United States
Environmental Protection
Agency
Office of
Reseach and
Development
Environmental Sciences RPS<-'.
Ldboititor v
EPA-600-7-77-033
Research Tn.inqlc Park, North Carolin,) 77711 Apri! 1977
COMPACT, IN-STACK, THREE
SIZE CUT PARTICLE
CLASSIFIER
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven 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 seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-033
April 1977
COMPACT, IN-STACK, THREE SIZE CUT PARTICLE CLASSIFIER
by
George E. Lacey
Kenneth M. Cushing
Wallace Bo Smith
Southern Research Institute
2000 Ninth Avenue South
Birmingham, Alabama 35205
Contract Number 68-02-1736
Project Officer
Kenneth T0 Knapp
Emission Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the ILS. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for usec
ii
-------
ABSTRACT
The goal of this research project was to design and con-
struct a sampling system which could be used in programs to
characterize particulate emissions from stationary pollution
sources.
A particle size classifier (PSC) impactor was designed
to accomplish this. It is a two-stage impactor made of anodized
aluminum with stainless steel jet and collection plates. This
PSC impactor is designed to measure the particle emissions in
three size ranges: nonrespiratory (>3 ym) , upper respiratory
tract (^1 to 3 ym) , and lower respiratory tract (<1 ym) .
Three sets of jet plates (two jet plates per set) are in-
cluded. Each set is designed for a different flow rate but with
all particle size cutpoints to give data in the three size ranges
of interest. A choice of flow rates is desirable to allow reason-
able sampling times at particulate emission sources with both
high and low mass loadings.
A complete sampling system was constructed for the PSC
impactor including probe, pitot tube, temperature controller,
magnehelic pressure guages, and flow rate metering orifices.
The impactor and sampling probe are designed to fit into four
inch sampling ports.
The PSC impactor has been calibrated in the laboratory with
monodisperse aerosols from a vibrating orifice aerosol generator .
Particle size distributions from PSC impactor data have been com-
pared to Andersen and Brink impactor particle size distributions
at three power plants.
With this compact sampling system, measurements of the three
respiratory-related size fractions are possible over a wide range
of test conditions.
This report was submitted in fulfillment of Contract No.
68-02-1736 by Southern Research Institute under the sponsorship
of the U.S. Environmental Protection Agency. This report covers
the period June 21, 1974 to March 31, 1976 and work was completed
on March 31, 1976.
i.ii
-------
CONTENTS
Abstract , iii
Figures vi
Tables ix
Acknowledgments x
1. Introduction 1
2. Design of the Particle Size Classifier Impactor 6
Description of calibration procedure 8
Sampling configuration , 18
References 76
Appendix 77
-------
FIGURES
Number Page
1. Particle size classifier (PSC)impactor 2
2. Two-stage prototype impactor 3
3. Two-stage PSC impactor showing new outer shell 4
4. Two-stage PSC impactor disassembled 5
5. Particulate deposition patterns for different
flow rates 7
6. Schematic representation of the vibrating
orifice aerosol generator 9
7. Ammonium fluorescein aerosol particles generated
using the vibrating orifice aerosol generator 12
8. Wall losses versus particle size for the PSC
impactor 14
9. Typical sampling setup for PSC impactor 19
10. Two-stage impactor sampling case 20
11. Two-stage impactor sampling case showing major
operational components 21
12. Top of sampling case containing pressure gauges and
substrate punches 22
13. Nomograph for selecting nozzles for isokinetic
sampling 24
14. Sampling time determination for total mass collection
of 25 milligrams 25
15. Sampling orifice calibration 27
16. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 1 - Stage 1 31
vi
-------
17. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 1 - Stage 2 32
18. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 2 - Stage 1 33
19. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 2 - Stage 2 34
20. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 3 - Stage 1 35
21. Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 3 - Stage 2 36
22. Collection efficiency vs. /? PSC impactor - Plate
Set 1 38
23. Collection efficiency vs. v PSC impactor - Plate
Set 2 39
24. Collection efficiency vs. /¥ PSC impactor - Plate
Set 3 40
25. Cumulative mass loading versus particle diameter
March 11, 1975 50
26. Cumulative mass loading versus particle diameter
March 12, 1975 51
27. Cumulative mass loading versus particle diameter
March 13, 1975 52
28. Cumulative grain loading versus particle size 57
June 25-26, 1975
29. Cumulative grain loading versus particle size 62
July 30-31, 1975
30. Cumulative particle size distribution for full load 68
January 28, 1976
31. Cumulative particle size distribution for half load 69
January 29, 1976
32. Differential particle size distribution on mass 70
basis for full load, January 28, 1976
33. Differential particle size distribution on mass 71
basis for half load, January 29, 1976
Vii
-------
34. Differential particle size distribution on numerical 72
basis for full load, January 28, 1976
35. Differefttial particle size distribution on numerical 73
basis for half load," January 29, 1976
36. Cumulative particle size distribution for ultrafine 74
region using the electrical aerosol analyzer,
January 28-29, 1976
viii
-------
TABLES
Number Page
1. Jet stage data for the PSC impactor 15
2. Viscosity of air vs. H20 content 41
3. Two-stage impactor test data, March 11-13, 1975 49
4. Two-stage impactor - Jet plate set I, June 25, 1975 54
5. Two-stage impactor - Jet plate set II, June 25, 1975 55
6. Two-stage impactor - Jet plate set III, June 26, 1975 56
7. Two-stage impactor - Jet plate set I, July 31, 1975 59
8. Two-stage impactor - Jet plate set II, July 31, 1975 60
9. Two-stage impactor - Jet plate set III, July 30, 1975 61
13. Two-stage impactor - Jet plate set I, January 28, 1976 64
11. Two-stage impactor - Jet plate set II, January 28, 1976 65
12. Two-stage impactor - Jet plate set I, January 29, 1976 66
13. Two-stage impactor - Jet plate set III 67
January 29, 1976
IX
-------
ACKNOWLEDGMENTS
The cooperation of Florida Power and Light Company, Alabama
Power Company, and the Tennessee Valley Authority in the field
testing phase of the project is gratefully acknowledged. The
assistance given by the plant personnel at the test sites is also
greatly appreciated.
Dr. Kenneth Knapp, EPA Project Officer is acknowledged for
his guidance and direction of this program.
-------
SECTION 1
INTRODUCTION
The particle size classifier (PSC) impactor is design-
ed to measure the particle emissions from stationary pollution
sources in three size ranges. These are the nonrespiratory
(>3 ym), upper respiratory (^1-3 ym), and lower respiratory tract
ranges (<1 ym). With this compact sampling system, measurements
of three size fractions within the flue gas over a wide range
of test conditions are possible.
The impactor consists of a basic housing, a set of nozzles,
a set of collection plates, and three sets of jet plates (two
jet plates per set). The impactor is shown in Figures 1 through
4. It is an in-stack device that samples particulate emissions
under stack or flue conditions.
-------
RETAINING
RING
IMFACTOR
INLET
NOZZLE
ALUMINUM
HOUSING
JET PLATE I
COLLECTION
PLATE I
JET PLATE 2
COLLECTION
PLATE 2
FILTER
HOLDER
Figure 1. Particle size classifier (PSC) impactor.
-------
Figure 2. Two-stage prototype impactor.
The impactor body is anodized aluminum.
The jet stages and collection stages are stainless
steel.
-------
Figure 3. Two-stage PSC impactor showing new outer shell with
retaining ring and V slots in base for pitot probe
attachment.
4
-------
Figure 4. Two-stage PSC impactor disassembled to show black
anodized aluminum body and stainless steel collection
plates and jet stages.
-------
SECTION 2
DESIGN OF THE PARTICLE SIZE CLASSIFIER IMPACTOR
The purpose of this contract was to devise and design a man-
ual sampling train to measure the particle emissions from station-
ary sources in three size ranges. The design goal was to provide
a compact, reliable system which could be used to obtain an accu-
rate measurement of the three size ranges over a wide range of
test conditions. Two general types of sampling systems were
deemed capable of providing the desired measurements: a two-stage
inertial impactor and a series cyclone system.
Tests with two cyclones assembled in series with a Gelman
47 mm filter holder showed the feasibility of a three-stage series
cyclone. The collection efficiency curves for the cyclones were
sufficiently sharp for good estimates of the amount of mass con-
tained in each size range. The cyclone system, however, requires
larger samples for accurate weighing, and at control device out-
lets with low mass concentrations the sampling time could become
excessively long. Therefore, the two-stage inertial impactor was
chosen.
Initial testing in the development of the two-stage impactor
involved using various jet plates from either the Andersen Model
III or University of Washington (Pilot) iiTipactors. Jet stages
were assembled into mock-up two-stage ir.jcc^or configurations and
tested with various monodisperse aerosols. Several combinations
of flow rates and jet sizes were tested to study particle deposi-
tion, bounce, blow off, and reentrainment.
The basic result of all these tests was that the lower jet
velocities resulted in more uniform particle deposition and great-
ly reduced particle reentrainment. Figure 5 shows data and illus-
trates this point. Jet stages from the particle size classifier
(PSC) impactor were used. Figures 5A and 5C show impactor con-
figurations where the D so for each stage was 1.8 pm. In Figure
5A the jet velocity was 11.4 m/sec while in the case shown in
Figure 5C the jet velocity was 4.2 m/sec. These jet stages were
used to collect 2.8 pm diameter ammonium fluorescein particles.
In the results illustrated in Figure 5A, 73% of the particles
were collected, although ideally 100% would have been collected.
In the results shown in Figure 5C, 92% of the particles were col-
lected. In Figure 5A the deposited patterns of particulate matter
-------
a. Vj = 11.4m/sec
73% COLLECTION
b. Vj = 45.1 m/sec
79% COLLECTION
32% WALL LOSSES
IN a AND b
c. Vj = 4.2 m/sec
92% COLLECTION
Figure 5.
d. Vj = 9.5 m/sec
94% COLLECTION
4% WALL LOSSES
INcANDd
Particulate Deposition Patterns for Different Flow Rates,
In all cases the particles were 2.8 ym diameter ammonium
fluorescein spheres. a. Dso=1.8 ym, b. Ds0=0.83 ym,
c. D5o=1.8 ym, d. Dso=0.38 ym.
-------
are not sharply defined but are blurred and smeared on the sub-
strate while those in Figure 5C are nice, circular, compact de-
posits .
Figures 5B and 5D show stages that were downstream of those
shown in Figures 5A and 5C. In the sample shown in Figure 5B, the
D so for that stage was 0.38 ym. For the configuration shown in
Figure 5D, the D 50 was 0.83 urn. Thus, collection of 100% of the
particles should be expected. The jet velocity in Figure 5B was
45.1 m/sec and only 79% of the particles were caught. The jet
velocity in Figure 5D was 9.5 m/sec and 94% of the particulate
matter was caught. At the higher jet velocities, scouring and
reentrainment were found to be significant.
