United Slates EPA-600 /7-84-016
Environmental Protection
Agency February 1984
&EPA Research and
Development
A NEW CHARGED FOG GENERATOR
FOR INHALABLE PARTICLE CONTROL
Prepared for
Office of Air Quality Planning and Standards
Prepared by
industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine .broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants.associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-relate'd pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-84-016
February 1984
A NEW CHARGED FOG GENERATOR
FOR INHALABLE PARTICLE CONTROL
by
C.V. Mathai
AeroVironinent, Inc.
145 N. Vista Avenue
Pasadena, California 91107
Contract No. 68-02-3145
EPA Project Officer: William B. Kuykendal
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
Prepared for:
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
A review of the literature shows that control efficiency of inhalable particles using
water droplets can be significantly improved if the droplets are electrically charged. A
spinning cup fog thrower was developed initially to generate electrically charged water
droplets. The poor performance of this device in wind tunnel tests was attributed to the
short lifetime of the fine droplets generated, the ineffective ionizer ring method of
charging the droplets, and the nonuniform charge distribution observed along different
regions of the fog spray pattern.
A new charged fog generator (CFG) was then developed by modifying a commercial
rotary atomizer. In this device, the droplets generated are contact charged to provide a
high charge-to-mass ratio of 1.2 uC/g. The droplets have a number concentration median
diameter of about 100 ym and a mass median diameter of about 200 ym. The water flow
rate is variable (4 to 70 1/h) and the fog spray pattern can be easily changed from long and
narrow to broad and short, with a typical spray coverage of 16-24 m . The device uses
about 1 kW power (110 V ac) and is portable.
Extensive field tests of the CFG (at a bentonite ore unloading operation) were
performed to determine the dependence of its inhalable particle control efficiency (PCE)
on various instrument settings and field conditions. These tests show that the overall
mean PCE is 78% higher than the corresponding value for uncharged fog. Individual PCEs
as high as 88% were achieved. The lifetime of the droplets seems to be the dominant
factor determining the PCE; and PCE values were higher for higher applied voltages and
higher water flow rates. The data suggest that, under optimum instrument settings, PCE
of water droplets could be doubled by charging the droplets.
This report was submitted in fulfillment of contract No. 68-02-3145 by AeroViron-
ment Inc. under sponsorship of the U.S. Environmental Protection Agency.
-------
CONTENTS
Abstract ii
Figures iv
Tables vi
Acknowledgements vii
1. Introduction 1
2. Conclusions 3
3. Recommendations 5
4. Theoretical Background 7
5. Development and Test Results of the Initial Spinning Cup Fog Thrower ... 20
Introduction 20
The Spinning Cup Fog Thrower (SCFT) 22
Initial Wind Tunnel Test Results and Discussion 29
Wind Tunnel Test Modifications 35
6. The Charged Fog Generator (CFG) 37
Introduction . 37
Charged Fog Generator Design 37
Size Distribution of Water Droplets 40
Charge-to-Mass Ratio of the Droplets 46
General Remarks on the CFG 50
7. Evaluation of the Inhalable Particle Control Efficiency of the CFG . . . .51
Field Test Site and Experimental Setup 51
Estimation of Particulate Matter Measurement
Accuracy and Precision 56
Field Test Design 57
Data Processing 59
Data Analysis, Discussion, and Results 65
Conclusions of the CFG Field Tests on Bentonite 77
References 78
Appendices
A. Reprint on Charged Fog Technology: Part I 84
B. Reprint on Charged Fog Technology: Part II 91
111
-------
FIGURES
Number
1 Definition of collision efficiency
2 Calculated single droplet collection efficiency as a function of particle
radius with droplet radius as a varying parameter
3 Calculated single droplet collection efficiency as a function of particle
radius with relative humidity as a varying parameter .......... 13
4 Calculated single droplet collection efficiency as a function of particle
radius for various droplet and particle charges ............ 1^
5 Comparison of calculated and experimentally measured single droplet
collection efficiency as a function of droplet radius for uncharged aerosol
particles ............................. 15
6 Lifetime of water droplets traveling at their terminal velocities ..... 16
7 Calculated Rayleigh limit of water droplet charges as a function of
droplet diameter ......................... 18
8 Means of producing a charged water spray .............. 21
9 Schematic of the spinning cup fog thrower .............. 23
10 Original spinning cup fog thrower .................. 2^
1 1 Schematic of University of Arizona wind tunnel ............ 26
12 Particulate sampling train used during wind tunnel studies ....... 28
13 Sample train to determine the charge-to-mass ratio of water droplets . . 30
13 Summary of test data for light dust loading .............. 33
15 Summary of test data for medium dust loading ............ 3^
16 Cross-sectional diagram of a type "AG" Ray Oil Burner ........ 3g
17 Schematic diagram of the Charged Fog Generator ........... 39
IV
-------
Number
18 Rotating cup and air cone of the Charged Fog Generator 41
19 Rear of the Charged Fog Generator showing rotating seal area 42
20 Side view of the Charged Fog Generator 43
21 Typical long and narrow spray pattern from the Charged Fog Generator . 44
22 Typical short and broad spray pattern from the Charged Fog Generator. . 45
23 Water droplet number concentration as a function of droplet diameter
measured using a cloud optical array probe and a precipitation optical
array probe 47
24 Water droplet mass as a function of droplet diameter measured using a
cloud optical array probe and a precipitation optical array probe .... 48
25 Experimental setup at the bentonite unloading operation at Worland,
Wyoming 52
26 Typical dust plume generated when front-end loader unloads the
bentonite on the hopper grill 54
27 View of the rotating cup of the Charged Fog Generator, inlet of the
cyclone preseparator and the hopper grill from inside the hopper .... 55
28 Mean inhalable particle control efficiency of all the test runs for
charged fog and uncharged fog 69
29 Particle control efficiency of the CFG plotted as a function of
ambient relative humidity for broad and narrow spray patterns 71
30 Comparison of total particle control efficiency of the CFG for a broad
spray and a narrow spray under identical or nearly identical conditions . . 72
31 Comparison of particle control efficiency of the CFG for positively
charged fog and negatively charged fog, with all other parameters
nearly identical 74
32 Comparison of total particle control efficiency of the CFG for two water
flow rates, with all other parameters nearly identical 76
-------
TABLES
Number
1 Initial Spinning Cup Fog Thrower Wind Tunnel Test Plan 27
2 Summary of the Initial Spinning Cup Fog Thrower Wind Tunnel Test
Data 31
3 Proposed Charged Fog Generator Field Test Plan 58
4 Measured Particle Concentrations Inside the Bentonite Unloading Hopper
During CFG Field Tests 60
5 Measured Inhalable Particle Control Efficiency, Meteorological
Conditions and CFG Settings During the Field Tests on
Bentonite Ore 66
6 Comparison of Total Particle Control Efficiencies of CFG for Two Pairs
of Applied Voltages, With All Other Parameters Nearly Identical .... 75
VI
-------
ACKNOWLEDGEMENTS
On behalf of AeroVironment Inc., the author gratefully acknowledges the contribu-
tions of the following persons for the successful completion of this project:
William B. Kuykendal and Dennis C. Drehmel, EPA Project Officers, for
their technical guidance, encouragement and understanding; Stuart A. Hoenig for his
consulting services and for making available the University of Arizona Wind Tunnel
Facilities for the initial device tests; John Kinsey for his contributions to this project
as the Principal Investigator from the beginning until July 1980; Carol Lyons for her
contributions as the AeroVironment Project Manager until November 1980; Lyle A.
Rathbun, who performed all the charged fog generator field tests in Worland, Wyoming;
and the personnel of Kaycee Bentonite Corporation, Worland, Wyoming, for their
cooperation and assistance throughout the field test program.
The author is also grateful to R.C. McCrillis and R.V. Hendriks of the U.S.
EPA for their critical reviews of this report and their valuable comments and suggestions.
vii
-------
SECTION 1
INTRODUCTION
Since the National Ambient Air Quality Standards were established in 1971,
significant progress has been made in the U.S. in controlling total suspended particulate
matter (TSP) emitted from conventional industrial sources. However, various regulatory
agencies are becoming increasingly concerned because ambient TSP standards are being
exceeded at many urban and industrial areas, mainly because of the lack of appropriate
control measures for particles from non-stack sources -- that is, fugitive emissions.
Although the total particle mass loading from human activities is only about ten percent
of that from natural sources, its effects are significant and largely detrimental. Recent
reports (Cowherd, 1980; NRC, 1979; Friedlander, 1977; and Natusch, 1974) indicate that
inhalable particles (10-15 um and smaller in aerodynamic diameter), in general, and fine
particles (2-3 um and smaller), in particular, may be a human health hazard and degrade
atmospheric visibility. Therefore, the U.S. Environmental Protection Agency (EPA) is
emphasizing the importance of finding new methods of controlling inhalable particles,
especially from fugitive emission sources.
Spraying fine water droplets is a well-known method to control dust. Various types
of scrubbers rely on water droplets to sweep dust from gases. Although spraying water
has been the most common dust control method used in material handling and mining, it
has been marginally successful, with particle collection efficiencies of only 30-^0%
(Emmerling and Seibel, 1975; Courtney and Cheng, 1976). In addition, equipment to
implement this method has drawbacks; water spray nozzles become clogged and large
quantities of water are needed. The latter is critical, especially in the arid western
United States and in applications where addition of large quantities of water is forbidden.
Studies by McCoy et al. (1983), Yung et al. (1980), Hoenig (1979 and 1977), Hassler
(1978), Prem and Pilat (1978), Drehmel (1977), Lear (1976), and Pilat (1975) have shown
that the collection of fine particles can be greatly enhanced using electrostatic forces in
particle control devices. Hoenig (1979 and 1977), Hassler (1978), Suck et al. (1981),
Walkenhorst (1971), and Schutz (1967) have shown that most industrial pollutants and
naturally occurring dust particles acquire electrical charges as they are dispersed into the
air. They have also shown that the polarity and magnitude of the charges depend on the
size and origin (coal, soil, minerals, etc.) of the particles. Therefore, if oppositely
charged water droplets are sprayed on the dust to be suppressed, the particles which
collide with the droplets will agglomerate rapidly and settle out of the atmosphere. The
particle collection efficiency of water sprays can thus be improved significantly if
charged water droplets (fog) of appropriate sizes can be generated.
This report details the development and testing of a new Charged Fog Generator
(CFG) under the sponsorship of the Industrial Environmental Research Laboratory (IERL)
of the U.S. EPA. The objective of this research project was to develop a portable unit to
control dust where conventional methods of dust control cannot be applied such as in
-------
material handling areas with conveyor belt transfer points, front-end loaders, and
material loading and unloading. This project consisted of two phases.
1. Phase I
The four objectives of this phase were:
a. Theoretical Calculations. To (1) define and identify the theory behind this new
concept of controlling inhalable particles using electrostatically charged fog;
(2) predict the control efficiency of the method based on the charge-to-mass
ratio of the water droplets and the size distribution of the droplets, together
with the mass collected and its size distribution — both before and after the
dust was treated with charged fog; and (3) estimate the power requirements as
a function of fractional efficiency and the minimum power required to achieve
90% efficiency in control of inhalable particles.
b. Preliminary Experiments. To design a bench-scale prototype instrument to
generate fog and to charge the water droplets to a high charge-to-mass ratio.
The size distributions of the droplets were to be appropriate for controlling
dust particles. The electrostatically charged fog was thus to be applied to
inhalable dust particles in a controlled atmosphere (wind tunnel) to verify the
fog's collection efficiency and this result compared with the predictions of the
theoretical calculations.
c. Economic Analysis. To perform an economic analysis of the method for
controlling inhalable particles with charged fog. Cost and power requirements
were to be estimated for the new instrument and compared with the cost of
any other instrument with the same capabilities.
d, Recommendations. To report on the work done under Tasks 1, 2, and 3 with
recommendations for work to be done under Phase II.
2. Phase II
Based upon the Phase I report, AV was asked to proceed with Phase II of the
program. The objectives of the second phase were:
a. To construct a working model of the instrument, applying the findings of.
Phase I;
b. To design an experiment to field test the instrument constructed;
c. To obtain data on the collection efficiency of the instrument and
evaluate the instrument's effectiveness for controlling inhalable
particles.
All field work under Phase II was completed by july 1981. This report presents the
results of work done under the two phases. Section 2 gives the conclusions and Section 3,
recommendations. Section 4 outlines the theoretical background, and Section 5 describes
the development^ test results and setbacks of the original spinning cup fog thrower.
Section 6 summarizes the development of the CFG. Section 7 describes the extensive
CFG field tests and relates the findings. Appendices A and 3 are reprints of two articles
on work related to this report.
-------
SECTION 2
CONCLUSIONS
Since most industrial pollutants and naturally occurring aerosol particles are
electrically charged, theoretical studies have shown that inhalable particle control
efficiency of water sprays can be improved significantly if well-charged water droplets of
100-200 ym in diameter are applied on these particles. Initially, charged fog was
generated using a spinning cup fog thrower, developed in association with the University
of Arizona in Tucson. This device created fog with a, median droplet diameter of about
20 vim and a charge-to-mass ratio of about 1 x 10~ C/g. Wind tunnel tests of this
spinning cup fog thrower at Tuscon showed about 50% inhalable particle control
efficiency. The poor performance of this device was attributed to the very short lifetime
of the fine droplets generated, ineffective ionizer ring method of charging the droplets,
and the nonuniform charge distribution observed along different regions of the fog
pattern.
A new Charged Fog Generator (CFG) was then developed by modifying a commercial
rotary atomizer. The CFG generates water droplets with a number concentration median
diameter of about 90-100 ym and a mass median diameter of about 200 urn. These larger
droplets will not evaporate as rapidly and will therefore remain in the air long enough to
provide sufficient time for interaction between the droplets and particles. These droplets
are electrically charged by the method of contact charging, wherein a 15-kV dc potential
is connected to the inflowing water. This charging process provides a typical charge-
to-mass ratio of 1.2 x 10~ C/g. The water flow rate in the device can be varied from
about 4 to 70 1/h. The total power required to operate the system is less than one
kilowatt and, therefore, can be easily operated from a remote-location with a small
generator. The fog spray pattern covers a volume of 16 to 24 m and the spray pattern
can be varied from broad and short to long and narrow, thus adapting to the extent of the
emission source to be controlled. This device uses no water-spray nozzles, thus
eliminating nozzle clogging problems. The CFG, mounted on a small platform, is
portable.
Extensive field tests of the CFG were performed to determine the dependence of its
inhalable particle control efficiency under various instrument settings and field
conditions. The field tests were conducted at a bentonite ore processing plant in Worland,
Wyoming, during 1981. Bentonite was unloaded from front-end loaders into a semi-
enclosed hopper and the dust generated in the hopper was treated with charged fog.
Inhalable particle concentrations were measured as fine and coarse fractions using a
combined cyclone preseparator/cascade impactor high-volume sampler. Ninety-six runs
were made under three test scenarios: uncontrolled dust, uncharged fog applied to dust,
and charged fog applied to dust. From these measurements, the CFG's inhalable particle
control efficiencies were calculated for various combinations of instrument settings and
field conditions. These test results show that:
-------
1. Charged fog was shown to be an effective means of controlling fugitive dust
emissions. The mean value of the inhalable particle (fine and coarse fractions
combined) control efficiency of charged fog measured under all instrument settings
and field conditions is 78% higher than the corresponding value for uncharged fog.
2. The relative humidity (in this particular experimental setup) seemed to play a
significant role in determining the overall particle collection efficiency. It appears
that the lifetime of the droplet is the dominant factor in determining particle
control efficiency. The droplets should be large enough not to evaporate too
quickly, yet small enough to yield a high particle control efficiency.
3. Under identical or nearly identical field conditions and instrument settings, nega-
tively charged droplets gave higher values of particle control efficiency than did
positively charged droplets, suggesting that inhalable bentonite particles carry a net
excess positive charge.
4. Measured inhalable particle control efficiencies were higher for higher applied
voltages in the 4-10 kV range. At the upper end of this range, the particle control
efficiency seemed to attain a saturation value.
5. Measured inhalable particle control efficiencies were higher when charged droplets
could cover more of the dust-laden air in the hopper. In the experimental setup
used, higher water flow rates and a broad spray pattern generally resulted in higher
collection efficiencies, although the key element appeared to be how many particles
were treated by the droplets.
6. Because of the type of particle sampling method used and the field setup, the effect
of wind speed and direction on particle control efficiency cannot be quantified with
the available data.
7. The optimum CFG instrument settings are found to be 60 1/h water flow rate; a
spray pattern which will cover maximum volume of the dust-laden air (broad or
narrow spray depending on the extent of the source); an applied voltage of 8-10 kV;
and positive or negative charge, depending on the charges on the dust particles.
Ideal field conditions are high relative humidity (to ensure long droplet lifetime),
and calm or low wind conditions.
Ordinary water sprays — the closest technique to charged fog — are inefficient in
collecting fine particles. By the addition of electric charges on the droplets, it seems
possible to at least double the inhalable particle control efficiency of water sprays. Other
control methods using evacuation and hooding are roughly ten times more expensive than
are charged-fog devices in capital cost and in operating cost. Charged-fog devices are
currently available on the market and the Charged Fog Generator is expected to be
commercially available soon.
-------
SECTION 3
RECOMMENDATIONS
The capabilities of the Charged Fog Generator demonstrated in this report can be
utilized in several applications, such as
controlling dusts in material handling (loading, unloading, conveyor belt
transfers of various materials, including coal and grain);
controlling inhalable particles in the mining industry to reduce the fine
fraction of particles inhaled and to recover, and subsequently recycle airborne
particles of precious metals, such as gold;
controlling particles generated by a mobile source (such as a road sweeper or
construction equipment);
decontamination or collection of biological organisms using charged droplets.
To facilitate the possible use of charged fog devices in areas listed above, the
following studies need to be carried out.
1. Although some qualitative work on the electrical characterization of airborne
particles has been done, quantitative information is needed on the polarity and
magnitude of the electrical charges of airborne particles and their dependence on
particle sizes and chemical composition. Also, the potential of the CFG to control
particles of various origin is required. An instrument capable of measuring
electrical charges and aerodynamic diameters of inhalable particles simultaneously
has recently been reported by Mazumder et al. (1982) and Renninger et al. (1981).
2. The CFG and other commercially available charged fog devices use high voltages
and their present designs are such that they could be a human safety hazard and a
fire hazard in an inflammable, gaseous atmosphere. Therefore, additional work is
needed to make the devices safe and explosion-proof before they can be used for
inhalable particle control in mines.
3. Use of charged fog to collect airborne submicron- and micron-sized organisms
should be of interest to the industrial hygienist. Control of germs and possible
decontamination of an area using proper chemical additives in the water spray is of
interest to the army. Preliminary investigation of this application could be coupled
with 1. above.
-------
4. The potential of the CFG to control dust from a mobile source is worth examining.
No alternative method of control for such emissions is available, other than ordinary
water sprays which do not efficiently control fine particles.
5. Possible application of charged fog in recovering airborne precious metals should be
of great interest to the gold and other precious metal and mineral mining industries.
Studies to evaluate the possibility of such a method could be undertaken.
-------
SECTION *
THEORETICAL BACKGROUND
Removing fine particles from a gas stream in an open area is difficult because of
the particles' low mobility, unfavorable inertial properties and uncontrollable external
factors. The use of electrostatics to control fine particles in a controlled atmosphere is
well documented (White, 1963). However, only during the last few years have electro-
statics been used to augment particle collection efficiency of water droplets (Yung et al.,
1980; Hoenig, 1979; Prem and Pilat, 1978; Drehmel, 1977; Wang et al., 1978; Lear, 1976;
Pilat, 1975). Hoenig (1977, 1979, 1980) and Hassler (1978) have shown that most industrial
pollutants and naturally-occurring dust particles acquire electric charges as they are
dispersed into the air. Therefore, particle collection can be enhanced via electrostatic
attraction if the water droplets are charged to the opposite polarity. Unlike conventional
water sprays, droplets in charged fog sprays carry electrostatic charges.
An electrostatically charged water droplet moving with reasonable speed in a
medium containing aerosol particles sweeps out a volume equal to its pathlength times its
projected area. Aerosol particles in this volume, which are not swept out by aerodynamic
forces as the droplet moves along its trajectory, collide with the droplet. Figure 1 shows
the definition of collision efficiency, E , which may be obtained from a knowledge of the
particle's trajectory around the drop as
2
(1)
where y is the largest initial horizontal offset a particle can have and still collide with
the droplet (Wang and Pruppacher, 1980). Further, we assume that once an aerosol
particle and a droplet collide, they adhere to each other; in other words, the collision
efficiency is identical to the collection efficiency. The agglomerated particles and the
droplet then settle out of the air according to the Stokes relationship.
