EPA 600/2-76-202
July 1976
Environmental Protection Technology Series
LABORATORY EVALUATION OF THE CLEANABLE
HIGH EFFICIENCY AIR FILTER (CHEAF)
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/2-76-202
July 1976
LABORATORY EVALUATION OF THE
CLEANABLE HIGH EFFICIENCY
AIR FILTER (CHEAF)
by
Manuel T. Rei and Douglas W. Cooper
GCA Corporation
Burlington Road
Bedford, MA 01730
Contract No. 68-02-1487
ROAP No. 21ADL-004
Program Element No. 1AB012
EPA Project Officer: D. L. Harmon
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
A novel scrubber, the CHEAT (Andersen 2000, Atlanta, Georgia) was tested
as part of a program to identify novel high efficiency fine particle
control devices. The tests were performed at room temperature using
iron oxide aerosols of concentrations near 0.2 g/m (0.1 grain/ft ),
mass median aerodynamic diameter of 1.1 pm. Inlet and outlet samples
were taken with cascade impactors, total mass filters, a condensation
nuclei counter, and an optical particle counter. These tests were per-
formed with different filter media, at different face velocities, and at
different water spray rates and water recycle rates. Efficiency increased
with increases in: foam pores; per inch, pressure drop, flow rate, spray
rate, make-up water addition. The results were consistent with the
hypothesis that impaction was the major collection mechanism and re-
entrainment a substantial contributor to penetration. Total mass effi-
ciency was approximately 95 percent at normal conditions, for which the
pressure drop across the CHEAF was 80 cm (31.5 inches) WC. The particle
aerodynamic cut diameter, for which the efficiency would be 50 percent at
these conditions, was determined from cascade impactor data to be below
0.5 um. This indicates that the 50 percent cut diameter for the CHEAP
is smaller than for a venturi scrubber operating at the same pressure drop.
iii
-------
CONTENTS
Page
iii
Abstract
vi
List of Figures
viii
List of Tables
ix
Acknowledgments
Sections
I Conclusions
3
II Recommendations
III Introduction
Q
IV Test Equipment and Procedure
13
V Results and Discussion
VI Theoretical Prediction of Efficiencies 57
VII References
Appendices
A Manufacturer's Description of CHEAP
B Data Reduction for Andersen Impactors
79
C Andersen Impactor Data
D Optical Particle Counter Measurements and Data 117
E Condensation Nuclei Counter Measurements and Data 120
-------
FIGURES
No.
1 Test System for CHEAP Evaluation 9
2 Sampling Equipment Arrangement 10
3 Penetration Versus Face Velocity and Pressure Drop for 20
the 65 ppi Foam
4 Inlet Particle Size Distribution, CHEAP Test Number 3222-3 22
5 Inlet and Outlet Concentration Distributions for Run 3222-3 23
6 Penetration Versus Particle Size for Nominal Conditions 26
(Average of 3222 Series of Runs), Corrected for Reentrainment
and Uncorrected
7 Inlet and Outlet Mass Concentration Distributions 27
8 Penetration Versus Particle Size for 65 ppi Foam at Three 29
Levels of Face Velocity
9 Penetration Versus Particle Size for 65 ppi Foam at Two 31
Water Spray Rates (Two Liquid-to-gas Ratios)
10 Particle Size Versus Penetration for Two Different Solids 33
Loadings in Recirculation Water
11 Penetration Versus Pressure Drop and Face Velocity for the 34
45 Pores per Inch Foam
12 Penetration Versus Particle Size for the 45 ppi Foam at 35
Three Levels of Face Velocity
13 Penetration Versus Particle Size for the 80 Pores per Inch 33
Foam
14 Penetration Versus Particle Size for 80 ppi and 65 ppi Foam 40
at Approximately the Same Pressure Drop
vi
-------
25
FIGURES (continued)
15 Penetration Versus Particle Size for the 65 ppi Foam at 42
Nominal Conditions
16 Penetration Versus Particle Size for the 65 ppi Foam at the 43
Lowest Face Velocity
17 Penetration Versus Particle Size for the 65 ppi Foam at 44
Nominal Conditions
18 Penetration Versus Particle Size for the 45 ppi Foam at 45
Nominal Conditions
19 Penetration Versus Particle Size for the 45 ppi Foam at 46
the Lowest Face Velocity
20 Penetration Versus Particle Size with Metal Foam 48
21 Predicted Aerodynamic Cut Diameter Versus Pressure Drop and 55
Power Consumption
22 Schematic of Jungle Gym Model of Porous Foam 58
23 Predicted Particle Size Versus Penetration for Three Foam 62
Porosities at a Low Face Velocity, 100 cm/s
24 Predicted Particle Size Versus Penetration for Three Foam 63
Porosities at a High Face Velocity, 1000 cm/s
CHEAP Unit, Cutaway Drawing 69
26 Temperature Correction Factor for Aerodynamic Size of 75
Particles Captured in Andersen Stack Sampler
27 Aerodynamic Diameter Versus Flow Rate Through Andersen 77
Stack Sampler
28 Diffusion Denuder Size Cutoffs As A Function of Flow Rate 121
vii
-------
TABLES
No. Page
1 Test Program for CHEAF Evaluation 5
2 Actual Value of Parameters in Test Matrix 6
3 Overall Penetration of the CHEAF at Nominal Conditions 15
4 Overall Penetration of CHEAF at Water Higher Consumption 16
Rate of 5.75 Liters per Minute
5 Overall Penetration of CHEAF at Three Levels of Solids 18
in Recirculation Water
6 Overall Penetration for the 65 ppi Foam at Three Levels 19
of Face Velocity
7 Sample Data Sheet for Impactor Tests of CHEAF Collection 24
Efficiency
8 Overall Penetration of the CHEAF with the 45 Pores per 35
Inch Foam at Three Levels of Face Velocity
9 Grand Mean, Marginal Means, and Elements of the Analysis 50
of Variance for Efficiencies at Different Particle Sizes,
Gas Flows, Foam Pore Densities
10 Root Mean Square Residuals Estimates of ao and Replication 52
Estimate of a
11 Summary of Mean Percentage Efficiencies and Standard 53
Deviation of These Means for the CHEAF Under the Normal
Operating Conditions
12 Predicted Penetrations Under Assumptions Noted in Text 61
13 Optical Particle Counter Data 118
14 CNC Data 122
viii
-------
ACKNOWLEDGMENTS
The cooperation received from Jack Brady and John Golutnbeski of Andersen
2000, Inc. is gratefully acknowledged. Also, the efforts of the two
major contributors to the testing effort from the GCA staff, Steve Brenan
and Mark Daniels, are also acknowledged.
ix
-------
SECTION I
CONCLUSIONS
This evaluation was one of a series of such evaluations being conducted
by the Control Systems Laboratory of the Environmental Protection Agency
to identify and test novel devices which are capable of high efficiency
collection of fine particles. The overall efficiency of the CHEAP was
improved by increasing the spray rate, the make-up liquid rate, the foam
pores per inch, the pressure drop and/or the face velocity at the filter
drum. Face velocity was not as important as pressure drop, however, when
various filter media were compared. The mean efficiencies and estimated
standard deviations of these means are listed below (for tests near
normal operating conditions, as specified by the manufacturer):
Estimated
standard
Aerodynamic Mean deviation of
diameter, MJTI efficiency, % the mean, %
9 90.8 1.9
7 93.5 1.6
5 93.4 1.2
3 97.4 0.5
1 97.1 0.4
0.5 90.4 2.8
The penetration actually increased as particle size increased, which is
the opposite behavior expected from theory and found in normal practice.
This effect appears to be due to reentrainment and possibly leakage around
seals which may not perform as well in the small pilot scale unit as in
-------
a full scale Industrial unit. The major collection mechanism was impac-
tion and the major penetration mechanism was probably reentrainment, for
the aerosol studied. Total mass efficiency was approximately 95 percent
at normal conditions, for which the pressure drop across the CHEAP was
80 cm (31.5 inches) WC. The particle aerodynamic diameter, for which th«
efficiency would be 50 percent at these conditions, was determined from
cascade impactor data to be below 0.5 m. This indicates that the aero-
dynamic 50 percent cut diameter for the CHEAT is smaller than for a
venturi scrubber operating at the same pressure drop.
-------
SECTION II
RECOMMENDATIONS
For operations for which scrubbers are practical, the CHEAP system should
be considered as an alternative which may represent a savings in power
consumption at equivalent collection efficiency.
-------
SECTION III
INTRODUCTION
The CHEAP is a particulate control device which utilizes a wetted polymer
foam material to capture particulate matter. The water acts as an aid in
particulate collection at the foam, and as a continuous cleaning agent.
A more complete description of the CHEAP by the manufacturer (Andersen
2000, Inc.) is given in Appendix A.
The goals of the CHEAP evaluation were to determine its fractional and
overall efficiency, its power and water consumption, and the relative
importance of different collection mechanisms. A balanced factorial test
matrix was designed utilizing three levels of foam pore grade (pores per
inch) and three levels of face velocity. At the normal operating con-
ditions (labeled 3222 in Table 1 for the third level of face velocity, and
the second levels of pore grade, spray flow, and make-up flow, respectively),
four tests were made with impactors, two tests with total mass filters at
the duct center-line, two tests with total mass filters taken at traverse
points called for in EPA Method 5, one test with the optical particle
counter, and one test with the condensation nuclei counter. The three
extreme conditions of pore grade and face velocity (1122, 3122, 1322) were
each tested with one run using impactors, total mass filters, an optical
particle counter and a condensation nuclei counter. The two remaining
runs in the matrix which were run at the second level of face velocity
with the extremes of pore grade (2122, 2322) were tested with one run
each using impactors and total mass filters.
-------
Table 1. TEST PROGRAM FOR CHEAP EVALUATION
PrIncipal test matrix
Face
velocity
250 cm/ s
(500 fpm)
500-600 cm/a
(1000-1200
fpm)
1000 cm/ s
(2000 fpm)
Foam pores per inch
45 ppi
1122b
2122
3122&
65 ppi
1222b
2222
3222»>b
80 ppi
1322b
2322
Secondary test series
Spray/gas,
gpm/1000 cfm
1.5
1.1
Test number
3212d
3222C
Make-up/gas,
gpm/1000 cfm
0.03
0.10
0.30
Equivalent
volume fraction,
%
1
0.3
0.1
Test
number
3221d
3222C
3223d
"Four total mass tests: two EPA Method 5; two in-stack filter.
bTests with condensation nuclei counter, diffusion battery, and optical
particle counter, in addition to impactbrs and filters.
°Tests already done for principal matrix.
vto tests each*-
-------
Secondary conditions were chosen to include two water spray rates and
three solids loadings in the recirculation water. The use of dif-
ferent solids loadings simulated operating at different bleed and make-up
flow rates. In this manner we can determine actual water consumption,
exclusive of recirculation rates. Two impactor and total mass tests were
made at each of the three water use rates (3212, 3221, 3223) which were
different than nominal conditions. Table 1 displays the test matrix and
conditions we intended to utilize for the tests which were performed.
Table 2 contains the actual values used for the tests.
Table 2. ACTUAL VALUE OF PARAMETERS IN TEST MATRIX
Run no.
3222 series
3212 pair
3221 pair
3223 pair
2222
1222
3122
2122
1122
2322
1322
Foam type
pores per
inch
65
65
65
65
65
65
45
45
45
80
80
Water solids
content}
% by volume
0.3
0.3
1.0
0.1
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Face
velocity,
cm/s
797
759
680
733
530
243
840
548
262
331
361
Liquid-
to-gas
ratio,
10-3m3/m3
0.19
0.26
0.21
0.23
0.18
0.39
0.18
0.18
0.37
0.29
0.27
The face velocity was determined by dividing the exposed area of the
polymer filter into the total gas flow through the filter. The liquid-
to-gas ratio was determined by dividing the water spray rate through
the nozzles by the gas volume flow rate at the foam.
-------
Tests were also performed using a metal foam of approximately 60 pores
per inch to determine the overall and fractional penetration of the
CHEAP utilizing this foam at otherwise nominal conditions.
-------
SECTION IV
TEST EQUIPMENT AND PROCEDURE
EXPERIMENTAL SET-UP
The general schematic of the test system for the CHEAP evaluation is
shown in Figure 1. The iron oxide aerosol output from a screw-type dust
feeder was aspirated into the inlet of the test ducting where it mixed
33 5
with room air. (The Reynolds number at 0.5 m /s (10 cfm) was 2 x 10 ,
and earlier work with a similar set-up indicated good mixing under these
conditions.) The ducting diameter was 20 cm (8 inches). There were more
than 8 diameters of duct length before the sampling probes, and about 2
after the probe before a right-angle bend. The CHEAF unit has a cyclone
with a spray preceding the wetted foam part of the unit. After, the
wetted foam are two types of demisters and a blower (still part of the
CHEAF), followed by more ducting with a sampling configuration which was
nearly identical with the upstream configuration.
The sampling equipment was arranged as shown in Figure 2, with long
straight probes placed in the elbows of the inlet and outlet ducting. The
reason for using the straight sampling probe was to minimize probe losses
which normally occur when utilizing impactors with the standard button•
hook type nozzles. This avoided probe bends and yet allowed sampling at
least two diameters upstream from the duct bend. The sample was pulled
isokinetically through the nozzle, then through the probe, to the filter
holder or Andersen Mark III impactor, followed by a column of dessicant
-------
( T }THERMOMETER
PRESSURE( P )
GAUGE V-X
ROOM
AIR
-*-EXHAUST
OPTICAL
PARTICLE
COUNTER
CONDENSATION
NUCLEI
COUNTER
Figure 1. Test system for CHEAP evaluation
-------
47mm FILTER
OR 1MPACTOR
PROBE
NOZZLE
41cm
TEMPERATURE SENSOR
MANOMET!
Figure 2. Sampling equipment arrangement
(8">
GAS FLOW
DIRECTION
-------
(silica gel), a calibrated orifice, a dry test meter, and a pump. The
temperature and pressure of the gas were measured between the calibrated
orifice and the dry test meter.
The probe nozzles were placed two duct diameters from an upstream distur-
bance and greater than 8 duct diameters downstream from a disturbance•
Therefore the sampling was done at an excellent location in the duct.
All of tb,e impactor and total mass determinations, except for two runs
for which traversing was utilized, were made at the center of the duct.
A detailed description of our procedure for the use of the Andersen MK III
impactors is given in Appendix B. Also, the results of a detailed inter-
comparison of our impactors with each other was presented in a previous
GCA report entitled, "Dynactor Scrubber Evaluation", EPA-650/2-74-0830a,
2
June 1975. All impactor data is given in Appendix C.
It was determined that sufficient impactor sampling times were 15 minutes
at the inlet and 2 hours at the outlet to produce loadings per impactor
stage < 15 mg.
For impactors and filters both the upstream and downstream samples were
observed to be dry. Downstream drying was aided by the gas temperature
increase produced by the heat from the blower of the CHEAP.
We also made particle size measurements with a Bausch and Lomb Dust
Counter Model 40-1, and a Rich Model 100 Condensation Nuclei Counter.
Details concerning our use of these instruments are given in Appendix D
and Appendix E.
PROCEDURE
The estimated equilibrium concentration of solids was achieved by adding
sufficient iron oxide to the recirculating water in the CHEAP, and the
spray flow rates were adjusted. The blower was turned on and the unit
11
-------
allowed to reach an approximate equilibrium pressure drop. The dust
feeder was started and more time was allowed to elapse to allow the con-
centration at inlet and outlet to approach equilibrium. (The sampling
probes were covered to prevent them from catching dust when not sampling.
Sampling was begun downstream. Upstream sampling was done thrice fof
impactor runs, centered on the midpoints of three equally-spaced time
intervals. Total mass filter samples were obtained simultaneously for
20 minutes upstream and downstream. The filters and the impactor sub-
strates were predried and preweighted. After a 24-hour drying period,
they were weighed again to determine weight of material captured. For
the condensation nuclei counter and the optical particle counter, samples
were taken at the inlet and at the outlet at somewhat different times,
and these samples were often drawn through a dilution system before going
to the measurement devices.
The total gas flows through the ducts were determined at the inlet and
outlet sampling locations for each run, utilizing velocity traverses.
These measurements indicated approximately 20 percent leakage between the
CHEAF inlet ducting and the outlet ducting. The effect of the leakage
on results was factored into the analysis of data by multiplying the
impactor dM/d(log d) by an appropriate correction factor. Total mass
efficiency was determined utilizing a mass flux calculation which uses
the total flow and concentration at each sampling point; therefore, no
further correction was necessary.
12
-------
SECTION V
RESULTS AND DISCUSSION
INITIAL FINDINGS REGARDING OPERATION OF CHEAF
Particle collection occurs in the CHEAF at a reticulated polyurethane
foam filter which is placed over a rotating drum for support. The thick-
ness of the foam is approximately 2.5 cm (1 inch). The maximum drum
width is 10 cm (4 inches). Smaller drum widths can be obtained by cover-
ing a portion of the drum with strips of impermeable rubber. In this
manner we were able to adjust the area of the foam through which the
gas flowed, hence the face velocity at the filter.
Our initial plans to test the CHEAF at 3 different face velocities were
hampered for two reasons. First, we were informed that in order for the
demisters to operate efficiently it would be necessary to maintain the
nominal flow of approximately 0.47 m^/s (1000 cfm) through the unit. This,
coupled with the limited maximum drum width, made it impossible to attain
the low face velocities we initially intended to run. Second, the fan in
the, CHEAF unit was designed for one foam and could not pull 0.47 m3/s (1000
cfm) through the two denser foams at the highest intended face velocity
of 1000 cm/s (2000 f pm). In fact, the fan was unable to pull 0.47 m3/s
(1000 cfm) through the densest 80 pores per inch foam even at the inter-
mediate, face velocity of 500 cm/s (1000 f pm). Consequently, we were
- "' ' --V.. ,'"-•"'•.' ' • ' ;, ''.'- " ' .: " ' • •
forced to run at approximately 20 percent lower face velocities than
originally intended and were unable to run the 80 ppi foam at more than
one face velocity.
