Particle-Size-Dependent Efficiency of Air Cleaners
Research Triangle Inst., Research Triangle Park, NC
Prepared for:
Environmental Protection Agency, Research Triangle Park, NC
1991
-------
AEJiRL-P-815
TECHNICAL REPORT DATA
(Iteasf read famvctiaiu on thf tcven? be
1. REPORT NO,
EPA/60Q/A-93/179
1. TITLE AND SUBTITLE
Particle-size-dependent Efficiency of Air Cleaners
B. REPORT DATE
0, PERFORMING ORGANIZATION CODE
7. AUTHOHIS)
D. S. Ensor and J. T. Hanley (RTI), and L. E. Sparks
(EPA)
8, PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Center for Aerosol Technology
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, North Carolina 21109
10. PROGRAM ELEMENT NO.
11, CONTRACT/CHANT NO,
CR814169
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
13. TNr-t OF REPORT AND PERIOD COVERED
Published paper; 9/89-12/90
14. SPONSORING AGENCY CODE
EPA/600/13
is,SUPPLEMENTARYNOTIS AEERL prelect officer is Leslie E. Sparks. Mail Drop 54, 919/
541-2458. Presented at ASHRAE IAQ 91. Washington, DC, 9/2-5/91.
ie, ABSTRACT The paper gives results of tests with media filters, electrostatic filters,
and electronic air cleaners, It also discusses results from system qualification
tests to detect system artifacts. The collection efficiency of air cleaners as a func-
tion of particle diameter must be known to predict their ability to control specific
sources. Collection efficiency from 0^01 to 3 micrometers has been measured with
condensation particle counters and white-light and laser' particle counters. The
challenge aerosol is generated by nebulizing solutions to allow stability with respect
to size distribution and concentration.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution
Air Cleaners
Aerosols
Efficiency
Particle Size
Fibers
Filters
b, IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
Electronic Air Cleaners
Fibrous Filters
c. COSATI Field/Group
13 B
13A. 131
07 D
1.4G
HE
18. DISTRIBUTION STATEMENT
Release to Public
1», SECURITY CLASS (This Report I
Unclassified
•>i. NO. or PAGES
20. SECURITY CLASS jThispagt)
Unclassified
22,
EPA f'jrm 1220 I (t-73|
-------
EPA/600/A-93/17Q
Particle-Size-Dependent Efficiency
of Air Cleaners
D& Eusor, Ph.D.
J.T. Hanta?
L.E. Sparks, Pk.D,
ABSTRACT
The collection efficiency of air cleaners Of a function of
particle diameter must be Known to predict their ability to con-
trol specific sources. Collection efficiency from &OI tojpmku
been measured with condf root ion particle counters and white-
li$ht and laser panicle counters, The thallente aerosol is ten-
anted by nebulivnj solutions re allow stability with respect»
size distribution and concentration. The results of teas with
media filters, electrostatic filters, and electronic air cleaners are
reported. In addition, results from system qualification resit tc
detect system artifacts are discussed.
INTRODUCTION
The demand for energy-efficient buildings has changed
requirement! for initiation design. Air cleaning by removing
pollutants from recirtulated indoor air is a potential option for
using outskk air (ASHRAE 1989). Also, it may be necessary to
remove pollutants and allergens from incoming outdoor air to
protect sensitive individuals. Air cleaner (*rformatice depends
on the removal efficiency of the air cleaner, the volume of aii
treated, and the ability of ihe heating, ventilating, and air-
conditioning (HVAC) system to transport air to the cleaner,
These effects can be estimated using an indoor air quality simu-
lator (Owen et al. 1919).
The efficiency of an air cleaner depends in part on the
panicle diameter of the contaminant. Therefore, both the
particle-size distribution of the contaminant and the particle*
titt-dependeat collection efficiency of the air cleaner must be
known to compute the probability of its success in controlling a
problem.
TEST PROGRAM
The experiment*! protocol examines two test variables:
volumetric flow (face velocity) and particle diameter. The volu-
metric flow may vary widely in installations depending on venti-
lation system design. The range of particle diameter (0.01 to 3
pro) provides information on how well a niter performs at
reducing aerosols in the respirabk size range.
Air cleaner efficiency was measured for units up to 2 ft by 2
ft (0.6 m by 0.6 m) ai shown in Figure I. The test procedure is
described in detail by Hanky et *L (1990). The test section and
attached transitions are constructed of stainless steel, and the
ducting of the system is 8 in. (20 cm) diameter PVC pipe. The
aerosol concentration was measured upstream and downstream
of the test section to obtain the challenge and penetrating aero-
sol concentrations, respectively. The penetration aerosol sample
tap is located sufficiently far downstream of the test section to
allow for the complete mixing of any penetrating aerosol with
the entire airstream. Because the downstream ducting runs back
under the test section, the challenge and penetrating aerosol
taps are located near each other, thereby facilitating aerosol
sampttni and reducing sample line length.