The collection of an aerosol particle by a charged droplet is the result of a number
of simultaneous mechanisms of interaction between them such as inertial impaction;
direct interception; Brownian diffusion; and electrostatic, diffusiophoretic and thermo-
phoretic forces (Wang et al., 1978; Prem and Pilat, 1978; Grover et al., 1977; Nielsen and
Hill, 1976a, 1976b). When an aerosol particle approaches a water droplet with a relative
velocity, it may directly collide with the droplet, barely touch the droplet, or entirely
miss the droplet. In this process, the first case is called an impaction and the second case
is called an interception. The relative effect of the mechanisms of interaction between
the droplet and the particle depends upon the size of the particle. For large particles
(aerodynamic diameter greater than 2-3 um), the dominant mechanisms of particle
collection by droplets are impaction and interception. For particles smaller than 0.1 ym,
-------
COLLECTOR DROP
O
G
01
n
<
OH
H
a
NM
N
<
a*
a
R = DROPLET RADIUS
r = PARTICLE RADIUS
yr = LARGEST INITIAL
C HORIZONTAL OFFSET
Figure 1. Definition of collision efficiency.
-------
Brownian diffusion becomes very important, and for particles between these two ranges,
electrostatic forces are the dominant interaction mechanism. Phoresis is the process by
which particles move because they are subjected to a temperature gradient (thermo-
phoresis) or vapor pressure (diffusiophoresis). These processes are significant only when
the droplets are evaporating fast and when the particles are small.
The particle collection efficiency of uncharged water sprays (where inertial
impaction is the major collection mechanism) is given by Cheng (1973) as,
E=l-exp - i.-^. fc.rj (2)
where Q, and Q_ are the volume flow rates of the water and air component of the dust
cloud, respectively, L is a characteristic length for the total capture process, D is the
mean droplet diameter and 17 is the single droplet collection efficiency.
For a given particle size, single droplet collection efficiency due to inertial
impaction is proportional to the relative velocity between the droplet and the particle and
inversely proportional to the droplet diameter. Cheng (1973) also showed that for a given
quantity of sprayed water and dust cloud, 300-ym droplets have the maximum collection
efficiency for 3-um particles, 200-um droplets have the maximum collection efficiency
for 2-um particles, and 150-um droplets have the maximum collection efficiency for
1-um particles.
When particles are small or when the relative velocity between the particle and the
droplet is very small, then the particle is considered to be inertialess. Electrostatic force
is the dominant mechanism of interaction of particle collection for particles in this size
domain. For the single droplet collection efficiency of a charged particle by an
oppositely-charged droplet, Kramer and Johnstone (1955), Nielsen and Hill (1976a), and
Prem and Pilat (1978), gave the following relation:
r]= -4Kc = - 4 C Qc QP/247T2 er R2 PUQ (3)
t '
where
K = electrostatic parameter
Cc = Cunningham slip correction factor
Q = droplet charge
Q = particle charge
e P = dielectric constant
r = particle radius
R = droplet radius
p = viscosity
U = free-stream velocity.
o
This relation shows that for a given particle size the electrostatic forces are
directly proportional to the magnitude of the charges on the droplet and particle, and
-------
inversely proportional to the square of the droplet radius and its free-stream velocity.
Physically, the last item indicates that when the droplet free-stream speed is slower, the
particle spends a longer time near a droplet and thus the electrostatic forces between the
droplet and particle act for a longer time thus enhancing particle collection.
Prem and Pilat (1978) showed that for a 200-ym and a 50-um charged droplet and
oppositely-charged particles in the size range 10-20 urn, the particle collection efficiency
due to electrostatic forces and inertia are comparable. However, for particles less than
10 urn, and for the same droplet sizes, electrostatic forces progressively become the
dominant collection mechanism, compared to particle inertia, as the particle sizes
decrease.
The value of the single droplet collection efficiency given by Equation (3) may be
substituted in Equation (2) and integrated over ail droplet and particle sizes to obtain the
total particle control efficiency of an electrostatically charged water spray system.
Recall, however, that the relation thus obtained for E is an idealized version of the
complex interactions between droplets and a moving dust cloud. Therefore, with a
spectrum of droplet radii and a dust cloud of particles with various sizes, the particle
control efficiency of a charged spray device becomes difficult to predict. Still, it is
instructive to examine some of the details of the results of the theoretical studies on
single droplet collection efficiency.
The mechanisms of interaction between water droplets and aerosol particles, and
the collection efficiency of a single droplet falling under its terminal velocity have been
topics of many detailed investigations (Yung et al., 1979; Wang et al., 1978; Grover et al.,
1977; Beard and Grover, 1974; Beard, 1974; Slinn and Hales, 1971; Greenfield, 1957). In a
series of papers, Pruppacher, Grover, Beard, and Wang at the University of California at
Los Angeles have examined, both theoretically and experimentally, the single droplet
collection efficiency, and some of their results are summarized graphically in Figures 2
through 5.
Figure 2 shows the theoretically computed, overall collection efficiency, E, of water
droplets on aerosol particles of various radii for the case of 75% relative humidity,
900 mb pressure and 10° C temperature. Greenfield (1957) was the first to demonstrate
that in the absence of an electric charge on either the droplets or the aerosol particles,
the graph of the collection efficiency of water droplets plotted against the radii of the
aerosol particles has a minimum for particles with radii near 1 urn. The minimum
becomes more pronounced for progressively smaller droplets, and becomes a "gap" (i.e.,
the collection efficiency goes to infinitely low values) for droplets of radii approximately
55 um and less. This gap, known as the "Greenfield Gap," and the minimum in the
collection efficiency shifts to smaller particle radii as the radius of the collecting water
droplet increases. This minimum (or gap) results from the fact that Brownian diffusion
dominates particle scavenging for smaller particles (whose radii are less than the value
corresponding to the minimum in collection efficiency), while inertial impaction
dominates particle scavenging for particles with radii greater than the radius corres-
ponding to the minimum in collection efficiency. For particles near 1-um radii, neither of
these two processes is very effective, thereby yielding the minimum; however, phoretic
and electric forces play significant roles in the collection efficiency in this region, as will
be shown shortly. The minimum in E and the Greenfield Gap, described above, is very
clearly seen in this figure. Also, the collection efficiency of larger particles is greater
10
-------
10
10.0
100.0
PARTICLE RADIUS (um)
Figure 2. Calculated single droplet collection efficiency (adapted from Wang et al.,
1978) in air of 10 C and 900 mb as a function of particle radius at a relative
humidity of 75% with droplet radius as a varying parameter (A) 42 um,
(B) 72 um, (C) 106 um, (D) 173 um, and (E) 310 um.
11
-------
for larger droplets, while the collection efficiency of smaller particles is greater for
smaller droplets, and the minimum collection efficiency (Greenfield Gap) occurs for
particle radius, ~ 1.0 vim.
Figure 3 plots the calculated collection efficiency versus aerosol particle radius for
a water droplet of 72 ym radius, with relative humidity (RH) of the medium as a varying
parameter. When the RH changes from 100% to 50%, the phoretic forces could raise E by
an order of magnitude or more. As the figure also shows, for larger particles (>2-3 urn
and larger in aerodynamic diameter) the collection efficiency is independent of RH. As
noted previously, the evaporation rate increases as RH decreases which, in turn, reduces
the lifetime of the droplet in the medium and consequently reduces the collection
efficiency.
The discussion so far has not considered the effect of charging the droplets or the
aerosol particles. Figure 4 shows three curves for the collection efficiency, E, versus
particle radius for a droplet of radius, R = 106 um, and 75% RH. Curve C at the bottom is
similar to Figures 2 and 3 without any charges. Curve B gives collection efficiency as a
function of particle radius, r, with droplets and aerosol particles both charged to the same
magnitude of 2.0 esu cm, but of opposite sign. And Curve A is similar to B, except the
magnitude of the charge is 20.0 esu cm" ; curve A is obtained by a less ri&orous
calculation method of George and Poehlein (1974). The charge of 2.0 esu cm" was
chosen for Curve B because that is the mean charge found on water droplets during
thunderstorms, as shown by Takahashi (1973).
Figure 4 shows that the introduction of electric charges on the droplets and the
aerosol particles completely eliminates the Greenfield Gap and, depending upon the
amount of charge, the collection efficiency of submicron particles is increased by more
than an order of magnitude. This effect is the fundamental principle on which the
charged fog technology is based. Figure 4 also shows that the addition of electric charges
does not significantly affect the collection efficiency for particles larger than about 5 vim
in aerodynamic diameter.
Figure 5 compares the theoretical prediction and the experimental data of
Wang et al. (1978) for collection efficiency of charged water droplets of various radii for
aerosol particles of radii, r = 0.25 urn, at about 20% relative humidity. The prediction
agrees fairly well with the measured data. This agreement is also compatible with the
conclusion of Walton and Woolcock (1960) that the overall collection efficiency of a given
quantity of sprayed water should be fairly independent of droplet sizes, in the 100 to
500 ym radius range.
When sprayed into the air, the charged droplets will evaporate unless the air is
saturated with water vapor. The droplet lifetime determines the effective contact time
between the droplet and particles and thus strongly influences the overall particle control
efficiency of the system. The lifetime of a water droplet depends upon the temperature
and relative humidity of the medium into which it is introduced, as shown in Figure 6
(Daugherty and Coy, 1979).
The choice of droplet size is, at best, a compromise. On the one hand, 50- to 60-ym
diameter droplets have a higher particle collection efficiency than do larger droplets; on
the other hand, smaller droplets have a much shorter lifetime if they are introduced into a
12
-------
lO'
10
-1
u
z
UJ
u
E
u.
UJ
O
UJ
-j
8
01
j
CU
O
<*
Q
UJ
a
10
-2
10
-3
10
-5
0.01
0.1 1.0
PARTICLE RADIUS (ym)
10.0
100.0
Figure 3. Calculated single droplet collection efficiency (adapted from Wang et al.,
1978) in air of 10 C and 900 mb as a function of particle radius for 72 urn
radius droplet with relative humidity as a varying parameter (A) 50%,
(B) 75%, (C) 95%, and (D) 100%.
13
-------
10
0.1 1.0
PARTICLE RADIUS (urn)
10.0
Figure 4. Calculated single droplet collection efficiency (adapted from Wang et al.,
1978) in air of 10°C and 900 mb as a function of particle radius for 106 ym
radius droplet at 75% relative humidity for droplet and particle charges of
(A) +20 esu cm, (B) +2 esu cm, and (C) zero charge.
-------
ttJ
D
E
u-
tu
'2
lO
UJ
U
io"3
a
UJ
10
-5
J I
30 *0 50 100 200 300 500 1000
DROPLET RADIUS (urn)
3000
Figure 5. Comparison of calculated (smooth line) and experi-
mentally measured (shaded) single droplet collection
efficiency (Wang et al., 1978) as a function of droplet
radius for uncharged aerosol particles of 0.25 um radius in
air of 22°C, 1000 mg, and a relative humidity of ~20%.
15
-------
100
2 5 10 20 50 100 200
INITIAL DROPLET DIAMETER (pm)
500 1000
Figure 6. Lifetime of water droplets traveling at their terminal
velocities (Daugherty and Coy, 1979).
16
-------
fairly low RH and high temperature medium. Thus, to obtain the best collection
efficiency, the droplets must be small enough to provide both an adequate spray rate per
volume of gas treated and sufficient contact time, yet large enough not to evaporate too
quickly. Droplets in the range 100-200 ym are expected to be a reasonable choice for
ambient applications.
Another important consideration in charged fog technology is the limitation on the
amount of charge on particles and droplets. As noted earlier, Hoenig (1979), Hassler
(1978), and others, showed that aerosol particles in the air are generally charged. The
maximum electric charge a particle or droplet can carry is limited by the physical
properties of the particle or the droplet. Once these limits are reached, a particle will
spontaneously emit some of the charge and a droplet will disintegrate into smaller
droplets. For a spherical particle, the limiting charge is given by Whitby and Liu (1966) as
DD2ES
"p- ^T- <»>
where E^ is the surface field intensity at which charges are emitted, D is the particle
diameter, e is the elementary unit of charge, and n is the number of elementary units of
charge corresponding to the limit. ^
In the case of a water droplet, the maximum charge it can carry before it
disintegrates is reached when the outward pressure produced by the electric field at the
surface of the drop is equal to the inward pressure produced by the surface tension (Leong
et al., 1982). This limiting charge is called the Rayleigh limit (after the physicist Lord
Rayleigh) and is given by
(5)
where
QR = limiting charge on the droplet (Coulombs)
e a^ = permittivity of the medium in which the droplet is located
a° = surface tension of the liquid
R = droplet radius, in meters.
2
.^Assuming,
-------
ID''F
Rayleigh Limit
for Water Droplets
10 100
DROPLET DIAMETER (pm)
1000
Figure 7. Calculated Rayleigh limit of water droplet charges as a function of droplet
diameter.
18
-------
In the studies outlined above, collision of a particle with the droplet was assumed to
always result in the collection or removal of that particle from the air. Pemberton (1960)
and McDonald (1963) argue that the particle collection efficiency should increase as the
surface tension of the droplet decreases. Indeed, increased particle collection efficiency
has been reported (Hesketh, 1974; Rabel et al., 1965) with the addition of surfactants to
lower surface tension. Drees (1966) and others suggest quite the opposite result due to
surfactants; yet others claim no effect at all.
Woffinden et al. (1978) suggested that the change in the droplet size distribution
upon addition of surfactants, and the consequent effect on particle collection is a possible
reason for the discrepancy in the observed results of the addition of surfactants.
McCully et al. (1956) and Goldshmid and Calvert (1963) have shown that non-wettable
particles are collected less efficiently by droplets than are wettable particles.
Stulov et al. (1978) conclude that non-wettable particles adhere to the droplet surface
upon impaction and, consequently, form a layer on the droplet surface causing particle
bounce-off later. The Stulov et al. study was restricted to particles larger than 5 ym. In
any case, data are lacking from any controlled experiment to study the effect of adding
surfactants to water droplets used in collecting fine particles.
Finally, a comment on the possible enhancement of the particle collection
efficiency by charging the droplets is appropriate. Although Figure 4 shows that the
addition of electric charges on particles and droplets completely eliminates the minimum
in particle collection efficiency (around r = 1 ym) and shows single particle collection
efficiencies 5 to 10 times higher than for uncharged sprays, the overall collection
efficiency of an operating system may not be that high. Increases in collection
efficiencies of about 15% for 1-um particles to over 45% for 0.3-ym particles compared
to uncharged droplets were reported by Pilat (1975). Hoenig (1979) reported particle
control efficiencies of 50 to 80% with charged droplets under controlled experimental
conditions using a device called Fogger I, which he had patented. Under the sponsorship
of the U.S. EPA, AV has developed a new charged fog device to control inhalable
particles. This new device, called the Charged Fog Generator, was extensively field
tested during 1981 to evaluate its particle control efficiency. These tests are described in
Section 7 of this report. To our knowledge, this is the first attempt to examine the
relationship of various parameters described earlier in this chapter and the actual
measured particle control efficiency of an operating charged fog device in a field setup.
19
-------
SECTION 5
DEVELOPMENT AND TEST RESULTS OF THE INITIAL
SPINNING CUP FOG THROWER
INTRODUCTION
Many researchers have attempted to enhance dust agglomeration by wetting the
dust, but have obtained limited success. Although as much as 60% collection efficiency
has been claimed for this method, about 30 to 40% is the generally accepted value. The
difficulties of generating small enough water droplets and inducing the dust particles to
attach to the droplets are quite well known. These problems become more complicated
when the dust to be controlled is in open areas with uncontrollable ambient conditions of
temperature, pressure, relative humidity and, most significantly, wind speed and wind
direction. Equations (2) and (3) (in Section 4) for the collection efficiency are for an
ideally controlled situation. The actual control efficiency will depend on the ambient
conditions, the electric charges on the droplets, the size distribution of the droplets and
particles, and the lifetime of the droplets.
Devices to generate charged sprays have been used extensively in paint spray
equipment for some time. Electrically charged sprays have been generated in applying
agricultural pesticides, as reported by Law (1978) and Carlton and Bouse (1980). The use
of charged sprays for dust control was initiated by Hoenig (1979) at the University of
Arizona and Hassler (1978) at the Royal Institute of Technology, Stockholm, Sweden. All
these devices use a pressure nozzle to atomize the water.
Water droplet charging is generally achieved by one of three principles: (a) electro-
static induction charging, (b) ionized field charging, (c) contact charging.
Hassler (1978) showed that the droplets may be charged by the water-to-metal
frictional forces inside the nozzle during atomization. This method requires no high
voltage supply; however, it requires very pure deionized water. The limitation to
deionized water precludes the use of this method of charging for dust control in the field.
Figure 8 schematically shows these droplet charging mechanisms in conjunction with
a spray nozzle. In the most frequently used method, electrostatic induction charging, a
high voltage potential is maintained between the water spray nozzle and the induction
ring, as shown in Figure 8a. Positive or negative charges are induced on the droplets due
to the charges on the induction ring, depending on the polarity of the high voltage applied
to the ring. A detailed analysis of the induction charging of droplets is given by Law
(1978). In ionized field charging, as shown in Figure 8b, an electrode at a sufficiently high
dc potential is placed near the water nozzle. This causes dielectric breakdown of the air
immediately surrounding the electrode. Water droplets traveling through this ionized
field region can acquire electric charges by ion attachment.
20
-------
Water
Feedline
Insulated
Metal Ring + Charged
\ ??'•'••'"\ .Droplets
Spray Nozzle
5-10 KV/
- Charge
a. CHARGE INDUCED VIA METAL RING
5-10 KV
+ Charge
Water
Feedline
Electrically
Isolated
Needle
£ "f + Charged
."'%&*. Droplets
- ' •* "
Spray
Nozzle
b. CHARGING VIA NEEDLE
Plastic Water
Feedline JM-.
Air Injected
to Segment
Water Column
Insulated
Spray
Nozzle ..••••:?•::•'•:•. + Charged
f-':X: Droplets
5-10 KV
+ Charge
c. DIRECT CONTACT WATER CHARGING
Feedline for
De-ionized Water
Grounded
Spray
Nozzle
.•:••' .. + Charged
-'*' -?:• Droplets
d. AUTOGENOUS CHARGING TO DE-IONIZED WATER
Figure 8. Means of producing a charged water spray (Source: Daugherty and
Coy, 1979).
21
-------
Charge transfer by conduction to the water and subsequently to the droplets at their
instant of formation can be achieved by connecting a source of charges (high voltage
supply) to the inflowing water. This method has been shown by many to be very effective
in charging the droplets to a high degree. However, this method requires that the entire
water supply and associated tubings be electrically isolated so that there is no current
leakage. Injecting bubbles into the water tube, as shown in Figure 8c, was suggested by
Hoenig (1979) as a possible solution for this problem.
The commercially available fog device ("Fogger") appropriate to our study was
originally developed by Dr. Hoenig, primarily under sponsorship by the U.S. Environmental
Protection Agency. The design of the Fogger had certain inherent problems, which not
only limited its application to certain types of fugitive sources, but also limited the
degree of control of fine particles. These problems are related to the need for high
pressure air or water to properly atomize water, poor charge-to-mass ratio of droplets
generated at reasonably acceptable waterflow rates, and the possibility of the nozzle
clogging if the water supply contains high concentrations of dissolved salts and suspended
solids. Thakur (1980) has reported dust and fog buildup on the induction ring and
subsequent spark formation.
In order to overcome these difficultites, the U.S. EPA contracted with
AeroVironment Inc. (AV), with Dr. Stuart Hoenig as AV's consultant, to design, develop,
and field test a new charged fog device. During the beginning of the project, a prototype
device, called a "spinning cup fog thrower" (SCFT) was built with Dr. Hoenig's assistance
at the University of Arizona in Tucson. This device, described in this section, attempted
ionized field charging of droplets. The SCFT was wind tunnel-tested in Tucson, and after
a complete analysis of the data collected, the instrument was abandoned. Contact
charging was thereafter used as the mechanism to charge the water droplets in the new
charged fog device. This chapter discusses the development and early setbacks of the
SCFT and gives its test results. The new charged fog device is discussed in detail in
Sections 6 and 7.
THE SPINNING CUP FOG THROWER (SCFT)
Basic SCFT Design
Figure 9 is a schematic representation of the first prototype SCFT designed and
built (with the assistance of Dr. Stuart Hoenig) at the University of Arizona. Figure 10 is
a photograph of the SCFT. The SCFT has three main components: a rotary atomizer, an
ionizer and a vane-axial blower. The rotary atomizer is composed of a small, hollow shaft
motor and a spinning cup. Water from a low-pressure source is introduced into the hollow
shaft and flows toward the other end where the spinning cup is mounted. Upon entering
the rear of the cup, the water stream strikes a rotating spider which deflects the water to
the sides. A sheet of water then flows toward the lip of the cup where droplets are
formed by centrifugal force and by the air stream striking the thin water layer, normally
leaving the edge of the cup tangentially. In an initial attempt, the droplets were passed
through a screen mesh maintained at a high voltage. However, this method of charging
the droplets had to be abandoned, because any mesh small enough to provide good charge
on the droplets also seriously altered the fog pattern.
