[iwronntenta! Protection Technology Series
PRO
strlt. Env
ental RBsearcJi
of Reseanh and
Offi
'U.S.
Research Tri'aifiB Pa^k, N,C. 21711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed
to develop and demonstrate instrumentation, equipment and
methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the
new or improved technology required;for the control and.'treatment
of pollution'sources to'meet environmentajl quality standards.
-.?. i - ^
EPA REVIEW NOTICE
This report has been reviewed by EPA and approved for publi-
cation. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
This document'is available to the public through the National
Technical Information.Service, Springfield, Virginia 22161.
-------
EPA-600/2-75-054
OVERVIEW OF
EPA/IERL-RTP
SCRUBBER PROGRAMS
by
Douglas W. Cooper, Lee W. Parker, and Eugene Mallove
GCA Corporation
GCA/Technology Division
Bedford, Massachusetts 01730
Contract No. 68-02-1316, Task 10
ROAPNo. 21ADL-002
Program Element No. 1AB012
EPA Task Officer: Leslie E. Sparks
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
September 1975
-------
CONTENTS
Page
List of Figures iv
List of Tables vi
Acknowledgments , vii
Sections
I Conclusions 1
II Recommendations 2
III Introduction 3
IV Topical Overview 15
V Chronological Overview 35
VI References 102
iii
-------
LIST OF FIGURES
No. Page
1 Generalized Description of Scrubber System 4
2 Predicted Aerodynamic Cut Diameter Versus Pressure
Drop and Power Consumption (Adapted from Scrubber
Handbook).4 Lines la and Ib are for Sieve Plates;
Line 3 is for Impingement Plate; Line 4 is for a
Packed Column; Lines 2a and 2b are for Venturi
Scrubbers with f = 0.25, 0.50 Respectively. See Text. 8
3 Penetration Calculated From a Venturi Scrubber Model
as a Function of Pressure Drop and Particle Aerodynamic
Diameter (Corrected Version of Scrubber Handbook^ Figure) 19
4 Particle Collection Efficiency of Electrostatic Spray
Droplet Scrubber as a Function of Particle Size 25
5 Particle Penetration Versus Water Vapor Condensed 43
6 FF/C Scrubber Performance Comparison (APT, Inc.) 48
7 Collection Efficiency Versus Gas Velocity in Tube Bank. 58
8 Summary of Scrubber Performance Curves for Adiabatic
Saturation, Condensation, and Vaporization Scrubbing-
Aerosol B 64
9 Generalized Two-Phase Jet Scrubber System 74
10 Fractional Efficiency of the Aronetics Scrubber Based
on Optical, Diffusional, and Impactor Data 75
11 Fractional Efficiency of the Lone Star Steel Steam-
Hydro Scrubber 77
12 Single-Stage Dynactor Diffusion System Cross-Sectional
View 80
iv
-------
LIST OF FIGURES (Continued)
No. Page
13 Dynactor Scrubber Collection Efficiency Versus Particle
Aerodynamic Diameter, Effects of Loading and Dust Type 83
14 Dynactor Scrubber Collection Efficiency Versus Particle
Aerodynamic Diameter, Effects of Flow Rate and Inlet
Temperature 84
TM
15 Fractional Efficiency of Idealized Centrifield Scrub-
ber (9 = 65) 88
16 Fractional Efficiencies as Determined by the Four Methods
Used in the Test Program. The Particle Sizes Shown for
the Impactor Data are Stokes Diameters Based on a
Particle Density of 2.5 grams/cm^ 91
-------
LIST OF TABLES
No. . Page
1 Ten Largest Sources of Particulates and Scrubber
Systems Used, If Any 6
2 List of Projects Related to Scrubber Research and Develop-
ment Sponsored by Industrial Environmental Research Labor-
atory-Research Triangle Park (IERL-RTP) 12
3 Publications Available From NTIS Related to IERL-RTP
Sponsored Work in Scrubber Technology 13
4 FF/C Scrubber Performance Comparison 49
5 Selected Comparisons of Scrubber Performances at
Different Operating Conditions 67
vi
-------
ACKNOWLEDGMENT
The direction and content of this work was strongly influenced by the
guidance and assistance of the Project Officer, Dr. Leslie E. Sparks,
Industrial Environmental Research Laboratory - Research Triangle Park
(IERL-RTP), of the Environmental Protection Agency, Research Triangle
Park, North Carolina, to whom we extend our thanks.
vii
-------
SECTION I
CONCLUSIONS
This report is descriptive rather than evaluative, and no conclusions
have been formulated.
-------
SECTION II
RECOMMENDATIONS
No recommendations are presented because this report is descriptive
rather than evaluative.
-------
SECTION III
INTRODUCTION
In this section we present an introduction to the particulate scrubbing
programs sponsored by the Industrial Environmental Research Laboratory-
Research Triangle Park of the U.S. Environmental Protection Agency
(EPA). The components of a particulate pollutant scrubber system are
briefly described, as an aid to the understanding of the short discussion
of the status of the Industrial Environmental Research Laboratory-Research
Triangle Park scrubber programs which follows. Finally, a.tabular pre-
sentation is made of all the IERL-RTP scrubber programs. Section IV .
shows the interrelations between the scrubber programs on a topic-by-
topic basis, serving also as a review of the state of the art. Section V
describes the chronological development of the scrubber program, high-
lighting the reasons behind that evolution.
DESCRIPTION OF GENERALIZED SCRUBBER SYSTEM
Figure 1 is a schematic of a scrubber system. An emissions source pro-
duces airborne pollutants at the mass rate of m , described in units
such as pounds per hour or kg/s. The scrubber uses water and power to
remove some of the pollutant, material from the air; conservation of mass
requires that at steady state the air-borne mass emission rate down-
stream from the scrubber, m and the water-borne mass emission rate,
m *, will equal the mass rate into the scrubber, m,. The air from the
scrubber will usually have water droplets entrained with it; a demister
(entrainment separator) will be used to remove these droplets and what-
ever pollutant mass they carry with, them, leaving an air flow with a
-------
"FUEL"
AIR
WATER
EMISSIONS
SOURCE
AIR
V V
POWER
SCRUBBER
m
^
-t-m.
WATER
4- mo
POWER
DEMISTER
AIR
+ m
3
CLARIFIER
WATER
+ m3*
POWER
SOLIDS
Figure 1. Generalized description of scrubber system
-------
mass rate of emissions m^. .The collection efficiency of the scrubber
plus dcmister would then be 1 - m /n^ on a mass basis. The water from
the scrubber will contain captured material; generally it will be pumped
to a clarifying system which will use power to remove some of the par-
. i
ticulate material as solid waste (with a mass rate-in ). The water leaves
the clarifier with a mass rate of water-borne emissions m*. Mass con-
servation means that the pollutants generated will leave the scrubber
as gases, liquids or solids either air-borne, liquid-borne or solid-borne.
As for clarifiers and their effluents, IERL-RTP has sponsored a 4-year
study of environmental acceptability and economics of techniques for
treating the sludge from scrubbing methods of flue gas desulfurization,
with a final report due this year. This is not discussed further in
this document, as the research is more related to gas scrubbing than to
fine particulate scrubbing and the amount of solid material to be treated
in flue gas desulfurization will differ greatly from that encountered in
most particulate scrubbing contexts.
IMPORTANCE OF SCRUBBERS
Scrubbers have certain advantages as particle control devices: They
often collect both particles and gases. Gas temperatures are reduced
by evaporation. The explosive characteristics of certain dusts may be
quenched by the scrubber moisture. Finally, scrubbers are usually more
compact than fabric filter or electrostatic precipitator installations
of comparable volume flow rate capacity.
