement.
Howard-Reed et al. (2003); Wallace et al.
(2004)
100
>%
o
c
CD
'o
y=
LJJ
c
o
After 2 months
light usage
After 1 month
heavy use
After 3 months;
2 months full-time
plus 1 month partial
0
0.1 1 10
Particle Diameter (micrometers)
Figure 22. Collection Efficiency as a Function of Particle Diameter After HVAC Use
(Hanleyetal., 2002)
-------
o
"ro
"0
c
0
Q.
0
120
100
80
60
40
20
_ Voltage, +kV
the best fit curve
Challenged with polydisperse particles
Flow rate = 30 L/min
0.01
0.1 1
Particle diameter, |jm
10
Figure 23. Aerosol Penetration as a Function of Particle Diameter at Various Applied Voltages
(Huang and Chen, 2001)
Several studies have shown that collection efficiency may
actually increase slightly for some small particles, as seen
by U-shaped efficiency or penetration curves in Figures
23 to 28. Figure 23 shows the aerosol penetration (1 -
collection efficiency) for particle diameters of 0.03 to about
7 um. Aerosol penetration decreases (collection efficiency
increases) as particle diameters decrease below about
0.3 um. In Figure 24, several curves of collection efficiency
for particles ranging from 0.5 to 10 um in diameter show
increasing efficiency below 1 um. Figure 24 also shows a
peak in collection efficiency around 5 um particle diameter,
with lower collection efficiencies above that threshold.
The increased collection of very small particles is also
demonstrated in Figure 26, with particle diameters ranging
from 0.03 to about 10 um.
There are significant differences in the experimental ESP
systems used to generate the data in Figures 22 through
26. Park et al. (2002, Figure 24) studied a small filter for a
window air conditioner. Hanley et al. (2002, Figure 22) and
Morawska et al. (2002, Figure 25) both studied duct-mounted
commercially available two-stage ESP devices. Huang
and Chen (2001, Figures 23 and 26) studied a miniature
ESP device (less than 11 cm long) from a commercial air
cleaning device. Huang and Chen (2002, Figures 27 and
28) used a longer ESP device (30 cm long) in a later study.
The measurement devices used in the various studies also
varied significantly. The wide variety of testing environments
increases the likelihood that the most penetrating particle
diameter would be different for the various ESP devices. For
any particular device, there will be an optimum combination
of voltage, face velocity, and other variables.
Huang and Chen (2002) assessed the penetration of
nanoparticles with nominal diameters ranging from 0.01
to 0.06 um through an ESP. In previous studies by Huang
and Chen, the smallest particle size evaluated was 0.03 um.
The results are shown in Figures 27 and 28. Consistent with
the previous work, the penetration is observed to decrease
with decreasing particle size over the range of 0.1 to 0.03
um. However, a minimum is observed at about 0.015 um,
and penetration begins to increase as particle size decreases.
The effect was minimized at the highest applied voltages as
penetration of particles less than 0.03 um was less than 1%.
However, the authors attributed the poorer performance at the
lower voltages to partial charging of the nanoparticles.
-------
0
-^
*>
u
1
OS
06
0.7
04
05
04
OJ
Q2
01
0
01
w /o dtfccbi filer
* ISO few 0.192 » /s}
1
dp ^ire)
10
Figure 24. Collection Efficiency as a Function of Particle Diameter for Several Face
Velocities (Park et al., 2002)
50.00
-4—1,0501/s, SMPS
••— 1,024 1/s, SMPS
-fr— 944 1/s, SMPS
X 800 1/s, SMPS
• 708 1/s, SMPS
•••*•••• 560 1/s, SMPS
—t 472 1/s, SMPS
•••O 1024 1/s, APS
A 944 1/s, APS
X 800 1/s, APS
X 708 1/s, APS
-»—472 1/s, APS
Figure
0.10 1.00
Aerodynamic particle diameter (pm)
25. Collection Efficiency as a Function of Particle Diameter for Various Face Velocities
(Morawska et al., 2002)
10.00
-------
0
c
0
Q.
"o
(n
2
0
120
100
80
60
40
20
close symbol - monodisperse
open symbol - polydisperse
Flow rate, L/min
Voltage = +8 kV
0.01
0.1
10
Particle diameter, Lim
Figure 26. Aerosol Penetration as a Function of Particle Diameter for Several Face Velocities
(Huang and Chen, 2001)
80
60
-i—•
CD
1 40
0
Q.
0 20
Single-stage ESP
Air flow rate, L/min Ni, 1/cm3 t. sec
Applied voltage = 15.5 kV
Wire diameter = 0.3 mm
o 50
» 100
o 150
3.41 x104 3.28
1.70x104 1.64
1.14x104 1.09
10
*
,%
100
Particle diameter, nm
Figure 27. Aerosol Penetration as a Function of Particle Diameter for Several Face Velocities
(Huang and Chen, 2002)
-------
The face velocity or flow rate of air through an ESP device
will significantly affect the collection efficiency. In general,
collection efficiency is reduced as the face velocity increases.
The higher particle velocities create a shorter residence time
for the particles to be attracted to the collection plates. In
Figures 24 and 25, collection efficiency clearly increases as
the face velocity or flow rate decreases. Figures 26 and 27
demonstrate the same effect, though aerosol penetration is
shown rather than collection efficiency. As flow rate through
the device increases, aerosol penetration increases (collection
efficiency decreases) due to a reduction in residence time.
Again, it is difficult to directly compare the results of the
different studies because of the diversity of devices tested,
as well as the differences in experimental conditions and
measurement devices. However, all studies indicated that
collection efficiency decreases as face velocity increases.
Increasing voltage will generally increase the collection
efficiency. In Figures 23 and 28, aerosol penetration curves
for several applied voltages are shown. As applied voltage
is increased, the aerosol penetration decreases (meaning
the collection efficiency increases). Figure 29 establishes
a similar trend of increasing collection efficiency with
increasing voltage.
80
60
o
I
"0
Q.
"o
O
40
20
Single-stage ESP
Applied voltage, -kV
Air flow rate = 100 L/min
Wire diameter = 0.3 mm
100
Particle diameter, nm
Figure
28. Aerosol Penetration as a Function of Particle Diameter at Various Applied Voltages
(Huang and Chen, 2001)
-------
c
o
1
03
04
0.7
0£
DJ
0,2
01
48kV/Qi
hokdfc.*4R*
Q -100 km 003***}
v /o dkbcofc
01
10
Figure 29. Collection Efficiency as a Function of Particle Diameter for
Several Applied Voltages (Park et al., 2002)
The polarity of the corona used to ionize particles for
collection affects the collection efficiency. As noted by earlier
authors of ESP theory, particle collection can be higher
with negative corona ionization under identical operating
conditions because of the greater electrical currents generated
at the same voltage (Oglesby and Nichols, 1970; White,
1963; Rose and Wood, 1956). An example of this effect from
a recent study can be seen in Figure 30. In Figure 30, best
fit lines for positive corona voltages are shown with data for
negative corona voltages. Negative corona voltages of 4 kV
and 6 kV perform as well as positive corona voltages of 5 kV
and 8 kV, respectively.
In the literature reviewed, only Park et al. (2002) measured
pressure drop of the ESP device studied. These authors had a
specific interest because they were creating an ESP filtering
device for a room air conditioner with specific pressure
drop limitations. This device differs significantly from other
ESP devices for room air cleaners and duct-mounted ESP
devices. Measured pressure drops for this device can be
seen in Figure 31. A number of mechanical configurations
relating to the size and spacing of holes in the collection plate
were tested over a range efface velocities. In general, fewer
holes and, to a lesser extent, smaller holes resulted in lower
pressure drops. Other authors simply state that the pressure
drop is much lower than traditional or HEPA filtration.
As with traditional high-efficiency filtration, the performance
of ESP devices can degrade over time. Performance
degradation with ESP devices can be due to several effects.
With residential ESP devices, dust loading (the amount of
dust collected) may affect performance. Howard-Reed et al.
(2003) and Wallace et al. (2004) both state that "frequent
cleaning" was required to maintain high efficiency for the
ESP device studied. No indication of the type of cleaning was
given. Figure 32 shows the decrease in collection efficiency
with time for fine (0.3 to 2.5 um diameter) and coarse (2.5 to
10 um diameter) particles. Arrows at the top indicate when
the unit was cleaned.
-------
c
o
?
0)
Q.
"5
(0
2
0)
120
100
80
60
40
20
the best fit curve of
the data in Figure 5
Flow rate = 30 L/min
Voltage -kV
0.01
0.1
1
10
Particle diameter, |jm
Figure 30. Aerosol Penetration as a Function of Particle Diameter for Several Positive and Negative
Corona Voltages (Huang and Chen, 2001)
F«c«
Figure 31. Pressure Drop as a Function of Face Velocity for Various Filter Configurations
(Parketal., 2002)
-------
0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000
time (h)
Figure 32. Overall Collection Efficiency as a Function of Time With Cleaning (Wallace et al., 2004)
Hanley et al. (2002) discovered that dust loading was a poor
predictor of performance degradation of ESP devices with
time. Performance of the ESP devices studied by Hanley
et al. did decrease with time, as seen in Figure 33, but the
observed degradation was due to decreased corona ionization
attributed to silicon deposits on the ionization wires. A new
testing protocol was suggested to replicate the degradation
of the device with time in use. The ESP device was to be
operated in a sealed test chamber with a source of liquid
silicon to allow significant amounts of silicon to deposit on
the ionization wires. Figure 33 demonstrates the degraded
performance observed for two different ESP devices after
exposure in the sealed silicon test chamber. Note that the
performance of the second device (on the right) did not
degrade as quickly as the first.
Though some authors have indicated a significant
dependence of collection efficiency on dust loading, others
have found none. Howard-Reed et al. (2003) and Wallace
et al. (2004) indicated degraded performance with time,
but it is not known what caused the degraded performance.
Cleaning restored the efficient operation. Hanley et al. (2002)
observed no significant reduction in ESP performance with
increasing dust on the device collection plates but observed
that deposition of impurities on the corona wires occurred
over time and did degrade performance. It is possible that the
effect seen by Howard-Reed et al. and Wallace et al. is the
same as the effect identified by Hanley et al. (2002).
0.1 1 10 0.1 1 10
Particle Diameter (micrometers) Particle Diameter (micrometers)
Figure 33. Degradation of Filtration Efficiency After Exposure in Silicon Vapor Chamber for Two ESP Devices
(Hanley etal., 2002)
-------
Several studies specifically address collection of biological
particles. In Mainelis et al. (1999), collection of biological
particles was determined to occur in the same ways as
collection of nonbiological particles. An ESP device
was modified to use three different sampling media
(agar, water, and filter material). In this study, biological
organisms were collected in order to identify them,
meaning that the organisms must remain culturable (or
alive). Three different organisms were tested, a spore-
forming organism (Bacillus subtilis var niger, or BG, a
relative of the organism causing anthrax), a bacterium
with high resistance to drying, disinfecting, and other
environmental processes (Mycobacterium bovis, a
relative of the organism causing tuberculosis), and a more
sensitive bacterium (Pseudomonasfluorescens). BG was
collected with greater than 90% efficiency on the filter
substrate, about 55% on agar, and about 20% in water. M.
bovis was recovered with an efficiency ranging from 0
to 8% in all three media. Very little of the more sensitive
P.fluorescens was recovered. The physical collection
efficiency of the ESP device was about 90% for both
biological and nonbiological (polysterene latex) particles.
In subsequent studies, Mainelis et al. (2001 and 2002b)
examined the electrical charging on airborne microorganisms
and the effects of electrical charging and fields on airborne
microorganisms. Based on those studies, Mainelis et al.
(2002a and 2002c) designed a new ESP device to maximize
the bio-recovery of microorganisms by collection on agar.
This device was able to achieve about 90% overall physical
collection efficiency (2002c). In Mainelis et al. (2002a),
BG, P.fluorescens, and Penicillium brevicompactum
(a fungal spore causing respiratory infections and
allergies) had biological collection efficiencies of about
70%, 20%, and 75%, respectively. These bioefficiencies
compared well with results from another biological
sampling device, the BioSampler (SKC, Inc., Eighty
Four, Pennsylvania). The ESP device showed nearly
equivalent collection for BG and P. brevicompactum
and higher collection efficiency for P.fluorescens.
Yao and Mainelis (2006) investigated whether natural
electrical charges on airborne organisms can be used for
their collection without the need for active charging. The
specific application was development of a low-volume
novel air sampler (< 20 L/min). The resulting collection
efficiency was independent of particle size over the range
tested (0.3 to 3.0 um) but decreased dramatically from about
80 to 30% as flow rate increased from 1.2 to 10 L/min. The
study demonstrates the potential of the bioaerosol sampler,
as the collection mechanism does not stress the organism
as much as inertial or impaction methods. However, the
authors also note considerable "day-to-day" variability due to
differences in the charge levels of the organisms, potentially
due to differences in weather conditions or organism source/
generation method. Thus, actively charging an aerosol would
be preferred for consistency in an HVAC application.
For both of the above devices, the biological collection
efficiency is lower than the overall collection efficiency for
several reasons. First, some of the collected microorganisms
were shown to be collected on surfaces other than the agar.
Only organisms collected on the agar were counted towards
the bioefficiency. Second, some of the organisms collected
can be injured during or after collection, further increasing
the difference between the biological collection efficiency
and the overall collection efficiency. Also, any losses due
to the aerosolization process are not accounted for in this
study, though P. fluorescens in particular may be subject to
degradation during aerosolization.