22
-------
0.6 H.P. Motor
Vane Axial Fan
+15 KV
N>
Hollow Shaft Motor
Charged
Fog
Flow Straightener
Spinning
Cup
Figure 9. Schematic of the Spinning Cup Fog Thrower.
-------
N)
Figure 10. Original spinning cup fog thrower.
-------
The droplets were then charged by a stream of positive ions produced by an ionizer
containing numerous small discharge needles. It was expected that the ions produced in
the region of the ionizer ring would follow the airflow from the vane-axial fan and mix
with the droplets and charge them. Current to the ionizer ring was supplied by a vacuum
tube high-voltage dc power supply. The charged droplets were then deflected and
projected forward by a stream of air supplied by the vane-axial blower. However, this
method of charging of the droplets was later found to be inefficient.
Initial Wind Tunnel Test Design
In order to evaluate the particle control efficiency of the SCFT, a series of tests
was conducted in a 1.2mxl.2mxl8m wind tunnel provided by the University of
Arizona. Figure 11 schematically represents the wind tunnel. A standard dust (AC
coarse) of known particle size and density was pneumatically fed into the entrance of the
tunnel by a vibra-screw feeder. Compressed air was homogeneously mixed with the
incoming ambient air by a diffuser cone. The SCFT was mounted approximately 3 m
downstream of the tunnel entrance with the fog and air in concurrent flow. Plexiglas
viewing ports on both sides of the tunnel at the discharge end of the fogger allowed the
wet/dry interface to be observed closely. Two sets of sampling ports were positioned in
the front vertical wall of the tunnel; one set, approximately 2.5 m downstream of the
discharge end of the SCFT was to determine charge-to-mass ratio and droplet size, and
the other set at the far end of the tunnel was to determine particle mass concentration
and size. An axial fan pulled approximately 12,000 dscf/min of air through the tunnel.
To determine the control efficiency of the SCFT under various experimental
conditions, particle samples were collected according to the test plan shown in Table 1.
The test runs were completed for each scenario of Table 1 at approximately 10-30%
ambient relative humidity. An isokinetic sample was collected for each run. A sample
train was made especially for this purpose to isokinetically extract a representative
sample of particulate matter from the wind tunnel. Figure 12 is a photograph of the
sampling train. This train consisted of a pitot tube and associated inclined manometer
connected by a flexible sample line to a horizontal elutriator. The elutriator was designed
to have a 15-ym particle diameter cut-point (50% efficiency) at a flow rate of
0.55 m /min (20 scfm). Behind the elutriator was mounted a standard Sierra Instruments
five-stage cascade impactor which discharged to a silica gel trap to determine moisture
content. The flow was measured by a National Bureau of Standards (NBS) traceable
Laminar Flow Element manufactured by Merriam Instrument Company. The Laminar
Flow Element (LFE) was located downstream of the silica gel trap and connected to a
standard Dwyer magnahelic gauge for reading differential pressure. The flow rate was
metered by a large capacity needle valve downstream of the LFE followed by a Cadillac
centrifugal blower acting as prime mover. The particle sample collected was later
analyzed gravimetrically to determine both the total mass concentration and particle size
distribution for each test scenario.
For sample recovery, the front portion of the sample train and the horizontal
elutriator were rinsed with deionized water and placed in separate plastic sample
containers for transport back to the laboratory. There, the liquid/particle mixture was
transferred to tared beakers and evaporated to dryness to determine the total mass
collected. The samples were then resuspended in distilled water and deagglomerated
ultrasonically. A Leeds
-------
N>
Air Inlet
Plenum,
D iff user
Cone
7 /
Dust from
Vibrascrew Feeder
Spinning Cup /
Fog Thrower
Viewing
Port
Viewing
Port
r-0
3m
O
\ Particle
O~*-5ampl
C/m Sampling / Ports
Port O
ft
2.5m
18m
3m
Axial ( )
Blade ~X
Fan
0
I
Figure 11. Schematic of University of Arizona wind tunnel.
-------
TABLE 1. INITIAL SCFT WIND TUNNEL TEST PLAN
Medium Dust Loading
1
No Fog
Charged
Fog
1
Uncharged
Fog
Light Dust Loading
1
No Fog
Charged
Fog
I
Uncharged
Fog
*High or low liquid-to-gas ratio.
27
-------
NJ
00
Figure 12. Particle sampling train used during the wind tunnel studies.
-------
ruby laser and forward-scatter optical technique. The samples collected on the various
cascade impactor stages were also weighed using a standard analytical balance.
The charge-to-mass ratio of the droplets was determined using a special sample
train developed by AV, as shown in Figure 13. This train consisted of an insulated
stainless steel probe tip mounted on a standard glass midget impinger. The probe was
connected electrically to copper wool packing placed inside the midget impinger which
was also connected by a shielded cable to an electrometer (Hewlett-Packard Model 425A)
and associated strip chart recorder. The impinger was immersed in a Dewar flask
containing dry ice. An isokinetic sample of droplets was then extracted from the tunnel.
As the droplets moved through the impinger they were condensed out of the gas stream
and frozen, thus transferring their charge to the copper wool packing. The charge was
then measured by the electrometer and recorded by the chart recorder. Isokinetic
conditions were maintained by a Hastings mass flowmeter and needle valve with a
standard diaphragm pump acting as prime mover. A standard pitot tube and associated
inclined manometer were mounted next to the impinger/Dewar flask assembly with
adequate distance between them to avoid aerodynamic interference. The mass of water
collected in the impinger was determined gravimetrically at the end of each test run using
a portable triple-beam balance. Knowing the mass of droplets collected and the current
produced by them, the charge-to-mass ratio was calculated.
The size distribution of the droplets was measured simultaneously with the charge-
to-mass ratio by a KLD Model DC-2 droplet counter. This unit uses a hot-wire technique
to measure the size of the droplets in 14 incremental ranges from 1 um to greater than
450 urn. The hot-wire probe was mounted far enough from the impinger to avoid
aerodynamic interference with it.
INITIAL WIND TUNNEL TEST RESULTS AND DISCUSSION
Data from the first series of tests, as described above, are presented in Table 2.
The total mass concentration measured for various experimental conditions for a light
dust feed rate to the tunnel is shown in Figure 14, and for a medium dust feed rate in
Figure 15. It was found that the larger, agglomerated particles were being collected in
the first bend of the sample probe and were not allowed to settle out of the gas stream
due to the high velocity. For this reason, Figures 14 and 15 also show a second value,
where available, for the total amount of particulate matter caught by the sample train
less the larger particles collected in the probe.
The droplet size distribution measured using the KLD droplet counter showed a
mean droplet diameter of about 5 um. If this were indeed the case, the poor collection
efficiency may be partly due to the very short lifetime of the droplets due to evaporation
in the wind tunnel. The charge-to-mass ratio was observed to be only of the order of
10 C/g. It was also noticed that big satellite droplets of water, constituting as much
as 70% of the SCFT water output, were too large for particle agglomeration and were just
wetting down the wind tunnel.
A detailed analysis of the wind tunnel tests has demonstrated that a number of
conditions were responsible for the poor collection efficiency of the SCFT. The first of
these relates to the charging of the droplets. The charge on the droplets varied
drastically across the cross-section of the spray cone. With a positive charge applied to
29
-------
DESIGN SKETCH
Stainless Steel Nozzle
Glass Tube
Wire to
Electrometer
Midget
Impinger
Large Hose Clamp -
^
PVCPipe -:
,-— Dewer
Copper Mesh
Dry Ice
Figure 13. Sample train to determine the charge-to-mass ratio of water
droplets.
30
-------
TABLE 2. SUMMARY OF INITIAL SPINNING CUP FOG THROWER WIND TUNNEL TEST DATA.
Run
Scenario No.
Uncontrolled- 2
Med. Dust
Uncontrolled- 3
Light Dust
Uncontrolled- 4
Light Oust
Med. Dust; 5
Fogi No Chg.
Med. Dust; 6
Fog No Chg.
Light Dust; 7
Fog; No Chg.
Light Dust; 8
Fog; No Chg.
Med. Dust; High 9
L/C; High C/m(->
Med. Dust; High 10
L/C; High C/m(-)
Med. Dust; High II
L/C; Low C/m(-)
Med. Oust; High 12
L/G-, High C/m(t)
Med. Dust; High 13
L/G; High C/m(t)
Med. Dust; High 14
L/C; Low C/m(+)
Med. Dust; High 15
L/G; Low C/m(t)
Particle Mass Collected
Mass of Washings
A B Total
(gm) (gm) (gm)
0.899 -- 0.899
0.055 0.770 0.826
0.120 0.524 0.644
0.039 0.401 0.440
0.595 -- 0.595
0.653 - 0.653
0.602 - 0.602
0.749 0.049 0.798
0.585 - 0.585
0.473 - 0.473
0.669 - 0.669
0.739 - 0.739
0.675 0.085 0.760
0.611 - 0.611
Mass Collected by linpactor Stages
12345 Total
(gm) (gm) (gin) (gm) (gm) (gm)
0.0062 0.0236 0.0282 0.0370 0.0249 0.1199
0.0095 0.0368 0.0392 0.0442 0.0268 0.1565
0.0062 0.0266 0.0298 0.0356 0.0228 0.1210
0.0078 0.0292 0.0342 0.0387 0.0236 0.1335
0.0126 0.0392 0.0372 0.0410 0.0240 0.1540
0.0094 0.0294 0.0350 0.0402 0.0244 0.1384
0.0122 0.0376 0.0416 0.0470 0.0317 0.1701
0.0134 0.0372 0.0377 0.0431 0.0271 0.1585
0.0162 0.0410 0.0380 0.0399 0.0243 0.1594
0.0101 0.0294 0.0284 0.0320 0.0204 0.1203
0.0106 0.0295 0.0280 0.0304 0.0178 0.1163
0.0148 0.0434 0.0416 0.0441 0.0278 0.1717
0.0142 0.0404 0.0442 0.0448 0.0287 0.1723
0.0166 0.0419 0.0421 0.0464 0.0263 0.1733
Total
Mass Collected
in in
grams grains
J.OI89 15.7241
0.9825 15.1623
0.7656 11.8155
0.5735 8.8505
0.7490 11.5589
0.7914 12.2132
0.7721 11.9154
0.9565 14.7611
0.7444 11.4879
0.5933 9.1560
0.7853 12.1191
0.9107 14.0543
0.9323 14.3876
0.7843 12.1036
Average
Ftowrate
Through
Sampling Grain
Sampling Train Loading
Time (dscj/ (gr/
(min.) min) dscf)
37 19.08 0.0221
60 17.23 0.0147
60 18.03 0.0109
36 19.92 0.0123
36 16.62 0.0193
60 17.53 0.0116
60 18.18 0.0109
36 16.93 0.0242
36 17.60 0.0181
36 20.01 0.0127
36 22.00 0.0153
36 18.55 0.0210
36 20.92 0.0191
36 19.83 0.0170
Testing
Date
11/14/79
11/15/79
11/15/79
11/15/79
11/16/79
11/16/79
11/16/79
11/17/79
11/19/79
11/19/79
11/19/79
11/20/79
1 1/20/79
11/20/79
(continued)
-------
TABLE 2/continued
Run
Scenario No.
Unconlrolled- 16
Med. Oust
Med. Dusl; 17
Fog No Chg.
Unconlrolled- 18
Light Dust
Light Dust; 20
Fog; No Chg.
Light Dust; High 21
L/C; High C/m
Light Dusl; Low 22
L/C; Low C/m
Light Dust; Hifh 23
L/G; Low C/m
Uncontrolled- 21
Light Dust
Light Dust; Low 25
L/C; High C/m
High Dust; Low 26
L/C; High C/m
Med. Dust; Low 27
L/G; Low C/rn
Uncontrolled- 28
Med. Dusl.
Particle Mass Collected
Mass of Washings
A B Total
(gm) (gm) (gin)
0.858 - 0.858
0.353 0.365 0.718
0.781 - 0.781
0.127 0.311 0.738
0.1(21 0.434 0.855
0.326 0.357 0.683
0.337 0.221 0.558
0.724 - 0.72*
0.363 0.233 0.596
0.446 0.295 0.741
0.348 0.293 0.641
0.849 - 0.849
Mass Collected by Impactor Stages
12345 Total
(gin) (gin) (gm) (gm) (gm) (gm)
0.0134 0.0386 0.0400 0.0435 0.0277 0.1632
0.0197 0.0409 0.0392 0.0416 0.0269 0.1683
0.0120 0.0353 0.0373 0.0416 0.0268 0.1530
0.0120 0.0321 0.0360 0.0406 0.0274 0.1481
0.0131 0.0362 0.00&9 0.01)10 0.0328 0.1350
0.0111 0.0298 0.0342 0.0424 0.0270 0.1445
0.0101 0.0304 0.0342 0.0380 0.0267 0.1394
0.01201 0.0288 0.0316 0.0409 0.0305 0.1438
0.0092 0.0304 0.0356 0.0388 0.0257 0.1397
0.0094 0.0296 0.0310 0.0354 0.0250 0.1304
0.0089 0.0267 0.0292 0.0338 0.0251 0.1237
0.0120 0.0252 0.0276 0.0352 0.0282 0.1282
Total
Mass Collected
in in
grams grains
1.0212 15.7596
0.8863 13.6777
0.9340 14.4139
0.8861 13.6746
0.9900 15.2781
0.8275 12.7703
0.6974 10.7626
0.8678 13.3922
0.7357 11.3536
0.8714 13.4478
0.7647 11.8012
0.9772 15.0805
Average
Flowrale
Through
Sampling Grain
Sampling Train Loading
Time (dscf/ (gr/
(mm.) min) dscf)
36 16.63 0.0263
36 17.27 0.0220
60 17.38 0.0138
60 17.93 0.0127
60 25.14 0.0101
60 24.03 0.0089
60 24.52 0.0073
60 23.59 0.0095
60 26.03 0.0073
36 24.00 0.0156
36 22.65 0.0145
36 23.90 0.0175
Testing
Date
11/28/79
11/28/79
11/30/79
12/03/79
12/04/79
12/04/79
12/05/79
12/05/79
12/06/79
12/06/79
12/06/79
12/06/79
IsJ
-------
U)
U.U1 J
V
5' 0.010
z
o
5
H
Z
8
0 0.005
U
Q
0.000
-
—
;
—
_
| Total Catch
(1 Total -
u Probe Catch
-
-
NO UNCHARGED HIGH L/C LOW L/C HIGH L/C LOW L/C
FOG FOG HIGH C/m LOW C/m LOW C/m HIGH C/m
Figure 1*. Summary of test data for light dust loading.
-------
z
o
0.030
0.025
0.020
0.015
ft!
2
O
g 0.010
L>
to
Q 0.005
0.000
Total Catch
D Total -
Probe Catch
NO UNCHARGED HIGH L/G HIGH L/G HIGH L/G HIGH L/C
FOG FOG HIGH C/m HIGH C/m LOW C/m HIGH C/m
LOW L/G
LOW C/m
(0
Figure 15. Summary of test data for medium dust loading.
-------
the droplets, the outer edges of the spray seemed to exhibit a slight negative charge,
whereas the droplets in the center of the cone exhibited a strongly positive charge.
The size of the droplets themselves also seemed to contribute to the poor collection
efficiency of the SCFT. Given the size distribution measured by the KLD droplet counter,
the median droplet diameter of approximately 5 ym, and the low relative humidity in the
wind tunnel, the lifetime of the droplets was too short for effective control of inhalable
dust particles. Several attempts were made to increase the relative humidity of the gas
stream in the tunnel for the final series of runs. A header was constructed and mounted
with a series of spray nozzles which produced fine droplets. The header was then supplied
with high-temperature, high-pressure (200° F/800 psig) water from a spray cleaning
device. Even placing the header in a specially constructed extension to the wind tunnel at
its entrance did not give the water adequate time to vaporize and thus it wetted down the
entire tunnel. A mist evaporator was constructed from four evaporative cooler pads to
eliminate the water droplets and to allow more surface area for evaporation, and was
placed downstream of the spray header in the tunnel extension. This setup yielded
relative humidities of up to 80%; however, the increased pressure drop due to the mist
evaporator caused the flow rate in the tunnel to drop to approximately 25% of normal.
WIND TUNNEL TEST MODIFICATIONS
The data in Figures 14 and 15 show only about 50% reduction in the dust, which is
far short of the 90%+ targeted efficiency under controlled conditions. It was clear that
the efficiency could be improved by
increased charge-to-mass ratio of the droplets,
decreased air velocity through tunnel allowing more time for interaction
between the particles and droplets, and
countercurrent flow of dust versus droplets.
Therefore, to facilitate the capture and removal of dust from the air stream, these
changes were incorporated into the test setup. Several attempts were made to redesign
the ionizer to increase the number of positive ions produced and thus increase the charge
on the droplets. The first attempt was to install small wire brushes in addition to the
needles on the ionizer ring. This arrangement did not significantly increase the charge on
the droplets and was eventually discarded. The most successful method was to isolate the
spinning cup itself at a high voltage and to supply water from a separate line into its
center. The original ionizer ring, without brushes, was retained to produce a constant
stream of positive ions, as before. Both the cup and the ionizer ring were to have
separate high-voltage power supplies.
The velocity of the air moving through the tunnel was decreased from approximately
1200 fpm to approximately 475 fpm by a larger pulley on the fan in conjunction with the
mist eliminator used for humidifying the gas stream. Further reductions in velocity were
not practical due to limitations of the velocity-measuring instrumentation.
In the original experiment, the SCFT was installed inside the tunnel, projecting fog
concurrently with the air/dust mixture. Arrangements were made to reverse the direction
35
-------
of the fog 135° relative to the air stream, with the SCFT unit itself outside the wind
tunnel. A complete 180° shift was not possible due to potential dust buildup on the
insulators. In addition to these modifications, a new screw was installed in the Vibra-
Screw Dust Feeder. Thus, more dust could be fed to the tunnel and the test run time
decreased substantially.
After these modifications a number of new problems were encountered. The first
was current leakage from the cup through the external water line. Several expedient
methods to solve this problem were attempted, such as installing a pressure nozzle on the
copper water line to break up the stream into coarse droplets, but measurements showed
that significant current still followed the water flow depending on the distance from the
pressure nozzle to the cup. The most effective method to reduce current leakage was to
inject air bubbles into the water stream before atomization, using a large-bore
hypodermic needle and compressed air. This technique did not completely eliminate the
problem but did reduce the leakage to a tolerable level.
The second major problem concerned actually generating the fog. The rotary
atomizer tended to produce large satellite drops which impinged on the tunnel walls and
thus did not contribute to the capture of dust particles. In addition, the small droplets
entrained in the airstream provided by the blower evaporated very quickly at ambient
relative humidity (10% to 30%). Both these factors severely limited the effectiveness of
the SCFT to control fine particles. To solve these problems and to provide droplets of an
effective size and concentration, it was necessary to redesign the whole device.
Associated with this new design would be a modified method for charging the droplets
which would achieve a high charge-to-mass ratio, while remaining compatible with the
new atomizer.
36
-------
SECTION 6
THE CHARGED FOG GENERATOR (CFG)
INTRODUCTION
The wind tunnel tests of the spinning cup fog thrower at the University of Arizona
revealed the need for a new system capable of producing well-charged fog. To assure the
highest possible particle control efficiency with charged droplets, the new device had to
achieve the following goals:
(1) It must be able to produce sufficient fog to control particles at typical
material handling areas.
(2) The size distribution of the fog generated should be most suitable for inhalable
particle control.
(3) It must incorporate an efficient method to charge the droplet so that the
charge-to-mass ratio would be suitable for inhalable particle control.
(4) It must project the charged droplets to the dust particle-laden area and
provide for sufficient residence and contact time between the droplets and the
particles, and
(5) The unit should be somewhat small, preferably portable, and of fairly low
power consumption for operation in areas where commercial power is not
available and preferably operable from a mobile vehicle.
A commercially available device using a spinning cup and blower configuration for
furnace oil atomization was determined to be adaptable to achieve the above goals. Thus,
as the core of the new CFG, a Ray Oil Burner was chosen.
CHARGED FOG GENERATOR DESIGN
The type AG Ray Oil Burner is a horizontal, rotary atomizing unit consisting of only
one movable part, a hollow steel shaft upon which are mounted the atomizing cup, the fan
and the motor rotor. These parts comprise the shaft assembly which rotates as a unit on
two ball bearings, as Figure 16 shows. Figure 17 is a schematic representation of the
CFG. A stationary tube extending through the center of the hollow main shaft of the
burner into the atomizing cup introduces the fuel — in this case, water. The atomizing
cup is long and tapered, which permits the water to form a thin film on the inside of the
cup before reaching the rim, thus discharging evenly.