Midwest Research Institute (MRI) in its document, Particulate Pollutant
2
System Study, ranked the sources of particulate emissions. The
Industrial Environmental Research Laboratory-Research Triangle Park staff
has identified the kinds of scrubbers used, if any, in the MRI list of
major source categories, an abridged version of which appears here as
Table 1. The top 10 source types are listed along with scrubbers em-
ployed by installations in the industry. Nine of the 10 most significant
-------
Table 1. TEN LARGEST SOURCES OF PARTICIPATES AND SCRUBBER SYSTEMS
USED, IF ANY
Source
type
Crushed stone
Coal combustion (elec-
trical utility)
Basic oxygen furnace
(steel)
Kraft pulp recovery
furnace
Cement rotary kiln
Hot-mix asphalt dryer
Ferroalloy electric
furnace
Open hearth furnace
(steel)
Coal combustion
(industrial)
Lime plants, rotary
kilns
Scrubber type
High b
energy
X
X
X
X
X
Moderate
energy
Low ,
energy
X
X
X
X
Novel
X
X
X
Flux
force
X
X
Charged
droplet
X
X
Ranked according to MRI Study, Particulate Pollutant System Study,
(Table 7,'Vol. 2)
, 32
~250 cm H 0 or ~ 15 hp/1000 acfm or ~ 25 x 10 N/m
2 2
C -60 cm HJ) or ~ 3 hp/1000 acfm or ~ 60 x 10 N/m
O *)
d ~20 cm H 0 or ~ 1 hp/1000 acfm or ~ 25 x 10 N/m
-------
sources have installations using scrubbers. Nine of the 10 next most
significant sources also employ scrubbers of one type or another.
Clearly, scrubber technology has an important position in particulate
pollution control.
DESCRIPTION OF IERL-RTP SCRUBBER STUDIES
Objective
The broad objective of the fine particle scrubber program is the develop-
ment of energy-efficient and cost-efficient scrubber systems for fine
particle control. For various industries, scrubbers have been found to
be the cost-effective control technology. Put into quantitative terms:
the objective is to develop a scrubber with power consumption per volume
flow equivalent to 30-50 cm WC (12-20 in. ELO) pressure drop capable of
collecting at least 90 percent by mass of particles having aerodynamic
diameters <_ 1 um.
State of the Art
Figure 2 is adapted from a paper by Calvert which in. turn is partially
based on results presented in the Scrubber Handbook. It shows the
cut diameter of the scrubber system plotted against its actual pressure
drop, its hydraulic power consumption per unit volume of air flow. Hy-
draulic power is volume flow rate times pressure drop. Calvert showed
how total efficiency can also be obtained from scrubber cut diameter
3
and aerosol size parameters. The cut diameter is the particle aero-
dynamic diameter for which the scrubber collection efficiency is 50 per-
cent. This can be obtained from data on scrubber collection efficiency
versus particle aerodynamic diameter. (The particle aerodynamic diameter
is the diameter of a spherical particle of unit density that has the same
settling velocity.) The figure shows sieve plate, impingement plate, and
packed scrubber characteristics as well as curves, for venturl scrubbers
under a variety of conditions (denoted by different empirical curve
-------
00
4.O
0.25
POWER, hp/IOOOocfm
0.8 1.0 2.0 3.0
5.0
8.0 10
VENTURI
AND
NOVEL
SCRUBBERS
70 9O 100
200
30O
Figure 2.
PRESSURE DROP, cm WC
Predicted aerodynamic cut diameter versus pressure drop and power
consumption (adapted from Scrubber Handbook). Lines la and Ib
are for sieve plates; line 3 is for impingement plate; line 4 is
for a packed column; lines 2a and 2b are for venturi scrubbers
with f = 0.25, 0..50 respectively. See text.
-------
fitting factors, f). With the exception of two TCA (Turbulent Contact-
ing Absorber) scrubbers, the novel scrubbers tested by Industrial En-
vironmental Research Laboratory-Research Triangle Park can be repre-
sented by points along or above the line 2b (f = 0.5) for venturi scrub-
bers. The TCA scrubbers are represented by the circle labeled TCA.
Agreement between scrubber theory and experimental evaluation shows that
inertial impaction is at present the primary collection mechanism at
work. To improve scrubbers dramatically, new forces will need to be
introduced into the scrubbing process or the fine particles will have
to be grown to a size where they can be caught inertially.
Status of the IERL-RTP Program
The major thrust of the IERL-RTP's scrubber program has been aimed at
developing and demonstrating Flux Force/Condensation (FF/C) Scrubbers.
In an FF/C scrubber, water vapor is condensed in the scrubber. When the
water vapor condenses, additional forces and particle growth contribute
to the particle collection process. When the water vapor or steam is
"free," FF/C scrubbers are low energy users. However, when waste heat
is not available, FF/C scrubbers require additional energy inputs for
efficient particle collection. A rough idea of the energy consumption/
performance relationship for FF/C scrubbers is shown in Figure 2. The
left hand region of the FF/C regime is for free steam and the right
hand region is for purchased steam. Note that when steam is free FF/C
scrubbers approach the program objective. The questions of how much
steam is needed and how much is free are major unknowns at present. The
answers to both questions are likely to be source-specific. Thus, pilot
demonstrations on a variety of sources are necessary to provide the re-
quired data. A pilot demonstration is underway and an additional one
is planned for FY76.
With two possible exceptions, performance data for all the non-FF/C
scrubbers (i.e., all conventional and novel scrubber work) confirm
-------
Figure 2. The first possible exception came out of the "Wet Scrubber
Liquid Utilization" program at Stanford Research Institute. This
research indicated that a series of low-energy, low-efficiency scrub-
bers might achieve much higher total efficiency at a given energy con-
sumption than could a single high energy scrubber, allowing a trade-off
between capital costs and operating costs. This lead, plus final con-
firmation of Figure 2, will be followed up in research funded in FY75.
The other possible exception to Figure 2 is the TCA or mobile bed
scrubber. Data from field tests of TCA scrubbers are shown in
Figure 2. Note that the TCA point is below the venturi and other
scrubber lines by a significant amount (TCA cut diameter is about
one-half that of a venturi at similar power consumption). At present
there is no explanation for the observed performance of these mobile bed
scrubbers. In fact, one scrubber vendor claims that mobile bed scrubbers
(single-stage or multi-stage) are less efficient than venturi scrubbers.
An FY75 program will seek the explanation for the observed performance
and design equations and theoretical models for mobile bed scrubbers.
FY76 funds will be used to investigate effects of slurry scrubbing in
mobile bed scrubber performance. Pilot plant work will be coordinated
with the full-scale tests at the Shawnee Wet Limestone Scrubbing Test
Facility, a coal-fired boiler instrumented for research into flue gas
3
desulfurization at realistic flow rates (e.g.,20,000 acfm, 10 m /s).
The overall efficiency of a scrubber system is determined by the
efficiency of the scrubber and the efficiency of the entrainment
separator. Recent field data indicate that in some cases inefficient
entrainment separator operation is a major cause of poor collection of
fine particles by scrubbers. IERL-RTP has nearly completed a systematic
study of entrainment separators. FY76 funds are provided to develop
and demonstrate, in cooperation with SO scrubbing research and develop-
ment, efficient and trouble-free entrainment separators.
10
-------
We close this section with two tables. Table 2 lists the scrubber pro-
jects sponsored by IERL-RTP, listed in order of starting date. Table 3
gives publications available from NTIS concerning IERL-RTP work in scrub-
ber technology. (Order from: National Technical Information Service,
U.S. Department of Commerce, Springfield, Virginia 22151). Next, in
Section IV, the major themes of the IERL-RTP work are discussed. In
Section V is presented a chronological overview of the IERL-RTP programs
concerning particulate scrubbing technology.