7.3.2 Assessment in an HVAC System
Measurements of collection efficiency are useful in
determining the performance of a device, but another
measure of performance is the effectiveness of the device in
real applications. Several authors have examined ESP/EAC
devices from this perspective.
Emmerich and Nabinger (2001) and Fugler et al. (2000)
measured the performance of several in-duct filtration
devices, including ESP devices. Emmerich and Nabinger
reported collection efficiencies ranging from about 96%
for particle diameters of 1 um and less to 91% for 1-5 um
diameters. Fugler et al. measured collection efficiencies of
84% for PM1 and 90% for PM10 particles. In both studies,
the ESP device clearly outperformed the other devices
tested, which included a range of mechanical, electret, and
electrostatically enhanced filters.
Several studies examined the effectiveness of in-duct
mounted ESP devices. Howard-Reed et al. (2003)
and Wallace et al. (2004) both presented results of the
reduction in indoor paniculate concentrations within a
three-story townhouse. The townhouse was occupied and
both mechanical and ESP filtration were studied. Both
authors calculated the deposition rates of several aerosols
using a mathematical model. Howard-Reed et al. reported
that the in-duct ESP reduced particle concentrations by
57-85% for 0.3 to 10 um particles. Wallace et al. studied
0.01 to 0.1 um and 0.54 to 2.5 um particles in the same
way (though different particle measurement devices
were used). In this second study, a reduction of 44-59%
was reported for the particle sizes studied. Both studies
concluded that simply running the HVAC central fan would
significantly reduce concentrations by 14-50%. Fugler
et al. also presented data on the reduction in indoor dust
levels for in-duct ESP with values of 31% when occupants
were active and 71% when occupants were inactive.
The literature search also revealed several other studies in
which the use of ESP was studied in office spaces. Croxford
et al. (2000) and Richardson et al. (2001) both studied
the use of several ESP devices located within the office
spaces themselves, though it is not clear whether these
devices were simply ionizers or also included collection
plates. Croxford et al. found that using ESP devices
within the "breathing zone" results in a 49% reduction in
particles 2 um and less, about a 46% reduction in particles
10 um and less, and an overall reduction of 37% for all
particle sizes. The authors concluded that the devices used
-------
were more effective at removing the smaller particles.
Richardson et al. reported a 21% reduction in indoor
particle concentrations for particles of 3 um and less.
There is a wide difference in the reduction of particle
concentrations reported in these two studies. The incomplete
description of the devices used in Croxford et al. (2000)
and Richardson et al. (2001) makes it difficult to determine
the reason for the differences. The locations, buildings, and
measurement devices were also different in these studies.
7.3.3 Additional Factors
While ESP remains an effective technology for air cleaning,
there are some negative effects. Ozone concentration and
generation of excessive ionization are possible problems.
Power consumption in general has not been well examined,
at least in the studies reviewed for this effort. It is not clear
whether the electrical power required for ESPs to function is
offset by the reduced pressure drop of these devices.
ESP devices form ozone, and to a lesser extent, other
nitrogen by-products. In fact, though negative polarity corona
results in more advantageous operation, much more ozone is
produced than with positive corona, as much as 5 to 6 times
as much (Huang and Chen, 2001). As a result, most indoor
air cleaning applications use positive corona for this reason.
Note that the time-weighted average (TWA) for ozone is 0.1
ppm (National Research Council, 1984).
Measurements of ozone concentration vary and will be
device and experiment dependent. Huang and Chen (2001)
measured ozone concentrations over 0.2 ppm with positive
corona devices and over 1.6 ppm with negative corona in 30
L/min of air, as shown in Figure 34. When mixed into the air
of a room (this device was from a room air cleaner), these
levels would likely become lower than the NIOSH TWA of
0.1 ppm. Grabarczyk (2001) reported that ozone exceeded
the smell threshold (about 0.005 ppm) after two hours of
operation. Fugler et al. (2004) measured ozone levels in
houses equipped with ESP devices and concluded that indoor
levels of ozone were similar to those measured outdoors.
Another unintended consequence of using ESP technology
for indoor air is the accumulation of ions in the air. As noted
by Grabarczyk (2001) and Lee et al. (2004a), objects within a
room can become charged and result in static electric shocks.
Furthermore, the accumulation of charged dust particles will
result in significantly increased deposition on indoor surfaces
(Grabarczyk, 2001). These effects are more significant with
whole-room ionization technologies that use devices that do
not also collect the charged particulates. These effects also
may occur with normal ESP devices that are collecting at
very low efficiency (ASHRAE, 2004).
6,000
.Q 5,000
Q.
Q.
O 4,000
?
5
•+•»
0) 3,000
O
c
O
O
0) 2,000
O
N
O
1,000
Flow rate = 30 L/min
0.0
• Positive
O Negative
0.1
0.2
0.3
0.4
0.5
Current, mA
Figure 34. Ozone Production as a Function of Current for Positive and Negative Corona (Huang and
Chen, 2001)
-------
7.4 Critical Assessment
7.4.1 Technology Assessment
ESP devices generally offer high filtration efficiency at
a low pressure drop. A lower pressure drop (than that
of mechanical filtration) is associated with lower power
consumption, but it is unclear whether this benefit is nullified
by the power required to run the ESP. The several studies
comparing residential ESPs with other types of residential
filtration have demonstrated the superior performance of the
ESPs (Emmerich and Nabinger, 2001; Fugler et al., 2000).
Residential ESP units are limited, as most ESP units are for
commercial applications.
In the review of the research regarding ESP collection
efficiency, it was clear that test aerosol challenges and
conditions (either environmental or equipment operation
such as field strength) are not standardized. Thus, data are
not easily compared between studies, and therefore not easily
comparable between units or technologies. Recent work by
EPA as part of the ETV program has made an effort to test
ESPs for residential use in a consistent manner. The EPA
data, as well as data from many other authors, suggest that
the composition of the test aerosol used to measure collection
efficiency is not critical. It does appear as though the particle
composition and size used for any loading or preconditioning
of the ESP prior to collection efficiency measurements is
very important. As previously mentioned, the work of Hanley
(2002) has led to the recommended use of nano-sized KC1
particles for loading.
The purpose of conducting performance tests with ESPs
that have been conditioned is to simulate the rapid and
significant decrease in collection efficiency with use. Hanley
et al. (2002) have attributed degradation to the formation of
silicon deposits on the ionization wires. Hence, the practice
of preconditioning an ESP by exposure to silicon vapor as a
means to simulate in-use operation is recommended.
As long as the selected ESP is cleaned regularly, performance
can be maintained at a relatively high level (Hanley et al.,
2002; Howard-Reed et al., 2003; Wallace et al., 2004).
Cleaning of the ESP collection surfaces has demonstrated
the ability to "regenerate" the initial operating performance
of ESPs. There is no definitive work regarding the number
of cleaning cycles that may be used before the degradation
is permanent. The literature did not suggest specific cleaning
techniques to be used.
The recent work conducted by Battelle as part of this current
project has shown an increase in aerosol penetration of a
commercial ESP unit for particles less than about 0.05 um.
The increase in penetration for the nanoparticles is believed
to be due to particle charging efficiency. The charge that can
be maintained on the particles is relatively low, and thus the
capture efficiency begins to drop.
Finally, ESPs can have a biocidal effect on collected
microorganisms under some operating conditions. In general,
the survival of organisms decreases as the operating voltage
increases, especially for relatively delicate organisms such as
viruses or vegetative cells. Robust organisms are relatively
unaffected by typical operating voltages. Survival of the
organisms also depends to some extent on the collection
media. In a standard ESP with metal collection plates, only
the most robust organisms will not be injured by the dry
conditions. High collection voltages and harsh collection
conditions can be used to disinfect collected particulates to
some degree.
Because ESPs can also operate in a range that allows for the
survival of many organisms, ESPs may be used as effective
bio-sampling devices. Mainelis et al. (2002c) demonstrated
that about 70% of robust organisms and 20% of vegetative
organisms can be recovered unharmed using an ESP
collection device. Compared to a typical biosampling device,
the ESP bio-efficiencies were equal and, in fact, greater for
the vegetative organism.
Because ESP devices can generate ozone, ozone levels
should probably be monitored, particularly when used in the
homes of people with particular sensitivity to ozone. For this
reason, most devices intended for occupied spaces operate
with positive polarity coronas rather than the more efficient
negative polarity corona since the negative polarity corona
generates more ozone.
7.4.2 Impact on HVAC System
As described above, ESP devices offer much lower pressure
drops than traditional high-efficiency filters. Because of
the lower pressure drops, the effect on HVAC systems is
minimal. An indirect impact might be the electric power
required to run a particular ESP device. The higher the
voltage used to ionize and collect particles, the greater the
collection efficiency will be. However, greater collection
efficiencies will come with the addition of higher electrical
power cost. Note that applied voltage is something typically
determined by the manufacturer as part of product design and
is probably not something that would be adjustable by the
end user of an ESP device.
7.4.3 Cost Analysis
Electrostatic precipitators, because of their design, typically
will not fit into an existing air handler without major
modifications. For this reason they may be installed outside
the air handler, in the ductwork. This arrangement would
require modifications to the existing ductwork, which may
demand additional services for redesign of the ductwork
system. These filters would also require new electric service.
For the above reasons, the installation and initial purchase
costs of these filters are very high.
-------
Despite their high initial costs, these niters have small building includes the labor required to clean the prefilters and
operating and maintenance costs. In comparison to electrostatic precipitators. The increase in operating cost of a
mechanical filtration, there is a small increase in static typical office building would be 3%.
pressure and a small increase in the electricity used to power An in.depth CQSt ^^ of dectrostatic precipitation is
the ESP, but they do not require penodic changing. An provided is Section 9.5.
increase of 12% in maintenance costs for a typical office
-------
8.0
Critical Assessment of
Ultraviolet Germicidal Irradiation
8.1 Technology Description
The uses of ultraviolet (UV) radiation can be classified into
two general categories, air and surface disinfection. A variety
of systems have been developed to accomplish these two
applications. Air disinfection systems fall into three general
categories: (1) in-duct air disinfection systems for reducing
the circulation of infectious material (e.g., tuberculosis) in
a facility, (2) recirculation systems used to treat the air in a
room, and (3) "upper air" disinfection systems, which consist
of UV lights mounted in a room so that the air above them
is irradiated. To limit exposure of the occupants, upper air
disinfection systems are installed at heights greater than
7 ft (~2 m) above the floor and/or some type of shielding
panels are used. Surface disinfection systems can be
classified into four general applications: (1) microbial growth
control systems such as UV exposure of a filter surface,
(2) laboratory disinfection such as the UV lights used in
biosafety cabinets, (3) portable disinfection systems, and
(4) mail room decontamination systems. Both air purifying
and surface decontamination systems are used in hospitals,
shelters, prisons, and clinics (Kowalski and Bahnfleth,
2000b). Unlike other previously discussed filter technologies,
UV radiation is used to kill the biocontaminant as opposed
to removing it from the air stream. While it is possible that
these systems could be used in commercial and residential
buildings, their application is not yet common.
8.2 Theory of UVGI
UV radiation in wavelengths of 225 to 302 nm is frequently
used for microbial disinfection (Kowalski and Bahnfleth,
2000b). UV radiation kills microorganisms by damaging
their DNA and, to a lesser extent, causing oxidation of their
proteins. DNA absorption of UV radiation is maximal at 254
nm and leads to the hydrolysis of cytosine and the formation
of thymine dimers (Snustad et al., 1997). Thymine dimers
prevent DNA replication while the hydrolysis of cytosine
can lead to base pair mismatches. Protein oxidation occurs
when reactive oxygen species are generated by UV radiation
and the addition of a chemical such as titanium oxide, which
releases significant amounts of oxygen upon exposure to
UV light and can facilitate the oxidation process (Lele and
Russell, 2005).
Whether or not UV radiation is lethal to a microorganism
depends on the dose that microorganism receives. Doses are
calculated from the average radiation intensity and exposure
time; the dose needed to kill a microorganism is specific to
that microorganism (Kowalski and Bahnfleth, 2003). This
dose can be approximated mathematically by the following
equation (Memarzadeh et al., 2005).
%Survival=100exp(-z / /)
where z is the susceptibility factor for the microorganism
(cm2/uW-s), / is the average radiation intensity (uW/
cm2), and / is the exposure time (seconds). Effective UVGI
doses have also been experimentally determined for many
microbial species; however, many were determined for
organisms on surfaces rather than in their aerosolized form.
Since it is easier to inactivate airborne organisms, some of
the published data may overestimate the dose required for an
air cleaning UVGI system (Brickner et al., 2003).
To be effective, UV radiation requires direct "line of
sight" exposure; therefore, having a fully developed light
field is critical. For example, most "in-duct" systems use
multiple light sources and reflective panels to create an
evenly illuminated exposure zone. Prefilters and routine
cleaning of the light sources and reflective panels may also
be incorporated to maintain a fully developed light field.
Encapsulation of microorganisms in other debris or material
can decrease the efficacy of UV radiation. This decrease in
efficacy was shown with Serratia marcescens suspended in
various solutions prior to aerosolization (Lai et al., 2004).
The need for direct exposure can be problematic for surface
decontamination as well because microorganisms in shaded
cracks are not killed.
8.3 Summary of Relevant Studies
Studies in the use of UVGI for air cleaning applications
are summarized in Table 17, where the main focus was
on those conducted since 1999. A large number of articles
discussed the potential application of UVGI in HVAC
systems, although the majority of the performance data were
for systems using UVGI for upper air inactivation or surface
decontamination. Computer modeling and articles with
design basics illustrate parameters that factor into selecting a
UVGI system specific to each building. Safety concerns over
UV exposure were also covered.