37
-------
. ATOMIZING CUP
• ANGULAR VANE NOZZLE /
• FURNACE HINGE PLATE MOUNTING
FAN
^•TOTALLY ENCLOSED MOTOR
rAIR COOLED MOTOR JACKET
FRONT BEARING
STATIONARY FUEL TUBE
ROTATING HOLLOW MAIN SHAFT
REAR BEARING
WORM
WORM GEAR
Figure 16. Cross-sectional diagram of a type "AG" Ray Oil Burner.
38
-------
Air Fan
Nonconductive Air Cone
Water-Deflecting
VO
Nonconductive
Spinning Cup
Hollow Shaft Motor
DC Power Supply
Nonconductive
Water Tube V
X
Spinning Shaft
Nonconductive
Water Line
Flowmeter
DC Water
Figure 17. Schematic diagram of the Charged Fog Generator.
-------
Water is introduced through the water tube into the 3600-rpm spinning cup, whose
inside is fabricated to a gradual smooth taper (Figure 18). A small deflecting baffle is
attached to the open end of the water tube so that the water will be deflected 90 and
strike the rear end of the spinning cup. Because of the centrifugal forces, the water
becomes a thin film and moves forward into a high velocity airstream from the axial fan.
The impact of the high velocity air on the thin water film instantly breaks the water film
into fine water droplets.
The fuel tube, air cone, and the rotating cup are made of nonconducting materials.
The water tube is firmly attached to the rotating cup, thus rotating with the cup. The
other end of the water tube is attached to the water supply through a rotating seal (see
Figure 19). The water for atomization is stored in an electrically isolated reservoir
(120-liter capacity) and a small pump is used to pump it to the inlet of the rotating seal.
The water flow rate can be varied from about 4 liters to 70 liters per hour. The flow
meter can be seen in Figure 20.
The oil burner is equipped with Ray adjustable angular vane nozzles which allow
spray pattern adjustments. The vanes of the nozzle can be angled from 50 to 90°. The
50° nozzle produces a short, wide spray, while the 90° nozzle produces a long, narrow
spray. The shape of the spray is also controlled by the airflow around the cup. The
airstream from the fan can be controlled by an air butterfly setting (1 to 9 scale), and this
airstream projects the fog forward. By adjusting this setting, the airflow speed can be
controlled, thereby controlling the shape of the fog spray pattern. The spray pattern
covers a volume of 16 to 24 cubic meters. Figures 21 and 22 show two possible spray
patterns — one long and narrow, and the other broad and short. Thus, the spray pattern
can be easily adjusted to better conform to the shape and size of the source of dust where
the charged fog is to be applied.
To achieve high charge-to-mass ratio for the droplets, this and other studies have
shown that directly charging the inflowing water is the best alternative (Pilat, 1975;
Kearns and Harmon, 1979). Thus, the water from the reservoir tank to the rotating cup
(until it becomes fog) has to be electrically isolated and maintained at a high potential. A
15-kV dc power supply was directly connected to the inflowing water near the rotating
seal. Although the voltage applied was very high, the current used was only fractions of
milliamperes and therefore, in our experience, posed no significant safety hazard, even to
someone accidentally touching the exposed high-voltage areas.
During the initial development stages of this method of droplet charging, the water
tube was stationary, causing the water to leak to the casing of the oil burner and thereby
the current to leak to the ground. This problem was solved by attaching the water tube to
the rotating cup and connecting the rotating seal at the other end. Under the present
configuration, the CFG is capable of producing well-charged fog (Mathai, 1983a), with a
droplet size distribution suitable for inhalable particle collection (see discussion below on
size distribution and charge-to-mass ratio).
SIZE DISTRIBUTIONS OF WATER DROPLETS
Initially, the size distribution of the droplets was determined by collecting the
droplets on greased (silicon oil) glass slides and observing them under a microscope. As
the spherical water droplets undergo a shape deformation when they are collected on
slides, a conversion factor of 1.26 (as suggested by Pilat et al., 1974) was used to obtain
-------
Figure 18. Rotating cup and air cone of the Charged Fog Generator.
-------
Figure 19. Rear of the Charged Fog Generator, showing rotating seal area.
-------
Figure 20. Side view of the Charged Fog Generator.
-------
•p-
-c-
Figure 21. Typical long and narrow spray pattern from the Charged Fog Generator.
-------
Figure 22. Typical short and broad spray pattern from the Charged Fog
Generator.
-------
the real droplet diameter. By repeating this method several times, a typical mean droplet
diameter of about 90 ym was obtained.
Later, a KLD Model DC-2 droplet counter was used to measure the droplet
diameters. The range of values given by this device was not in agreement with the above
measurements. The droplet counter was sent to its manufacturer, KLD Associates, and
they confirmed that the DC-2 counter was not operating properly. Therefore,
Meteorology Research, Inc. (MRI), Altadena, California, was hired to accurately
determine the droplet size distribution.
The droplet size measurements were made at AeroVironment using a cloud optical
array probe (manufactured by Particle Measuring Systems Inc. (PMS), Boulder, Colorado)
for droplets in the aerodynamic size range of 30 to 300 vim and using a precipitation
optical array probe for droplets in the range of 125 to 1875 ym. The probes were
calibrated with glass beads of known diameter just before data were taken.
In the PMS optical array probes, a collimated laser beam is projected at right angles
to the air and fog flow (using mirror systems) and focused on a photo diode array. A
droplet entering the light beam casts a shadow over part of the diode array, causing the
recording system to register a count. The size of the droplet is determined from the
number of elements in the diode array which are shadowed. The droplets are sized into 15
channels in both instruments.
From the probe measurements, droplet concentration in unit volume and unit
micrometer intervals, and droplet mass median diameter were calculated. Figures 23 and
24 show the concentration and mass distributions (provided by MRI), respectively.
Logarithmic scales are used for both figures since the values vary over an order of
magnitude. These measurements gave a mass median diameter (MMD) of about 200 vim.
This MMD is a good indicator of the central point in the distribution about which most of
the mass is clustered. The concentration distribution gives a median droplet diameter of
about 100 ym. This is an upper limit because the probe was not able to count droplets
smaller than about 30 ym. This value is consistent with the value obtained using the
microscope method.
It may also be possible to mechanically control the size distribution generated by
the CFG to a certain degree. The best size distribution can be selected only after
completing the actual field tests and optimizing different parameters for the best possible
collection efficiency for the CFG.
CHARGE-TO-MASS RATIO OF THE DROPLETS
The charge-to-mass ratio of the water droplets generated by the CFG was
determined using the method described earlier in Section 5. This study yielded a typical
charge-to-mass ratio of 1.2 x 10~ C/g, when the water was charged by direct contact at
15 kV. The Rayleigh limit of charges on a 200-ym droplet is about 4.2 x 10 C/g. This
value may be compared with a typical value of 0.56 x 10 C/g obtained by Pilat et al.
(1974) and Pilat (1975) when the droplets were charged with a 5000-volt inductance
charging method, and a value of 0.1 x 10 C/g for the largest of the commercially
available fogger devices, Fogger IV (Brookman et al., 1981, 1982).
46
-------
£
Z
2 10
'1
QJ
H
uj
U
Z
8
OS
UJ
cfl
? 10
l-
UJ
8
-2
10
1 1 1 1 1 1 N
1 11111)11
1 1 1 1 II
1 1 1
1 1 1 1 1
ll
101 102
DROPLET DIAMETER (ym)
1 1 1 1 1 1 1 1
10-
Figure 23. Water droplet number concentration as a function of droplet
diameter measured using a cloud optical array probe and a precipi-
tation optical array probe.
-------
10
-1
I0
~2
in
e/i
a
g io-3
Q
10'
IO1
I I I 11 li
I I
I II ll
101 102
DROPLET DIAMETER (ym)
i i MM
10-
Figure 24. Water droplet mass as a function of droplet diameter measured
using a cloud optical array probe and a precipitation optical array
probe.
-------
To estimate the charge-to-mass ratio of the droplets, another experiment was
performed. A standard 8-inch diameter, No. 120 sieve (125-ym mesh size) was mounted
in the path of the droplets. The electrical charges transferred to the mesh from the
charged droplets are allowed to leak to the ground through a 1-megaohm resistor. The
voltages generated across this resistor were measured with a Hewlett-Packard HP 419A
dc null voltmeter.
With a water flow rate of about 60 1/h and an applied high voltage of 10-kV to the
inflowing water, the current measured in the circuit described above was about 4 uA. It
was determined, for this measurement setup, that about one-fourth of all droplets
generated by the CFG were colliding with the 8-inch diameter sieve. Again, using the
200-ym mass median droplet diameter value, the charge-to-mass ratio for this particular
arrangement was calculated to be about 1 x 10 C/g. The experiment was repeated for a
different water flow rate and the result was consistent with the above value. However,
when the applied voltage was reduced to 5 kV, the observed current was reduced to
one-half the value obtained with 10 kV applied voltage. These values are quite consistent
with the gravimetric method for determining the charge-to-mass ratio of the droplets of
1.2 x 10~6 C/g reported above with 15 kV applied voltage. It is clear that the charge-to-
mass ratio increases with increases in the applied voltage (at least in the range used in
this study).
It is interesting to estimate the number of elementary charges carried by a typical
droplet. As explained above, &uA current was generated by 10° droplets (see below),
which corresponds to 2.5 x 10 elementary charge transfers. Therefore, assuming the
typical mass median droplet diameter to be-about 200 vim, the number of elementary
charges carried by a typical droplet is 2.5 x 10 unit charges.
The following discussion illustrates the electric energy supplied to the droplets
under various water flow rates and droplet charges.
At a water flow rate of 60 1/h flowing into the rotating cup of the CFG, and
assuming the mass median diameter of the droplets to^be about 200 ym, the number of
droplets generated per second is approximately * x 10 droplets. Now, the maximum
charge a droplet of radius R can carry (Rayleigh limit) is given by Equation (6)
QRay = (2 x ID'5) R3/2 (6)
where Q is in Coulombs and R is in meters.
Assuming no current leaks, therefore, the upper limit of current drawn from the
high voltage power supply is given by
IL = 4xl06(2xlO-5)R3/2 (7)
where limiting current, L, is in amperes, and R is in meters.
-------
For a 200-um diameter droplet, this gives a limiting current of 80 uA. If the water
flow rate is reduced to half (30 liters per hour), the corresponding value of this limiting
current drawn will be half, or 40 uA.
If the droplet/liameter is halved, eight times as many droplets will be generated per
second i.e., 32 x 10 droplets at a water flow rate of 60 liters per hour. In this case, the
corresponding limiting current will be 226 uA. Similarly, for 50-ym, 25-um, and 12.5 ym
mass median droplet diameter size distributions, the maximum possible currents will be
0.54 mA, 1.82 mA, and 4.2 mA, respectively. Again, these values are directly propor-
tional to the water flow rates. Since 60 to 70 liters per hour is the upper limit of the CFG
water flow rates, these values are the upper bound for various hypothetical mass median
droplet diameters.
GENERAL REMARKS ON THE CFG
The CFG is fairly small, portable, and mounted on a movable platform. The total
power requirement is about 1 kW (110 Vac power supply). It has no nozzle-clogging
problems and does not require compressed air. With the use of a small inverter, the unit
can be operated at a remote location where commercial power is not available. This
feature may be important, because the system has high potential application at sources
where commercial electric power is not available.
These measured droplet size distributions, charge-to-mass ratios, and droplet charge
estimations indicate that the goals laid down for the new CFG at the beginning of this
section were achieved. The physical characteristics of the charged fog droplets from the
CFG are such that high inhalable particle control efficiency is expected. The inhalable
particle control efficiency of the CFG in a field situation is described in the next section.
50
-------
SECTION 7
EVALUATION OF THE INHALABLE PARTICLE
CONTROL EFFICIENCY OF THE CFG
As indicated in Section 4, an exact theoretical analysis of the inhalable particle
control efficiency of any charged fog device is very difficult because:
(1) Although the typical size distribution of the droplets is known, size variation
of droplets and particles at various locations of the test volume may not be
fully known.
(2) Concurrent with uncertainty about the size distribution of droplets and
particles, as in (1) above, the magnitude of charges on the droplets and the
magnitude and nature of charges on the particles may also not be fully known;
(3) Since charged fog technology is applied primarily in open areas containing a
dust cloud, seldom are external variables, such as wind speed and direction,
relative humidity, temperature and diffusive characteristics of the dust cloud,
controllable or can they be accurately determined.
Therefore, in order to fully evaluate the particle control efficiency of a charged fog
device, extended field tests under various combinations of possible field conditions and on
dust clouds of various materials (coal, soil dust, mineral dust, etc.) need to be performed.
As a first step toward this goal, a detailed field test plan for the project's second phase
was prepared for the EPA project officer's approval.
Specifically, the objective of the Phase II field tests was to evaluate the particle
control efficiency of the Charged Fog Generator on bentonite ore and under various
instrument settings and meteorological conditions. From these results, one could assess
the optimum instrument settings and field conditions. Ideally, this task would have been
performed under a controlled experimental setup (wind tunnel study). However, because
of financial and time constraints, it was decided to go ahead with the field program.
FIELD TEST SITE AND EXPERIMENTAL SETUP
The Charged Fog Generator was field tested at the Kaycee Bentonite Corporation's
bentonite processing plant in Worland, Wyoming, during the first half of 1981 (Mathai,
1983b). Worland is about 250 km northwest of Casper, Wyoming. Bentonite is a highly
water-absorbing material used mainly in sealing water leaks in oil wells. Bentonite ore is
unloaded from front-end loaders onto the grill of a hopper which is attached to the west
wall of the plant building (see Figure 25). From the bottom of the hopper (approximately
4 m below the hopper grill level, inside the plant) the ore is carried by conveyor belts to
51
-------
K>
Figure 25. Exp>erimental setup at the bentonite unloading operation at Worland, Wyoming.
E is the location of the Charged Fog Generator and D, the particle sampler.
-------
processing areas. The hopper is completely enclosed except on one side through which the
front-end loaders unload. The hopper is 6.7 m wide, about 2 m deep and, from the grill
level, about 3 m high. The grill is inclined about 30° to the horizontal.
Bentonite ore from two large piles, approximately 100 m to northwest and southwest
of the hopper, are carried to the hopper by front-end loaders and dropped on the grill
(Figure 26). The bucket of the front-end loader is about 2.4 m wide and successive loads
are unloaded uniformly over the 6.7-m wide hopper. It takes about ten front-end loader
dumps to fill the hopper to the grill level. These ten dumps are accomplished in about 25
minutes. The bucket is removed from inside the hopper area in 20-25 seconds. One full
hopper of ore will be carried away by the conveyor belt in about an hour.
Bentonite samples collected from the hopper area were resuspended to measure
particle size distribution at IERL, EPA. This analysis showed that only 8% of the
particles were smaller than 37 um (Drehmel, 1981). The fraction of inhalable particles
will, therefore, be even lower (maybe 3-5%). A source with a higher fraction of inhalable
particles would have been preferable. It was hoped that this fraction would be higher for
airborne bentonite samples than for the one collected from the hopper wall.
The land around the plant is fairly flat. A railroad track lies on the east side of the
plant and a paved road (very little traffic) beyond that. Bottom-dump trucks filled with
bentonite ore arrive near the storage piles from the south and, therefore, the dust in the
hopper area due to transport of ore to the plant area is negligible compared to the dust
generated in the hopper itself. In other words, the hopper can be considered as a fairly
isolated source. The mean wind speed and direction in the area was estimated (using
National Weather Service data) for the February-May period to be about 2.8-3.6 m/s.
Daytime high temperatures during the test period varied from 9°C to 35°C.
The Charged Fog Generator and the particle sampling instruments were mounted on
the outside of the south wall of the hopper, marked as E and D in Figure 25, about 4 m
above ground level. A platform was built to mount these instruments. Ideally, the
particle sampler inlet should have been mounted on the east (rear) wall of the hopper, but
for practical reasons could not be. The CFG sprayed water droplets across the hopper
above the grill. The total volume to be treated by the charged fog was about
(6.7 m x 2 m x 3 m) 40 cubic meters. This volume is somewhat larger than the maximum
coverage of the CFG fog. Unfortunately, at that time only one prototype CFG was
available for tests. To have had a second unit mounted on the north wall of the hopper
and operated concurrently would have been ideal.
Because of the physical layout and the nature of the operation, a new particle
sampler system, different from that described in Section 5, was installed. Particle
samples were collected using a Sierra Model 230 CP cyclone preseparator followed by a
Sierra two-stage cascade impactor. The cyclone's air inlet protruded 0.3 m into the
hopper at location B in Figure 27. The CFG's spinning cup and aircone, marked as A in
Figure 27, are 1 m away from cyclone's air inlet. This particular experimental setup was
used because of practical limitations; this was done with the approval of the EPA project
officer. The relative positions of A and B with respect to the hopper grill (C) can also be
observed from this figure.
53
-------
•
Figure 26. Typical dust plume generated when front-end loader unloads the bentonite on the
hopper grill.
-------
Figure 27. View of the rotating cup (A) of the Charged Fog Generator,
inlet of the Cyclone preseparator (B) and the hopper grill (C)
from inside the hopper.
55
-------
The particle sampling system was operated at a flow rate of 0.85 m /min (30 CFM).
At this flow rate, the cyclone has a particle cut-point of about 7.3 um. The impaction
plates of the cascade impactor were chosen so that the upper filter would collect
particles larger than 1.8 pm. According to the manufacturer, because the collection
efficiency curves of the cyclone and impactor are not perfectly sharp, and because
approximately 30% of the particles larger than the cyclone's cut-point of 7.3 jim will pass
through the cyclone to the impactor, the size range of particles collected on the upper
slotted filter will be from 1.8 um to somewhat larger than 7.3 ym (Sierra Instruments,
1981). The lower back-up filter collects all particles smaller than 1.8 vim. For the
purpose of this report, particles collected on the back-up filter will be characterized as
the fine particle fraction and those on the upper filter will be characterized as the coarse
fraction. The mass of coarse fraction collected was often an order of magnitude smaller
than the fine fraction, possibly indicating a severe particle bounce problem, with particles
larger than the cut-point of 1.8 vim reaching the back-up filter. Subdividing the samples
into more size ranges would have required either much longer sampling time or much
higher flow rates to obtain acceptable filter loadings. Neither of these alternatives was
desirable. The sampler flow rates were calibrated at regular intervals during the entire
test program.
ESTIMATION OF PARTICULATE MATTER MEASUREMENT ACCURACY AND
PRECISION
All the particle control efficiency evaluations were derived from particulate matter
concentration measurements made during various test runs. Particulate matter concen-
trations were determined by dividing the mass of particulates collected on the filter
media by the volume of air sampled. Therefore, the uncertainty in mass concentration
values involves the uncertainties introduced by the filter weighing process and sampling
volume measurements. At the mass concentrations encountered in this experiment, the
uncertainty due to the filter weighing process was negligible compared to the corres-
ponding value due to sample volume measurements.
If we define the precision to be given by the relative standard deviation, precision of
volume measurement is given by Mueller and Hidy (1983) as,
CTv
-F2
• F2
(FJ + F2)2
(8)
where Fj and ^^ are tne initial and final sample flow rates, V is the sample volume, T is
the sampling duration, and CT'S refer to the corresponding standard deviations. Assuming
that the flow rates remain constant throughout the sampling period (i.e. F, = F- = F), and
that T can be measured precisely, we get,
56
-------
It is reasonable to assume that the precision of the flow rate values during flow
calibrations is representative of the precision of the actual sample flow rate measure-
ments; one can then calculate ffv/V using the above equation and the flow calibration
records. From our experience in various mass concentration measurement projects, it is
estimated that the precision of sampling volume measurements reported here is in the
3-5% range (Mathai, 1983c; Mathai, et al., 1983). Therefore, it is estimated that the
precision of particulate matter data presented in this report is in the range of 3-5%.
The accuracy of the mass concentration measurements can be estimated from the
accuracies in filter weighing and sample volume measurements. Again, at mass
concentrations encountered in this experiment, the overall accuracy is dominated by that
of sample volume measurement. The latter value can be estimated using the calibration
records. Again, from our prior experience, we estimate that the accuracy of particulate
matter mass concentrations reported are in the 1-3% range.
Particle size cut-point data given earlier in this section were based on information
provided by the instrument manufacturer. No attempt was made during this experiment
to verify these cut-points independently, mainly due to financial and time constraints of
the project.
FIELD TEST DESIGN
To determine the particle control efficiency of the Charged Fog Generator, three
test scenarios were designed. In the first scenario, no attempt was made to control the
dust inside the hopper, except that the CFG's fan blew continuously. This test scenario
was necessary to ensure that the mixing of the dust cloud was nearly identical among
various runs in order to make direct comparison of tests with and without charged fog. In
the second scenario, uncharged fog was applied on the dust clouds; and in the third
scenario, charged fog was applied. Other governing parameters were varied under each
scenario to yield a statistically acceptable set of test data.