11
-------
Table 2. LIST OF PROJECTS RELATED TO SCRUBBER RESEARCH AND DEVELOP-
MENT SPONSORED BY INDUSTRIAL ENVIRONMENTAL RESEARCH LABOR-
ATORY-RESEARCH TRIANGLE PARK (IERL-RTP)
Starting date
Title
1970
1971
September 1972
December 1972
June 1973
September 1973
September 1973
October 1973
December 1973
June 1974
1974
June 19.74
July 1974
July 1974
August 1974
September 1974
January 1975
February 1975
April 1975
Proposed
Proposed
Proposed
Wet Scrubber Systems Study
Flux Force/Condensation Scrubber Feasibility
Fine Particle Scrubber Performance Tests
Systems of Charged Droplets and Electric Fields for the
Removal of Submicron Particulates
Wet Scrubber Entrainment Separation
Pilot Scale Demonstration of Charged Droplet Scrubbing
Wet Scrubber Development II (FF/C Scrubbers)
Wet Scrubber Liquid Utilization
Design and Fabrication of Mobile Wet Scrubber Facility
Foam Scrubbing for Fine Particle Control
Growth of Fine Particles by Condensation .
Fine Particle Collection with University of Washington
Electrostatic Scrubber
Evaluation of Novel Control Devices
FF/C Scrubber Pilot Plant
Evaluation of Wet Scrubbers for Control of Particulate
Emissions from Utility Boilers
Operation of EPA-Owned Mobile Units, Aerodynamic Test
Chamber, and Pilot Scrubbers
Rocket Motor Emissions Control
Entrainment Characteristics of Mobile Bed Scrubbers
Evaluation of Horizontal Scrubber
Evaluation of Systems for Control of Rocket Motor Test
Pad Emissions
Evaluation of Electrostatic Scrubber
Effects of Interfacial Properties on Fine Particle
Scrubbing
Mobile Bed FF/C Scrubbers for Collection of Fine Particles
Wet Scrubber Energy Utilization
12
-------
Table 3. PUBLICATIONS AVAILABLE FROM NTIS RELATED TO IERL-RTP
SPONSORED WORK IN SCRUBBER TECHNOLOGY
Novel devices
Contractor
(contract number)
NTIS number
Report title and date
Southern Research
Institute for
M.W. Kellog
(68-02-1308)
GCA
(68-02-1316)
Southern Research
Institute
GCA
(68-02-1487)
PB 232-436/AS
EPA-650/2-74-028
PB 234-146/AS
EPA-650/2-74-036
PB 239-422/AS
EPA-650/2-74-129
PB 240-397/AS
EPA-650/2-75-024-6
Lone Star Steel Steam-Hydro Air
Cleaning System Evaluation (Task
11) 4/74 . .
Braxton Sonic Agglomerator Evalua-
tion (Task l)--5/75
Evaluation of Aronetics Two-Phase
Jet Scrubber--12/74
Pentapure Impinger Evaluation3/75
New concepts
Contractor
(contract number)
NTIS number
Report title and date
MIT
(68-02-0018)
CSL
PB 205-188
(APTD 0869)
PB 205-187
(APTD 0868)
PB 236-676/AS
EPA-650/2-74/081
Electrical Induction of Particulate
Agglomeration (Final Report, Task
Order 7)--8/71
Charged Droplet Technology for Removal
of Particulates from Industrial Gases
(Final Report, Task Order 8)--8/71
Seminar on Electrostatics and Fine
Particles, September 1973--8/74
Wet scrubbers
Contractor
(contract number)
NTIS number
Report title and date
Ambient Purifica-
tion Technology
(CPA 70-95)
APT
PB 213-016
EPA-R2-72-118a
PB 213-017
EPA-R2-72-118b
PB 227-307
EPA-650/2-73-036
Jet Scrubber Systems Study, .
(Handbook) Vol. I--8/72
Same Vol. II-8/72 (Final report &
Bibliography)
Feasibility of Flux Force/Condensa-
tion Scrubbing for Fine Particulate
:ollection--10/73
13
-------
Table 3 (continued).
PUBLICATIONS AVAILABLE FROM NTIS RELATED TO
IERL-RTP SPONSORED WORK IN SCRUBBER TECHNOLOGY
Wet scrubbers
Contractor
(contract number)
NTIS number
Report title and date
APT, Inc.
(68-02-1328)
APT, Inc.
(68-02-0637)
APT, Inc.
Stanford Research
Institute
(68-02-1079)
PB 239-335/AS
EPA-650/2-74-112
PB Not assigned
EPA-650/2-74-119-;
PBr240-235/AS
EPA-650/2-74-093
PB 237-749/AS
EPA-650/2-74-108
EPA Fine Particle Scrubber Symposium
(San Diego, 5/74)10/74
Entrainment Separators for Scrubbers-
Initial Report--10/74
Fine Particle Scrubber Performance
Tests--10/74
Jet Scrubber Liquid Utilization--
10/74
14
-------
SECTION IV
TOPICAL OVERVIEW
INTRODUCTION
In general terms, the work of the Industrial Environmental Research
Laboratory-Research Triangle Park in scrubber technology has focused on
understanding the operation of scrubbers and using that understanding to
improve their performance as fine particle control devices. A review
of scrubber technology under IERL-RTP sponsorship produced the Scrubber
Handbook written by Calvert and co-workers at APT, Inc. The models
presented there have been generally borne out by experimental work per-
formed on scrubbers by IERL-RTP contractors and others. As noted in
the preceeding section, efficiency versus power consumption has been
found to be quite similar for conventional scrubber types, with two
interesting exceptions, low-energy scrubbers in series and TCA mobile-
bed scrubbers. Unconventional approaches, specifically flux force/
condensation scrubbing and electrostatic scrubbing, also produce high
efficiencies at atypically low power consumptions. In this section we
describe the operation of a spray scrubber in terms of a simple model
to indicate the promise of these two unconventional approaches. After
a discussion of FF/C and electrostatic scrubbing investigations, two
other subjects are focused on: demisting (entrainment separation) and
interfacial effects between particle and scrubber liquid, both poten-
tially important obstacles to high efficiency scrubbing.
15
-------
GENERAL COLLECTION MODEL
An approximate general formula for collection of particles of diameter
d by scrubbers (and electrostatic precipitators) can be shown to be:
Pt = n/nQ = exp ( - w A/Q) (1)
where: n/n = penetration, one minus the efficiency (1 - E)
. n = outlet number concentration;
n = inlet number concentration;
w = net particle velocity toward collecting surface, the
migration velocity;
A = total collection surface area;
Q = gas volume flow rate.
(This equation applies to a single particle size. For polydisperse aero-
sols, the equation should be integrated over the particle size distribu-
tion.) Collection efficiencies are high when the magnitude of the argu-
ment of the exponent is large. The migration velocity will depend upon
the collection mechanisms. For collection due to inertial impaction,
which depends on the inability of particles (due to their inertia) to
follow the flow streamlines around an obstacle such as a scrubber dro-
plet, the migration velocity is given by:
w = i\ (u - ud) A./A (2)
where: T\ = single droplet collection efficiency;
u = particle velocity;
u, = droplet velocity;
A, = droplet cross-sectional area.
Particles in a moving stream will impact upon a perpendicular surface
provided that the inertia of the particles is sufficient to overcome
16
-------
the drag exerted by the air stream deflected by the impaction surface.
The impaction process can be characterized by the impaction parameter
which is defined by the following expression:
C p v d
* - E - E
18
where: p = particle density;
v = gas velocity;
d = particle diameter;
u = gas viscosity;
D = obstacle diameter;
C = Cunningham correction factor defined empirically as;
C = 1 + 2.492 X/d +0.84 X/d exp (-0.435 d /X);
where: X is the molecular mean free path.