-------
Table 17. Summary of UVGI Studies
Overview
Design Basics and Modeling
Effectiveness of UVGI
Safety Considerations
UVGI is a viable HVAC option; focus has been on
tuberculosis (TB) studies, but UVGI can be used
to affect other microorganisms; most applications
promote combination with mechanical filtration or
other to cover range of particle sizes.
Both general and computationally intensive
calculations of effectiveness and application
specifics.
Most were upper-room installations; some effects
of temperature and relative humidity; calculated
efficiencies for in-duct applications.
Minimizing UV exposure and ozone production.
Brickner et al., 2003; Kowalski and Bahnfleth,
2005; Kowalski and Bahnfleth, 2003; Kowalski and
Bahnfleth, 2002; Kowalski and Bahnfleth, 2000a;
Kowalski and Bahnfleth, 2000b
Brickner et al., 2003; Kowalski, 2003; Kowalski and
Bahnfleth, 2000a; Kowalski and Bahnfleth, 2000b
Miller and Macher, 2000; Miller, 2002; Peccia et
al., 2001; Koetal., 2002; Xu et al., 2000; Xu et
al., 2003; VanOsdell and Foarde, 2002; Kowalski,
2003; Kowalski and Bahnfleth, 2003; Menzies et
al., 2003
Talbot et al., 2002; Nardell, 2002; Kowalski, 2003
8.3.1 Performance and Variables That Affect
Performance
Environmental and design variables that affect the
performance of UVGI systems are relative humidity,
temperature, air velocity and air mixing, lamp selection, the
use of reflectors, and combination of UVGI with filtration.
Each of these variables is discussed in the following sections.
Relatively humidity, especially at levels greater than 50%,
has been documented to impair the UVGI "kill" rate of
some microorganisms (VanOsdell and Foarde; 2002; Peccia
et al., 2001). Others have reported that relative humidity is
not a factor — at least in the 20% to 80% range (Ko et al.,
2002). Increased operating temperature can affect biological
inactivation by negatively impacting the output of the UV
lamps. Temperature was shown by Ko et al. (2002) to have a
measurable effect on kill rates.
Air velocity and air mixing can affect the effectiveness of
UVGI. The majority of the literature with experimental
results dealt with UVGI in upper air cleaning applications.
UVGI alone was found to reduce the culturable airborne
bacteria between 46% and 80% for B. subtilis spores,
between 93% and 98% forM. parafortuitum and between
96% and 97% forM. bovis BCG cells, depending on
the ventilation rate (Xu et al., 2003). Incomplete mixing
decreased effectiveness by 80% compared to complete
mixing conditions (Xu et al., 2000). Air velocity and air
mixing in an in-duct system would need to be sufficient to
supply the demands of the building but not in such excess
as to reduce the effective dose of the incorporated UVGI
system. This theory is discussed in more detail by Kowalski
(2003), but no experimental data were found in the literature.
The combination of UVGI with filtration is discussed
as a viable option to combine the efficiencies of
each to compensate for the areas in which the other
performs less effectively. Using experimental data on
filtration efficiency and calculated UVGI performance,
Kowalski (2003) illustrated the effectiveness of
combining mechanical filtration with UVGI, as shown
in Figure 35. Only microbes with known UVGI rate
constants were included, ordered in size from smallest
(1) to largest (33), including many BW agents.
100-
fl I f I ( f i / ( I I I I I -I—L ! 1 I I \ t { \
/ .-'i C£-^~~^^-vtlX^r~r- Initial Concentrations
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
Microorganism
Figure 35. Microbial Populations Before and After Filters and a UVGI System
(Kowalski, 2003)
-------
A study performed by Menzies et al. (2003) assessed the
reduction in microbial contamination after UVGI was applied
to drip pans and cooling coils within the ventilation system
of an office building. Although the use of UVGI led to a 99%
reduction of microbial contamination on exposed surfaces,
airborne microbial levels did not decrease significantly.
Kujundzic et al. (2006) characterized the performance of
six in-room air cleaners, including a HEPA filter device,
electrostatic filter device, ESP, and air ionizer. Tests were
also performed with the HEPA filter and ESP in combination
with UVGI. The cleaners were challenged in an 87 m3 test
room with three biological aerosols, two bacterial and one
fungal. Cleaner performance was described using the clean
air delivery rate (CADR), which represents the amount of
particle-free air produced. The HEPA filter and ESP provided
the highest removal rates. These rates were increased by
a factor of 2 to 3 when used in combination with UVGI.
Several of the filtration-based air cleaners were equipped with
internal UV lamps to kill bioaerosols that penetrated the filter
or to prevent growth on the filter. The authors note that these
UV lamps had no effect on the removal rates from the room.
8.3.2 Assessment in an HVAC System
The merits of including a UVGI system in an HVAC
system are discussed in detail by Kowalski and
Bahnfleth (2000a, 2000b 2002, 2003). Kowalski and
Bahnfleth (2003) indicated that a UVGI system was
going to be retrofitted into the ventilation system of
the administration building of the Memphis Light Gas
and Water (MLGW) company to augment the existing
air cleaning system; however, no experimental data
from this installment were found in the literature.
Design considerations must be made when integrating UVGI
into an HVAC system. UVGI in-duct systems need to be
appropriately designed to the specific building. Overdesign
results in prohibitive costs and high energy consumption,
and underdesigned systems are rendered ineffective. Lamp
and reflector selection are key in obtaining the appropriate
dose for the stated disinfection goal. Economics dictate that
the lamp be appropriately sized to the building/application.
Kowalski and Bahnfleth have written numerous articles
detailing UVGI design basics for air and surface disinfection
as well as a model for predicting the rate of air stream
disinfection to improve system design (Kowalski, 2003;
Kowalski and Bahnfleth, 2003; Kowalski and Bahnfleth,
2002; Kowalski and Bahnfleth, 2000a; Kowalski and
Bahnfleth, 2000b). UVGI systems are commonly located
downstream from the air intake filter bank but upstream
of the cooling coils. Larger particles are more efficiently
removed by filters than killed by UVGI, and the use of
filters upstream helps maintain a fully developed light
field by reducing deposition on the lamps and reflectors.
The additional heat created by the UVGI lights must be
dissipated through modified HVAC design or an increase in
cooling coil performance. Although UVGI systems can be
installed in the return air ducts to inactivate any recirculated
microorganisms, this type of installation is less common
except in specific medial applications.
Kowalski and Bahnfleth (2000a) discuss in detail the
four computational aspects of UVGI essential for the
accurate modeling of air stream disinfection systems — the
exponential decay curve, the lamp intensity field, the direct
reflected intensity field, and the inter-reflected intensity field.
Model results have been corroborated with laboratory tests
within ±15%.
8.3.3 Additional Factors
Safety. Two safety considerations apply when using
UVGI for air cleaning applications: UV exposure and ozone
production. Literature related to UV exposure cited accidental
occupational exposure with unshielded upper-room air
installations (Talbot et al., 2002). An upper-room air system
needs to be installed at a minimum height with or without
additional shielding to reduce UV exposure (Nardell, 2002;
Miller et al., 2002). For in-duct applications, UV exposure is
less of a concern since personnel and UV lamps would not
occupy the same space. Simple precautions would need to
be taken by maintenance personnel servicing the system.
In other applications, UV lamps are used specifically
to generate ozone, which aids in the destruction of
microorganisms. Unfortunately, high levels of ozone can
be harmful to health and constitute a respiratory irritant.
Non-ozone-producing lamps are available for UVGI systems
(Kowalski, 2003), although no other mention of ozone
production or the possible effects was encountered in the
reviewed literature. Further investigation of the levels of
ozone created during UVGI in-duct applications is warranted.
8.4 Critical Assessment
8.4.1 Technology Assessment
The combination of UVGI and mechanical filtration appears
to be the most likely use of UVGI due mainly to the fact
that UVGI systems would probably be added to a current
HVAC system that already employs some type of mechanical
filtration. This application is advantageous since UVGI is
most effective against biocontaminants in the particle size
range where mechanical filtration is less efficient (lum and
smaller). Since UVGI kills the biocontaminant but does
not capture the actual particle, further mechanical filtration
downstream of the UVGI system may be necessary.
Retrofit UVGI systems would have to be installed
downstream of the original mechanical filtration
to aid in maintaining the fully developed light
field. Periodic cleaning of the lamps will also be
important in establishing the light field in order to
maintain the effectiveness of the UVGI system.
Unfortunately, little to no experimental data exist on
HVAC applications of UVGI. It is highly recommended
that any research in the future be specifically directed
to this area so that the benefits of UVGI can be
determined. Due to the high initial, operating, and
maintenance costs of a UVGI system, the benefits of
the system must be profound to outweigh the costs.
-------
8.4.2 Impact on HVAC System
A UVGI air cleaner can be installed in-duct in existing
ventilation systems, although modifications are required.
The addition of UV bulbs and reflective material in a
crossflow configuration adds a negligible pressure drop to
the existing HVAC system. Any additional pressure drop
on the HVAC system would come from including a higher
MERV level prefilter than currently exists in the HVAC
system. This additional filter would impact the system as
discussed previously under Mechanical Filtration, Section
4.4. Additional power would be required to operate the UVGI
lamps as well as cool the air stream from the heat generated
by the UVGI system.
8.4.3 Cost Analysis
UVGI, because of the design requirements, typically will not
fit into an existing air handler without major modifications.
For this reason, the UVGI may be installed outside the air
handler, in the ductwork. This arrangement would require
redesign of the ductwork system. These systems would also
require new electric service to a greater extent than the other
filters analyzed here. For the above reasons, the installation
cost of these filters is very high. Because of the more
complex design of these systems, their initial purchase cost is
extremely high. The initial purchase and installation costs for
these systems are much higher than for any of the other filters
analyzed in this report.
Maintenance costs for these systems are also high as they
include cleaning and/or changing the bulbs periodically. The
increase in maintenance costs after installing these filters in
a typical office building would be 58%. Because they use
a large amount of electricity, the operating costs for these
systems are also very high; an increase in operating costs of
22% would be observed after installing a UVGI system into a
typical office building.
An in-depth cost analysis of UVGI is provided in Section 9.6.
-------
9.0
Cost Analysis
9.1 Approach
When estimating the cost (in dollars) of replacing standard
air handler filters with varying types of more protective
filters, there are six main factors to consider: (1) the initial
(manufacturer's) cost of the new filters, (2) the cost for
installation of new filters, (3) the cost of retrofits to the
existing system, (4) the cost of other services, (5) yearly
operating costs, and (6) yearly maintenance costs.
It is helpful to create a typical office building to provide a
model for this approach. Assuming a 100,000 ft2 (9,290 m2)
office building with four stories and 25,000 ft2 (2,323 m2) per
story (a typical modern suburban-type office building), and
assuming 1 cfm/ft2 (0.3 mVmin/m2) (Bell, 2000), results in
a 100,000 cfm (2,832 mVmin) system. Because selection of
a system is based on factors too numerous to list and input
on the system selection comes from both engineers and
owners, there are numerous possible HVAC systems that a
building of this size could have. One system for a building
of this size would be a variable air volume (VAV) central
system, which is becoming increasingly popular because
it saves energy. In a VAV system, all heating and cooling
is done at centralized air handlers, which consist of VAV
terminal boxes for individual zone control, a hot water boiler
for heating, and a packaged air cooled chiller for cooling.
Since a 100,000 cfm (2,832 mVmin) air handler unit would
be an extremely large custom unit (single custom units are
not typical in office buildings in the U.S. because of cost),
assume that there are four air handling units (one unit for
each floor) of 25,000 cfm (708 mVmin) each. This system
would have both return and outside air mixed within the unit
(all the return air from the building comes back to the units).
Basing each unit on a Trane Climate Changer (Trane, 2004)
unit size 50, the cross-sectional area of each unit would be
75 x 120 inches (191 x 305 cm). Assume that the unit has
two fans positioned in parallel, with both operating at 12,500
cfm (354 mVmin), 5.0 in. w.g. (1.2 kPa), and 1000 RPM,
with a 25 HP motor. By consulting the Trane manual, one
can also see that the fan's belt can be safely adjusted to 1200
RPM maximum (selections outside the bold lines are unsafe),
which gives a static pressure of 7.5 in. w.g. (1.9 kPa). The fan
would have 2.5 in. w.g. (0.6 kPa) of extra capacity. The entire
HVAC system would be controlled by a direct digital control
system, currently the most commonly accepted method of
HVAC controls. Assume that the office building is less than
one year old and is located in Columbus, Ohio. Equipment
of this type is typically located on the roof if space permits.
Assume that space permits and all equipment will be located
on the roof with adequate space for additional equipment
as well. The total first year cost is determined by summing
these terms: Cp + Q + Cs + Cr + C0 + Cm. These terms are
discussed below.
The initial purchase cost (Cp = Purchase Cost) of the
filters depends largely on the types of materials, size,
and complexity of filter design. ASHRAE Applications
(ASHRAE, 2003) Ch. 36 states, "A reasonable estimate of
the capital costs of components may be derived from cost
records of recent installations of comparable design or from
quotations submitted by manufacturers and contractors or
by consulting commercially available cost-estimating guides
and software." One such cost estimating guide is provided by
R.S. Means (2004). These cost-estimating guides are widely
accepted within the construction industry. Manufacturers
often give filter costs on a dollars-per-square-foot of filter
cross sectional area basis (the area of a filter perpendicular to
the air stream).
The installation cost (C: = Installation Cost) depends on
the size and complexity of the filters and how well the
filters fit into the existing air handler (how many man-hours
required) and who is doing the installation (labor rates).