Tests such as those described above indicate that a jet velo-
city of about 10 m/sec is the maximum which will give useful re-
sults for stage Dso's of 1 micron and longer. In order to have
Dso's of 0.5 micron or less, a 10 m/sec limit is impractical, al-
though as low a velocity as possible should be used. A lower ve-
locity results in lower particle momentum, and lower particle mo-
mentum minimizes particle bounce and reentrainment. Wall losses
are also reduced significantly when the lower jet velocities are
used.
Careful attention was given to selection of size and number
of jets for the PSC impactor to insure that jet velocities were
under 10 m/sec or as low as practical.
A total of six jet stages were designed and constructed in
an effort to make the impactor as versatile as possible. (See
Table 1 for jet stage data.) Combinations of jet stages are
available to permit sampling at flow rates from about 19 cm3/sec
for high grain loading situations to 94 cm3/sec for low grain load
situations while maintaining jet velocities below 10 m/sec.
DESCRIPTION OF CALIBRATION PROCEDURE
Laboratory evaluation of the PSC two-stage impactor involved
the use of a Vibrating Orifice Aerosol Generator (VOAG) as a source
of mpnodisperse aerosols. The VOAG used in this study was design-
ed and built at Southern Research Institute, although similar de-
vices have been reported by several authors previously1'2'3 and
a commercial version is available.* Figure 6 is a schematic dia-
gram which illustrates the operating principle of the VOAG. A sol-
ution of known concentration [in our case, a solution of fluores-
cein (CaoHisjOs) in 0.1N NtUOH] is forced through a small orifice
5, 10, 15, or 20 ym in diameter. The orifice is attached to a
*Thermo-Systems, Inc., 2500 Cleveland Ave., N., St. Paul,
MN 55113.
-------
Plexiglass Drying
Chamber
Vibrating
Orifice
Flow
Meters
Control
Valves
210
Po Charge Neutralizer
Signal Generator
Dry Air
Figure 6. Schematic representation of the vibrating orifice
aerosol generator
-------
piezoelectric ceramic which, under electrical stimulation, will
vibrate at a known frequency. This vibration imposes periodic
perturbations on the liquid and causes it to break up into uni-
formly-sized droplets. The droplet size can be readily calculated
fioji the liquid flow rate and the perturbation frequency. The
solvent evaporates from the droplets leaving the non-volatile
solute as a spherical residue. The final dry particle size can
be calculated from the droplet size through use of the known con-
centration of the liquid solution.
The dry particle diameter, dp, is calculated from the ex-
pression
where C is the solution concentration or volume of solute/volume
of solution,
Q is the solution flow rate (cm3/min), and
F is the perturbation frequency (hz).
By using smaller orifices, one can obtain higher operating
frequencies. This yields higher particle number concentrations
and allows a shorter sampling time to collect the same mass per
stage. The sampling time must be sufficiently long in each test
to allow accurate determination of the stage collection efficien-
cies and wall losses. The 20 pm orifice was consistently easier
to use in particle generation, primarily because of fewer clogging
problems.
Prior to particle generation the orifices were washed in
detergent with ultrasonic agitation and then rinsed several times
in distilled water, also with ultrasonic agitation. After the
filter and liquid handling system was flushed several times with
the aerosol solution to be used, an orifice was placed, still wet
with distilled water, into the crystal holder and the syringe
pump turned on. A jet of air was played over the orifice to keep
the surface clean until enough pressure was built up behind the
orifice to form a jet.
After a stream of particles was generated, a determination
of monodispersity had to be made. Two methods were used to accom-
plish this. By using a small, well-defined air jet to deflect
the stream of particles, it was possible to tell when the aerosol
was mono- or polydisperse. Depending on the droplet size, the
stream was deflected by the air at different angles. If the
aerosol was polydisperse, several streams could be seen at one
time. By adjusting the oscillation frequency of the crystal, the
10
-------
system could be fine tuned to give only a single deflected parti-
cle stream, thus indicating monodispersity. Polonium-210 alpha
sources were placed near the air stream as charge neutralizers to
reduce the loss of particles due to electrostatic forces. A
three-foot-high plexiglass cylinder was placed on the generator
and dispersion and dilution air turned on to disperse, dilute
and loft the particles into a plenum with several sampling ports.
Because amonium fluorescein is nonhygroscopic and has physical**
properties similar to fly ash, it was used as the test aerosol,
although in principle, any material that will dissolve readily
in an evaporable solvent could be used. On several occasions,
the aerosol tended to drift from monodispersity, and in order to
protect against this occurrence, periodic filter samples were
taken and checked by optical microscopy. This also provided a
good check on the sphericity of the aerosol because the final
particles were investigated instead of the primary liquid droplets.
Optical microscopy thus served as a check on proper drying, satel-
lites, correct size, and multiplets. Figure 7 shows one of the
test aerosols generated. In general, about 8% or less by mass of
the particles were of twice the volume (1.26 x diameter) of the
primary particles.
After it was determined that particles of the correct size
were being generated, samples were taken from the plenum with the
two-stage impactor containing the appropriate jet stages. Noniso-
kinetic sampling was performed; however, a series of tests in-
dicated that this did not affect the collection efficiency of
the impactor stages as compared to isokinetic sampling results.
The nozzle losses were probably influenced by anisokinetic samp-
ling.
At the conclusion of each test, the impactor was carefully
disassembled and all internal surfaces cleaned with a solution
of 0.1N NHi»OH. Each plate and surface was washed with a known
amount of the solution to dissolve and rinse off the ammonium
fluorescein particles.
The quantity of material collected on each surface was de-
termined by absorption spectroscopy. A Bausch and Lomb Spectron-
ic 88 Spectrophotometer, calibrated with solutions of known con-
centration of ammonium fluorescein, was used to measure the concen-
tration of ammonium fluorescein in each wash. From a knowledge of
the amount of wash solution, the dilution factor, if any, and the
absolute concentration, the mass of particles on each surface
could be calculated. With the mass on each plate and surface known,
the wall losses and stage collection efficiencies could be calcu-
lated.
Several particle sizes were used to measure the stage col-
lection efficiency curves. These were 15, 10, 7.0, 5.0, 3.8,
3.0, 2.0, 1.0, 0.7, and 0.5 micrometers diameter.
11
-------
Figure 7. Ammonium fluorescein aerosol particles generated
using the vibrating orifice aerosol generator.
The particle diameters are 5.4 ym.
12
-------
Wall Losses Versus Particle Size
Each surface of the impactor was washed individually after
each test to obtain data on losses occurring in nozzles, inlet
cones, jet plates, etc. Such losses can be attributed to parti-
cle settling, diffusion, electrostatic forces, bounce, and re-
entrainment. In the majority of the tests the greatest losses
occurred in the nozzles and inlet cones. These losses tend to
decrease to a minimum at a diameter of about 3 microns and then
remain low for smaller particles. The percentage total wall
losses for the three jet plate sets are shown in Figure 8.
It should be pointed out that the majority of these particles
are not irrevocably lost, but would be brushed onto the appropriate
collection surfaces, or retrieved by washing.
PSC Impactor Description
The PSC impactor is a two-stage sampler with a backup filter,
designed to fit through four-inch ports. The three sets of jet
stages are all the round jet type and made of stainless steel.
The first stage of each set has ten jets, and the second stage
has either fifty or ninety jets. Table 1 contains the speci-
fications for each set of jet stages. Each set of jet stages
has a different designed flow rate as noted in the table. Each
set of jet stages also has different nominal cutpoints to give
the operator some latitude in selecting the cutpoints he wishes.
Further cutpoint selections can be obtained by changing the
flow rates over a limited range. The use of different flow
rates for the stage sets also enables the operator to sample
sources of high or low mass loading and still maintain reasonable
sampling times. Jet velocities at these flow rates are all less
than 10 m/sec.
The collection plates are made of stainless steel and are
designed to accept doughnut-shaped collection substrates. The
first collection plate has a larger center hole and a larger
thickness than the second plate. The larger hole reduces the
collection area; however, a large area is not needed for the
single row of jets in the first jet stage. This larger hole is
designed to allow smoother gas flow from the collection plate to
the second jet stage with less loss of material. The greater
thickness is necessary for the correct jet-to-collection-plate
distance.
The final filter for the impactor is located just beneath
the second collection stage. A standard 47 mm filter is used.
The impactor housing is made of anodized aluminum. Aluminum
was chosen for its good heat transfer characteristics, which
enables quick in-stack heating of the impactor to prevent conden-
13
-------
I I I I I I I
O PLATE SETI
D PLATE SET H
A PLATE SET m
L_i.J I I I I
4 5 6 7 8 9 10
PARTICLE DIAMETER, Dp (microns)
20
30
40
Figure 8,
Wall losses versus particle size for the PSC impactor,
14
-------
TABLE 1. JET STAGE DATA FOR THE PSC IMPACTOR
At designed flow rate, 22°Cf and 749 mm Hg
for unit density spheres
Ol
Jet
II
III
Dia. of Jet-to-plate Jet
age
1
2
1
2
1
2
Designed No. of
flow rate jets
94
(0
47
(0
18
(0
cm3 /sec
.20 cfm)
.2 cm3/sec
.10 cfm)
.9 cm3/sec
.04 cfm)
10
50
10
90
10
50
0
0
0
0
0
0
jets,
cm
.325
.065
.167
.028
.088
.026
spacing,
cm
0
0
0
0
0
0
.56
.37
.56
.37
.56
.37
D5o ,
microns
6
1
4
0
3
1
.40
.63
.80
.99
.00
.22
velocity,
cm/sec /fys o
114
569
215
852
311
712
0.195
0.280
0.310
0.370
0.295
0.360
-------
sation in the impactor. The anodized finish gives good cor-
rosion resistance and makes the clean-up of the housing easier.
The construction of the housing is such that the nozzle, which
is threaded to fit the impactor inlet, can be aimed in any di-
rection the operator wishes. Thus, for a horizontal port with
either left-to-right or right-to-left gas flow, it is possible
to have the nozzle pointed upstream and still have the impactor
in a vertical position.
Ten stainless steel nozzles are included with the impactor
for isokinetic sampling. These nozzles allow isokinetic sampling
at flow rates from 19 cm3/sec (0.04 cfm) to 94 cm3/sec (0.20 cfm)
in ducts with gas velocities of about 3 to 20 meters per second
(10 to 70 feet per second). The nozzles' diameters are 1.0, 1.5,
2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, and 6.0 mm. They are designed
to allow the impactor to be inserted into four-inch ports.
Impactor Assembly and Preparation for Sampling
The collection substrates are specially shaped discs of
glass fiber filter material. Other materials can be used to
make substrates, although at present the glass fiber material
seems to be the best for collecting dry particulate matter. In
situations where the particulate matter is of a wet nature,
aluminum or stainless steel foil substrates can be used. Grease
can also be used with the metal substrates to aid particle
retention. Teflon was found to be unsuitable for collecting
dry laboratory particles, but worked well enough in field tests.
A special punch for cutting the glass fiber substrates is in-
cluded with this sampling system. Since both collection sub-
strates have the same external dimension, a single punch works
for both, and two interchangeable inner punches fit into the outer
one for cutting the different center holes.