Various functions of the impaction parameter have been found, empirically,
to approximate the single droplet efficiency for impaction on spheres.
As the impaction parameter increases, so does collection efficiency.
A
In the Scrubber Handbook, the following equation is derived to corre-
late fractional penetration, with pressure drop, AP, in spray scrubbers:
0.61 C p dW f2
n/no = Pt = exp [ 1-f-* *J . (4)
where: AP = pressure drop, cm WC
3
pT = scrubbing liquid density, g/cm
L
y = gas viscosity, poise
17
-------
(The parameter f is determined' empirically; it includes effects such
as mean droplet/gas relative velocity, particle wettability, and venturi
throat liquid distribution. Usually f is between 0.5 and 0.25.)
Figure 3 shows how strongly penetration depends upon particle diameter
and venturi pressure drop, based upon equation (4). For a given par-
ticle size the penetration decreases rapidly with increased pressure
drop. For a given pressure drop, the penetration decreases even more
rapidly as particle size increases.
The collection efficiency curves in Figure 2 and 3 have generally been
borne out by experimental data obtained on conventional scrubbers and
novel scrubber types. The agreement serves as support for equation (4)
and analogous equations for other scrubber types presented in the Scrub-
ber Handbook. In turn, this indicates the predominant mechanism for
capture of particles larger than about 0.5 ym in scrubbers is impaction,
as postulated there. (For much smaller particles, Brownian diffusion
produces effective collection.) Equations (1) and (4) and their ana-
logues help clarify the ways in which scrubber efficiency might be
increased.
It is highly desirable to increase scrubber efficiency without increasing
the power consumption, a major cost component. Candidate methods for
increasing scrubber efficiency are; increase impaction or other types of
particle capture, remove interferences to capture, and prevent the escape of
scrubber liquid which contains particulate material or soluble solids.
Particle collection by impaction would be increased by increasing the
following (refer to equations (3) and (4):
Cunningham correction
Particle density
Particle size
Gas velocity
f factor.
18
-------
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.09
0.08
0.07
0.06
0.05
0.04
O 0.03
S
Of
in 0.02
Ul
o.
0.01
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
5 10 15 JO
30
INCHES H2O
40 50 60 70
eo
90 100 110 120 130 MO
0.001
10
I
I
I
I
I
I
VENTURI
PENETRATION.
PRESSURE DROP
5«I03
IOM01
15*10*
20»I05
25«IOJ
N/mz
Figure 3. Penetration calculated from a venturi scrubber model
as a function of pressure drop and particle aerodynamic
diameter (corrected version of Scrubber Handbook figure)
19
-------
Unfortunately, particle density is usually inconvenient to increase.
The Cunningham correction is also difficult to alter favorably, as it
decreases with increasing particle size. One can increase particle size
fairly easily by condensing water vapor on the particles, so this is an
attractive direction to go for improvement. From Figure 3 it can be
seen that dramatic changes can result from doubling the particle size.
Although increasing the gas velocity will improve impaction, it is costly
in power consumption, the power depending upon gas velocity linearly or
even more strongly. The f factor may be amenable to change through im-
proved wetting of particles or through improvements in flow profiles
and scrubbing liquid distribution. Collection efficiency due to im-_
paction can also be improved by decreasing the following:
Scrubbing droplet size
Gas viscosity.
The drawback to decreasing droplet size is that it takes more energy to
produce a small spray droplet than a big one (compare their surface free
energies), so that improved collection by impaction by this method has
an energy penalty. Gas viscosity does decrease with decreased tempera-
tures , so that cooling the gas can save energy and improve collection
efficiency. The most promising approaches to improved collection effi-
ciency due to impaction without major energy cost seem to be: increasing
the particle size by condensation of water vapor and increasing the f
factor, perhaps by improving wettability.
Having discussed the possibilities for increasing collection by impaction,
we turn our attention to other methods for improving particle capture
by the scrubbing liquid. The following mechanisms might be employed
to enhance collection by producing larger migration velocities toward
the liquid:
Electrostatic forces
Diffusiophoresis
Gravitational sedimentation
20
-------
Thermophoresls
Brownian diffusion
Electrical attraction between particle and scrubbing liquid will enhance
collection; it can be shown that this should be possible with very
little additional energy consumption, making this an attractive direction
for research and development. Diffusiophoresis is particle motion pro-
duced by vapor gradients, the most common being the motion toward or
away from a liquid surface due to condensation or evaporation. As will
be seen, diffusiophoresis due to condensation can be used to improve
scrubbing efficiency. Condensation also releases heat energy rather
than requiring an energy input. Sedimentation can only be effective
if the gravitational force has time to produce a settling distance which
is comparable in dimension to the collector geometry. For particles
smaller than about a micron, there is negligible gravitational settling.
For larger particles, the mechanism could be useful, but the residence
times in the control system would still have to be relatively long, re-
quiring large structures with high initial construction costs; thus,
sedimentation is not very promising. Thermophoresis is the motion of
particles in a temperature gradient due to greater transfer of momentum
from the gas on the hotter side than on the cooler side of the particles;
the force is relatively weak in comparison to other forces normally operant
in control contexts. Diffusion occurs due to Brownian motion; it increases
as particle size gets smaller and as gas temperature increases; to heat
the gas to improve diffusive deposition would carry a substantial energy
cost. As this discussion indicates, two promising methods of efficiency
improvement would be the addition of electrostatic forces and the use of
condensation to produce diffusiophoresis. (Recall that condensation also.
causes particle growth, thus improvement in impactive collection.)
Collection by the scrubber liquid is necessary but not' sufficient. No
matter how well the scrubbing liquid captures particulate material, the
liquid and its entrapped particulate matter must be prevented from
being passed to the atmosphere. This requires minimizing the formation
21
-------
of unwanted droplets and assuring the capture of droplets through the use
of mist eliminators, topics of relatively little prior investigation.
In the light of this discussion, it is understandable why IERL-RTP has
selected the following areas for research and development in scrubber
technology:.
Enhanced impaction by increasing particle size through
the use of water vapor condensation
Increased migration velocities by addition of dlffu-
siophoretic force through condensation
Increased migration velocities by addition of electro-
static forces
Higher f factor values, perhaps by improved wettability
Reduction of droplet-borne emissions by the improvement
of mist eliminators.
The logic of this course is to improve the major mechanism (impaction),
as feasible; increase collection by incorporating other mechanisms (diffu-
siophoretic and electrostatic), where practicable; diminish any inter-
ference with collection which might be due to wettability (the f factor);
and finally, prevent those particles which have been captured from es- '
caping to the air (entrainment minimization and mist elimination). Eacb
of these areas will be discussed in more detail in what follows.
ELECTROSTATIC AUGMENTATION
Electrostatic precipitators produce relatively high efficiency collec-
tion at relatively high capital expenditures and relatively low power
costs, in comparison with scrubbers and filters of comparable particle
collection efficiency. Augmenting the inertial impaction mechanism with
electrical attraction was viewed by IERL-RTP as a means for improving
scrubber efficiency with little additional power cost, and three major
22
-------
programs have been sponsored, toward this goal, one each at MIT, TRW, and
the University of Washington.
Equations for the various electrostatic forces which are important in
particle collection have been presented by Kraemer and Johnstone and
are also available in the classic work by Fuchs. Basically, these
forces are of two types:
Attraction or repulsion of a charged particle due to charge on
other particles or collectors (Coulomb force)
Attraction of uncharged bodies to charged bodies (dipole
force or image force)
Recently, Cooper evaluated these forces for reasonable values of the
relevant parameters, as part of a lERL-RTP-sponsored review of electro-
static augmentation, and found the Coulomb force to be quite clearly pre-
dominant, generally. The migration velocity, w, produced by these forces
is the product of the force and the particle mobility, available from
tables, such as those given by Fuchs. (Particle mobility is terminal
velocity per unit applied force.) The influence of such forces will be-
come important when the migration velocities they produce are comparable
to the product of the single target impaction efficiency and the relative
gas/droplet velocity.