13
-------
We were also unable to run any of the foams at the lowest face velocities
without modification of the CHEAF unit itself. This was accomplished by
introducing filtered dilution air through a replacement plate which was
placed downstream of the drum and upstream of the mist eliminator. In
this manner we were able to pull the reduced volume of air through the
CHEAF filter drum and maintain the needed air flow through the demister,
thus insuring its proper operation.
Utilizing a combination of varying filter drum width and addition of
dilution air beyond the drum, we were able to achieve 3 levels of face
velocities, varying between 243 cm/s and 840 cm/s (478 and 1650 fpm).
TESTS WITH 65 PORES PER INCH FOAM
CHEAF Set-Up
The majority of the tests performed were with the 65 pores per inch foam
since it is the usual foam used in the CHEAF. Tests performed with this
foam included impactor and total mass filter runs up and downstream as
well as optical particle counter and condensation nuclei counter
measurements.
Initial flow testing with the 65 ppi foam determined that a 6,35 cm
(2.5 inches) wide strip of exposed foam at the CHEAF drum yielded the
highest face velocity possible with a corresponding total volume flow
3
at the fan inlet of approximately 0.47 m /sec (1000 acfm). The initial
pressure drop across the filter drum at these conditions was
61 cm WC (24 in. WC). Running the CHEAF at these conditions, with the
addition of 6 g/min of iron oxide at the inlet and 0.3 percent by volume
of iron oxide in the recirculation water, rapidly increased the pressure
drop across the filter drum to 66 cm WC (26 in. H_0). This increase in
14
-------
pressure drop across the filter drum caused a corresponding decrease in
the total volume flow of air through the CHEAP, bringing the nominal
conditions of face velocity down to approximately 800 cm/sec or 20 percent
lower than initially planned.
We were forced to settle for this compromise because the fan was somewhat
underpowered and was not capable of pulling more than approximately
0.38 m3/s (800 acfm) through the 65 ppi foam with a face velocity of
800 cm/s.
Overall Efficiency Runs
Overall Efficiency Runs at Nominal Conditions (3222 Series) - Two sets
of total mass and impactor runs were made at the nominal condition (3222)
before the pressure drop across the filter drum rose noticeably. Two
more pairs of runs were made with impactors and total mass filters. This
time the total mass filters were attached to button hook type nozzles
and collected by traversing in the duct as called for in EPA Method 5.
During these tests the pressure drop across the filter drum had risen to
70 cm WC (27.5 in. WC), and the flow through the CHEAP was consequently
diminished somewhat along with the face velocity. The overall penetration
(total mass) for these runs at nominal conditions are given in Table 3.
Table 3. OVERALL PENETRATION OF THE
CHEAP AT NOMINAL CONDITIONS
(3222 SERIES OF RUNS)
Run no.
3222-1
3222*2
3222-3
3222-4
[Penetration^
%
5.87
5.04
5.65
4.99
; Pressure drop,
cm (in.) WC
66 (26)
66 (26)
70 (27.5)
7.0 (27.5)
15
-------
Overall Efficiency Runs at Increased Water Spray Rate (3212 pair) - A
secondary set of tests were performed to evaluate the effect of water
usage of the CHEAF upon overall and fractional efficiency. These in-
cluded a higher water spray rate and both higher and lower water recycle
rates.
The nominal water spray rate was chosen as 4.2 liters per minute
(1.1 gpm) or approximately 11 Ipm/m3/sec (1.4 gpm/1000 cfm). This
is the water rate at which the 3222 series of tests were run. Two
runs were performed at a higher water spray rate of 5.75 liters per
minute (1.5 gpm) or approximately 15 Ipm/m3/sec (1.9 gpm/1000 cfm).
Two total mass filter and impactor efficiency runs were performed at this
higher water spray rate, and are numbered as the 3212 series of runs.
The penetration for the total mass filter runs are given in Table 4, and
it can be seen that they average somewhat lower than for the 3222 series
of runs; penetration decreased as liquid-to-gas ratio increased.
Table 4. OVERALL PENETRATION OF CHEAF AT
HIGHER WATER CONSUMPTION RATE OF
5.75 LITERS PER MINUTE (1.5
gpm) (3212 SERIES OF RUNS)
Run no.
3212-1
3212-2
Penetration,
%
5,24
4,89
Pressure drop,
cm (in.) WC
70 (27.5)
70 (27.5)
Overall Efficiency Runs at Simulated High and Low Water Recycle Rates
(3221 and 3223) - The second water-related parameter which was investi-
gated was the rate of recycling water through the CHEAF system. It is
common scrubber practice to maintain a tank or reservoir of water from
which water is drawn for the sprays and into which is charged the col-
lected spray water. This substantially reduces water consumption and
hence reduces the cost of water and its subsequent treatment. Therefore,
16
-------
it is important to determine how high a solids loading (corresponding to
a high water recycle rate) can be tolerated before the solids in the water
begin to substantially contribue to the penetration. If a constant
amount of particle-bearing water is entrained by the scrubber off-gases,
the amount of solids escaping into the off-gases with the water would be
proportional to the solids loading in the scrubber water.
In order to determine the effect of solids loading on the CHEAP's per-
formance, we ran a series of two paired runs each at a lower and higher
than nominal solids loading. These tests spanned a tenfold difference
in solids loading which corresponded directly to a tenfold difference in
water recycle rate. Since the CHEAP unit was not set up to run with a
continuous water addition and water removal flows, we ran tests on a batch
type basis, periodically removing solids-laden water and replacing it
with clean water to maintain a fairly steady known solids level in the
recirculation tank.
At nominal conditions (3222), we maintained a solids level in the tank
of 0.3 percent volume fraction, which was to be equivalent to a water
removal/replacement rate of 1 Ipm/m3/s (0.125 gpm/1000 cfm). We ran
two total mass filter and impactor tests for penetration at a higher
and lower solids in water volume fraction. The lower volume fraction
was equivalent to 3 Ipm/m3/s (0.375 gpm/1000 cfm), the 3223 series.
The higher volume fraction was equivalent to 0.3 Ipm/m3/s (0.0375 gpm/
1000 cfm), the 3221 series.
The pressure drop across the drum remained the same for the lower volume
fraction runs (3223) as it had been for the previous nominal condition
runs (3222), namely, 70 cm WC (27.5 in.); however, for the higher
volume fraction runs (3221), the pressure drop across the drum rose to
72 cm WC (28.5 in.), thus clouding somewhat the explanation of the re-
sults presented here in Table 5.
17
-------
Table 5. OVERALL PENETRATION OF CHEAT AT
THREE LEVELS OF SOLIDS IN RECIR-
CULATION WATER
Run no.
3222
Series
3221-1
3221-2
3223-1
3223-2
Pene-
tra-
tion,
'%
5.38
4.15
4.64
3.07
3.76
Pressure drop,
cm (in.) WC
68 (26.75)
72 (28.5)
72 (28.5)
70 (27.5)
70 (27.5)
Volume fraction
of solids
in water,
7.
0.3
1.0
1.0
0.1
0.1
Although the results are somewhat obscured by the variations in pressure
drop across the filter drum, it can be seen clearly from runs 3221-1 and 2
and 3223-1 and 2, that the penetration depends upon volume fraction of
solids in the recirculation water. Even with a higher pressure drop, the
higher solids runs (3221-1 and 2) had nearly a 29 percent higher average
overall penetration than the lower solids (3223-1 and 2) runs. This in-
dicates that entrained water droplets which escape the demisters are a
substantial source of penetration in the CHEAP.
Overall Efficiency Runs at Lower Face Velocities (2222 and 1222) - Two
other tests were run with the 65 ppi foam, and these were considered to
be primary tests belonging in the principal test matrix. These two
tests involved operating the CHEAP at lower-than-nominal face velocities
of 530 and 243 cm/s (1040 and 480 fpm). The first or higher velocity tes^
(2222) were accomplished by exposing 10 cm (4 in.) of the foam filter
material on the drum. With this new increased drum width, the fan in the
CHEAP was able to pull 0.387 m3/s (820 cfm) through the filter drum at a
pressure drop of 48 cm WC (19 in. WC), yielding a 530 cm/s face velocity.
A total mass and impactor efficiency test was run at these operating
conditions.
18
-------
The lowest face velocity tests (1222) were accomplished with much more
difficulty, requiring a modification of the CHEAP itself, as explained
in the Introduction. Efficiency tests using total mass filters and
impactors were run as well as measurements with an optical particle
counter and a condensation nuclei counter. The face velocity for this
test was 243 cm/s and the pressure drop across the filter drum was
18 cm WC (7 in.). Results of the total mass filter efficiency tests
are given in Table 6, and as would be expected, the penetration rises
with decreasing face velocity and/or pressure drop.
Table 6. OVERALL PENETRATION FOR THE 65 PPI
FOAM AT THREE LEVELS OF FACE
VELOCITY
Run no.
3222
Series
2222
1222-A
Penetration,
%
5.38
6.20
22.45
Pressure drop,
cm (in.) WC
68 (26.75)
48 (19)
18 (7)
Face velocity,
cm/ sec
797
530
243
As can be seen in Figure 3, however, the total penetration appears to
level off at the intermediate face velocity, indicating perhaps some
optimum face velocity exists where the CHEAF will achieve the maximum
particulate collection per unit of energy. These results are somewhat
clouded by the lowest face velocity run (1222-A) since we were unable
to decrease the water spray rate per unit time, hence the ratio of water
spray to volume of gas treated is elevated by over a factor of 2 for
this run. The effect of using a higher water spray rate in our earlier
tests (3212) indicated that a somewhat lower penetration might be ex-
pected. Since the penetration at this low face velocity is very high
compared to the higher face velocity runs, the effect of the excess water
spray is unlikely to change the overall appearance of curves in Figure 3.
19
-------
100
90
80
70
6O
50
40
30
20
2
2L
z"
o
QC 9
8
7
6
Ul
Ld
CL
10
T"
20
PRESSURE DROP.cmW.C.
30 40 50 60 70
80
T
T
T
T
T
90
—T~
100
o PRESSURE DROP
A FACE VELOCITY
0 100 200 300 400 500 600 700 800 900 1000
FACE VELOCITY, ctn/t
Figure 3. Penetration versus face velocity and pressure drop
for the 65 ppi foam
20
-------
Fractional Efficiency Tests With Andersen Impactors
Fractional Efficiency for Runs at Nominal Conditions - As we have stated,
Andersen impactors were run for each of the conditions previously mentioned.
The irapactor data were ultimately utilized to derive fractional efficiency
information presented here as curves of penetration versus particle size.
Appendix B gives the details.
Figure 4 shows one of the inlet particle size distribution results, as
obtained from the Andersen impactor, using the manufacturer's calibration.
The mass median aerodynamic diameter is about 1.1 um and the geometric
standard deviation is about 2.2, although the distribution is clearly
truncated in the upper particle size range, thus substantially different
from a log-normally distributed aerosol. Because about 85 percent of
the particle mass was less than 3 um aerodynamic diameter, the iron
oxide was a convenient test dust for fine particle efficiency measurements.
A typical data sheet for the tests is shown in Table 7, and the data sheets
for all other impactor runs are given in Appendix C. This sheet cor-
responds to the size distribution results shown in Figure 4. From the
cumulative distributions one can calculate the fraction of particles in
any size interval, and thus compare inlet and outlet concentrations as a
function of particle size to obtain penetration versus particle size.
Figure 5 gives the inlet and outlet concentrations for the same test
(again number 3222-3) at the usual operating conditions for the CHEAF
(nominal flow rate, foam pores-per-inch rating, water-spray-to-gas flow
ratio) in terms of the mass concentration per logarithmic interval, dm/d
(log dp). On this basis the penetration of any particle size is just the
ratio of the outlet value of dm/d (log dp) to the inlet value. The outlet
value of dm/d (log dp) has been corrected for air leakage into the CHEAF,
and it is the corrected value which has been plotted in Figure 5.
21
-------
100
M
at
u
o
tc
u
»-
Ul
10
z
§ 1-0
K
0.1
I I
2 5 10 20 40 60 60 90 96 96
PERCENTAGE OF MASS LESS THAN OR EQUAL TO STATED SIZE
Figure 4. Inlet particle size distribution, CHEAF test number 3222-3
22
-------
(J
o»
(T
UJ
UJ
N
o
2
X
H
o£
or
Id
O.
-------
Table 7. SAMPLE DATA SHEET FOR MPACTOR TESTS OF CHEAP
COLLECTION EFFICIENCY
ANDERSEN IMPACTOR
Date 11/12/75 Run # (3222-3)
Sample volume at STP (ft3) = 10.9630
Moisture (%) = 0.9492
Concentration (grains/ft3) = 0.1040
I.mpactor flow rate (acfm) = 0.5529
Location Inlet
Orifice E
Bar. press. ("Hg)
Avg. Pm (-"Hg)
Avg. Ttn (OF)
H20 (grams)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (OF)
Time (minutes)
Correction factor
29.93
29
1.5
. 76
2.2
11.56
-1
74
20
1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight,
grams
0.0032
0.0012
0.0010
0.0012
0.0014
0.0053
0.0203
0.0227
0.0137
0.0032
0.0732
7. on
stage
4.37
1.64
1.37
1.64
1.91
7.24
28.73
31.01
18.72
4.37
Size cutoff,
[00
12.91
8.07
5.41
3.75
2.34
1.22
0.74
0.51
% < stated
size
95.63
93.99
92.62
90.98
89.07
81.83
54.10
23.99
4.37
dm/d log D
0.0070
0.0098
0.0125
0.0366
0.1018
0.1503
0.1170
0.0027
Geo. mean,
Mm
10.21
6.61
4.51
2.96
1.69
0.95
0.61
0.07
These abbreviations are explained in Appendix B.
24
-------
If the predominant collection mechanisms for the CHEAP are interception
and inertial impaction, one would expect the penetration to decrease
with increasing particle size. If diffusion is important, one would
expect perhaps an improvement in collection efficiency (or at least a
change in slope of collection efficiency versus particle size) at the
smallest particle sizes ($ 0.3 urn). One more complication, at least,
will arise. Some fraction of particulate material which has been captured
in the scrubbing liquid may escape the demisters, either because the
droplets were too fine to be captured or because the liquid has been
atomized from some surface because of air shear. Figure 6 shows the
penetration versus particle size curve for a test at the nominal con-
ditions with and without correction for spray solids. For the uncorrected
values, up to about 3 ym, one has the expected increased efficiency (de-
creased penetration) versus increasing particle size. The penetration then
rose to a nearly constant level for particles larger than about 5 ym aero-
dynamic diameter. Because we thought such behavior might occur, we made a
test near the very beginning without a dust feed with the scrubber in the
nominal operating mode (including suspended material in the recirculating
water) to have an estimate of the extent to which reentrained droplets (or
those not captured at all) would affect penetration.
Curve (4) in Figure 7 shows the results of the test without inlet aerosol,
just described. Curves (1) and (2) in Figure 7 are the inlet and outlet
concentrations for the test with dust generated at the inlet, and curve (3)
is the outlet concentration for that test minus the outlet concentration
found when no aerosol was generated at the inlet. The penetration calcu-
lated from the corrected curve (3) and the inlet curve (1) is also shown
in Figure 6. This is strong evidence for the hypothesis that most penetra-
tion of the particles larger than 3 urn or so was due to reentraihment and/
or imperfect demisting. When dust is present at the inlet, the contribu-
tion from reentrainment and imperfect demisting would be somewhat greater.
This is because the concentration of the particulate material in the drops
25
-------
§
5
a
t-
u
z
III
a
AVERAGE
MEASURED PENETRATION
AVERAGE PENETRATION
CORRECTED FOR PARTICIPATE IN OUTLET
WITHOUT OUST FEED
3 5 7 9 II
PARTICLE AERODYNAMIC DIAMETER, jun
Figure 6. Penetration versus particle size for nominal conditions
(average of 3222 series of runs), corrected for re-
entrainment and uncorrected
26
-------
1.0
o
9
o
O.I
'5
-i
ui
M
in
o
5
I
_/
0.01
0.00
K
h-
w
u
0.000
0.01
AVERAGE INLET
> AVERAGE OUTLET
I AVERAGE OUTLET CORRECTED FOR PARTICULATE
IN OUTLET WITHOUT DUST FEED
OUTLET WITHOUT OUST FEED
0.1 I
PARTICLE AERODYNAMIC DIAMETER ,/»m
Figure 7. Inlet and outlet mass concentration distributions
27
-------
and on the collection surfaces will be somewhat higher than the average
equilibrium concentration in the recirculating water on account of the
particulate material just captured. This may well explain the change in
slope of percent penetration for particles larger than 0.5 ym in Figure <»
In reporting the CHEAF penetration we have not subtracted the material
produced by reentrainment and/or imperfect demisting, for this is an
integral part of the CHEAF performance.
Fractional Efficiency at Different Face Velocities - Figure 8 contains
penetration versus particle size curves for the average of the tests at
nominal conditions, the 3222 series, and for the two lower face velocity
runs, namely, 2222 and 1222-la. It can be seen that curves (3) and (2)
have a very similar shape with their main difference being that they are
at different levels of penetration. These two curves correspond to the
highest and intermediate face velocities, therefore, the overall pene-
tration being somewhat higher for curve (2) than curve (3) is as ex-
pected if impaction is the predominant collection mechanism. The simi-
larity of the curves in their shape indicates that the collection mecharti
are essentially unchanged for these two conditions.