The substrates are prepared for field use by baking at
200 to 300°C for two to four hours to remove volatile matter
and then desiccating for approximately 24 hours before weighing.
A balance with a sensitivity of 0.01 milligrams is required to
weigh the substrates since the amount of particulate matter to
be collected could be less than a milligram. The substrates
are placed in the collection plates after they have been desic-
cated and weighed.
The impactor is assembled with the collection substrates
in place and a 47 mm backup filter on the filter holder plate
as illustrated in Figure 1. Two types of o-rings (Parker No.
2-030) have been furnished with the impactor: Teflon and silicone.
The silicone o-rings have an upper temperature limit of 232°C
(450°F) and the Teflon o-rings are rated slightly higher. The
silicone o-rings probably provide the best seal but they tend to
-------
stick to the 47 nun backup filter more than the Teflon. A combina-
tion of the two types can be used.
Before the impactor is assembled, the housing, collection
plates, and especially the jet stages should be thoroughly cleaned,
preferably ultrasonically in a detergent solution. The jets must
be clear and free of any obstructions.
The impactor is assembled from "the bottom up", as follows:
1. Place the impactor base on a flat surface.
2. Insert a silicone (or Teflon) o-ring in the base's
o-ring groove.
3. Place the filter support on the base.
4. Center a pre-weighed 47 mm backup filter on the filter
support between the guide pins.
5. Insert an o-ring in the groove on the bottom of the
second collection stage and place this on the filter
and filter support making sure that the guide pins
on the filter support are inserted into their corres-
ponding holes.
6. Insert a pre-weighed collection substrate into the
second collection stage.
7. Place the selected second jet stage (with o-ring in-
serted in its groove) on the second collection stage.
8. Place the first collection stage (with o-ring inserted)
on the second jet stage.
9. Insert a pre-weighed collection substrate into the
first collection stage.
10. Place the selected first jet stage (with o-ring inserted)
on the first collection stage.
11. Place the outer shell around the assembly and screw
it to the base.
12. Attach the inlet cone with the retaining ring to the
outer shell making sure that the intake is pointed in
the proper direction.
13. Install the nozzle.
17
-------
SAMPLING CONFIGURATION
t
The basic sampling train arrangement is shown in Figure 9.
The stack gas passes in sequence through the impactor and probe,
a cooling coil-condenser, an ice bath, a desiccant, an orifice
flowmeter, and the vacuum pump. The condenser and drying column
are used to insure proper measurement of the flow rate and gas
volume and to protect the vacuum pump from damage. An insulated
ice box for immersing the cooling coil-condenser is furnished.
Two drying columns are mounted in the top of the insulated box.
In order to measure the flow rate, a calibrated orifice is
used. Three calibrated orifices have been furnished with the
system, one orifice for each of the designed flow rates. Proced-
ures for the calculation of flow rates using orifice pressure
drops are outlined below with all necessary equations given. The
total pressure drop across the impactor and system components is
needed for the flow rate calculation and is measured at a point
just upstream of the orifice. Magnehelic gauges with various
ranges are furnished in the impactor sampling case for measure-
ment of these pressure drops. The impactor sampling case is
shown in Figures 10, 11, and 12.
A gas meter can be used in place of, or in addition to, the
orifice to measure the flow rate. Gas meter flow rate calcula-
tion procedures are given below also.
A vacuum pump is enclosed in the sampling case with a hose
fitting for easy connection.
Sampling with the PSC Impactor
Isokinetic Sampling and Sampling time—
Included with the impactor is a reversed-type (S-type)
pitot tube which attaches to the bottom of the impactor housing.
The pitot should be attached to the housing so that its tips point
parallel to the gas stream. Sampling and pitot readings should
be done at a point in the duct where the velocity profile is
uniform, preferably several duct diameters upstream and down-
stream of any obstructions or bends.
To determine a gas velocity from the pitot tube pressure
reading, the following standard equation8 can be used:
Gas Velocity = C (1096.2 )*\\ J
where C = Coefficient for S-type pitot =0.86
P = Velocity pressure in inches of water
p = Gas density in Ib/cu ft
18
-------
Impactor
Probe
VO
Drying Column
Cooling Coil-Condenser
Magnehelic for system
pressure drop
Figure 9. Typical sampling setup for PSC Impactor.
Orifice
Magnehelic for orifice
pressure drop
-------
10
o
Figure 10. Two-stage impactor sampling case
-------
Storage Box for Accessories
Impactor Nozzles
Two-stage
Impactor
Impactor
Temperature Control
Sampling
Orifices r-HP-65 Calculator
Pump
On-Off
Switch
Power
Flow Adjust Valve
120 v Accessory Outlets
Pump Inlet
— Pump (Hung Below Plate)
Figure 11. Two-stage impactor sampling case showing major operational components
-------
to
tO
Sampling
Orifice
Pressure Gauge
Pitot
Pressure
Gauges
Impactor
Substrate
Punches
r—System Pressure
\ Drop Upstream of
Orifice
Pressure Gauge Connections
Figure 12. Top of sampling case containing pressure gauges and
substrate punches.
-------
Figure 13 is a chart for selecting the correct nozzle
for isokinetic sampling at a particular flow rate. Isokinetic
sampling is a must to obtain a representative sample and care
should be taken to insure that it is accomplished.
The length of the sampling time is dictated by mass loading
and particle size distribution. An estimate for initial tests
can be made from Figure 14. This figure is used to obtain an
estimate of the time required to collect a total of 25 milligrams
at any of several flow rates which might be used. Tests subse-
quent to the first should have sampling times adjusted such that
all stage loadings are kept below 10 milligrams.
Flow Rate Selection--
In high mass loading situations such as the inlet to control
devices, a low flow rate is preferable because it will permit a
reasonable long sampling time and process averaging. At the out-
let of the control devices, the mass loading may be moderate to
very low, and this situation will require the use of a high flow
rate to avoid excessively long sampling times. The two-stage im-
pactor with its three jet stage sets has the capability to handle
both of these situations. For the high- mass loading condition,
one would choose the set with the designed flow rate of 18.9
cm3/sec (0.04 cfm); and for the lower loadings, either the
47.2 cm3/sec (0.10 cfm) or the 94.4 cm3/sec (0.20 cfm) sets would
work well.
After the mass loading, sampling time, nozzle, and jet
stage set have been determined, the impactor can be assembled
as described above, and then attached to the probe.
Before sampling can begin, the orifice pressure drop must
be determined so that the flow rate can properly be set. The
following equation gives the pressure drop across the orifice
as read on the appropriate Magnehelic gauge:
fe)'
MA
where
23
-------
0.01
3 456789 10 2
Gas Velocity (ft./sec.)
4 5 6 7 8 9 100
Figure 13. Nomograph for selecting nozzles for isokinetic
sampling.
24
-------
Grain Loading (grains/acf)
0.001 2 345 6789 2 345 6789 2 3 4 56789
2 3 4 5 6789 2 345 6789
Selected Flow Rates (scfm)
Figure 14. Sampling time determination for total mass collection of 25 milligrams,
-------
MM = Mean molecular weight of flue gas
MA = Mean molecular weight of air
AP = Calibrated orifice pressure drop (from Figure 15) "H20
AP = Pressure drop at which orifice calibrated "H2O
Q = Impactor flow rate chosen for isokinetic sampling ACFM
Q = Calibration flow rate for orifice ACFM
Fu « = Volume fraction of water in the flue gas
112'-'
PS = Ambient stack pressure ps = pa + APS "H9
P = Pressure upstream of orifice referred to ambient "Hg
P = Ambient pressure when orifice calibrated "Hg
c
T = Temperature of the orifice OR
T = Stack temperature OR
This equation has been incorporated into a program for the
Hewlett-Packard HP-65 calculator. The program sheet included
in this report details the use of the program to compute the
pressure drop.
Notice that some of the input data are used with engineer-
ing units. This is for convenience in using the gauges and
meters. The end results from all calculations, however, are
in metric units.
A dry gas test meter can also be used to monitor impactor
flow rate, either alone or in conjunction with a calibrated
orifice. The following equation is applicable:
Ta Ps
Q = Q — —
wm ws T P
26
-------
1000
Orifice - Flow Rate
029-23.0 cmVsec
042-46.5 cm3/sec
059-86.5 crnVsec
-id ::••_!• .—0.059p
:::!.! I :;• ;->fr.i:-..-qzn
tirtti-mttti
1 10
AP, Pressure Drop (inches H20)
Figure 15. Sampling orifice calibration.
100
27
-------
Q = Flow rate indicated by the dry gas meter ACFM
T = Temperature of metered gas o_
3.
T = Flue gas temperature °R
p = Pressure upstream of the meter referred to ambient "Hg
ci
P = Ambient stack pressure "Hg
Fu ~ = Volume fraction of water in flue gas
H20
If condensable vapors are not desired in the collection,
and if the stack temperature is not high enough, auxiliary
heating may be needed.
Heating the Impactor—
If flue streams above approximately 175°C (350°F),
auxiliary heating is not usually required. The auxiliary
heating is accomplished by wrapping the impactor with heater
tape, which can be controlled by the temperature controller
in the sampling case. A thermocouple placed in the impactor
exit gas stream monitors the flue gas temperature immediately
after it passes through the impactor. This temperature is
needed for impactor cutpoint calculations. This thermocouple
is connected to the temperature controller. If the impactor
is wrapped with heater tape (also connected to the controller)
the exit gas temperature can be controlled by simply setting
the temperature controller to the desired temperature.
The impactor should be heated for at least 30 minutes
(either in the duct or by external heater tape) before beginning
sampling to insure that the entire impactor is at the desired
temperature.
Taking the Sample—
After the impactor has reached its operating temperature,
sampling can commence. If the impactor has been heated in the
stack, the nozzle can be turned upstream and the correct flow
rate quickly set. For short sampling times, typical of those
that are necessary at control device inlets, this is especially
important. If the impactor was heated outside the duct, some
time should be allowed for the impactor temperature to come to
an equilibrium with its new surroundings before the nozzle is
turned upstream and sampling is begun. The flow rate should be
maintained constant for the entire test to assure that the
cutpoints do not change.
28
-------
After the sample has been taken, the hose to the probe
should be pinched off and the nozzle turned downstream before
the impactor is removed from the duct. This procedure is
especially important where there is a negative duct pressure.
A negative duct pressure can cause a backflow through the
impactor which might draw condensed water into the impactor from
the probe and tubing. Therefore, care must be taken to insure
that no gas flow through the impactor takes place except when
sampling. It is also important to carefully remove the impactor
from the duct to prevent any scraping or jarring and dislocation
of particulate matter.
Disassembly of the Impactor
The post-test procedure is very important in obtaining
useful results. The crucial part of this procedure is to make
sure that the material on the collection substrates stays where
it originally impacted, and that all particles not on these sub-
strates are correctly cleaned onto the appropriate collection
stage. A pair of fine tweezers and a small brush are essential
in accomplishing this.
The first step in disassembly is to remove the inlet cap
and the middle part of the housing. This exposes the stack of
jet and collection plates. Remove the first jet plate to reveal
the first collection substrate. All of the collected material
above the first substrate should be brushed onto this substrate.