8
Under a contract with IERL-RTP, Melcher and Sachar of MIT studied experi-
mentally and theoretically the various combinations of charged droplets
and particles which might be used in electrostatically augmented spray
scrubbers. They defined characteristic times for various particle and
droplet behavior, and from these characteristic times the collection
efficiency of such a system could be determined. They confirmed their
theoretical predictions experimentally.
Q
Lear, Krieve, and Cohen of TRW published the results of their IERL-RTP-
sponsored work with a novel control device, the Charged Droplet Scrubber.
The scrubber accelerates the scrubbing droplets by using electrostatic
23
-------
forces rather than hydraulic or aerodynamic forces. The droplets cap-
ture some of the particulate material by conventional impaction, etc.,
and the droplets charge other particles, those which come sufficiently
close to cause an electrical discharge from the charged droplets. The
droplets and the charged particles are then caught on the conductive
surfaces of the scrubber due to the electric field which is set up by
the high voltage nozzle and grounded surfaces.
The collection efficiencies for the laboratory-scale Charged Droplet
Scrubber were higher than those for conventional scrubbers of similar
power consumption. The work done to accelerate the droplets should be
very much the same for this kind of scrubber and for a conventional scrub-
ber; thus the power requirements would be much the same with the exception
of pump or fan efficiency factors versus the electrostatic spray nozzle
efficiency factors. The major source of improvement on conventional
scrubber technology would be the addition of considerable electrostatic
force, especially useful in the range 0.1 to 1 urn, where scrubbers nor-
mally have their efficiency minima.
Research at the University of Washington under the direction of Dr. Michael
Pilat had shown that charging particles and spray droplets to opposite
polarities can improve scrubber efficiency dramatically, especially in
the submicron range. Figure 4 is from an article by Pilat.. It shows
collection efficiency as a function of particle size for two conditions:
oppositely charged particles and droplets, and electrically neutral par-
ticles arid droplets. The power consumption was 0.5 hp/1000 acfm (0.8 kW/
o
(m /s)), quite low for a scrubber of high efficiency (90 percent) at 1 um.
IERL-RTP is sponsoring further work on this scrubber: the development and
3
testing of a 1000 acfm (0.47 m /s) model, with the testing specifically
directed toward answering the questions that would be necessary for con-
struction of a full-size device. Efficiencies, particle characteristics,
and gas characteristics will be determined as part of this program.
24
-------
100
80
60
40
«
20
Droplets and
particles
charged
oppositely
No charge
Liq./gas = 15.7 gal./1000 acf
Mean drop diameter by
number = 50 microns
Drop geometric stand, dev. => 1.9
.2 .4 .6 .81.0 2 4 6 8 10
Particle diameter (microns)
Figure 4. Particle collection efficiency of electrostatic
spray droplet scrubber as a function of par-
ticle size
10
In summary, the option of increasing scrubber efficiencies at moderate
power levels by the addition of electrical forces is being pursued in
a number of programs sponsored by Industrial Environmental Research
Laboratory-Research Triangle Park.
FLUX FORCE/CONDENSATION (FF/C)
In particle collection by wet scrubbers, the role of condensation effects
associated with the presence of vapor in the gas stream is presently the
subject of intensive investigation. The idea of using these effects,
however, is not new, dating from the turn of the century. These effects
augment the usual mechanisms of Brownian diffusion and inertial impaction.
When vapor condenses onto a scrubber droplet, a diffusiophoretic force
is exerted on particles in the direction toward the droplet surface.
This force has two sources: the net flow of water molecules to the
droplet, a "wind" of sorts called "Stephan flow," and a subtle component
due to the gradient of the water vapor and the difference between its
molecular weight and the molecular weight of air. When vapor condenses
25
-------
onto the particles, the particles become effectively larger and are
therefore more easily captured by settling or by inertial impaction.
These effects seem to be important for submicron particles. Still un-
certain is the importance of the thermophoretic force, associated with
heat flow and in the same direction as the heat flow: the particles are
driven toward the drop surface when the drop is colder than the gas.
The terminology "flux force" applies here to the forces of diffusio-
phoresis and thermophoresis. As opposed to conventional scrubbers, FF/C
scrubbers often require the introduction of steam, which increases their
expense of operation if the steam is to be purchased.
Available theoretical and experimental evidence is apparently not yet
conclusive regarding degree of improvement of collection efficiency by
condensation effects. Sparks and Pilat performed calculations of par-
ticle collection efficiency by droplets under conditions assuming that
(1) condensation, or (2) evaporation, or (3) neither, is occurring. The
mechanisms were assumed to be inertial impaction and diffusiophoresis.
Simultaneous condensation was shown to enhance, while simultaneous eva-
poration was shown to diminish, the collection by inertial impaction,
the enhancement - diminution effect being more pronounced for the smaller
particles. (Thermophoresis was assumed to be negligible). Lancaster and
12
Strauss concluded that diffusiophoresis is not significant in con-
ventional scrubbers in which steam is injected. On the other hand, Fuchs
13
and Kirsch showed that high condensation rates lead to significant
improvement. lERL-RTP-sponsored studies by APT, Inc. and Stanford Re-
search Institute showed that diffusiophoresis is significant in FF/C
scrubbers and in conventional scrubbers with steam added.. Collection is
also enhanced by additional stages, whereby particle enlargement can
contribute. (A list of major references is available in a .report by
APT, Inc.14)
In 1951, Shauer conducted experiments in which the flue gas from
the combustion of radioactive wastes was drawn into a steam exhauster
26
-------
and the resulting aerosol-steam mixture was expanded through a nozzle,
causing condensation of the aerosol particles. The nozzle system was
followed by a spray scrubber. Efficiencies up to 99.9 percent were
obtained for 0.3 ym DOP (di-octyl phthalate) particles, as opposed to
zero efficiency without steam addition. (Energy consumption was great,
however.) Schauer attributed the improvement to particle enlargement
by condensation.
The diffusiophoresis "flux force" mechanism is discussed in detail by
Waldmann and Schmitt. The existence of this mechanism is evident
18
from the experimental results of Lapple and Kamack (1955) and Semrau
19
et al. (1958). Semrau et al., for example, noted a large difference in
efficiency between wet scrubbers operating with hot and cold water sprays.
They suggested the differences could be caused by evaporation from the
hot water droplets, which would produce a diffusiophoretic force away
from the drop surface and therefore would result in reduced efficiency.
(This would correspond to "evaporation scrubbing" as opposed to "conden-
sation scrubbing").
The possible dominance of thermophoresis over diffusiophoresis under
20
some conditions is proposed by Slinn and Hales in contradiction to
21
others such as Goldsmith and May who argue that thennophoretic forces
should be negligible. Two kinds of experimental results seem to support
the latter: improved scrubbing under condensing conditions and the
formation of a "dust-free" space observable near the surface of an evap-
orating droplet.
In studies supported by IERL-RTP (Calvert et al. and Calvert and
22
Thaveri ), APT, Inc. showed that FF/C scrubber efficiency is insensitive
to particle size. Therefore FF/C scrubbing is potentially competitive
with high-energy scrubbers when high collection efficiencies for fine
(submicron) particles are required. In the APT, Inc. studies it was also
found that FF/C collection increases as the concentration of particles
27
-------
decreases. This can be understood on the basis that the available
moisture is shared by fewer particles, which thereby grow larger and
are collected more easily than otherwise. Thermophoresis was shown
to be of minor importance compared with diffusiophoresis and the effect
of particle enlargement by condensation.