The cost can be incurred by the building owner, an owner-
retained mechanical contractor, or in some cases even the
manufacturer. These costs can be estimated by multiplying
the number of man-hours by the labor rate and adding an
overhead rate if a contractor is used. Estimates can also be
obtained from mechanical equipment installation estimating
books such as R.S. Means Mechanical Cost Data (2004).
Service (Cs = Service Cost) from other professionals may
be needed, depending on the complexity and size of the
installation. Because equipment data can sometimes be
inaccurate as fan conditions change over time (dirty filters)
and good fan performance data for existing systems are
often hard to locate, an air balance contractor may be needed
to determine the actual cfm and operating conditions of
an existing fan/system. This information is very important
as it may determine whether the existing fan will have
the capacity to overcome the pressure increase from the
new filters (this information would determine whether a
new fan is needed or whether the existing fan just needs a
speed adjustment). As with any work involving changing
some aspect of the air handling system, an air balance is
also needed after the installation work is complete. The
air balance would be performed by a certified air balance
contractor. The services of a design engineer may also be
needed if the air handler unit or system has changed or
expertise is needed to retrofit the air handler. These costs
can be obtained from the professional in question or from
estimating books such as R.S. Means (2004).
The cost of retrofits (Cr = Retrofit Cost) to the existing
system would include the cost of any equipment that must be
added to the system or any replacements or changes that must
be made to the system to accommodate the new filters. Due
to the increase in static pressure that a new filter will cause,
-------
a new fan may be needed as a replacement for the existing
fan or as a booster fan (to boost the static pressure). In some
cases the existing fan may need only a speed adjustment,
whether by adjusting the belt drive or in cases of variable
frequency drive fans, adjusting the speed directly through
a graphical user interface. Depending on the type of new
filter, the filter may be larger in size or require a different
face velocity than the existing filter(s). The existing air
handler housing may need to change or a special housing
may be needed for the new filter. The change in filter size
from smaller to larger may also make it necessary to add
directional or straightening air vanes to existing air handlers
because of the reduced space in the air handler housing.
Some filters may require electric service with the addition
of new electrical equipment. With some filters, a new filter
monitoring system might be required to monitor the filter's
cleanliness. In buildings where humidity monitoring is
important, a new humidifier may be needed to compensate
for the loss of humidity that some filters may cause. The
costs of the above-mentioned changes to the system can
be estimated from the installers (of the equipment) or from
estimating books such as R.S. Means (2004).
The operating costs (C0 = Operating Costs) include costs
incurred by the operation of the new filters or adjustments
that must be made to the system as a result of the new filters.
Electrical power usage cost will increase if equipment such
as a new fan, electronic air filter, or humidifier is installed.
The cost of other utilities such as steam and/or water for a
humidifier could increase. Operating costs can be obtained
from a calculation of the expected energy usage multiplied
by the utility rate estimated or obtained from local utilities
companies. Operating (as well as maintenance) costs can
also be calculated by computer programs marketed by HVAC
equipment manufacturers as well as government agencies
such as the Department of Energy (DOE). The National
Institute of Science and Technology (NIST) has created
a program called the Building Life-Cycle Cost Program
(BLCC) (DOE, 2007). Operating costs can be determined by
comparing data from previous studies.
To determine the overall building energy cost for
comparison purposes, it is helpful to consider the DOE/
Energy Information Administration (EIA) 1998 Commercial
Buildings Energy Consumption Survey, which reports the
consumption and expenditures of commercial buildings
during 1995. Data from this study (reprinted in Wang, 2001)
reveal that for a central system of the type described above,
the heating energy intensity (usage per year) is 29.0 kBTU/
fWyear. This study also reveals that the cooling energy
intensity for a central system is 10.0 kBTU/ft2/year.
Applying these numbers to our model:
Heating energy: 100,000 ft2 x 29 kBTU/fWyear = 2,900,000
kBTU/year
Assuming natural gas is used with a cost of $0.00007/BTU,
the total heating cost is:
2,900,000,000 BTU/year x $0.00007/BTU = $203,000/year
Cooling energy: 100,000 ft2 x 10 kBTU/ft2/year = 1,000,000
kBTU/year
Convert to KWH (cooling will be an electrical utility):
1.000.000.000 BTU/year = 292,997 KWH/year
3,413 BTU/KWH
Assuming $0.13/KWH, the total operating cost of the cooling
system is:
292,997 KWH/year x $0.13/KWH = $38,090
Total building operating cost: $38,090 + $203,000 =
$241,090/year
Maintenance (Cm = Maintenance Cost) includes the cost
of upkeep of the filters or any equipment that was added to
accommodate the filters. Some filters may need to be cleaned
or have parts replaced on a regular basis. These costs can
be incurred by the owner, through a maintenance contract
with a mechanical contractor, or in some cases by the
manufacturer. They can be estimated using data provided by
the filter manufacturer, using data from various studies, or by
multiplying the number of man-hours by the labor rate.
To determine overall building maintenance costs for
comparison purposes, one can use the following equation
from ASHRAE (2003):
C83 = A + h + c + d + SB
where:
A = age adjustment = O.OOlSn
n = number of years the system has been in use
h = heating adjustment
c = cooling adjustment
d = distribution system adjustment
SB = mean maintenance cost
C83 = cost in 1983 dollars
The cost in 1983 dollars can be converted
(ASHRAE, 2004) to 2005 dollars (C05) by using
consumer price index data (CPI) provided by the
U.S. Department of Labor (DOL, 2007).
Cos = CPI05 /CPI83 x C83
Values for the preceding equations were found in works by
Bell (2000), DohrmannandAlereza (1986), and ASHRAE
(2003) and are listed below.
A = 0, n=0 (new building)
h = 0.0077 $/ft2 or 0.083 $/m2 (fire tube boilers)
c = -0.04 $/ft2 or -0.43 $/m2 (reciprocating chiller)
d = -0.0446 $/ft2 or -0.48 $/m2 (multi-zone distribution
system)
SB = 0.32 $/ft2 (3.4 $/m2) per year (mean maintenance cost)
C83 = cost in 1983 dollars
-------
C83 = 0 + 0.0077 - 0.04 - 0.0446 + 0.32 = 0.2431 $/ft2
(2.617 $/m2)
C05 = 195.3 799.6 x 0.2431 = 0.4767 $/ft2 (5.129 $/m2)
Considering the 100,000 ft2 (9,290 m2) building, the total
maintenance cost is:
5.129 $/m2 x 9,290 m2 = $ 47,648/year
9.2 Model Estimation for Mechanical Filtration
Using the typical office building model described above,
assume that the existing filters are replaced by high efficiency
filters with MERV 14 rating and prefilters with MERV 8. Two
new pressure gauges would be used to measure pressure drop
across the filter. These gauges would have interfaces to the
building control system. Because the existing filter would
have roughly the same static pressure as the prefilter, the only
addition in static pressure to the system is from the MERV
14 filter. This filter would typically be replaced when the
pressure drop through the filters reached 1 in. w.g. (249 Pa).
The 1 in. w.g. (249 Pa) is safely below the excess of 2.5 in.
w.g. (623 Pa) inherent in the fan selection, so a new booster
fan would not be needed (the belt would be adjusted to 1100
RPM, which would give 6 in. w.g. [1.5 kPa] static pressure
capacity). In this case, the fan is on a variable frequency
drive, which means that the fan does not operate at full
capacity. The filters installed in the existing air handling unit
have a face velocity of 400 fpm, which is below the generally
recommended maximum face velocity for commercial air
filters of 500 fpm.
9.2.1 Initial Purchase Cost (CD)
9.2.3 Service Cost (Cs)
Item Cost
MERV 8 filters (n=16)
MERV 14 filters (n=16)
New pressure gauges (2) and controls
Delivery
Total 1 unit
Total whole building (4 units) Cp
$ 258
$ 1,687
$ 820
$ 100
$ 2,865
$11,460
The above data were acquired from confidential discussions
with a leading air filter manufacturer. It was assumed that 4
air handlers would be procured, each delivering 25,000 cfm,
so that the 100,000 cfm required for the building could be
achieved.
9.2.2 Installation Cost (C;)
Item Cost
Install MERV 8 filters
Install MERV 14 filters
Install gauges/controls
Total 1 unit
Total whole building (4 units) Cj
$ 72
$ 240
$ 600
$ 912
$3,648
Filter installation costs were estimated using data from
the report entitled "Performance and Costs of Particle Air
Filtration Technologies" (Faulkner et al., 2002).
Item Cost
Engineering fees
Air balance
Total whole building (4 units) Cs
$3,650
$4,500
$8,150
The air handling units must be rebalanced and possibly
have their belts adjusted (to speed up the fan) because of the
additional static pressure added by the filters. Engineering
fees are for the investigation of the system, planning, and
design of air handler modifications. The engineering fees
estimation assumes a $50,000 total project cost. Both
numbers were estimated using R.S. Means (2004).
9.2.4 Retrofit Cost (Cr)
Item Cost
Remove existing filters
New access door
Total 1 unit
Total whole building (4 units) Cr
$ 72
$ 202
$ 274
$1,096
Labor cost to remove existing filters was based on data
from "Performance and Costs of Particle Air Filtration
Technologies" (Faulkner et al., 2002). Assuming the existing
access door is not located correctly for the new filter or big
enough to accommodate it, adding a new access door would
be necessary. The cost of a new access door was estimated
fromR.S. Means (2004).
9.2.5 Operating Cost (C0)
Total whole building (4 units) C0
The fan power operating cost increase is due to the additional
electrical power required to operate the fan at its new
pressure. This power increase is calculated from eq. 25 of
the Handbook of Air Conditioning and Refrigeration (Wang,
2001), given below:
ApV
Pf =
6356r|fr|mr|d
where:
Pf = fan energy use in hp (increase due to new filter)
Ap = static pressure drop of filter (in. w.g.) = 5 in. w.g. (1.2
kPa)
V = total system volume flow rate (cfm) = 25,000 cfm (708
mVmin)
nf nm nd = combined fan, motor, and drive efficiency (for
a VAV central system, this value is 0.55, from Table 25.1
[Wang, 2001])
-------
Power is then converted to yearly dollars using eq. 5 from
Faulkner etal. (2002):
Energy Cost =
(Fan Power)(Fan Operating Time)(Electricity Price)
In the energy cost calculation, the fan is assumed to be
running continuously 24 hours a day, 365 days per year. The
electricity price is the local price in Columbus, Ohio, using
data from the DOE's Energy Information Administration
Table 5.6.A., "Average Retail Price of Electricity to Ultimate
Customers by End-Use Sector" (DOE, 2004).
9.2.6 Maintenance Cost (CJ
Item Cost
MERV8
MERV 14
Total 1 unit
Total whole building (4 units) Cm
$ 330
$1,927
$2,257
$9,028
Maintenance costs include the cost of changing both filter
banks. It is assumed that the filters will be changed once
per year.
9.2.7 Mechanical Filtration Cost Summary
Item Cost
Total 1st year
Total yearly operations and maintenance
(0+M)
Energy increase
Maintenance increase
$47,458
$23,104
6%
19%
The cost for the first year, including initial costs and
operation (maintenance only requires yearly changing of the
filter) in our model is:
Total 1st Year Cost = Cp + Q + Cs + Cr + C0 + Cm
Using this equation and the above data, the first-year
cost for installing mechanical filtration in the model
building is $47,458.
The yearly costs (for subsequent years after the first year) can
be estimated by the following:
Total yearly cost = C0 + Cm
The total yearly cost in the above example is $23,104. This
yearly cost incurred after the first year can be useful in life-
cycle cost analysis.
For the purpose of extrapolating these data to other buildings
with similar HVAC systems, it is helpful to compare yearly
cost to total building yearly operating and maintenance cost.
By calculating the energy increase due to the filters as a
percentage of the above total building energy consumption,
one can see that the increase in operating costs is 6% of
the total building energy costs. In a similar manner, the
maintenance cost increase is 19% of the estimated total
building maintenance cost.
9.3 Model Estimation for Electrostatically En-
hanced Filtration (EEF)
Using the typical office building model described above,
it was assumed that the existing filters were replaced by
an electrostatically enhanced filtration system. A MERV 8
prefilter would be used before the EEF to keep the EEF clean
from large particles. The EEF would add 1 in. w.g. (249 Pa)
of static pressure when dirty, which allows the same fan to
be used, per the above-mentioned criteria (under 2.5 in. w.g.)
(623 Pa). The filters installed in the existing air handling unit
have a face velocity of 400 fpm, which is acceptable because
it is below the maximum recommended velocity of 500 fpm.
9.3.1
Initial Purchase Cost (CJ
Item Cost
EEF final filter
MERVS(prefilter)
Delivery
New pressure gauges (2) and controls
Total 1 unit
Total whole building (4 units) CD
$22,500
$ 258
$ 100
$ 820
$23,678
$94,712
The above data were obtained from confidential discussions
with a leading manufacturer. The EEF final filter cost
includes the cost of control modules within the filter system.
9.3.2 Installation Cost (C,)
Item Cost
Install EEF filter
Install MERV 8 filter
Install gauges/controls
Total 1 unit
Total whole building (4 units) Cj
$18,000
$ 72
$ 600
$18,672
$74,688
Estimates shown above were obtained from confidential
information provided by a leading manufacturer. The "Install
EFF filter" includes the cost of modifications to the existing
system to fit the control modules.
9.3.3 Service Cost (C,)
Item Cost
Engineering fees
Air balance
Total whole building (4 units) Cs
$4,800
$4,500
$9,300
It is assumed that the air handling units must be rebalanced
and possibly have their belts adjusted (to speed up the fan)
because of the additional static pressure added by the filters.