Cleaning the nozzle completely is important, especially if it is
a small bore nozzle. The inlet cap and both sides of the first
jet stage should be cleaned and all the material placed on the
first substrate. The first collection substrate can now be re-
moved from the collection plate and stored in a suitable container
to prevent any of the particulate matter from being lost. In
lieu of a container, a foil square could be used. Use of a foil
is a good method if there is a heavy loading of larger particles.
All of the particulate matter in the nozzle and inlet cap can be
brushed directly onto the foil, the collection substrate placed
on the foil, and then the foil folded to prevent loss of any of
the sample. This method though requires that the foil be pre-
weighed with the collection substrate.
All of the material on the second jet plate is brushed
onto the second collection substrate, and this substrate and
the backup filter handled in a manner similar to that of the
first substrate.
Once all the substrates have been removed and placed in their
containers (foil or whatever)/ they should be placed in a desic-
cator and stored there for at least 24 hours. This desiccation
brings the water content of the substrate to a level comparable
to their initial level. After this desiccation period, the
29
-------
substrates are weighed and the results recorded.
Cascade Impactor Data Reduction6
After an impactor run, it is necessary to obtain a particle
size distribution from the mass loadings on each stage. The
conditions at which the impactor was run determine the stage D
outpoints. Theoretical outpoints can be calculated by an
iterative solution of the following two equations:
yD3 P X(I)
= 1.43 x 10"
QT P C 472.0
p I o
and
C = 1 + „ 2L. ,,-.. 1.23 + 0.41 EXP IK-0.44 D50)/L x
[(-0.
where
D50 is the stage cutpoint (urn) ,
y = gas viscosity (poise) ,
D = stage jet diameter (cm) ,
C
P = local pressure at stage jet (atm) ,
p = particle density (gm/cm3),
Q_ = impactor flow rate (cfm) ,
P = ambient pressure at impactor inlet (atm) ,
C = Cunningham Correction Factor,
L = gas mean free path (cm) , and
X(I) = number of holes per stage.
It is preferable however, to calibrate the impactor to deter-
mine the DSQ'S, and in Figures 16 through 21, the calibration
curves for the jet stage sets are given. These results were de<-
termined with an ammonium fluorescein aerosol which has a density
of 1.35 gm/cm3. By using the data from these curves and the
theory presented by Ranz and Wong, 5 an equation can be developed
to calculate the 50% collection efficiency size, or D so , from the
30
-------
100
dP
0)
•H
U
•H
M-l
»W
H
§
•H
43
U
0)
H
U) H
H O
U
1.0 10.0
Particle Diameter, Micrometers
50.0 100.0
Figure 16.
Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 1 - Stage 1.
Calibration Dso: 5.5 ym, 94 cm3/sec, p = 1.35 gin/cm3
6.4 pm, 94 cm3/sec, p = 1.00 gm/cm3
-------
100
fi
0)
•H
O
•H
«H
-------
100
U)
>1
u
g
•H
U
•H
W
G
O
t-4
O
U
1.0 10.0
Particle Diameter, Micrometers
100.0
Figure 18.
Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 2 - Stage 1
Calibration Dso: 4.1 ym, 47 cm3/sec, p = 1.35 gm/cm3
4.8 vim, 47 cm3/sec, p = 1.00 gm/cm3
-------
loop-
90
80
70
O
0)
•H
O
•H
U-l
M-l
W
c
O
2 1
0)
H
O
O
0.1 10.0
Particle Diameter, Micrometers
Figure 19. Collection efficiency vs, particle diameter for the
PSC impactor: Plate Set 2 - Stage 2
Calibration D5o: 0.85 ymr 47 cm3/sec, p = 1.35 gm/cm3
0.99 um, 47 cm3/sec, p = i.QO gm/cm3
100.0
-------
0.1
1.0 10.0
Particle Diameter, Micrometers
100.0
Figure 20.
Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 3 - Stage 1
Calibration D50: 2.6 pm, 18.9 cra3/sec, p = 1.35 gin/cm*
3.0 urn, 18.9 cm3/sec, p = 1.00 gm/cm
-------
100 ,-T
OJ
Figure 21.
1.0 10.0
Particle Diameter, Micrometers
Collection efficiency vs. particle diameter for the
PSC impactor: Plate Set 3 - Stage 2
Calibration D5o: 1.05 ym, 18.9 cm3/sec, p = 1.35 gm/cm3
1.22 ym, 18.9 cm3/sec, p = 1.00 gm/cm3
100.0
-------
calibration curves for the range of sampling conditions which are
normally encountered in source testing. Ranz and Wong studied
the effects of various forces on a particle which cause it to
move through a gas stream onto a collection body. For inertial
impaction, these forces can be presented in a dimensionless
parameter ty given by
ij; = C FD 2/4.5 ?ryD 3
P P c
where C = Cummingham Correction Factor, dimensionless,
pp = Particle density, gm/cm3,
F = Gas flow rate through impactor, cm3/sec,
Dp = Particle diameter, cm,
y = Gas stream viscosity, gm/cm sec, and
DC = Diameter of jet, cm.
ty is the ratio of the stopping distance of a particle with
velocity of VQ = 4F/irDc3 to the jet diameter, D . By holding
all parameters in the equation constant except for D , as is
done in calibrating the impactor, a plot of collection efficiency
versus /iJT can be obtained. From this graph, the value of /ij7
at which the collection efficiency is 50% can be found. If this
value for /\|» so is used in the equation for ip, the D so or cut-
equation is obtained by solving for D so.
/4.
\
Dpso =
This equation will furnish the correct stage D so point for
sampling conditions different from those at calibration. Figures
22-24 show the laboratory calibration data presented in this form
for each of the three plate sets. These calibration curves have
been used to obtain values of /^ so for each stage. These con-
stants are given in Table 1.
Table 2 contains values for the viscosity of air at tem-
peratures between 10°C and 300°C and for gas water content by
volume from 0% to 10%.9 These values are to be used in the
HP-65 program for calculating the impactor stage D 50.
37
-------
0.1 0.2 0.3 0.4 0.5
0.6 0.7
0.9 1.0
Figure 22.
Collection Efficiency vs.
PSC Impactor - Plate Set 1
Constant Flowrate and Variable Particle Size
• Stage 1
A Stage 2
38
-------
<*>
u
e
0)
•H
O
•H
M-J
M-l
W
G
O
•rH
-p
O
(1)
O
O
Figure 23.
Collection Efficiency vs. /*F
PSC Impactor - Plate Set 2
Constant Flowrate and Variable Particle Size
• Stage 1
A Stage 2
39
-------
o
e
0)
•H
O
•H
W
C
O
•H
4J
O
0)
O
O
100
90
80
70
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Figure 24.
Collection Efficiency vs.
PSC Impactor - Plate Set 3
Constant Flowrate and Variable Particle Size
• Stage 1
A Stage 2
40
-------
TABLE 2. VISCOSITY OF AIR VS. H20 CONTENT
10°C - 300°C
H20
°c
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
0
1.767
1.810
1.854
1.900
1.938
1.979
2.020
2.059
2.099
2.137
2.175
2.213
2.250
2.286
2.321
2.356
2.390
2.424
2.457
2.489
2.521
2.552
2.583
2.613
2.642
2.671
2.699
2.727
2.754
2.780
1
1.758
1.801
1.844
1.887
1.929
1.970
2.011
2.050
2.090
2.129
2.167
2.204
2.241
2.277
2.313
2.348
2.382
2.416
2.449
2.482
2.513
2.545
2.575
2.606
2.635
2.664
2.692
2.720
2.747
2.773
2
1.748
1.792
1.835
1.878
1.920
1.961
2.002
2.042
2.081
2.120
2.158
2.195
2.232
2.269
2.304
2.339
2.374
2.408
2.441
2.474
2.506
2.537
2.568
2.598
2.628
2.657
2.685
2.713
2.740
2.767
3
1.739
1.783
1.826
1.869
1.911
1.952
1.993
2.033
2.072
2.111
2.149
2.189
2.224
2.260
2.296
2.331
2.366
2.400
2.433
2.466
2.498
2.530
2.560
2.591
2.621
2.650
2.678
2.706
2.734
2.761
4
1.730
1.774
1.817
1.860
1.902
1.943
1.984
2.024
2.063
2.102
2.140
2.178
2.215
2.252
2.288
2.323
2.358
2.392
2.425
2.458
2.490
2.522
2.553
2.583
2.613
2.632
2.671
2.700
2.727
2.754
5
1.721
1.765
1.808
1.850
1.892
1.934
1.975
2.015
2.054
2.093
2.132
2.169
2.207
2.243
2.279
2.315
2.349
2.383
2.417
2.450
2.482
2.514
2.545
2.576
2.606
2.636
2.663
2.693
2.720
2.748
6
1.712
1.755
1.799
1.841
1.883
1.925
1.966
2.006
2.046
2.085
2.123
2.161
2.198
2.235
2.271
2.306
2.341
2.375
2.409
3.442
2.475
2.507
2.538
2.569
2.599
2.628
2.657
2.686
2.714
2.741
7
1.702
1.746
1.790
1.832
1.874
1.916
1.957
1.997
2.037
2.076
2.114
2.152
2.189
2.226
2.262
2.298
2.333
2.367
2.401
2.434
2.467
2.499
2.530
2.561
2.592
2.621
2.650
2.679
2.707
2.734
8
1.693
1.737
1.780
1.823
1.865
1.90;
1.948
1.988
2.028
2.067
2.105
2.143
2.181
2.218
2.254
2.289
2.325
2.359
2.393
2.426
2.459
2.491
2.523
2.554
2.584
2.614
2.643
2.b72
2.700
2.728
9
1.684
1.728
1.771
1.814
1.856
1.898
1.939
1.979
2.019
2.058
2.097
2.135
2.172
2.209
2.245
2.2bJ
2.311)
2.351
2.385
2.418
2.451
2.483
2.515
2.546
2.577
2.607
2.l.3i.
2.1,65
2.694
2.721
10
1.675
1.719
1.762
1.805
1.847
1.888
1.930
1.970
2.010
2.049
2.088
2.126
2 . 1 1, 4
2.201
2.237
2. 2 73
2. JOB
2 . 3 4 3
2.377
2.410
2.443
2.47t,
2. 507
2.5.M
2. b70
2.600
2 . (, 2 9
2 . u 5 8
2.b87
2.715
X 10"" poise
-------
The most feasible way to calculate these cupoints is to
write a computer program. Otherwise, a slow and tedious process
results. The size parameter reported can be either aerodynamic
diameter, that is, diameter based on the behavior of unit
density particles, or approximate physical diameter, based on
the estimate of the true particle density. In either case,
the particles are assumed to be spherical.
A program for the HP-65 has been written to calculate the
D 50 cutpoints for each of the jet plate sets. The program sheet
gives a breakdown of the parameters needed to calculate the
cutpoints. Once the D 50 cutpoints have been calculated, the
particle size distribution may be presented on a differential
or cumulative basis.