Other studies (multiple plates, etc.) have been sponsored by IERL-RTP,
including those of APT, Inc. and of Stanford Research Institute.
In a study by APT, Inc. on FF/C (flux-force/condensation) scrubbing,
14
Calvert et al. showed that collection increases with the quantity of
water vapor condensed per unit quantity of gas, and that diffusiophoresis
was the major collection mechanism for a single-sieve-plate scrubber.
22
In a subsequent study of multi-stage collection, Calvert and Jhaveri
found collection was much greater than on single-stage devices; this is
presumably due mainly to effects whereby particles enlarged by conden-
sation can be collected by downstream stages. Effects of water vapor
condensation and evaporation in conventional scrubbers (cocurrent-contact
orifice type) were studied by Semrau and Witham. At a given pressure
drop, condensation of substantial amounts of water was found to increase
the' collection, whereas the opposite effect was achieved by vaporization.
This implies that diffusiophoresis is significant. Particle growth ef-.
fects were minor in this single-stage scrubber.
ENTRAINMENT SEPARATOR TECHNOLOGY
All parti'culate control scrubbing involves the energetic interaction of the
gas stream and the scrubbing liquid, usually water. Often this means that
entrained water droplets containing contaminants will exit from the
scrubber unless some means of demisting, or entrainment separation, is
employed. The dried residue from these water droplets form an air pollu-
tion source. (In the novel control device evaluation program of IERL-RTP
there have occasionally been observed net increases in particulate mat-
ter in certain particle size ranges, thought to be due to evaporation of
scrubbing liquid containing dissolved solids, such as hard water.)
28
-------
Increased emphasis on entraihment separation has accompanied the develop-
ment of scrubbers which remove SO with drops containing limestone or
other reactants. Emission of these droplets to the atmosphere would,
produce substantial quantities of particulate pollution, for example
in the use of limestone scrubbing to remove S0_ from flue gases.
23
Nannen et al. indicated that for a packed tower scrubber there would
be a liquid-to-gas ratio of about 50 gal/1000 ft3 or 6.7 x 10~3 m3/m3
and the liquid would be about 10 percent solids by weight. Thus the
o
ratio of scrubbing solids to air volume would be would be about 670 g/m
o o
or 290 gr/ft , and the escape of only 10 of this solid material as
part of droplets would produce heavy emissions. It is desirable both to
minimize reentrainment and to use mist eliminators to capture the liquid
which is entrained. Demisters are effective for drop sizes above a few
microns in diameter. Their effectiveness may be assessed in terms of
three criteria:
e Primary collection efficiency (should be high)
Pressure drop (should be low)
Reentrainment rate (should be very low)
Primary collection of mist exceeds total collection by the amount of the
reentrainment, which normally does not occur at low-gas velocities. There
is a critical value of gas velocity at which reentrainment occurs, and
above which the total collection efficiency is thereby reduced.
24
Until recent work by APT, Inc. under IERL-RTP support, little was known
about entrainment separator technology, in particular the reentrainment
rate. In the scrubber industry 100 percent overall efficiency has been
generally assumed for entrainraent separators.
The APT, Inc. work included theory for primary collection efficiency,
pressure drop, and reentrainment in baffles and cyclones. This
work also included pilot-plant experiments with various types of separa-
tors. The theoretical models generally agreed with experimental data,
29
-------
except that the reentrainment theory was found to underestimate the ob-
served reentrainment. Above the critical velocities, reentrainment val-
ues about 0.5 to 1.0 percent were observed, and the minimum drop size
25
in the reentrainment was 40 pm, although Statnick and Drehmel, among
others, have reported capturing droplets on cascade impactor stages
with size cut-offs on the order of micrometers, which may be due to en-
trainment or, more likely, to condensation. The entrainment separator
is now considered an important part of any scrubber system, representing
significant capital and operating costs. Contributing to these costs is
the fact that the entrainment separators must be operated at low velo-
cities, leading to large sizes and therefore large costs.
Such work is to be carried further by APT, Inc., under sponsorship by
IERL-RTP, in ascertaining the characteristics of the entrainment from a
mobile bed scrubber (TCA), similar to those found to be unusually
energy-efficient in producing high efficiency control of fine particu-
lates. Thus, two important areas of Industrial Environmental Research
Laboratory-Research Triangle Park research are merged.
EFFECTS OF INTERFACIAL PROPERTIES ON SCRUBBING
Up to this point we have discussed mechanisms such as impaction, elec-
trostatic migration, and diffusiophoresis for bringing particles into
contact with the scrubbing liquid. Contact and capture are not synony-
mous, however. Particles may rebound from the surface, and difficulties
observed in some scrubber applications have been ascribed to the non-
wettability of the particles by the droplets. The importance of surface
properties in scrubbing is a matter of scientific debate, some of which
we will present here, but IERL-RTP hopes to provide definitive answers
through its recent sponsorship of a new investigation of these effects.
Wetting, a seemingly simple process, is actually a subtle physico-
26
chemical interaction. According to Berg et al. the wetting of solids
30
-------
by water consists of the formation of hydrogen bonds between ILo mole-
cules in the droplets and OH on the surface of the solid. The rate at
which the process proceeds can be critical, and it depends upon many
factors. The removal of particulate material from gases by moving
drops has been of considerable interest in the contexts of scrubbing and
of the precipitation scavenging from the atmosphere of radioactive debris,
in which context Engelmann described the effects of wettability on par-
27
tide raindrop collisions. Nonretention of particles can be due to
(a) rebound from the drop surface, (b) rebound from another particle
already on the surface, or (c) reentrainment by the drag of the air.
There are two somewhat different cases: liquid-solid and liquid-liquid
contact.
Coalescence of water drops with one another has been studied for almost
28
a century, beginning with the work of Rayleigh. However, it has been
apparently much more difficult to obtain reliable experimental data on
the retention of solid particles (dust, for example) by water drops. In
29
recent experimental work reported by Fuchs et al., dust particles
(quartz, carborundum, and coal) 5 Jim to 80 ^m in diameter were observed
to be almost entirely retained upon contact with a water drop 1 to 2 mm
in diameter, at relative velocities of 1 to 6 m/s. However, these
results must be considered tentative; there is room for doubt regard-
ing the high degree of retention because it was not certain that all
particle rebounds were observable, particularly at the smaller particle
sizes.
29
It was also observed by Fuchs et al. that the captured particles remain
on the surface when the drop is pure water, but penetrate into the
interior when the drop is an aqueous solution of surfactant. The effects
of wetting seemed to become important only at small relative impact velo-
cities. Wettability of a particle is sometimes assumed to be the same
as the wettability of the solid bulk material, which is defined in terms
of the equilibrium contact angle between a water drop and a plane surface
31
-------
of the material. The interaction between a particle and a drop depends
on the dynamic rather than the static surface tension because the liquid
surface undergoes rapid extension during the collision, according to
Fuchs.6
There are other difficulties, such as the fact that minute amount of
impurities radically change the surface properties. The adhesion be-
tween a drop and a solid surface is different for a dry surface and
for a surface which has previously been wetted, even for a very short
time; a layer of grease 01
contact angle significantly.
time; a layer of grease only one molecule thick can increase the
28
In Rayleigh's experiments in 1879, colliding water drops were observed
to bounce apart in the absence of electric charge; that is, charge was
found by him to be essential for coalescence between water drops. This
31
was confirmed by Berg, who found that the time required for coalescence
of water drops, and for the wetting of solid particles, depends on elec-
trical potentials.
26
In later work, Berg et al. showed that:
1. Wetting is a rate process; if the rate is not large
enough, colliding solid particles and water drops
bounce apart.