The entire test system is located within a Gass 100 clean-
room (fewer than 100 particles greater than O.S jun/ft1 [O.S
pm/0.0283 m3)) to minimize possible measurement artifacts.
The cleanroom also served as the source of clean air used tc
establish near-zero concentration conditions on the penetration
samples before the cleaner was challenged with the text aerosol.
Because the system operates with p »uti« pressure (blower on
the upwind side of the air cleaner) and the system is in a clean-
room, the downwind access door eta be opened while the sys-
ttm is operating to allow tar downwind face of the filter to be
scanned with a sampling prone to detect areas of aerosol pene-
tration. The ability to scan the ur cleaner while it is mounted in
the test duct has proved to be a great advantage, because it
allows for investigation of aerosol penetration due to sneakage
vs. penetration due to low-efficiency collection.
Aerosol instrumentation used for the filtration efficiency
tests included a differential mobility panicle tizer (DMPS), a
laser aerosol spectrometer (LAS), arid an optical particle coun-
ter (OPC). The DMPS was operated at particle diameters of
0.01,0.03,0,1, ami 0.3 jun. The DMPS sampling rate was 0.071
cfm (2.0 L/min), of which 0.11 cfm (0.30 L/min) was directed to
the instrument's particle counter. The LAS operated over the
size range of 0.09 to 3 jun in 16 sizing channels with a flow rate
of 0.0021 cfm (0.060 L/min). The OPC covered the size range
from 0.3 to 10 urn in 16 sizing channels with a flow rate of 0.2S
cfm (7.1 L/min). The downstream face of the cleaner was
scanned with a second laser particle counter covering the size
range from OJ. to 5 jun.
The lest aerosol was solid potassium chloride (KG) gener-
ated from an aqueous solution. Laskin and Collisoo nozzles
D& £*MH is DirectM tod IT. Mtatey is Xaearch Environmental Sdeoiia, Center for AetmtA Teehadof y, Reuuca "Mangle Institute, Bacardi Tri-
angle Park, NC. 'LJL Sptfkf a Senkx Chemical Engineer, Ait ud Energy Ensineerifif Research laboratory, US. Emifonmeniai Protection Afeacy,
Rttea/cto Triui|k Park.
-------
Schematic diagram qfw dtan* tat MI /aeiUiy
were used to |en«rtt« the challenge aerosol. Upon generation,
the aerosol wu passed through a charge neutralizer to neutralue
any eleciro*atic charge on tlie aerosol Potassium chloride was
selected as the test aerosol because of its relitively high water
solubility, high deliquescent humidity, known cubic shape, and
low toxiciiy, By varying the saline concentration of the aerosol
solution and the operating air pressure of the generator, the
mean diameter of the challenge aerosol was controlled. Beer use
the tests spanned a Urge range of panicle silts (0,01 to 3 Mm),
the challenge aerosol was generated using a range of generator
operating conditions and saline concentrations.
PC* each test, nitration efficiency was calculated from
measurements of panicle concentration upstream and down-
stream of the air cleaner. The sampling sequence consisted of
obtaining three consecutive upstream samples, followed by six
consecutive downstream samples, followed by three additional
upstream samples. The ratio of the aveuge of the six down-
stream countt to the average of tne six upstream counts was then
used to compute the filtration efficiency. Each sample was of
1 minute's duration. Approximately 3 minutes was allowed
between switching the sample line bom upstream to down-
stream (and vice vena) to allow purging of the sample line
and particle counters. Tbe testa were performed with 50%,
100%, and 200% of rated nominal face velocities. Tbe
coefficient of variation (standard deviation divided by the
mean) for most of tne panicle diameters measured is less
than 5% for repetitive measurements.