Engineering fees encompass the investigation of the system,
planning, and design of air handler modifications. The
engineering fees estimation assumes a $100,000 total project
cost. The project cost may actually be more, but typically
engineering fees for projects of this size would not be based
on the cost of the equipment (only a small portion of the
square footage of the building is being worked on). Both
numbers are estimated using R.S. Means (2004).
-------
9.3.4 Retrofit Cost (Cr)
Item Cost
Remove existing filters
New access door
New electric service
Total 1 unit
Total whole building (4 units) Cr
$ 72
$ 202
$ 496
$ 770
$3,080
Labor cost to remove existing filters was based on data
from "Performance and Costs of Particle Air Filtration
Technologies" (Faulkner et al., 2002). Assuming the existing
access door is not located correctly or the access door is not
big enough to accommodate a new filter, a new access door
would have to be added. The cost of a new access door is
estimated from R.S. Means (2004). New electric service
includes the cost of wiring, junction boxes, and disconnect
switches, and assumes that there is a spare breaker or breaker
space in a panel relatively close to the air handler.
9.3.5 Operating Cost (C0)
Item Cost
Fan power
EEF power
Total 1 unit
Total whole building
; (4 units) C0
$ 3,519
$ 198
$ 3,717
$14,868
The EEF requires 15 watts per filter with 20 filters. All
calculations were made in a manner similar to Section 9.2.5.
9.3.6 Maintenance Cost (Cm)
Item Cost
EEF filters (change pads)
MERVS(prefilter)
Total 1 unit
Total whole building (4 units) Cm
$ 2,576
$ 330
$ 2,906
$11,624
The above cost includes the cost of changing the EEF filter
pads and the cost of changing the prefilters.
9.3.7 EEF Filtration Cost Summary
Item Cost
Total 1st year
Total yearly operations and maintenance
(0+M)
Energy increase
Maintenance increase
$208,272
$26,492
6%
24%
The above data were calculated in the same manner as
described in Section 9.2.7.
9.4 Model Estimation for Electret Media Filtration
(EMF)
Using the typical office building model described above, it
was assumed that the existing filters were replaced by electret
media filters with prefilters. A MERV 8 prefilter would be
used before the EMF to keep the EMF clean from large
particles. The EMF would add 1.5 in. w.g. (374 Pa) of static
pressure when dirty, which allows the same fan to be used per
the above-mentioned criteria (under 2.5 in. w.g.) (623 Pa).
The fan would have its speed adjusted to 1150 RPM, which
would give 6.5 in. w.g. (1.6 kPa) total static pressure. The
new filters installed in the existing air handling unit have a
face velocity of 390 fpm, which is acceptable because it is
below the maximum recommended velocity of 500 fpm.
9.4.1 Initial Purchase Cost (CJ
Item Cost
Electret filters (n=16)
MERV8(prefilters)(n=16)
Delivery
New pressure gauges (2) and controls
Total 1 unit
Total whole building (4 units) CD
$ 2,500
$ 258
$ 100
$ 820
$ 3,678
$14,712
The above data were obtained through confidential
discussions with a leading manufacturer.
9.4.2 Installation Cost (C,)
Item Cost
Install electret filters
Install MERV 8 (pref liters)
Install gauges and controls
Total 1 unit
Total whole building (4 units)
$ 160
$ 72
$ 600
$ 832
$3,328
Filter installation costs were estimated using data from
the report entitled "Performance and Costs of Particle Air
Filtration Technologies" (Faulkner et al., 2002).
9.4.3 Service Cost (C,)
Item Cost
Engineering fees
Air balance
Total whole building (4 units) Cs
$3,650
$4,500
$8,150
The air handling units must be rebalanced and possibly
have their belts adjusted (to speed up the fan) because of the
additional static pressure added by the filters. Engineering
fees are for the investigation of the system, planning, and
design of air handler modifications. The engineering fees
estimation assumes a $50,000 total project cost. The project
cost may actually be more, but typically engineering fees for
projects of this size would not be based on the cost of the
equipment (only a small portion of the square footage of the
building is being worked on). Both numbers are estimated
usingR.S. Means (2004).
9.4.4 Retrofit Cost (Cr)
Item Cost
Remove existing filters
New access door
Total 1 unit
Total whole building (4 units) Cr
$ 72
$ 202
$ 274
$1,096
-------
Labor cost to remove existing filters was based on data
from "Performance and Costs of Particle Air Filtration
Technologies" (Faulkner et al., 2002). Adding a new access
door would be necessary, assuming the existing access door
is not located correctly or is not big enough to accommodate
new filters. The cost of a new access door is estimated from
R.S. Means (2004).
9.4.5 Operating Cost (C0)
9.5.1 Initial Purchase Cost (CD)
Total whole building (4 units) C0
$21,112
All calculations were made in a manner similar to
Section 9.2.5.
9.4.6 Maintenance Cost (CJ
Item Cost
Change electret (every 2 years)
Change prefilters (every year)
Total 1 unit
Total whole building (4 units) Cm
$1,330
$ 330
$1,660
$6,640
The above electret filter cost represents the average yearly
cost to change the filters every two years as recommended
(Manz, 2005).
9.4.7 Electret Media Filtration Cost Summary
Item Cost
Total 1st year
Total yearly operations and maintenance
(0+M)
Energy increase
Maintenance increase
$55,038
$27,752
9%
14%
The above data were calculated in the same manner as
described in Section 9.2.7.
9.5 Model Estimation for Electrostatic
Precipitation
In the electrostatic precipitation example, the electrostatic
precipitators or electronic air cleaners (as they are commonly
called) are mounted in the ductwork (outside of the air
handling unit). The size of the air cleaners and their geometry
make it difficult to fit into an existing air handler. Because
they are mounted in the existing ductwork, the ductwork
would have to be reconfigured. Fourteen air cleaners would
be used with a velocity of 515 fpm (2.62 m/s), which is
acceptable according to the manufacturer's recommendation
of 576 fpm (2.93 m/s) maximum. The air cleaners in this
example are based on electronic air cleaners from a leading
manufacturer. The pressure drop for these filters is 0.5 in.
w.g. (125 Pa) when dirty. This pressure drop requires the fan
to be adjusted to 1049 RPM to give 5.5 in. w.g. (1.4 kPa) of
total static pressure.
Item Cost
Electronic air cleaners (n=14)
Delivery
New pressure gauges (2) and controls
Total 1 unit
Total whole building (4 units) CD
$11,410
$ 1,000
$ 820
$13,230
$52,920
Electronic filter cost data were estimated using R.S. Means
(2004) and retail prices for the electronic air cleaner.
9.5.2 Installation Cost (C,)
Item Cost
Install electronic air cleaners (n=14)
Install new pressure gauges and controls
Total 1 unit
Total whole building (4 units)
$ 4,340
$ 600
$ 4,940
$19,760
Installation cost data were estimated using R.S. Means (2004).
9.5.3 Service Cost (Cs)
Item Cost
Engineering fees
Air balance
Total whole building (4 units) Cs
$ 9,000
$ 4,500
$13,500
The air handling units must be rebalanced and possibly
have their belts adjusted (to speed up the fan) because of the
additional static pressure added by the electronic air cleaners.
Engineering fees are for the investigation of the system,
planning, and design of air handler modifications. The
engineering fees estimation assumes a $100,000 total project
cost and a more complex design (because of the installation
of the air cleaners in the ductwork). The project cost may
actually be more, but typically engineering fees for projects
of this size would not be based on the cost of the equipment
(only a small portion of the square footage of the building
is being worked on). Both numbers are estimated using R.S.
Means (2004).
9.5.4 Retrofit Cost (Cr)
Item Cost
Re-work/install new duct transitions
New electrical service
Total 1 unit
Total whole building (4 units) Cr
$ 7,375
$ 1,488
$ 8,863
$35,452
The above estimations were based on data provided by R.S.
Means (2004). New electric service includes the cost of
wiring, junction boxes, and disconnect switches and assumes
that there is a spare breaker or breaker space in a panel
relatively close to the air handler.
-------
9.5.5 Operating Cost (C0)
9.6.1 Initial Purchase Cost (CD)
Item Cost
Power to units
Fan power
Total 1 unit
Total whole building (4 units) C0
$ 332
$1,715
$2,047
$8,188
Power to units is the energy required to operate the
air cleaner at 36 watts per filter. Data were taken from
manufacturer's data. All calculations were made in a manner
similar to Section 9.2.5.
9.5.6 Maintenance Cost (Cm)
Item Cost
Wash prefilters
Total 1 unit
Total whole building (4 units) Cm
$1,400
$1,400
$5,600
The air cleaners used in this model had washable (reusable)
filters built into the air cleaners. Data were estimated using
R.S. Means (2004).
9.5.7 Electrostatic Precipitation Cost Summary
Item Cost
Total 1st year
Total yearly operations and maintenance
(0+M)
Energy increase
Maintenance increase
$135,420
$13,788
3%
12%
The above data were calculated in the same manner as
described in Section 9.2.7.
9.6 Model Estimation for UVGI
This estimation is based on a commercial UVGI system
that is 36 x 48 x 72 inches (91 x 122 x 183 cm) in size
and has built-in prefilters. Because the size of the system
is different from the size of the air handler and the system
cannot fit into the air handler, the system would have to be
located outside of the air handler (on the supply side). This
arrangement requires reworking the ductwork and assumes
that there is enough space in the supply duct exiting the air
handler. The pressure drop across the UVGI is low enough
to be negligible; however, the prefilter would add 1 in. w.g.
(249 Pa) static pressure to the system. The 1 in. w.g. (249 Pa)
is safely below the excess of 2 in. w.g. (498 Pa) inherent in
the fan selection, so a new booster fan would not be needed
(the belt would be adjusted to 1100 RPM, which would give
6 in. w.g. [1.5 kPa] static pressure capacity).
Item Cost
UVGI filter
MERV 13 prefilter (n=16)
Delivery
New pressure gauges (2) and controls
Total 1 unit
Total whole building (4 units) Cp
$ 253,125
$ 1,927
$ 1,000
$ 820
$ 256,872
$1,027,488
The above costs were based on data provided during
confidential discussions with a commercial vendor.
9.6.2 Installation Cost (Ci)
Item Cost
Install UVGI filter/housing/prefilter
Install gauges/controls
Total 1 unit
Total whole building (4 units)
$ 7,500
$ 600
$ 8,100
$32,400
The above costs were based on data provided during
confidential discussions with a commercial vendor.
9.6.3 Service Cost (C,)
Item Cost
Engineering fees
Air balance
Total whole building (4 units) Cs
$4,800
$4,500
$9,300
The air handling units must be rebalanced and possibly
have their belts adjusted (to speed up the fan) because of the
additional static pressure added by the filters. Engineering
fees are for the investigation of the system, planning,
and design of air handler/ductwork modifications. The
engineering fees estimation assumes a $100,000 total project
cost. The project cost may actually be more, but typically
engineering fees for projects of this size would not be based
on the cost of the equipment (only a small portion of the
square footage of the building is being worked on). Both
numbers are estimated using R.S. Means (2004).
9.6.4 Retrofit Cost (Cr)
Item Cost
Re-work/install new duct transitions
New electrical service
Total 1 unit
Total whole building (4 units) Cr
$ 9,375
$ 3,600
$12,975
$51,900
The above costs were based on data provided during
confidential discussions with a vendor and assumes that there
is adequate power in the building as well as a spare breaker
or breaker space in a panel close to the existing air handler.
-------
9.6.5 Operating Cost (C0)
9.6.7 UVGI Filtration Cost Summary
Item Cost
UVGI filter electric power
Fan power
Total 1 unit
Total whole building (4 units) C0
$ 9,894
$ 3,519
$13,503
$53,652
UVGI filter electric power is the power required to operate
the UVGI system and is based on data provided during
confidential discussions with a vendor.
9.6.6 Maintenance Cost (Cm)
Item Cost
Clean UV bulbs
Replace bulbs
Replace prefilter
Total 1 unit
Total whole building (4 units) Cm
$ 375
$ 4,623
$ 1,927
$ 6,925
$27,700
The above costs represent regular maintenance as
recommended by a leading vendor. The vendor recommends
cleaning the bulbs every 9 months and replacing the bulbs
every 18 months (costs shown above are average per year
based on this schedule).
Item Cost
Total 1st year
Total yearly operations and maintenance
(0+M)
Energy increase
Maintenance increase
$1,202,440
$ 81,352
22%
58%
The above data were calculated in the same manner as
described in Section 9.2.7.
-------
10.0
Conclusions
An open literature survey of air cleaning technologies for
building HVAC applications was conducted. We reviewed
the literature and conducted a critical assessment, which
included an analysis of the cost and physical impacts of the
air cleaning device on building operation. Summaries of each
technology, including recent research, were presented.
The critical review focused on air cleaning technologies
relevant for building HVAC applications, primarily (but not
exclusively) those for particle removal. The five technologies
assessed in the report are as follows:
• Fibrous filters (mechanical filters)
• Electrostatically enhanced filters
• Electrets
• Electrostatic precipitators
• Ultraviolet germicidal irradiation
The typical HVAC filtration system in a building is a
relatively low (<90%) efficiency fibrous filter that is intended
to remove particles to keep the remainder of the HVAC
system clean and to remove nuisance dust for the occupants.
Technologies exist that can improve paniculate removal,
which will be needed if the HVAC filtration system is to
mitigate the hazard associated with an intentional biological
agent release. The analysis to determine which technology
to use for a particular building HVAC system will need to be
made on a case-by-case basis and will include factors such as
cost (initial and operating), level of protection desired, and
coverage (how much of the building to protect).