Curve (1) is very dissimilar to both curves (2) and (3). First of all,
the data indicated more large particles at the outlet than at the inlet.
Unfortunately, the impactor runs from which curve (1) was derived con-
tain some zero weight gains for the upper stages of the inlet impactor,
which was most unusual for our tests. These zero weight gains may well
be the result of a somewhat different particle size distribution at the
CHEAF inlet sampling point, due to the very low flow for this test.
The very low flow or velocity in the inlet duct could have caused the
larger particles to settle out in the ducting before reaching the inlet
sampling point. If this was the case, then the appearance of large
particles at the outlet due to reentrainment of solids-laden spray water
28
-------
4>
O
UJ
z
UJ
a.
10
9
8
7
6
(3)
D I222-IA.243 cm/s
O 2222,530 Cm/s
A 3222 SERIES AVERAGE, 797cnyg
6
10
AERODYNAMIC PARTICLE DIAMETER,
Figure 8. Penetration versus particle size for 65 ppi foam at three
levels of face velocity
29
-------
could cause the greater than 100 percent penetration observed, which in
fact we plotted as 100 percent penetration for convenience.
Looking at the bulk of curve (1), below 5 microns, we see penetration
behaving oppositely to curves (2) and (3). The lowest penetration in
curves (2) and (3) at 2 microns, corresponds to the highest penetration of
greater than 100 percent for curve (1). Extrapolating from the changes in
going from (3) to (2), we expected somewhat higher penetrations in all
size classes for (1) but with a minimum in the 1 to 3 ym range— we observe
the former but not the latter. Either the CHEAT is operating very unlike
nominal conditons at this low face velocity or the change in conditions
seriously affected our ability to obtain accurate and meaningful data.
Later the results for 45 ppi foam at three flow rates will be presented;
these curves show no such anomaly.
Fractional Efficiency at Higher Water Spray Rate — Figure 9 contains frac^
tional efficiency curves for the CHEAP at nominal conditions (3222) and
with a higher water spray rate (3223). Curve (1) is for the nominal con-
-33 3
ditions, liquid-to-gas ratio 0.19 x 10 m /m and curve (2) is for the
—3 3 3
higher ratio (L/G) of 0.26 x 10 m /m . The two curves are very similar,
both in shape and magnitude. The two points of major difference are at
7 microns where the penetration is much higher at the higher water spray
rate, and at 1/2 micron where the penetration is much lower at the higher
flow rate. The higher penetration at 7 microns may be due to poor mist
eliminator performance, in which case it is a problem which could be cor-
rected by better design. The lower penetration for the 1/2 micron particle
size seems to represent an improvement in the CHEAP's performance on fine
particles. Only one impactor run of two contained useable data, therefore
we cannot place as much confidence in curve (2) of Figure 9 as in curve (1)
which is the average of four impactor runs.
30
-------
100
90
80
70
60
50
40
30
20
o>
a.
5 10
* 2
UJ 8
u
a
A 3222 SERIES AVERAGE.0.19 [lO"3m3/m2]
o 3212-1 SINGLE RUN,0.26|jO-3m3/m3]
23 4 56 7 8 9
AERODYNAMIC PARTICLE DIAMETERt/im
Figure 9. Penetration versus particle size for 65 ppi foam at two
water spray rates (two liquid-to-gas ratios)
31
-------
The total mass runs indicate only a very slight improvement in efficiency
for the higher water spray rate (3212 series) runs, indicating that the
penetration should be lower at least somewhere on curve (2), further
evidence for the better fine particle particle capture at higher L/G.
Fractional Efficiency for Runs at High and Low Simulated Water Recircula*-
tion Rates - Figure 10 contains the particle size versus penetration
curves for the two sets of runs, 3221-1 and 2, and 3223-1 and 2, where
the solids loading in the tanjc recirculation water was made both higher
and lower than nominal conditions. The two curves have a similar shape
for particles smaller than 3 ym. Above 3 microns, curve (1), which cor-
responds to a lower solids loading, begins to show a lesser degree of pens
tration than curve (2). These results from these two pairs of runs furthe
strengthen the argument that reentrained particle made up most of the
larger particles found at the outlet. The curves are as expected if there
is a given fraction of the water spray droplets by-passing the demisters
and being carried through to the outlet. Under these circumstances, we
would expect that the overall occurrence of larger particles at the outlet
would be increased if the spray water contains a higher solids loading,
and this is what was observed.
TESTS WITH THE 45 PORES PER INCH FOAM
Overall Efficiency at Three Face Velocities
The 45 pores per inch foam was tested at three levels of face velocity at
the nominal water spray rate and water solids loading. These tests in-
cluded one set of impactors and total mass filters per condition, and
optical particle counter and condensation nuclei counter measurements of
the highest and lowest face velocity. The three runs were numbered 3122,
2122, and 1122 from highest to lowest face velocity, respectively. The
results of the total mass tests were as shown in Table 8 and presented
graphically in Figure 11.
32
-------
to
9
8
7
q>
o
a>
z
O
.
tr
i-
UJ
z
Ul
a.
O 3223 -1 + 2, AVERAGE, 0.1% VOLUME
SOLIDS
A 3221 -1-1-2, AVERAGE, 1.0% VOLUME
SOLIDS
JL
J_
JL
_L
2345678
AERODYNAMIC PARTICLE DIAMETER ,
Figure 10. Particle size versus penetration for two different
solids loadings in recirculation water
33
-------
100
90
80
70
60
50
40
3O
10
-------
Table 8. OVERALL PENETRATION OF THE CHEAF WITH
THE 45 PORES PER INCH FOAM AT THREE
LEVELS OF FACE VELOCITY
Run no.
3122
2122
1122
Penetration,
%
9.61
17.45
41.52
Pressure drop,
cm (in.) WC
53 (21)
28 (11)
14 (5.5)
Face velocity,
cm/s
840
548
262
Comparing Figure 11 with Figure 3 (for 65 ppi foam) we note that both show
decreasing penetration as pressure drop (or face velocity) increases, as
one expects for collection dominated by impaction. For a comparable
pressure drop, however, the 65 ppi foam has a higher efficiency than the
45 ppi foam, an important advantage for the 65 ppi material.
Fractional Efficiency at Three Face Velocities
Figure 12 contains the fractional penetration curves for the three dif-
ferent face velocity runs with the 45 ppi foam. Curves (2) and (3) have
very similar shapes with a minimum penetration occurring at 2 microns.
The penetration to the right of the minima on curves (2) and (3) is believed
due to reentrainment. As the face velocity increases, smaller droplets are
reentrained, becoming smaller particles once dry. Curve (1) appears to
have a contribution from reentrained particles which is predominant even at
1 ym. As the face velocity increases, the capture of the droplets before
the demister section has apparently increased (the detnisters were operated
at constant flow).
Generally the curves occur at varying overall penetration levels are as
expected, that is, penetration was highest for the lowest face velocity,
which is consistent with the hypothesis that impaction was the dominant
particle collection mechanisms for particles & 1 ym aerodynamic diameter.
35
-------
100
90
8-0
70
60
50
40
30
- 20
0>
Q.
o:
H
UJ
10
9
8
7
6
5
A 3122 (840 cm/s)
o 2122 (548 cm/s)
O | 122 (262 cm/s)
JL
23456 78
AERODYNAMIC PARTICLE DIAMETER.
10
Figure 12. Penetration versus particle size for the 45 ppi
foam at three levels of face velocity
36
-------
Overall, the penetration is very high for each particle size, indicating
that the 45 ppi foam is not as efficient as the 65 ppi foam for the CHEAP
system.
TESTS WITH THE 80 PORES PER INCH FOAM
Overall Efficiency Run
We were unable to test the 80 ppi foam at more than one face velocity, as
previously explained, due to the high resistance of that foam compared with
the fan power in the CHEAP unit. Consequently, we ran a total mass ef-
ficiency test and an impactor or fractional efficiency test for the low
face velocity we were able to achieve, 331 cm/s.
The results of the total mass test indicated an overall penetration of
3.67 percent at a pressure drop of 72 cm (28.5 in.) WC, and a face
velocity of 331 cm/s. This low overall penetration was as expected for
the noticeably denser foam and the high pressure drop. The low face
velocity of 331 cm/s did not seem to have as important an effect upon
the penetration with the 80 ppi foam as the high pressure drop across
that foam. The 80 ppi foam had lower penetration than the 65 ppi foam for
comparable pressure drop, but the volume flow per filter area was also
lower.
Fractional Efficiency Run
The fractional efficiency curve for the impactor run is plotted in Fig-
ure 13. The curve displays the general shape of the same curves for the
65 and 45 ppi foams, indicating that the collection mechanisms are similar.
It is interesting to note that the lowest penetration occurs at 2 microns
for essentially all of the foams and face velocities tested, including
this most dense foam.
37
-------
10
9
8
7
u
O>
o.
O
Ul
z
u
a.
o
10
AERODYNAMIC PARTICLE DIAMETER,
Figure 13. Penetration versus particle size for the 80 pores
per inch foam
38
-------
Figure 14 contains the 2 fractional penetration curves for the 65 ppi
foam and the 80 ppi foam at nearly the same pressure drop. Since the
face velocity was not the same for any runs with either foam, the
similar pressure drop is the best remaining common parameter for com-
parison. It can be seen that the penetration is lower between 1 micron
and A microns for the 65 ppi foam while it tends to be higher for the
below 1 micron and above 5 micron particles. The overall penetration
and shape of the curves is somewhat similar while the penetration is lower
for the 80 ppi foam, perhaps due to the disproportionately higher mass of
the larger particles which penetrated the 65 ppi foam.
RESULTS OF, MEASUREMENTS WITH AN OPTICAL PARTICLE COUNTER AND A
CONDENSATION NUCLEI COUNTER
As previously indicated in the test plan, we made both optical particle
counter measurements and condensation nuclei counter measurements of some
runs. These measurements were made for four of the five runs indicated
in the test plan, since we were unable to make the CHEAP operate at con-
dition 1322 long enough to make the measurements.
Both types of measurement yielded data in terms of number of particles equal
to or greater than a given size, per volume. Therefore, the data had to
be reduced to number of particles between a given size interval, and this
number of particles was assigned to the geometric mean particle diameter
of that interval. In this manner we were able to compare the number of
particles assigned to a geometric mean particle diameter of the inlet and
outlet of the CHEAP from which we calculated the penetration. As with our
other calculations, we also corrected the outlet for the effects of air
leakage at the CHEAP outlet. The data for these runs are given in
Appendices D and E.
39
-------
10
9
8
7
6
£ 4
v
if
V
O.
zf
O
UJ
z
UJ
Q.
A 2322,80ppi FOAM, P=72 cmW-C.
o 3222,65pp| FOAM, P = 68cmW.C.
10
AERODYNAMIC PARTICLE DfAMETER,/ifn
Figure 14. Penetration versus particle size for 80 ppi and 65 ppi
foam at approximately the same pressure drop
40
-------
Tests With the 65 Pores Per Inch Foam
The data from the optical particle counter and the condensation nuclei
counter were combined to yield penetration versus particle size data from
1.4 microns down to 0.006 microns. The results for run 3222 are plotted
in Figure 15 along with the penetration versus particle size curve for the
impactor runs. The results are in good agreement with the impactor data,
where they overlap. The points which go beyond the lower limits of the
impactor data indicate a rapid rise in penetration with decreasing particle
size which is as expected for particle impaction.
Similar tests for the lowest face velocity run 1222, given in Figure 16,
indicate similar behavior for submicron particles. However, there is,
substantial disagreement between the impactor and optical counter data for
the two points near 1 micron. Recall that the 1222 impactor data seemed
anomalous (Figure 8).
Figures 15, 16, 18, and 19 were plotted with a linear scale for the par-
ticle size to facilitate their comparison with the preceding figures of
penetration versus particle size. However, since the very fine particle
size data is difficult to plot accurately on a linear scale, we have
plotted Figure 17, the nominal conditions run (3222), with a log scale for
geometric particle size. Here we can see more clearly that the penetra-
tion rises steeply below the 1 micron particle size, as would be expected
from the penetration theory given below.
Tests With the 45 Pores Per Inch Foam
Figures 18 and 19 are very similar in nature to those for the 65 ppi foam
just discussed. The submicron particles show the expected dramatic rise
in penetration while the overlapping impactor data are in good agreement
for the highest face velocity run and in poor agreement for the lowest face
velocity run. Again we can offer no good explanation for these anomolous
results near the 1 micron size.
-------
100
90
80
70
60
SO
40
30
20
u
w
a
O
<
(C
H
UJ
UJ
0.
IO
9
8
7
6
5
. ©
o OPTICAL DATA
B CMC DATA
* OPTICAL PLUS CNC
AIMPACTOR DATA
DATA
234567
PARTICLE DIAMETER,Mm
8 9
10
Figure 15. Penetration versus particle size for the
65 ppi foam at nominal conditions
42
-------
IOOQ
90
80
70
60
50
40
30
20
z
o
K
t-
hJ
Z
tu
CL
10
9
8
7
6
5
O OPTICAL DATA
0 CNC DATA
* OPTICAL PLUS CNC
AIMPACTOR DATA
DATA
4 5 67
PARTICLE DIAMETER,/*™
10
Figure 16. Penetration versus particle size for the 65 ppi
foam at the lowest face velocity (1222)
-------
100
111
tu
0.
-n
D
_
-
-
—
_
"
-
-
i i
TT| 1 — i — i i i ii 1 1 1 — i — i i i 1 1 f| 1 — i — n —
CD +x
O
O OPTICAL DATA
Q CNC DATA A ./
X OPTICAL PLUS CNC DATA \ /
vv »»
A IMPACTOR DATA \ /
•f THEORETICAL VALUES V /
O A
ill i iiiiiiil i iiiiiiil i iii
^^^m
-
-
^
-
-
_
O.OI
O.I 1.0
PARTICLE DIAMETER ,/im
5.O
IO
Figure 17. Penetration versus particle diameter for the 65 ppi
foam at nominal conditions
-------
01
a.
o
Ul
z
I OOP
9O
80 (D
70
60
50
40 |-
30
10
9
8
7
o OPTICAL DATA
Q CNC DATA
x OPTICAL PLUS CNC DATA
AlMPACTOR DATA
-A,
PARTICLE DIAMETER,
0
Figure 18. Penetration versus particle size for the
45 ppi foam at nominal conditions (2122)
-------
1009
90
80
70
60
50
40
30
V
O.
O
Ul
z
lu
a.
IO
9
8
7
o OPTICAL DATA
Q CNC DATA
x OPTICAL PLUS CNC
AIMPACTOR DATA
DATA
JL
JL
J_
234 567
PARTICLE DIAMETER,/
10
Figure 19. Penetration versus particle size for the 45 ppi
foam at the lowest face velocity (1122)
46
-------
TESTS WITH METAL FOAM
Overall Efficiency Run
The metal foam supplied by Andersen 2000, Inc. was made by plating the
45 ppi foam with nickel and burning out the polymer, leaving the. structure
with some 60 pores per inch, similar to the foam utilized for our nominal
conditions. In all, four runs were made with the metal foam: two impactor
runs, and two total mass runs. Each of the runs were at nominal conditions,
similar to the 3222 series of runs. The most notable differences between
the metal and urethane foam was the much smaller pressure drop across the
metal foam (38 cm or 15 in. WC) at the nominal face velocity.
The two total mass efficiency runs yielded penetrations of 13.7 percent
and 16.1 percent considerably higher than for the polyurethane foam of
approximately the same pore structure, namely, the 65 pores per inch foam.
The penetration versus pressure drop is more consistent with what we have
observed from Figure 8; we estimate the 65 ppi foam would have had about
10 percent penetration at 38 cm WC.
Fractional Efficiency Run
The fractional penetration is plotted in Figure 20 from the average penetra-
tions determined by the two impactor runs. The higher penetration may mask
the effect of spray carryover, as there is no sign here of penetration in-
creasing with particle size as it did in the more efficient case, the
polyurethane foam.
STATISTICAL ANALYSIS: FLOW RATE, FOAM, MEASUREMENT ERROR
The data from the efficiency tests were subjected to a two-way analysis of
variance, using a standard program, "BMD02V-Analysis of Variance for
Factorial Design" (Revised 12 September 1969, UCLA Health Services Com-
^
puting Facility). Such a two-way analysis works as follows:
47
-------
100
:
i*
•
&
8
u
w
1.0.
'\
0 I2346«789
PARTICLE AERODYNAMIC DIAMETER,Mm
Figure 20. Penetration versus particle size with metal foam
48
-------
1. The two factors (flow and foam) are arranged in a row and
column format with r rows and c columns (here, r = 3 levels
of flow and c = 2 foams).
2. It is assumed that each measurement value, xij, comes from
the mean value y, as altered by a row (flow) effect, a±, and
a column (foam) effect, g j , and experimental error, eij,
with eij assumed to be "independently and normally distributed
with mean zero and common variance oo^." That is:
a
3. The mean squares (sum of squares divided by degrees of freedom)
for the rows and columns are divided by the residual mean
square to form the F ratio.
A. This ratio is compared with the F value from tables; the F
chosen is Fp ((r - 1), (r - l)(c - 1)) for the row comparison
and Fp ((c - 1), (r - 1)(c - 1)) for the column comparison,
with p being the level of statistical significance (i.e.,
p - 90 percent means that ratios as large as Fgo% would happen
by chance only 10 percent of the time). From Table 5 of Crow,
et al.3 (r - 3, c - 2):
Fg() (2, 2) = 9.00
F9Q (1, 2) = 8.63
Fg5 (2, 2) = 19.00
(1, 2) - 18.51
Note that F (L, 2) means the F ratio for a variable with 1
degree of freedom (foam pore density) versus one with 2
(residual) .