2. A solid salt that is hydrated such as to have a
saturated solution at the surface is wetted at
the same rate as that of two colliding water drops;
this rate is determined by the rate at which hydro-
gen bonds are reoriented across the interface, and
this requires an orienting electric field, that is,
charged colliding partners.
3. A solid covered with OH is more rapidly wetted than
one covered with H20; this is because reorientation
of bonds is required on the part of both partners
in the latter case, as opposed to the water drop
alone in the former case (electric charge required).
32
-------
32 33
In theoretical treatments to date (McCully et al., Pemberton,
34
McDonald ), wetting is assumed to be associated with the mechanical
problem of the deceleration of the impacting particle by the drop
33
surface tension. Pemberton equated the particle kinetic energy to
the work necessary to overcome surface tension and penetrate to the
drop interior. He found a minimum velocity u for penetration for a
completely unwettable particle:
. 1/2
where T is surface tension, and p and d are the particle density and
34 P
diameter, respectively. McDonald extended this theory to arbi-
trary contact angles (equilibrium or static), and considered the condi-
tions under which the particle would "shoot through" the drop. These
theories do not consider the dynamic surface tension as suggested by
Fuchs.
29
Fuchs et al. concluded from their observation of captured un-
wetted particles remaining on the surface that Pemberton1s concept is
unrealistic: namely, in order to be captured, particles must be completely
26
engulfed by the drop. On the other hand, the work of Berg et al.
shows that charge is probably present, and this would.explain the ob-
29
servations of Fuchs et al. . as well as tl
unwetted particles remain on the surface.
29 32
servations of Fuchs et al. . as well as those of McCully et al. that
Because of the remaining questions related to wettability, etc., IERL-RTP
has initiated a program for determining interfacial property effects on
scrubbing.
SUMMARY
Scrubber technology seems to have reached the practical limits of per-
formance for the use of inertial impaction. (Possible exceptions may be
33
-------
represented by the mobile bed scrubber, TCA, and low energy scrubbers
used in series.) Attention has turned to improving scrubber performance
by introducing new forces (electrostatic, diffusiophoretic) or by in-
creasing impaction and interception by using condensation to increase
the physical size of the particles being scrubbed. Removing the en-
trained water from scrubber exhaust is needed to achieve maximum bene-
fits from the control system. In some cases, nonwettability of the par-
ticulate matter may cause a scrubber system to work much less effectively
than predicted. All of these subjects are being covered or have been
covered by the Industrial Environmental Research Laboratory-Research
Triangle Park scrubber program.
More detail on the specific projects in the IERL-RTP scrubber program
are provided in the next section, which presents a chronological over-
view of the program.
34
-------
SECTION V
CHRONOLOGICAL OVERVIEW
This section describes the projects sponsored by Industrial Environmental
Research Laboratory-Research Triangle Park in chronological order, by
starting date. Where our information was sufficient, we have followed
our introductory remarks for each study with material describing its
goals, methods, results, conclusions and recommendations.
WET SCRUBBER SYSTEMS STUDY (APT, INC., 1970)
Introduction
IERL-RTP contracted with Ambient-Purification Technology, Inc. (now
APT, Inc.) to perform a study and review of scrubber technology. The
study was to yield a thoroughly documented engineering handbook (Scrub-
4
ber Handbook) suitable for. scrubber design, which would also aid future
development by setting forth the state of the art.
Goals
As presented in the Scrubber Handbook, the goals of the APT, Inc.
study were:
1. Evaluate current engineering technology
2. Evaluate existing scrubber systems
3. Investigate present usage problems
4. Determine potential new applications
5. Develop specific research recommendations
35
-------
Methods
The study involved analysis, but not new experimental work. The pro-
ducts of the review were: (1) the Scrubber Handbook, (2) Final Report,
(3) Bibliography, and (4) Research and Development Plan. The handbook
involved both selection from the literature and development of design
criteria and theoretical and empirical relationships for wet scrubbers
intended to cleanse both particulates and gases from the effluents of
stationary sources, with emphasis on particulate scrubbing. Much in-
formation on scrubbers was gathered by questionnaires submitted to
scrubber manufacturers and actual on-site observation of scrubber
operations. Compilation of the bibliography was assisted by computer
search. The Scrubber Handbook treated scrubbers by analyzing "unit
mechanisms" for collection of particulates and gases, a significant con-
tribution to analysis. Particular scrubbers were then analyzed as being
dependent upon a group of unit mechanisms for collection. For collection
of particles, the unit mechanisms considered were: inertial collection,
interception, diffusion, electrostatic migration, diffusiophoresis,
thermophoresis, and condensation. A method was developed for calculating
the penetration for a polydisperse aerosol conveniently. For gases the
unit mechanisms were: transfer to drops, transfer to liquid films, and
transfer within bubbles.
Results
i
The results of the study are summarized by a portion of the abstract
to the Scrubber Handbook; "The Scrubber Handbook brings together
previously scattered material and makes clear its applicability
to scrubber technology. It discusses the various aspects of scrubber
use and presents engineering design methods based on a unifying concept.
Actual experience on hundreds of scrubber installations is presented in
a condensed form. Many related topics such as auxiliaries, cost estima-
tion and optimization techniques, and the disposal of liquid and solid
36
-------
wastes are all covered " The bibliography contains about 1,700
references discussing various aspects of wet scrubbers.
The subject areas and organization of the Scrubber Handbook are:
Chapter 1 - Introduction
Chapter 2 - Guide to the Handbook
Chapter 3 - Scrubbers Available
Chapter 4. - Basic Concepts
Chapter 5 - Design Methods
Chapter 6 - Auxiliaries
Chapter 7 - Scrubber Performance on Industrial Emissions
Chapter 8 - System Analysis, Costs, and Optimization
Chapter 9 - Design Examples
Chapter 10 - Physical and Chemical Data1
Chapter 11 - Materials Data
Chapter 12 - Liquid and Solid Waste Disposal
Conclusions and Recommendations
The Final Report of the study presents two R & D plans; one at a $2
million, 5-year level, and the second at a $7 million, 5-year level.
The major objectives of the R&D effort proposal were:
1. Basic Concepts Research - Experimental verification of par-
ticle collection mechanisms theory; study of mechanisms of
agglomeration; experimental study of mass transfer coeffi-
cients in two-phase flow; experimental and theoretical
study of interfacial areas.
2. Physico-Chemical Data - Study of reacting solutions and
slurries to determine reliable gas- and liquid-phase dif-
fusivities for SO- and other pollutants.
3. Design of New Equipment - Utilization of additional forces
and agglomeration; use of smaller scrubbers and higher gas
velocities; alteration of liquid properties to increase
collection efficiency.
37
-------
4. Engineering Design-and Optimization of Scrubber Systems-
Development of standard methods for evaluating wet scrub-
bers; formulation of collection efficiency equations for
multiple collection mechanisms; development of calcula-
tion method for converting collection efficiency data from
hygroscopic to nonhygroscopic particles; optimization
of scrubber systems.
5. Dynamic Behavior and Control Instrumentation Technique -
Theoretical and experimental study of scrubber system
response to start up, shut down, and overload surges.
FEASIBILITY OF FLUX FORCE/CONCENTRATION SCRUBBING FOR
FINE PARTICIPATE COLLECTION (APT, INC.)
Introduction
This work represents a follow-up to the preceding scrubber study by
APT. In the previous work it was learned that, in fine particle col-
lection by liquid drops, collection by the usual well-known mechanisms
(such as Brownian diffusion and inertial impaction) is aided by addi-
tional less well-known mechanisms, "flux-force" and condensation
effects. The term "flux-force" is defined to include thermophoresis,
diffusiophoresis (and Stefan flow), and photophoresis, all of which
result in body forces acting on particles; of interest are conditions
under which these forces drive the particles toward the collector
drop surface. One condensation effect is to cause particles to become
enlarged, by the accretion of liquid due to vapor condensation; the en-
larged particles are subsequently more easily collected by impaction
and interception.