The parameters varied were water flow rates, fog pattern (short or long), applied
high voltage, wind conditions, and relative humidity. In addition, a group of test runs was
performed with the polarity of charges on the droplets reversed. Table 3 shows the
planned test program.
An alternate combination of field conditions was given by:
Low wind/Low humidity/Broad spray
Low wind/High humidity/Long spray
Medium wind/Low humidity/Long spray
Medium wind/High humidity/Broad spray
as the column headings in Table 3, with the same water flow rates and charge conditions.
Either of these test protocols called for a total of 32 test runs.
57
-------
TABLE 3. PROPOSED CFG FIELD TEST PLAN.
Run No.
1-4
5-8
9-12
13-16
17-20
21-24
25-28
29
30
31
32
Medium Wind,
High Humidity,
Narrow Spray
Water
Flow
(1/h) Charge*
No water
30 none
60 none
30 V.
30 M12
60 V.
60 V*
Medium Wind,
Low Humidity,
Broad Spray
Water
Flow
(1/h) Charge*
No water
30 none
60 none
30 V.
30 V^
60 V.
60 V^
30 V.
30 V,
60 Vf
60 V^
Low Wind,
High Humidity,
Broad Spray
Water
Flow
(1/h) Charge*
No water
30 none
60 none
30 V.
30 V*
60 V.
60 V*
Low Wind,
Low Humidity,
Narrow Spray
Water
Flow
(1/h) Charge*
No water
30 none
60 none
30 V.
30 V*
60 V.
60 V*
and V? are two voltages.
58
-------
However, in actual practice, data had to be collected under prevailing field
conditions and, therefore, more test runs were performed under some columns of Table 3,
and fewer runs were performed under others. Also, because of fluctuations in the dust
level observed from day to day, several times more than the four planned 'no water1 runs
had to be recorded. In total, 96 runs were performed instead of the planned 32.
Particle samples were collected on preconditioned and preweighed glass fiber
filters. Each particle sample was collected during 12 to 15 front-end loader dumps
(roughly 30-40 minutes). Simultaneously, wind speed and direction and relative humidity
were also recorded. After each sample was collected, the filters were transferred to
special envelopes and brought to AV's laboratory for analysis.
DATA PROCESSING
Particle samples collected on glass fiber filters were returned to the laboratory on a
weekly basis along with the field logs. These filters were conditioned in a constant
temperature and humidity chamber for at least 24 hours before final weighing The
sampling time, the number of front-end loader dumps, and the sampling flow rates, as
well as meteorological data were obtained from the field log. As discussed in the previous
section, particle samples were collected as two fractions, fine and coarse. The mass of
each fraction and sum of the two were obtained from the final and initial weights of the
filters. Using known values of the sample time, number of dumps, and flow rate, particle
mass concentration for a single dump is calculated from:
M. x 106 -
Ci =1^T7F
where
C. = particle concentrations; i = 1, 2, or 3 for fine and coarse fraction and the
1 sum of these, respectively
M. = particle mass collected on filter media, in grams
N = number of front-end loader dumps in each sample
T = sample time in minutes
F = sample flow rate in m /min.
Particle concentrations were calculated for all the 96 test runs (fine and coarse
fractions separately and their sum) and they are presented in Table 4.
As we indicated earlier, the amount of dust generated in the hopper fluctuated from
dump to dump. To overcome this variation, each sample was collected during a total of
12 to 15 dumps and the mean value for each dump was calculated. The amount of dust
generated also varied from day to day. This variation was caused by factors such as
changes in ambient conditions and the moisture content of the bentonite ore itself. The
59
-------
TABLE 4. MEASURED PARTICLE CONCENTRATIONS (yg/rri ) INSIDE THE BENTONITE UNLOADING
HOPPER DURING CFG FIELD TESTS.
Test Run
Date No.
4/22/81 1
2
3
4/23/81 4
5
6
4/24/81 7
8
9
10
4/25/81 11
12
13
4/28/81 14
15
16
17
18
4/29/81 19
20
21
22
Particle Concentrations (ug/m )
Fan Only
Fine Coarse Total
1198 120 1318
1940 208 2148
2382 123 2505
1323 133 1456
1936 126 2062
3435 205 3640
Uncharged Fog
Fine Coarse Total
426 59 485
1702 107 1809
1936 119 2055
980 72 1052
997 122 1119
1704 163 1867
1904 148 2052
1622 190 1812
Charged Fog
Fine Coarse Total
341 55 396
988 196 1184
923 94 1017
713 165 878
113 131 244
562 230 792
1523 197 1720
1556 116 1672
(continued)
-------
TABLE it (continued)
Test Run
Date No.
5/20/81 23
24
25
26
5/21/81 27
28
29
30
31
5/22/8 1 32
33
6/2/81 34
35
36
37
6/3/81 38
39
40
6/4/81 41
42
43
44
45
46
47
Particle Concentrations (yg/m )
Fan Only
Fine Coarse Total
1186 197 1383
1491 102 1593
1825 298 2123
2225 133 2358
1783 151 1934
588 96 684
1345 174 1519
1579 116 1695
Uncharged Fog
Fine Coarse Total
658 143 801
707 193 900
507 355 862
1481 111 1592
1506 104 1610
711 75 786
1462 101 1563
Charged Fog
Fine Coarse Total
544 167 711
506 125 631
469 178 647
294 199 493
133 122 255
405 58 463
74 130 204
149 220 369
1210 73 1283
1277 72 1349
(continued)
-------
TABLE 4 (continued)
Test Run
Date No.
6/5/81 48
49
50
6/8/81 51
52
53
54
6/9/81 55
56
57
58
59
60
6/10/81 61
62
63
64
6/11/81 65
66
67
68
69
Particle Concentrations (ug/m )
Fan Only
Fine Coarse Total
1291 84 1375
3097 285 3382
1091 50 1141
2196 207 2403
2384 111 2495
1672 105 1777
Uncharged Fog
Fine Coarse Total
987 67 1054
1197 156 1353
930 260 1190
1469 107 1576
1035 128 1163
Charged Fog
Fine Coarse Total
619 96 715
640 55 695
1079 96 1175
648 146 794
492 111 603
920 71 991
943 68 1011
1207 245 1452
1140 271 1411
1134 103 1237
1111 74 1185
(continued)
-------
TABLE 4 (continued)
Test Run
Date No.
6/12/81 70
71
72
73
6/15/81 7*
75
7/7/81 76
77
78
79
80
81
82
7/8/81 83
84
85
7/9/81 86
87
88
89
90
91
92
Particle Concentrations (ug/m )
Fan Only
Fine Coarse Total
1256 166 1422
546 238 784
1144 105 1249
6918 284 7202
2537 305 2842
2896 197 3093
2689 139 2828
Uncharged Fog
Fine Coarse Total
1393 174 1567
894 42 936
1566 133 1699
Charged Fog
Fine Coarse Total
901 275 1176
731 85 816
197 119 316
862 145 1007
832 101 933
2655 190 2845
2920 185 3105
1977 143 2120
2013 260 2273
1916 94 2010
1761 123 1884
2210 144 2354
1570 109 1679
OS
VjJ
(continued)
-------
TABLE 4 (continued)
Test Run
Date No.
7/10/81 93
94
95
96
Particle Concentrations (ng/m )
Fan Only
Fine Coarse Total
2253 152 2405
Uncharged Fog
Fine Coarse Total
1862 117 1979
Charged Fog
Fine Coarse Total
1724 264 1988
1784 129 1913
ON
-------
effect of the latter is not too significant since the ore stored in each pile comes from the
same mine and a pile is completely processed before a new pile is started. To overcome
the day-to-day dust level fluctuations, sample concentrations were normalized for each
day with respect to the background value (the "fan only" value) and a percentage particle
collection (control) efficiency was calculated. Percentage particle control efficiency is
obtained from:
C -C.
E. = ° ' xlOO (11)
J ^r,
where
E. = particle control efficiency (percentage); j = 1 for uncharged fog and
J j = 2 for charged fog;
CQ = particle concentration when no fog is applied (fan only — dry run); and
C. = particle concentration when the fog (charged or uncharged) is applied.
Values of E. are calculated, as before, for the fine fraction, the coarse fraction and
the sum of fine and coarse. Table 5 gives these percentage particle control efficiencies
and the corresponding field condition data, water flow rate, applied high voltage, and
comments, if any, for the whole field test program.
DATA ANALYSIS, DISCUSSION, AND RESULTS
As indicated earlier, data were collected under prevailing field conditions and
various combinations of instrument settings. This resulted in a data base with a large
number of variable parameters making data analysis very difficult.
Comparison Between Charged and Uncharged Fog
Figure 28 shows the mean values of the measured percentage fine particle control
efficiency and total particle control efficiency for charged (striped bars) and uncharged
(solid bars) fog. For this comparison, all test runs under all instrument settings and field
conditions are included. The mean and standard deviations of the fine particle control
efficiency are 48.1% and 23.0%, respectively, for the charged fog and 27.8% and 25.3%,
respectively, for the uncharged fog. The corresponding values for all the particles (fine
and coarse together) are 44.5%, 21.8%, 25.0%, and 24.4%, respectively. These numbers
show that, even under average field conditions and instrument settings, inhalable particle
control efficiency can be almost doubled by electrically charging the water droplets.
However, under optimum instrument settings and favorable field conditions, the improve-
ment in inhalable particle control efficiency can be expected to be higher. It may also be
pointed out that the volume of the dust cloud treated was somewhat larger than the
maximum coverage of the CFG; therefore, the observed improvement in particle control
efficiency is encouraging.
65
-------
TABLE 5. MEASURED INHALABLE PARTICLE CONTROL EFFICIENCY (PERCENTAGE), METEORO-
LOGICAL CONDITIONS AND CFG SETTINGS DURING THE FIELD TESTS ON BENTONITE
ORE.
Test Date
4/22/81
4/23/81
4/2'l/XI
4/25/81
4/28/81
4/29/81
5/20/81
5/21/81
5/22/81
6/2/81
Uncharged Fog
Fine Coarse Total
64 51 63
12 49 16
19 3 18
60 42 58
25 9 15
12 -30 10
2 -18 1
53 7 50
45 27 42
53 -89 44
66 -248 46
26 23 26
25 28 25
Charged Fog
Fine Coarse Total
72 54 70
49 6 45
61 24 59
46 -24 40
94 -4 88
71 -83 62
56 4 53
55 43 54
54 15 49
57 37 54
69 -75 59
80 -95 69
93 59 88
Meterological Conditions
Wind
Temp. & Direction R.H.
°C (kmph) %
9 N&NW 13 64
12 W 19 60
9 S 5-8 60
23 NE 0-3 45
25 SW 3-8 32
24 SW 8-16 35
26 NNE 8 30
17° W 5-16 64
18 NNE 16-25 48
6 Calm 84
14 Calm 67
15 S 3 60
23 NE 3-8 33
17 Calm 60
19 NE 3-15 63
25 N 5 44
18 Calm 37
17 SSW 8-25 53
17 SW 8 69
14 E 8 63
14 W 5 63
16 E 8 64
15 E 8-15 65
13 W 3 90
24 NE 5 39
25 E 5 31
CFG Settings
Water
Flow Applied
Rate Spray Voltage
(Iph) Patterns (kV)
60 B 4
60 B 0
60 N 4
60 N 0
60 D 0
60 U 4
60 B 0
60 N 4
60 N 0
60 B 4
60 B 4
60 B 0
60 F\ 0
60 N 4
60 N 0
60 N 4
60 N 4
60 N 0
60 N 4
60 B 0
60 B 4
60 B 0
60 B 4
60 B 4
60 B 0
60 B 0
(continued)
-------
TABLE 5 (continued)
Test Date
6/3/81
6/4/81
6/5/81
6/8/81
6/9/81
6/10/81
6/11/81
6/12/81
Uncharged Fog
Fine Coarse Total
-21 22 -15
7 13 8
68 77 69
-10 -212 -19
15 -420 -H
36 33 36
38 -22 35
-11 -5 -10
Charged Fog
Fine Coarse Total
31 00 32
95 25 87
89 -26 76
23 37 24
19 38 19
52 -14 48
50 35 50
65 77 65
79 49 77
55 -120 17
16 -42 13
14 -36 11
47 -54 41
32 -158 21
32 2 30
34 30 33
28 -66 17
42 49 43
Meterological Conditions
Wind
Temp. & Direction R.H.
°C (kmph) %
24 N 16 45
20 N 32 72
7 W 3 95
10 SW 8 89
22 S 3 44
23 S 3 32
24 Variable 1-2 30
27 Variable 1-12 37
28 Calm 34
21 NE 5-8 35
27 NE 8 31
24 N 20 46
16 NE 8 74
21 NW 16 39
22 NW 8-11 37
21 N 19 27
19 N 5 28
17 S 2 60
26 Calm 25
11 Calm 73
15 N 3 60
20 NW 16 61
28 Calm 27
11 Calm 73
26 E 3 48
28 E 3 43
CPC. Settings
Water
Flow Applied
Rate Spray Voltage
(Iph) Patterns (kV)
30 D 0
30 B -7
30 B -7
30 B -7
30 B 0
30 B -10
30 B -10
30 B -10
30 B -10
30 B 0
30 B -10
30 B -10
30 B 8
30 B 0
30 B 8
30 B 8
30 B 0
30 N 8
30 N 0
60 N 8
60 N 0
60 N -7
60 N -9
60 B -5
60 B 0
60 B +8
ON
VI
(continued)
-------
TABLE 5 (concluded)
Test Date
6/13/81
7/7/81
7/8/81
7/9/81
7/10/81
Uncharged Fog
Fine Coarse Total
22 60 25
44 21 43
17 23 18
Charged Fog
Fine Coarse Total
64 50 60
25 -38 19
27 * 25
62 33 61
58 35 57
22 53 25
21 15 20
31 44 32
37 27 36
21 If 17
44 39 43
24 -74 17
21 22 21
Meterological Conditions
Wind
Temp. & Direction R.H.
°C (kmph) %
13 E 3 86
24 S 5 50
33 N 16 29
33 S 16 29
33 S 16-25 20
33 S 30 21
29 N 13 22
28 N 8 29
15 SW 3 49
30 Calm 23
34 Calm 21
36 Calm 27
33 S 8 33
21 Calm 65
29 Calm 36
33 Calm 29
CrC. Settings
Water
Flow Applied
Rate Spray Voltage
(Iph) Patterns (kV)
60 B 8
30 B 4
30 B 0
30 B 8
30 N 4
30 N 8
60 B -8
60 B -8
60 N 8
60 N 0
60 N 8
60 B 4
60 B 8
60 N 4
60 N 0
60 N 8
00
*B - Broad; *N - Narrow.
-------
50
UJ
u
<
2
UJ
a
u
z
UJ
D
N-»
to-
ll,
UJ
8
UJ
30
20
10
I
Uncharged Fog
Charged (+ & -) Fog
FINE
PARTICLES
ONLY
TOTAL
PARTICLES
Figure 28. Mean inhalable particle control efficiency of all the test
runs (for various CFG settings and field conditions) for
charged fog and uncharged fog.
69
-------
Although the effect is not as strong as for the fine fraction, the mean value of the
particle control efficiency of coarse particles increased modestly when the droplets were
charged. This result is in good agreement with the theoretical predictions. However, the
size range of particles collected in the coarse mode was fairly narrow, and the mass
collected was often an order of magnitude smaller than the fine fraction. Another
problem which may have inhibited coarse particle collection is the particle bounce effect,
by which some of the larger particles pass the upper impaction plate and settle on the
back-up filter with the fine particles. Therefore, this particular experiment could not
demonstrate the full effect of charged fog on these particles. Consequently, most of the
ensuing discussion will concentrate on the sum of fine and coarse fractions of the
particles collected, and the subscript on E will be dropped hereafter.
For further detailed analysis, the data base was divided into four groups on the basis
of water flow rate (60 1/h or 30 1/h) and the spray pattern (broad or narrow). Each group
contained particle control efficiencies (percentages) measured for two or more applied
voltages, and existing meteorological conditions. Data in each one of these four groups
were examined to determine the dominant variable parameters affecting the value of the
particle control efficiency, while other variable parameters are constant or nearly
identical. Findings of these analyses are presented below.
Spray Pattern and Relative Humidity
Figure 29 shows particle control efficiency, E (for all particles — fine and coarse),
plotted as a function of ambient relative humidity for two sets of instrument settings for
an applied high voltage of 4 kV (positive charges) and a water flow rate of 60 1/h. The
circles represent a broad spray pattern and squares represent a narrow spray pattern. The
method of least squares was used to fit a straight line to the data sets, shown in the
figure, yielding a correlation coefficient of 0.93 for the broad spray and -0.19 for the
narrow spray. The corresponding slopes are 0.99 and -0.09, respectively. Although the
wind conditions were not identical for the data points, it can be seen that the particle
control efficiency increases with increases in ambient relative humidity for a broad spray,
while it is fairly independent of RH for a narrow spray.
Figure 30 compares the total particle control efficiencies for broad spray (A, B, and
C) and the corresponding values for narrow spray (A1, B', and C1) with identical or nearly
identical field conditions and an applied voltage of 4 kV at a water flow rate of 60 1/h.
The broad spray pattern provides higher particle control efficiencies.
The difference in the dependence of E on RH for the broad spray and narrow spray
can be explained as follows. For a narrow spray, most of the droplets occupy a volume
away from the open side of the hopper, and when fog is continuously applied, this area
becomes more saturated with water droplets and water vapor than outside the hopper or
near the hopper opening, if the wind is not too strong. Thus, there is less droplet
evaporation and, consequently, fairly steady particle collection efficiency. However, in
the case of a broad spray, the droplets are distributed from the rear wall of the hopper to
outside the open side of the hopper. Thus, if the ambient relative humidity is high, a
smaller number of droplets will evaporate leaving more droplets to collect dust particles;
on the other hand, when the ambient relative humidity is low, more droplets are lost due
to evaporation near the open side of the hopper and outside the hopper.
The increase in particle control efficiency with increased relative humidity may
appear to contradict the theoretical predictions shown in Figure 3. In practice, however,
70
-------
95
ILl
U
<
1
C*
UJ
CL
u
z
UJ 7*
D 75
E
u.
UJ
65
Z
o
U
UJ
g55
U
UJ
Z
UJ
d
35
CFG Settings
Water flow:
Voltage:
Spray Type
Broad O
Narrow D
I
601ph
Slope: 0.995
Corr. Coeff.: 0.927
Slope: -0.087
Corr. Coeff.: -0.189
_L
I
I
I
35
*5 55 65 75 85
AMBIENT RELATIVE HUMIDITY (PERCENTAGE)
95
Figure 29. Particle control efficiency of the CFG plotted as a function of
ambient relative humidity for broad (O) and narrow (d) spray
patterns.
71
-------
so
1 1 T
UJ
O
UJ
U
UJ
0-
U
z
UJ
C
C
u-
UJ
o
o
U
UJ
d
B
60
B
50
30
30
W 50 60 70
AMBIENT RELATIVE HUMIDITY (PERCENTAGE)
80
Figure 30. Comparison of total particle control efficiency of the CFG for a
broad spray (A, B, and C) and narrow spray (A1, B', and C') under
identical or nearly identical conditions.
72
-------
the effect of the longer droplet lifetime in a higher relative humidity atmosphere
increases the particle control efficiency. Our test observations are thus consistent with
the conclusion that the droplets should be small enough to provide high particle collection
efficiency, yet large enough not to evaporate too quickly.
Particle Control Efficiency and Applied Voltage
Hoenig (1977) and Hassler (1978) have shown that most mineral particles, especially
those in the fine fraction, are negatively charged. It was observed, however, that the
particle control efficiency of charged fog, when applied on bentonite ore, was higher when
the fog was charged negatively than when the fog was charged positively, with all other
parameters being identical. Figure 31 shows particle control efficiencies for negatively
charged fog (stripes tilted to the left) and positively charged fog (stripes tilted to the
right) for the fine fraction, and for both fine and coarse particles, with the same water
flow rate, same applied voltage, and nearly identical field conditions.
Table 6 compares the effect of applied high voltages on the observed particle
control efficiency, with all other variable parameters identical or nearly identical. It is
evident that a higher applied voltage results in a higher value of E. As we have shown in
the last section, an increase in the applied voltage generated droplets with more charges
(and charge-to-mass ratio). Consequently, the electrostatic forces of attraction between
the droplet and particles were increased, resulting in higher values of E.