RESUU1
The analysis of the panicle concentration dau wai directed
M determining the nitration efficiency at seven specific particle
diameters; 0,01,0,03,0.10,0.3$, 0.90,1.7J, and 2.50inn. Several
of these sizes wr*1 measured with more than one of the aerosol
instruments. For example, at 0.3 pin, all three instruments
(DMPS. LAS, and OPQ provided data. In these cases, the data
from the instrument yielding the greatest number of panich
counts per minute were used to compute the filtration effi-
ciency. In practice, this resulted in the DMPS being used
for the 0.01- and 0.03-ura particle sizes, the LAS lor the
0.1 um ate, sod the OPC for the remaining larger sizes.
to cover this range of particle sizes, each test con-
sisted of two separate runs. For the first run, the aerosol
generator condiuooa (i.e., air pressure and salt concentra-
tion in the aqueous nebulizer solution) were set to generic
a challenge aerosol having a mean diameter of approxi-
mately 0.02 un. With this aerosol, the filtration efficiency
measurements at 0.01 and 0.03 urn were performed. .*or
the second run, the aerosol generator was adjusted to pro-
vide an aerosol with a mean diameter of about 0,1 um and
the remaining particle sizes were evaluated. Using this
two-run approach, the challenge aerosol concentration
yielded a minimum of 500 particle counts per minule at
each of the seven panicle sizes and kept the maximum
count rates well below the instrument's upper counting
limits.
Sample results for five sir ckaners are presented. Pour of the
air cleaners wen pleated paper media types fe ivtag ASHRAE
atmospheric dust spot efficimc % of 4C%, 6S% 85%, and 95%.
The fifth air cleaner was an electronic ait cleaner consisting of a
coarse metal mesh prefllter and a corona discharge ionization
section followed by collection plates. The nitration -fficiencws
of the electronic air cleaner at the three test face velocities are
presented in Figure L Ov« the size range from 0.03io 3 0n, the
efficiency was approximately 15% for the lower two flow rates
(90 and ISO ft/min [0.46 and 0.91 mAJ). Over this same site
range, however, the efficiency at the high flow rate (360 ft/min
(1.12 m/sl) was significantly lower. At all three flow rates, the
filtration efficiency decreased sharply as particle size decreased
below 0.03 no. lie lower efficiency at these small particle sizes
is currently attributed to inadequate charting of snail particles
by the specific air deaner tested. A smsEst minimum efficiency
at 0.3->»rn diameter was also observed as a resuh of the transition
between the boundary of diffusional charging and field
charging.
135
-------
*0.
10
IB
•0
to
*0
• «0 «/m« (0 U m/i|
• XC Uifun (1 U «*)
Od* 01 * *o
PurbCl* r*m*»« (,ml
Flfitff i Filtration efficiency vi lotiriihm of particle diameter for
I He tttcironic air cleaner us various fan velocities
The filtration tfficif neks of « aeries of ventilation filters
are shown in Figure 3, The ASHR AE rating is for the atmo-
spheric dust spoi test (ASHRAE 1^6), the characteristic mini-
mum efficiency at about O.I ion is at the boundary of diffusion*]
collection for very small panicles and of interception and
impaciion for large particles. Aerosols of interest, such as
tobacco smoke, have the maximum number of panicles in the
same size range (O.I |im).
SUMMARY
Air cleaner efficiency depends on particle diameter and
flow rate for reduction of indoor pollutants. The measured
particle-siie-dependent collection efficiencies for both elec-
tronic and fibrous media Tillers show the effect of collection
mechanisms.
REFERENCES
ASHRAE. 1916. ASHRAE Standard $21976, MtlHod of f«*Mf
aircltenint dtocts v»d on ttiurtl witilttion for nmovlng
partinJoitmaittr. AilMiu. Anerkan Society of Haiiaf, Rifriger-
ating, and Aw-CoMlitio&iDg Engineers, Inc.
Ftttmuon ffftcmmiy vs. lotanthm of par licit diamtter for
f At jbur ASHRA E-reitd filttn at a nominal fact velocity
ASHRAE, \W.ASHMESia«leHt63l9».
blt Moor tir quility. Ailioi*: Amerkao Society of Heating.
Rtfriger»um, and AirCowUiioniag Engineen, Inc.
Hanky. J.T., FAD. Smith, PA Uwless. D.S. Ensor, and LJE. Sparks.
1990. "A hisdanefktai evaluation of to electronic air cleantr?'
Pncttdtnii of ilu Flfili Inltrnotwnai Con}trt*(i on l*dow Air
Quality and Climate, pp. 143-1 JO Toronto, Ctaada, July 29-Au|uit
3. Otiawi: Interoatioi. .1 Conkrtoct on Indoor Air Quality and
Climate, Inc.
Owen, NX., PA Lawless, T. \anuunoto, D,S Eawr, and I.E. Sparks.
19g9. "I AQPC An indoor air qu»i"y umulatorT TV Human tqva-
iion. Htalih aid comfort, pp 158-163 Mlana: American Society
of Huting, Refnieraiing, and Air-Conditio.vn| Engineen, Inc.
136
------- |