Based on the assessment discussed in this report, mature
technologies exist to enhance particle removal without
impacting the entire HVAC system—that is, without
requiring extensive retrofits or significant duct modifications.
Operating costs may increase because the technology is more
expensive to maintain or operate. The specific impacts of
these technologies on operation and cost are discussed and
summarized in each technology's critical assessment in the
body of this report.
Many data gaps need to be addressed. Two examples of gaps
in data are in-use operational data and the impact that particle
reduction in an HVAC system has on paniculate levels within
the building. Another gap or question is related to how air
infiltration or leakage impact building contaminant levels.
With respect to electrostatic precipitators for residential
and commercial applications, some data gaps exist in
the literature. Because ESP devices generally have lower
pressure drops than traditional mechanical filtration, pressure
drop is not generally reported. Many studies also do not
report ozone concentrations or generation rates, important
factors when considering an ESP technology because of
human exposure issues. Further difficulties with the available
literature were the diverse types of ESP devices considered;
many were not designed for in-duct use in residential and
commercial HVAC applications and even those that were
specific for this application varied in design. Authors
also disagreed about the effects of dust loading on the
performance degradation of ESPs. The work of Hanley et al.
(2002) should be continued to further quantify performance
degradation as a function of service life.
The UVGI literature lacked experimental data in HVAC
applications. Kowalski and Bahnfleth (2003) indicated
that data are being collected in a commercial setting for a
retrofit UVGI application. However, those data were not
located or are not yet available. Again, standardizing some
test parameters may help generate comparable data on other
factors that affect UVGI performance.
Electrostatically enhanced filter technologies appear to
be a relatively new and small player in the residential and
commercial HVAC market. Additional literature review
may locate more studies of this technology. Three studies
obtained in the current effort were helpful but not exhaustive.
There is a gap in data for filter pressure drops for these types
of devices. Further quantification of fiber diameters for the
filters used and comparison with mechanical filters that are
not electrostatically enhanced should be examined to quantify
the reduction in pressure drop at equal collection efficiencies
that can be achieved with electrostatically enhanced filtration.
Many data exist regarding the application and performance
of air cleaning technologies, but data specific to the impact
of air cleaning technologies on residential and commercial
environments are sparse. As more studies are conducted to
examine the impact of HVAC air cleaning technologies on
indoor air quality, it would be helpful to develop a sort of
classification or standard for this type of testing. While there
will always be site-specific differences, certain factors such
as duration of the test, baseline conditions, and flow rates
could be standardized for air cleaner evaluation. Studies that
incorporate a wide range of filters or air cleaner technologies
with different efficiencies would prove more valuable.
-------
-------
11.0
References
Agranovski, I.E., R.D. Braddock, and T. Myojo. "Comparative Study of the Performance of Nine Filters
Utilized in Filtration of Aerosols by Bubbling," Aerosol Science and Technology 35(4): 852-859 (2001).
Agranovski, I.E., R. Huang, O. Pyankov, I. Airman, and S.A. Grinshpun. "Enhancement of the Performance
of Low-Efficiency HVAC Filters Due to Continuous Unipolar Ion Emission," Aerosol Science and Technology
40(11): 963-968 (2006).
ANSI/ASHRAE (American National Standards Institute/American Society of Heating, Refrigerating and Air-
Conditioning Engineers). ANSI/ASHRAE Standard 52.1-1992, "Gravimetric and Dust-Spot Procedures for
Testing Air-Cleaning Devices Used in General Ventilation for Removing Paniculate Matter,"
Atlanta, GA, 1992.
ANSI/ASHRAE (American National Standards Institute/American Society of Heating, Refrigerating and
Air-Conditioning Engineers). ANSI/ASHRAE Standard 52.2-1999, "Method of Testing General Ventilation
Air-Cleaning Devices for Removal Efficiency by Particle Size," Atlanta, GA, 1999.
Arnold, B. and D. Myers. "Electret Media for HVAC Filtration Applications," Paper presented at the AFS
Indoor Air Quality-Filtration Conference Proceedings. Cincinnati, OH, November 14-15, 2002.
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers). ASHRAE Handbook
2003: HVAC Applications. Atlanta, Georgia: ASHRAE, Inc., 2003.
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers). ASHRAE Handbook
2004: HVAC Systems and Equipment. Atlanta, Georgia: ASHRAE, Inc., 2004.
Balazy, A., A. Podgorski, and L. Gradon. "Filtration of Nano-sized Aerosol Particles in Fibrous Filters. I -
Experimental Results," Journal of Aerosol Science 35: S967-S980 (2004).
Balazy, A., M. Toivola, T. Reponen, Z. Podgorski, A. Zimmer, and S.A. Grinshpun. "Manikin-Based
Performance Evaluation of N95 Filtering-Facepiece Respirators Challenged with Nanoparticles," Annals of
Occupational Hygiene 50(3): 259-269 (2005).
Barrett, L.W. and A.D. Rousseau. "Aerosol Loading Performance of Electret Filter Media," American
Industrial Hygiene Association Journal 59: 532-539 (1998).
Bell, A. A., Jr. HVAC Equations Data and Rules of Thumb. New York, New York: McGraw-Hill, 2000.
Brickner, P.W., R.L. Vincent, M. First, E. Nardell, M. Murray, and W. Kaufman. "The Application
of Ultraviolet Germicidal Irradiation to Control Transmission of Airborne Disease: Bioterrorism
Countermeasure," Public Health Reports 118: 99 (March-April 2003).
Brosseau, L.M., MJ. Ellenbecker, and J.S. Evans. "Collection of Silica and Asbestos Aerosols by Respirators
at Steady and Cyclic f\ovi" American Industrial Hygiene Association Journal 51(8): 420-426 (1990).
Brosseau, L.M., S.K. Chen, D. Vesley, and J.H. Vincent. "System Design and Test Method for Measuring
Respirator Filter Efficiency Using Mycobacterium Aerosols," Journal of Aerosol Science 25(8): 1567-1577
(1994).
Brown, R.C. "Capture of Dust Particles in Filters by Line-Dipole Charged Fibers," Journal of Aerosol Science
12: 349-356 (1981).
Brown, R.C. "Filtration in Industrial Hygiene," American Industrial Hygiene Association Journal 62(5):
633-643 (September-October 2001).
Brown, R.C., D. Wake, R. Gray, D.B. Blackford, and GJ. Bostock. "Effect of Industrial Aerosols on the
Performance of Electrically Charged Filter Material," Annals of Occupational Hygiene 32(3): 271-294 (1988).
-------
Burroughs, B.H. "The Role of Filtration Efficiency in the Cleanliness of HVAC Systems Phase III -An
Extended Field Study of the Cleanliness of HVAC Systems Based Upon the Efficiency Level of the Air
Filtration System," Paper presented at the Ninth World Filtration Congress, New Orleans, LA,
April 18-24, 2004.
Carlsson, T. "Indoor Air Filtration: Why Use Polymer Based Filter Media?" Filtration & Separation 38(2):
30-32 (March 2001).
Carpin, I, R. Doherty, and P. Gardner. "Fractional Penetrations for Select Commercial Electrostatic Filter
Media in the Sub-Micron Particle Size Range," Proceedings of the 1996 Edgewood Research, Development
and Engineering Center (ERDEC) Scientific Conference on Chemical and Biological Defense Research, Paper
CB-103985, AD-A334105, pp. 511-517, 1997.
Carrera, M., R.O. Zandomeni, J. Fitzgibbon and J.L. Sagripanti. "Difference Between the Spore Sizes of
Bacillus anthracis and Other Bacillus Species," Journal of Applied Microbiology (2006).
Centers for Disease Control and Prevention (CDC). "Filtration and Air Cleaning Principles," Guidance for
Protecting Building Environments from Airborne Chemical, Biological, or Radiological Attacks. http://www.
cdc.gov/niosh/docs/2003-136/2003-136c.html (May 2005).
Chambers, F.W., A. Al-Sarkhi, and S. Yao. "Velocity Distribution Effects in Air Filter Testing," Particulate
Science and Technology 19(1): 1-21 (2001).
Chen, C., S. Huang, C. Chang, and C. Chen. "Penetration of 4.5 nm to 10 um Aerosol Particles Through
Fibrous Filters," Poster presented at American Industrial Hygiene Conference and Exhibition, Chicago, IL
(2006).
Chen, S-K., D. Vesley, L.M. Brosseau, and J.H. Vincent. "Evaluation of Single-use Masks and Respirators for
Protection of Healthcare Workers Against Mycobacterial Aerosols," American Journal of Infection Control
22(2): 65-74 (1994).
Chimack, M. J. and D. Sellers. "Using Extended Surface Air Filters in Heating Ventilation and Air
Conditioning Systems: Reducing Utility Maintenance Costs While Benefiting the Environment," In the
American Council for Energy Efficient Economy's Proceedings of the American Council for an Energy
Efficient Economy Summer Study on Energy Efficiency in Buildings 3: 377-388, Washington, D.C., 2000.
Cooper, C. D. and Alley, F. C. "Air Pollution Control: A Design Approach". PWS Publishers, 1986.
Croxford, B., K.W Tham, A. Young, T. Oreszczyn, and D. Wyon. "A Study of Local Electrostatic Filtration
and Main Pre-Filtration on Airborne and Surface Dust Levels in Air-Conditioned Office Premises," Indoor Air
10(3): 170-177 (September 2000).
DOE (Department of Energy), Energy Information Administration. "Table 5.6. A. Average Retail Price of
Electricity to Ultimate Customers by End-Use Sector, by State, July 2004 and 2003 (Cents per kilowatt-
hour)," Electric Power Monthly Data For: July 2004, Report Released: October 19, 2004. http://www.eia.doe.
gov/cneaf/electricity/epm/table5_6_a.html (accessed April 2005).
DOE (Department of Energy), Federal Energy Management Program. "Building Life-Cycle Cost (BLCC)
Programs," Information Resources. http://wwwl.eere.energy.gov/femp/information/download_blcc.html
(accessed April 2007).
DOL (Department of Labor), Bureau of Labor Statistics. "2005 Consumer Price Index Detailed Report
Tables," Consumer Prices Indexes, http://www.bls.gov/cpi/cpid05av.pdf (accessed April 2007).
Drouin, B. "Triboelectric Blend Enhances Filtration," Filtration & Separation 37(8): 20-23 (November 2000).
Drouin, B. "Enhanced Air Filtration Performance with a New Triboelectric Blend," Paper presented at the
American Filtration and Separations Society (AFS) Indoor Air Quality-Filtration Conference. Cincinnati, OH,
November 14-15, 2002.
Emmerich, S.J., and S.J. Nabinger. "Measurement and Simulation of the IAQ Impact of Particle Air Cleaners
in a Single-Zone Building," InternationalJournal of Heating, Ventilating, Air-Conditioning and Refrigerating
Research 7(3): 223-244 (July 2001).
-------
Faulkner, D., W. Fisk, J. Palonen, and O. Seppanen. "Performance and Costs of Particle Air Filtration
Technologies," Indoor Air 12(4): 223-234 (December 2002).
First, M. W., E.A. Nardell, W. Chaisson, and R. Riley. "Guidelines for the Application of Upper-Room
Ultraviolet Germicidal Irradiation for Preventing Transmission of Airborne Contagion—Part I: Basic
Principles," In ASHRAE Transactions: Symposia, Paper CH-99-12-1, The 1999 Winter Meeting, Chicago
105(1): 869-876 (1999).
Fuchs, K., A. Fuchs, and S. Koeger. "NBC-Building Protection System and Method," U.S. Patent US
2004/0144104 Al (2004).
Fugler, D., D. Bowser, and W. Kwan. "The Effects of Improved Residential Furnace Filtration on Airborne
Particles," ASHRAE Transactions: Research 106(1): 317-326 (2000).
Grabarczyk, Z. "Effectiveness of Indoor Air Cleaning with Corona Ionizers," Journal of Electrostatics 51-52:
278-283 (May 2001).
Hanley, J.T., D.L. Frank, M.K. Owen, D.S. Ensor, andL.E. Sparks. "Improved Test Methods for Electronic
Air Cleaners," Paper presented at Indoor Air 2002, 9th International Conference on Indoor Air Quality and
Climate. Monterey, California, June 30 - July 5, 2002.
Hanley, J.T., D.S. Ensor, K.K. Foarde, and L.E. Sparks. "The Effect of Loading Dust Type on the Filtration
Efficiency of Electrostatically Charged Filters," Indoor Air 99 in Edinburgh, Scotland, August, 1999.
Hanley, J.T and M.K. Owen. "Develop a New Loading Dust and Dust Loading Procedures for the ASHRAE
Filter Test Standards 52.1 and 52.2, Final Report," ASHRAE Project No. 1190-RP, August 2003.
Hanley J.T., and K.K. Foarde. "Validation of Respirator Filter Efficacy," RTI International Project No. 08621,
2003.
Harrington, M. "Experimental Study of Nanoparticle Penetration Through Various Filter Media," 2005
National Nanotechnology Infrastructure Network (NNIN) Research Experience for Undergraduates (REU)
Research Accomplishments, 42-43 (2005).
Hawkins, B.E. and K.C. Hofacre. "Tasks 7 & 8: Experimental and Theoretical Analysis of Important
Parameters for Determining Impact of a Chemical/Biological Attack on a Building," Contract No. GS-10F-
0275K, Task No. 8, Battelle report submitted to U.S. Environmental Protection Agency for final review, 2007.