Table 6 also shows the particle control efficiency when two negative voltages are
applied to the inflowing water. These data show only a slight increase in particle
collection efficiency when the applied voltage was increased from -7 kV to -9 kV. This
result may be because, with either of the voltages applied, the droplets reach their
Rayleigh limit of charges and they begin to evaporate. Therefore, increased high voltage
above about 8 kV does not significantly increase the particle collection. It should also be
recalled that the absolute value of the particle control efficiency in this particular case is
low because of the effect of other variable parameters, such as water flow rate and
unfavorable meteorological conditions.
Effect of Water Flow Rate on Particle Control Efficiency
Figure 32 is a good example of the effect of water flow rate in the CFG on its
ability to control particles. The percentage particle control efficiency in this case
decreased by *2% when the water flow rate was reduced from 60 1/h to 30 1/h, with all
other instrument settings and field conditions nearly identical. This effect is expected as
increased water flow rate increases the number of charged droplets available for particle
collection. Water flow rate was particularly significant in our experiment because the
CFG was being applied to control dust in a volume larger than its maximum coverage;
therefore, a decrease in the water flow rate decreased its particle control efficiency.
Effect of Wind Conditions on Particle Collection
Under a controlled experimental setup one would expect the particle control
efficiency of charged droplets to decrease if the wind speed in the volume being treated
were increased. The increased wind would reduce the time available for the droplet and
particle to interact. However, in our experimental situation, because of the location and
method of particle sampling, this effect cannot be quantified.
73
-------
50
UJ ZtQ
O
H
Z
UJ
O
UJ
a*
\ 3°
UJ
y
a.
UJ
o
g 20
Z
O
u
UJ
-4
y
P
K;
CL
10
0 Charged (-) Fog
^ Charged (+) Fog
-
1
i
I
i
FINE
PARTICLES
ONLY
TOTAL
PARTICLES
Figure 31. Comparison of particle control efficiency of the CFG for positively
charged fog and negatively charged fog, with all other parameters
nearly identical (Run Numbers 69 and 96).
-------
TABLE 6. COMPARISON OF TOTAL PARTICLE CON-
TROL EFFICIENCIES OF CFG FOR THE
TWO PAIRS OF APPLIED VOLTAGES, WITH
ALL OTHER PARAMETERS NEARLY IDEN-
TICAL.
Applied
High
Voltage
4 kV
8 kV
-7 kV
-9 kV
CFG Particle
Control
Efficiency (%)
17
21
30
31
75
-------
50
U
UJ
U
OS
UJ
3°
z
UJ
U
E
u.
UJ
20
Z
O
u
UJ
J
U
P
o:
£
10
60 1/h 30 1/h
Figure 32. Comparison of particle control efficiency of the CFG for two water
flow rates, with all other parameters nearly identical.
76
-------
Because the hopper is completely enclosed except on the side through which the
bentonite ore is dumped and because the particle sampler inlet is located near the
southeast corner of the hopper, slow winds from the southeast, east, and northeast will
not significantly affect the particle concentration measured. If the wind is from the
north and northwest, more dust will be blown toward the sampler inlet, and the sampler
may record a higher concentration than when there are no such winds. This wind
condition enables the dust to remain in the hopper area longer, giving more time for the
droplets to interact with the particles.
CONCLUSIONS OF THE CFG FIELD TESTS ON BENTONITE
As was indicated at the beginning of this chapter, the experiment to evaluate the
particle control efficiency of the CFG and to determine the optimum instrument settings
and meteorological conditions ideally would have been done under controlled cond-
itions — or a much larger data base obtained. Nevertheless, the following conclusions can
be drawn from the available data.
1. Charged fog was shown to be an effective means of controlling fugitive dust
emissions. The mean value of the particle control efficiency of charged fog
measured under all instrument settings and field conditions showed a 78% increase
when compared with the corresponding value for uncharged fog.
2. The relative humidity (in this particular experimental setup) seemed to play a
significant role in determining the overall particle collection efficiency. It appears
that the lifetime of the droplet is the dominant factor in determining what the
particle control efficiency will be. Therefore, the droplets should be large enough
not to evaporate too quickly, yet small enough to yield a high particle control
efficiency.
3. Under identical or nearly identical field conditions and instrument settings,
negatively charged droplets gave higher values of particle control efficiency than
did positively charged fog, suggesting that inhalable bentonite particles carry a net
excess positive charge. •
4. Measured inhalable particle control efficiencies were higher for higher applied
voltages in the * to 10 kV range. At the upper end of this range, the particle control
efficiency seemed to attain a saturation value.
5. Measured inhalable particle control efficiency was higher when charged droplets
could cover more of the dust-laden air in the hopper. In the experimental setup
used, higher water flow rates and broad spray patterns resulted generally in higher
collection efficiencies, although the key element appeared to be how many particles
were treated by the droplets.
6. Because of the type of particle sampling method used and the field setup, the effect
of wind speed and direction on particle control efficiency cannot be quantified with
the available data.
7. The optimum CFG instrumental settings are found to be 60 1/h water flow rate, a
spray pattern which will cover a maximum volume of dust-laden air (broad or
narrow spray depending on the extent of the source), an applied voltage of 8-10 kV,
and positive or negative charge depending on the charges of the dust particles. Ideal
field conditions are high relative humidity (to ensure long droplet lifetime), and
calm or low wind conditions.
77
-------
REFERENCES
Beard, K.V. (1974): Experimental and numerical collision efficiencies for submicron
particles scavenged by small raindrops. J. of Atmos. Sci. 31, 1595-1603.
Beard, K.V., and S.N. Grover (1974): Numerical collision efficiencies for small raindrops
colliding with micron size particles. J. of Atmos. Sci. 31, 543-550.
Brookman, E.T., R.C. McCrillis, and D.C. Drehmel (1981): Demonstration of the use of
charged fog in controlling fugitive dust from large-scale industrial sources.
Presented at the Third Symposium on the Transfer and Utilization of Particle
Control Technology, Orlando, Florida, March 9-12.
Brookman, E.T., and K.J. Kelly (1982): Demonstration of two electrostatic fog devices on
fugitive dust sources within Armco Plants. Draft Final Report submitted to Armco
Inc., TRC Environmental Consultants Inc., TRC Project No. 1725-L52, East
Hartford, Connecticut.
Calvert, S., J. Goldschmid, D. Leith, and D. Mehta (1972): Wet scrubber system study;
Vol. 1, Scrubber handbook. EPA-12-72-118a, Ambient Purification Technology, Inc.,
Riverside, California.
Carlton, J.B., and L.F. Bouse (1980): Electrostatic spinner-nozzle for charging aerial
sprays. Transactions of the ASAE 23, 1369-1378.
Cheng, L. (1973): Collection of airborne dust by water sprays. Ind. Eng. Chem. Process
Des. Develop. 12, 221-225.
Courtney, W.G., and L. Cheng (1976): Control of respirable dust by improved water
sprays. Bureau of Mines Information Circular 1977, 1C 8753, U.S. Department of the
Interior, 92-106.
Cowherd, C. (1980): The technical basis for a size-specific particulate standard. 3APCA
30, 971-982.
Daugherty, D.P., and D.W. Coy (1979): Assessment of the use of fugitive emission control
devices. EPA-600/7-79-045, Research Triangle Institute, Research Triangle Park,
North Carolina.
Drees, W. (1966): Investigating the use of surface-active substances in dust control by
water. Staub-Reinhalt Luft. 26, 31.
Drehmel, D.C. (1977): Advanced electrostatic collection concepts. 3APCA 27, 1090-
1092.
Drehmel, D.C. (1981): Personal communication.
78
-------
Emmerling, 3.E., and R.J. Seibel (1975): Dust suppression with water sprays during
continuous coal mining operations. Bureau of Mines Report on Investigations 1975,
RI 8064, U.S. Department of the Interior, 1-12.
Friedlander, S.K. (1977): Smoke, Dust, and Haze; Fundamentals of Aerosol Behavior.
Wiley-Interscience, New York, 317.
Frost, W., F.G. Collins, and D. Koepf (1981): Charged particle concepts for fog
dispersion. NASA Contract Report 3440, Marshall Space Flight Center, Huntsville,
Alabama.
George, H.F., and G.W. Poehlein (1974): Capture of aerosol particles by spherical
collectors: electrostatic, inertial, interceptional, and viscous effects. Envir. Sci.
and Tech. 8,
Gillespie, T. (1955): The role of electric forces in the filtration of aerosols by fiber
filters. J. Colloid & Interface Sci. H), 299.
Goldshmid, Y., and S. Calvert (1963): Small particle collection by supported liquid drops.
AIChE. 3. 9. 352-358.
Greenfield, S. (1957): Rain scavenging of radioactive particulate matter from the
atmosphere. J. of Met. 14, 115-125.
Grover, S.N., H.R. Pruppacher, and A.E. Hamielec (1977): A numerical determination of
the efficiency with which spherical aerosol particles collide with spherical water
drops due to inertial impaction and phoretic and electrical forces. 3. of Atmos.
Sci. 34, 1655-1663.
Grover, S.N., and K.V. Beard (1977): A numerical determination of the efficiencies with
which electrically charged cloud drops and small raindrops collide with electrically
charged spherical particles of various densities. J. of Atmos. Sci. 32, 2156-2163.
Gunn, R. (1955): The statistical electrification of aerosols by ionic diffusion. 3. Coll. &
Int. Sci. 10. 107-119.
Hassler, H.E.B. (1978): A new method for dust separation using autogenous electrically
charged fog. Journal of Power and Bulk Solids Technology, Spring.
Hesketh, H.E. (1974): Fine particle collection efficiency related to pressure drop,
scrubbant and particle properties and contact mechanisms. 3APCA 2_4, 939-942.
Hoenig, S.A. (1977): Use of electrostatically charged fog for control of fugitive dust
emissions. EPA-600/7-77-131, Available from NTIS, Springfield, VA 22161.
Hoenig, S.A. (1979): Fugitive and fine particle control using electrostatically charged fog.
EPA-600/7-79-078.
Hoenig, S.A. (1980): The control of dust using charged water fogs. EPA-600/9-80-039d,
201-216.
79
-------
Kearns, M.T., and D.L. Harmon (1979): Demonstration of a high field electrostatically -
enhanced Venturi scrubber on a magnesium furnace fume emission. Paniculate
Control Devices, Vol. Ill, EPA-600/69-80-039C, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 27711.
Kinsey, 3.S., C.E. Lyons, S.A. Hoenig, and D.C. Drehmel (1979): New concepts for the
control of urban inhalable particulate by the use of charged fog. Paper presented at
the 73rd Annual Meeting of the Air Pollution Control Association, Montreal,
Canada, 3une 22-27.
Kramer, H.F., and H.F. Johnstone (1955): Collection of aerosol particles in the presence
of electrostatic fields. Ind. Eng. Chem. 47,
Kunkel, W.B. (1950a): The static electrification of dust particles on dispersion into a
cloud. 3. Appl. Phys. 21, 820-832.
Kunkel, W.B. (1950b): Charge distribution in coarse aerosols as a function of time. 3.
Appl. Phys. 21, 833-837.
Law, S.E. (1978): Embedded-electrode electrostatic -induction spray -charging nozzle:
Theoretical and engineering design. Transactions of the ASAE 21, 1097-1104.
Lear, C.W. (1976): Charged droplet scrubber for fine particle control: Laboratory study.
TRW Systems Group, Redondo Beach, California. Available from NTIS, PB-258823.
Leong, K.H., 3.3. Stukel, and P.K. Hopke (1981): Effect of surface properties of
collectors on the removal of charged and uncharged particles from aerosol suspen-
sions. U.S. Environmental Protection Agency, ORD Publication No. 81-2. IERL,
Cincinnati, Ohio
Leong, K.H., 3.3. Stukel, and P.K. Hopke (1982): Limits in charged-particle collection by
charged droplets. Env. Sci. & Tech. 16, 386-387.
Mathai, C.V. (1983a): Charged fog technology. Part I: Theoretical background and
instrumentation development. 3. of Air Pollut. Contr. Assoc. 33, 664-669.
Mathai, C.V. (1983b): Charged fog technology. Part II: Prototype tests of a new charged
fog generator for fugitive emission control. 3. of Air Pollut. Contr. Assoc. 33,
756-759. ~
Mathai, C.V. (1983c): Preliminary evaluations of WRAQS aerosol measurements. Report
Prepared for the Electric Power Research Institute by AeroVironment Inc.,
Pasadena, CA.
Mathai, C.V., D.V. Pankratz, and D.M. Wilbur (1983): Acid fog and filter artifact
formation studies in Kern County, 1983. Final Report Prepared for the Western Oil
and Gas Association by AeroVironment Inc., Pasadena, CA.
Mazumder, M.K., R.E. Ware, and W.G. Hood (1982): Simultaneous measurements of
aerodynamic diameter and electrostatic charge on a single particle basis. Measure
ment of Suspended Particles by Quasi-Elastic Light Scattering. 3ohn Wiley and
Sons, Inc., 328-341.
80
-------
McCoy, 3., T. Melchar, 3. Valentine, D. Monaghan, T. Muldoon, and 3. Kelly (1983):
Evaluation of charged water sprays for dust control. Final Report by Foster-Miller
Inc., Waltham, MA for the U.S. Bureau of Mines, Contract H0212012.
McCully, C.R., M. Fisher, G. Langer, 3. Rosinski, H. Glaess, and E. Werle (1956):
Scavenging action of rain on air-borne particulate matter. Ind. Engrg. Chem. 48,
1512-1516.
McDonald, 3.E. (1963): Rain washout of partially wettable insoluble particles. 3.
Geophys. Res. 68, 4993-5003.
Mueller, P.K., and G.M. Hidy (1983): The Sulfate Regional Experiment: Report of
findings Vol. 1. E A-1901. Final Report Prepared for Electric Power Research
Institute, by Environmental Research and Technology, Westlake Village, CA.
National Research Council (1979): Controlling Airborne Particles. National Academy of
Sciences, Washington, D.C.
Natusch, D.R.S. (1974): The chemical composition of fly ash. Proc. of Symposium on
Control of Fine-Particulate Emission from Industrial Sources, U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711.
Nielsen, K.A. and 3.C. Hill (1976a): Collection of inertialess particles on spheres with
electrical forces. Indus. Eng. Chem. Fund. 15, 143-157.
Nielsen, K.A. and 3.C. Hill (1976b): Capture of particles on spheres by inertial and
electrical forces. Indus. Eng. Chem. Fund. 15, 157-163.
Pemberton, C.S. (1960): Scavenging action of rain on non-wettable particulate matter
suspended in the atmosphere. Int. 3. Air Pollut. 3, 168-178.
Pilat, M.3., S.A. 3aasund, and L.E. Sparks (1974): Collection of aerosol particles by
electrostatic droplet spray scrubbers. Envir. Sci. and Tech. 8, 360-362.
Pilat, M.3. (1975): Collection of aerosol particles by electrostatic droplet spray
scrubbers. 3APCA 25, 176-178.
Pilat, M., and D.F. Meyer (1976): University of Washington electrostatic spray scrubber
evaluation. EPA-600/2-76-100.
Prem, A., and M. 3. Pilat (1978): Calculated particle collection efficiencies by single
droplets considering inertial impaction, Brownian diffusion, and electrostatics.
Atmos. Envir. 12, 1981-1990.
Proceedings of the APCA Specialty Conference on the Technical Basis for a Size-Specific
Particulate Standard - Parts I and II (1980): March and April.
81
-------
Rabel, G., H. Neuhaus, and K. Vettebrodt (1965): The wetting of dusts and fine ores for
the purpose of reducing dust formation. Staub-Reinhalt Luft. 25, 4-8.
Renninger, R.G., M.K. Mazumder, and M.K. Testerman (1981): Particle sizing by
electrical SPART analyzer. Rev, of Sci. Instruments 52 (2), 242-246.
Schutz, A. (1967): The electrical charging of aerosols. Staub-Reinhalt. Luft. 27, 24-32.
Sierra Instruments (1981): Product literature for model 230CP Cyclone preseparator.
Sierra Instruments Inc., Carmel Valley, California.
Slinn, W.G., and 3.M. Hales (1971): A reevaluation of the role of thermophoreses as a
mechanism of in and below cloud scavenging. 3. of Atmos. Sci. 28, 1465-1471.
Stulov, L.D., F. I. Murashkevich, and N. Fuchs (1978): The efficiency of collision of solid
aerosol particles with water surface. 3. Aerosol Sci. 9, 1-6.
Suck, S.H., J.L. Kassner, Jr., R.E. Thurman, P.C. Yue, and R.A. Anderson (1981):
Theoretical predictions of ion clusters relevant to the atmosphere: size and
mobility. 3. Atmos. Sci. 38, 1272-1278.
Takahashi, T. (1973): Measurement of electric charge of cloud droplets, drizzle and
raindrops. Review of Geophysics and Space Physics 11, 903-924.
Thakur, P. (1980): Personal Communication.
Walkenhorst, W. (1971): Charge measurements of dust particles. Staub-Reinhalt. Luft.
3J_, 8-16.
Walton, W.H., and A. Woolcock (1960): The suppression of airborne dust by water spray.
Aerodynamic Capture of Particles, 129-153, Pergamon Press, New York.
Wang, P.K., and H.R. Pruppacher (1977): An experimental determination of the efficiency
with which aerosol particles are collected by water drops in subsaturated air. 3. of
Atmos. Sci. 34, 1664-1696.
Wang, P.K., S.N. Grover, and H.R. Pruppacher (1978): On the effect of eJectric charges
on the scavenging of aerosol particles by clouds and small raindrops. 3. of Atmos.
ScL. 35, 1735-1743.
Wang, P.K., and H.R. Pruppacher (1980): The effect of an external electric field on the
scavenging of aerosol particles by cloud drops and small rain drops. 3. of Colloid
and Interface Science 75, 286-297.
Whitby, K.T., and B.Y.H. Liu (1966): The electrical behaviour of aerosols. Aerosol
Science, Edited by C.N. Davies, Academic Press, New York.
82
-------
White, H.3. (1963): Industrial electrostatic precipitation. Addison-Wesley, Reading,
Massachusetts.
Woffinden, G.3., G. R. Markowski, and D.S. Ensor (1978): Effects of interfacial properties
on collection of fine particles by wet scrubbers. EPA-600/17-78097, NTIS
PB-284073, U.S. Environmental Protection Agency, Research Triangle Park, N.C.,
65 pp.
Yung, S.C., S. Calvert, and D.C. Drehmel (1979): Fugitive dust control using charged
water spray. Paper presented at the 72nd Annual Meeting of the Air Pollution
Control Association, Cincinnati, Ohio, June 24-29.
Yung, S.C., S. Calvert and D.C. Drehmel (1980): Spray charging and trapping scrubber for
fugitive particle emission control. JAPCA 30, 1208-1211.
83
-------
APPENDIX A
Charged Fog Technology. Part I: Theoretical
Background and Instrumentation Development
84
-------
Reprinted from APCA JOURNAL, Vol. 33, No. 7, July 1983
(with permission)
Charged Fog Technology
Part I: Theoretical Background
and Instrumentation Development
C. V. Mathai
AeroVironment Inc.
Pasadena, California
Water sprays, UM most common method of controlling dust In
mines and from fugitive emission sources, do not control Inhalable
particles very effectively. Since most Industrial pollutants and natu-
rally occurring fugitive dust particles acquire electric charges as they
are dispersed into the air, the inhalable particle control efficiency of
water sprays can be significantly Improved If the water droplets are
also electrically charged to the opposite polarity. Commercially
available charged fog devices have several disadvantages. They
require high pressure air and/or water, their nozzles are prone to
dogging, and they generate charged droplets which are poorly suited
for good inhalable particle control. This paper reviews the basic
principles of charged fog technology and describes a new charged
fog generator which overcomes the problems of commercial charged
fog devices.
In the charged fog generator (CFG), water from a reservoir Is In-
troduced Into a rotating cup where the water forms a thin layer due
to centrifugal forces. As the water moves towards the lip of the cup,
high speed air from an axial fan strikes the thin water film, breaks It
into fine droplets and projects the droplets forward. The droplets thus
generated have a typical mass median diameter of about 200 fun and
a concentration median diameter of about 100 /tm. The droplets are
electrically charged by contact charging the Inflowing water, providing
a typical charge to mass ratio of 1.2 X 10~* C/g with an applied
voltage of about 15 kV. The water flow rate in the CFG can be varied
from 4-70 Uk and the spray pattern can be easily adjusted to conform
to the size and shape of the dust source to be treated. Other advan-
tages of this device are that It uses only about 1 kW power, H Is por-
table, and tt Is easily adaptable for use at remote areas where com-
mercial electrical power supply Is not available.