Differential Particle Size Distributions—
For the purpose of analysis, the assumption is made that
all of the mass caught on an impaction stage consists of material
having diameters equal to, or greater than the D 50 for that
stage. For the first stage, it is assumed that all the material
caught has diameters greater than, or equal to, the D &0 for
that stage, but less than the largest particle size that has
been sampled.
Because the intervals between the stage D so's are loga-
rithmically related, and to minimize scaling problems, the
differential particle size distributions are plotted on log-log
or semilog paper with
M
n
(log D)
as the ordinate and log D as the abscissa. (D &Q is the
geometric mean of DI and D2fD = /DiD2.) The mass on stage
"n" is designated by Mn. The9S (log D) associated with Mn is
log (Dso)n+1 - log (D5o)n- The total mass having diameters
between (Dso),,, and (D50)M is equal to the area under the curve;
i.e.,
m
n
n
or
M.
Mass =
m
(D
so
D
r
d.M
d (log D)
42
d dog D)
-------
for a near continuum.
The procedure outlined above is used to construct a
histogram. In practice, a smooth curve is frequently drawn
through the points, yielding an approximation to the real particle
size distribution. Such a curve is needed to calculate fractional
efficiencies of control devices if the D50's differ between
inlet and outlet measurements. The accuracy of the approximation
is limited by the number of points, and by the basic inaccuracy
of neglecting the nonideal behavior of the impactors, especially
overlapping collection efficiencies for adjacent stages.
Cumulative Particle Size Distributions—
The data may be presented on a cumulative basis by summing
the mass on all the collection stages and back-up filter, and
plotting the fraction of the mass below a given size, versus
size. This is frequently done on special log-probability paper.
This paper is especially convenient for log normal distributions,
but semi-log paper may be preferable for interpretation, especial-
ly if the distribution is not log normal. In general, cumulative
distributions are more difficult to interpret than differential
plots. The abscissa is the logarithm of the particle diameter,
and the ordinate is the percentage smaller than this size. The
value of the ordinate at a given (D so) would be
t = n
t
Percent less than stated size = - j - x 100%
t = o
or t =n
EM
t = o
43
-------
where
$
t = o corresponds to the filter,
t = n corresponds to the stage under consideration, and
t = N corresponds to the coarsest jet or cyclone.
Alternately, an analytical curve might be fitted to the
cumulative distribution obtained above, and values of dM/d(logD)
obtained by differentiation of the analytical expression:
A program for the HP-65 has been written to calculate
the cumulative particle size distribution. This is given in
the program sheet.
Sample Calculations with the HP-65 Programs
Sampling Orifice Pressure Drop Calculation—
This program calculates the pressure drop across a par-
ticular sampling orifice for a desired impactor flow rate. The
following data are required:
1. Desired sampling flow rate (cm3/sec).
2. Calibration flow rate at 10" HaO (cm3/sec).
3. Volume fraction of water in the flue gas.
4. Ambient pressure at the impactor inlet ("Hg).
5. Ambient pressure when orifice was calibrated ("Hg).
6. Orifice temperature when sampling (°C).
7. Impactor gas exit temperature (°C).
8. Orifice temperature when orifice was calibrated (°C).
The calibration information for the orifice is found on
the orifice calibration sheet. The other data are for the
particular stack conditions under consideration.
For the following sample calculation, the orifice in-
formation has been taken from the calibration sheet and the
stack conditions assumed (for a typical source). They are
listed below:
1. Desired flow rate =50 cm/sec,
44
-------
2. Calibration flow rate = 46.5 cm/sec.
3. Volume fraction of water = 0.10.
4. Ambient pressure at impactor inlet =28.5 "Hg.
5. Ambient pressure at calibration = 29.34 "Hg.
6. Orifice temperature when sampling = 29°C.
7. Gas exit temperature = 200°C.
8. Orifice temperature when calibrated = 24.5°C.
When these conditions are inputed to the calculator in
the sequence indicated on the user instruction sheet and the
program is run, a value for the orifice pressure drop will appear
in the x-register. For the above conditions, the orifice pressure
drop is
AP = 3.65" H20
Particle Size Distribution Calculation—
This program calculates the cumulative mass loading for the
two outpoints and the total mass loading in both actual cubic
meters and standard dry cubic meters. Also calculated is the
cummulative percent for the stages. Necessary inputs are as
follows:
1. Collected mass on first stage (milligrams).
2. Collected mass on second stage (milligrams).
3. Collected mass on filter (milligrams).
4. Sampling flow rate (cm3/sec).
5. Sampling duration (minutes).
6-^ Gas temperature (°C) .
7. Ambient pressure at impactor inlet ("Hg).
8. Volume fraction of water.
Assuming typical field test data (given below) and in-
puting the data into the program, the particle size distri-
bution can be found.
1. Mass on first stage =12.0 mg.
45
-------
2. Mass on second stage = 7.0 mg.
3. Mass on filter = 8.5 mg.
4. Flow rate = 50 cm3/sec.
5. Sampling duration = 90 min.
6. Gas temperature = 200°C.
7. Impactor ambient pressure = 28.5 "Hg.
8. Volume fraction of water = 0.10.
Total mass loading = 101.9 mg/ACM = 190.5 mg/DNCM*
Cumulative mass loading to first stage cutpoint = 57.4 mg/ACM =
107.4 mg/DNCM.
Cumulative mass loading to second stage cutpoint = 31.5 mg/ACM =
58.9 mg/DNCM.
Cumulative percent to first stage = 56.4%.
Cumulative percent to second stage = 30.9%.
Impactor Stage Cutpoint Calculation—
This program calculates the cutpoint for a particular jet
stage for a given impactor flow rate and flue conditions. The
following data are used:
1. Flue gas viscosity (poise).
2. Ambient pressure at impactor inlet ("Hg).
3. Impactor exit gas temperature (°C),
4. Jet diameter (cm).
5. Number of jets.
6. Particle density (gm/cm3).
7. Impactor flow rate (cm3/sec).
8. Jet plate constant.
*Dry Normal Cubic Meter - 760,mmHg, 21.1°C, dry.
46
-------
The flue gas viscosity can be taken from the table of
viscosities given. The jet plate information (jet diameter and
number of jets) is found in Table I. The jet plate constant is
located on the HP-65 program form.
Selecting, for example, the jet plate designed for a
nominal cutpoint of 5 microns at a flow rate of 47 cm3/sec,
typical values for the program are as follows (assuming typical
stack conditions):
1. Flue gas viscosity = 2.225 x 10 ** poise.
2. Ambient pressure at impactor inlet = 30.42 "Hg.
3. Exit gas temperature = 146°C.
4. Jet diameter = 0.167 cm.
5. Number of jets = 10.
6. Particle density = 1.00 gm/cm3.
7. Impactor flow rate = 57 cm3/sec.
8. Jet plate constant = 1.54 x 108.
For these values the cutpoint for this jet stage is
D so = 5.18 micrometers.
47
-------
Field Testing of the PSC Impactor
Bull Run Steam Plant—
In March, 1975 the first field test of the PSC impactor
took place at the TVA Bull Run Steam Plant in Oak Ridge,
Tennessee. The tests were conducted on the outlet duct of
Precipitator A about 20 feet upstream from the stack. Con-
currently with each test a Brink Cascade Impactor was run to
obtain a comparative size distribution. The mass loading at
this point was approximately 2.3 grams per actual cubic meter
(1.0 grain per actual cubic foot) which is high for a coal-
fired boiler precipitator outlet. The high mass loading
limited the sampling time of the impactor considerably.
All three jet sets of the PSC impactor were tested. In
these tests, jet plate set I was run at 94 cm3/sec (0.2 cfm),
jet plate set II at 47 cm3/sec (0.1 cfm), and jet plate set III
at 193/sec (0.04 cfm). A 2.5 mm nozzle was required in order
to obtain isokinetic sampling for the 94 cm3/sec flow rate, a
1.75 mm nozzle for the 47 cm3/sec, and a 1.0 mm nozzle for the
19 cm3/sec. However, the smallest nozzles were 2.5mm for both
the two-stage impactor and the Brink Cascade Impactor which was
run at 14 cm3/sec. Thus, the two-stage impactor with jet plate
set I sampled isokinetically while the other two jet plate sets
and the Brink sampled anisokinetically.
The tests were performed on three days. On the first day
jet plate set I was tested, on the second day plate sets II and
III, and on the third day plate sets I and III. A Brink Cascade
Impactor was run each day in the same port and at the same depth.
Table 3 shows the pertinent data for each test including the
stage weight gains. During the first two days Teflon collection
substrates were used and on the last day Gelman Type A Glass
Fiber substrates were used. From visual inspection, each sub-
strate appeared to have good particle retention qualities but
the Teflon was easier to remove after sampling without collected
dust being dislodged. Sampling times were undesirably short due
to the high particulate grain loading.
Figures 25, 26, and 27 show the results of this prelim-
inary test for each day of testing. In Figure 25 the curve for
jet plate set I is lower in total loading than the Brink. This is
expected since the Brink sampled anisokinetically and collected
a greater than normal quantity of large particles. The plate
set I test sampled isokinetically. The Brink test should be .re-
liable below 2 or 3 microns however and it can be seen that the
plate set I curve is above the Brink indicating possible bounce
and reentrainment. In Figure 26 agreement in total loading be-
tween the Brink and plate set III is seen when both types of im-
pactor were operated anisokinetically at very nearly the same
48
-------
TABLE 3. TWO-STAGE IMPACTOR TEST DATA
DATE
PORT
"Hg Amb. Pres .
"H20 Stack
Pres.
m/sec Gas Velo-
city
"H20 Orifice AP
in. Orifice
I.D.
11 Hg Imp. AP
°C Imp . Temp .
°C G.M. or
Orifice
cm3/ Flow rate
sec
% H20
Start Time
Min . Duration
mm Nozzle
Impactor
Plate Set
STAGE
WEIGHT GAIN
mg 1
mg 2
mg Filter
3/11/75
2
29.31
-1.6
20
5.7
.059
0.7
143
13
99
7.5
4:15
3
2.5
I
7.78
8.16
2.90
3/12/75
2
28.99
-1.2
20
5.3
.042
0.4
138
19
47
7.5
12:45
6
2.5
II
12.46
5.46
3.42
49
3/12/75
2
38.99
-1.2
20
3.8
.029
0.2
160
17
20
7.5
3:15
15
2.5
III
33.18
7.66
5.28
3/13/75
2
29.00
-1.0
20
3.8
.029
0.2
149
18
20
7.5
11:50
10
2.5
III
63.14
14.26
9.54
3/13/75
2
29.09
-1.0
20
5.7
.059
0.7
138
18
98
7.5
2:50
3
2.5
I
27.34
20.34
7.62
-------
10'
<
O
a
>
<
s
o
10-
10'
10
=183
mm
A TWO-STAGE IMPACTOR PLATE SET I 94 cm3/sec 7*
TEFLON SUBSTRATES
O BRINK CASCADE IMPACTOR - 14 cm3/sec
fi HI.
10-
UPPER SIZE LIMIT ( micrometen )
Figure 25. Cumulative Mass Loading versus Particle Diameter
March 11, 1975
50
-------
"I
o
5
<
o
2
ui
S
u
• TWO-STAGE IMPACTOR PLATE SET II 47 cm3/sec
TEFLON SUBSTRATES
• TWO-STAGE IMPACTOR - PLATE SET III - 19 crn3/sec
TEFLON SUBSTRATES
O BRINK CASCADE IMPACTOR -14 cm3/sec
i—[~z:=~=
10'
UPPER SIZE LIMIT ( micrometer* )
Figure 26. Cumulative Mass Loading versus Particle Diameter
March 12, 1975
51
-------
10'
<
o
I
S
o
10 3
10'
10
A TWO-STAGE IMPACTOR - PLATE SET I - 94 cm3/sec
GLASS FIBER SUBSTRATES
• TWO-STAGE IMPACTOR PLATE SET III -19 cm3/sec
GLASS FIBER SUBSTRATES
O BRINK CASCADE IMPACTOR - 14cm3/sec
-t, j_4.uu;.:J *-i-4
Trffis; *~ ^' -
rt:
10'
10
10'
UPPER SIZE LIMIT ( micrometers )
Figure 27. Cumulative Mass Loading versus Particle Diameter
March 13, 1975
52
-------
conditions. The curve for plate set II is lower because this set
was used nearer isokinetic conditions. In both cases the sub-
micron end of the cumulative size distribution is higher than the
Brink, perhaps indicating passage of particles to lower stages
which should have remained on an earlier stage. Figure 27 shows
results similar to Figure 26.
Gorgas Steam Plant—
On June 25-56, 1975, a test of the two-stage impactor was
performed at Alabama Power's Gorgas Steam Plant. Sampling was
conducted at the outlet duct of the Unit 10 coal boiler, downstream
of the precipitator. Each of the three pairs of jet plates were
used during this series of tests. An Andersen Stack Sampler was
run simultaneously with the two-stage impactor to obtain a com-
parable size distribution. Glass fiber substrates were used in
both impactors for all tests. The two-stage impactor and Andersen
were located at the same vertical level in the duct three feet
apart, but were offset in depth to avoid any interference. The
two-stage impactor pitot was used to measure the gas velocity in
the duct which was 21 m/sec. On the first test day the two-stage
impactor with jet plate set I was run for two hours at a flow rate
of 94 cm3/sec and with jet plate set II for three hours at
47 cm3/sec. Simultaneously an Andersen Stack Sampler was run at
236 cm3/sec for five hours. On the second day the two-stage im-
pactor with jet plate set III was run at 19 cm3/sec for five hours
and an Andersen Stack Sampler was also run for five hours at
236 cm3/sec. On the first day the gas temperature was 166°C while
on the second day it was 177°C. In all cases each impactor was
allowed to warm up for 45 minutes before sampling was initiated.
Appropriate nozzles were chosen to obtain as nearly isokinetic
sampling as possible.
All sampling data are shown in Tables 4, 5, and 6. Figure
28 shows the two-stage impactor and Andersen Impactor data pre-
sented on a cumulative mass basis. The discrepancies in total
grain loading are not fully understood but may be partly due to
sulfate absorption by the Andersen and two-stage impactor sub-
strates.7 Sulfate determinations were run on all substrates used
in these tests. Significant amounts of sulfate were found on the
Andersen substrates and some was found on the two-stage impactor
substrates as shown in Tables 4, 5, and 6. Results shown in
Figure 28 are based on substrate weights gained corrected for
sulfate weight gains. The flatness of the two-stage impactor
curves may be due to either bounce or reentrainment of particles
which passed to the filter.
On July 30-31, 1975 a second test of the two-stage impactor
took place at Gorgas Steam Plant. Sampling was done on the out-
let of Unit 10 again and all three jet plate sets were used. An
Andersen Stack Sampler was run simultaneously with the two-stage
53
-------
TABLE 4. TWO-STAGE IMPACTOR-JET PLATE SET I
Nominal Flowrate
Stage
Total Weight
Gain (mg)
1st 0.90
2nd 0.98
Back-up 5.05
Filter
94 cm3/sec
June 25, 1975
SOX (mg)
0.06
0.10
1.14
Corrected Weight Cumulative
Gain (mg) Loading (mg/ACM)
0.84
0.88
3.91
TOTAL 8.28
7.05
5.75
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
94.3 cm3/sec
29.5 "Hg
21.3 m/sec
0.059"
7.0" H20
8.0" H20
41°C
166°C
11:15 A.M.
1:15 P.M.
120 minutes
2.5 mm
-2.0 "H20
10%
Calculated
Calculated
1st Stage D50 - 4.24 ym
2nd Stage Dso - 1.14 ym
54
-------
TABLE 5. TWO-STAGE IMPACTOR-JET PLATE SET II
Nominal Flowrate
47 cm3/sec
June 25, 1975
Total Weight
Stage Gain (mg)
1st
2nd
Back-up
Filter
1.66
2.46
4.86
SO (mg)
0.07
0.10
1.98
Corrected Weight Cumulative
Gain (mg) Loading (mg/ACM)
1.59
2.36
2.88
TOTAL 13.40
10.28
5.65
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
45.1 cm3/sec
29.5 "Hg
21.3 m/sec
0.042"
4" H20
8" H2O
41°C
166°C
2:30 P.M.
5:30 P.M.
180 minutes
1.5 mm
-2.0" H20
10%
Calculated
Calculated
1st Stage Dso - 3.91
2nd Stage D50 - 0.82 ym
55
-------
TABLE 6. TWO-STAGE IMPACTOR-JET PLATE SET III
Nominal Flowrate
19 cm3/sec
June 26, 1975
Stage
1st
2nd
Filter
Total Weight
Gain(mg)
1.83
0.65
3.88
Corrected Weight
S0x(mg) Gain(mg)
0.07 1.76
0.23 0.42
1.76 2.12
Cumulative Loading
(mg/ACM)
TOTAL
12.65
7.47
6.24
SAMPLING DATA
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
20.0 cm3/sec
29.5" Hg
21.3 m/sec
0.029"
3" H20
6" H20
41°C
177°C
11:00 A.M.
4:00 P.M.
300 minutes
1 mm
-2" H20
10%
Calculated
Calculated
1st Stage D50 - 1.83 ym
2nd Stage D50 - 0.78 ym
56
-------
103
(9
5
iu
I
i
o
102
101
100
O ANDERSEN STACK SAMPLER -JUNE 25,1975
D ANDERSEN STACK SAMPLER • JUNE 26. 1975
A TWO-STAGE IMPACTOR - JET PLATE SET I - JUNE 25.1975
• TWO-STAGE IMPACTOR - JET PLATE SET II - JUNE 25.1975
• TWO-STAGE IMPACTOR - JET PLATE SET III JUNE 26.1975
10-'
10° 101
UPPER SIZE LIMIT ( micrometer* )
10'
Figure 28. Cumulative grain loading versus Particle Size
57
-------
impactor to obtain a comparable size distribution. In this
series of tests, preconditioned substrates were used in both
impactors to lessen the magnitude of the SOa interference. The
substrates were conditioned by exposure to filtered flue gas
for five hours before desiccation and initial weighing.
The two-stage impactor and Andersen were located at the
same vertical level in the duct, three feet apart, but were
offset in depth to avoid any interference. The two-stage im-
pactor pitot was used to measure the gas velocity which was
21.3 m/sec. On the first test day the two-stage impactor with
jet plate set III was run at a flow rate of 19 cm3/sec for five
hours and an Andersen Stack Sampler was also run for five hours
at 236 cm3/sec. On the second day the two-stage impactor with
jet plate set I was run for two hours at 94 cmvsec and with jet
plate set II for three hours at 47 cm3/sec. Simultaneously an
Andersen Stack Sampler was again run at 236 cm3/sec for five
hours. On both days uie flue gas temperature was 160°C. In all
cases the impactors were allowed to warm up 45 minutes before
sampling began. Appropriate nozzles were chosen to obtain as
nearly isokinetic sampling as possible.
All two-stage impactor sampling data are shown in Tables
7, 8, and 9. Figure 29 shows the two-stage impactor and
Andersen Stack Sampler data presented on a cumulative mass basis.
Reasonable agreement can be seen between the two-stage
impactor and Andersen impactor data. This is in contrast to the
poor agreement between the two impactor data presented for the
June 25-26 test. The preconditioning of the substrates has aided
in this agreement and it appears that reliable information with
this two-stage impactor is attainable when care is taken in all
aspects of the sampling, including preconditioning the collection
substrates.
58
-------
TABLE 7. TWO-STAGE IMPACTOR — JET PLATE SET I
Nominal Flowrate — 94 cm3/sec
July 31, 1975
Stage
1st
2nd
Back-up
Filter
Weight
Gain(mg)
0.42
2.46
2.17
Cumulative Loading
(mg/ACM)
TOTAL
5.83
5.34
2.50
Calculated Stage
D5o
3.69
0.99
SAMPLING DATA
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
120.36 cra3/sec
29.3" Hg
21.3 m/sec
0.059"
7.94" H20
19" H20
21°C
160°C
2:20 p.m.
4:20 p.m.
120 minutes
2.5 mm
-1.8" H20
10%
59
-------
TABLE 8. TWO-STAGE IMPACTOR — JET PLATE SET II
Nominal Flowrate — 47 cm3/sec
July 31, 1975
Stage
1st
2nd
Back-up
Filter
Weight
Gain(mg)
0.45
1.53
2.59
Cumulative Loading
(mg/ACM)
TOTAL
9.57
8.60
5.42
Calculated Stage
D5o
3.91
0.82
SAMPLING DATA
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
44.23 cm3/sec
29.3" Hg
21.3 m/sec
0.042"
4.2" H20
8.0" H20
21°C
160°C
10:10 a.m.
1:10 p.m.
180 minutes
1.5 mm
-1.8" H20
10%
60
-------
TABLE 9. TWO-STAGE IMPACTOR — JET PLATE SET III
Nominal Flowrate — 19 cm3/sec
Stage
1st
2nd
Back-up
Filter
Weight
Gain(mg)
0.93
0.83
1.41
Cumulative Loading
(mg/ACM)
TOTAL
8.48
5.99
3.77
July 30, 1975
Calculated Stage
D50 (mg)
1.75
0.75
SAMPLING DATA
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Metering Orifice Temperature
Sampling System AP
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
20.77 cm3/sec
29.3" Hg
21.3 m/sec
0.029"
3.73" H20
27°C
6^ H20
160°C
11:35 a.m.
4:35 p.m.
300 minutes
1.0 mm
-1.8" H20
10%
61
-------
100.0
^
to
C
•H
•s
(0
D)
0)
•H
•P
1.0
10.0 illlKllNlUii;
10.0
Particle Diameter, ym
Figure 29. Cumulative Grain Loading versus Particle Size
o Andersen Stack Sampler - July 30, 1975
D Andersen Stack Sampler - July 31, 1975
• Two-Stage Impactor - Jet Plate Set I - July 31, 1975
• Two-Stage Impactor - Jet Plate Set II - July 31, 1975
A Two-Stage Impactor - Jet Plate Set III - July 30, 1975
100.0
-------
Oil-Fired Power Plant—
On January 28-29, 1976 tests of the two-stage impactor were
made at an oil-fired power plant. No pollution control device
was installed at this facility. Two adjacent ports were used
for sampling. Each of the three sets of jet plates were run at
least once during the testing period. Glass fiber collection
substrates were used in all impactor tests. An Andersen Stack
Sampler was run at approximately the same time intervals to
obtain a comparable size distribution. The collection substrates
for both the two-stage and the Andersen impactors were precon-
ditioned instack by exposure to filtered flue gas. Blank runs
for the Andersen were made and collection weights accordingly
adjusted for SO weight gains.
J\.
Ultrafine particle measurements were also done as part
of these tests. These measurements were made with Thermo-
Systems1 Electrical Aerosol Size Analyzer (EAA). A dilution
system developed at SRI was used to lower the particle concen-
tration to a level suitable for the EAA. This device is used
to obtain a particle size distribution over the 0.013 to 0.31
micrometer range, which is just below that of the inertial
impactor.
The power plant was operated at three load conditions
during the testing period: full load, half load, and maximum
load. The first day of the test was at full load, the morning
of the second day was at maximum load and the afternoon of the
second day was at half load.
Two measurements were made with the Andersen Stack
Sampler at the full load condition. The PSC impactor was used
with plate sets I and II at the full load condition. Two
Andersen tests were also made at the half load condition, and
only plate set III of the PSC impactor was used at this load.
For the maximum load condition, only the PSC impactor with
plate set I was run and because of a miscalculation, an aniso-
kinetic sample was taken. Large particles were probably over
sampled during this test. The results of this test are plotted
on the" graphs for the full load condition. Again, the impactors
were allowed at least 45 minutes warm-up time before sampling
was begun.
All sampling data for the PSC impactor runs, including
flowrates, stack temperatures, and sampling time, are shown in
Tables 10, 11, 12, and 13. Figures 30 and 31 show the cumulative
particle size distributions for the full and half load conditions.
Figures 32 and 33 show the differential particle size distribution
on a mass basis and Figures 34 and 35 show the differential size
distributions on a numerical basis. Figure 36 is a cumulative
63
-------
TABLE 10. TWO-STAGE IMPACTOR — JET PLATE SET I
Nominal Flowrate
Stage
1st
2nd
Back-up
Filter
Weight
Gain (mg)
10.66
0.82
3.79
94 cm3/sec
Cumulative Loading
(mg/ACM)
TOTAL: 139.84
42.22
34.71
January 28, 1976
Calculated Stage
D50 (ym)
6.39
1.77
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H2O by Volume
91.0 cra3/sec
30.2" Hg
18.3 m/sec
0.059"
5" H20
1.25" Hg
21°C
138°C
11:18 a.m.
11:38 a.m.
20 min.
2.5 mm
+3.0 "H2O
10.5%
64
-------
TABLE 11. TWO-STAGE IMPACTOR — JET PLATE SET II
Nominal Flowrate
47 cm3/sec
January 28, 1976
Stage
1st
2nd
Back-up
Filter
Weight
Gain (mg)
4.75
0.63
2.88
Cumulative Loading
(mg/ACM)
TOTAL: 48.47
20.60
16.90
Calculated Stage
Dso
5.19
1.12
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H2O by Volume
56.8 cm3/sec
30.2" Hg
18.3 m/sec
0.042"
6.3" H20
0.37" Hg
216C
147°C
3:25 p.m.
4:15 p.m.
50 min.
2 mm
+ 3.0" H O
10.5%
65
-------
TABLE 12. TWO-STAGE IMPACTOR — JET PLATE SET I
Nominal Flowrate
Stage
1st
2nd
Back-up
Filter
Weight
Gain (mg)
4.51
0.61
1.55
94 cm3/sec
Cumulative Loading
(mg/ACM)
TOTAL: 106.08
34.35
24.65
January 29, 1976
Calculated Stage
Dso
8.52
2.39
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
52.4 cm3/sec
30.1" Hg
18.3 m/sec
0.042"
5" H20
0.22" Hg
7.2°C
149°C
9:29 a.m.
9:49 a.m.
20 min.
2.5 mm
+3.0" H20
10.5%
66
-------
TABLE 13. TWO-STAGE IMPACTOR — JET PLATE SET III
Nominal Flowrate
Weight
Stage Gain (mg)
1st
2nd
Back-up
Filter
2.47
0.22
3.03
19 cm3/sec
Cumulative Loading
(mg/ACM)
TOTAL:
45.94
26.10
24.34
January 29, 1976
Calculated Stage
Dso
3.48
1.55
Sampling Data
Calculated Flowrate
Ambient Pressure
Gas Velocity
Metering Orifice
Metering Orifice AP
Sampling System AP
Metering Orifice Temperature
Impactor Temperature
Start Time
End Time
Duration
Nozzle
Stack Pressure
Flue Gas % H20 by Volume
12.5 cm3/sec
30.1" Hg
8.5 m/sec
0.029"
1.2" H20
0.15" Hg
206C
143°C
12:06 p.m.
2:52 p.m.
166 min.
1 mm
0
10.5%
67
-------
1000
.
HieSMtemilUnUIHHIIIILiifllHIIIIIIIIIlii
••lfllllllllllll!l!!ll!h:
OY
00
o>
CJ
Z
a
§
ID
O
100
10
- ANDERSEN .•
!!- ANDERSEN i"
- PSC - PLATE SET I H1
- PSC - PLATE SET ll!i
- PSC - PLATE SET I :..
IUU1MI (MAXIMUM LOAD; ANISOKINETIC)
- ELECTRICAL AEROSOL ANALYZER
0.01
0.1
1.0 10
PARTICLE SIZE (MICROMETERS)
100
1000
Figure 30. Cumulative Particle Size Distribution for Full Load
January 28, 1976
-------
1000
k«»!iiiniiiiii!i
so
••MiliiiuiiiiiiiiiiirtiuriiiHiiMiiiiiwrv Bi«!ni:i!in!hiii;!!v:i iiimmi1! J,, ,: ••••ittiiininiim
HUim
(9
I
- ANDERSEN
- ANDERSEN
1-PSC -PLATE SET III
- ELECTRICAL AEROSOL ANALYZER
aoi
1.0 10
PARTICLE SIZE (MICROMETERS)
100
1000
Figure 31.
Cumulative Particle Size Distribution for Half
Load, January 29, 1976
-------
1000
- ANDERSEN
- ANDERSEN
- PSC - PLATE SET I
- PSC - PLATE SET Ilij
-PSC - PLATE SET I
(MAXIMUM LOAD;ANISOKINETIC)
- ELECTRICAL AEROSOL ANALYZER -
100
1000
PARTICLE SIZE (MICROMETERS)
Figure 32. Differential Particle Size Distribution on Mass Basis
for Full Load, January 28, 1976
-------
[I I'llHIIIHIinilHIIHIWIIIHilHIUUIIIDIIIItl'
- ANDERSEN
- ANDERSEN
-SET ml!!
- ELECTRICAL AEROSOL ANALYZER
Figure 33,
1 10
PARTICLE SIZE (MICROMETERS)
Differential Particle Size Distribution on Mass Basis
for Half Load, January 29, 1976
100
1000
-------
K>
K>
10
10
A - J
I I I I I I I I I I I I M I I I I I
A
A
o
I E *
«
-. o Andtrsm
10 h Q Andtrsm
• P8C-Plot«S*t I
• P8C-PlattS«t II
A P8C-Plate Stt I (Maximum Load)
A Eltctricol Aerosol Anolyztr
i i I i 1 1 1 1 _ L i I I i 1 1 1 1 i i i I 1 1 1 1 1
a -
i i
0.01 o.i 1.0 10 too
Portielfl Sbt (Microm«ttr«)
Figure 34. Differential Particle Size Distribution on Numerical
Basis for Full Load, January 28, 1976
72
-------
K)M
IO
K
0
09
_o
i
.o
6
1 I I | "I M| 1 ' I | I llf[ 1 1 I | I III
A A
A
O
Q
O
Q
0 Q
O
Q o
* O
^ 0°
o
10
o Andersen
a Andersen
^ PSC-Plate Set III
A Electrical Aerosol Analyzer ° _I
i i I 11 n I i i I I I n li i i i I M ii I i ill
0.01 O.I - 1.0 10.0 100
Particle Size (Micrometers)
Figure 35. Differential Particle Size Distribution on
Numerical Basis for Half Load, January 29, 1976
73
-------
0.1 1.0
PARTICLE SIZE (MICROMETERS)
Figure 36. Cumulative Particle Size Distribution for Ultrafine
Region Using the Electrical Aerosol Analyzer,
January 28-29, 1976
10
74
-------
size distribution of the ultrafine particles for all three load
conditions as seen by the Electrical Aerosol Analyzer. There was
no detectable changes in the ultra-fine particle size distribu-
tion with the plant load condition.
The differential size distributions for the PSC impactor
were derived from the cumulative distributions. For accurate
comparison with the Andersen runs, intervals from the PSC cumula-
tive curves were chosen such that they were approximately the
same as those on the Andersen curves. The PSC differential curves
were then determined from these cumulative loadings.
There is reasonable data agreement between the Andersen
Stack Sampler and the PSC impactor. Data from both impactors
were in good alignment with the EAA ultrafine data.
From the results of this field test and the previous two
tests, it is seen that good agreement of the PSC impactor with
commercial instruments can be obtained. The only restrictions
are those that apply to the commercial instruments also—that the
impactor is carefully used as it was designed to be. This means
that isokinetic sampling is a must; the correct nozzle must be
used; the flow rate must be accurate and constant; and all other
basic rules of stack sampling must be adhered to. Only when these
rules are closely and carefully followed, can good, reliable,
particle-size data be obtained.
75
-------
REFERENCES
1. Berglund, R. N. and Liu, B. Y. H., "Generation of Mono-
disperse Aerosol Standards," Environmental Science and
Technology, Vol. 6, No. 2, 1973.
2. Lindblad, N. R. and Schneider, J. M., "Production of Uni-
form-Sized Liquid Droplets," J. Sci. Instru., Vol. 42, 1965.
3. Strom, L., "The Generation of Monodisperse Aerosols by
Means of a Disintearc-.ted Jet of Liquid," Rev. Sci. Instr. ,
Vol. 40, No. 6, 19o9.
4. Stober, W. and Flachsbart, H., "An Evaluation of Ammonium
Fluorescein as a Laboratory Aerosol," Atmos. Environ.,
Vol. 7, 1973.
5. Ranz, W. E. and Wong, J. B., "Impaction of Dust and Smoke
Particles," Ind. and Eng. Chem., Vol. 44, No. 6 (June, 1952)
6. Smith, W. B., K. M. Gushing, and J. D. McCain. Particulate
Sizing Techniques for Control Device Evaluation, Special
Summary Report to the Environmental Protection Agency,
Report No. 21, July 12, 1974.