Heim, M., B.J. Mullins, M. Wild, J. Meyer, and G. Kasper. "Filtration Efficiency of Aerosol Particles Below
20 Nanometers," Aerosol Science and Technology 39(8): 782-789 (2005).
Hofacre, K.C., A. W. Richardson, A. Wang, and B.L. Sheridan. "First Generation Mask Filter Prototype Design
Analysis, Final Report," Contract No. SP0900-94-D-0002, Task No. 229, Submitted to U.S. Army Edgewood
Chemical Biological Center, May, 1999.
Hofacre, K.C., P.M. Schumacher, R.T Hecker, and T.L. Forney. "Filtration Efficiency Assessment of HEPA
Filters Against a Bioaerosol Challenge," Battelle Contract No. SP0900-94-D-0002, 1996.
Homeyer, S.T., J. Walenga, and C. Cavalca. "Indoor Air Purification System," U.S. Patent US 6,280,691 Bl
(2001).
Homonoff, E. "North American Filtration Markets: 2003-2008," Paper presented at Filtration 2004
International Conference and Exposition. Philadelphia, PA, December 7-9, 2004.
Howard-Reed, C., L.A. Wallace, and S.J. Emmerich. "Effect of Ventilation Systems and Air Filters on Decay
Rates of Particles Produced by Indoor Sources in an Occupied Townhouse," Atmospheric Environment 37(38):
5295-5306 (December 2003).
Huang, S.H. and C.C. Chen. "Filtration Characteristics of a Miniature Electrostatic Precipitator," Aerosol
Science and Technology 35(4): 792-804 (October 1, 2001).
Huang, S.H. and C.C. Chen. "Ultrafine Aerosol Penetration Through Electrostatic Precipitators,"
Environmental Science and Technology 36: 4625-4632 (2002).
-------
IEST (Institute of Environmental Sciences and Technology). "HEPA and ULPA Filters," Recommended
Practice (RP) CC001.3, 1993.
Jankowska, E., T. Reponen, K. Willeke, S.A. Grinshpun, and K. Choi. "Collection of Fungal Spores on Air
Filters and Spore Reentrainment from Filters into Air," Journal of Aerosol Science 31(8): 969-978, 2000.
Janssen, L.L., J.O. Bidwell, H.E. Mullins, and T.J. Nelson. "Efficiency of Degraded Electret Filters: Part I -
Laboratory Testing Against NaCl and DOP Before and After Exposure to Workplace Aerosols," Journal of the
International Society for Respiratory Protection 20: 71-80 (Fall/Winter 2003a).
Janssen, L.L., J.O. Bidwell, H.E. Mullins, and T. J. Nelson. "Efficiency of Degraded Electret Filters: Part II -
Field Testing Against Workplace Aerosols," Journal of the International Society for Respiratory Protection 20:
81-90 (Fall/Winter 2003b).
Japuntich, D., L.M. Franklin, D.Y. Pui, T.H. Kuehn, S.C. Kim, and A.S. Vmer. "A Comparison of Two Nano-
Sized Particle Air Filtration Tests in the Diameter Range of 10 to 400 Nanometers," Journal ofNanoparticle
Research, 10 November (2006).
Ji, J.H., G.N. Bae, S.H. Kang, and J. Hwang. "Effect of Particle Loading on the Collection Performance of an
Electret Cabin Air Filter for Submicron Aerosols," Journal of Aerosol Science 34: 1493-1504 (2003).
Kemp, PC., H.G. Neumeister-Kemp, G. Lysek, and F. Murray. "Survival and Growth of Micro-Organisms on
Air Filtration Media During Initial Loading," Atmospheric Environment, 35(28): 4739-4749 (2001).
Kim, C.S., L. Bao, K. Okuyama, M. Shimada, and H. Niinuma. "Filtration Efficiency of a Fibrous Filter for
Nanoparticles,' 'Journal ofNanoparticle Research 8: 215-221 (2006).
Kim, S., M. Harrington, and D. Pui. "Experimental Study of Nanoparticles Penetration Through Commercial
Filter Media," Journal ofNanoparticle Research 9(1): 117-125 (2007).
Ko, G., M.W. First, and H.A. Burge. "The Characterization of Upper Room Ultraviolet Germicidal
Irradiation in Inactivating Airborne Microorganisms," Environmental Health Perspectives 110(1): 95-101
(January 2002).
Kogan, V, C. Harto, D.J. Hesse, and K. Hofacre. "Task 4: Evaluation of In-Room Air Cleaners," Contract
No. GS-10F-0275K, Task No. 4, Battelle report submitted to U. S. Environmental Protection Agency for final
review, 2007.
Kowalski, W.J. Immune Building Systems Technology. New York: McGraw-Hill, 2003.
Kowalski, W.J. and W.P Bahnfleth. "Effective UVGI System Design Through Improved Modeling," In
ASHRAE Transactions: Symposia, Paper MN-00-11-1, The 2000 Annual Meeting, Minneapolis 106(2):
721-730 (2000a).
Kowalski, W.J. and W.P. Bahnneth. "UVGI Design Basics for Air and Surface Disinfection," HPAC
Engineering 72(1): 100-110 (January 2000b).
Kowalski, W.J. and W.P. Bahnfleth. "Airborne-Microbe Filtration in Indoor Environments," HPAC
Engineering 1'4(1): 57-69 (January 2002).
Kowalski, W.J. and W.P. Bahnfleth. "Immune-Building Technology and Bioterrorism Defense," HPAC
Engineering 75(1): 57-62 (January 2003).
Ktihn, K.P and I.F. Chaberny. "Ultraviolet Germicidal Irradiation of Office Ventilation Systems Reduces
Work-Related Mucosal and Respiratory Symptoms," Evidence-based Healthcare 8: 148-149 (2004).
Kujundzic, E., M. Fatimah, C.J. Howard, M. Hernandez, and S.L. Miller. "UV Air Cleaners and Upper-Room
Air Ultraviolet Germicidal Irradiation for Controlling Airborne Bacteria and Fungal Spores," Journal of
Occupational and Environmental Hygiene 3: 536-546 (2006).
Lai, K.M., H.A. Burge, and M.W. First. "Size and UV Germicidal Irradiation Susceptibility of Serratia
Marcescens When Aerosolized From Different Suspending Media 1," Applied and Environmental
Microbiology 70: 2021-2027 (2004).
-------
Lathrache, R., and H. Fissan. "Enhancement of Particle Deposition in Filters Due to Electrostatic Effects,"
Filtration and Separation 24: 418-422 (1987).
Lee, M., Y. Otani, N. Namiki, and H. Emi. "Prediction of Collection Efficiency of High-Performance Electret
Filters," Journal of Chemical Engineering of Japan 35(1): 57-62 (2002).
Lee, B.U., M. Yermakov, and S.A. Grinshpun. "Removal of Fine and Ultrafme Particles from Indoor Air
Environments by the Unipolar Ion Emission," Atmospheric Environment 38: 4815-4823 (2004a).
Lee, B.U., M. Yermakov, and S.A. Grinshpun. "Unipolar Ion Emission Enhances Respiratory Protection
Against Fine and Ultrafme Particles," Journal of Aerosol Science 35: 1359-1368 (2004b).
Lehtimaki, M. "Development of Test Methods for Electret Filters," Nordtest, NT Technical Report 320. 57 p.
NT project No. 1164-94, 1996.
Lehtimaki, M. and K. Heinonen. "Reliability of Electret Filters," Building and Environment 29(3): 353-355
(1994).
Lele, B.S. and A. J. Russell. "Enhancing Enzyme Stability Against TiO2-UV Induced Inactivation,"
Biomacromolecules 6: 475-482 (2005).
Letts, G.L., PC. Raynor, and R. L. Schumann. "Selecting Fiber Materials to Improve Mist Filters," Aerosol
Science 34: 1481-1492 (2003).
Lifshutz, N. "Performance Decay in Synthetic Electret Filter Media," Advances in Filtration and Separation
Technology 11: 307-311 (1997).
Liu, B.Y.H. and F J. Romay. "Assessment of Methodologies for the Test and Evaluation of Electret Filter
Media," Final Report for Edgewood Research, Development and Engineering Center (ERDEC) U.S. Army
Research Office (1997).
Mainelis, G., S.A. Grinshpun, K. Willeke, T Reponen, V. Ulevicius, andPJ. Hintz. "Collection of Airborne
Microorganisms by Electrostatic Precipitation," Aerosol Science and Technology 30(2): 127-144 (February
1999).
Mainelis, G., K. Willeke, P. Baron, T. Reponen, S.A. Grinshpun, R.L. Gorny, and S. Trakumas. "Electrical
Charges on Airborne Microorganisms," Journal of Aerosol Science 32: 1087-1110 (2001).
Mainelis, G., A. Adhikari, K. Willeke, S. Lee, T. Reponen, and S.A. Grinshpun. "Collection of Airborne
Microorganisms by a New Electrostatic Precipitator," Aerosol Science 33: 1417-1432 (2002a).
Mainelis, G., R.L. Gorny, T. Reponen, M. Trunov, S.A. Grinshpun, P. Baron, J. Yadav, and K. Willeke.
"Effect of Electrical Charges and Fields on Injury and Viability of Airborne Bacteria," Biotechnology and
Bioengineering 79: 229-241 (2002b).
Mainelis, G., K. Willeke, A. Adhikari, T. Reponen, V. Ulevicius, and P.J. Hintz. "Design and Collection
Efficiency of a New Electrostatic Precipitator for Biological Aerosol Collection," Aerosol Science and
Technology 36: 1073-1085 (2002c).
Manz, A., Freudenberg Nonwovens. Personal Communication, April to June 2005.
Mayya, Y.S., B.K. Sapra, A. Khan and F. Sunny. "Aerosol Removal by Unipolar lonization in Indoor
Environments," Aerosol Science 35: 923-941 (2004).
McCullough, N.V, L.M. Brosseau, and D. Vesley. "Collection of Three Bacterial Aerosols by Respirator and
Surgical Mask Filters Under Varying Conditions of Flow and Relative Humidity." Annals of Occupational
Hygiene 41(6): 677-690 (1997).
Memarzadeh, F., Z. Jiang, and W., Xu. "Analysis of Efficacy of UVGI Inactivation of Airborne Organisms
Using Eulerian and Lagrangian Approaches," March 24, 2005, http://www.vigilairsystems.com/what/
uvgilibrarypage.
-------
Menzies, D., J. Popa, J.A. Hanley, T. Rand, and D.K. Milton. "Effect of Ultraviolet Germicidal Lights
Installed in Office Ventilation Systems on Workers' Health and Wellbeing: Double-Blind Multiple Crossover
Trial," The Lancet 362(9398): 1785-1791 (November 29, 2003).
Miller, J.D. "Defensive Filtration," ASHRAE Journal 44(12): 18-23 (December 2002).
Miller, S.L. and J.M. Macher. "Evaluation of a Methodology for Quantifying the Effect of Room Air
Ultraviolet Germicidal Irradiation on Airborne Bacteria," Aerosol Science and Technology 33: 274-295
(2000).
Miller, S.L., M. Hernandez, K. Fennelly, J. Martyny, J. Macher, E. Kujundzic, P. Xu, P. Fabian, J. Peccia,
and C. Howard. "Efficacy of Ultraviolet Irradiation in Controlling the Spread of Tuberculosis," Final Report
submitted to Centers for Disease Control and Prevention, NIOSH under NIOSH Contract No. 200-97-2602.
pp. 1-80 (October 14, 2002).
Morawska, L., V. Agranovski, Z. Ristovski, and M. Jamriska. "Effect of Face Velocity and the Nature of
Aerosol on the Collection of Submicrometer Particles by Electrostatic Precipitator," Indoor Air 12(2): 129-
137 (June 2002).
Musser, A., W. Kowalski, and W. Bahnfleth. "Stack and Mechanical System Effects on Dispersion
of Biological Agents in a Tall Building," In Proceedings ofIMECE2002, American Society of
Mechanical Engineers (ASME) International Mechanical Engineering Congress and Exposition, Paper
IMECE2002-33862: 119-126. New Orleans, LA, November 17-22, 2002.
Myers, D.L. and B.D. Arnold. "Electret Media for HVAC Filtration Applications," Paper presented at INTC
2003, International Nonwovens Technical Conference, Conference Proceedings, pp. 602-630 (2003).
Nardell, E. A. "Use and Misuse of Germicidal UV Air Disinfection for TB in High-Prevalence Settings," The
InternationalJournal of Tuberculosis and Lung Disease 6(8): 647-648 (August 2002).
National Research Council Committee on Toxicology. "Emergency and Continuous Exposure Limits for
Selected Airborne Contaminants." Volume 1. Washington, DC: National Academy Press, 1984.
Nifuku, M., Y. Zhou, A. Kisiel, T. Kobayahsi, and H. Katoh. "Charging Characteristics for Electret Filter
Materials," Journal of Electrostatics 51/52: 200-205 (2001).
Oglesby, S., and G.B. Nichols. A Manual of Electrostatic Precipitator Technology. Part I. Final Report from
Southern Research Institute to National Air Pollution Control Administration, Division of Process Control
Engineering, APTD-0610 (August 25, 1970).
Otani, Y, H. Emi, and J. Mori. "Initial Collection Efficiency of Electret Filter and Its Durability for Solid and
Liquid Particles," KONA Powder and Particles 11: 207-214 (1993).
Owen, M.K., D.S. Ensor, and L.E. Sparks. "Airborne Particle Sizes and Sources Found in Indoor Air,"
Atmospheric Environment 26A(12): 2149-2162 (1992).