Table 9 shows the results of such an analysis. The first column has the
aerosol size fraction; the second has the mean of all efficiency tests for
that size; the third through fifth give the mean efficiencies for that size
49
-------
Table 9. GRAND MEAN, MARGINAL MEANS, AND ELEMENTS OF THE ANALYSIS OF VARIANCE FOR
EFFICIENCIES AT DIFFERENT PARTICLE SIZES, GAS FLOWS, FOAM PORE DENSITIES
No.
1.
2.
3.
4.
5.
6.
7.
8.
Total
Total
9 urn
7 pm
5 nn
3 um
1 HB
0.5
V*
Grand
m^an
82.897
84.038
69.950
68.175
81.780
81.975
80.913
77.808
Marginal wans
Flow
250
68.015
77.183
29.730
25.000
62.335
59.135
59.165
51.895
500
88.175
90.893
88.445
86.435
88.500
92.010
88.895.
89.235
800
92.500
—
91.675
93.090
94.505
94.780
94.680
92.295
Foam pore density
45
77.140
70.515
80.083
76.183
78.223
86.783
84.947
70.467
65
88.650
85.675
59.817
60.167
85.337
77.167
76.880
85.150
80
_
95.925
—
—
—
—
_
—
Source of variation
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Flow
Foam density
Residual
Total
Degree of
freedom
2
1
2
5
1
2
2
5
2
1
2
5
2
1
2
5
2
1
2
5
2
1
2
5
2
1
2
5
2
1
2
5
Mean
squares
342
199
27.6
282
327
70.0
2,431
616
577
2,818
385
433
585
75.9
1.13
786
139
406
726
97.6
210
1,012
323
319
F ratio
12.34
7.21
4.02
4.67
4.22
1.07
6.51
0.89
518
67.2
1.94
0.34
3.47
0.466
3.17
1.01
Significance
> 0.90
_
_
—
_
> 0.995
> 0.975
_
—
—
—
—
Ul
o
-------
at different flows; the sixth through seventh (and in one case eighth)
columns give the mean efficiencies for different foams for that fraction.
finally, the ninth through thirteenth columns give the values and results
in connection with the statistical analysis.
From the analysis we conclude:
1. Although it is evident that increasing the flow increased
efficiency in every case, this was demonstrated statistically
significant (p >_ 90 percent) only for the total collection
efficiency and for the 5 ym particle size aerosol fraction.
2. Higher foam densities (pores per inch) sometimes produced
improved collection efficiency and sometimes did not, with
a statistically significant improvement demonstrated only for
the 5 urn size. (This lack of demonstrable improvement is
puzzling. Explanations might be that improved particle
capture is being offset by increased particle release in the
finer foam or that capture is nearly complete for both and
release is similar for both.)
The two-factor model assumes that the variation due to flow is independent
of that from foam density and vice-versa. All effects other than the row
(flow) and column (foam) are included in the residual eij (thus in ao).
Table 10 has listed the root mean square (RMS) residuals, which are esti-
mates of o0, sometimes (loosely) called the experimental error. The values
of o0 for 5 urn, 1 um, and 0.5 pm are quite like those for very similar
tests done by GCA in evaluating the Dynactor scrubber (Cooper and Anderson,
1975); for 3 Mm, the 20 percent is much larger than the (approximately)
1 percent for the Dynactor tests and the values for the total filter tests
are much larger than the 1.3 percent value for such tests in the Dynactor
study. These figures indicate there were other sources of variation
(beyond flow and foam) for the CHEAP evaluation, but that the sources of
variation might well not be "experimental error," as they were often much
larger than the "experimental error" pf the Dynactor.
51
-------
Table 10. ROOT MEAN SQUARE RESIDUALS ESTIMATES OF ao AND REPLICATION
ESTIMATE OF a
Aerosol aerodynamic
diameter fraction
Total lb
Total 2C
9 urn
7 ym
5 ym
3 ym
1 ym
0.5 ym
Root mean square
residuals, oo
5.3%
8.4%
24.0%
20.8
1.1
20.1
14.5
17.7
Measurement error estimate
from four replicate runs
Range (w)
0.9d
7.6
6.6
4.9
1.9
1.7
11.5
a - (0.49) w*
0.44
3.7
3.2
2.4
0.93
0.83
5.6
Best estimate of oo from range of four measurements.
Three flows times two foam densities, total filter values.
Three foam densities times two flows, total filter values.
Total filter results for the four replicate runs.
Table 10 also has an error estimate from the four replicate tests made at
nominal CHEAP operating conditions. The estimates of experimental error,
a, from the replicate runs are (except at 5 urn) much smaller than the RMS
residuals from the factorial tests. This is strong evidence for concluding
that the simple model
a
does not account for all the effects of flow rate and foam density. A
more detailed statistical test model would take into account the inter-
actions of flow and foam (as done for the Dynactor evaluation)— for ex-
ample the degree to which the difference between high and low flow
velocities is different for the various foams.
52
-------
In Table 11, we have listed for the CHEAF at normal operating conditions
the mean efficiencies and the standard deviation of the means (a- =
x
OQ//N" ), the latter derived from the last column in Table 10. These
results are in general agreement with those of Calvert, Rowan, and Lake,6
who found efficiencies in the range 91 to 96 percent for particle dry
aerodynamic diameters from 0.5 ym to 2.0 ym (nearly independent of par-
ticle size) for a CHEAF installation on a diatomaceous earth calcining and
drying process operating at - 50 cm WC (20 in. H20) and 0.11 x 1Q-3 m3
water per m gas.
Table 11. SUMMARY OF MEAN PERCENTAGE EFFICIENCIES AND STANDARD
DEVIATION OF THESE MEANS FOR THE CHEAF UNDER THE
NORMAL OPERATING CONDITIONS (TEST SERIES 3222)
Aerosol aerodynamic
diameter fraction
Totala
9 ym
7 ym
5 ym
3 ym
1 ym
0.5 ym
Mean
efficiency,
percent
94.6
90.8
93.5
93.4
97.4
97.1
90.4
Estimated standard
derivation of the mean,
percent
0.2
1.9
1.6
1.2
0.5
0.4
2.8
Mean
penetration,
percent
5.4
9.2
6.5
6.6
2.6
2.9
9.6
aTotal efficiency for iron oxide 'aerosol at normal conditions.
POWER CONSUMPTION
The power used by a scrubber is predominantly the pressure drop of the
air going through the scrubber (Ap) times the flow rate of the air (Qg) in
actual volume terms divided by the fan/motor power efficiency.
We measured the pressure drop across the fan in the CHEAF unit using a
rin* • - .. . . . .__.._ . .. ^°
Magnehelic. For the 65 pores per inch foam at nominal conditions the
53
-------
pressure drop across the entire CHEAT, including ducting, cyclone, drum,
and demisters, was 1.5 cm, 5 cm, 68 cm, and 5.5 cm, respectively, for a
total of 80 cm (31.5 inches) WC. The gas flow at the fan inlet was mea-
3
sured to be approximately 0.47 am /s (1000 acfm). Thus the power con-
2
sumption of the CHEAT is the pressure drop in N/m times the actual volume
2
flow in m /s, or:
Power = Q AP
&
= (0.47 m3/s) (7.85 x 103 N/m2)
= 3.7 kW
This gives the air power consumption in kilowatts. A good estimate of
the electrical power consumption is to assume motor plus fan efficiency
of 0.6. Using this assumption we get an overall electrical power con-
sumption of 13.1 kW/m3/s (8.3 HP/1000 acfm). The CHEAF fan motor was
rated at 10 horsepower; therefore, it is in reasonably good agreement with
our estimate.
The CHEAF also used pumps for water sprays and a very small motor to turn
the filter drum. The combined power rating of these three motors was
under 746 watts (1 horsepower).
Comparison of CHEAF with Other Scrubbers
One method of comparison between the performance of various particulate
control devices with regard to their power consumption is to compare their
energy consumptions for given aerodynamic cut diameters, as presented in
Figure 21. Aerodynamic cut diameter is the aerodynamic diameter for which
collection efficiency is 50 percent. Pressure drop is equivalent to air
power used per volume flow rate and it serves as a useful measure of power
consumption per flow capacity for different devices (we have not included
for the CHEAF the small pump power consumption).
54
-------
Ul
O.I
0.25
0.5
POWER, hp/tOOOcefm
0.8 LO 2.0 3.0
5.0
8.0 10
90 100
ZOO
300
PRESSURE DROP, cm HtO
Figure 21. Predicted aerodynamic cut diameter versus pressure drop and power consumption (adapted from
Scrubber Handbook). Lines la and Ib are for sieve plates; line 3 is for impingement plate;
line 4 is for a packed column; lines 2a and 2b are for venturi scrubbers with f = 0.25,
0.50 respectively. See text
-------
The last size interval we obtained using the irapactors was 0 to 0.5 urn.
Since the aerodynamic cut diameter for the CHEAF occurred at greater than
one-half micron for only one condition, we were only able to plot one
point. However, for the remainder of the runs we were able to indicate
a line which starts at 0.5 microns and goes straight down parallel to the
y-axis which defines where, at the corresponding pressure drop, the cut
diameter must fall. For example, the line for run 2122 should be read
as, the aerodynamic cut diameter for the CHEAF operating at 28 cm H»0
pressure drop is below 0.5 microns. Extrapolation of impactor data for
the 3222 runs indicates that the aerodynamic cut diameter is much less
than 0.5 micron, indicating that the point should be placed very low on
the line.
It should also be noted here that the use of the cyclone may have detracted
from the overall performance of the CHEAF per unit of energy expended.
Because we generated a very fine test aerosol, the low-pressure-drop
cyclone would not have been very efficient and the CHEAF might have
yielded essentially the same penetration for our fine test aerosol with
or without the cyclone. However, the cyclone contained a water spray
which may have conditioned the aerosol somewhat previous to seaching the
foam-filter drum, which would aid collection.
It is evident from the figure that the CHEAF may be substantially more
energy efficient than many conventional and novel scrubbers.
56
-------
SECTION VI
THEORETICAL PREDICTION OF EFFICIENCIES
The pore structure of the CHEAF collection material was modeled as an
array of equally-spaced cylinders intersecting each other and parallel
to the x, y, or z axis, a "jungle gym" configuration (see Figure 22).
The void volume fraction of the material is e = 0.97.7 To determine fiber
diameter, consider a volume LxLxL in dimension (see Figure 22). Each
face, front, side, top, has (L A)2 rods (fibers) perpendicular to it
(!L is number of pores per length) so the total number of fibers is
2 3
Nf « 3U L) • The volume is L and the fraction of that which is solid
is (1-e). The solid volume is just that of the fibers: (l-e)L3 =
3(«. L)2 (ir/4)df2L. The fiber diameter is thus df = r4(l-e)/3Tr]1/2/Jl.
Using Jl not in pores per inch but in pores per cm, one obtains fiber
diameters 64 um, 44 urn, 36 urn for 45, 65, 85 ppi foam.
A first approximation to the open face area in a layer is just L x L minus
the area of fibers parallel to the face. The fractional open area becomes
L x L
- ['2 (L £) (Ldf) - (L £)2 (df)2 ] / L x L
where the last term in brackets comes from discounting overlapping. The
open fraction is thus
•
1 - 2 (£ df) + (£ df)2
and we can obtain (X, d.) from the equation for df above:
57
-------
Figure 22. Schematic of jungle gym model of
porous foam
58
-------
* df - [4(1 - e)/3Ti]1/2
= 0.113 for e = 0.97
So the fractional open area of each layer is 0.787, and the fractional
fiber area is 0.214. (The values used in the calculation below were
0.78 and 0.22, respectively.)
1. Diffusion (from Stafford and Ettinger,8 after Stairmand)
nD » (8D/vo df)1/2
where nD «• single fiber efficiency due to diffusive collection
D - particle diffusivity, cm2/s
v « foam face velocity, cm/s
d^ « fiber diameter, cm
9
2. Interception
nc - 1+R-U+R)"1 A 2R
where nc " single fiber efficiency due to interception
R - ratio of particle diameter to fiber diameter, d /d
3. Impaction (curve fit to data of May and Clifford)
Hj - <|>2/OJ»2 + 0.64)
where ty =* impaction parameter
il) • Co., d v /18 u d- ;
P P o f
where C « Cunningham correction factor (l-fO.16 ym/d )
3 ^
p_ - particle density, g/cm
V » gas viscosity, poise.
59
10
-------
4. Combined efficiencies, assumed independent
n = l-(l-nI)(l-nc)(l-nD)
Efficiency of a single layer was calculated from
nL - 0.22 n,
the fractional area of fibers in a layer times the efficiency of a fiber.
Efficiency of the whole foam (1-inch thick) was calculated from
E = l-(l-nT)N
Jj
one minus the probability that a particle would escape capture N times.
where N = pores per inch times thickness in inches.
Table 12 gives the results of these calculations. Figures 23 and 24
show the penetrations as a function of particle diameter (assuming
p = 1 g/cm3). These results are derived from a model which assumes all
particles which strike the (wet) surface are caught (probably nearly true)
and none are reentrained (not true). The following trends are apparent
for the cases studied:
a. Penetration decreased as particle size increased, for
d > 0.3 urn.
P *
b. Penetration decreased as velocity increased, for d > 0.3 ym.
c. Penetration decreased as pore grade (pores per inch) increased,
for a given thickness of foam.
d. At 1000 cm/s, impaction was the dominant mechanism for particles
d > 1 ym. At 100 cm/s, impaction dominated for d > 3 urn.
P ~ v
e. Interception was nearly as important as or more important than
impaction for d = 0.3 ym, and diffusion was negligible. Inter-
ception was alsB significant for d » 1 ym, especially at the
lower flow velocity.
60
-------
Table 12. PREDICTED PENETRATIONS UNDER ASSUMPTIONS NOTED IN TEXT
PORE GRADE FACE PARTICLE
(PORES PER VELOCITY IIIAMLTtR
INCH) (CM/SEC) (MICRONS)
45.
100.
".I
f>.3
1.0
3.0
10.0
1000.
0.1
0.3
1.0
3.0
10.0
65.
100.
O.I
0.3
1.0
3.0
10.0
1000.
0.1
0.3
1.0
10.0
100.
0.1
0.3
1.0
3.0
10. 0
1000.
f'.l
0.3
1.0
3.0
10.0
SINGLE FIBER EFFICIENCIES
*«*************»*+«**»*»»»«»»*
INTERCEPTION DIFFUSION IMPACTHIN
0.0032 ,_ 0.009} 0.0001
O.0095 0.0040 ~ 0.0001
0.0315 0.0019 O.0049
0.0943 0.0011 0.2460
0.3141 0.0006 0.9740
0.0032 0.0030 (T.OOOl "
0.0095 0.0013 0.0069
0.0315 0.0006 0.3281
0.0943 0.0004 0.9703
0.3141 0.0002 0.9998
0.0046 0.0111 0.0001
0.0137 0.004B 0.0002
0.04?4 6.0023 ~"070101
0.1361 0.0013 0.4050
0.4536 0.0007 0.9874
0.0046 0.0035 0.0006
0.0137 ""6. 0015 0."6l43
0.0454 0.0008 0.5047
0.1361 0.0004 0.9856
0.4536, 0.0003 0.9999
0.0056 0.0123 0.0001
0.0168 0.0053 0.0003
0.0559 0.002S 0.0152
0.1675 " 0.0014 6.56Y6
0.5503 0.0008 0.9917
0.0056 0.0039 0.0008
0.0163 0.0017 0.0214
0.0559 O.OOOa O."6fl69
0.1675 0.0005 0.9904
0.5583 0.0003 1.0000
TIITAL
t
b Ft* 1C !*• NT f
0.1147
0.1239
0.3131
0.9613
1.0000
6.0602
0.1564
0.9724
1.0000
1.0000
0.1992
0.114o
O.9997
l.onon
1.0000
0.26U7
0.3221
0 . 7«: «: 3
i .bofio
1.0000
n.lt.42
0.5002
1 .0000
1 .0001)
1.0000
Pf NETRAT10N
0.8854
0 .8762
0.637(1
0.038ft
0.0001
0.9399
0.8417
0.0277
0.0001
0.0001
0.8009
0.76U9
6.4409
0.0007
0.0001
O.BB55
O.OOO4
0.0001
0.0001
0.7314
0.67PO
o.fiorti
0.0001
0.8359
0.4999
"6.8601
0.0001
0.0001
61
-------
100
o 45 ppi, lOOcm/s
•»- 65 ppi, 100cm/s
x SOppi, lOOcm/s
PARTICLE DIAMETER,
Figure 23. Predicted particle size versus penetration for
three foam porosities at a low face velocity,
100 cm/a
62
-------
100
50
20
- 10
V
;•
O
UJ
a
0.5
0.2
© 45 ppijlOOOcm/s
•f 65pp|,IOOOcm/s
x 80ppj,lOOOcm/s
O.I
JL
JL
4 6
PARTICLE DIAMETER,
8
10
Figure 24. Predicted particle size versus penetration for
three foam porosities at a high face velocity,
1000 cm/a
63
-------
f. Diffusion was about as important as or more important than
interception for particles ~ 0.1 ym and both were much
greater than impaction.
g. At 1000 cm/8 much less than 1 percent of all particles of
d > 3 ym would penetrate.
Recall that Figure 17 had the experimental values of penetration at
nominal conditions graphed against our estimates of particle diameter
(converting impactor aerodynamic diameter to geometric diameter); the two
crosses on that figure were obtained by doing the theoretical calculations
described above for the case of 65 ppi foam at 800 cm/s face velocity
O
for particles of density 5.2 g/cm (iron oxide). For particles 1 ym
in diameter or larger, the predicted penetrations were less than 10" ,
much lower than the observed penetrations (more evidence for the reen-
trainment explanation). At 0.3 ym the penetration was predicted to be
0.045, in nearly perfect (probably fortuitous) agreement with interpola-
tion of the impactor measurements; at 0.1 ym geometric diameter, the
predicted penetration was 0.78, again quite close to the measured value.
Except for the optical counter data at 0.4 ym, the information shown in
Figure 17 shows a consistent picture for particles less than about
1/2 ym: penetration increases from a few percent to nearly 1QQ percent
as particle size goes from tenths of microns to hundredths. Interception
is more important than diffusion even at 0.1 ym. If diffusion were
becoming an important collection mechanism, we should expect a decrease
in penetration as particles became much smaller than a few tenths microns,
unless perhaps there is another source of particulate matter in this
finest size range. The theory, which did not take into account reen-
trainment, agreed with the experimental results over the limited size
range of several hundredths of a micron to several tenths of a micron.
64
-------
SECTION VII
REFERENCES
1. Determination of Particulate Emissions from Stationary Sources
(Method 5). Fed Regist. Reference Methods (39 FR 20790)
June 14, 1974.
2. Cooper, D.W. and D.P. Anderson. Dynactor Scrubber Evaluation.
,GCA Corporation, GCA/Technology Division, Bedford, Massachusetts.
Prepared for U.S. Environmental Protection Agency, Office of
Research and Development, Washington, D.C. Contract No. 68-02-1316,
Task Order No. 6. June 1975. 107 p.
3. Crow, E.L., F.A. Davis, and M.W. Maxfield. Statistics Manual.
Dover, New York, 1960. 287 p.
4. Calvert, S. Engineering Design of Fine Particle Scrubbers. J Air
Pollut Control Assoc. 24(10):924. October 1974.
5. Wilson, E.B. Jr. An Introduction to Scientific Research. New York.
McGraw Hill, Inc., 1952.
6. Calvert, S. J. Rowan, and L. Lake. Final Report on the CHEAP
Cleanable High Efficiency Air Filter (Draft Final). U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C. EPA Contract
No. 68-02-1496. June 1975.
7. Technical specifications, Scott Paper Co., Chester, Pa.
8. Stafford, R.G. and H.J. Ettinger. Filter Efficiency as a Function of
of Particle Size and Velocity. Atmos Environ. 6:353-362, 1972.
9. Fuchs, N.A. Mechanics of Aerosols. Pergamon, New York, 1964.
408 p.
10. May, K.R. and R. Clifford. The Impaction of Aerosol Particles on
Cylinders, Spheres, Ribbons, and Discs. Ann Qccup Hyg. 10:83-95,
1967. —
11. Strauss, W. Industrial Gas Cleaning. Pergamon, New York, 1966.
65
-------
APPENDIX A
MANUFACTURER'S DESCRIPTION OF CHEAF
ANDERSEN 2000 INC.
CHEAF AIR POLLUTION CONTROL EQUIPMENT
The Andersen 2000 Inc. Cleanable Media High Efficiency Air Filter (CHEAF)
was introduced commercially to the air pollution control market in early
1975. The CHEAF system combines features of filtration equipment, wet
scrubbers, and Brownian motion or diffusion controlled mist eliminators.
The CHEAF unit is mechanically simple, inherently reliable and is physic-
ally quite small in comparison with other fine particle emission control
equipment. It can be built of any stainless steel alloy, mild steel, plas-
tic, fiber glass, reinforced plastic, or rubber-lined steel.
The CHEAF system requires an extremely low liquid-to-gas ratio, possibly
the lowest of any type of wet emission control equipment available today.
This, of course, reduces the size of pumps, tanks, connecting piping, and
other liquid handling equipment associated with the air pollution control
installation.
The CHEAF system is well suited to the collection of subtnicron water
soluble particulates and aerosols, including ammonium nitrate and urea
prill tower emissions, soda and borosilicate glass furnace emissions,
phosphoric acid mists and phosphorus pentoxide fumes, emissions from
inorganic chemical calciners and dryers, food product spray dryers,
66
-------
galvanizing fumes and sulfuric acid mists. It is also well suited to
control of insoluble particulate matter which has a mean diameter of
less than 1.0 micron. This makes it an excellent control device for
metallurgical fumes. The CHEAF system is not capable of handling large
quantities of particulate matter exceeding 5 microns in diameter, although
low energy pre-collectors can often be installed to remove these compo-
nents prior to entry into the CHEAF system. The CHEAF system is normally
considered for applications where high energy scrubbers or Brownian
motion controlled mist eliminators might also be considered.
The CHEAF air pollution control system utilizes a porous urethane or
metallic foam material as the filter medium. The foam is installed as a
1-inch thick mat over the perforated filtration drum. The foam is
secured to the drum using screw clamps. The foam material can be removed
and replaced in a very short time on the largest CHEAF unit.
The perforated drum, with the reticulated foam in place, is rotated in
the CHEAF housing and is continuously wetted by spray nozzles mounted
at locations around the drum housing. The sprays around the drum keep
the pores of the foam lined with a film of liquid so that as particulate
matter, aerosols, and fumes pass through the pores, they are contacted
with the scrubbing solution and are either dissolved or are agglomerated
and captured in liquid drops. Most of the gas stream contaminants pass
through the pores in the filter material with the liquid. Particulate
matter, which is larger than the pores, will deposit on the surface of the
foam. In small concentrations, this material is cleaned from the surface
by turbulent washing action from the sprays as the drum rotates past the
nozzles. However, if the concentration of this large particulate matter
(greater than 5 microns in diameter) exceeds about 0.10 grains/standard
dry cubic foot, pluggage to the filter media will occur. A pre-collector
should be used where this condition exists.
67
-------
The foam filter media, which is supplied for the CHEAP unit, typically has
60 pores per linear inch. The velocity through the filter material is
normally between 1,500 and 2,500 ft/min. This corresponds to a pressure
drop across the CHEAP unit of between 10 and 30 inches W.G. The total
liquid flow rate to the CHEAP unit for the sprays around the drum is
1.0 gallon per thousand cubic feet of saturated gas exiting the CHEAP
unit. This is between 7 and 10 times less than required for a Venturi
type scrubber.
In addition, the CHEAP unit can be operated with a highly viscous scrub-
bing liquid, since atomization of the water is not critical to operation
of the unit. The spray nozzles, which feed scrubbing liquid over the
CHEAP drum, are operated at low pressures, typically 20 psig.
Once the contaminated gas has passed through the foam filter medium and
has been contacted with the scrubbing liquid, the gas and the liquid drop-
lets exit through one end of the drum into the entrainment separation
section. A conventional, chevron-type mist eliminator is used to remove
the bulk of the liquid droplets. A mesh type mist eliminator is then
used to remove the extremely small liquid droplets and any residual mois-
ture which is not collected by the chevron mist eliminator. The gas is
then exhausted by an induced draft fan and discharged to the atmosphere.
Spent scrubbing liquid is discharged by gravity from the entrainment se-
paration section into a recirculation tank. This liquid can be recir-
culated back to the CHEAF filtration drum and used for scrubbing the
incoming gas. A small bleed stream is taken to product recovery or to
disposal. A cutaway drawing of the CHEAF system is shown in Figure 25.
The CHEAF system depends on three collection mechanisms for its operation:
1. Filtration
2. Impaction
3. Brownian motion (diffusion)
68
-------
Inlet
Filtration Drum Spray
Discharge
Mist Separator Chamber
Chevron Type Mist Separator
Mesh Type Mist Separator
Reticulated Polymeric
Foam Filter Media
Filtration Drum Spray
Perforated Filtration
Drum
Drum Rinse Tank
Figure 25 . CHEAF unit, cutaway drawing
-------
For large insoluble particulate matter and for undlssolved large soluble
particulate matter, the foam acts as a filter, physically impeding the
flow of the particles and separating them from the gas stream. Filtra-
tion, however, contributes only slightly to the CHEAF system's high effi-
ciency for fine particles. For particles of about 3 microns in diameter
down to about 0.3 microns in diameter, impaction is the dominant collec-
tion mechanism. Impaction results when a projectile (the dust particle
or aerosol droplet) runs into (impacts on) a collection object (the
wetted fibers in the foam) and loses enough of its energy to separate
from the bulk of the gas stream. Filtration and Brownian motion add
slightly to the collection efficiency resulting from impaction.
Direct comparisons with other wet scrubbers indicate the CHEAF unit is
capable of achieving higher collection efficiencies at lower energy input
levels than conventional scrubbers. This is probably due to the filtra-
tion and diffusion effects being additive with impaction collection.
70
-------
APPENDIX B
DATA REDUCTION FOR ANDERSEN IMPACTORS
Data reduction for the Andersen Impactors was done as follows. Once the
weight gain on each impactor stage had been determined, a Hewlett Packard
Model 9810 A calculator was used to calculate the following parameters:
• Total mass collected (grams)
o
• Total volume of gas sampled (ft at NTP)
• Moisture content of gas stream (percent)
• Particulate mass concentration (grains/dscf)
• Flow rate through impactor., (acfm)
• 50 percent effective cutoff diameter for each stage
for unit density spherical particles (nm)
« Mass percent on each stage (percent)
* Cumulative mass percent £ each stage cutoff diameter
(percent) .
e Geometric mean diameter for each stage interval (ntn)
• dM/d log D for each stage interval (grains/dscf).
The following data are entered into the calculator to obtain the above
parameters:
» Barometric pressure (inches of mercury)
o Molecular weight of gas stream at meter (Ib/lb-mole)
• Average static pressure at meter (inches of mercury)
71
-------
• Average temperature at meter ( F)
• Water collected in condenser and silica gel (grams)
3
• Gas volume at meter (ft )
• Average static pressure in duct (inches of water)
• Average duct temperature ( F)
• Sampling duration (minutes)
• Temperature correction factor for Andersen impactor
(see Figure 26)
• Particulate mass collected in probe and expander
section prior to top stage (grams)
• Particulate mass collected on each impactor stage
(grams)
• Particulate mass on back-up filter (grams).
The following calculations are performed by the calculator:
1. Total mass collected
where M = total mass collected (grams)
M = mass collected in probe and impactor prior
^ to top impactor stage (grams)
M. = mass collected on impactor stage (grams)
i = t = corresponds to the top impactor stage
i = b = corresponds to the bottom impactor stage
Mf = mass collected on back-up filter (grams).
72
-------
2. Volume tUO vapor @ NTP
Q - (0.0473) (W)
where V Q ». volume of water vapor @ NTP (ft3)
W = water collected (grams)
NTP - 29.92 in. Hg and 70°F.
3. Meter volume @ NTP
V -V I — 2>
ms m \ T + 460J
where V » meter volume @ NTP (ft3")
ms N '
o
V » meter volume (ft )
m N '
Pfe « barometric pressure ("Hg)
Pm * avera8e static pressure at meter ("Hg)
Tm * avera8e meter temperature (°F).
4. Total volume sampled @ NTPT y^ (ft3)
a. Outlet impactor only
VT ' VH20
b. Inlet impactor only
VT = Vms/F
where F « dry gas fraction.
..73
-------
5. Percent moisture,
H2°
6. Mass concentration, C (grains/dscf )
*""I~B~™T \±\ ^ — •_
) (15.43 grains/gram)
7. Volume sampled at duct conditions
tns
V = V
s VT
Ts + 46° \/29.92\
-Ps/13-596/\530/
where V = volume sampled at duct conditions (ft )
s
T = average duct temperature (°F)
S
P = average duct static pressure ("H-0).
S f.
8. Impactor flow rate
Vg/t
where Q = impactor flow rate (acfm)
t « time (minutes).
9. Impactor stage size cutoffs
Dg - (tc) (k)
where D = impactor stage size cutoff (um)
S
tc = temperature correction factor from Figure 26.
74
-------
Ln
t
1.60
1.50
1.40
o 1-30
8 1.20
o
o
1.10
1.00
0.00
III I I I I I
i i i r
i i i i
, I
I l i
IO £0 30 4O 50 70 100 ZOO 30O 800 700 WOO
STACK TEMPER ATURE,d*flr««s F
Figure 26. Temperature correction factor for aerodynamic size
of particles captured in Andersen stack sampler
2OOO
-------
k = constant for each particular stage curve
from Figure 27 for the Andersen Impactor.
y = exponent for each particular stage curve
from Figure 27 for the Andersen Impactor.
10. Percent mass collected on back-up filter, P (7,)
Pp = (M£/Hj) 100
11. Percent mass collected on each impactor stage, P (7.)
Pi =
12. Percent mass collected in probe and impactor prior
to top impactor stage, P (%)
P - (M /M_,) 100
p p l
13. Cumulative percent mass < stage cutoff diameter
for each impactor stage, C (%)
- - u — < - — i —
where L = b = corresponds to the bottom impactor stage
L = n = corresponds to the impactor stage of interest.
14. dM/d log D (graina/dacf)
dM/d - - ~
(log Di+1) - (log Dt)
where P£ = percent mass collected on a particular stage(s) (%).
76
-------
LL
FLOW RATE.cfm
•
o
£5?
a. M
fD O
i-» a.
a vi
2 ?
P §
CD H.
n CL
^ S"
en g
1 »
•3 n»
fe"1
M •<
(fi
(0
CO
l-h
H
P
rt
ffi
rt
H
§
?•
-------
15. Geometric mean diameter, D (urn)
lXDi
To get fractional efficiency, we plot the dM/d log D versus the geometric
mean diameter for each stage interval for the inlet and outlet impactors.
The curves from these plots yield the concentration for any given size
particle. The ratio of the outlet concentration divided by the inlet
concentration for any size particle is the penetration for that size.
Because the CHEAP unit exhibited significant leakage between our inlet
and outlet sampling points, we corrected the calculated dM/d log D for
the effect of leakage before plotting curves. The correction was made
by multiplying the outlet dM/d log D by the ratio of the total outlet
flow of gas divided by the total inlet flow of gas, in dry standard cubic
feet per minute.
78
-------
APPENDIX C
ANDERSEN IMPACTOR DATA
79
-------
ANDERSEN IMPACTOR
Date 11/5/75 Run # 2 (no dust feed) Location Outlet
Sample volume at STP (ft3) = 8.4388 Orifice D
Moisture (%) *= 1.6876 Bar. press. ("Hg)
Concentration (grains/ft3) = 0.0006 *Iw „ , „„ .
Avg. Pm (- Hg)
Impactor flow rate (acfm) = 0.6886 Avg. Tm (OF)
H20 (grams or %)
Meter volumes (ft3)
Avg. Ps (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
= 29.94
= 29.0
a 1.2
" 80.0
" 28.7
= 83.90
» 1.0
" 84.0
= 120
" I
Stage
Probe 6,
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0000
0.0003
0.0006
0.0004
0.005
0.0001
0 . 0004
0.0001
0.0006
0.0000
0.0030
% on
stage
0.00
10.00
20.00
13.33
16.66
3.33
13.33
3.33
20.00
0.00
Size cutoff
(um)
11.56
7.23
4.83
3.35
2.10
1.08
0.66
0.44
7, < stated
size
100.00
90.00
70.00
56.66
40.00
36.66
23.33
20.00
0.00
!
dm/d log D
0.0006
0 . 0004
0.0006
0 . 0001
0.0003
0 . 0001
0 . 0007
0.0000
Geo. Mean
(um)
9.14
5.91
4.03
2.65
1.51
0.84
0.54
0.06 ,
80
-------
ANDERSEN IMPACTOR
Date 11/11/75 Run * (3222-1)
Sample volume at STP (ft3) = 12.4140
Moisture (%) ; = 0.8382
Concentration (grains/ft3) = 0.0806
Impactor flow rate (acfm) = 0.6247
Location Inlet
Orifice E
Bar. press. ("Hg) = 30.00
Mw B 29
Avg. Pm (-"Hg) = It47
Avg. Tm (op) = 74
H20 (grains or 70) = 2.2
Meter volume (ft3) = 13.01
Avg. Ps (+H20) = i
Avg. Ts (OF) = 74
Time (minutes) = 20
Correction factor =
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0034
0.0011
0.0004
0.0010
0.0007
0.0035
0.0192
0.0220
0.0109
0.0021
% on
stage
5.28
1.71
0.62
1.55;
1.08
5.44
29.86
34.21
16.95
3,26
i
Size cutoff
(um)
12.14
7.59
5.08
3.52
2.20
1.14
0.69
0.47
% <. stated
size
94.71 .
93 . 00
92.37
90.82
89.73
84.29
54.43
20.21
3.26
dm/d log D
0. 0025
0.0072
0.0055
0.0214
0.0846
0.1276
0.0812
0.0016
Geo. Mean
(um)
9.60
6.21
4.23
. 2.78
.1.58
0.89
0.57
0.06
81
-------
ANDERSEN IMPACTOR
Date 11/11/75 Run # (3222-1)
Sample volume at STP (ft3) = 80.8596
Moisture (%) = 1.5794
Concentration (grains/ft3) = 0.0053
Impactor flow rate (acfm) = 0.6781
Location Outlet
Orifice
<"Hg>
Bar. press,
Mw
Avg. Pm (-"Hg)
Avg. Tm (op)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
30.00
28.7
2.42
79
27
87.82
1
74
120
1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0022
0.0000
0.0002
0.0001
0.0003
0.0002
0.0031
0.0049
0.0078
0.0086
0.0274
7. on
stage
8.02
0.00
0.72
0.36
1.09
0.72
11.31
17.88
28.47
31.38
Size cutoff
(|im)
11.65
7.28
4.87
3.38
2.11
1.09
0.66
0.45
% < stated
size
91.97
91.97
91.24
90.87
89.73
89.05
77.73
59.85
31.38
dm/d log D
0.0002
0.0001
0.0004
0.0002
0.0021
0.0044
0.0089
0.0010
Geo. Mean
(nm)
9.21
5.96
4.06
2.67
1.52 -
0.85
0.54
0.06
82
-------
ANDERSEN 3MPACTOR '
Date 11/12/75 Run # (3222-2)
Sample volume at STP (ft3) = 10.776
Moisture (%) . 0.895
Concentration (grains/ft3) = 0.10258
Impactor flow rate (acfm) » 0.54225
Location Inlet
Orifice E
Bar. press. ("Hg) = 30.00
Mw . 29.00
Avg. Pm (-"Hg) . 1.5
Avg. Tm (op) a 75
H20 (grams or 7.) a 2.04
Meter volume (ft3) = 11.32
Avg. Ps (+H20) = 1
Avg. Ts (OF) = 74
Time (minutes) m 20
Correction factor » 1
Stage
Probe &
expander
0 I
1
2
3
4
5
6
7
F
Total
Net weight
(gra)
0.0021
0.0008
0.0008
0.0016
0.0013
0.0041
0.0215
0.0231
0.0130
0.0027
0.0710
7. on
stage
2.95
1.12
1.12
2.25
1.83
5.77
30.28
32.53
18.30
3.80
Size cutoff
(|im)
13.03
8.14
5.46
3.79
2.36
1.22
0.75
0.51
% <. stated
size
97.04
95.92
94.79
92.54
90.70
84.93
54.65
22.11
3.80
dm/d log D
0.00566
0.01333
0.01183
0.02879
0.109682
0.15566
0.11315
0.00228
Geo. Mean
(um)
10.30
6.67
4.55
2.99
1.70
0.96
0.62
0.07
• M .
83
-------
ANDERSEN IMPACTOR
Date 11/12/75 Run # (3222-2)
Sample volume at STP (ft3) = 78.7093
Moisture (7.) = 1.4423
Concentration (grains/ft3) = 0.0052
Irapactor flow rate (acfm) = 0.6601
Location Outlet
Orifice D
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (op)
H20 (grams or %)
Meter volume (ft3)
Avg. Ps (±H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
= 30.00
- 28.7
= 2.45
= 76
- 24
= 85.22
= 1
= 74
- 120
- 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0028
0.0008
0. 0007
0.0005
0.0009
0. 0008
0.0025
0.0029
0.0057
0. 0084
0.0260
. % on
stage
10.76
3.07
2.69
1.92
3.46
3.07
9.61
11.15
21.92
32.30
Size cutoff
(urn)
11.81
7.38
4.94'
3.43
2.14
1.11.
0.67
0.45
% < stated
size
89.23
86.15
83.46
81.53
78.07
75.00
65.38
54.23
32.30
dm/d log D
0.0007
0.0006
0.0011
0. 0008
0.0017
0.0027
0.0067
0.0010
Geo. Mean
(nm)
9.34
6.04
4.11
2.71
1.54
0.86
0.55
0.06
84
-------
ANDERSEN IMPACTOR
Date 11/12/75 Run # (3222-3)
Sample volume at STP (ft3) a 10.9630
Moisture (7.) » 0.9492
Concentration (grains/ft3) a 0.1040
Impactor flow rate (acfm) •* 0.5529
Location Inlet
Orifice E
Bar. press. ("Hg) * 29.93
Mw a 29
Avg. Pm (-"Hg) = 1.5
Avg. Tm (op) a 76
H2° (grams or 7.) = 2.2
Meter volume (ft3) = 11.56
Avg. PS (+H20) a -1
Avg. Ts (op) a 74
Time (minutes) a 20
Correction factor a 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0032
0.0012
0.0010
0.0012
0.0014
0.0053
0.0203
0.0227
0.0137
0.0032
0.0732
7. on
stage
4.37
1.63
1.36
1.63
1.91
7.24
27.73
31.01
18.71
4.37
— —
Size cutoff
12.9101
8.0688
5.4112
3.7541
2.3381
. 1.2175
0.7427
0. 5064
7. < stated
size
95.62
93.98
92.62
90.98
89.07
81.83.
54.09
23.08
4.37
• " "
dm/d log D
0.0070
0.0098
0.0125
0.0366
0.1016
0. 1503
0.1170
0. 0027
i
Geo. Mean
10.20
6.60
4.50
2.96
1.68
0.95
0.61
0.07
1
'•' :
85
-------
ANDERSEN IMPACTOR
Date 11/12/75 Run #(3222-3)
Sample volume at STP (ft3) - 80.3666
Moisture (%) = 1.3537
Concentration (grains/ft3) = 0.0051
Impactor flow rate (acfm) = 0.6756
Location Outlet
Orifice D
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (op)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (±H20)
Avg. Ts (OF)
Time (minutes)
a 29.93
a 28.7
2.50
a 78
a 23
» 87.80
a 1
a 74
a 120
Correction factor a
Stage
Probe &
expander
0
I
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0022
0.0007
0.0008
0.0006
0. 0008
0.0006
0.0024
0.0040
0.0061
0.0079
0.0261
% on
stage
8.42
2.68
3.06
2.29
3.06
2.29
9.19
15.32
23.37
30.26
Size cutoff
(jim)
11.67
7.29
4.88
3.39
2.12
1.09
0.66
0.45
% < stated
size
91.51
88.88
85.82
83.52
80.45
78.16
68.96
53.63
30.26
dm/d log D
0.0008
0.0007
0.0010
0.0006
0.0016
0.0036
0.0070
0.0009
Geo. Mean
(nm)
9.23
5.97
4.07
2.68
1.52
0.85
0.54
0.06
86
-------
ANDERSEN IMPACTOR
Date 11/13/75 Run * (3222-4)
Sample volume at STP (ft3) a 10.2335
Moisture (7.) = 1.0169
Concentration (grains/ft3) « 0.1190
Impactor flow rate (acfm) » 0.5260
Location Inlet
Orifice E
Bar. press. ("Hg) » 29.37
Mw a 29
Avg. Pm (-"Hg) a 1.5
Avg. Tm (OF) a 76
H2° (S^ams or 7.) a 2.2
Meter volume (ft3) « 11.00
Avg. Ps (±H20) = -1
Avg. Ts (OF) a 74
Time (minutes) a 20
Correction factor a l
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0016
0.0023
0.0008
0.0006
0.0014
0.0017
0.0202
0.0230
0.0232
0. 0033
0.0781
7. on
stage
2.04
2.94
1.02
0.76
1.79
2.17
25.86
29.44
29.70
4.22
Size cutoff
(um)
13.2365
8.2728
5.5507
3.8503
2.3959
1.2489
0.7630
0.5212
7. < stated
size
97.95
95.00
93.98
93.21
91.42
89.24
63.38
33.93
4.22
dm/d log D
0.0060
0. 0053
0.0134
0.0126
0.1088
0.1637
0.2136
0.0029
Geo. Mean
-------
ANDERSEN MPACTOR
Date 11/13/75 Run # (3222-4)
Sample volume at STP (ft3) m 73.3436
Moisture (%) » 1.1286
Concentration (grains/ft3) = 0.0052
Impactor flow rate (acfm) '« 0.6319
Location Outlet
Orifice D
Bar. press. ("Hg) • 29.37
Mw => 28.7
Avg. Pm (-"Hg) = 2.4
Avg. Tm (OF) = 78
H20 (grams or 7.) » 17.5
Meter volume (ft3) « 81.68
Avg. Ps (+H20) - 1
Avg. Ts (°F) » 77
Time (minutes) » 120
Correction factor » 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0018
0.0000
0.0002
0.0003
0.0005
0.0011
0.0025
0.0053
0.0040
0.0076
0.0243
% on
stage
7.40
0.00
0.82
1.23
0.05
4.52
14.40
21.81
16.46
31.27
Size cutoff
(um)
12.0771
7.5482
5.0553
3.5086
2.1905
1.1374
0.6913
0.4690
% < stated
size
92.59
92.59
91.76
90.42
88.47
83.95
69.54
47.73
31.27
dm/d log D
0. 0002
0.0004
0. 0007
0.0011
0.0026
0.0052
0.0051
0.0010
Geo. Mean
(UN)
9.54
6.17
4.21
2.77
1.57
0.88
0.56
0.06
88
-------
ANDERSEN IMPACTOR
pate 11/24/75 Run # 2222
Sample volume at STP (ft3) m 7.4563
Moisture (7.) » 1.0150
Concentration (grains/ft3) a 0.0721
Impactor f}.ow rate (acfm) a 0.4934
Location Inlet
Orifice E
Bar. press. ("Hg) • 30.20
Mw =29
Avg. Pm (-"Hg) a 1.8
Avg. Tm (op) m 69.5
HaO (grama or %) - 1.6
Meter volume (ft3) a 7.77
Avg. Ps (±H20) a -0.37
Avg. Ts (OF) a 71
Time (minutes) » 15
Correction factor » 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gn>)
0. 0028
0. 0003
0. 0004
0. 0004
0. 0009
0. 0029
0.0119
0.0051
0.0079
0.0019
0. 0345
% on
.. stage
8.11
0.86
1.15
1.15
2.60:
8.40
34.49
14.78
22.89
5.50
Size cutoff
(urn)
13.6671
8.5419
5.7350
3.9774
2.4721
1.2904
0.7897
0.5408
% < stated
size
91.88
91.01
89.85
88.69
86.08
77.68
43.10
28.40
5.50
dm/d log D
0.0041
0.0048
0.0118
0.0294
0.0881
0.0500
0.1004
0.0023
Geo. Mean
(um)
10.80
6.99
4.77
3.13
1.78
1.00
0.65
0.07
89
-------
ANDERSEN IMPACTOR
Date 11/24/75 Run # 2222
Sample volume at STP (ft3) o 79.7884
Moisture (%) - 1-3635
Concentration (grains/ft3) = 0.0046
Impactor flow rate (acftn) = 0.6661
Location Outlet
Orifice D
Bar. press, ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (OF)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
30.20
28.7
2.62
73
23
85.88
0.34
76
120
1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0009
0.0002
0.0004
0.0005
0.0007
0.0012
0.0034
0.0056
0.0041
0.0064
0.0234
7. on
stage
3.8462
2.8547
1.7094
2.1368
2.9915
5.1282
14.5200
23.9316
17.5214
27.3504
Size cutoff
(um)
11.7623
7.3515
4.9209
3.4159
2.1346
1.1072
0.6719
0.4549
7. < stated
size
96.1538
95.2991
93.5897
91.4530
88.4615
83.3333
68.8034
44.8718
27.3504
dm/d log D
0.0004
0.0006
0.0009
0.0012
0.0023
0.0051
0.0047
0.0008
Geo. Mean
(urn)
9.2989
6.0146
4.0999
2.7003
1.5374
0.8625
0.5529
0.0674
90
-------
ANDERSEN IMPACTOR
Date 12/3/75 Run # 1222-1A
Sample volume at STP (ft3) a 7.5612
Moisture (7.) • 1.0001
Concentration (grains/ft3) a 0.0946
Impactor flow rate (acfm) « 0.5074
Location Inlet
Orifice E
Bar. press. ("Hg) « 29.97
Mw a 29
Avg. Pm (-"Hg) » 1.3
Avg. Tin (op) a 74.5
H20 (grams or 7.) «. 1.5
Meter volume (ft3) a 7.88
Avg. PS (±H20) » -0.04
Avg. Ts (op) „ 75
Time (minutes) « 15
Correction factor « 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
~~
Net weight
(gm)
0.0016
0. 0000
0.0000
0.0000
0.0018
0.0003
0.0101
0.0004
0.0083
0.0160
0.0459
7. on
stage
3.4858
O.OOOC
O.OOOC
O.OOOC
, 3.9216
0.6536
22.0043
16.122C
18.9542
34.8584
Size cutoff
(urn)
13.4762
8.4226
5.6533
3.9210
2.4383
1.2719
0.7778
0.5321
7. < stated
size
96.5142
96.5142
96.5142
96.5142
92.5926
91.9389
69.9346
53.8126
i
34.8584
t— — — . — _ _
dm/d log D
0. 0000
0. 0000
—^— —•••••••••••«.
0. 0233
0.0029
0.0737
0.0714
0. 1087
0.0191
Geo. Mean
(urn)
10.6539
6.9004
«>*>*»^«»«MMBMM»
4.7082
3.0921
1.7611
0. 9947
0.6433
————•-»•.•,•
0.0729
91
-------
ANDERSEN IMPACTOR
Date 12/3/75 Run # 1222-lA
Sample volume at STP (ft3) a 73.0685
Moisture (%) = 1.0357
Concentration (grains/ft3) = 0.0093
Impactor flow rate (acfm) = 0.6239
Location Outlet
Orifice D
Bar. press. ("Hg) = 29.97
Mw = 28.7
Avg. Pm (-"Hg) - 2.1
Avg. Tm (OF) m 76.5
H20 (grams or %) a 16
Meter volume (ft3) a 78.60
Avg. Ps (+H20) » 0.34
Avg. Ts (OF) = 84
Time (minutes) » 120
Correction factor = 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gin)
0.0012
0.0000
0. 0005
0.0008
0.0011
0.0020
0.0102
0.0163
0.0092
0.0022
0.0435
% on
stage
2.7586
0.0000
1.1494
1.8391
2.5287
4.5977
23.4483
37.4713
21.1494
5.0575
Size cutoff
(urn)
12.1540
7.5963
5.0881
3.5312
2.2041
1.1448
0.6960
0.4724
% < stated
size
97 . 2414
97.2414
96.0920
94.2529
91.7241
87.1264
63.6782
26.2069
5.0575
dm/d log D
0.0005
0.0010
0.0015
0.0021
0.0077
0.0161
0.0117
0.0003
Geo. Mean
(urn)
9.6086
6.2170
4.2388
2.7898
1.5885
0.8927
0.5734
0.0687
92
-------
ANDERSEN IMPACTOR
Date 11/14/75 Run # (3212-1)
Sample volume at STP (ft3) » 7.4943
Moisture (7.) « 0.9467
Concentration (grains/ft3) a 0.1328
Itnpactor flow rate (acfm) » 0.5136
Location Inlet
Orifice E
Bar. press. ("Hg) - 29.37
Mw a 29
Avg. Pm (-"Hg) a 1.72
Avg. Tm (OF) a 73
H20 (grams or 7.) a 1.5
Meter volume (ft3) a 8.08
Avg. Ps (+H20) a -1
Avg. Ts (OF) - 74
Time (minutes) •» 15
Correction factor a 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gro)
0.0033
0.0007
0.0007
0.0006
0.0014
0.0038
0.0175
0.0198
0.0116
0.0045
0.0639
% on
stage
5.1643
1.0955
1.0955
0.9390
2.1909
5.9468
27.3865
30.9859
18.1534
7 . 0423
Size cutoff
(pro)
13.3952
8.3720
5.6187
3.8972
2.4240
1.2642
0.7728
0.5284
% < stated
size
94.8357
93.7402
92.6448
91.7058
89.5149
83.5681
56.1815
25.1956
7.0423
dm/d log D
0.0071
0.0072
0.0183
0.0383
0.1267
0.1925
0.1460
0.0054
Geo. Mean
(H«n)
10.5898
6.8585
4.6794
3.0736
1.7505
0.9884
0.6390
0.0727
93
-------
ANDERSEN IMPACTOR
Date 11/14/75 Run # (3212-1)
Sample volume at STP (ft3) a 80.2594
Moisture (7.) = 1.3614
Concentration (grains/ft3) = 0.0046
Impactor flow rate (acfm) =» 0.6966
Location Outlet
Orifice D
Bar. press. ("Hg) = 29.37
Mw = 28.7
Avg. Pro (-"Hg) = 2.33
Avg. Tm (op) « 76
H20 (grams or 7.) =. 23.1
Meter volume (ft3) = 88.61
Avg. Ps (±H20) = 1
Avg. Ts (OF) = 81
Time (minutes) = 120
Correction factor =» 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gin)
0.0009
0.0004
0.0005
0.0006
0.0008
0.0011
0.0042
0.0058
0.0042
0.0053
0.0238
7. on
stage
3.7815
1.6807
2.1008
2.5210
3.3613
4.6218
7.6471
24.3697
17.6471
22.2689
Size cutoff
(urn)
11.5023
7.1889
4.8100
3.3393
2.0884
1.0822
0.6560
0.4434
7. < stated
size
96.2185
94.5378
92.4370
89.9160
86.5546
81.9328
64.2857
39.9160
22.2689
dm/d log D
0.0005
0.0007
0.0010
0.0011
0.0029
0.0052
0.0048
0.0006
Geo. Mean
(urn)
9.0933
5.8803
4.0077
2.6408
1.5034
0.8426
0.5393
0.0666
94
-------
ANDERSEN XMPACTOR
Date 11/15/75 Run # (3212-2)
Sample volume at STP (ft3) » 7.1999
Moisture (7.) = 0.8869
Concentration (grains/ft3) » 0.1784
Impactor flow rate (acfm) » 0.4858
Location Inlet
Orifice E
Bar. press. ("Hg) = 29.82
Mw » 29
Avg. Pm (-"Hg) » 1.4
Avg. Tm (OF) » 72
H20 (grams or 7.) « 1-35
Meter volume (ft3) « 7.54
Avg. Ps (+H20) « -1
Avg. Ts (op) . 74
Time (minutes) » 15
Correction factor » 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gn>)
0.0015
0.0010
0.0008
0.0008
0.0144
0.0011
0.0082
0.0130
0.0116
0.0306
0.0825
7. on
stage
1.8182
1.2121
0.9697
0.9697
17.4545
1.3333
9.9394
15.7576
14.0606
36.4848
Size cutoff
(urn)
13.7732
8.6083
5.7804
4.0087
2.4909
1.3006
0.7963
0.5456
7. < stated
size
9J.1818
96.9697
96.0000
95.0303
77.5758
76.2424
66.3030
50.5455
36.4848
dm/d log D
0.0085
0.0100
0.1959
0.0115
0.0628
0.1319
0.1528
0.0375
Geo. Mean
(urn)
10.8887
7.0540
4.8137
3.1599
1.7999
1.0177
0.6591
0.0739
95
-------
ANDERSEN IMPACTOR
Date 11/15/75 Run # (3212-2)
Sample volume at STP (ft3) « 71.5421
Moisture (7.) = 1.0909
Concentration (grains/ft3) = 0.0044
linpactor flow rate (acfm) » 0.6093
Location Outlet
Orifice D
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tin (OF)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
= 29.82
a 28.7
= 2.25
„ 72
a 16.5
a 77.10
a 1
a 79
a 120
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0009
0.0004
0.0007
0.0006
0.0005
0.0009
0.0029
0.0048
0.0035
0.0051
0.0203
7. on
stage
4.4335
1.9704
3.4483
2.9557
2.4631
4.4335
14.2857
23.6453
17.2414
25.1232
Size cutoff
(nn>)
12.2989
7.6868
5.1500
3.5739
2.2208
1.1587
0.7050
0.4789
7. < stated
size
95.5665
93.5961
90.1478
87.1921
84.7291
80.2956
66.0099
42.3645
25.1232
dm/d log D
0.0007
0.0008
0.0007
0.0010
0.0022
0.0048
0.0045
0.0007
Geo. Mean
(urn)
9.7231
6.2918
4.2902
2.8230
1.6074
0.9038
0.5810
0.0692
96
-------
ANDERSEN IMPACTOR
Date 11/22/75 Run # (3221-1)
Sample volume at STP (ft3) - 6.5745
Moisture (7.) «. 1.2231
Concentration (grains/ft3) B 0.1226
Impactor flow rate (acfm) ** 0.4408
Location Inlet
Orifice E
Bar. press. ("Hg) = 29.87
Mw a 29
Avg. Pm (-"Hg) m 1.45
Avg. Tin (OF) a 74
H20 (grains or 7.) a 1.7
Meter volume (ft3) a 6.89
Avg. Ps (±H20) « -1
Avg. Ts (OF) a 74
Time (minutes) a 15
Correction factor a 1
Stage
probe &
expander
0
1
2
3
4
5
6
—
7
F
~~
Total
•*"~
Net weight
(gm)
0.0028
0.0000
0.0008
0.0006
0.0009
0.0020
0.0113
0.0210
0.0096
0.0028
0.0516
7. on
stage
5.4263
0.0000
1.1628
1.1628
1.7442
3.8759
21.8992
40.6977
18.6046
5.4263
Size cutoff
(lira)
14.4585
9.0366
6.0739
4.2110
2.6120
'•'
1.3666
0.8390
ni
0.5770
i
7. < stated
size
9 v. 57 36
94.5736
— — — — _ _
93.4108
92.2481
90.5039
86.6279
64.7287
24.0310
.