7. Forrest, J. and L. Newman. "Sampling and Analysis of
Atmospheric Sulfur Compounds for Isotope Ratio Studies,"
Atmospheric Environment, Vol. 7, 1973.
8. Dwyer Co., Inc., Bulletin No. H-ll.
9. Wilke, C. R., "A Viscosity Equation for Gas Mixtures,"
Journal of Chemical Physics, Vol. 18, No. 4 (April, 1950).
10. May, K. R., "The Cascade Impactor: An Instrument for
Sampling Coarse Aerosols," J. Sci. Instr., 22 (October,
1945).
11. Pilat, M. J., Ensor, D. S., and Busch, J. C., "Cascade
Impactor for Sizing Particulates in Emission Sources,"
Am. Ind. Hygiene Assoc. J., Vol. 32, No. 8 (August, 1971).
12. Cohen, J. J. and Montan, D. N., "Theoretical Considerations,
Design, and Evaluation of a Cascade Impactor," Am. Ind.
Hygiene Assc :. J., (March-April, 1967).
13. Gelman, C. and Marshall, J. C. , "Hiv.-h Purity Fibrous Air
Sampling Media," Industrial Hygiene Association Annual
Meeting, Thursday, May 16, 1974, Miami, Florida.
.76
-------
APPENDIX
CALCULATOR PROGRAMS FOR THE PSC IMPACTORS
77
-------
T.,lf>
HP-65 User Instructions
Sampling^ Orifice Pressure Drop Calculation— Two-Stage Impactor
PRESSU
RE DRO
' '
STEP
1
2
3
4
5
6
7
8
9
10
11
12;
13
14
15
16!
17
18
INSTRUCTIONS
Input desired sampling f lowrate ._
Store desired sampling^ f lowrate..
Input calibration f lowrate at 10" H20. (
Store calibration f lowrate.
Input flue gas volume fraction of water.
Store flue gas volume fraction of water.
Input ambient pressure at impactor inlet
Store ambient pressure at impactor inlet
Input orifice calibration ambient- preset,
Store orifice calibration ambient pr*»ssu
Input Orifice temperature when samnl i no
Jtore orifice temperature when sampling.
Cnput impactor gas exit temperature.
Jtore impactor gas exit temperature.
itore orifice temperature whan calibrate
Jegin calculation.
lead orifice pressure drop.
INPUT
DATA/ UNITS
cm3/sec)
cmVseci.
- ("Hgl_
-e, ("Ha)
"6
_ _(°cj_
(°ci_
t. (°c)
KEYS
1 II 1
STO II 1 1
II 1
STO 1 2 1
II 1
n^nn i
i ii
STO II 4
1 II
STO II 5
II
STO ILe
II
STO || 7
II
LSTO J|_s
A II
II 1
II 1
II 1
II 1
II i
II 1
II 1
II 1
1 II 1
OUTPUT ,
DATA/UNITS |
-
-- • •--
("H20)
78
-------
HP-65 Program Form
Tltte Sampling Orifice Pressure Drop Calculation— Two-Stage Impactor
SWITCH to * PRCU PUSS 4 l] fPBCM~| 1O CLCAR u£ut»V
KEY
ENTRY
LBL
A
1
0
Enter
KCi, 1
X
RCL 1
X
vRCL 2
f
RCL 2
RCL 3
CHS
1
+
X
RCL 3
» CHS
1
+
X
RCL 4
X
RCL 5
i
RCL 6
2
30 7
3
+
X
RCL 8
2
7
3
+
X
4RCL 7
2
7
3
+
T
RCL 7
2
7
3
so +
?UO-06«
CODE
SHOWN
23
11
01
00
41
34 01
71
34 01
71
34 02
81
34 02
81
34 03
42
01
61
71
34 03
42
01
61
71
34 04
71
34 05
81
34 06
02
07
03
61
71
34 08
02
07
03
61
71
34 07
02
07
03
61
81
34 07
12
07
03
61
COMMENTS
KEY
ENTRY
i
RTN
CODE
SHOWN
81
24
r-
70
80
90
no
COMMENTS
REGISTER
Ri£Lamplin<
Flowra_te_
Flovfffafce
_cm /^etf \
R3Fractioi
of H20
R4Pressura
At Inlet
(;Ha)
R Ca libra--
t J.Oii
Pcesaure
("Hg) 1
RsOrifice
Temp.
(°C)
Rrlmpactoi
Temp.
^
-------
HP-65 User Instructions
Particle Si»e Distribution Calculation^— Two-Stage Impactor
Size Distribution
tribution^ __^J
1 ' ' •
-L.
VI
STEP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
INSTRUCTIONS
[nput 1st stage collected mass.
Store 1st stage collected mass.
[nput 2nd stage collected mass.
Store 2nd stage collected mass.
[nput filter collected mass.
Store filter collected mass.
Enput sampling flowrate.
Store sampling flowrate.
[nput sampling duration.
Store sampling duration.
[nput impactor gas stream temperature.
Store impactor gas stream temperature.
[nput impactor inlet ambient pressure.
Store impactor inlet ambient pressure.
[nput flue gas volume fraction of water.
Store flue gas volume fraction of water.
Jegin calculation.
Cumulative mass loading to 2nd stage cut
:umulative mass loading to 1st stage cut
Cumulative mass loading to 2nd stage cut
Cumulative mass loading to 1st stage cut
22!Total mass loading.
23JFotal mass loading.
24
25
Zumulative percent to 2nd stage cut poin
Cumulative percent to 1st stage cut poin
INPUT
DATA/ UNITS
(rag)
(mg)
(mg)
(cm3/secj
(min)
(°C)
("Hg)
point.
point.
point.
_point.
t.
t.
KEYS
1 II
STO JL 1
II
STO II 2
H
STO II 3
II
STO II 4
II
STO II 5
1
STO II a
II
STO II 7
II
STO II 8
A II
prr. 1 i
RCL II 2
I RCL || 3
RCL II 4
RCL | 5
RCL II 6
RC-T. 11 7
prr. 1 g
1
OUTPUT
DATA/UNITS
.
(mg/ACM)
(mg/ACM)
(mg/DNCM)l
I
(mg/DNCM.)j
(mg/ACM)
j
(mg/bNCM )j
(%)
(%)
1
80
-------
HP-65 Program Form
ie Particle Size Distribution Calculation—Two-Stage Impactor
SWITCH IO W'PKGJK PKESS [l ] [ PHCM~ IO ClEM UCMORV
KEY
ENTRY
LBL
A
RCL 1
RCL 2
+
RCL 3
+
RCL 4
T
iRCL 5
T
6
0
T
1
EEX
6
X
STO 4
?RCL 6
2
7
3
+
RCL 4
X
2
9
.
jo 9
2
1
X
2
9
5
T
RCL 7
T
.RCL 8
CHS
1
+
T
STO 5
RCL 1
RCL 2
+
RCL 3
CODE
SHOWN
23
11
34 01
34 02
61
34 03
61
34 04
81
34 05
81
06
00
81
01
43
06
71
33 04
34 06
02
07
03
61
34 04
71
02
09
83
09
02
01
71
02
09
05
81
34 07
81
34 08
42
01
61
81
33 05
34 01
34 02
61
34 03
61
COMMENTS
KEY
ENTRY
STO 6
RCL 3
_BCL_6
STO 7
RP.T. 3
+
RCL 6
c- T
STO 8
RCL 5
STO 6
RCL 4
STO 5
RCL 7
RCL 5
X
STO 1
'RCL 8
RCL 5
X
STO 2
RCL 7
RCL 6
X
STO 3
RCL 8
RCL 6
80 X
STO 4
RCL 7
1
0
0
X
STO 7
RCL 8
1
'0 0
0
X
STO 8
RTN
K.J
CODE
SHOWN
!33 06
4J4 03_
34 06
81
33 07
3d n3
34 02
61
34 06
81
33 08
34 05
33 06
34 04
33 05
34 07
34 05
71
33 01
34 08
34 05
71
33 02
34 07
34 06
71
33 03
34 08
34 06
71
33 04
34 07
01
00
00
71
33 07
34 08
01
00
00
71
33 08
24
COMMENTS
REGISTER:
RI 1st SLaj
weight..
RaFilter _
Weight
(mg) j
R4Flowrat
(cmVsec)
RsDuratioi
(min. )
Re Temp . .
R7Pressur<
("Hg)
RgFractiojji
of H?0
Rg
LABELS
A .St£Xt_
B
C
D
E
0
1
2
3
4
5
6
7
B
Q
FLAGS
1
2
81
-------
HP-65 User Instructions
T.n,. Impactor Stage Cut_Point Calculatipnrr Two-Stage Impactor 0,
Proqi.'immer - - — •— ——-— '""'
^rapactorStajjeCatPo^ntl /
" ' ' • ' • Q
i • i i
i
STEP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
INSTRUCTIONS
Input flue gas viscosity.
Store flue gas viscosity. - .
Input ambient pressure at impactor inlel
Store ambient pressure at impactor inlel
Input impactor exit gas temperature.
Store impactor exit gas temperature.
Input diameter of stage jets.
Store diameter of stage jets.
Input number of jets.
Store number of jets.
Input particle density of particulate.
Store particle density of particulate.
Input impactor sampling flowrate.
Store impactor sampling flowrate.
Input jet plate constant. (See other she
Store jet plate constant.
Begin calculation.
Observe resulting stage cut point.
!
INPUT
DATA/UNITS
(poise)
:. ("Hg)
• •
<°0
(cm)
(gm/cm3)
i
(cm3/sec)
set.)
KEYS
II
- _ -
STO II 1
1 II
STO II ">
II
|_STOjl_3j
II 1
STO || 4 |
II 1
STO 1 5
II
STO 1 6
II
1 STO II 7
1 II
STO || 8
i II
II
II
II
II
II
H
II
II
ii ~T
OUTPUT '
DATA/UNITS!
.
(micro-
meters)
(micro-
meters)
82
-------
HP-65 Program Form
Impactor Stage Cut Point Calculation— Two-Stage Impactor
.01
SWITCH TO W-PRGM PRESS JT] f PHQM I TO CLEM UCMOftr
KEY
ENTRY
LBL
A
RCL 3
•
0
0
3
6
7
c X
1
+
2
9
•
9
2
1
X
20 a
1/X
•
0
0
1
2
9
3
X
3
-------
TLCHNICAL REPORT DATA
(Please read lasirtii'lions on the reverse before completing)
1. REPORT NO.
EPA-600/7-77-033
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
COMPACT, IN-STACK, THREE SIZE CUT PARTICLE
CLASSIFIER
5. REPORT DATE
April 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
George E. Lacey, Kenneth M. Cushing, and
Wallace B. Smith
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 - 9th Avenue South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
1NE625
11. CONTRACT/GRANT NO.
68-02-1736
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory-RTP,NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711
13. TYP.E OF RiPftRJ AND pEBIPD COVERED
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A compact, in-stack, three size cut particle classifier was designed, fabricated
and tested. The classifier consists of a two-stage impactor and back-up filter
designed to measure the particulate emissions from sources in three size ranges:
>3ym, ^l-3ym,
------- |