Owen, M.K., J.T. Hanley, D.S. Ensor, and L.E. Sparks. "Residential and Commercial Air Filter Efficiencies,"
Paper presented at the Annual Air and Waste Management Association Conference, San Diego, CA, 2003.
Park, H.S., H.J. Chang, and K.W Lee. "Analytic Solution for Collection Efficiency of Electrostatic
Precipitators for Polydisperse Aerosols," Environmental Engineering Science 21(4): 451-461 (2004).
Park, Y.O., SJ. Park, I.K. Shin, and H.U. Jung. "Development of a Cartridge Filter Combining a Dielectric
Filter with an Electrostatic Precipitator," Presentation 2002.
Peccia, J., H.M. Werth, S. Miller, and M. Hernandez. "Effects of Relative Humidity on the Ultraviolet Induced
Inactivation of Airborne Bacteria," Aerosol Science and Technology 35: 728-740 (2001).
Peters, T., M. Boundy, and D. Leith. "Influence of Upstream Flow Characteristics on Filter Efficiency,"
Filtration & Separation 38(10): 41-47 (December 2001).
Pich, J. Fundamentals of Aerosol Science, (D.T. Shaw, ed.). New York: Wiley Interscience, 1978.
-------
Pich, I, H. Emi, and C. Kanaoka. "Coulombic Deposition Mechanism in Electret Filters," Journal of Aerosol
Science 18: 29-35 (1987).
Pierce, M.E. and N. Lifshutz. "New Developments in Synthetic ASHRAE Filtration Media," Tappi Journal
80: 142-145 (1997).
Qian Y, K. Willeke, S.A. Grinshpun, and J. Donnelly. "Performance of N95 Respirators: Reaerosolization of
Bacteria and Solid Particles," American Industrial Hygiene Association Journal 58(12): 876-80 (1997a).
Qian, Y, K. Willeke, V Ulevicius, and S. Grinshpun. "Particle Reentrainment from Fibrous Filters," Aerosol
Science and Technology 27: 394-404 (1997b).
Qian Y, K. Willeke, S.A. Grinshpun, J. Donnelly, and C.C. Coffey. "Performance of N95 Respirators:
Filtration Efficiency for Airborne Microbial and Inert Particles," American Industrial Hygiene Association
Journal 59: 128-132 (1998).
Raynor, PC. and S.J. Chae. "Effects of Particle Loading on Electrostatically Charged Filters in an HVAC
System," Paper presented at the AFS Indoor Air Quality-Filtration Conference Proceedings. Cincinnati, OH,
November 14-15, 2002.
Raynor, PC. and S.J. Chae. "Dust Loading on Electrostatically Charged Filters in a Standard Test and a Real
HVAC System," Filtration & Separation 40(2): 35-39 (March 2003).
Richardson, G., D.J. Harwood, S.A. Eick, F. Dobbs, and K.G. Rosen. "Reduction of Fine Airborne Particulates
(PM3) in a Small City Centre Office, by Altering Electrostatic Forces," The Science of the Total Environment
269: 145-155 (2001).
Rivers, R.D. and D.J. Murphy, Jr. "Air Filter Performance Under Variable Air Volume Conditions," ASHRAE
Transactions 106: 131-144 (2000).
Romay, F.J. and B.Y.H. Liu. "Degradation of Electret Filters During DOP Aerosol Loading," InAdvances in
Filtration and Separation Technology, v!2, Advancing Filtration Solutions, St. Louis, Missouri: American
Filtration & Separations Society, pp. 193-200, May 4-7, 1998a.
Romay, F. J., B.Y.H. Liu, and S.J. Chae. "Experimental Study of Electrostatic Capture Mechanisms in
Commercial Electret Filters," Aerosol Science and Technology 28: 224-234 (1998b).
Romay, F. J., B.Y.H., Liu, and S.J. Chae. "Electrostatic Particle Capture in Commercial Electret Filters,"
Advances in Filtration and Separation Technology 11: 126-133 (1997).
Rose, H.E., and A.J. Wood. An Introduction of Electrostatic Precipitation in Theory and Practice. London,
Great Britain: Constable and Company Ltd., 1956.
Rousseau, A.D., M.E. Jones, and B.Z. Mei. "Electret Articles and Filters with Increased Oily Mist Resistance,"
European Patent Application EP 1402934 Al (1998).
Roth, C., E. Karg, K. Sommerer, and J. Wiegand. "Ultrafine Ambient Particles Can Be Used to Measure the
Efficiency of High Volume Filters," Journal of Aerosol Science, 30, S751-S752 (1999).
R.S. Means Company. R.S. Means Mechanical Cost Data, 25th Annual Edition. Kingston, Massachusetts: R.S.
Means Company, Inc., 2004.
Rudnick, S. N. "Optimizing the Design of Room Air Filters for the Removal of Submicrometer Particles,"
Aerosol Science and Technology 38(9): 861-869 (2004).
Snustad, D.P, M.J. Simmons, and J.B. Jenkins. Principles of Genetics. John Wiley and Sons, Inc., 1997.
Stafford, R.G., H. J. Ettinger, and T. J. Rowland, "Respirator Cartridge Filter Efficiency Under Cyclic- and
Steady-Flow Conditions" American Industrial Hygiene Association Journal May: 182-192 (1973).
StrionAir, Inc. "Products: How It Works," http://www.strionair.com/operation.asp. (April 2005).
Talbot, E. A., P. Jensen, H. J. Moffat, and C.D. Wells. "Occupational Risk from Ultraviolet Germicidal
Irradiation (UVGI) Lamps," International Journal of Tuberculosis and Lung Disease 6: 738-741 (2002).
-------
Tennal, K.B., M.K. Mazumder, and A. Siag. "Effect of Loading with an Oil Aerosol on the Collection
Efficiency of anElectret Filter," Particulate Science and Technology 9: 19-29 (1991).
Thorpe, A. and R.C. Brown. "Performance of Electrically Augmented Fibrous Filters, Measured with
Monodisperse Aerosols," Aerosol Science and Technology 37: 231-245 (2003).
Trane Company, The. "Quick Selection Guide, July 2004, M-Series and T-Series Climate Changer Air
Handlers, Fan Performance Data for Sizes 3 Through 30, December 2001," www.trane.com/commercial/
equipment/ProductDetails.aspx?prod=19.
Tsai, P.P., H. Schreuder-Gibson, and P. Gibson. "Different Electrostatic Methods for Making Electret Filters,"
Journal of'Electrostatics 54: 333-341 (2002).
US Environmental Protection Agency, National Homeland Security Research Center. Verification reports for
the evaluation of filtration media in HVAC systems. Available at http://www.epa.gov/nhsrc/tte_ventmediaair.
html
VanOsdell, D. and K. Foarde. "Defining the Effectiveness of UV Lamps Installed in Circulating Air
Ductwork," RTI International Final Report ARTI-21CR/610-40030-01 prepared for the Air Conditioning and
Refrigeration Technology Institute (November 2002).
Wake, D., A.C. Bowry, B. Crook, and R.C Brown. "Performance of Respirator Filters and Surgical Masks
Against Bacterial Aerosols," Journal of Aerosol Science 28(7): 1311-1329 (1997).
Wallace, L.A., SJ. Emmerich, and C. Howard-Reed. "Effect of Central Fans and In-duct Filters on Deposition
Rates of Ultrafine and Fine Particles in an Occupied Townhouse," Atmospheric Environment 38(3): 405-413
(2004).
Walsh, D.C. and J.I.T. Stenhouse. "The Effect of Particle Size, Charge, and Composition on the Loading
Characteristics of an Electrically Active Fibrous Filter Material," Journal of Aerosol Science 28(2): 307-321
(1997).
Walsh, D.C. and J.I.T. Stenhouse. "Parameters Affecting the Loading Behavior and Degradation of Electrically
Active Filter Materials," Aerosol Science and Technology 29: 419^32 (1998).
Wang, A. and K.C. Hofacre. "Task 6: Advanced Air Filtration System Development," Contract No. GS-10F-
0275K, Task No. 6, Battelle report submitted to U.S. Environmental Protection Agency for final review, 2007.
Wang, C.S. "Electrostatic Forces inFibrous Filters - A Review," Powder Technology 118: 166-170 (2001).
Wang, H. C. and G. Kasper. "Filtration Efficiency of Nanometer-size Aerosol Particles," Journal of Aerosol
Science 22(1): 31-41 (1991).
Wang, J., D. Chen, andD. Pui. "Modeling of Filtration Efficiency of Nanpoarticles in Standard Filter Media,"
Journal of Nanoparticle Research 9(1): 109-115 (2007).
Wang, S.K. Handbook of Air Conditioning and Refrigeration, Second Edition. New York, New York:
McGraw-Hill, 2001.
Ward, M., J.A. Siegel, and R.L. Corsi. "Stand-Alone Air Cleaners: Evaluation and Implications," Paper
presented at the Annual Air and Waste Management Association Conference. San Diego, CA, 2003.
White, H.J. Industrial Electrostatic Precipitation. Reading, Massachusetts: Addison-Wesley Publishing
Company, Inc., 1963.
Willeke, K., Y Qian, J. Donnelly, S. Grinshpun, and V Ulevicius. "Penetration of Airborne Microorganisms
through a Surgical Mask and a Dust/Mist Respirator," American Industrial Hygiene Association Journal 57:
348-355 (1996).
Xu, P., J. Peccia, P. Fabian, J.W Martyny, K.P Fennelly, M. Hernandez, and S.L. Miller. "Efficacy of
Ultraviolet Germicidal Irradiation of Upper-Room Air in Inactivating Airborne Bacterial Spores and
Mycobacteria in Full-Scale Studies," Atmospheric Environment 37: 405-419 (2003).
-------
Xu, P., J. Peccia, M. Hernandez, and S.L. Miller. "The Efficacy of Upper Room Ultraviolet Germicidal
Irradiation in Inactivating Airborne Microorganisms Under Incomplete Mixing Conditions," Paper presented
at the U.S. EPA's Engineering Solutions to Indoor Air Quality Problems Conference, Proceedings, Research
Triangle Park, NC, July 2000.
Yang, S. and G.W.M. Lee. "Filtration Characteristics of a Fibrous Filter Pretreated with Anionic Surfactants
for Monodisperse Solid Aerosols," Journal of Aerosol Science 36: 419-437 (2005).
Yao, M. and G. Mainelis. "Utilization of Natural Electrical Charges on Airborne Microorganisms for their
Collection by Electrostatic Means," Aerosol Science 37: 513-527 (2006).
Zukeran, A., PC. Lioy, A. Chakrabarti, A.A. Berezin, S. Jayaram, J.D. Cross, T Ito, and J. Chang. "Collection
Efficiency of Ultrafine Particles by an Electrostatic Precipitator Under DC and Pulse Operating Modes,"
Institute of Electrical and Electronics Engineers (IEEE) Transactions on Industry Applications 35(5): 1184-
1191 (September/October 1999).
-------
-------
Appendix A
Databases and Conferences Searched
for Relevant Information
Published technical literature
Source Reviewed
1 Defense Technical Information Center's (DTIC's) Database
1 Technical Report (TR) Database
1 The Research Summaries (RS) Database
1 Chemical and Biological Defense Information Analysis Center (CBIAC)
Bibliographic Database
1 Dialog Database (including but not limited to):
Journal of Aerosol Science
Atmospheric Environment
Aerosol Science & Technology
Building & Environment
ASHRAE Journal
Journal of the Air & Waste Management Association
Journal of the Air Pollution Control Association
(JAPCA)
Air & Waste
Indoor Air
Filtration and Separation
HPAC Heat. Piping. AirCond. Eng.
Journal of the Chemical Engineering of Japan
Journal of the Institute of Environmental Sciences and Technology
(IEST)
Powder Technology
Particulate Science and Technology
1 Internet resources available from the following agencies:
American Society of Heating, Refrigerating and Air-Conditioning
Engineers (ASHRAE)
American Society of Mechanical Engineers (ASME)
Occupational Safety and Health Administration (OSHA)
National Institute for Occupational Safety and Health (NIOSH)
Environmental Protection Agency (EPA)
Department of Health and Human Services (DHHS)
Centers for Disease Control and Prevention (CDC)
Edgewood Chemical Biological Center (ECBC)
Conference proceedings
' 2002 Indoor Air Quality-Filtration Conference, November 14-15, 2002,
AFS Society, Cincinnati, OH.
• World Filtration Congress 9 Proceedings CD-ROM, AFS Society, April
18-24, 2004, New Orleans, LA.
1 First NSF International Conference on Indoor Air Health Proceedings,
National Sanitation Foundation, May 3-5, 1999.
1 Second NSF International Conference on Indoor Air Health Proceedings,
National Sanitation Foundation, January 2001, Miami, FL.
Manufacturer's literature
1 Market Survey and Evaluation of Filters for Enhanced SIP Applications,"
U.S. Army Soldier Biological and Chemical Command (SBCCOM),
Aberdeen Proving Ground, MD, October 2001.
1 Market Survey and Evaluation of Filters for Large Area Shelter-in-Place
Applications," U.S. Army Research Development and Engineering
Command (RDECOM) Chemical Stockpile Emergency Preparedness
Program (CSEPP), Aberdeen Proving Ground, MD, May 2004.
1 U.S. Commercial and Industrial Air Filtration Markets, Frost & Sullivan,
March 29, 2001.
' World HVAC Equipment to 2006, Freedonia Group, May 1, 2002.
1 Growing Markets for Nonwoven Filter Media, Business Communications
Company, December 1, 2002.
-------
&EPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development
National Homeland Security Research Center
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |