EPA-650/2-74-117
NOVEMBER 1974
Environmental Protection Technology Series
ft"$:$:$:"l$i:!:::S-i:ivi::;
-------
EPA-650/2-74-117
SOURCES AND CHARACTERIZATION
OF FINE PARTICULATE TEST DUSTS
by
W. H. Hedley, S. M. Mehta, C. M. Moscowitz,
A. D. Snyder, H. H. S. Yu, andD. L. Zanders
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45410
Contract No. 68-02-1320 (Task 8)
ROAP No. 21ADM-021
Program Element No. 1AB012
EPA Project Officer: J. H. Turner
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
November 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
To assist the research efforts in applied R&D in fine
particle collection and simulation of industrial sources by
the Control Systems Laboratory of EPA and its contractors,
an investigation was undertaken to select suitable charac-
terization criteria for test dusts; determine procedures for
obtaining, handling, and characterizing the dusts; and
establish potential suppliers of test dusts. Seventeen
suitable characterization criteria for test dusts were
identified, and techniques were selected for obtaining
values for each property defined. The criteria considered
included size distribution; shape, surface area, and pore
volume; chemical composition; density; wettability and
moisture content; solubility; hardness, abrasiveness, and
grindability; charge properties; dielectric properties;
corrosiveness; optical properties; magnetic susceptibility;
plus two other properties, carrier gas composition and
solids loading. Potential industrial suppliers of test dusts
for simulation purposes were identified for eleven industries;
pulverized coal combustion, stoker-fired coal combustion,
basic oxygen furnaces, open hearth furnaces, electric arc
furnaces, metallurgical coke ovens, cement plants, municipal
incineration, steel foundries, Kraft pulp mill recovery
furnaces, and asphalt plants. The problem of producing a
particulate-laden test flue gas was evaluated. Redispersion
of collected dusts is favored over generation of fresh par-
ticulate. The establishment of a central coordinating
logistics network for the acquisition and characterization
of test dusts is recommended.
iii
-------
TABLE OF CONTENTS
1. CONCLUSIONS AND RECOMMENDATIONS 1
2. INTRODUCTION 5
3. POTENTIAL DUST SOURCES 7
4. PRELIMINARY COST AND TIME REQUIREMENTS FOR 11
ACQUISITION AND CHARACTERIZATION OF TEST DUSTS
5. SELECTION OF DUST DISPERSION METHOD 15
5.1 GENERATION AND REDISPERSION 15
5.2 PREFERABLE DUST DISPERSION TECHNIQUES 19
5.2.1 Fluidized Bed 19
5.2.2 Jet Mill 21
5.2.3 Combustion or Explosion of Dust- 21
Fuel Mixture
5.2.4 Electric Arc Furnace 21
5.2.5 Rotary Kiln 24
5.3 JUSTIFICATION FOR REDISPERSION OF 24
TEST DUSTS
5.3.1 Aging and Agglomeration 24
5.3.2 Handling 26
5.4 QUALITATIVE EXPERIMENT ON FLUIDIZED BED 27
6. REFERENCES 33
APPENDIX A SELECTION OF DUST CHARACTERIZATION 37
CRITERIA'
A.I SIZE DISTRIBUTION 39
A.2 SHAPE, SURFACE AREA, AND PORE VOLUME 39
A.3 CHEMICAL COMPOSITION 4l
A.4 DENSITY 41
A.5 WETTABILITY AND MOISTURE CONTENT 42
A.6 SOLUBILITY 43
A.7 HARDNESS, ABRASIVENESS AND GRINDABILITY 43
A.8 CHARGE PROPERTIES 43
A.9 DIELECTRIC PROPERTIES 44
A.10 CORROSIVENESS 45
-------
TABLE OP CONTENTS - Cont'd
A.11 OPTICAL PROPERTIES
A.12 MAGNETIC PROPERTIES
A.13 CARRIER GAS COMPOSITION AND SOLIDS LOADING
APPENDIX B
B.I
B.2
B.3
B.4
B.5
B.6
B.7
B.8
B.9
B.10
B.ll
B.12
APPENDIX C
APPENDIX D
Particle Sizing Techniques
Primary Particle Sizing
Aerosolized-Size-Distribution
DUST CHARACTERIZATION TECHNIQUES
PARTICLE SIZE DISTRIBUTION
B.I.I
B.I.2
B.I.3
SHAPE, SURFACE AREA, AND PORE VOLUME
CHEMICAL COMPOSITION
DENSITY
WETTABILITY AND MOISTURE CONTENT
B.5.1 Wettability
B.5.2 Moisture Content
SOLUBILITY
HARDNESS, ABRASIVENESS, AND GRINDABILITY
CHARGE ANALYSIS
DIELECTRIC PROPERTIES
CORROSIVENESS
OPTICAL PROPERTIES
MAGNETIC PROPERTIES
TYPICAL FLUE GAS CHARACTERISTICS FOR
INDUSTRIAL SOURCES
LABORATORY TESTING OF REDISPERSIBILITY
OF BULK POWDERS
APPENDIX E METRIC CONVERSION TABLE
Paf
46
46
48
49
49
50
52
53
59
62
64
67
67
68
70
70
73
75
76
78
81
83
95
98
vi
-------
LIST OF FIGURES
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5-
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Potential Dust Acquisition System
Size Distribution of Coal Dust, Flyash
and Test Dust Dispersed by Fluidized
Bed and Water
Schematic of Fluidized Bed Dust Dispersion
System
Schematic of Jet Mill Installation
Schematic of Bench-Scale Fluidized Bed
Photomicrograph of Fluidized Bed Beads
Dispersion Using Glass Beads ('v/lOOy)
Dispersion Using Nickel Beads
Electron Photomicrographs of Aerosol
Particles in Smokes
Figure 10. Calculated Effect of Particle Refractive
Index on the Calibration of the Optical
Counter
Figure 11.
Figure 12.
Schematic Diagram of Particle Size
Comparator
Schematic of Potential Particle Size
Distribution Measurement System Utilizing
a Diffusion Battery
Figure 13. Schematic of Potential Particle Size
Distribution Measurement System Utilizing
an Electrical Particle Counter
Figure 14. Size Distributions of Indoor and Outdoor
Aerosols
Figure 15.
Figure 16.
Figure 17.
Air Pycnometer - Piston Action
Block Diagram Moisture. Analyzer
Gillespie Apparatus for Measuring Particle
Size and Charge
Page
2
20
22
23
28
29
30
31
40
47
54
56
57
60
66
71
74
vii
-------
LIST OF FIGURES - Con't
Page
Figure 18. Resistivity Probe 77
Figure 19. Schematic of Optical Arrangement 80
Figure 20. Principle of Gouy Balance 82
Figure 21. MRC Glass Reaerosolizability Apparatus 97
viii
-------
LIST OP TABLES
Table 1.
Table 2.
Table 3.
Table I*.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Page
Relative Importance of Pine Particle i\
Characterization
Potential Dust Suppliers 8
Dust Acquisition Costs for 1,000 Pounds 12
of Test Dust
Dust Characterization Costs for Each 14
Sample
Advantages and Disadvantages of Generation 17
and Redispersion of Pine Dusts
Suitable Dust Dispersion Techniques 18
Forces Influencing Partial Collection 38
in Air Pollution Control Devices
Summary of Types of Size Analysis 51
Techniques
Operable Size Ranges of Practical 55
Combinations of Size Classifiers and
Sensors
Protocol for Analytical Analysis 63
Wettability Determination Results by 69
Cylinder Method
ix
-------
SECTION 1
CONCLUSIONS AND RECOMMENDATIONS
CONCLUSIONS
As a result of our investigation during this task, MRC
concludes that securing adequate quantities of well-charac-
terized particulates to fulfill the testing requirements of
the Control System Laboratory can be accomplished by utili-
zing collected dust for redispersion from most industries
of interest.
The test dusts can be characterized on many levels. Values
for these characterization criteria may be obtained from a
well-equipped analytical laboratory. The dust may be charac-
terized to any level desired, but to keep cost at a minimum
the variables and particulate properties required for each
specific experiment must be carefully defined.
To provide a suitable test dust in the most efficient and
economical way requires a well organized network of indus-
trial contacts and a wide variety of analytical capabilities.
Assuming that no one company can serve as a source of supply
for all types of dust required, the most efficient system
would establish a central coordination logistics network
to arrange for the acquisition and characterization of the
needed particulate samples. This operation would optimize
utilization of the talents and capabilities of the test dusts
sources, the EPA, and the appropriate Industrial contractors.
Working relationships with the various potential dust sources,
which are discussed in Section 3, could be established ac-
cording to a framework similar to that shown in Figure 1.
Within this framework several alternatives exist to maximize
efficiency and minimize cost. Depending on the requirements
'for the test dusts, samples might be sent directly from their
source to the Control Systems Laboratory or its contractors,
-------
Figure 1. Potential Dust Acquisition System
Pulverized Coal
Fired Combustion
Stoker Fired
Coal Combustion
Basic Oxygen
Furnace
Open Hearth
Furnace
Electric Arc
Furnace
Metallurgical
Coke Oven
Cement Plant
Municipal
Incineration
Steel Foundry
Kraft
Pulp Mill
Asphalt Plant
Other Dusts
I^HHMB
— —
— —
(Direct Shipment)
Dust Characterization
Center
Industrial Laboratory
i 1
i t
i '
Particle Size
Subcontracting f
Chemical
(Large Scale) Composition
Determination
t
All Other Necessai
Criteria
Establishment Wor
(Direct Shipment)
1
1
t
EPA
^ Duct [Icjnn
Facility
t
ry
•k
_,__ T._ ,^_ . ^_
1
(Direct Shipment) ^
-------
they might go through a central coordinating facility and
Industrial laboratory for characterization, or for special
requirements beyond the capabilities of the central facility,
the test dusts might be channeled to an appropriate subcon-
tractor enroute to their destination.
RECOMMENDATIONS
Creation of a flue gas suitable for fine particulate investi-
gation appears to be feasible by redispersion of collected
particulate using fluidized bed techniques. However, to
ensure the applicability of this technique for the varied
particulate sources of interest, further investigation is
recommended. The effect of the radioactive charge neutral-
izer, the influence of a metallic fluidized bed redispersion
column, the grain loading obtainable, and the applicability
of alternative redispersion techniques should be identified.
Once the aforementioned reservations concerning redispersion
of the dust are removed, it is recommended that the Control
Systems Laboratory of EPA seek a contractor to establish a
central coordination logistics network to acquire and charac-
terize test dusts. The system could be similar to that
described in Figure 1, and should be capable of acquiring,
characterizing, and distributing ample quantities of test
dust for the experimental programs of the Control Systems
Laboratory and its contractors.
The contractor should be equipped to characterize particulate
with regard to most if not all of the properties listed in
Table 1, and described in Appendix A. A contractor with
contacts among members of the industries listed in Figure 1
would also be beneficial in expediting the establishment and
operation of a dust acquisition system. An industrial coor-
dinator/supplier with numerous established Industrial contacts
for dust acquisition, extensive analytical capabilities for
dust characterization, and a history of working with EPA and
other government agencies would be ideal for this task.
-------
Table 1. RELATIVE IMPORTANCE OP PINE PARTICLE CHARACTERIZATION
£_l
o>
Xi
XI
3
£_i
0
CQ
CD
0
•H
CD
Q
£
O
•H
-P
0
CD
H
H
0
O
•P
0>
fe
£»s
M
CD
£
CD
X!
60
•H
O
-P
cd
-P
•H
a
•H
O
0>
PH
a
o
•H
-P
Cd
-P
CO
O
fn
•P
O
CD
rH
W
P |
0)
•P
H
•H
Cj_|
O
•H
£_,
o
cd
C
o
•H
4J
rt
4^
C
0)
g
3
£
o
•H
CO
PH
0)
D,
co
•H
T)
PH
J^
I •>
co
C
H
c
o
R
g
O
o
co
•H
CO
>J
rH
cfl
C
cd
(U
ts)
•H
co
^
O
t-l
r-H
cd
0
•H
C
c«
X!
O
0)
s
H
CtJ
o
•H
-P
a
o
H
cd
o
•H rH
-P S
O PH
0) CD
H X!
W EH
H
cd
o
•H
g
0)
X!
o
1
0
o
•H
CO
>3
X!
Property
Size distribution
Shape, surface area, pore
volume
Chemical composition
Density
Wettability and moisture
content
Solubility
Hardness, abrasiveness,
grindability
Charge Properties
Dielectric properties
Corrosiveness
Optical properties
Magnetic susceptibility
P P P P
S S P S
S S S S
P P P P
S P P P
S S S L
L L S P
S S S P
S P P P
P S S L
L L L L
L L L P
P P P P P
S P P L S
S S S S S
P L S S L
L L L L P
L L L L L
L L L L L
S L P L L
S L P L L
L L L L S
L P L L S
L L L L L
P * primary importance
S = secondary importance
L = little or no importance
-------
SECTION 2
INTRODUCTION
The Control Systems Laboratory of the Environmental Protection
Agency (EPA) has been involved in evaluation and improvement
of particulate measuring and control devices for several years.
Now, with increased emphasis being placed on particle size,
(especially fine particles less than five microns in size),
it is necessary to establish sources of reproducible, well
characterized, fine particulates for experimentation in this
field. The test dusts are to be used for two general pur-
poses: (1) applied R&D in fine particle collection, and
(2) simulation of industrial sources. Test particles are
presently .available from Dow, glass spheres from NBS, AC test
dusts, and ASME test dusts, but due to their costs, quantity
purchase is prohibitive. Consequently, an alternative source
of test dusts is desirable.
To assist in this effort, an investigation was undertaken
to select suitable criteria for test dusts; to determine
procedures for obtaining, handling, and characterizing the
dusts; and to establish potential suppliers of test dusts.
Suitable dust characterization criteria were selected by con-
sidering the relative importance of the various criteria on
conventional collection equipment, redlspersion requirements,
and common size-measuring instrumentation. The dust proper-
ties considered were: size distribution; shape, surface area,
and pore volume; chemical composition; density; wettability
and moisture content; solubility; hardness, abrasiveness, and
grindability; charge properties; dielectric properties; cor-
rosiveness; optical properties; and magnetic susceptibility.
Carrier gas composition and solids loading were also consid-
ered. Discussion of these properties and their relative
importance to the three areas mentioned above appears in
Appendix A, Selection of Dust Characterization Criteria, while
Appendix B, Dust Characterization Techniques, identifies po-
tential techniques for measurement of each dust characteristic.
-------
Section 3 entitled Potential Dust Sources identifies possible
suppliers of test dusts for the simulation of industrial
sources. Potential suppliers were identified for the follow-
ing sources: (1) pulverized coal combustion. (2) stoker-fired
coal combustion, (3) basic oxygen furnace, (4) open hearth
furnace, (5) electric arc furnace, (6) metallurgical coke
oven, (7) cement plant, (8) municipal incineration, (9) steel
foundry, (10) Kraft pulp mill recovery furnace, and (11) asphalt
plant.
Two examples of the cost and time required to provide quanti-
ties of well characterized test dusts follow in Section 4,
Preliminary Cost and Time Requirements for Acquisition and
Characterization of Test Dusts. Estimates of the cost to
acquire and characterize test dusts for each of the character-
ization criteria are provided for samples of flyash from
pulverized coal combustion and particulate from municipal
Incineration.
The final section, Selection of Dust Dispersion Method,
addresses the problem of establishing a particulate laden
test gas. Two techniques for creating a dust dispersion,
generation or redispersion, were considered and redispersion
is recommended. As support for this position a simple quali-
tative experiment was conducted using a fluldized bed of
^100 vi glass beads or nickel beads to successfully redisperse
flyash or iron oxide particulate.
6
-------
SECTION 3
POTENTIAL DUST SOURCES
For much of the fine particulate research being proposed by
EPA, a representation or simulation of industrial effluents
is required. To produce these simulated flue gases, sources
of dust supplies must be secured. The fine particulate
sources of greatest interest to EPA at present are combustion
processes and iron and steel furnaces. For combustion pro-
cesses both pulverized coal-fired and stoker-fired coal com-
bustion methods are of interest, and in the iron and steel
industry basic oxygen furnaces, open hearth furnaces, electric
arc furnaces, and metallurgical coke ovens are also of inter-
est. Other sources of fine particulate emissions of interest
are cement plants, municipal incinerators, steel foundries,
Kraft pulp mill recovery furnaces, and asphalt plants.
Several potential industrial sources of test dusts were
contacted. The results of these contacts are summarized
in Table 2.
The Dayton Power and Light Company operates both pulverized
and stoker-fired coal combustion electric power generation
facilities, and is a potential dust supplier. No difficulties
are anticipated in obtaining quantities of test dust for redis-
persion from both type facilities.
Bethlehem Steel Corporation operates several basic oxygen fur-
naces in Sparrow's Point, Maryland; Bethlehem, Pennsylvania;
and Burns Harbor, Indiana. Particulate control is accomplished
with electrostatic precipitation. Quantities would be avail-
able if handling and shipping problems can be solved.
Open hearth furnaces operated by Bethlehem Steel Corporation
in Sparrow's Point, Maryland; Johnstown, Pennsylvania; and
Lackawanna, New York utilize electrostatic precipitators for
particulate emission control. One shop at Sparrow's Point
used fabric filtration but is no longer in operation. The
collected material may be,acquired for redispersion testing.
-------
Table 2
POTENTIAL DUST SUPPLIERS
Industry
Pulverized and stoker fired
coal combustion
Basic oxygen furnace
Open hearth furnace
Electric arc furnace
Company
Cement plant
Municipal incineration
Steel foundry
Kraft pulp mill recovery
furnace
Asphalt Plant
Dayton Power and Light Company
Dayton, Ohio
(513) 222-0441
Bethlehem 'Steel Corporation
Bethlehem, Pennsylvania
(215) 694-3676
Bethlehem Steel Corporation
Bethlehem, Pennsylvania
(215) 694-3676
Bethlehem Steel Corporation
Bethlehem, Pennsylvania
(215) 694-3676
Inland Steel Company
East Chicago, Indiana
(219) 392-1200
Southwestern Portland Cement Company
Pairborn, Ohio
(513) 878-8651
City of Braintree
Waste-Disposal Department
Braintree, Massachusetts
(617) 843-6209
Chicago Northwest Incinerator
Chicago, Illinois
(312) 744-4571
Dayton Steel Foundry Company
Dayton Walther Corporation
Dayton, Ohio
(513) 296-3113
Mead Corporation
Chlllicothe, Ohio
(614) 772-3111
Valley Asphalt Corporation
Dayton, Ohio
(513) 293-4119
Contact
Mr. H. Palmer
Mr. J. Leming
Mr. J. Leming
Mr. J. Leming
Mr. J. Brough
Mr. J. Dorn
Mr. Courchene
Mr. Groszek
Mr. F. Fensel
Mr. W. Lapp
Mr. M. Levy
8
-------
Fabric filters are used to control particulate emissions from
several of Bethlehem Steel Corporation's electric arc furnaces.
Another potential source of fine particles from an electric arc
furnace is the Inland Steel Company's East Chicago, Indiana
plant. The collected material is stored until an adequate
amount is accumulated at which time it is wetted to cause
agglomeration and then removed to a landfill. Quantities
may be collected before wetting and are available as needed
from either source.
The search for a potential supplier of collected coke parti-
culate emissions was abandoned after numerous contacts and
continued discouraging comments on the subject. (Contacts
were made with American Iron and Steel Institute, Bethlehem
Steel Corporation, National Steel Corporation, Inland Steel
Company, and Great Lakes Carbon Corporation). Collection
equipment is not common and where used is generally of the
wet scrubbing variety.1 Consequently, simulation of metal-
lurgical coke ovens may require generation of test dusts.
Southwestern Portland Cement Company of Fairborn, Ohio is a
dry process Portland cement manufacturer. Their emissions
are controlled by an electrostatic precipitator. Unlimited
quantities of their collected dust are available.
Both the Braintree, Massachusetts and Chicago Northwest muni-
cipal incinerators utilize electrostatic precipitation for
particulate collection. The collected material is currently
landfilled. Quantities of the collected material are avail-
able from either of these sources.
The Dayton Steel Foundry Company of Dayton Walther Corporation
operates an electric arc furnace controlled by a high effi-
ciency fabric filter. The collected material is discarded
as waste. Consequently, they are willing to provide quantities
of the collected material upon request.
The Mead Corporation operates a kraft pulp mill in Chillicothe,
Ohio. The emissions from the recovery furnace are controlled
by a low efficiency electrostatic precipitator. Samples of
the collected material could be obtained.
Valley Asphalt Corporation of Dayton, Ohio operates sand and
gravel based asphalt plants. Three of their plants in the
Cincinnati area use fabric filters to collect fine particulate
emissions. Quantities of the collected material could be
obtained.
1. Anon., "A Search for Clean Coking Processes," Steel
Facts, Number 224, 10-11 (January 1974).
-------
For the industries where the collected particulate does not
adequately represent the small end of the particle size distri-
bution, auxiliary particulate collection is possible. For
example, a high efficiency mobile fabric filter could be
applied to a side stream to collect fine particulate for
augmentation of the already collected material. Also, to
meet specific size requirements classification of collected
dusts as shown in Figure 1 may also be performed.
10
-------
SECTION 4
PRELIMINARY COST AND TIME REQUIREMENTS
FOR ACQUISITION AND CHARACTERIZATION OP TEST DUSTS
Cost and time requirements for MRC to provide quantities of
test dusts are estimated below. These figures are preliminary
estimates and are to be used only to establish a cost range
for the acquisition of test dusts suitable for research and
development efforts.
Two dust sources have been arbitrarily selected as typical
examples of a request for test dust so that costs may be esti-
mated. The two sources are (1) a pulverized coal fired uti-
lity boiler in Dayton, Ohio, and (2) a municipal incinerator
in Chicago, Illinois. The cost of the test dust on a per
pound basis is proportional to the quantity of dust provided,
with smaller quantities being more expensive. To establish
a basis for this cost estimate we have assumed that 1000 *
pounds of each test dust as collected at the industrial site
have been requested.
Table 3 presents the cost of obtaining a 1000-pound quantity
of test dust from its source. For the pulverized boiler the
cost for obtaining the dust is $228 while the cost for muni-
cipal incineration dust is $1525. Notice that the bulk of
the dust acquisition cost is encounted in the manpower re-
quirements to collect the dust and return it to the charac-
terization center. Depending on the geographical orientation
of the source, characterization center, and destination of
the test dust; the dust might be returned to the characterir-
zation center or sent directly to its destination with small
samples retained for characterization purposes. The dust may
be carried in the dust acquisition vehicle or may be sent by
common carrier.
*See Appendix E, Metric Conversion Table, for metric equival-
ents of English units.
11
-------
Table 3. DUST ACQUISITION COSTS FOR 1,000 POUNDS OP TEST DUST
r\j
Charges for material
and container
Labor
Subsistence
Truck
Total
($)
Plyash from Pulverized
Coal Utility Boiler
Dayton, Ohio
ESP Waste from
Municipal Incinerator
Chicago, Illinois
$0.50 per 50 Ib
$1 per 50 Ib, flyash $ 20 particulate
2 men for h day
($25/hr)
no charge
$15/day, 15
-------
We have assumed for our estimate that the 1000 pounds of test
dust are returned to the dust characterization center by the
dust acquisition team.
The dust must then be characterized as specified by the user's
request. Table 4 presents estimated typical charges for char-
acterization of each sample. Depending on user specifications,
multiple samples might be required, increasing the per pound
cost of the test dust. However, the increased cost will not
be linear in nature due to the economy achieved in scale.
The total cost for 1000 pounds of flyash or incinerator ESP
waste will be a combination of the figures in Tables 3 and 4
depending on the level of characterization and number of test
samples required for characterization.
The time required to provide quantities of adequately charac-
terized dusts suitable for experimental work depends on sev-
eral factors. The quantity of dust requested and the location
of the source of the dust will influence the time required
for actual dust acquisition. The level of dust characteriza-
tion requested will also influence time requirements; more
characterization criteria or multiple samples would require
additional time. In any case, three weeks would be adequate
to acquire, characterize, and prepare for shipment a quantity
of test dust suitable for experimentation. The time required
between receipt of a request for test dust and shipment of
the product dust may possibly be shortened by the establishment
and maintenance of a stock of test dusts to be drawn upon as
needed.
-------
Table 4. DUST CHARACTERIZATION COSTS FOR EACH SAMPLE
Property
Size Distribution (primary particles)
Shape**
Surface Area (BET)
Pore Volume
Chemical Composition
Density
Wettability (qualitative)
Moisture Content
Solubility
Hardness
Abrasiveness*
Corrosiveness*
Grindability
Dielectric Properties
Charge Properties
Magnetic Susceptibility
Optical Properties
Cost
($/sample)
(June
60
170
70
210
270
40
10
40
40
40
70
200
110
110
410
*Due to the combined influence of flue gas and particulate on
determination of abrasiveness and corrosiveness, these properties
are best measured on an actual flue gas at the test facility.
**Cost for shape measurement includes the cost for size distribution
surface area, and density measurements.
14
-------
SECTION 5
SELECTION OF DUST DISPERSION METHOD
5.1 GENERATION AND REDISPERSION
The test dusts will be used for two general purposes:
(1) applied R&D in fine particle collection, and (2) simu-
lation of industrial sources.
Two methods of providing dust dispersions for the simulation
of industrial sources are (a) generation of dust from a par-
ent material, and (b) redispersion of an existing dust col-
lected from industrial effluent of interest. The objective
is to produce a dust stream which is almost identical in all
respects to those existing in the stack. It appears that
neither generation nor redispersion would reproduce a dust
stream (in the laboratory) which is identical to large-scale
industrial effluent.
If dust is generated from-a parent material in a small-scale
furnace or other facility (e.g. plasma jet, rotary kiln, etc.),
the resultant product may not simulate exactly the industrial
effluent. The dust characteristics can change with the size
of furnace and process design. For example, the effect of
process design (firing method) on the size distributions from
pulverized coal combustion and stoker-fired coal combustion
(see Appendix C) are the result of different gas velocities,
temperature, and residence time in the furnace. The effect
also extends to density and chemical composition because the
time-temperature relationships influence the burnout of car-
bon and the fusion of the mineral components. In the iron
and steel industry, emissions from open hearth furnaces and
basic oxygen furnaces are very different in size, shape, den-
sity, and resistivity as the result of the design and opera-
tion of the two types of furnaces. Reported particle size
distribution measurements from several power plant installa-
tions indicate that boiler furnace and operating conditions
15
-------
are the most important factors in determining fly ash particle
size.2
In addition to difficulty in industrial effluent simulation,
dust generation is also complex and expensive. This is not
to say that redispersion of an existing dust is without any
problem. On the other hand, it would be difficult to redispers3
extremely fine particles (as found in basic oxygen furnace and
electric arc furnace operations) or sticky particles (as prob-
ably found in metallurgical coke oven operation) by any of
the redispersion methods.
To ease the comparison of dust handling techniques, the advan-
tages and disadvantages of generation and redispersion are
listed in Table 5« The primary advantage of generation is
freshness of generated particles. This is important when
certain surface characteristics become important or when extra^'
fine dust is not obtainable by redispersion of existing dust.
However, for the majority of industrial effluents, the labora-
tory simulation can be achieved by means of a redispersion
technique. Based on our past experience (e.g. sampling and
laboratory redispersion) and the analysis of the problem,
handling techniques suitable for simulation of various indus-
tries of interest were classified and are listed in Table 6.
Among the many ways to redisperse existing dust are the
following:
(a) fluidization
(b) jet mill dispersion
(c) combustion or explosion of dust-fuel mixture
(d) atomization of liquid suspension and evaporation
(e) air-blast powder dissemination
Applicability of these dispersion methods to various Industrie^
is also shown in Table 6.
Several generation techniques which may be suitable for simu-
lation of industrial effluent are:
(a) electric arc furnace
(b) rotary kiln
(c) plasma jet, etc.
Their applicability is also listed in Table 6.
2. Anon., "Criteria for the Application of Dust Collectors
to Coal Fired Boilers," IGCI/ABMA Joint Technical
Committee Survey, April, 1965-
16
-------
Table 5. ADVANTAGES AND DISADVANTAGES OF GENERATION AND
REDISPERSION OP PINE DUST
Generation
Re dispersion
1. Dusts are freshly formed
2. Similar carrier gas
(not necessarily exact)
Advantages 3. No need for dispersion
1. Equipment relatively inexpensive
2. Versatile, same equipment for many different dusts
3. Easy for government contractors to duplicate the system
ll. Same chemical composition
5. Same primary particle size
6. Most of dusts are readily available from bag house or
preeipitator of various industries
Dl sadvantage s
5.
6,
Expensive Cequipment
and operation)
Not exactly the same
aggregate size
distribution
Need different equipment
for different dust
Need exact composite raw
materials for furnaces
Not necessarily the same
chemical composition
Not necessarily the same
primary particle size
Expensive to duplicate
the system
Not fresh dust surface (aging effect)
Not exactly the same aggregate size distribution
Need dispersion and sometimes difficult to disperse
extra fine or sticky dust
Need gas conditioning {e.g. composition and humidity,
etc. )
Need collection of dust from sources
-------
Table 6. SUITABLE DUST DISPERSION TECHNIQUES
Industry
1 Pulverized Coal Combustion
2 Stoker Fired Coal Combustion
3 Open Hearth Furnace
4 Basic Oxygen Furnace
5 Electric Arc Furnace
6 Metallurgical Coke Oven
7 Cement Plant
8 Kraft Pulp Mill
9 Asphalt Plant
10 Municipal Incineration
11 Steel Foundry
P • Preferred
L Less Preferred Redispersion
Generation Redispersion
L (K) P (F, J)
L (K)
L (K)
P (E)
P (E)
P (E)
L (K)
L (K)
L (K)
L (K)
L (E)
P (F, J)
P (F, J)
L (C, J, F)
L (C, F)
L
P (F, J)
P (F, J)
P (F, J)
P (P, J)
P (F, J)
F Fluldlzed bed
J Jet mill
C Combustion
Generation
/E Electric arc furnace
IK —
Rotary kiln
18
-------
5.2 PREFERABLE DUST DISPERSION TECHNIQUES
In the previous section various methods of generation and
redispersion techniques were listed. It would be very diffi-
cult for a single piece of equipment to disperse the dust
(either by generation or redispersion) and simulate all 11
industries listed in Table 6. It would be desirable to use
a single redispersion technique (e.g. fluidized bed) for a
majority of the industries. For the simulation of one or
two other industries, a different generation scheme may be
needed. For example, it would be very difficult to redisperse
a mixture of tar-like or oily dust by a redispersion technique,
Among the many techniques listed in the previous section, some
of the more promising ones are discussed below.
5.2.1 Fluidized Bed
The fluidized bed appears to be particularly suitable for
dust dispersion because it deagglomerates the dust and at the
same time acts as a capacitor, thereby damping out unsteady
dust feedings into the bed. When a mixture of dust and spher-
ical beads is exposed to upward airflow, the dust will mix
and eventually elutriate out of the fluidized bed. The dust
should have an average aerodynamic diameter one order of
magnitude smaller than the bed solids.
In order to minimize electrostatic charging, the fluidized
bed column and the bed solids should be made from electrically
conducting materials (e.g. metal). Bed solid materials should
also have the following properties: (a) smooth, spherical
shape, (b) hard material, low abrasion, (c) inert at operation
temperatures and pressures, no corrosion or oxidation, (d) no
gas or liquid adsorption, (e) low electrical charging effects.
Successful dispersion of several dusts (fly ash, coal dust,
and air cleaner test dust) by fluidized bed is reported by
K. Willeke and co-workers at the Particle Technology Labora-
tory of University of Minnesota.3 The results of their
work are shown in Figure 2.
The quality of dispersion by fluidized bed was determined by
comparing the dr,y dispersed dust from a fluidized bed with
the same dust dispersed in the liquid. Considering an accu-
racy of 10 to 15% for the determination of the particle
Willeke, K., Lo, C. S. K., and Whitby, k. T., "Dispersion
Characteristics of a Fluidized Bed," Particle Technology
Laboratory Publication No. 217, Particle Technology
Laboratory, Mechanical Engineering Department, University
of Minnesota, December 1973.
19
-------
99.9
99
90
CO
to
Ul
C
LJ
CD
50
» 10
r-
j
CLASSIFIED AIR CLEANER TEST DUST -i
CLASSIFIED COAL DUST—7 /
FLY ASH-7 / /
FLU1DI2LED BED.
n/ o— WATER'
I 5 10 50 100
PARTICLE DIAMETER (MICROSCOPE) , /im
99
ui
W
9-80
z
10
IT
i ,,,I
CLASSIFIED
COAL DUST
FLYASH
FJ-.UIII.BEDr
HEIGHT, cm.
1
2
Figure 2.
.51 5 10 50
PARTICLE DIAMETER (OOP-CALIBRATED OPC)j*ftm
*OPC - optical particle counter
Size Distribution of Coal Dust, Fly Ash, and
Test Dust Dispersed by Fluidized Bed and Water
20
-------
number distribution under an optical microscope, the two dis-
persion techniques result in equal size distributions by
number within experimental error. The fluidized bed appears
to disperse well those dusts which do not cake appreciably.
However, a fluidized bed may not sufficiently deagglomerate
some aggregates such as. carbon black.
A typical fluidized bed dust dispersion set-up is shown in
Figure 3« Since the dynamics of the fluidized bed always
electrically charges the effluent dust particles, a radioactive
charge neutralizer is shown above the fluidized bed. In case
the formation of charged strings of particle aggregate is
desired, the neutralizer could be removed.
5.2.2 Jet Mill
In work on Brink® fiber beds and wet scrubbers, we have suc-
cessfully redispersed submicron iron oxides and fly ash in
our dust testing facility using Jet mills and various dust
feeders (such as the BIP Omega grooved disc dry feeder, Metco
plasma flame spray feeder, TAP A powder feeder, SYLCO powder
feeder, and vibrating trough feeder). Figure 4 is a schematic
of our jet mill installation. Our past experience indicates
that dispersing and feeding of submicron iron oxides is much
more difficult than that of fly ash.
5.2.. 3 Combustion or Explosion of Dust-fuel Mixture
For extremely fine particles often it is much easier to dis-
perse in a liquid phase than in a gas phase. To facilitate
dispersion fuel (e.g. acetone) or some other explosive mate-
rial may be used. The technique has been widely used in the
dissemination of fine particulate matter (e.g. silver iodide)
in work on atmospheric trace or weather modification studies.
Desired dust may be mixed in a solution or flare material at
proper concentration for dust redispersion purpose in the
laboratory.
5.2.4 Electric Arc Furnace
Electric arc furnace or arc welding techniques have been used
occasionally for producing fine metal fumes in the laboratory
(e.g. at the University of Minnesota and the West Virginia
University). ** Fresh and highly charged fine particles (thread-
like aggregate can be produced by. this method. However, it
Personal Communication, Professor B. Linsky, West Virginia
University, 4 February 1974.
21
-------
Sampling Port
Vibrating
Screw Feeder
(enclosed)
Dispersed Dust
to Test Duct
Support
Screen
Radioactive Neutralizer
(when needed)
Conductive Column
Fluidized Bed
(Nickel Beads)
Observation Port
Air In
Air Filter
Ground
Figure 3. Schematic of Fluidized Bed Dust Dispersion System
-------
LARGE
PARTICLES
DISCHARGE
(Variable Sizes)
CLASSIFICATION
CHAMBER
FINE PARTICLE
SEPARATION
UPSTACK
COMPRESSED
AIR
IONIZER
P TUBE IMPACT
(Variable Sizes) CHAMBER
COMPRESSED
AIR
FRONT VIEW
TEST DUCT
DIFFUSER PLATE
JET MILL
"ORIFICE PLATE
VIBRATING DUST FEEDER
K 85 NEUTRALIZER
SIDE VIEW
Figure 4. Schematic of Jet Mill Installation
23
-------
would be difficult to produce a large quantity of dust to
simulate grain loadings of an industrial stack, especially
for a large test facility. Metals of special interest can
be used as consumable electrodes to produce desired metal
fume s.
5.2.5 Rotary Kiln
A small rotary kiln could be used to generate a variety of
dusts by burning mixtures of materials of interest. As dis-
cussed in an earlier section, even with proper control of
burning conditions (e.g. temperature) the resultant fresh
dusts may not be identical to industrial effluent in particle
size distribution or chemical composition. However, the
equipment could be valuable in certain aspects of laboratory
study.
5.3 JUSTIFICATION FOR REDISPERSION OF TEST DUSTS
5.3.1 Aging and Agglomeration
In order to understand the effect of aging on the production
of fine particles, it may be helpful to examine the formation
of particles in the combustion process. For example, when
fumes and smokes are generated by combustion, the basic mecha-
nism operating is usually that of vapor condensation. Because
of the high temperatures involved and the rapidity of the
reaction taking place, it is impossible to analyze the process
in detail. Two types of primary particles form the reaction
zone. One is crystalline in nature (such as perfect crystals
of some metal oxides) and the other amorphous in nature (such
as carbon smoke). After the particles leave the reaction
zone, coagulation begins and aggregates are formed. Fine
particles of submicron size tend to coagulate quite rapidly
owing to their lively Brownian motion. Freshly formed fumes
or smokes, therefore, exhibit great Brownian activity which
brings individual particles into frequent contact. If the
particles adhere on such contact, coagulation has occurred,
and the particle size is thereby increased. As the average
size increases, it is clear that the rate of coagulation
decreases and the particle-size structure of the particulates
becomes more stable.
When carbon smoke is produced by incomplete combustion, the
small particles tend to form long chains or filaments. In
the case of electric arc furnaces, metal vapor evolved at a
very high temperature is cooled in the air stream and con-
denses to form a smoke. Oxidizable metals (e.g. cadmium,
-------
lead, copper, manganese, chromium, magnesium, and aluminum)
readily give oxides, while platinum, silver, and gold yield
metal smoke. The smoke from copper and iron consists of mixed
oxides. The smoke generated in electric arc furnaces will be
electrically charged, and since the initial concentration is
high and is reduced comparatively slowly by the diluting air,
one should expect that aggregates of wide range of complexity,
electrical charge, and size would be present in arc smoke.
Microscopic examination shows this to be the case, the com-
plexes consisting of long chains, frequently of hundreds
of fine particles, most of which are too small to be resolved
by the optical microscope.
After the formation of particles, often an adsorbed film is
formed on the surface of the particles during the process of
cooling. If ample moisture (or S03) is present in the flue
gas, it will be adsorbed on the surface of particles, thus
changing the characteristics (such as resistivity) of collected
particles. Adsorption of moisture is often influenced by the
existence of other species. For example, a weak basic particle
such as ZnO would exhibit improved adsorption capacity with
the presence of HC1 or H2SOi» vapors. Similarly, weak acidic
particles such as AlaOs would show improved adsorption char-
acteristics with the presence of ammonia in the flue gas. In
coal-burning furnaces, at the temperatures that exist down-
stream of air heaters, the S03 probably combines with moisture
to form adsorbed E2SO^ on the fly ash. The . I^SOi, is highly
hygroscopic and attracts free moisture. Multiple adsorption
systems of this type occur with either weak acidic (e.g. ZnO)
or weak basic (As203) particles with the presence of strong
acids or strong bases as just described.
Hydratlon and hydrolysis can also occur during the process of
aging thus changing the characteristics of particles. There-
fore, unless freshly collected dust particles are prevented
from being directly exposed to moist gas (or ambient air),
the aging of dust particles could occur, thus changing some
of the particle characteristics. In the simulation of an
industrial effluent, use of film adsorbed aged fine particles
could result in different experimental results, for instance,
if the collection efficiency of control equipment is strongly
influenced by resistivity.
s
However, with proper handling (e.g., collection, bagging,
shipping, storage, etc.) of bulk dust, it is possible to
prevent or minimize the aging of fine particles for later
use. The factors that we must consider are how the aging
would alter the properties of particles listed in Table 1.
Among the properties listed in the table, the properties
which may be affected by improper handling of particles are
chemical composition, moisture content, dielectric properties,
-------
charge properties, density, size distribution, or shape.
Therefore, it is of utmost importance that care be exercised
from the point of source collection to the end use.
The particles in many bulk dusts and fine powders tend to
clump together to a remarkable degree, and although it is pos-
sible to redisperse them, many of the airborne particles will
be found to be aggregates composed of varying number of indi-
vidual particles unless special precautions are taken to mini-
mize this. However, sometimes it may be desirable to have
aggregates in simulating certain industrial effluents. Since
the molecular attraction becomes increasingly Important as
the particles become smaller, the small particles will be
harder to separate than larger particles. Additional attrac-
tive force between particles may arise from surface tension
effects, if vapor films are presented on the surface of parti-
cles. Powder particles dispersed in an airstream are known
to carry electric charges, and in some cases electrical forces
will contribute to the mutual adhesion of particles in forming
aggregates. All these factors and the method of redisperslon
certainly would greatly influence the dispersed dust size
distribution in the airstream.
5.3.2 Handling
To minimize the effect of aging as discussed in the previous
section, utmost care must be exercised in the handling of fine
dusts. Several steps involved in the handling of dust are:
(1) Collection — dust must be collected on-site directly
into moisture proof bags or containers with minimum
exposure to atmospheric air. A high efficiency bag-
house or electrostatic precipitator are the preferred
dust collection methods because they maintain the ori-
ginal primary particle size distribution that exists
in the industrial effluents.
(2) Shipping and storage — after the collection of dust
from a desired source the bagged dust should be sent
to the central distribution center for storage in a
humidity controlled room for added protection until
it is needed.
(3) Characterization — every batch of dust returned to
the central distribution center must be characterized
for needed essential Information. The bag must be
tagged for Identification.
Delivery — upon request from user, the bags of cer-
tain dusts could be sent immediately from the distri-
bution center warehouse.
26
-------
5.4 QUALITATIVE EXPERIMENT .ON PLUIDIZED BED
Dispersion of fly ash and submicron iron oxide by a simple
bench-scale fluidized bed was performed in the laboratory for
qualitative study of fluidized bed dispersion techniques. A
schematic of the experimental set-up is shown in Figure 5.
In this system fly ash or iron oxide was injected directly
batchwise into the fluidized bed prior to start-up of air
flow. Plyash was obtained from Dayton Power & Light Company,
while submicron iron oxide was obtained from the Pfizer Company,
Two types of beads (i.e. glass and nickel) both about 100 y
in diameter were used in the experiment (see Figure 6). No
attempts were made to use an electrically conductive column
or radioactive source for charge neutralization. However, a
ground wire was provided whenever nickel beads were used for
the bed.
Air flow was adjusted so that the upward air flow velocities
were much less than the gravitational setting velocity of the
beads in the fluidized bed. The dust had an average aerody-
namic diameter at least one decade smaller than the bed solids
and should have eventually elutriated out of the fluidized
bed. No attempt was made to optimize the fluidized bed design
at this time.
As expected, flyash was relatively easy to redisperse by
fluidized bed using either glass beads or nickel beads as bed
materials. The fluidized bed deagglomerated the flyash and
redispersed most of it out of the bed during the first minute.
There was no tendency for flyash fines to coat or attach to
the glass or nickel beads.
There was visual evidence that the red iron oxides (<1 y)
remained electrostatically attached to (or coated on) the
white fresh glass beads. A coating effect of iron oxides
on nickel beads was not observed.
Photomicrographs of redispersed dust samples on Millipore
filters are shown in Figures 7 and 8. No attempts were made
to measure size distributions of the redispersed particles.
With batchwise feeding of dust into bed solids, the fluidized
bed could have shifted the size distribution of the dust.
However, in a continuously operating dust feeder the degree
of particle classification by a fluidized bed should decrease
with time. Because of the nature of this project, no further
experiments were performed at that time.
Qualitatively speaking, it is possible to redisperse most of
the collected bulk dusts (if not all) from industrial efflu-
ents by fluidized bed. Because of the vastly different dust
27
-------
CO
Vacuum
Regulator
Ground
Wire
Flowmeter
Compressed ^
Air (Dry) Regulator
Millipore Flowmeter
Filter
-2" I.D. Glass Column
Fluidized Bed (Glass or Nickel Beads)
Porous Support (Glass)
Figure 5. Schematic of Bench-Scale Fluidized Bed
-------
I
(a) Glass Beads, 225X
(Microbeads Div., Cataphote Corp.)
(b) Nickel Beads, 225X
(Sherritt Gorden Mines, Toronto, Canada)
Figure 6. Photomicrograph of Pluldlzed Bed Beads
29
-------
(a) Flyash, 500X
WT *.> **
(b) Iron Oxide, 500X
Figure 7. Dispersion Using Glass Beads (^100 u)
30
-------
*•*.
v
(a) Flyash, 500X
'
(b) Iron Oxide, 500X
Figure 8. Dispersion Using Nickel Beads (^100 y)
31
-------
characteristics involved, the extent of difficulty in redis-
persing different collected fine particles from various indus-
tries is not known at this time without further experimental
investigation.
The effects of use of a radioactive charge neutralizer or use
of metallic fluidized bed columns on the fluidized bed redis-
persion technique are also unknown. Further study on the
maximum dust loading obtainable on both a bench-scale or
scaled-up unit is needed.
If the fluidized bed is found to be unsuitable for redisper-
sion of certain types of fine dust, other types of redisper-
sion techniques (such as the jet mill) should be tested.
Systematic testing and evaluation of various dust sources
and redispersing techniques is thus warranted.
As mentioned previously, the characterization of a test dust
under dynamic conditions can be a measure of the efficiency
of redispersion of a bulk powder. A test method employed
previously by us could be of interest in characterization
of dispersability or aerosolizability of fine test dust.
Our re-aerosolization apparatus is described in Appendix D.
32
-------
6. REFERENCES
1. Anon., "A Search for Clean Coking Processes," Steel
Pacts, Number 224, 10-11 (January 1974).
2. Anon., "Criteria for the Application of Dust Collectors
to Coal Fired Boilers," IGCI/ABMA Joint Technical
Committee Survey, April, 1965.
3. Willeke, K., Lo, C. S. K., and Whitby, K. T., "Dis-
persion Characteristics of a Fluidized Bed," Particle
Technology Laboratory Publication No. 217, Particle
Technology Laboratory, Mechanical Engineering Depart-
ment, University of Minnesota, December 1973.
4. Personal Communication, Professor B. Linsky, West
Virginia University, February 1974.
5. Air Pollution Manual Part II - Control Equipment.
American Industrial Hygiene Association, Detroit,
Michigan, 1968.
6. Vandergrift, A. E., et al., "Particulate Pollutant
System Study," Vol. Ill: Handbook of Emission
Properties, Midwest Research Institute, 1 May 1971,
PB 203 522.
7. Davies, C. N., Aerosol Science. Academic Press, Inc.,
New York, N.Y., 1966.
8. Fuchs, N. A., The Mechanics of Aerosols. Pergamon
Press, New' York, N.Y., 1964.
9. Whitby, K. T., Liu, B. Y. H., Husar, R. B., and
Barsic, N. J., "The Minnesota Aerosol-Analyzing
System Used in the Los Angeles Smog Project," Journal
of Colloid and Interface Science, 3J9(1), 136-64 (1974).
10. Irani, R. R. and Callis, C. P., Particle Size; Meas-
urement Interpretation, and Application. John Wiley
and Sons, Inc., New York, N.Y., 1956.'
33
-------
11. Selwood, P. W., Magnetochemistry, second edition,
Interscience Publishers, New York, N.Y., 1956.
12. Davies, R., "Rapid Response Instrumentation for
Particle Size Analyzer - A Review," Part I, American
Laboratory, p. 17-23, December 1973.
13. Davies, R., ibid, Part II, January 1974, pp 72-86.
14. Davies, R., ibid. Part III, February 1974, pp 47-55.
15. Sem, G. J., et al., Thermo-Systems, Inc., "State-of-
the-Art: 1971 Instrumentation for Measurement of
Particulate Emissions from Combustion Sources," Vol.
I, April 1971.
16. Sem, G. J., ibid. Vol. II.
17. Sem, G. J., ibid, Vol. III.
18. Sem, G. J., ibid, Vol. IV.
19. Lapple, C. E., "Particle-Size Analysis and Analyzers,"
Chemical Engineering, 75, 149-156 (1968).
20. Hedley, W. H., et al., "Studies of the Surface Chem-
istry of Solids in Dissemination," Final Report,
Contract DAAA15-68-C-0006, U. S. Army, Edgewood
Arsenal, 1969.
21. Gellman, I., "Methods and Experience in the Measurement
of Submicron Particles in Source Emissions and the
Ambient Air," Atmospheric Quality Improvement Technical
Bulletin No. 62, National Council of the Paper Industry
for Air and Stream Improvement, New York, N.Y., August
1972.
22. Green, H., "The Effect of Non-uniformity and Particle
Shape on Average Particle Size," Journal of the
Franklin Institute, 204, 713-29 (1927).
23. Martin, G., et al., "Researches on the Theory of Fine
Grinding, Law Governing the Connection Between the
Number of Particles and their Diameters in Grinding
Crushed Sand," Transactions British Ceramic Society,
23., 61-109 (1923).
24. Whitby, K. T., and Clark, W. E., "Electric Aerosol
Particle Counting Size Distribution Measuring System
for the 0.015 to ly Size Range," Tellus, XVIII(2),
573-86 (1966).
34
-------
25. Brunauer, S., Emmett, P. H., Teller, E., "The Adsorp-
tion of Gases in Multimolecular Layers," Journal of
American Chemical Society, 6p_, 309 (1938).
26. Hedley, W. H., et al., op. cit., ref. 20.
27. Binning, R. C., et al., "Physical and Colloid Chemical
Research on Agents," Final Report, contract DA18-035-
AMC-136(A), U. S. Army, Edgewood Arsenal, January 1969.
28. Snyder, A. D., et al., "Physical and Colloid Chemical
Research on Agents," Final Report, contract DAAA-15-
68-C-0316, U. S. Army, Edgewood Arsenal (1970).
29. Vandergrift, A. E., et al., op. cit., ref. 6.
30. Hedley, W. H., et al., op. cit., ref. 20.
31. Strauts, C. R. N., Gilfellan, J. H., and Wilson, H. H.,
Analytical Chemistry, Vol. II, Oxford, Great Britain,
The Clarendon Press, 1966, p. 253.
32. Salisburg, S. K., Section 27, "Pulverizers," Kent's
Mechanical Engineer's Handbook, 12th Edition, Wiley
Engineering Handbook Series, New York, N.Y. (I960).
33. Gillespie, T. and Langstroth, G. 0., "An Instrument
for Determining the Electric Charge Distribution on
Aerosols," Canadian Journal of Chemistry, 30, 1056-68
(1952).
34. Kunkel, W. B. and Hansen, S. W., "A Dust Electricity
Analyzer," Review of Scientific Instruments, 21, 308-
14 (1950).
35. Vandergrift, A. E., et al., op. cit., ref. 6.
36. White, H. J., Industrial Electrostatic Precipitation,
Addlson-Wesley Publishing Co., Inc., Reading, Massa-
chusetts (1963).
37. Oglesby, S. and Nichols, G. B., "A Manual of Electro-
static Precipitator Technology, Part I - Fundamentals,"
prepared for NAPCA, contract CPA 22-69-73, Cincinnati,
Ohio, 25 August 1970.
38. Perry, J. H., Chemical Engineers' Handbook, 4th edition,
McGraw-Hill, Inc., New York, N.Y. (1963).
35
-------
39. Kallen, H. P., "Corrosion," a Power special report,
New York, N.Y., December 1956, pp 73-108.
40. Willis, C., "The Complex Refractive Index of Particles
in a Flame," Journal of Physics, Series 2, Part D,
Vol. 3, 1944-56 (1970).
41. Volz, P. E., "Infrared Refractive Index of Atmospheric
Aerosol Substances," Applied Optics, 1^(4), 564-8 (1972)'
42. Volz, F. E., "Infrared Optical Constants of Ammonium
Sulfate, Sahara Dust, Volcanic Pumice, and Flyash,"
Applied Optics, L2(3), 755-9 (1973).
43. Bhatnagar, S. S. and Mathur, K. N., Physical Principles
and Applications of Magnetochemistry, Macmillan and
Co., Ltd., London, 1935.
44. Trade Literature, LDJ Electronics, Inc., Troy,
Michigan, Vibrating Sample Magnetometer.
45. Trade Literature, Princeton Applied Research, Princeton*
New Jersey, Vibrating Sample Magnetometer.
46. Selwood, P. W., et al., op. cit., ref. 11.
47. Vandergrift, A. E., et al., op. cit., ref. 6.
36
-------
APPENDIX A
SELECTION OP DUST CHARACTERIZATION CRITERIA
Selection of criteria for the characterization of fine test
dusts is based primarily on the requirements of the experi-
ment for which the dust will be used. A test dust may be
characterized to virtually any level desired. However,
overclassification can become time consuming, expensive,
and consequently wasteful. While all properties of a test
dust are significant in characterizing a sample, a level
of importance must be established for each property with
regard to the mechanism by which the device being tested
operates. Those properties influencing the production of
the test flue gas (generated or redlspersed) also require
consideration in characterizing the dusts.
Six mechanisms can contribute to the particle collection
efficiency of a control device: gravity, centrifugal force,
inertial impaction, direct .interception, diffusion, and
electrostatic effects. The magnitude of the collection
effect of each mechanism is generally related to particle
size, with smaller particles requiring more effort and cost.
The primary collection mechanisms associated with various
collection devices are listed in Table 7.
Several particle properties potentially useful in charac-
terization are: size distribution; shape, surface area,
and pore volume; chemical composition; density; wettability
and moisture content; solubility; hardness, abrasiveness,
and grindability; charge properties; dielectric properties;
corrosiveness; optical properties; and magnetic suscepti-
bility. Although carrier gas composition and solids load-
ing are not characteristics of fine particles, they too
influence the particle properties and must be considered
in the overall operation of a control device. Typical
values for these properties are listed in Appendix C for
the following sources; pulverized coal combustion, stoker-
fired coal combustion, basic oxygen furnace, open hearth
37
-------
Table 7. FORCES INFLUENCING PARTICLE COLLECTION
IN AIR POLLUTION CONTROL DEVICES5
Class of Air
Pollution Control Device
Settling Chamber
Cyclone
Large Diameter
Small Diameter
Mechanical Cent. Rotor
Scrubber
Simple Spray Tower
Packed Tower
Wet Cyclone
Inertlal-Power Driven
Self-Induced Spray
Venturl
'Filter
High Velocity Impingement
Spun Glass Prefllters
Deep Fiber Bed
High Efficiency Cellulose-
Asbestos or All Glass
Superfine Fiber
Plastic Fiber-Superfine
Cellulose Ester Membrane
Bag or Screen Woven
Fabric
Reverse-Jet Felt
Electrostatic Precipatator
Single Stage High Voltage
Two Stage Low Voltage
Force or Mechanism
Gravity
Centrifugal + Impaction
Centrifugal + Impaction
Centrifugal + Impaction
Impaction + Direct Interception
Impaction + Direct Interception
Impaction + Direct Interception + Centrifugal
Impaction + Direct Interception + Centrifugal
Impaction + Direct Interception
Impaction + Direct Interception
Impaction + Direct Interception
Impaction + Direct Interception
Impaction + Direct Interception + Diffusion
Impaction + Direct
Impaction + Direct
Electrostatic
Impaction + Direct
Electrostatic
Impaction + Direct
Electrostatic
Impaction + Direct
Electrostatic
Electrostatic
Electrostatic
Interception + Diffusion
Interception + Diffusion +
Interception + Diffusion +
Interception + Diffusion +
Interception + Diffusion +
"Minimum particle size collected at approximately 90!? efficiency under
usual operating conditions.
5. Air Pollution Manual Part II - Control Equipment, America!?
Industrial Hygiene Association, Detroit, Michigan, 1968.
38
-------
furnace, electric arc furnace, metallurgical coke ovens,
cement plants, municipal incineration, iron foundries,
kraft pulp mill recovery furnaces and asphalt and other
crushed stone operations.
In many cases particle properties are interrelated; a
change in one results in a corresponding change in others.
Table 1 summarizes the relative importance of each property
with respect to conventional control devices, redispersion,
and size determination mechanisms. Following are discus-
sions of each of the listed properties and their relation-
ship to the three conventional methods of fine particle
collection, redispersion, and common instrumentation types
for size analysis.
A.I SIZE DISTRIBUTION
Particle size distribution is the most important property
for the characterization of a test dust. The efficiency
and applicability of a particle removal technique depends
to a great extent on the size of the particles to be col-
lected. Size distribution is also of primary importance
for particulate redispersion efforts and size-measuring
instrumentation.
Particle size generally cannot be specified by one para-
meter. For irregular particles, the average dimension along
three mutually perpendicular axes may be used, or the dia-
meter of a sphere of the same volume or same surface area
may be used. For extremely irregular particles other prop-
erties such as settling velocity are of more significance
than actual size and shape. Some idea of the variety of
shapes and sizes of primary particles in smokes is given
by the electron photomicrographs shown in Figure 9.
When dealing with dust effluents from industrial sources,
distinction must be made between the size and shape of
primary particles and that of the aggregates formed-from
them. Aggregates in effluents arise from the coagulation
of individual particles, and also from incomplete disinte-
gration of powders during their dispersion. Size distri-
bution should be determined by sampling a gas stream where
the actual level of agglomeration is similar to that of
the original flue gas.
A.2 SHAPE. SURFACE AREA, AND PORE VOLUME
As mentioned above, particle shape can Influence the meas-
urement of particle size. Shape and surface condition can
also influence particle handling characteristics, chemical
reactivity, adsorption potential, and flammability limits.
39
-------
MgO
ZnO
Al
3722-1
Aggregates in Fe203
Smoke
t •
-i
1C*
Exploding Wire
3722-2
C (Carbon black)
Figure 9. Electron Photomicrographs of Aerosol Particles
in Smokes
-------
As particle size is reduced, surface properties predominate
over bulk properties in controlling the physical and chem-
ical behavior of finely divided solids. As a consequence,
the surface area and distribution of pore sizes becomes of
utmost importance owing to the exaggerated tendencies for
adsorption of gases and moisture, and the extent to which
the particle can be electrically charged by triboelectric
phenomena.
Shape is of secondary importance for high energy wet scrub-
bing, electrostatic precipitation, and redispersion, while
it is of primary importance for fabric filtration where the
shape of the. collected particles is critical to the cleaning
cycle of the filter operation. Pine needle-shaped particles
have a tendency to remain trapped in the filter fabric dur-
ing the cleaning cycle, thereby plugging the filter and re-
ducing its operating efficiency.
Shape, surface area, and pore volume area of varied impor-
tance to the operation of different types of particle sizing
equipment. The orientation of irregularly shaped particles
is of primary importance to the size distribution obtained
from an optical measuring device. Surface area, as it in-
fluences a particle's ability to be charged, is very impor-
tant to electrical size measuring equipment, utilizing a
particle charging mechanism. For mechanical and physico-
chemical size analyzers, shape, surface area, and pore
volume are only of secondary importance. The shape of a
particle can influence its aerodynamic behavior in a mech-
anical sizing instrument, while the surface properties of
a particle may either aid or hinder vapor condensation on
particles in a physico-chemical sizing technique.
A.3 CHEMICAL COMPOSITION
The chemical composition of a dust particle is a very reliable
method of characterizing the dust and identifying its origin,
but the mechanisms of particle collection are physical and
electrostatic in nature, not chemial. Consequently, chemical
composition is only important to the extent that it influences
the physical and electrical properties of the dust particle.
The same is true for size measuring techniques. The mechan-
isms are based on physical properties of. the particles, not
chemical, but the chemical composition of the particle is a
prime factor in determining what the important physical
properties are.
A'21 DENSITY
Both bulk and real particle density are of primary impor-
tance for handling and collection of fine dusts. The
-------
effort required to redisperse collected dusts will be
influenced by the particle density. Once dispersed,
density plays an important role in the collection mech-
anisms of impaction, direct interception, and diffusion,
all of which are important to the three collection devices
being considered.
Mechanical, optical, and thermal size measurement techniques
all utilize inertial impaction. Consequently, particle
density is important for their operation. For mechanical
devices, density is more important than for electrical or
thermal devices because inertial impaction is the primary
sizing mechanism. The effective density of the particles
is utilized for size classification by accelerating the
particles (e.g., centrifugally or through an orifice),
while for the other two types of device, inertia acts only
as an assist to the electrical or thermal force exerted on
the particle.
A.5 WETTABILITY AND MOISTURE CONTENT
Moisture content is the result of the wettabillty of a
particle. Wettability, in turn, depends on the nature of
the dust particle, and is related to boundary surface ener-
gies. Wetting occurs when solid/liquid adhesion energy
exceeds liquid cohesion energy.
This phenomenon can affect particle collection in wet
scrubbers, although it is generally only a second-order
effect. The collection efficiency may be improved by the
increased particle size resulting from moisture collection
on the particle surface.
Moisture in dust particles will contribute to increased
agglomeration of particles collected for redispersion, ad-
'ding potential difficulties to the redispersion operation.
Once dispersed, a dust of high moisture content will have a
tendency to cake on the collection surface, thus inhibiting
further collection. Moisture content also exerts a strong
influence on the particle resistivity, flammability, and
handling characteristics.
Wettability is of prime importance for particle sizing via
physico-chemical techniques. For example, the operation of
a nuclei condensation particle sizing instrument is based
on the condensation of water vapor on fine particles, en-
larging their size for ease of observation. Should a part-
icle not be wettable, e.g. DOP particles, this type of size
measuring device will not be applicable.
-------
A.6 SOLUBILITY
Solubility of fine particles In water is of secondary im-
portance for conventional fine particulate collection
devices. For wet collection devices, solubility is of most
concern with regard to operating conditions, i.e., recycle
ratio. Another possible Influence of particle solubility
on wet collection is a change in the collection efficiency
due to soluble components of collected particulate In the
recycle liquor. Dry collection devices can be adversely
affected by solubilized particulate. Soluble particulate
components can accelerate equipment failure by corroding
collection surfaces of an electrostatic precipitator or by
contributing to bag failure in a fabric filter.
A.7 HARDNESS. ABRASIVENESS, AND GRINDABILITY
These three properties are not related to any collection
mechanism and consequently are not of primary importance.
However, they are of importance with respect to equipment
lifetime and selection of materials of construction. Of
the conventional collection devices, fabric filters are
most susceptible to deterioration and failure from contin-
uous bombardment by abrasive particles.
Another area in which hardness, abrasiveness, and grinda-
bility play an important role is in the production of a
test flue gas. To redisperse a collected dust sample re-
quires an input of energy. Supplying adequate energy for
redispersal may also provide enough energy to break the
particles, yielding an aerosol of different size distribu-
tion from the original flue gas.
A.8 CHARGE PROPERTIES
Most industrial dusts are charged to an appreciable extent,
but the fractions of positively and negatively charged
particles are similar so that an overall flue gas is gen-
erally neutral. Mechanisms that contribute to the charging
of dusts include electrolytic mechanisms, contact potential,
spray electrification, contact separation, ion diffusion,
and electric field charging processes in gases. The con-
trolling factor limiting the level of charging of solid
particles is the surface area of the particle, while for
liquid droplets the controlling factors are size and surface
tension. The electrostatic characteristics of the charged
particles are utilized to facilitate collection. Both the
primary particle capture and consequent collection are en-
hanced by electrostatic attraction.
-------
The ability of a particle to become charged is of secondary
importance for conventional particle collection equipment.
For electrostatic precipitators particles must be able to be
charged for collection but the ability to drain that charge
is more critical. Collection efficiencies for both wet
scrubbers and fabric filters may be enhanced by both natural
and induced particle charging. The charge properties of
particulates play an important role in redispersion attempts.
Charged particles tend to agglomerate when redispersion is
attempted, and when fluidized bed dispersion techniques are
used, charged particles tend to adhere to the bed material.
Both electrical and mechanical size measuring techniques
are influenced by particle charge properties. The primary
importance of charge properties on electrical sizing equip-
ment is self explanatory; the mechanism by which an elec-
trical particle sizing device operates depends on the force
exerted on charged particles in an electric field. Mechan-
ical sizing equipment operation can be adversely affected by
charged particles. In devices where particles have a tenden-
cy to become charged (e.g., cascade impactor where high
speeds and particle collisions occur), collection efficiency
may be hampered by collection surfaces becoming charged and
subsequently repelling particles. Particle charging char-
acteristics are important to the study of fine particle be-
havior and collection.
A.9 DIELECTRIC PROPERTIES
Of more critical importance is the capability of the parti-
cles to conduct electricity and drain their charge once
collected. If the charge cannot be drained, an electro-
static barrier to further particle deposition may form.
For normal electrostatic precipitation a maximum resistivity
of 2 x 1010 ohm-cm (measured in situ} is required.6 There
are a number of factors that influence the apparent resis-
tivity of particulates. Some of the important factors are
particle size distribution and shape, temperature, surface
energy characteristics, packing configuration, and chemical
composition. The resistivity is also a function of carrier
gas temperature and composition.
The electrical resistivity of a bulk layer of particles
depends on both surface and volume factors. Below 300-350°P
surface conductivity predominates, with electrical charges
being conducted along the particle surface through adsorbed
6. Vandergrift, A. E., et al., "Particulate Pollutant
System Study," Vol. Ill: Handbook of Emission
Properties, Midwest Research Institute, 1 May
1971, PB 203 522.
-------
chemical films. At higher temperatures volume conduction
due to temperature excitation of internal electrons pre-
dominates.
While size measuring techniques are similarly affected by
both charge and dielectric properties, dielectric properties
are more important with regard to particle collection mech-
anisms. The ability to drain charge is important to all
particle collection equipment but is of primary importance
for dry collection mechanisms. Particle charge enhances
collection, but if the charge is not easily drained from
the particle on the collection surface of either an electro-
statis precipitator or fabric filter, an electrostatic bar-
rier, to further particle collection may form.
A. 10 CORROSIVENESS
Corrosiveness is of primary importance for wet scrubbers,
and only of secondary importance for dry collectors. This
is due to the fact that in dry collectors the solid/solid
contact between a corrosive material and the collection
surface is not as intimate as it is in the solutions formed
in wet scrubbers. Also, corrosiveness is primarily due to
carrier gas components (for the industries of interest) with
particle composition contributing a minor effect. The po-
tential corrosive nature of a wet slurry must be considered
in the selection of appropriate materials of construction.
However, if the carrier gas has high moisture content, cor-
rosive compounds adsorbed on the particle surface may come
into close contact with the collection surfaces of dry
collectors, producing a corrosive effect, reducing precip-
itator life, or contributing to bag failure in a fabric
filter. Corrosiveness is a minimal problem for redispersion
efforts since dusts are,generally kept dry to inhibit ag-
glomeration.
Corrosiveness will have minimal effect on most sizing instru-
ments. However, the corrosive nature of particulate could
be enhanced in physico-chemical measuring equipment due to
the high moisture content required by nuclei condensation
devices to enlarge the particles by condensation.
-------
A.11 OPTICAL PROPERTIES
Optical properties are of interest primarily in particle
sizing instrumentation. Photophoresis is the force
imparted to fine particles when exposed to illumination.
The photophoretic velocity at illumination of an intensity
comparable to sunlight at sea level is of the order of
10~* cm/sec.7
Optical instruments may utilize either the transmission
or the scattering of light to determine either the con-
centration or particle size of an aerosol. The particle
refractive index is very important in the determination
of particle size using light scattering techniques. As
can be seen in Figure 10 the refractive index of a par-
ticle can influence the apparent size measurement for
fine particles and introduce a significant error in size
determination efforts. Physico-chemical sizing techniques
are somewhat influenced by optical properties due to the
use of optical techniques after the particles have been
enlarged by vapor condensation on the particle surface.
A.12 MAGNETIC PROPERTIES
Magnetic susceptibility may indirectly influence particle
collection efficiency by enhancing the tendency for
magnetic particles to agglomerate, increasing their size
and improving collection efficiency. The use of magneto-
phoretic mechanisms for novel particulate collection
equipment has also been suggested.
Any increased tendency toward agglomeration will be
critical with regard to redispersion of collected dusts.
The force required to break an aggregate in which there
are magnetic forces between particles is 100-1000 times
greater than for an aggregate held together by only
molecular forces.8
The use of magnetic susceptibility in instrumentation
as a method for size classification is of limited appli-
cability. Studies of the effect of particle size on
7. Davies, C. N., Aerosol Science, Academic Press, Inc.,
New York, N.Y., 1966.
8. Puchs, N. A., The Mechanics of Aerosols, Pergamon
Press, New York, N.Y.,
-------
0 2.4 6 8 10
EQUIVALENT LATEX-SPHERE DIAMETER OF
PARTICLE INDICATED BY OPTICAL COUNTER,/zm
*Polystyrene Latex
Figure 10.
Calculated Effect of Particle Refractive Index
on the Calibration of the Optical Counter9
9. Whitby, K. T., Liu, B. Y. H., Husar, R. B.s and Barsic,
N. J., "The Minnesota Aerosol-Analyzing System Used in
the Los Angeles Smog Project," Journal of Colloid and
Interface Science, 39.CD, 136-64 (197*0.
-------
magnetic susceptibility have shown a direct relationship
for some materials (e.g., graphite), but in general the
magnetic susceptibility of powdered metals is independent
of particle size.10'11
A.13 CARRIER GAS COMPOSITION AND SOLIDS LOADING
These two items are not properties used to characterize
test dusts, but they are both of concern for the testing
of fine particle control devices.
The carrier gas composition is of interest mainly as it
affects the physical characteristics of the particles.
For example, allowing sufficient contact time, the humi-
dity of the carrier gas is directly related to the moisture
content of the dust particles. The moisture content and
other adsorbed species, especially S03, also affect the
electrical properties and handling characteristics.
Finally, components of the carrier gas may be corrosive
in nature and contribute to the corrosiveness of a wet
scrubber slurry.
Although solids loading is not a particle property it is
essential in the selection of a control device or system.
Solids loading is of particular interest for redispersal
of collected dusts with regard to the capacity required
for the redispersal unit.
10. Irani, R. R. and Callis, C. P., Particle Size;
Measurement Interpretation, and Application, John
Wiley and Sons, Inc., New York, N.Y., 195b.'
11. Selwood, P. W., Magnetochemistry, second edition,
Interscience Publishers, New York, N.Y., 1956.
-------
APPENDIX B
DUST CHARACTERIZATION TECHNIQUES
The following is a discussion of potential methods for
characterizing test dusts. Techniques for obtaining
values for the properties discussed in Appendix A are
presented.
B.I . PARTICLE SIZE DISTRIBUTION
Particle size is of great importance for the characteri-
zation of test dusts since size is usually the character-
istic that most markedly affects the behavior of particulate
materials. The need for characterization of the particle
size distribution of test dusts could occur in two alternate
stages of use:
a. Laboratory analysis for primary particle size
distribution
b. Size distribution for aerosolized (or stack)
conditions
Characterization in the first case is'addressed to obtain-
ing information concerning the size distribution of the
bulk test powder. This laboratory measurement yields
information on the size distribution of primary particles.
Characterization in the second case will define the test
dust behavior under dynamic test conditions, i.e., a
measure of the efficiency of redispersion of the bulk
powder and the extent of agglomeration or aggregation of
the primary particles. This analysis must be performed on
aerosolized dust, preferably at the test facility using
the actual test flue gas. For simulation of industrial
sources the actual size distribution is of secondary im-
portance compared to whether the size distribution is
similar to that which exists in the actual industrial
-------
effluent. This comparison requires previous knowledge
of particle size distribution in the actual industrial
effluent.
B.I.I Particle Sizing Techniques
Davies12*13'lu has recently presented a detailed review of
the physical principles employed to sense and size particles,
This review was directed to rapid response instrumentation
for particle size analysis and listed the following particle
sensing concepts:
1. Image replication
2. Electrical resistance charge
3. Light scattering
4. Light obscuration
5. $-ray attenuation
6. X-ray absorption
7. Acoustic interference
8. Ultrasonic attenuation
9. Differential flow classification
10. Pulsating gas flow
11. Pressure drop in nozzles
12. Transient and evaporative cooling of a hot wire
13. Impact and momentum measurement
14. Magnetic flux variations
15. Electrostatic ion capture
A four-volume publication by G. J. Sem et al.l5»1*>l7>l8
discusses the state of the art of instrumentation for the
measurement of particulate emissions from combustion
sources. A summary of types of size analysis techniques
appears in Table 8.
12. Davies, R. , "Rapid Response Instrumentation for Par-
ticle Size Analyzer - A Review," Part I. American
Laboratory, p. 17-23, December 1973-
13. Davies, R., ibid, Part II, January 1974, pp 72-86.
14. Davies, R., ibid, Part III, February 1974, PP 47-55-
.15. Sem, G. J., et al., Thermo-Systems, Inc., "State-of-
the-Art: 1971 Instrumentation for Measurement of
Particulate Emissions from Combustion Sources, Vol I,
April 1971.
16. Sem, G. J., ibid, Vol II.
17. Sem, G. J., ibid, Vol III.
18. Sem, G. J., ibid. Vol IV.
50
-------
Table 8. SUMMARY OP TYPES OP SIZE-ANALYSIS TECHNIQUES19
»io-«M>lo)l«Mlll| Pfagarly
TVF-
Goonwlric
Mtchoniccl
or Dynamic
(in liuidt)
Optical
Electrical
Magnetic
Thermal
Physico-
chemical
ClucKtator
MackMlHl
Intrtlo
Terminal settling
velocity (in
gravity or contri*
fugol lieldt)
(possibilities af
using tltetrical
litlot t>iil)
Diffusion
Imagintt
Tranemlseion
(>poc trail
Scattering
Diffraction
Rotlotonco
Copefltoriee
Charge
Applicable to
magnetic
material* only ,
Particle dopotltlon
Condensation
Tt
Sieving
ehnlqu* and Vorlitlont
Wei
Dry
Ullrafilnotion
Impoction on surface
Protsurt pulse (tonic)
Elutriation
Single
fraclionalion
Seriet
fraclionorion
Layer
mothads
(dilferonllal)
Suspension
method I
Decontotian (In liquid only)
Liquid
Gas
Liquid
Got
Liquid
Gas
Differential-
liquid
BX
Intogrol-gat
iPnoto-todlmentotlon (photography of porllclo irroakol
Particlo ditplocomtnt (random walk)
Particle dtpetlllen
Light mlcraieofr
Ultramlcrotcopy (gives mean lilt only)
Clactron microscopy
EntlneWen meawrod at function at wavotanglh
Single particle
count
Right angle (90')
Angular
Forward
Polarlialion
Mocrofcopic (gives moan tile only)
Light
X.'ray
Loser (holography.rtcons tructton of diffraction
patttrn)
Altorotlen of currant flow by paitleioi
Polantial pulto duo to particle deposition
Trlbt-chorglng
Induction charging
Corona charging
Particle migration In magnetic fields; magnetic
pulsti
Particle migration in thermal gradient
MICIMS
5-1000
0.01-5
0.1-100
10-10001?)
5-100
0.02- IWQ
0.002- 1(UC)
1-100(6)
1-100(0)
0.01-1
0.2-100
0.005-1
0.002-15
0.1-21?)
0.2-50
(higher with
microwave)
MOO
1-30
O.I.KX?)
Growth ol nocle! with controlled tupertaturotlon 0.01-0.1(7)
• Item. In parenmetot hava following ilgnljicance! C • in centrifugal (laid; G > in gravity Hold; UC » c...
t Ropllcat may be vied hi place ol portlclat that might avoporala or bo deitroyod during meetureiMnt, e.g., In electron mlcrotca|iy topvoM the effect of vacuum
or electron beams, and In magnesium onide film method for measuring ilia or drops
.19. Lapple, C. E.s "Particle-Size Analysis and Analyzers,"
Chemical Engineering 75., 1^9-156 (1968).
51
-------
Table 8 shows that the equipment required depends on the
size range of the dust particles. For simulation of
industrial sources it is apparent that the wide range of
sizes present from different sources and within each source
(See Appendix C) would necessitate the use of multiple
pieces of sizing equipment.
B.I.2 Primary Particle Sizing
The primary particle size distribution can be obtained in
the laboratory using a sample of collected dust. The most
common method for this type of measurement is microscopy.
For optical microscopy the size-limiting factor (^0.5 u)
is controlled by the wavelength of the light used. As
the particle size approaches the wavelength of light used,
the particle cannot be resolved. Smaller particles can,
however, be detected using electron microscopy.
In both cases counting aids may be used to facilitate size
distribution determination. A mechanical aid, shown in
Figure 11, has been designed by MRC for simple size dis-
tribution calculations using negatives obtained by micro-
scopic techniques.io The images of the particles are
projected down upon tracing paper by a spot projector
located above the table using one of several calibrated-
diameter beams of light. The sizes of the projected beam
of light are matched to the projected particle through the
paper. When the size of the projected spot matches that
of the particle, a foot switch is pressed to record the
particle size in a bank of counters. Another method for
speeding the counting operation is to incorporate automatic
computer counting with the microscopic detection instru-
ment, for example, scanning electron micrograph.
Microscopy is an ultimate size measuring technique where
the actual physical size of the particle is the parameter
measured. Because of this direct observation, microscopic
size determination is frequently used as a primary standard
for the calibration of other sizing equipment.
Unfortunately, microscopy has disadvantages for generating
a particle size distribution. The primary disadvantage is
the time required to obtain results. Microscopy is not a
real-time particle sizing technique, and preparation of a
sample for analysis is an extensive and tedious operation.
20. Medley, W. H., et al., "Studies of the Surface
Chemistry of Solids in Dissemination," Final Report,
Contract DAA15-68-C-0006, U.S. Army, Edgewood Arsenal,
1969.
52
-------
LO
12 Position
Rotary Switch
Super imposition
Optics Assembly
Viewing Table
Tracing Paper
35 mm Projector
I 0
Slide Mount
12 Position Switch
Front Surface Mirror
o o a o o o
a a a ao a
o o o a
I Aperture
I Assembly
Foot Switch
Figure 11. Schematic Diagram of Particle Size Comparator
-------
B.I.3 Aerogglized-Slze-Distributlon
Potential combinations of sensing and sizing equipment
which appear to be technically feasible and which also
appear to be most practical for application to effluent
testing are listed in Table 9. Not all combinations In
Table 9 are equally applicable, however, and most combin-
ations have not yet been developed.
Figures 12 and 13 present potential combinations of
equipment for determining the size distribution of fine
particulates on a continuous basis.
To reduce the concentration of particles to within the
limits of the sensing equipment, flue gas dilution capa-
bility is required. The dilution chamber also prevents
condensation of moisture in the flue gas, which could
change particle size or damage equipment.
Photoelectric or optical particle counters function on
the principle of light scattering; that is, each particle
In a continuous flowing stream passing through a beam of
light scatters light onto a phototube during the time the
particle is illuminated. The amount of scattered light
is proportional to the square of the diameter of the
particle. The overall system is usually calibrated in
terms of particles per unit volume of unit size. A count
is indicated for all particles passing through the Instru-
ment whose sizes are larger than a selectable preset
minimum. The photoelectric particle counter will give
reliable size distribution information up to concentrations
of ^1000 particles per cubic centimeter. Above this
concentration limit, the frequency of occurrence of more
than one particle in the view volume of the photometer
will affect the measured size distribution; these particles
will be counted as a single larger particle. This range
can be extended with an Internal dilution system. This
type of equipment Is applicable for particles from 0.2 u
to the cyclone cut point of 10 y.
Condensation nuclei counters function on the ability of
particles to act as centers (nuclei) for the condensation
of water vapor in a supersaturated environment. The
technique Is used for particle measurements in the 0.002
to 0.3 micron range. The concentration of the grown par-
ticles is then measured by attenuation or scattering of
light. Particles in this size range are frequently
referred to as condensation or Aitken nuclei. In conden-
sation nuclei detectors, a sample is withdrawn from the
aerosol stream, humidified, and brought to a supersaturated
condition. In this supersaturated condition, condensation
-------
U1
Table 9- OPERABLE SIZE RANGES OP PRACTICAL COMBINATIONS OF SIZE CLASSIFIERS AND SENSORS
All numbers are the diameter, in microns, of a spherical particle
with a density of 1.0 g/cu cm13
01
u.
to
CO
u
UJ
M
CO
UJ
u
tc
AERODYNAMIC
IMPACTOR
0.2*"- 30
CYCLONE
05-3O
GRAVITATIONAL
ELUTRIATOR
I.O - 100
GRAVITATIONAL
SEDIMENTATION
1.0 - 50
ELECTROSTATIC
0.005-0.6
BROWNIAN
DIFFUSION
0.001-0.05
CONCENTRATION SENSING TECHNIQUE
MASS
BETA RADIATION
ATTENUATION
0.01 -100
0.2* -30
0.5* -30
1.0* -100
1.0* -50
-
-
PIEZOELECTRIC
QUARTZ CRYSTAL
0.01 -20
0.2* -20
05* -20
1.0* -20
l.0*-20
—
— .
OPTICAL
LIGHT
TRANSMISSION
0.2 -100
O.2 —30
0.5—30
1. 0-100
1.0 — 50
— -
- .
PHOTOMETRY
0.2 -50
_. — „ 1
0.2 —30
0.5 — 30
1.0—50
1.0—50
—
—
-i
Z Z |
o od
co a.
0.2 — 30
05—30
1.0— 5O
1.0—50
—
• —
ELECTROSTATIC ION
CAPTURE AND
ATTENUATION
0.005-100
0.2 — 30
05—30
1.0—100
1.0 — 50
0.005 — 0.6
- .
NUCLEI COUNTER
0.001 - 0.05
_
' -
—
-
0.005 — 0.05
0.001 — 0.05
*Although this is the lowest size cutoff of the classifier, the sensor can detect smaller particles below this
size cutoff, lumping them into one size range.
**Low pressure impactor may prove useful down to 0.05 microns for some applications where volatile particles are
not present.
13- Sem, G. J., ibid, Vol. III.