Recent reports1-3 indicate that inhalable particles (particles
smaller than 15 Mm in aerodynamic diameter), in general, and
fine particles (particles smaller than 2-3 jim in aerodynamic
diameter), in particular, are a human health hazard, con-
tribute to the formation of acid rain, and degrade atmospheric
visibility. Although significant progress has been made in
Copyright 198.1—Air Pollution Control Auociation
controlling emissions from conventional industrial stack
sources, effective and economically feasible methods are
lacking for controlling inhalable particles from nonstack
sources (fugitive emissions). Recognizing this situation, EPA
has been encouraging the development of new methods, such
as charged fog technology, to control fugitive emissions.
Removing fine particles from a gas stream in an open area
is difficult because of the particles' low mobility and unfa-
vorable inertial properties, and because of uncontrollable
external factors (meteorological parameters). The most
commonly used dust control method (ordinary water sprays)
in mining and other material handling areas is only 30-40%
efficient in controlling inhalable particles.4 Only during the
last few years have electrostatics been used to augment par-
ticle collection efficiency of water droplets.5-10 Numerous
studies8-18 have shown that most industrial pollutants and
naturally occurring dust particles acquire electric charges as
they are dispersed into the air. Walkenhorst,11 Schutz,12
Hoenig,8-9 and Hassler,10 have also shown that the polarity and
magnitude of the charges on these particles depend upon their
size and origin (coal, soil, mineral, etc.). Therefore, particle
collection efficiency of water droplets can be significantly
enhanced via electrostatic forces of attraction if the droplets
are charged to the opposite polarity.
In this paper the theoretical concepts of charged fog tech-
nology are outlined and currently available charged fog devices
are discussed, with special emphasis on a prototype charged
fog generator developed under the sponsorship of the Indus-
trial Environmental Research Laboratory of the EPA. Tests
of this prototype device for fugitive emission control are de-
scribed in a subsequent paper.19
Theoretical Background
The collection of an aerosol particle by a charged droplet
is the result of a number of simultaneous mechanisms of in-
teraction between them, such as inertial impaction, direct
interception, Brownian diffusion, and electrostatic, dif-
fusiophoretic and thermophoretic forces.20-24 When an aerosol
particle approaches a water droplet with a relative velocity,
it may directly collide with the droplet (impaction), barely
touch the droplet (interception), or entirely miss the droplet.
The relative effect of the mechanisms of interaction between
the droplet and the particle depends upon the size of the
particle. For large particles (aerodynamic diameter greater
than 2-3 nm), the dominant mechanisms of particle collection
by droplets are impaction and interception. For particles
664
85
Journal of the Air Pollution Control Association
-------
0.01
0.1 1.0
Particle radius.
10.0
100.0
Figure 1. Calculated single droplet collection efficiency in air
of 10°C and 900 mb as a function of particle radius at a relative
humidity of 75 % with droplet radius as a varying parameter (A)
42 Mm. (B) 72 Mm, (C) 106 Mm, (D) 173 Mm, and (E) 310 Mm.
smaller than 0.1 ^m, Brownian diffusion becomes very im-
portant, and for particles between these two ranges, electro-
static forces are the dominant interaction mechanism.
The particle collection efficiency of uncharged water sprays
(where inertia! impaction is the major collection mechanism)
is given by,25
(1)
where QL and Qc are the volume flow rates of the water and
air component of the dust cloud, respectively; L is a charac-
teristic length for the total capture process; D is the mean
droplet diameter; and i\ is the single droplet collection effi-
ciency (assumed to be identical to collision efficiency).
For a given particle size, single droplet collection efficiency
due to inertia! impaction is proportional to the relative ve-
locity between the droplet and the particle and inversely
0.01
1.0 1.0
Particle radius.
10.0
100.0
Figure 2. Calculated single droplet collection efficiency in air
of 10°C and 900 mb as a function of particle radius for a 72 pm
radius droplet with relative humidity as a varying parameter (A)
50 % .(6)75%. (C) 95 % . and (D) 1 00 % .
July 1983 Volume 33, No. 7
proportional to the droplet diameter. The single droplet col-
lection efficiency of a charged particle by an oppositely
charged droplet is given by20
TJ = -Cqcqp/6ir2KrR2nU0
(2)
where C is the Cunningham slip correction factor;
-------
nology is based. Figure 3 also shows that the addition of
electric charges does not significantly affect the collection
efficiency for large particles.
Prem and Pilat211 showed that for 200-^m and 50-jim
charged droplets and oppositely charged particles in the size
range 10-20 nm, the particle collection efficiency due to
electrostatic forces and inertia are comparable. However, for
particles smaller than 10 jim, and for the same droplet sizes,
the collection efficiencies due to the electrostatic forces pro-
gressively become the dominant collection mechanism
(compared to inertial impact ion) as the particle sizes are de-
creased.
When sprayed into the air, the charged droplets will evap-
orate unless the air is saturated with water vapor. The droplet
lifetime determines the effective contact time between the
droplet and particles and thus strongly influences the overall
particle control efficiency of a charged fog device. The lifetime
of a water droplet depends upon the temperature and relative
humidity of the medium into which it is introduced. To obtain
the best collection efficiency, the droplets must be small
enough to provide both an adequate spray rate per volume of
gas treated and sufficient contact time, yet large enough to not
evaporate too quickly.
The maximum electrical charge that can be carried by a
particle is limited by the physical properties of the particle.
Three such limits, the electron limit, the ion limit, and the
Rayleigh limit are discussed by Cohen.27 The maximum
charge a droplet can carry before it disintegrates is reached
when the outward pressure produced by the electric field at
the surface of the drop is equal to the inward pressure pro-
duced by the surface tension. This limiting charge (Rayleigh
limit) is given by28-29
QRay = 8?T(f0(rfl3)l/2
(3)
where Quay is the limiting charge on the droplet, e0 is the
permittivity of the medium in which the droplet is located, a
is the surface tension of the liquid, and R is the droplet ra-
dius.
Although Figure 3 shows that the addition of electric
charges on particles and droplets completely eliminates the
minimum in particle collection efficiency (around r = 1 ;um),
and shows single particle collection efficiencies 5-10 times
higher than for uncharged sprays, the overall collection effi-
ciency of the system may not be that high. In laboratory ex-
periments, investigators6 have reported increases in the col-
lection of efficiencies of charged droplets over uncharged
droplets of about 15% for 1-^m particles to over 45% for 0.3-
(tm particles. Hoenig8-9 reported particle control efficiencies
of 50-80% with charged droplets under controlled experi-
mental conditions. Brookman et al.'M have reported inhalable
particle control efficiencies of about 60% from measurements
at a sand and gravel operation using the largest commercially
available charged fog device.
Charged Fog Devices
Charged droplet devices are used in fugitive emission con-
trol and in agricultural spraying.'11"12 Here we will restrict our
discussion to those devices used in fugitive emission con-
trol.
Water droplets can be generated using a pressure nozzle or
a rotating cup. Droplets may be charged by electrostatic in-
duction charging, ionized field charging, or contact charging
methods. Hassler10 showed that droplets may be charged by
the water to metal frictional forces inside the nozzle during
atomization. This method requires no high voltage supply;
however, it requires very pure deionized water. This limitation
precludes using this method of charging water droplets for
dust control in the field.
In the most frequently used method, electrostatic induction
charging, a high voltage potential is maintained between the
water spray nozzle and the induction ring. Positive or negative
charges are induced on the droplets by the charges on the in-
duction ring, depending on the polarity of the high voltage
applied to the ring. In ionized field charging, an electrode at
sufficiently high DC potential is placed near the water nozzle,
causing dielectric breakdown of the air immediately sur-.
rounding the electrode. Water droplets traveling through this
ionized field can acquire electric charges by ion attach-
ment.
In contact charging, charges are transferred by conduction
to the water, and subsequently, to the droplets at their instant
of formation by connecting a source of charge (high voltage
supply) to the inflowing water. This method has been shown
to be very effective in charging the droplets to a high degree.33
However, this method requires that the entire water supply
and associated plumbing be electrically isolated to avoid
current leakage.
Charged fog devices are commercially available from Key-
stone Dynamics Inc., Villanova, PA ("Dustron"), and Sonic
Air Fan
DC Pow«r Supply
Nonconductlve
Spinning Cup
Nonconductlve
Water Tub*
Nonconductlv*
Wat«r Lin*
Nonconductlve Air Con*
Water-Deflecting
Baffl*
•MSkV
DC Water Pump.
Figure 4. Schematic diagram of the charged fog generator.
87
Journal of the Air Pollution Control Association
-------
Figure 5. Typical long and narrow spray pattern from the charged fog generator.
¥
Development Corp., Mawah, NJ ("Foggers I, II and IV"). The
charged fog generator developed by AeroVironment Inc.
currently exists in prototype form. The Dustron and Foggers
operate on the same principle—atomizing water in a nozzle
with high pressure air and/or water and passing the droplets
through an induction charging ring. On the other hand, the
charged fog generator uses rotary atomization and direct
contact charging of the inflowing water.
In the Fogger IV30 (the largest commercially available
charged fog device), water at a variable flow rate of 0-151 L/h
is atomized as it is ejected from a nozzle by a compressed
(5.6-8.8 kg/cm2) air supply. The droplets pass through an
induction ring maintained at 12.5 kV, where the droplets ac-
quire electric charges by induction. A flow of air around the
nozzle provided by a 79 m3/min centrifugal fan projects the
fog towards the dust source. The droplets generated by this
device are estimated to have an average diameter of 60 jim and
a charge to mass ratio of 0.11 X 10~€ C/g at a water flow rate
of 75 L/h. The Fogger IV requires 230 V (AC) and a maximum
power of 8 kW. Foggers I and II are similar to Fogger IV, ex-
cept they do not use fans to blow the fog forward toward the
dust source and their water handling capabilities are lower:
0-60 L/h for Fogger I and 0-120 L/h for Fogger II.
Charged fog devices using pressure nozzles require a sub-
stantial supply of high pressure water and/or air for proper
atomization and have a tendency to clog if the water supply
contains high concentrations of dissolved salts and suspended
solids. As indicated earlier, the induction ring method of
charging the droplets provides a lower droplet charge to mass
ratio than is needed for a high degree of inhalable particle
control. Most of the problems associated with pressure nozzles
can be eliminated by using rotary atomizers. Rotary atomizers
using either a spinning disc or cup have been used for many
years for generating fine droplets.34-35 The charged fog gen-
erator described below uses a spinning cup.
Charged Fog Generator
Figure 4 is a schematic representation of the charged fog
generator (CFG). Water is introduced through the water tube
into the 3,600 rpm rotating cup, whose inside is configured
with a gradual linear taper. A small deflecting baffle is at-
tached to the open end of the water tube so that the water will
be deflected 90° and strike the inner rear surface of the ro-
tating cup. Because of the centrifugal forces, the water spreads
into a thin film and moves forward into a high velocity air-
stream generated by the axial fan. The impact of the high
velocity air on the thin water film breaks the water film into
fine water droplets.
The water tube, air cone, and the rotating cup are made of
nonconducting materials. The water tube is firmly attached
to the rotating cup, thus rotating with the cup. The other end
of the water tube is attached to the water supply through a
rotating seal. The water for atomization is stored in an elec-
trically isolated reservoir (130-L capacity) and a low pressure
pump is used to pump it to the inlet of the rotating seal. The
water flow rate can be varied from about 4-70 L/h. The air-
stream from the fan can be controlled by a butterfly valve
Figure 6. Typical short and broad spray pattern from
the charged fog generator.
setting and this airstream projects the fog forward. By ad-
justing this valve, the airflow speed can be controlled, thereby
controlling the shape of the fog spray pattern to conform to
the shape and size of the source of dust upon which the
charged fog is to be applied (see Figures 5 and 6). The spray
pattern covers a volume of 16-24 m3.
To achieve the preferred high charge to mass ratio for the
droplets, contact charging by directly connecting a high voltage
source to the inflowing water was found to be the most effi-
cient method. However, this method requires that the entire
water supply (reservoir) and associated tubing be electrically
isolated to prevent current leakage.
July 1983 Volume 33, No. 7
88
667
-------
Droplet Size Distribution
The size distribution of the droplets was determined using
a cloud optical array probe and a precipitation optical array
probe manufactured by Particle Measuring Systems of
Boulder, CO (for droplets in the size range of 30-1875 pm).
These measurements gave a mass median droplet diameter
of about 200 nm (Figure 7) and a concentration median di-
ameter of about 100 Mm (Figure 8). Collecting the droplets on
greased glass slides and observing them under a microscope
also yielded values consistent with the above result.
Droplet Charge to Mass Ratio
The charge to mass ratio of the droplets was determined
using a special sampling train. This train consisted of an in-
sulated stainless steel probe tip mounted on a standard glass
midget impinger. The probe was connected electrically to
10"
n
6
§10'
10"
10°
103
10' 102
Droplet diameter, ion
Figure 7. Droplet mass as a function of droplet diameter
measured using a cloud optical array probe and a precipitation
optical array probe.
copper wool packing placed inside the midget impinger which
was also connected by a shielded cable to an electrometer. The
impinger was immersed in a Dewar flask containing dry ice.
An isokinetic sample of droplets was then extracted from the
fog spray. As the droplets moved through the impinger they
were condensed and frozen, thus transferring their charge to
the copper wool packing. The charge was then measured by
the electrometer. Knowing the mass of droplets collected and
the current produced by them, the charge to mass ratio was
calculated. This method gave a typical value of 1.2 X 10~6 C/g
with an applied voltage of 15 kV. This value is about one
fourth of the maximum allowed droplet charge value (Ray-
leigh limit) for 200 nm droplets.
The charge to mass ratio was also estimated using another
experimental setup. In this case, the droplets were allowed to
transfer their charges to a 125 jim mesh size standard sieve
placed in the path of the fog spray and measuring the current
generated through the mesh. Knowing the mass median
droplet size and the total number of droplets generated per
second by the CFG, the charge-to-mass ratio was estimated.
This method, though less reliable, gave values consistent with
that reported above for the gravimetric method. In a similar
manner, the charge per unit area of the droplets can be cal-
culated using the measured current generated in the mesh and
the surface area of the droplet with the median concentration
diameter. In terms of electrostatic units used in Figure 3, the
charges on the droplets generated by the CFG will fall ap-
proximately midway between cases A and B.
Summary and Conclusions
Since inhalable particles in the air are known to be elec-
trostatically charged, their control by water sprays can be
significantly enhanced if the water droplets are also electro-
statically charged to the opposite polarity. Commercially
available charged fog devices require high pressure air and/or
water and their nozzles are prone to clogging if the water
supply contains dissolved salts and a high concentration of
suspended solids. Moreover, their induction ring method of
droplet charging provides a lower droplet charge to mass ratio
than is needed for a high degree of inhalable particle con-
trol.
The new charged fog generator uses rotary atomization,
eliminating both fine nozzle and high pressure air require-
ments. Contact charging of droplets provides a high charge
to mass ratio for the droplets generated. The CFG's spray
pattern can be easily adjusted to conform to the extent of the
dust source upon which the charged fog is to be applied. The
size distribution of the droplets gives a mass median diameter
of about 200 pm and a concentration median diameter of
about 100 pm. The contact charging method provides a high
charge to mass ratio of 1.2 X 10~6 C/g. The CFG is small,
portable and requires only about 1 kW (ordinary 110 V AC line
voltage). With the use of small inverter the unit can be oper-
ated on battery power. This feature may be important, be-
cause the system has potential applications where commercial
electric power is not readily available.
For certain fugitive dust sources in open areas, charged fog
technology seems to be the most practical and cost-effective
method of dust control, and probably the only method for
controlling fine particles (<2.5 urn in aerodynamic diameter)
efficiently.
10'
\
10-'
10-'
ID'3
10°
10' 10s
Droplets diameter, itm
103
Figure 8. Droplet concentration as a function of droplet di-
ameter measured using a cloud optical array probe and a pre-
cipitation optical array probe.
Further research is required in two areas: 1) quantitative
information is needed on electrical charges (both polarity and
magnitude) carried by aerosols in the size range 0.5 /*m to 10
or 15 Mm; and 2) further improvements are needed in the in-
strumentation, especially to make it explosion proof, before
it can be used to control inhalable particles in gassy under-
ground mines.
Acknowledgments
This work was funded by the Industrial Environmental
Research Laboratory of the U.S. EPA under contract No.
68-02-3145. The author is grateful to EPA Project Officers Dr.
89
Journal of the Air Pollution Control Association
-------
D. C. Drehmel and Mr. W. B. Kuykendal for their technical
guidance during this research project. I am indebted to Mr.
J. S. Kinsey and Mr. L. A. Rathbun for their technical assis-
tance. I am also thankful to Dr. Ivar Tombach, Dr. R. Nininger
and Ms. Diane Miller (all of AeroVironment) for their tech-
nical and editorial review of this article before submission to
JAPCA.
References
1. C. Cowherd, "The technical basis for a size-specific particulate
standard," JAPCA 30:971 (1980).
2. S. K. Friedlander, Smoke, Dust, and Haze: Fundamentals of
Aerosol Behavior, Wiley-Interscience, New York, 1977.
.1. D. R. S. Natusch, "The Chemical Composition of Fly Ash," in
Proceedings of Symposium on Control of Fine Particulate
Emission from Industrial Sources, U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC, 1974.
4. W. G. Courtney, L. Cheng, "Control of Respirable Dust by Im-
proved Water Sprays," Bureau of Mines Information Circular
1C 8753, U.S. Department of the Interior, Pittsburgh, PA, 1977,
pp. 92-106.
5. C. W. Lear, "Charged Droplet Scrubber for Fine Particle Control:
Laboratory Study," (NTIS PB-258823), TRW Systems Group,
Redondo Beach, CA, 1976.
6. M. J. Pilat, "Collection of aerosol particles by electrostatic droplet
spray scrubbers," JAPCA 25:176 (1975).
7. D. C. Drehmel, "Advanced electrostatic collection concepts,"
JAPCA 27:1090 (1977).
8. S. A. Hoenig, "Use of Electrostatically Charged Fog for Control
of Fugitive Dust Emissions," EPA-600/7-77-131 (NTIS PB
276645), University of Arizona, Tucson, AZ, 1977.
9. S. A. Hoenig, "Fugitive and Fine Particle Control Using Elec-
trostically Charged Fog," EPA-600/7-79-078, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1979.
10. H. E. B. Hassler, "A new method for dust separation using au-
togenous electrically charged fog," J. Powder and Bulk Solids
Tfchnol. 2:10(1978).
11. W. Walkenhorst, "Charge measurement of dust particles,"
Siaub-Reinhalt. Luft 31: 8 (1971).
12. A. Schutz, "The electrical charging of aerosols," Staub-Reinhalt.
Luft 27: 24 (1967).
13. S. H. Suck, J. L. Kassner, Jr., R. E. Thurman, P. C. Yue, R. A.
Anderson, "Theoretical prediction of ion clusters relevant to the
atmosphere: size and mobility," J. Atmos. Sci. 38:1272 (1981).
14. K. T. Whitby, B. Y. H. Liu, "The Electrical Behaviour of Aero-
sols," in Aerosol Science, C. N. Davies, ed., Academic Press, New
York. 1966, p 59.
15. R. Gunn, "The statistical electrification of aerosols by ionic dif-
fusion," J. Colloid Interface Sci. 10:107 (1955).
16. W. B. Kunkel, "Charge distribution in coarse aerosols as a func-
tion of time," J. Appl Phys. 21:833 (1950).
17. W. B. Kunkel, "The static electrification of dust particles on
dispersion into a cloud," J. Appl. Phys. 21,820 (1950).
18. T. Gillespie, "The role of electric forces in the filtration of aerosols
by fiber filters," J. Colloid Interface Sci. 10: 299 (1955).
19. C. V. Mathai, "Charged fog technology. Part II: prototype tests
of a new charged fog generator for fugitive emission control,"
JAPCA, to be published, (1983).
20. A. Prem, M. J. Pilat, "Calculated particle collection efficiencies
by single droplets considering inertial impaction, Brownian dif-
fusion, and electrostatics," Atmos. Environ. 12:1981 (1978).
21. P. K. Wang. S. N. Grover. H. R. Prupparher. "On Ihe eltect ol
electric charges on the scavenging of aerosol particles by clouds
and small raindrops," J. Atmns. Sri, 35: 17H5 (197H).
22. S. N. Grover, H. R. Pruppacher. and A.E. Hamielec, "A numerical
determination of the efficiency with which spherical aerosol
particles collide with spherical water drops due to inertial im-
paction and phoreticand electrical forces."«/. Atmos. Sci. 34:1655
(1977).
23. K. A. Nielsen, J. C. Hill, "Collection of inertialess particles on
spheres with electrical forces," Ind. Entl. Chcm. Fundam. 15:14M
(1976).
24. K. A. Nielsen, J. C. Hill, "Capture of particles on spheres by in-
ertial and electrical forces," Ind. Eng. Chcm. Fundam. 15, 157
(1976).
25. L. Cheng, "Collection of airborne dust by water sprays," Ind. Enf!.
Chem. Process Des. Dev. 12: 221 (1973).
26. H. F. George, G. W. Poehlein, "Capture of aerosol particles by
spherical collectors: electrostatic, inertial, interceptional, and
viscous effects," Environ. Sci. Tech not. 8:46(1974).
27. E. Cohen, "Research on the Electrostatic Generation and Ac-
celeration of Submicron-Size Particles," Space Technology
Laboratories, Inc., Redondo Beach, California, (1963).
28. K. H. Leong, J. J. Stukel, P. K. Hopke, "Limits in charged-particle
collection by charged drops," Environ. Sci. Techno!. 16: 384
(1982).
29. A. J. Kelley, "Electrostatic metallic spray theory," J. Appl. Phvs.
47:5264(1976).
30. E. T. Brookman, R. C. McCrillis, D. C. Drehmel, "Demonstration
of the Use of Charged Fog in Controlling Fugitive Dust From
Large-Scale Industrial Sources," presented at the Third Sym-
posium on the Transfer and Utilization of Particle Control
Technology, U. S. Environmental Protection Agency. Orlando,
Florida, 1981.
31. J. B. Carlton, L. F. Bouse, "Electrostatic spinner-nozzle for
charging aerial sprays," Trans. Am. Soc. Automotive Eng. 23:
1369(1980).
32. S. E. Law, "Embedded-electrode electrostatic-induction spray-
charging nozzle: theoretical and engineering design," Trans Am.
Soc. Automotive Eng. 21:1097 (1978).
33. M. T. Kearns, D. L. Harmon, "Demonstration of a High Field
Electrostatically-Enhanced Venturi Scrubber on a Magnesium
Furnace Fume Emission," in Particulate Control Devices, Vol.
Ill, EPA-600/69-80-039C, U.S. Environmental Protection
Agency, Research Triangle Park, NC, 1979.
34. W. Balachandran, A. G. Bailey, "The Dispersion of Liquids Using
Centrifugal and Electrostatic Forces," presented at the IEEE/IAS
Annual Meeting, San Francisco, California (1982).
35. J. 0. Hinze, H. Milborn, "Atomization of liquids by means of
rotating cups," J. Appl. Mechanics 17: 145 (1950).
Dr. Mathai is a Senior Scientist with AerpVironment Inc.,
145 Vista Avenue, Pasadena, CA 91107. This paper was sub-
mitted for editorial review on November 3, 1982; the revised
manuscript was received on April 29,1983.
July 1983 Volume 33, No. 7
90
-------
APPENDIX B
Charged Fog Technology. Part II: Prototype Tests of
A New Charged Fog Generator for Fugitive Emission Control
91
-------
Reprinted frt>m APCA JOURNAL, Vol. :«, No. 8, Autiutt 19M
(with permission)
Charged Fog Technology
Part II: Prototype Tests of a New Charged Fog Generator
for Fugitive Emission Control
C. V. Mathai
AeroVironment Inc.
Pasadena, California
The prototype charged log generator described In a preceding
paper1 was field tested on a fugitive emission source al a bentonlte
ore processing plant In Worland, Wyoming, during 1981. Participate
matter samples were collected as fine and coarse fractions under
three different test scenarios: with no control, with partial control
(uncharged fog), and with full control (charged fog). Measured par-
tlculate matter sample concentrations were normalized for each test
day with respect to the background values so that particle control
efficiency of the device could be evaluated without any bias during
the entire test program.
These tests have shown that mean value of the Inhalable particle
control efficiency of charged fog measured under all Instrument
settings and field conditions Is Increased by 78% when compared
with uncharged fog. In particular, fine particle control efficiencies of
over 90% were recorded under optimum Instrument settings and Ideal
field conditions. The bentonlte particles seemed to carry a net positive
charge. The optimum Instrument settings were found to be: 60 L/h
water flow rate, an applied voltage of 10-15 kV and a spray pattern
which covers maximum volume of dust-laden air. Ideal field conditions
are high relative humidity and calm or low winds.
Charged fog technology appears to be an effective and economi-
cally feasible method to control sources of fugitive particle emission
in the inhalable size range. Further research Is needed to evaluate
the inhalable particle control efficiency of the device for emissions
of various chemical compositions.
The prototype charged fog generator (CFG) described in a
preceding paper1 was field tested at the Kaycee Bentonite
Corporation's bentonite ore processing plant in Worland,
Wyoming, during 1981. The plant is located in a remote area
in a fairly flat terrain. A railroad track and a paved road lie on
the east side of the plant. Bottom-dump trucks filled with
bentonite ore from mines, about 15 miles away, arrive from
the south and unload bentonite on two piles, approximately
Dr. Mathai is a Senior Scientist with AeroVironment Inc.,
145 Vista Avenue, Pasadena, CA 91107. This paper was sub-
mitted for editorial review on November 3, 1982; the revised
manuscript was received on April 29, 1983.
Ci'pvnithl 19»:l-Air Pnllutiun Control Aasoctaliun
756
92
100 m to the northwest and southwest of the plant building.
Front-end loaders carry bentonite from these piles and unload
onto the grill of a hopper which is attached to the west wall of
the plant building. The hopper is completely enclosed except
on one side through which the front-end loaders unload. From
the bottom of the hopper (approximately 4 m below the
hopper grill level) the ore is carried inside the plant by con-
veyor belts to processing areas. When the front-end loaders
dump the ore on the hopper grill, clouds of fugitive dust fill
the hopper and spread to outside the hopper. The hopper was
selected as the source of fugitive dust emission on which the
CFG would be tested.
The objectives of the field tests were to evaluate:
1. Optimum instrument settings for maximum value of in-
halable particle control under various combinations of
meteorological parameters in the field; and
2. Optimum field conditions for maximum value of inhalable
particle control efficiency of the device.
A secondary objective was to determine whether positively
or negatively charged fog would control particles more effi-
ciently.
Experimental Arrangements
The hopper is 6.7 m wide, about 2 m deep and about 3 m
high from the grill level. The bucket of the front-end loader
is about 2.4 m wide and successive loads are unloaded uni-
formly over the 6.7 m wide hopper. About 10 front-end loader
dumps fill the hopper to the grill level, and take about 25
minutes to complete. The bucket is removed from inside the
hopper area in 20-25 seconds. One full hopper of ore is carried
away by the conveyor belt in about an hour.
The CFG and the particle sampling instruments were
mounted on the outside of the south wall of the hopper, about
4 m above ground level. A platform was built to mount these
instruments. Ideally, the particle sampler inlet should have
been mounted on the east (rear) wall of the hopper, but for
practical reasons could not be done. The CFG sprayed water
droplets across the hopper above the grill. The total volume
to be treated by the charged fog was about 40 m3 (6.7 m X 2
m X 3 m). This volume is somewhat larger than the maximum
coverage of the CFG fog. Unfortunately, at that time only one
prototype CFG was available for tests. To have had a second
unit mounted on the north wall of the hopper and operated
concurrently would have been ideal.
Particle samples were collected using a Sierra Instruments
Model 230 CP cyclone preseparator followed by a Sierra In-
struments two-stage cascade impactor.- The cyclone's air inlet
protruded 0.3 m into the hopper. The particle sampling system
was operated at a flow rate of 0.85 nvVmin (30 cfm). The
Journal of the Air Pollution Control Association
-------
60
40
*0
to -
10
B uncharged Fofl
0 Chug** I* • ->
PvttelM (My Fti* ft Cotrs*
Figure 1. Mean inhalable particle control efficiency of all the
test runs (for various CFG settings and field conditions) for
charged fog (both positive and negative) and uncharged fog.
sampler flow rates were calibrated at regular intervals during
the entire test program. The cyclone has a particle cut-point
of about 7.3 nm at this flow rate. The impaction plates of the
cascade impactor were chosen so that the upper filter would
collect particles larger than 1.8 pm. The lower backup filter
collects all particles smaller than 1.8 fan. For the purpose of
this report, particles collected on the backup filter will be
characterized as the fine particle fraction and those on the
upper filter will be characterized as the coarse fraction. The
mass of coarse fraction collected was often an order of mag-
nitude smaller than the fine fraction, possibly indicating a
particle bounce problem, with particles larger than the cut-
point of 1.8 nm reaching the backup filter. Subdividing the
samples into more size ranges would have required either
much longer sampling time or much higher flow rates to obtain
acceptable filter loadings. Neither of these alternatives was
desirable.
Field Test Design
To determine the particle control efficiency of the CFG,
three test scenarios were designed. In the first scenario, no
attempt was made to control the dust inside the hopper, ex-
cept that the CFG's fan blew continuously. This test scenario
was necessary to ensure that the mixing of the dust cloud was
nearly identical among various runs in order to make direct
comparison of tests with and without charged fog. In the
second scenario, uncharged fog was applied on the dust cloud,
and in the third scenario, charged fog was applied. Other
governing parameters were varied under each scenario to yield
a statistically acceptable set of test data.
The parameters varied were water flow rates, fog pattern
(short or long), applied high voltage, wind conditions, and
relative humidity. In addition, a group of test runs were per-
formed with the polarity of charges on the droplets reversed.
This test protocol called for a total of 32 test runs. However,
in actual practice, data had to be collected under prevailing
field conditions; hence, a total of 96 runs were performed.
Particle samples were collected on preconditioned and
preweighed glass fiber filters. Each particle sample was col-
lected during 12-15 front-end loader dumps (roughly 30-40
min). Simultaneously, wind speed and direction and relative
humidity were also recorded. After each sample was collected,
the filters were transferred to special envelopes and brought
to AeroVironment's laboratory for analysis. Further details
of the experimental program are described elsewhere.3
Data Processing
The mass of the fine and coarse fraction and sum of the two
were obtained from the final and initial weights of the filters.
Using known values of the sample time, number of dumps, and
flow rate, particle mass concentration for each dump was
calculated.
The amount of dust generated in the hopper fluctuated
from dump to dump. To overcome this variation, each sample
was collected during a total of 10-15 dumps and the mean
value for each dump was calculated. The amount of dust
generated also varied from day to day. This variation was
caused by factors such as changes in ambient conditions and
the moisture content of the bentonite ore itself. The effect of
the latter is not very significant since the ore stored in each
pile comes from the same mine and a pile is completely pro-
cessed before a new pile is started. To overcome the day to day
dust level fluctuations, sample concentrations were normal-
ized for each day with respect to the background value (the
fan only value) and a percentage particle collection (control)
efficiency was calculated. Percentage particle control effi-
ciency E is obtained from:
C0-C
X100
where C0 and C are the particle concentrations corresponding
to fan only and fog test scenarios, respectively.
Values of £ are calculated for the fine fraction, the coarse
fraction and the sum of fine and coarse. Percentage particle
control efficiencies and the corresponding field condition data,
water flow rate, and applied high voltage for the whole field
test program are available from the author.
Results and Discussion
Figure 1 shows the mean values of the measured percentage
fine particle control efficiency and total particle control effi-
ciency for charged (striped bars)—both positive and nega-
tive—and uncharged (solid bars) fog. For this comparison, all
test runs under all instrument settings and field conditions
are included. The mean particle control efficiencies and
standard deviations for these tests are shown in Table I.
These numbers show that, even under average field con-
ditions and instrument settings, inhalable particle control
efficiency can be almost doubled by electrically charging the
water droplets. However, under optimum instrument settings
and favorable field conditions, the improvement in inhalable
particle control efficiency can be expected to be much higher.
It may also be pointed out that the volume of the dust cloud
treated was somewhat larger than the maximum coverage of
the CFG; therefore, the observed improvement in particle
control efficiency is remarkable.
Although the effect is not as strong as for the fine fraction,
the mean value of the particle control efficiency of coarse
particles increased modestly when the droplets were charged.
Table I. Mean particle control efficiencies of all test runs.
Particle Control Efficiency (%)
Charged Fog Uncharged Fog
Fine particles only
Mean
Standard deviation
Fine and coarse particles
together
Mean
Standard deviation
48.1
23
44.5
21.8
27.8
25.3
25
24.4
Auaust 1983 Volume 33, No. 8
93
757
-------
This result is in good agreement with the theoretical predic-
tions. However, the size range of particles collected in the
coarse mode was fairly narrow, and the mass collected was
often an order of magnitude smaller than the fine fraction.
Therefore, this particular experiment could not demonstrate
the full effect of charged fog on these particles. Another
problem which may have inhibited coarse particle collection
is the postulated particle bounce effect, by which some of the
larger particles pass the upper impaction plate and settle on
the backup filter with the fine particles. Consequently, most
of the ensuing discussion will concentrate on the fine fraction
of the particles collected.
Figure 2 shows fine particle control efficiency E plotted as
a function of ambient relative humidity (RH) for two sets of
instrument settings for an applied high voltage of 4 kV (pos-
itive charges) and a water flow rate of 60 L/h. The method of
least squares was used to fit a straight line to the data sets,
shown in the figure, yielding a correlation coefficient of 0.91
for the broad spray and -0.19 for the narrow spray. The cor-
responding slopes are 0.64 and —0.07, respectively. Although
the wind conditions were not identical for all the data points,
it can be seen that the fine particle control efficiency increases
with increases in ambient RH for a broad spray, while it is
fairly independent of RH for a narrow spray. The generally
lower values of E for the narrow spray can be easily explained.
A broad spray will cover most of the dust cloud in the hopper
while the narrow spray will cover a smaller portion, resulting
in slightly lower E values.
The difference in the dependence of £ on RH for the broad
spray and narrow spray can be explained as follows. For a
narrow spray, most of the droplets occupy a volume away from
the open side of the hopper, and when fog is continuously
applied, this area becomes more humid than outside the
hopper or near the hopper opening, if the wind is not too
strong. Thus there is less droplet evaporation, and conse-
quently, fairly steady particle collection efficiency. However,
in the case of a broad spray, the droplets are distributed from
the rear wall of the hopper to outside the open side of the
hopper. Thus, as the ambient RH increases, a smaller number
of droplets evaporate, leaving more droplets to collect dust
particles; on the other hand, when the ambient RH decreases,
more droplets are lost due to evaporation near the open side
of the hopper and outside the hopper.
The effect of the longer droplet lifetime in a higher RH at-
mosphere increases the particle control efficiency. This result
era s«tt»c»
W«l« flow: «0 Iph
. VolUe*: 4 kV
TO
S
I
s
40 50 60 70 80 90
Ambient Rclallv* HumMtty 1%)
Figure 2. Fine particle control efficiency of the CFG plotted as a function of
ambient relative humid'ty for broad (O) and narrow (D) spray patterns.
3
30
10
0 Ch«rg«d (-) Fog
ffl Cnirgcd <»> Fog
•
-
1
\
S
V/////////////////S
!
\
s
\
'
-------
Conclusions
Extensive field tests of the prototype CFG at a bentonite
ore unloading operation show that:
1) The mean inhalable particle control efficiency of charged
fog measured under all instrument settings and field
conditions shows a 78% increase when compared with the
corresponding value for uncharged fog.
2) Individual tests with charged fog show a particle control
efficiency (absolute value) as high as 87% over uncharged
fog. Overall fine particle control efficiencies of over 90%
were achieved under optimum instrument settings and
favorable field conditions.
3) The enhancement in particle control efficiency is signifi-
cant for the fine particles, and that for the coarse particles
is moderate, in agreement with theory.
4) The relative humidity (in this particular experimental
setup) seemed to play a significant role in determining the
overall particle collection efficiency. The lifetime of the
droplet is believed to be the dominant factor in deter-
mining what the particle control efficiency will be.
Therefore, the droplets need to be large enough not to
evaporate too quickly, yet small enough to yield a high
particle control efficiency. Mean droplet diameters ranging
from 100-200 /im appear to be reasonable for charged fog
devices.
5) Negatively charged droplets gave higher values of particle
control efficiency than did positively charged fog,
suggesting that inhalable bentonite particles carry a net
positive charge.
6) Measured inhalable particle control efficiencies were
higher for higher applied voltages. An optimum voltage for
good particle control seems to be 10-15 kV.
7) Measured inhalable particle control efficiency was higher
when charged droplets could cover more of the dust-laden
air in the hopper. It thus appears that if the source is large,
multiple units of CFG will have to be used. In this experi-
mental setup, higher water flow rate resulted generally in
higher collection efficiencies, although the key element was
how many particles were treated by the droplets. This
quantity depended, to a certain extent, on the spray pat-
tern for this experiment.
40
•0
10
10
0
m
/
I
trr
-
s
1 30
w
i
tlcl* Control E
M
O
1 to
IL
0
-
-
.
-
-
/
Ys
\
I
-
-
-« kV .4 kV
60 l/h 30 l/h
Figure 4. Comparison of fine particle control efficiency of the
CFG for two water flow rates, with all other parameters nearly
identical.
Flgwi 5. Comparison of fine particle control efficiency of the
CFG for two applied high voltages, with all other parameters
nearly identical.
8) Because of the type of particle sampling method used and
the field setup, the effect of wind speed and direction on
particle control efficiency could not be quantified with the
available data.
9) The optimum CFG instrument settings are found to be 60
L/h water flow rate, a spray pattern which will cover a
maximum volume of dust-laden air (broad or narrow spray-
depending on the extent of the source), an applied voltage
of 10-15 kV, and positive or negative charge of opposite
polarity to the charges on the dust particles. Ideal field
conditions are high relative humidity (to ensure long
droplet lifetime), and calm or low winds.
Although the charged fog method of dust suppression is
only a temporary form of inhalable particle control, there is
no other practical alternative method to control inhalable
particles effectively under certain material handling situations
and for other moveable emission sources like road sweepers,
front-end loaders, and construction activities. Thus, in those
cases where any other form of dust control is not economically
feasible and at locations where personnel exposure is haz-
ardous, charged fog technology appears to be the ideal dust
control method. Additional research work is required to
evaluate the inhalable particle control efficiency of the
charged fog generator on emission sources of different
chemical composition.
Acknowledgments
This work was funded by the Industrial Environmental
Research Laboratory of the U.S. EPA under contract 68-
02-3145. The author is grateful to EPA Project Officers Dr.
D. C. Drehmel and Mr. W. B. Kuykendal for their technical
guidance during the course of this research project. The author
is also indebted to Mr. J. S. Kinsey and Mr. L. A. Rathbun for
their technical assistance, and to the personnel of the Kaycee
Bentonite Corporation for their excellent cooperation during
our field tests in Worland, Wyoming.
References
1. C. V. Mathai, "Charged fog technology: Part I. Theoretical hack-
ground and instrumentation development." JAPCA 33: 664
(1983).
"> Sierra Instruments product literature. P.O. Box 909, Carmel Valley,
California 93924. 1981.
3. C. V. Mathai, "A New Charged Fog Generator for Inhalable Par-
ticle Control." Final Project Report submitted to I'.S. EPA.
Contract No. 68-02-3145, AeroVironment Inc., Pasadena, CA,
1982.
95
759
-------
TECHNICAL REPORT DATA .
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-84-016
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
A New Charged Fog Generator for Inhalable
Particle Control
5 REPORT DATE
February 1984
6. PERFORMING ORGANIZATION CODE
7. AUTHOH(S)
C.V. Mathai
8. PERFORMING ORGANIZATION REPORT NO
AV-R-82/505
9. PERFORMING ORGANIZATION NAME AND ADDRESS
AeroVironment, Inc.
145 N. Vista Avenue
Pasadena, California 91107
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3145
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13 TYPE OF REPORT AND PERIOD COVERED
Final: 7/79 - 4/82
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL_RTp project officer is William B. Kuykendal, Mail Drop 61,
919/541-7865.
is. ABSTRACT The report discusses the development of a new charged fog generator
by modifying a commercial rotary atomizer. In this device, the droplets generated
are contact-charged to provide a high charge-to-mass ratio of 1.2 microcoulombs/g.
The droplets have a number concentration median diameter of about 100 micrometers
and a mass median diameter of about 200 micrometers. The water flow rate is var-
iable (4-70 1/h) and the fog spray pattern can be changed easily from long and nar-
row to broad and short, with a typical spray coverage of 16-24 cu m. The device
uses about 1 kW of power (110 VAC) and is portable. Extensive field tests of the CFG
(at a bentonite ore unloading operation) were performed to determine the dependence
of its inhalable particle control efficiency (PCE) on various instrument settings and
field conditions. These tests show that the overall mean PCE is 78% higher than the
corresponding value for uncharged fog. Individual PCEs as high as 88% were achie-
ved. The lifetime of the droplets seems to be the dominant factor determining the
PCE; and PCE values were higher for higher applied voltages and higher water flow
rates. The data suggest that, under optimum instrument settings, the PCE of water
droplets could be doubled by charging the droplets.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Fogging
Aerosol Generators
Atomizers
Charging
Particles
Dust
Aerosols
Drops (Liquids)
Pollution Control
Stationary Sources
Fog Generators
Charged Fog
Inhalable Particles
Particulate
13B
14G
13 D
13K.07A
11G
07D
Release to Public
19. SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
103
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
96
------- |