51.4263
i
'
dm/d log D
0.0070
~ _
0.0083
0.0134
0.0229
0.0054
0.2355
— — ~— .
0.1403
0.0037
• *
———————
— — — ^ — — —
Geo. Mean
CM»)
11.4305
7.4086
5.0574
3.3165
1.8894
11.0708
0.6958
0.0760
97
-------
ANDERSEN IMPACTOR
Date 11/22/75 Run # 3221-1
Sample volume at STP (ft3) » 66.0387
Moisture (%) = 1.3394
Concentration (grains/ft3) » 0.0035
Impactor flow rate (acfm) ** 0.5667
Location Outlet
Orifice D
Bar. press. ("Hg) - 29.87
Mw = 28.7
Avg. Pm (-"Hg) o 1.8
Avg. Tm (op) • 76
H20 (grains or %) » 18.7
Meter volume (ft3) = 70.25
Avg. Ps (+H20) • 1
Avg. Ts (°F) - 84
Time (minutes) » 120
Correction factor » 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0003
0.0002
0.0002
0.0002
0.0005
0.0006
0.0017
0.0086
0.0034
0.0040
0.0147
% on
stage
2.0403
1.3605
1 . 3605
1.3605
3.4014
4.0816
11.5646
24.4896
23.1293
27.2109
Size cutoff
(urn)
12.7528
7.9705
5.3439
3.7077
2.3103
1 . 2024
0.7330
0.4993
% < stated
size
9 .9592
96.5986
95.2381
93.8776
90.4762
86.3946
74.8299
50.3401
27.2109
dm/d log D
0.0002
0.0003
0.0007
0.0007
0.0014
0.0040
0.0048
0.0006
Geo. Mean
(um)
10.0820
6.5264
4.4513
2.9267
1.6667
0.9388
0.6050
0.0707
98
-------
ANDERSEN IMPACTOR
pate 11/22/75 Run # 3221-2
Sample volume at STP (ft3) 9 6.5584
Moisture (7.) a 1.2261
Concentration (grains/ft3) = 0.1248
Iinpactor flow rate (acfm) - 0.4419
Location Inlet
Orifice E
Bar. press. ("Hg) » 29.87
Mw „ 29
Avg. Pm (-"Hg) « 1.45
Avg. Tm (OF) • 73
H20 (grams or %) = 1.7
Meter volume (ft3) B 6.86
Avg. Ps (+H20) « -1
Avg. Ts (opj =, 74
Time (minutes) a 15
Correction factor » 1
probe &
expander
99
-------
ANDERSEN IMPACTOR
Date 11/22/75 Run * 3221-2
Sample volume at STP (ft3) a 65.8084
Moisture (7.) - 1.3297
Concentration (grains/ft3) - 0.0038
Impactor flow rate (acfm) = 0.5647
Location Outlet
Orifice D
Bar. press. ("Hg) = 29.87
Mw » 28.7
Avg. Pm (-"Hg) - 1.9
Avg. Tm (op) - 74
H20 (grams or %) =» 18.5
Meter volume (ft3) « 70
Avg. Ps (±H20) - 1
Avg. Ts (OF) » 84
Time (minutes) «a!20
Correction factor » 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
m
Net weight
(gm)
0.0000
0.0003
0.0007
0.0004
0.0008
0.0009
0.0021
0.0035
0.0032
0.0039
0.0158
% on
stage
0.0000
1.8987
4.4304
2.5316
5.0633
5.6962
13.2911
22.1519
20.2532
24.6835
Size cutoff
(n«)
12.7751
7.9844
5.3535
3.7143
2.3142
1 . 2045
0.7344
0.5003
% < stated
size
1 JO. 0000
98.1013
93.6709
91.1392
86.0759
80.3797
67.0886
44.9367
24.6835
dm/d log D
0.0008
0.0005
0.0012
0.0010
0.0018
0.0039
0.0046
0.0005
Geo. Mean
(urn)
10.0996
6.5379
4.4592
2.9318
1.6696
0.9405
0.6062
0.0707
100
-------
ANDERSEN IMPACTOR
Date 11/21/75 Run # 3223-1
Sample volume at STP (ft3) «. 5.9974
Hoisture (7.) m 1.1491
Concentration (grains/ft3) • 0.12068
Impact or flow rate (acfm) =» 0.4736
Location Inlet
Orifice E
Bar. press. ("Hg) » 29.74
Mw a 29
Avg. Pm (-"Hg) . 1.45
Avg. Tm (op) a 76
H20 (grams or 7.) » 1.7
Meter volume (ft3) =» 7.4
Avg. Ps (+H20) a -1
Avg. Ts (OF) « 74
Time (minutes) =• 15
Correction factor • 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(g«n)
0.0033
0.0014
0.0008
0.0011
0.0015
0.0020
0.0111
0.0172
0.0119
0.0038
0.0541
7. on
stage
6.010
2.589
1.479
2.033
2.772
3.697
20.518
31.793
21.996
7.024
Size cutoff
(um)
13.9499
8.7187
5.8561
4.0608
2.5221
1.3176
0.8073
0.5537
7. < stated
size
T3.900
91.312
89.834
87.800
85.028
81.331
60.813
29.020
7.024
dm/d log D
0.0087
0.0142
0.0210
0.0216
0.0878
0.1803
29.020
0.0049
Geo. Mean
(um)
11.0283
7,1454
4.8765
3.2003
1.8230
1.0314
0.1621
0.0744
101
-------
ANDERSEN IMPACTOR
Date 11/21/75 Run # 3223-1
Sample volume at STP (ft3) * 70.9377
Moisture (7.) = 1.8670
Concentration (grains/ft3) a 0.0034
Impactor flow rate (acfm) =» 0.6136
Location Outlet
Orifice D
Bar. press. ("Hg) = 29.74
Mw « 29
Avg. Pm (-"Hg) - 2.1
Avg. Tin (op) a 78
H20 (grains or 7.) » 28
Meter volume (ft3) = 76.51
Avg. Ps (+H20) » 1
Avg. Ts (OF) - 86
Time (minutes) = 120
Correction factor » 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0011
0.0004
0.0003
0.0004
0.0004
0.0010
0.0010
0.0044
0.0031
0.0032
0.0153
7. on
stage
7.1895
2.6144
1.9608
2.6144
2.6144
6.5359
6.5359
28.7582
20.2614
20.9150
Size cutoff
(ura)
12.2552
7.6505
5.1313
3.5611
2.2220
1.1545
0.7023
0.4769
7. < stated
size
9' .8105
90.1961
88.2353
85.6209
83.0065
76.4706
69.9346
41.1765
20.9150
dra/d log D
0.0003
0.0005
0.0006
0.0011
0.0008
0.0045
0.0041
0.0004
Geo. Mean
(urn)
9.6886
6.2692
4.2747
2.8130
1.6017
0.9004
0.5787
0.0691
102
-------
ANDERSEN IMPACTOR
Date 11/23/75 Run # 3223-2
Sample volume at STP (ft3) =r 5.1961
Moisture (7.) » 1.4565
Concentration (grains/ft3) o 0.1211
Impactor flow rate (acfm) » 0.3439
Location Inlet
Orifice E
Bar. press. ("Hg) » 30.41
Mw <. 29
Avg. Pm (-"Hg) » 1
Avg. Tin (op) a 72
H20 (grams or %) o 1.6
Meter volume (ft3) =. 5.23
Avg. Ps (+H20) = -1
Avg. Ts (°F) =t 74
Time (minutes) » 15
Correction factor » 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gn>)
0.0015
0.0007
0.0005
0.0002
0.0005
0.0015
0.0067
0.0132
0.0096
0.0058
0.0402
7. on
stage
3.7313
1.7413
1.2438
0.4975
1 .2438
3.7313
16.6667
32.8358
23.8806
14.4279
Size cutoff
([Am)
16.3701
10.2313
6.8941
4.7761
2.9493
1.5512
0.9589
0.6658
7. < stated
size
9'..2687
94.5274
93.2836
92.7861
91.5423
87 . 8109
71.1443
38.3085
14.4279
dm/d log D
0.0074
0.0035
0.0095
0.0216
0.0723
0.1904
0.1825
0.0096
Geo. Mean
(nm)
12.9417
8.3985
5.7382
3.7531
2.1389
1.2196
0.7980
0.0816
103
-------
ANDERSEN IMPACTOR
Date 11/23/75 Run # 3223-2
Sample volume at STP (ft3) » 56.9631
Moisture (7.) • 1.3203
Concentration (grains/ft3) a 0.0035
Impact or flow rate (acfm) ** 0.4757
Location Outlet
Orifice
("Hg)
Bar. press.
Mw
Avg. Pm (-"Hg)
Avg. Tm (OF)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (OF)
Time (minutes)
Correction factor
» 30.41
- 28.7
" 1.6
» 72
- 15.9
o 58.61
» 1
- 79
o!20
= 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0011
0.0000
0.0000
0.0000
0.0001
0.0002
0.0007
0.0028
0.0028
0.0058
0.0129
% on
stage
8.5271
0.0000
0.0000
0.0000
0.7752
0.0000
5.4264
18.6047
21.7054
44.9612
Size cutoff
(am)
13.9192
8.6995
5.8429
4.0518
2.5167
1.3147
0.5523
7. < stated
size
U.4729
91.4729
91.4729
91.4729
90.6977
90.6977
85.2713
44.9612
dm/d log D
0.0000
0.0000
0.0002
0.0000
0.0007
0.0031
0.0047
Geo. Mean
(urn)
11.0041
7.1295
4.8656
3.1933
1.8190
1.0290
0.6669
104
-------
ANDERSEN IMPACTOR
Date 11/19/75 Run * 3122
Sample volume at STP (ft3) . 8.9309
Moisture (%) .0.9003
Concentration (grains/ft3) « 0.0824
Impactor flow rate (acfm) » 0.5922
Location inlet
Orifice E
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (op)
H2° (grams or 7.)