-------
Flowmeters
Cyclone Pump
Process
Exhaust
Line
Neutralizer
Diffusion
Battery
Nuclei Counters
Particulate
Sample Line
Optical Particle
Counter
L-- Diffusional Dryer
,7 |K (Optional)
// nilntinn X
T\ Device
Charge
Neutralizer
Manometer /^^\
Pressure
Balancing
Line
Recirculated
Clean Dilution
Air
Filter
Pump
Bleed
Figure 12. Schematic of Potential Particle Size Distribution
Measurement System Utilizing a Diffusion Battery
-------
Flowmeters
Cyclone Pump
Diffusion
Battery
Process
Exhaust
Line
Neutralizer
Electrical Particle
Counter
Diffusional Dryer
(Optional)
Partial late
Sample Line
Nuclei Counters
Charge
Neutralizer
Pressure
Balancing
Line
Recirculated
Clean Dilution
Air
Filter
Pump
Bleed
Figure 13.
Schematic of Potential Particle Size Distribution
Measurement System Utilizing an Electrical
Particle Counter
57
-------
will be initiated on all particles larger than a certain
critical size and will continue as long as the sample is
supersaturated. This condensation process forms a homo-
geneous aerosol, predominantly composed of water, containing
one water drop for each original particle (condensation
nucleus) whose size was greater than the critical size
appropriate to the supersaturation obtained. Greater
supersaturation is used to initiate growth on smaller
particles. The number of grown particles is determined
from the characteristic optical properties of the final,
grown aerosol, with the scattered light method described
earlier being commonly used.
A diffusion battery consists of a number of long, narrow,
channels, an example unit being about 20 parallel channels,
0.1 mm wide, 10-12 cm high and 10 cm or more long whose
dimensions are adjusted to permit variations in length
and spacing as a means of particle size measurement in a
selected size range. Gravity and particle density have
a negligible effect on particles below a diameter of
0.6 micron. As the aerosol moves in streamline flow through
the long narrow rectangular channels, the random Brownian
movement of the particles causes them to displace from
their original position in the air flow streamline. The
most probable displacement from a streamline is zero, but
the "average displacement is proportional to the square
root of the travel time. Consequently, some of the particles
are displaced sufficiently to reach the wall of the battery.
It is assumed that once the particle reaches the battery
wall, it will adhere. Therefore, a fraction P of the in-
fluent particles appear in the battery effluent. Since the
magnitude of the Brownian movement increases with decreasing
particle size, the smaller the particle size, the lower
the value of P. Prom -the value of P, the diffusion constant
D of the aerosol may be obtained and the particle size
calculated.
By selecting one of a series of sizes of diffusion batteries
all particles less than a selected minimum size can be
measured. The percent penetration of particles through
the diffusion battery (particle concentration in divided
by particle concentration out) is determined from particle
concentration measurements made using a condensation
nuclei counter described earlier.21
21. Gellman, I., "Methods and Experience in the Measurement
of Submicron Particles in Source Emissions and the
Ambient Air," Atmospheric Quality Improvement Technical
Bulletin No. 62, National Council of the Paper Indus-
try for Air and Stream Improvement, New York, N.Y.,
August 1972.
58
-------
Another potential sizing train replaces the diffusion battery
with an electrical particle counter. The sizing range of an
electrical particle counter is 0.005 to 0.6 y. It operates
on the principle that distinctly different charges can be im-
posed on different sized aerosols and the mobility of these
particles differs and can be measured. A combination optical
counter, electrical, and nuclei counter has been used by Uni-
versity of Minnesota personnel to determine the number size
distribution of an aerosol from 0.002 to 10 y (Figure 14.}
The time for a measurement cycle is about four minutes.
Numerous other alternatives are available, each having advan-
tages and disadvantages. However, the systems used at any
specific laboratory facility are usually based on the avail-
able equipment at that location to minimize capital expendi-
ture and labor.
B.2 SHAPE, SURFACE AREA, AND PORE VOLUME
H. Green22 and G. Martin, et al 2 3 have shown that for Irreg-
ularly shaped particles of a material having a statistical
diameter ds and a volume V, V/ds3 is constant. The various
particles of a sample to which ds applies will possess diff-
erent individual volumes but the total volume of all the
individual particles for any distribution of particles having
the same average diameter will be constant. If the particles
are individual spheres of different sizes, the value of the
constant will be ir/6 (e.g., V = ir/6 d3).
However, for irregularly sized particles, the value of this
constant is less than 7T/6. This reasoning may also be ex-
tended to surfaces, so that if S is the aggregate surface
area of a group of particles possessing an average diameter
ds, then S/ds is constant. The value of this constant for
spheres of different sizes would be ir (e.g., S = ud3), while
the value of the constant for irregular particles would be
less than IT. If the volume const'ant is denoted by ocv, and
the surface constant by as, then:
22. Green, H., "The Effect of Non-Uniformity and Particle
Shape on Average Particle Size," Journal of the Frank-
lin Institute, 204, 713-29 (1927).
23. Martin, G., et al., "Researches on the Theory of Fine
Grinding, Law Governing the Connection Between the
Number of Particles and their Diameters in Grinding
Crushed Sand," Transactions British Ceramic Society,
23, 61-109 (1923).
59
-------
10'
I06
Ul
CD
2
3
10*
10
1.0
O.I
O OUTDOOR AIR SAMPLE
A INDOOR AIR SAMPLE
€>A NUCLEI COUNTER
OA ELECTRICAL COUNTER
• A ROYCO COUNTER
.001 .01 O.I 1.0 10
PARTICLE DIAMETER (Dp) MICRON
100
Figure 14. Size Distributions of Indoor and
Outdoor Aerosols21*
Whitby, K. T., and Clark, W. E., "Electric Aerosol
Particle Counting Size Distribution Measuring System
for the 0.015 to ly Size Range," Tellus, XVIII(2),
573-86 (1966).
60
-------
The terms a and a are called volume and surface shape-
factors, respectively.
One method for determining a and a for test dust samples
would employ results of other characterization tests. These
tests include optical counting of average Martin's diameter
on a field of N particles (^400), air pycnometer measure-
ment of crystal density, and BET measurement of the surface
area in m2/g. The average Martin's diameter is defined as
the average of the diameters of a specified number of par-
ticles along a line equally dividing the bulk of the parti-
cles.
With a measure of crystal density, e, the total weight of
the equivalent spheres can be calculated;
wsph • "/6 V
The actual surface area in m2/g can then be determined by
the BET method. This value SBET, would then be normalized
to the weight of the N equivalent spheres and then divided
by the square of the statistical diameter to yield the
value for a , the surface-shape factor. For example:
SBET Wsph = S(N particles)
S(W particles) =
The shape factor does not represent a measurement of shape,
but from the manipulation of other property measurements,
already used to characterize the particles, a constant
indicative of particle shape can be obtained.
A classical method of surface area analysis is based on
the BET theory.25 This analysis is conducted by determin-
ing the quantity of gas necessary to form a single layer
of gas molecules on the surface of the powder. One instru-
ment utilized for this type of measurement is the Numinco-
Orr Model MIC-103 Surface Area-Pore Volume Analyzer. Using
nitrogen gas this instrument measures surface areas in the
25. Brunauer, S., Ernmett, P. H., Teller, E., "The
Adsorption of Gases in Multimolecular Layers,"
Journal of American Chemical Society, 60, 309 (1938).
61
-------
range from 1-200 m2/g. Employing krypton, surface areas
below 1 m2/g are accurately measured. By measuring the
rate of desorption of the gas, pore size distributions
between pore radii of ten to several hundred angstroms
can be measured.
B.3 CHEMICAL COMPOSITION
The chemical composition of test dusts simulating station-
ary source emissions is largely inorganic in nature, but
in some cases (such as asphalt batch plants and coke oven
emissions) a significant carbonaceous or organic chemical
component can be present. Most analytical chemical labor-
atories possessing a full range and variety of sophisticated
instrumental and wet chemical analysis capabilities can
adequately characterize test dusts. The challenge repre-
sented in chemical characterization of particulate is to
develop an analytical protocol that will yield the most
information with the least amount of effort. One potential
protocol is presented in Table 10.
The protocol accentuates instrumental rather than wet
chemical analysis and depends to a large extent on the use
of nondestructive analysis approaches, thereby minimizing
the total amount of effort in both analysis and sampling.
The nondestructive methods employed are x-ray fluorescence
and x-ray diffraction techniques.
An instrument such as the EDAX-EXAM computerized x-ray
fluorescence system is employed first for elemental analy-
sis. The EDAX-EXAM system provides simultaneous analysis
and display of all elements from fluorine.to uranium,
qualitative analysis in seconds, sensitivities to better
than 0.1 ug/ctn2, quantitative analysis in minutes (% to
ppm), matrix corrections of interelement effects and
conversion of raw data to percentages, ppm, yg/cm2, or
other relevant concentration values.
The x-ray fluorescence elemental analysis is followed by
molecular analysis of crystalline components by x-ray
diffraction employing a General Electric XRD-3 diffracto-
meter and Debye, Scherrer, and Laue back-reflection cameras.
This system has recently been used in our laboratory to
characterize the composition of particulate collected from
an asphalt batch plant where high-vol filter samples were
compared to baghouse dust and the filter, sand, and
aggregate feed materials. The relative percentages of
dolomite, quartz, and calcite were obtained via x-ray
diffraction.
62
-------
Table 10. PROTOCOL FOR ANALYTICAL ANALYSIS
I. Teflon Filter
II. Glass or Quartz Fiber Filter
Measure weight Increase
Non-destructive XRF
for metal and non-metal
elements
Measure weight increase
Non-destructive XRF* for
non-metal elements, e.g.,
S, and for some non-metal
elements
Non-destructive XRD* for
crystalline components
Portions of filter for micro
C-H-N analysis - Oxidation
and Pyrolysis Modes. Yield-
ing total C, H, N. carbonate,
nitrate, water, organic C-H-N
Selected portions** of filter
for wide-line NMR, SEM,
electron probe, DTA, TGA,
camera XRD, wet chemical
sulfur analyses, IR
*Non-destructive XRF and XRD analyses of particulate on
filters are optional, but recommend for initial studies.
**Measurements to be'made on selected specimens, if necessary,
after evaluation of basic XRF, XRD, and C-H-N analyses.
XRF - X-ray fluorescence
XRD - X-ray diffraction
NMR - Nuclear magnetic resonance
SEM - Scanning electron microscope
DTA - Differential thermal analysis
TGA - Thermo'gravametric analysis
IR - Infrared
63
-------
Following these nondestructive analyses the samples are
analyzed using an F&M Model 185 CHN Analyzer. Determin-
ations of total carbon, organic hydrocarbon, inorganic
carbon, nitrogen, and water are made by operating the micro
carbon-hydrogen-nitrogen analyzer in two modes, (a) oxida-
tion (normal), and (b) pyrolytic (non-oxidation). Meas-
urements with the analyzer are made by sequential catalytic
oxidation and reduction processes which permit gas chroma-
tographic separation and detection of C02, H20, and N2 .
The combustion is performed in a furnace at 1070°C. Approx-
imately 0.5 mg to 10 mg samples yield accurate measurements
of carbon, hydrogen, and nitrogen down to 0.1$ or less.
By operating in the oxidation mode, total C (organic and
inorganic), total N (organic and inorganic), and total H
(organic and inorganic, as hydrate, adsorbed H20, and
hydroxide) are obtained. In pyrolysis (1070°C) mode with-
out oxidizing catalyst, C02 from carbonate, N2 from nitrates,
and H20 from hydrates, adsorbed H20, and hydroxides are
measured.
Although the above tests are the basic measurements, other
optional measurements are also available. We have studied
adsorbed and hydrated water phenomena by wide-line proton
nuclear magnetic resonance, differential thermal analysis,
thermogravimetric measurements and pyrolysis. The DTA,
TGA, and pyrolysis (weight-loss) measurements with auxili-
ary monitoring of H20 by a sensitive hygrometer (Panametric
Model 1000) and by gas chromatographic measurements (direct-
ly with thermal conductivity detector and carbon molecular
sieve or conversion to acetylene and flame ionlzation
detection) are alternate thermal analysis methods. By
changing the matrix gas (inert gas to air) in these thermal
methods, measurements, of hydrocarbons, fixed carbon, and
water can be made. The major limitation of the TGA, DTA,
and pyrolysis (weight-loss) measurements is the continuous
evolution of H20 throughout the temperature range 100-1000°C«
Wide-line NMR is a nondestructive technique and can be used
to estimate the total water content and also aid in estab-
lishing the nature of the water bonding to the particulate.
Additional measurements by micro-infrared absorption tech-
niques can aid in defining the water, carbonate, nitrate,
and hydrocarbon contents of both crystalline and non-
crystalline particulate.
B.4 DENSITY
Crystalline density of test dusts can be measured with a
Beckman Air Comparison Pycnometer. The apparatus consists
of two equal-volume cylinders with a sample chamber provided
-------
in one of the cylinders for insertion of the test powder.
Pistons in both cylinders permit volume changes. The
pistons in both reference and sample cylinders are ad-
vanced so as to maintain a zero differential pressure
between the cylinders. Due to the presence of the powder
sample, the volume change in the specimen chamber is
smaller by a factor equal to the volume of the preweighed
dust sample. A schematic of the pycnometer is presented
in Figure 15 (ASTM D2856-70).
In previous studies of finely divided powders26»27»28 the
measurement of bulk, apparent, and fluid densities of
powder samples were obtained employing relatively simple,
but meaningful laboratory procedures.
Apparent density is defined as the weight of a sample of
powder divided by the volume it occupies when first placed
in a container. Bulk density is defined as the weight of
powder divided by the minimum volume into which it can be
compressed by vibration or tapping. Fluid density is
defined as the weight of the powder divided by its volume
after having been shaken with air and allowed to settle
to its fluid volume.
The procedures for measuring these densities are as follows
Fluid density - Approximately 5 g of a material are
sealed into a 50-ml graduated cylinder. The cylinder
is rocked back and forth in a horizontal position to
fluidlze the sample and then rapidly placed in an up-
right position. The volume as the meniscus is first
detected in the cylinder is taken as the fluid volume,
and this figure is divided into the sample weight to
find the fluid density.
Apparent density - After being fluidized, the sample
is allowed to settle until no further evidence of
settling is observed. The volume is taken as the
apparent volume; dividing into the sample weight
yields the apparent density. The fluid density and
26. Hedley, W. H. et al., op. cit., ref. 20.
27. Binning, FL C., et al., "Physical and Colloid Chemical
Research on Agents," Final Report, contract DA18-035-
AMC-136(A), U.S. Army, Edgewood Arsenal, January 1969.
28. Snyder, A. D., et al., "Physical and Colloid Chemical
Research on Agents," Final Report, contract DAAA-15-
68-C-0316, U.S. Army, Edgewood Arsenal (1970).
-------
o\
stop
Differential
Pressure
Indicator
Valve A
Cylinder B
Piston A
I I
I I
J I
\\\\\\\\\\\\\\\
\ \ \\ \~\ \V Reference
\ V \ \ \ . U Handwhppl
Position 2
Position 3
iPosition 1
N
I T
Piston B j |
Counter
(\ in cc.)
Measuring
HanoVheel
Figure 15. Air Pycnoraeter-Piston Action
-------
apparent density are determined a number of times (at
least three determinations) and average values calcu-
lated.
Bulk density - The graduate containing the sample is
placed on a Cenco-meinzer sieve shaker (control setting
of 6) for approximately 10 minutes or until a minimum
volume is occupied by the powder. This value is taken
as the bulk volume. Dividing it into the sample weight
yields the bulk density. This determination is repeated
by fluidizing the sample, compacting it on the shaker3
and reading the volume at least three times to yield
an average bulk density.
B.5 WETTABILITY AND MOISTURE CONTENT
B.5.1 Wettabillty
The classical method of determining wettability is to measure
the contact angle of a drop of liquid with a surface of the
material in question. A contact angle goniometer Instrument
can be applied to the measurement of the wettability of test
powders. For these measurements a disc of the powder is
pressed in order to develop a surface for contact angle
measurement. The repeatability of the measurement requires
assessment due to uncertainties in reproducibly producing
the compacted powder surfaces.
Another method for determining the wettability of test dusts
is to measure either the velocity of rise of a liquid in
capillaries packed with powders or the total time required
for a liquid to rise a designated height.29
A method for qualitative determination of wettability was
developed by Monsanto Research Corporation.30 Briefly,
the test consists of adding the samples of powder to vials
containing liquids of various surface tensions and deter-
mining which solutions wet the powder. The solutions are
prepared by adding different amounts of detergent to water.
The majority of studies were done with biodegradable
Sparkleen, but other surfactants were also used.
29.
30.
Vandergrift, A. E., et al., op. cit., ref. 6.
Hedley, W. H., et. al., op. cit., ref. 20.
67
-------
The test procedure for the determination of the wetta-
bility by this method is as follows: eight sample vials
are cleaned and numbered from 1 to 8. Distilled water
(20 milliliters) is placed in the first vial. Vials 2
through 8 contain 20 milliliters of a solution of varying
amounts of surfactant to yield surface tensions from 47
to 68 dynes per centimeter. (The surface tension of pure
water is 72 dynes per centimeter). Approximately 0.5 g of
powder is added to each vial and and all vials are shaken
for 5 min on a Cenco Meinzer sieve shaker. After this time
the amount of turbidity in each of the sample tubes is
observed visually. Samples of powder with a high wettabil-
ity show turbidity in nearly all of the sample vials.
Those with low wettability show turbidity only in the vials
with the lower surface tension (the higher numbered vials
of the series). The identity of the lowest numbered vial
showing turbidity is recorded. High numbers in this test
indicate a low degree of wettability. The results obtained
on 13 samples using both Sparkleen and Tween 80 detergents
are shown in Table 11. The surface tension of each of the
various samples is also shown in this table.
B.5.2 Moisture Content
Moisture content for collected test dusts may be determined
using physical, chemical, or electrical techniques. An
example of each type is described below.
The oldest and most common analytical method for determining
moisture content consists of heating the sample to ensure
complete drying. The moisture content is calculated based
on the loss of weight between the original and dried sample.
Care must be taken to ensure that the loss in weight is due
only to water loss and not caused by evaporation of other
volatile components in the dust.
The moisture content of fine powders can be determined
chemically using a modified Karl Fisher titration method.
A 1-gram sample of the powder is added to dry methanol in
a titration vessel and a measured quantity of Karl Fisher
reagent added. The excess reagent is immediately back-
titrated with a standard solution of water in methanol.
The end point is determined electrometrically using a
dead-stop end-point technique.31
An instrumental technique using a type 26-321A Moisture
Analyzer from Du Pont Instruments can also be used for
moisture content analysis. The instrument measures moisture
JT.Strauts, C. R. N., Gilfellan, J. H., and Wilson, H. H.,
Analytical Chemistry, Vol. II, Oxford, Great Britain,
The Clarendon Press, 1966, p. 253.
68
-------
Table 11. WETTABILITY DETERMINATION RESULTS
BY CYLINDER METHOD
Lowest Numbered Vial Showing Turbidity
Sparkleen Tween 80
Sample
Powder
1
2
3
4
5
6
7
8
9
10
11
12
13
Vial
No.
7
6
8
8
3
6
6
8
8
1
1
6
3
Y
(dynes/cm)
49.0
52.5
47.0
47.0
64.0
52.5
52.5
47.0
47.0
72.0
72.0
52.5
64.0
Vial
No.
6
8
8
8
3
4
6
7
8
3
3
6
6
Y
(dynes/cm)
48.5
45.8
45.8
45.8
57.7
54.3
48.5
47.5
45.8
57.7
57.7
48.5
48.5
69
-------
content In a powder sample using coulometric principles.
The sample to 'be tested is placed in an oven where the
temperature is raised to drive off any water present.
Dry nitrogen carrier gas flowing through the oven sweeps
the moisture into an electrolytic cell where it is elec-
trolyzed. The cell consumes an amount of current corres-
ponding to the quantity of moisture. The electrolysis
current drives an integrator, and because the cell current
is integrated as a function of time, a measurement of the
total moisture present is made which is independent of
the carrier gas flow rate.
The sensitivity of the instrument is 0.1 yg with an accuracy
of 2% of the reading or ±20 yg H20, whichever is greater.
The dynamic range of the moisture analyzer is 0.1 yg to
100 mg of H20. A block diagram of the instrument is pre-
sented in Figure 16.
This procedure is applicable for moisture which is removed
from the dust samples at temperatures less than 200°C (the
upper temperature limit of the moisture analyzer).
An alternate method would be required for determination of
water of crystallization or bound water that is not evolved
below 200°C. The F&M Model 185 CHN Analyzer can be adapted
for this purpose. Hydrogen analysis of (1) powder samples
burned under oxidizing catalyst conditions and (2) powder
samples heated at 450°C in the absence of an oxidizing
catalyst can be employed to estimate the volatile hydrogen
present in -the form of moisture.
B.6 SOLUBILITY
For the purpose of characterizing test dusts water solubil-
ity Is of concern. A convenient method for determining
solubility is by successive water extraction of a weighed
quantity of dust using a Soxhlet extractor. The rate of
extraction should be set at 15 to 20 times per hour and
continued for 4 hours (ASTM D-1739 and 0-49*0. The solu-
bility can be calculated after equilibrium solubility is
attained by drying the sample to constant weight and cal-
culating the loss In weight due to solubilized powder. An
alternate method is to collect the extracting water and
evaporate to dryness in a tared petri dish until constant
weight is achieved. The increase in weight of the petri
dish would yield the solubility of the test sample.
B.7 HARDNESS, ABRASIVENESS, AND GRINDABILITY
The hardness of test dusts or material from which test dusts
originate can be determined using either research-oriented
70
-------
CARRIER
GAS —*r!
GAS
INLET
DRIER
T_
r
NEEDLE
VALVE
FIXED
RESTR1CTOR
MOTOR
DRIVEN
VALVE
SAMPLE OVEN
CARRIER
GAS VENT
ELECTROLYTIC
CELL
rLn_n_n_r
OVEN
HEATER
OVEN
POWER
SUPPLY
AUTO.
MOTOR
CONTROL
CELL CURRENT
INTEGRATOR
JL
ZERO
CONTROL
CELL
POWER
SUPPLY
Figure 16. Block Diagram Moisture Analyzer
-------
or end-use-oriented test methods. The research method most
widely used is the standard test method for microhardness
of materials (ASTM E384-70). In this test (microindentation
hardness) a calibrated machine forces a diamond indenter of
specific geometry, under a test load of 1 to 100 gf (grams
force), into the surface of the test materials. After
indentation the mean diagonal produced on the test specimen
is measured by use of a microscope mounted on the machine.
The hardness can be measured with either Knoop or Vickers
indenters. The Knoop hardness number is expressed as:
where HK = the Knoop hardness
Pi = load, gf, and
di = length of the long diagonal, ym
For Vickers hardness tests the mean length of all diagonals
is employed and the hardness number is expressed as:
HV =
dV
where HV = the Vickers hardness
d~i = mean diagonal of indentation, ym
The abrasive effect of fine particles on materials of
construction cannot be easily separated from the corrosive
effect of the carrier gas and dust particles. Corrosiveness
is the result of a chemical reaction primarily due to car-
rier gas composition, "with particulate composition, contrib-
uting a small portion. Abrasiveness is a result of the
inertial collisions between many small particles and the
walls of the equipment. A test for abrasiveness would be
similar to that for corrosiveness. In fact, the test
described for corrosiveness could be used as a combined
abrasiveness/corrosiveness test.
The grindability of materials is often measured using the
Hardgrove method.32 In this test a 50-gram sample of dried
material (such as coal) sized to minus 16 to plus 30 mesh
is placed In a mortar of the test machine. A motor then
turns the grinder through 60 revolutions and the sample is
32. Salisbury, S. K., Section 27 "Pulverizers," Kent's
Mechanical Engineer's Handbook, 12th Edition, Wiley
Engineering Handbook Series, New York, N.Y. (I960)
72
-------
removed and screened. The quantity of sample passing a
200-mesh sieve is used to determine the Hardgrove grind-
ability index by the following empirical formula.
G = 6.93 W + 13
where W = weight (grams) passing the 200-mesh sieve. This
test is used as a primary method for assessing the grinda-
bility of coals and to predict the performance of a partic-
ular type mill with respect to its ability to grind the
coal.
B.8 CHARGE ANALYSIS
Methods for charge analysis could be either that of
Gillespie and Langstroth33 or that of Kunkel and Hansen31*
which employ a low velocity laminar stream of dilute aerosol
and determine the distribution of both particle size and
particle change.
A schematic view of Gillespie's charge analyzer is illus-
trated in Figure 17. The Gillespie apparatus measures
charge distribution and particle size distribution by
examination of the positions along the length of the micro-
scope slides on which particles of different size have been
deposited due to the electric field. Uncharged particles
pass through the apparatus, and are precipitated and examined
under the microscope. The charge analysis is carried out
by considering the interplay of the equations of motion
on a particle of known radius and initial velocity as it
enters an electric field of known strength.
The apparatus of Kunkel and Hansen does not depend upon
examination of the deposit, but photographs the particles
as they fall down into a field of known strength. The
particle sizes are measured on enlargement of the film from
the velocities measured in falling through an uncharged
field. The charges are calculated from the particle trajec-
tories in the charged field, which are filmed by light
interruption employing a rotating disc in front of the light
source. This method appears to be more convenient, but was
not applicable to diameters smaller than ly in Kunkel's work.
33. Gillespie, T. and Langstroth, G. 0., "An Instrument
for Determining the Electric Charge Distribution on
Aerosols," Canadian Journal of Chemistry, 30, 1056-68
(1952.) ^~
34. Kunkel, W. B. and Hansen, S. W., "A Dust Electricity
Analyzer," Review of Scientific Instruments, 21,
308-14 (1950).
73
-------
To Thermal
Precipitator
Electrodes
Aerosol
Duct
Air-Sheath
Inlet
Flushing
Exit
Three Way
Valves
Air Inlet
Microscope
Slides
Main Duct
Air-Sheath
Inlet
Flushing
Chamber
From Aerosol
Figure 17. Gillespie Apparatus for Measuring Particle
Size and Charge
-------
B.9 DIELECTRIC PROPERTIES
Electrical properties are known to directly influence the
behavior of fine powders because they relate basically to
the charging and relaxation times for charge transfer.
Quantitative determination of particulate resistivity can
be obtained using several techniques utilizing a variety
of equipment. Both laboratory and in situ measurement
techniques are available, but in situ determination is the
preferred method due to the predominance of surface con-
ductivity at typical lower flue gas temperatures (below
approximately 350°P). At these temperatures adsorbed
species from the flue gas control particulate resistivity.
Flue gas conditions may be simulated in a laboratory, but
the potential for chemically active compounds adsorbed on
the particles to modify the particulate properties between
time of collection and measurement is great.
The resistivity of fine dusts may be obtained in a labora-
tory situation and may be used to characterize a dust
sample. The standard technique for obtaining resistivity
measurements in the laboratory is described in the American
Society of Mechanical Engineers Power Test Code 28,
Determining the Properties of Fine Particulate Matter.
However, the measurements may not be representative of the
values that exist in an actual flue gas due to a difference
in the conditioning effects of the carrier gas composition.
The resistivity obtained in the laboratory may frequently
be two to three orders of magnitude higher than in situ
measurements.35 To obtain representative values of particle
resistivity for a flue gas, the measurement must be per-
formed in aitu. Equipment similar in principle to that
used in the laboratory may be utilized for this type of
measurement.36 In situ resistivity measurements are made
under conditions in which the only atmosphere to which the
dust is exposed is that present in the process. This
eliminates uncertainties in attempting to duplicate the
gaseous conditions arid generally gives more reliable and
reproducible results than those obtained by laboratory
techniques.
A number of instruments are available for making in situ
resistivity measurements; point to plane probe, cyclone
resistivity probe, Kevatron electrostatic precipitator
35. Vandergrift, A. E., et al., op. clt., ref. 6.
36. White, H. J., Industrial Electrostatic Precipitation,
Addison-Wesley Publishing Co., Inc., Reading, Massa-
chusetts (1963).
.75
-------
analyzer, and Lurgi electrostatic collection resistivity
device, with the point to plane probe preferred. The
instruments differ fundamentally in the method of sample
collection, degree of compaction of the dust sample, the
values of the electric field and current density used for
measurement, and the method of maintaining thermal equi-
librium in the measurement cell. These variances in
operation cause differences in the sample characteristics
and consequently differences in resistivity measurements
For example, the collection efficiency for fine particles
varies among the different types of measurement devices
thereby biasing the measurement sample toward larger par-
ticles to a varying extent. The resulting resistivity
measurement is correspondingly biased.
Figure 18 is a schematic diagram of a typical in situ
resistivity probe. The apparatus consists of a point-
plane cell at the end of the probe, which is inserted into
the duct through openings of from 2-1/2 inches to 4 inches
diameter. When a voltage sufficient to generate a corona
is applied to the cell, dust is precipitated onto the
plate. After sufficient time has elapsed to collect a
dust layer, the point-plane disc is lowered onto the dust
layer and the current and voltage are measured. The probe
is then removed and the dust thickness measured. Resis-
tivity can then be calculated from the dimensions of the
probe, thickness of the dust layer, and the electrical
data. Thermocouples on the probe are used to measure
temperature.37
B.10 CORROSIVENESS
Corrosiveness is primarily due to carrier (flue) gas com-
ponents. Particle composition contributes a minor effect.
Laboratory experiments to determine the corrosive contri-
bution of the fine particles and adsorbed gas species can
be performed, but to determine the true corrosive nature
of the entire flue gas, tests must be conducted in the
flue gas. There is no standard or preferred method for
performing a corrosion test.38 The method must be designed
to suit the purpose of the test. For particulate emission
control equipment the purpose of corrosion testing would
be to aid in the selection of materials of construction.
37. Oglesby, S. and Nichols, G. B., "A Manual of Electro- fj
static Precipitator Technology, Part I - Fundamentals>
prepared for NAPCA, contract CPA 22-69-73, Cincinnati*
Ohio, 25 August 1970.
38. Perry, J. H., Chemical Engineers' Handbook, 4th
McGraw-Hill, Inc., New York, N.Y.(1963).
76
-------
Hie
Co
i
1
I
j
j
1
Fi
h Voltage
nnection
A [1 f
~ Vs
j
i
A
c
berglas String
V
-
L
[ici
on
\v
Thermocouple Mater
Connection Point
n
Disc Plate | ! wj-
o =n- 1 n "-^
v» \ ^] /*Xx/vV" -\\
•^boooooocoopoc-o*- j ?^vvj /w^r-
Point- Plane CeU jj
Adapter
-oammeter ^
nection **
Flue
Wai
Figure 18. Resistivity Probe
-------
A series of experiments using a variety of test materials
(e.g., various grades of steel, plastic coupons, etc.) could
be designed to provide means for comparison. In the labor-
atory these experiments could be performed by making a wet
slurry and inserting the different samples for designated
time periods. The tests could also be accelerated by ele-
vating the temperature of the dust slurry. For flue gas
corrosiveness testing the different test materials could
be mounted in the gas stream and exposed to the corrosive
atmosphere for predetermined periods of time.
Several potential methods of measuring the extent of the
corrosion may be utilized. Direct observation is a simple
technique but only qualitative in nature. Mechanical tests
including dimensional measurement, change in weight meas-
urement, or measurement of a change in a physical property
(e.g., tensile strength) provide a quantitative measure of
corrosiveness. For laboratory measurements a simple indi-
cation of the corrosive nature of a dust slurry is the
pH. One final potential indicator of the extent of corrosion
on a test sample is a change in the electrical properties
of the sample.39
B.ll OPTICAL PROPERTIES
Refractive index is the physical property most frequently
associated with measurement of the optical properties of
macrosized samples of pure materials. However, there is
little information in the literature on determination of
the refractive index of fine particles of irregular shape
and varied composition.1*0
The complex refractive index, m, may be defined in terms
of two optical constants, n, the refractive index, and k,
the extinction coefficient:
m = n(l-ik)
These constants are not measured directly, but are cal-
culated from reflectivity measurements. In principle, the
method for measuring the complex refractive index of a
substance consists of measuring characteristics of reflect!0**
at various angles of incidence. If the measured values of
the reflection characteristic for plane polarized light
39. Kallen, H. P., "Corrosion," a Power special report,
New York, N.Y., December 1956, pp 73-108.
40. Willis, C., "The Complex Refractive Index of Particles
in a Flame." Journal of Physics, Series 2, Part D,
Vol 3, 19M-56 (1970).
78
-------
of azimuth, p, at angles of reflection 0i and 02 are
Ii and I2, then:
Ii = f(n,k,9!,p)
and
I2 = f(n,k,02,p)
where f is a function obtained from the appropriate Fresnel
equations for the reflections. The simultaneous solution
of the algebraic relations represented above cannot be ob-
tained directly for n and k, the optical constants, but
must be solved indirectly either graphically or by a least
squares computer analysis.
The complex refractive index for a particle can also be
defined as the complex refractive index of the material in
the form of a sphere having the same scattering pattern as
the sample particle. This definition, although not a uni-
versal parameter, is more useful in practice than the actual
refractive index of the material of the particle.
Some effort has been made to determine the refractive index
of some atmospheric aerosols. Volz1*l >"2 has determined the
complex refractive index for a variety of fine particulate
matter (atmospheric water-soluble substances collected from
rainwater, ammonium sulfate, Sahara dust, volcanic pumice,
and flyash) using the potassium bromide disc technique.