Meter volume (ft3)
Avg. P8 (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
30.31
29
2.07
75.7
1.7
9.48
- 0.42
74
15
1
Stage
probe &
expander
0
1
2
3
r
4
5
6
7
F
Total
Net weight
(en)
0.0051
0.0018
0.0006
0.0008
0.0009
0.0019
0.0070
0.0108
0.0140
0.0044
0.0473
% on
stage
10.7822
3.8054
1.2684
1.6913
1.9027
4.0169
14.7991
22.8329
29.5983
9.3023
Size cutoff
(urn)
12.4746
7.7966
5.2250
3.6257
2.2609
1.1756
0.7158
0.4868
% < stated
size
89.2177
85.4122
84.1437
82.4524
80.5496
—— •— — •••IM,^,
76.5327
61.7336
38.9006
9;3023
dm/d log D
0.0051
0.0080
0.0098
0.0161
— | ••••••••^•^
0.0429
0;.0873
0.1457
0.0045
Geo. Mean
(urn)
9.8620
6.3826
4.3525
2.8631
1.6303
0.9173
0.5903
0.0697
"••••—-•••••.—
-^— «^— — i«*^»^i™««
105
-------
ANDERSEN IMPACTOR
Date 11/19/75 Run # 3122
Sample volume at STP (ft3) » 75.8611
Moisture (7.) = 1.4208
Concentration (grains/£t3) = 0.0048
Impactor flow rate (acfm) « 0.6334
Location Outlet
Orifice D
Bar. press. ("Hg) = 30.32
Mw » 28.7
Avg. Pm (-"Hg) » 3.02
Avg. Tin (op) a 77.8
H20 (grams or %) » 22.5
Meter volume (ft3) = 83.2
Avg. Ps (±H20) - l.O
Avg. Ts (°F) = 80
Time (minutes) » 120
Correction factor «•
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gro)
0.0015
0.0005
0.0003
0.0005
0.0006
0.0010
0.0037
0.0059
0.0047
0.0046
0.0233
7. on
stage
5.4377
2.1459
1.2875
2.1459
2.5751
4.2918
15.8798
25.3218
20.1716
19.7424
Size cutoff
(um)
12.0616
7.5385
5.0486
3.5040
2.1877
1.1359
0.6903
0.4682
7. < stated
size
9J.5622
91.4163
90.1287
87.9828
85.4077
81.1158
65.2360
39.9141
19.7424
dm/d log D
0.0003
0.0005
0.000
0.0010
0.0026
0.0056
0.0057
0.0005
Geo. Mean
(urn)
9.5355
6.1692
4.2060
2.7687
1.5764
0.8855
0.5685
0.0684
106
-------
ANDERSEN IMPACTOR
Date H/18/75 Run # 2122
Sample volume at STP (ft3) a 6.8863
Moisture (7.) - 0.9866
Concentration (grains/ft3) a 0.09791
Impactor flow rate (acfm) a 0.4591
Location Inlet
Orifice
Bar. press. ("Hg) » 30.21
Mw •» 29
Avg. Pm (-"Hg) a 2.23
Avg. Tm (OF) a 73
H20 (grains or %) a 1.32
Meter volume (ft3) a 7.34
Avg. Ps (+H20) a - 0.48
Avg. Ts (OF) » 75
Time (minutes) a 15
Correction factor a 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gn>)
0.0037
0.0010
0.0007
0.0007
0.0010
0.0020
0.0094
0.0116
0.0108
0.0024
0.0433
7. on
stage
8.55
2.31
1.62
1.62
2.31
4.62
21.71
26.79
24.94
5.543
Size cutoff
(um)
14.1685
8.8553
5.9497
4.1254
2 . 5607
1.3386
0.8209
0.5637
7. < stated
size
31.46
89.146
87.529
85.912
83.603
78.984
57.275
30.485
5.543
dro/d log D
0.0078
0.0092
0.0142
0.0218
0.0754
0.1235
0.1486
0.0031
Geo. Mean
(nm)
11.2012
7.2585
4.9542
3.2502
1.8515
1.0483
0.6802
0.0750
107
-------
ANDERSEN IMPACTOR
Date 11/18/75 Run # 2122
Sample volume at STP (ft3) « 80.2753
Moisture (7.) - 0.9132
Concentration (grains/ft3) = 0.0106
Impactor flow rate (acfm) = 0.6773
Location Outlet
Orifice
Bar. press. ("Hg) = 30.21
Mw =» 28.7
Avg. Pm (-"Hg) => 2.9
Avg. Tm (OF) - 76
H20 (grains or 7.) =» 15.5
Meter volume (ft3) =» 88.15
Avg. Ps (±H20) - 1.02
Avg. Ts (OF) =» 81
Time (minutes) =• 120
Correction factor => 1
Stage
Probe &
expander
0
I
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0005
0.0003
0.0009
0.0013
0.0020
0.0028
0.0107
0.0174
0.0129
0.0059
0.0547
7. on
stage
0.9140
0 . 5484
1.6453
2.3766
3.6536
5.1188
19.5612
31.8098
23.5831
10 . 7 861
Size cutoff
(urn)
11 . 6645
7.2903
4.8792
3.3870
2.1172
1.0978
0.6659
0.4505
7. < stated
size
90.0850
98.5374
96.8921
94.5155
90.8592
85.7404
66.1791
34.3692
10.7861
dtn/d log D
0.0008
0.0014
0.0024
0.0026
0.0072
0.0155
0.0147
0.0006
Geo. Mean
(urn)
9.2216
5.9641
4.0652
2.6779
1.5245
0.8550
0,5477
0.0671
108
-------
ANDERSEN IMPACTOR
Date 12/5/75 Run # 1122
Sample volume at STP (ft3) = 9.0883
Moisture (%) = 0.8327
Concentration (grains/ft3) a 0.0877
Impactor flow rate (acfm) - 0.6062
Location Inlet
Orifice E
Bar. press. ("Hg) = 30.39
Mw => 29
Avg. Pm (-"Hg) » 2.2
Avg. Tm (op) a 74
H20 (grams or 7.) = 1.6
Meter volume (ft3) a 9.64
Avg. Ps (+H20) = -0.06
Avg. Ts (OF) = 79
Time (minutes) = 15
Correction factor =» 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0042
0.0015
0.0004
0.0006
0.0014
0.0037
0.0133
0.0172
0.0077
0.0012
0.0512
% on
stage
8.2031
2.9297
0.7813
1.1719
2.7344
7.2266
25.9766
33.5928
15.0391
2.3438
Size cutoff
(Mm)
12.3304
7.7065
5.1634
3.5832
2.2354
1.1618
0.7069
0.4803
% < stated
size
91.7969
88.8672
88.0859
86.9141
84.1797
76.9531
50.9766
17.3828
2.3438
dm/d log D
0.0034
0.0059
0.0151
0.0309
0.0801
0.1365
0.0785
0.0012
Geo. Mean
(urn)
9.7480
6.3081
4.3014
2 . 8302
1.6115
0.9062
0.5827
0.0693
109
-------
ANDERSEN 3MPACTOR
Date 12/5/75 Run # 1122
Sample volume at STP (ft3) a 37.5762
Moisture (7.) - 2.1342
Concentration (grains/ft3) m 0.0127
Impactor flow rate (acfm) - 0.6333
Location Outlet
Orifice D
Bar. press. ("Hg) =• 30.39
Mw = 28.7
Avg. Pm (-"Hg) =• 2.25
Avg. Tin (OF) » 75.5
H2<> (grams or %) = 17
Meter volume (ft3) = 39.52
Avg. Ps (+H20) = 0.36
Avg. Ts (OF) = 83
Time (minutes) = 60
Correction factor =» 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gn>)
0.0012
0.0001
0.0003
0.0006
0.0009
0.0011
0.0044
0.0087
0.0100
0.0030
0.0303
% on
stage
3.9604
0.3300
0.9901
1.9802
2.9703
3.6304
14.5215
28.7129
33.0033
9.9010
Size cutoff
(urn)
12.0629
7.5393
5.0492
3.5044
2.1879
1.1361
0.6904
0.4683
7. < stated
size
96.0396
95.7096
94.7195
92.7393
89.7690
86.1386
71.6172
42.9043
9.9010
dm/d log D
0.0006
0.0014
0.0024
0.0023
0.0065
0.0168
0.0248
0 . 0008
Geo. Mean
(n*)
9.5365
6.1699
4.2065
2.7690
1.5766
0.8856
0.5686
0.0684
110
-------
ANDERSEN IMPACTOR
Date 11/17/75 Run # (2322)
Sample volume at STP (ft3) = 7.2045
Moisture (7.) - I'0505
Concentration (grains/ft3) - 0.1359
Impactor flow rate (acfm) - 0.4818
Location Inlet
Orifice E
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (OF)
H20 (grains or %)
Meter volume (ft3)
Avg. Ps (±H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
30.16
29
1.5
74
1.6
7.5
- 0.3
76
15
1
111
-------
ANDERSEN IMPACTOR
Date 11/17/75 Run # (2322)
Sample volume at STP (ft3) = 65.1860
Moisture (%) = 0.8345
Concentration (grains/ft3) = 0.0038
Impactor flow rate (acfm) = 0.5428
Location Outlet
Orifice D
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (OF)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (+H20)
Avg. Ts (°F)
Time (minutes)
Correction factor
= 30.16
• 28.7
- 1.9
= 75
= 11.5
= 69.10
= 0.3
- 74
= 120
= 1
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0005
0 . 0003
0.0003
0.0005
0.0000
0.0009
0.0027
0.0053
0.0023
0.0032
0.0160
7. on
stage
3.1250
1.8750
1.8750
3.1250
0.0000
5.6250
16.8750
33.1250
14.3750
20.0000
Size cutoff
(urn)
13.0296
8.1435
5.4623
3.7893
2.3593
1.2290
0.7501
0.5118
% < stated
size
96.8750
95.0000
93.1250
90.0000
90.0000
84.3750
67 . 5000
34.3750
20.0000
dm/d log D
0.0004
0.0007
0.0000
0.0010
0.0023
0.0059
0.0033
0.0004
Geo. Mean
(UN)
10.3008
6.6695
v 4. 5495
2.9900
1.7028
0.9602
0.6196
0.0715
112
-------
ANDERSEN IMPACTOR
Date 12/13/76 Run # Metal Foam #1
Sample volume at STP (ft3) = 7.721717
Moisture (7.) = 0.918837
Concentration (grains/ft3) = 0.093176
Impactor flow rate (acfm) = 0.510891
Location inlet
Orifice E
Bar. press. ("Hg)
Mw
Avg. Pm (-"Hg)
Avg. Tm (OF)
H20 (grams or 7.)
Meter volume (ft3)
Avg. Ps (±H20)
Avg. Ts (°F)
Time (minutes)
=30.38
B 29
- 2.13
"72.3
=1.5
= 3. 14
"-0.45
= 74°p
= 15
Correction factor = l
Stage
Probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0046
0.0011
0.0004
0.0010
0.0018
0.0035
0.0037
0.0147
0.0062
0.0006
0.04620
7. on
stage
12 . 234
2.926
1.064
2.660
4.787
9.309
9.840
39.096
16.489
1.596
Size cutoff
(urn)
13.431
8.394
5.634
3.908
2.430
1.268
0.775
0.530
7. < stated
size
87.766
84.840
83.777
81.117
76.330
67.021
57.181
18.085
1.596
dm/d log D
0.00395
0.01165
0.02285
0.03422
0.08776
0.13874
0.07578
0.00070
Geo. Mean
(urn)
10.6181
6.8770
4.69210
3.08172
1.7552
0.99118
0.64093
0.073
113
-------
ANDERSEN IMPACTOR
Date 1/13/76 Run % Metal Foam #1
Sample volume at STP (ft3) = 75.041321
Location Outlet
Orifice D
Moisture «) - 1.040027 Bar. press. ("Hg) - 30.38
Concentration (grains/ft3) = 0.009101 Mw a 28-7
AVB Pro ( " "HE^ — 2 72
Impactor flow rate (acfm) = 0.62204 Avg." Tm (op) I 72^4
H20 (grams or 70) a 16. 5g
Meter volume (ft3) = 80.71
Avg. Ps (+H20) a +0.92
Avg. Ts (°F) = 77
Time (minutes) = 120
Correction factor =» 1
Stage
Probe &
expander
0
1
2
3
•
4
5
6
7
F
Total
Net weight
(gm)
— — — — — — — —
0.0024
0.0003
-I.,.-—
0.0003
0.0008
0.0008
0.0027
"
0.0078
0.0138
__
0.0102
' - _
0.0047
0.0438
% on
stage
5.48
0.68
0.68
1.83
1.83
6.16
111 •M^^_V
17.81
11 -i.--
31.51
23.29
—
10.73
Size cutoff
(urn)
12.17
7.61
5.10
3.54
2.21
1.15
— — — — — —
0.697
0.473
— — — — — — —
% < stated
size
94.52
93.84
93.15
91.32
89.50
—————.
83.33
65.53
34.02
10.73
dm/d log D
0.00031
0.00096
0.00105
0.00274
0.00570
0.01327
0.01260
0.00058
Geo. Mean
(urn)
- -
9.6228
6.2263
4.2452
2.7940
1.59085
0.89404
0.574
0.069
114
-------
ANDERSEN IMPACTOR
Date 1/13/76 Run # Metal Foam #2
Sample volume at STP (ft3) = 7.726561
Moisture (%) = 0.918261
Concentration (grains/ft3) = Q.109241
Impactor flow rate (acfm) = 0.511212
Location Inlet
Orifice £
Bar. press. ("Hg) =» 30.38
Mw a 29
Avg. Pm (-"Hg) • 1.9
Avg. Tm (op) a 730
H20 (grains or %) = 1.5
Meter volume (ft3) = 8.09
Avg. Ps (+H20) m -0.45
Avg. Ts (°F) a 74
Time (minutes) » 15
Correction factor = 1
Stage
probe &
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gin)
0 ,0051
0.0009
0 . 0009
0.0010
0.0016
0.0043
0.0135
0.0172
0.0083
0.0014
0.0542
% on
stage
9.41
1.66
1.66
1.85
2.95
7.93
24.91
31.73
15.31
2.58
Size cutoff
(ura)
13.427
8.39
5.63
3.91
2.43
1.27
0.775
0.530
% < stated
size
90.59
88.93
87.27
85.42
82.47
74.54
49.63
17.90
2.58
dm/d log D
0.00889
0.01164
0.203
0.0420
0.0963
0.16223
0.10138
0.00164
Geo. Mean
(nm)
10.615
6.875
4.691
3.081
1.755
0.991
0.641
0.073
115
-------
ANDERSEN IMPACTOR
Date 1/13/76 Run # Metal Foam # 2 Location Outlet
Sample volume at STP (ft3) = 38.606736 Orifice D
Moisture (7.) = 1.163916 Bar. press. ("Hg)
Concentration (grains/ft3) = 0.008249 Jj pm (_tlHg)
Impactor flow rate (acfm) = 0.640046 Avg. Tm (OF)
H20 (grams or %)
Meter volume (ft3)
Avg. Ps (±H20)
Avg. Ts (OF)
Time (minutes)
Correction factor
- 30.38
= 28.7
= 2.26
a 73.8
9.52
= 40.9
= 0.92
= 77
" 60
" 1
Stage
Probe {,
expander
0
1
2
3
4
5
6
7
F
Total
Net weight
(gm)
0.0015
0.0000
0.0001
0.0005
0.0005
0.0011
0.0054
0.0084
0.0047
0.0012
0.0204
7. on
stage
7.35
0
0.49
2.45
2.45
5.39
26.47
41.18
8.33
5.88
Size cutoff
(nm)
12.00
7.50
5.02
3.49
2.18
1.13
0.687
0.466
7. < stated
size
92.65
92.65
92.16
89.71
87.25
81.86
55.39
14.22
5.88
dm/d log D
0.00020
0.00116
0.00128
0.00218
0.00767
0.0157
0.00407
0.00029
Geo. Mean
(urn)
9.486
6.137
4.184
2.755
1.568
0.881
0.565
0.068
116
-------
APPENDIX D
OPTICAL PARTICLE COUNTER MEASUREMENTS AND DATA
GCA/Technology Division utilizes a Bausch and Lomb Dust Counter Model
40-1 which is capable of optically counting dust particles at particle
sizes equal to or greater than 0.3 urn, 0.5 um, 1.0 urn, 2.0 urn, 3.0 mn,
5.0 ^m, and 10.0 |xm. The dust counter was used in conjunction with the
dilution systems mentioned for use with the CNC.
Readings were taken at 0.3 ^im and were often high enough to be very close
to the maximum counting capabilities of the instrument. Readings at
0.5 fzm and 1.0 fim were normally well within the counting range of the
instrument while readings at 2 ^im and higher were often very low or
zero. Therefore, the data were considered valid only for the 0.3 um,
0.5 um, 1.0 jxin, and 2.0 jim settings. Table 13 contains the data
obtained with the optical particle counter.
117
-------
Table 13. OPTICAL PARTICLE COUNTER DATA
Run no.
3222
Inlet
Outlet
1222
Inlet
Outlet
3122
Inlet
Particle
diameter
Lower
limit, u,m
0.3
0.5
1.0
2.0
0.3
0.5
1.0
2.0
0.3
0.5
1.0
2.0
0.3
0.5
1.0
2.0
0.3
0.5
1.0
2.0
3.0
5.0
Concentration
Number of
particles
per 0.01 ft3
6
3.19 x 10
6
1.34 x 10
5.83 x 105
4530
6
1.15 x 10°
4
3.18 x 10
1.28 x 104
637
1.67 x 106
5
2.58 x 10J
4
6.92 x 10
31
7.90 x 10
4
1.31 x 10H
3887
0
3.50 x 106
6
1.12 x 10
5.04 x 105
4
3.74 x 10*
2770
1517
118
-------
Table 13 (continued). OPTICAL PARTICLE COUNTER DATA
Run no.
Outlet
1122
Inlet
Outlet
Particle
diameter
Lower
limit, urn
0.3
0.5
1.0
2.0
3.0
5.0
0.3
0.5
1.0
2.0
0.3
0.5
1.0
2.0
Concentration
Number of
particles
per 0.01 ffc3
1.57 x 105
6.53 x 104
2.88 x 104
85.5
8.1
0
1.67 x 106
2.58 x 105
6.92 x 104
31
7.1 x 105
7910
2336
53
119
-------
APPENDIX E
CONDENSATION NUCLEI COUNTER MEASUREMENTS AND DATA
GCA/Technology Division utilizes a Rich Model 100 Condensation Nuclei
Counter (CNC) in conjunction with the Diffusion Denuder (DD) for sub-
micrometer particle sizing measurements. The efficiency of particle
removal by the DD is a function of flow rate through the device, as
shown in Figure 28, so that the utilization of different flow rates will
allow size distribution measurements with the CNC.
To obtain successful measurements with the CNC it is often necessary
to dilute and modify the static pressure of the gas stream to be
sampled. Various diluters are employed by GCA/Technology Division.
Each system mixes the sample stream with filtered air, and the
measured flow rates of the sample and dilution streams are used to
calculate the total sample dilution.
Dilution systems were utilized to dilute the gas stream up to 200 times
and to raise the static pressure of the gas stream, since the CNC will
not operate properly if the static pressure of the sample is too far
below atmospheric.
Table 14 contains the data obtained with the CNC.
120
-------
Ivi
O.I
0.08
O.O6
0.04
O.O2
0.01
i i i
I I
i i
_L
3 4 5 6 7 8 9 10 20 30 40
FLOW RATE THROUGH DIFFUSION-DENUDER,cc/sec.
I I 1 1—L
60 80 100
Figure 28. Diffusion denuder size cutoffs as a function of flow rate
-------
Table 14. CNC DATA
Run no.
3222
Outlet
Inlet
1222
Outlet
Inlet
3122
Outlet
Inlet
1122
Outlet
Inlet
Particle
diameter
Lower
limit, um
0.0025
0.0145
0.048
0.0025
0.0145
0.048
0.0025
0.0145
0.0025
0.0145
0.0025
0.0145
0.048
0.0025
0.0145
0.048
0.0025
0.0145
0.0025
0.0145
Concentration
Number of
particles
per cnr*
o , -.n6
3.1 x 10
6
2.8 x 10
5
6.4 x 10
6
3.8 x 10°
6
3.6 x 10
8.4 x 105
6
5.3 x 10
7.7 x 105
6
2.2 x 10
1.4 x 106
6
2.6 x 10
6
2.1 x 10°
5.8 x 105
6
2.8 x 10
6
3.0 x 10°
6
1.1 x 10
6
3.9 x 10°
9.2 x 105
6
2.2 x 10
6
1.4 x 10
122
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-202
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Laboratory Evaluation of the Cleanable High Efficiency
Air Filter (CHEAP)
5. REPORT DATE
July 1976
6. PERFORMING ORGANIZATION CODE
7. AUTMOR(S)
Manuel T. Rei and Douglas W. Cooper
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA Corporation
Burlington Road
Bedford, MA 01730
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
I. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-76-9-G
10. PROGRAM ELEMENT NO.
1AB012; ROAP 21ADL-004
11. CONTRACT/GRANT NO.
68-02-1487
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; Through 5/76
14. SPONSORING AGENCY CODE
EPA-ORD
is. SUPPLEMENTARY NOTES project officer for this report is D. L. Harmon, Mail Drop 61,
Ext 2925. Previous report in this series was EPA-650/2-75-024a.
report gives resuits of testing a novel scrubber, the Cleanable High
Efficiency Air Filter (CHEAF), as part of a program to identify novel high efficiency
fine particle control devices. The scrubber was tested at room temperature, using
iron oxide aerosols of concentrations near 0.2 g/cu m (0.1 grain/cu ft), mass median
aerodynamic diameter of 1.1 micrometers. Inlet and outlet samples were taken with
cascade impactors, total mass filters, a condensation nuclei counter, and an
optical particle counter. These tests were performed with different filter media, at
different face velocities, and at different water spray rates and water recycle rates.
Efficiency increased with increases in: foam pores per inch, pressure drop, flow
rate, spray rate, and make-up water addition. The results were consistent with the
hypothesis that impaction is the major collection mechanism and re-entrainment
contributes substantially to penetration. Total mass efficiency was approximately
95 percent at normal conditions, for which the pressure drop across the CHEAF was
80 cm (31.5 inches) WC. The particle aerodynamic cut diameter, for which the
efficiency would be 50 percent at these conditions, was determined from cascade
impactor data to be below 0.5 /im. This indicates that the 50 percent cut diameter for
the CHEAF is smaller than for a venturi scrubber operating at the same pressure drop.
17.
a.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Air Filters
Scrubbers
Dust
Aerosols
Tests
Air Pollution Control
Stationary Sources
CHEAF
Particulate
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
13B
13K
07A
11G
07D
14B
18. DISTRIBUTION STATEMENT
Unlimited
19, SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
132
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
123
------- |