Reflection data from discs of pure aerosol are used to
obtain the real part of the refractive index while potassium
bromide discs with about l£ aerosol are .used to determine
the imaginary component. This technique may be utilized for
refractive index determination for fine particles from in-
dustrial sources.
However, as a result of his work with silicon dioxide par-
ticles and pulverized coal particles, Willis32 claims that
a particle stream does not have the same value of complex
refractive index as does a solid or compacted macrosized
sample of the same material. A schematic of the optical
arrangement utilized in his study appears in Figure 19.
Data were collected for both ambient and elevated (1700°K)
temperatures without observation of any significant change
41. Volz, P. E., "Infrared Refractive Index of Atmospheric
Aerosol Substances," Applied Optics, 11^(4), 564-8 (1972)
42. Volz, P. E., "Infrared Optical Constants of Ammonium
Sulfate, Sahara Dust, Volcanic Pumice, and Flyash,"
Applied Optics, 12(3), 755-9 (1973).
79
-------
Collimator
Scale
Receiver
Circular track
Figure 19. Schematic of Optical Arrangement
in the complex refractive index. With refinement this
type of device would be suitable for the charact-erization
of industrial fine particulate.
Since the refractive index is only important in its influ-
ence on the measurement of particle size, the determination
of the actual refractive index is less critical than know-
ing the effect of that refractive index on size measuring
techniques. For this purpose a relatively simple charac-
terization technique could be developed. Assuming optical
size measurement devices are similarly affected by a
particle's refractive index, the behavior of a particle in
any of these type devices could be predicted by its behavior
in one of them. Calibration curves for particles of varied
refractive index could be generated for one specific instru-
ment. Then collected industrial particles of several known
sizes could be tested to determine their "apparent" refrac-
tive index. The particle's actual refractive index may not
be known, but its behavior is more important for character-
ization purposes.
Two assumptions must be accepted for this type of measurement
to be meaningful. The refractive index must be the only
variable influencing an apparent change in the size measure-
ment, and the particles sampled must be typical of all the
collected particles. The result of these measurements will
be a curve showing refractive index as a function of particle
size for the material being tested.
80
-------
B.12 MAGNETIC PROPERTIES
There are two types of magnetic susceptibility. Substances
on which the molecules are arranged essentially at random
(gases and liquids) are magnetically isotropic, i.e., the
magnetic susceptibility is the same in all directions.
Solids too may be effectively isotropic if their components
are randomly oriented. The magnetic susceptibility of
these types of materials is referred to as the average sus-
ceptibility. The magnetic susceptibilities along each axis
(often unequal) are referred to as principal susceptibili-
ties. Since the magnetic susceptibility of test dusts
would be measured on bulk randomly oriented quantities,
average susceptibility measurements would be appropriate.
There are many methods available for measurement of mag-
netic susceptibility. Choice of method depends on the
problem to be solved. There is no single method of uni-
versal applicability-. Several techniques are described in
the literature. * ' »** "5 >* *
One method applicable to the testing of finely divided
powders is the Gouy method. The principle by which this
method operates is shown in Figure 20. A cylindrical
sample is suspended with one end in a uniform magnetic
field and the other end in a region where the field is
negligible. The force exerted on the sample by the mag-
netic field is measured by suspending the cylinder from the
arm of a sensitive balance and finding the change in weight
due to the field.
Instrumentation suitable for application to determining
magnetic susceptibility of fine particulate dusts is
available. ^'^
43. Bhatnagar, S. S. and Mathur, K. N., Physical Principles
and Applications of Magnetochemistry* Macmillan and
Co., Ltd., London, 1935.
44. Trade Literature, LDJ Electronics, Inc., Troy,
Michigan; Vibrating Sample Magnetometer.
45. Trade Literature, Princeton Applied Research, Princeton,
New Jersey, Vibrating Sample Magnetometer..
46. Selwood, P. W., op. clt., ref. 11.
81
-------
N
Figure 20. Principle of Gouy Balance
82
-------
APPENDIX C
TYPICAL FLUE GAS CHARACTERISTICS
FOR INDUSTRIAL SOURCES*17
Key to Typical Flue Gas Characteristics
Solids
Particle size - weight % < size (micron)
Solids loading - grain/SCF
Chemical composition - component - weight %
Bulk density - lb/ft3
Particle density - g/cm$
Electrical resistivity - ohm-cm (laboratory measurement)
Moisture content - weight %
Toxicity - self explanatory
Solubility - self explanatory
Wettability - self explanatory
Handling characteristics - self explanatory
Carrier Gas
Plow rate - M scfm
Temperature - °P
Moisture content - volume %
Chemical composition - component - volume %
47. Vandergrift, A. E., et al.,qp. cit., ref. 6.
83
-------
Pulverized Coal Combustion (electricity generation)
Particle size - 15<3, 25<5, 42<10, 65<20, 8l<40
Solids loading - 1.02 - 5.6 (mean 3.3)
Chemical composition - SiOp 17-64 (average 43)
2-36 (15)
9-58 (24)
CaO 0.1-22 ( 4)
MgO
Fe2°3
0,
0,
1- 5 (1.0)
Na20
0.3- 4 (0.9)
Bulk density - 30-50
Particle density - 0.6-3.0 (average 2.3)
Electrical resistivity - 108-1013
Moisture content - 0.23
Toxicity - not toxic but acidic
Solubility - contains H20 soluble components
Wettability - difficult
Handling characteristics - difficult, agglomerates
Carrier Gas
Flow rate - 292-402
Temperature - 251-314.5
Moisture content - 5.9-8.0
Chemical composition - CO^
°2
N2
CO
so
so
12.5-14.9
4.2- 6.6
balance
5-69ppm
I6l-526ppm
1080-178Oppm
3-66ppm
84
-------
Stoker Fired Coal Combustion
Particle size - 1K10, 23<20, 42<44
Solids loading - 1.5-2.3 (average 1.9)
Carrier Gas
Plow rate - 49-36?
Temperature - 396-426 (average 411)
Moisture content - 7.4-7.8 (average 7.6)
Chemical composition - COp 12.1
02 6.6-6.9
N£ balance
CO 13-29ppm
NOY 430ppm
JL
S02 1280-1380ppm
52-58ppm
85
-------
Basic Oxygen Furnace (iron & steel)
Particle size - 85-95 <1 (count median diameter 0.012,
mass median diameter 0.095)
Solids loading - 2-10
Chemical composition - FepO- 90
FeO 1.5
Mn 0.4-1.5
Si02 1.3-2.0
A120~ 0.2
CaO 3-5
MgO 0.6-1.1
Particle density - 3.44
Electrical properties - resistivity peak between 300-400°F
Solubility - Fe20~ soluble - 10% HC1
Wettability - difficult to wet
Handlingi characteristics - abrasive
Carrier Gas
Flow rate - 35-250
Temperature - 560-3000 depending on utilization of waste
heat boiler
Chemical composition -
Before combustion After combustion
with aspirated air with aspirated
C02 5-16 0.7-13.5
CO 74-91 0-0.3
N2 3-8 74-78.9
02 - - - balance
86
-------
Open Hearth Furnace (iron & steel)
Particle size - 50<5
Solids loading - 0.1-3.5
Chemical composition - F^O-
SiO ~
CaO
MnO
P2°5
85-90
0.9-1.6
0.5-0.7
0.85-1.0
0.6
0.5-1.2
0.4-1.0
Particle density - 5.0
Electric properties - f(temperature) resistivity peak ^300°P
Solubility - Fe20o soluble 10* HC1
Wettability - difficult to wet
Handling characteristics - abrasive
Carrier Gas
Plow rate - 25-100
Temperature - 460-1800 depending on utilization of waste
heat boiler
Moisture content - 7-15
Chemical composition - C02
°
so
SO
Fluorides
8-9
8-9
balance
2-5ppm
100-200ppm
500-800ppm
0-300ppm
87
-------
Electric Arc Furnace (iron & steel)
Particle size - highly variable, generally 60<5, BAHCO
analysis: 68<5, 84<10, 95<20, 99<40
Solids loading - 0.1-2.2
Chemical composition - dependent on charge
Fe20,, 19-44
FeO 4-10
Cr20, 0-12
Si02 2-9
A1203 1-13
CaO 5-22
MgO 2-15
MnO 3-12
FnO 0-44
CuO 0-1
NiO 0-3
PbO 0-4
C 2-4
Particle density - 3.8-3.9
Electrical properties - 6xl05 - 6.6xl013
Wettability - difficult
Handling characteristics - abrasive, fluid-cohesive, high
angle of repose, will bridge
and arch
Carrier Gas
Flow rate - 10-100
Temperature - 215-3000 depending on use of cooling techniques
Moisture content - 0.045 Ib/lb dry gas
Chemical composition - mainly C02, CO, 02, & N2 -
varies with operation; typically,
CO 8-85
C02 5-15
N2 5-85
88
-------
Metallurgical Coke Ovens (iron & steel)
Particle size - quency tower (Q.T.) 95-97>47
Solids loading - Q.T. 0.05-0.1 oven charging (O.C.) 1-15
Chemical composition - coke balls, coal dust, pyrolytic
carbon
Carrier Gas
Flowrate - Q.T. 900M ftVquench O.C. 2.1M ft "/charge
Temperature - Q.T. 140-150
Chemical composition - Q.T. - steam, air
O.C. - C02 2.2
02 0.8
N2 8.1
CO 6.3
H2 46.5
32.1
4 3.5
0.5
89
-------
Cement Plant (kiln)
Particle size - 15-5<5, 43<10, 64<20, 88<40 ; 30<5,
47.6
-------
Municipal Incineration
Particle size - 89.2<44, 69<30, 47<20, 2K10, 6<5
Solids loading - dependent on operation 0.1-0.9 (avg. 0.4)
e 12% co
2
Chemical composition - SiOp 14-52
8-28
CaO 9-22
MgO 2,8-3.4
Na20+K20 3-10
Ti02 0.7-2.8
Particle density - 1.8-3.8
Electrical resistivity - peak at approximately 300°P
(109-1013)
Carrier Gas
Flow rate - 14-288
Temperature - 250-700
Chemical composition
C02, Op, CO, Nps H90
£• ^ £, • a
Typical: C02 3.7
02 13.0
CO 0.2
N2 64.0
H2° 19
also aldehydes
Toxicity and corroslvity - dependent on waste composition
91
-------
Ir_p_n__Fgundry
Particle size - 10<5, 22<10, 25<20, 35<50
Solids loading - usual range 1-3, average 1.2
Chemical composition - Si02 20-40
CaO 3-6
MgO 1-3
PeO (Fe203, Pe) 12-16
MnO 1-2
Particle density - mean vlaue 2.5-3.1
Electrical resistivity - peak at approximately 300°P
Solubility - CaO soluble in H20, CaO, Si02, A1203 soluble
in 10$ HC1
Wettability - difficult to wet
Handling characteristics - abrasive, cohesive, corrosive
Carrier Gas
Plow rate - 10.8-92
Temperature - 210-1410
Chemical composition.- C02 2.8-12.3
02 11.8-12.7
CO 0-0.1
S00 0.002-0.013
92
-------
Kraft Pulp Mill (recovery furnace)
Particle size - 50-85<2
Solids loading - 3-8 dry (mean 3.8)
Chemical composition - Na^O^ - 14-90 (average 80)
Na2C03 - 2.6-73 (11.2)
NaCl - 0.6-14 (high end sea water)
Na2S - 3.3-5.4 (44)
Carbon, flyash
Resistivity - decreases with temperature Increase and
with moisture increase (109-1013)
Hygroscopic characteristics - hygroscopic
Handling characteristics - agglomerates, corrosive
Carrier Gas
Plow rate - 20-568 (average 432) standard ftVair-dried ton
Temperature - 270-650 (average 350)
Moisture content - 20-40
Chemical composition -
Typical orsat - C02 12.5
CO 0.1
°2 7'6
N2 79.8
Minor components - H2S 130-935ppm
S02 l-350ppm
Methyl mercaptan 60-1,400 ppm
Dimethyl sulflde 0-125ppm
Corrosivity - corrosive due to S compounds
93
-------
Asphalt Plants (rotary dryer)
Particle size - Highly variable, 13-9K10, 32-99<20,
55-100<44
Solids loading - extreme range 11-200 typical range 20-70
average - 30
Chemical composition - stone dust, flyash, soot, and
unburned oil
Particle density - 2.6 (average value)
Carrier Gas
Flowrate - 7.7-46 (average 20) 3.9-24.6 (average 9)
Temperature - 85-525 (average 240)
Chemical composition - C02, N0x, N,,, Op, CO
94
-------
APPENDIX D
LABORATORY TESTING OP REDISPERSIBILITY OP BULK POWDERS
The re-aerosolization apparatus developed by Monsanto
Research Corporation is shown in Figure 21. It consists
of a Pyrex tube 2-7/8 inches in diameter and 14 inches long
with a glass frit (Corning Glassworks, Catalog No. 360 60)
located near the bottom.
The sample is distributed on this frit and nitrogen is
passed through, causing aerosolization. The upper portion
of the apparatus contains a Pyrex filter holder (Millipore
Filter Corporation, Catalog No. XX1004700), which holds
47-mm diameter RAWP 0^700 Millipore filters having a small
pore size. The re-aerosolized portion of the sample is
collected on this filter for weighing. The apparatus, as
assembled for use, is shown in Figure 21. This figure
also shows the gauge and manometer used to indicate the
back pressure on this system through the frit.
To perform a test, approximately 0.25 gram of powder is
placed on the lower glass frit through one of the side arms
of the apparatus. A small burst of nitrogen gas distrib-
utes the powder over the frit. The gas flow is then turned
on slowly, and flow is increased (over a period of ten
seconds) to a previously designated flow rate. The gas
flow is continued for an appropriate time period to re-
aerosolize the sample. The re-aerosolized portion is col-
lected on the Millipore filter (previously tared). The
percent (by weight) re-aerosolizabillty for each sample is
computed as follows:
a « -i-t~«K.f m-,, wt of sample collected on filter v , nn
% Re-aerosolizability = wt of sample initially added x 10°
to apparatus
95
-------
Use of this system on a number of powders of similar compo-
sition but differing particle size distributions and dif-
ferent surface coating permitted differentiation in the
relative re-aerosolizability of powder samples.
A commercial product similar to the above apparatus is a
stainless steel elutriation device which was developed for
the U.S. Army by GCA Corporation, this system has been
modified and tested at Monsanto Research Corporation's
Dayton Laboratory. The major difference between the GCA
and MRC instruments is that the dispersion forces in the
stainless steel elutriator are somewhat greater.
96
-------
Nitrogen
Metering System
Elutriation
Chamber
Figure 21. MRC Glass Reaerosolizability Apparatus
97
-------
APPENDIX E
METRIC CONVERSION TABLE
1 inch = 2.54 cm
1 Ib = O.W kg
1 grain = 0.0648 gram
1 ft3 = 0.0283 m3
98
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-650/2-74-117
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Sources and Characterization of Pine
Particulate Test Dusts
5. REPORT DATE
November 1974
6. PERFORMING ORGANIZATION CODE
T.AUTHORIS) w.H.Heaiey, s.M.Mehta, C.M.Moscowitz,
A.D.Snyder,. H.H.S.Yu, and D.L.Zanders
8. PERFORMING ORGANIZATION REPORT NO,
MRC-DA-431
10. PROGRAM ELEMENT NO.
1AB012; ROAP 21ADM-021
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45410
11. CONTRACT/GRANT NO.
68-02-1320 (Task 8)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 1-6/74 '
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
is.ABSTRACT The repojfa gives results of an investigation: to select suitable
characterization criteria for test dusts; to determine procedures for
obtaining, handling, and characterizing the dusts; and to establish
potential suppliers of test dusts. Seventeen suitable characterization
criteria for test dusts were identified, with techniques for obtaining
values for each property defined. Potential industrial suppliers of
test dust for simulation purposes were identified for 11 industries. The
problem of producing a particulate-laden test flua gas was also eval-
uated; redispersion of collected dusts was favored over generation
of fresh particulate. The establishment of a central coordinating logis-
tics network for the acquisition and characterization of test dusts was
recommended. The industries contacted included: pulverized coal combus-
tion, stoker-fired coal combustion, basic oxygen furnaces, open hearth
furnaces, electric arc furnaces, metallurgical coke ovens, cement plants,
municipal incineration, steel foundries, kraft pulp mill recovery fur-
naces, and asphalt plants.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air pollution
Dust
Tests
Supplying
Properties
Logistics
Air Pollution Contro
Stationary Sources
Pine Particulates
13 B
11G
14B
15E
8* DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
108
20. SECURITY CL ASS jTHil pageJ
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
22
------- |