The phenomena of flux forces and of condensation enlargement had been
observed earlier, as discussed in Section IV herein. Simple theories
were available to describe them individually in simple cases.. There
did not exist, however, a quantitative model which (1) describes the
simultaneous action of phoretic forces and condensation enlargement
(the combination denoted by the symbol FF/C), and (2) predicts particle
38
-------
collection efficiency in practical scrubber systems incorporating FF/C.
Hence, the study was undertaken to determine the feasibility of FF/C
scrubbing. Its scope was restricted to the specific phenomena of diffu-
siophoresis, thermophpresis, and particle enlargement by condensation.
Goals
The purpose of the study was to determine whether FF/C effects are
sufficiently important to warrant further development, and, if so, to
14
recommend a program of further research. The main tasks included:
1. Review of existing theory on FF/C effects (see Scrubber
Handbook, for example) and extension thereof toward
developing engineering design methods suitable for eval-
uation of practical scrubber systems.
2. An experimental study to explore and evaluate the impor-
tant features of FF/C scrubbing.
3. Preliminary engineering and cost analyses of promising
practical cases.
4. Recommendations for future FF/C scrubber system development.
Methods
A literature review was carried out to obtain available background in-
formation on FF/C scrubbing. Engineering design involved analyses of
several factors: particle growth; energy balance, including latent
heat (to .determine the temperature) for the liquid interface; the
rate of vapor condensation (on particles and on collector surface);
and particle deposition rates. Based on available theories for depo-
sition velocities and for particle eniargment by condensation, mathe-
matical models (i.e., design equations) were developed for spray,
sieve-plate, impinging-jet, and liquid-sheet (wetted-wall or packed
bed) scrubbers.
39
-------
Theory - In the APT, Inc. study a mathematical model was developed for
sieve plate scrubbers as well as for other types of scrubbers, and experi-
ments were performed on a sieve plate scrubber. The sieve plate scrub-
ber is an alternative to scrubbers that use droplets. A bubble layer
on a plate represents a large collecting surface area. The principal
mechanism for gas/liquid contacting involves bubble formation and rise
through the liquid. The mathematical model involves:
1. Heat and mass transfer between bubbles and liquid.
2. Heat and mass transfer between bubbles and particles
suspended in the bubbles.
3. Particle growth due to condensation.
4. Particle deposition by (a) diffusiophoresis, (b) thermophoresis,
(c) centrifugation during bubble rise.
The temperature of the liquid interface depends on the gas/liquid heat
and mass transfer coefficients, the liquid-phase heat transfer coeffi-
cient, the nature of the interfacial area, and the flow patterns of
the two major phases. Simple models may be inaccurate because they
do not account for the manner in which local gas velocity and overall
gas flow patterns influence both local transfer coefficients and bulk
conditions for the phases in contact in any region. The latter are
also influenced by the liquid-phase hydrodynamics, which is very
complicated in bubbling systems. In plate-type scrubbers, particles
are collected during bubble formation, and this becomes important
after particle growth takes place. This effect is a function of foam
density, as well as particle properties, plate geometry, and gas
velocity. When vapor is condensing, the foam characteristics are
changed from normal (noncondensing) conditions.
Experiment - Based on theory, APT decided to verify assumed transfer
coefficients in an experiment using a sieve-plate scrubber. The par-
ticular transfer coefficients needed for predicting penetration under
40
-------
given experimental conditions are the gas-phase and liquid-phase heat
transfer coefficients, and the gas-phase mass transfer coefficient.
The major components of the experimental apparatus consisted of:
1. A test section with an aluminum sieve plate, of diameter
10.2 cm and thickness 1.6 mm, having 30 perforations
4.8 mm in diameter adding up to 6.6 percent free area.
2. An air flow system, controlling air flow rate (up to
0.71 m-Vmin and temperature), water vapor, and aerosol
particle concentration (up to 106/cm3).
3. A particle generator, using a Collison atomizer with di-
butyl-phthalate (DBF), an impactor, and an evaporation-
condensation generator. (Particle diameters were in the
range 0.5 urn to 1.0 ym.)
4, A particulate sampling system, using filters, a Climet
photocounter, and an eight-stage Andersen cascade impactor.
The object of the experiments was to infer values of the transfer coeffi-
cients required by the mathematical model.
In the experiments, the liquid-to-gas-flow-rate ratio (L/G) was kept
"3
high (average value = 7.7 Jl/m ) in order to maintain a low liquid tempera-
ture on the sieve plate, and therefore, a high diffusiophoretic force
due to vapor condensation of the bubble wall. The collection was attrib-
uted to the combined effects of:
1. Impaction during bubble formation
2. Diffusiophoresis
3. Thermophoresis
4. Diffusion
5. Particle condensation-enlargement with centrifugal force
The diffusiophoretic collection was found to depend 6n the amount of
vapor condensed per unit volume of gas, and the collection is higher
for higher water vapor concentrations, at a fixed temperature.
41
-------
Comparisons were made between the predicted and the measured values for
heat and mass transfer coefficients. The predicted value of the liquid-
phase heat transfer coefficient was of the order of 100 times larger than
the empirically determined value. This casts doubt on the assumptions
regarding the bubble area available for transfer, the bubble diameter, the
liquid temperature, the liquid surface renewal time, and the uniformity of
the foam density.
Results
One result of the work was the finding that diffusiophoresis was the
major collection mechanism for a single plate scrubber. Diffusiophoresis
depends on the amount of vapor condensed, which can be expressed as the
parameter (y. - y ), or moles of vapor condensed per mole of dry air.
Figure 5 shows experimental penetration, Ft", as a function of this
parameter. (The six data points are labeled by the run numbers.) The
variation of Pt is due to the diffusiophoresis, while the contributions
of the other mechanisms (impaction during bubble formation, thermo-
phoresis, diffusion, and condensation growth) are approximately the
same for all the runs. In the figure, PtF refers to impaction during
bubble formation. Particle enlargement by condensation was an unimpor-
tant effect, probably because the DBF particles were not very wettable.
Another reason is that this was a single stage device. Subsequent work
has shown this enlargement to be significant for multi-stage devices:
particles "grown" in the first stage are more easily caught in later
stages. It was found that the theoretical model satisfactorily predicts
the general form and magnitude of the experimental results if particular
values are used for the heat and mass transfer coefficients. Theoretical
values of the transfer coefficients obtained using penetration theory
were found to agree with the corresponding experimental values for the
gas-phase heat and mass transfer, but not for the liquid-phase heat
transfer, as noted above. It was thereby inferred that there may be
42
-------
0.01
0.02
0.03
0.04
0.05
0.06
(y±-ye),
mole vapor condensed
mole dry air
Figure 5. Particle penetration versus water vapor.condensed
-------
errors in the assumptions regarding the bubble area available for trans-
fer, the bubble diameter, the liquid temperature, the liquid surface
renewal time, and/or the consistency of the foam density. These results
were felt to indicate a need for further experimental and analytical
work. Moreover, the predicted particle collection values were found
to be higher than the experimental inferred values.
Conclusions
Calvert, et al. reached the following conclusions:^
"Overall, the comparison of experimental data with predictions based on
our mathematical model shows that the model is capable of giving use-
ful and realistic results. While additional refinement... is needed,
the model can account for what is observed experimentally and gives
us a good tool for interpreting what we see and utilizing this know-
ledge for engineering design of practical equipment ....
"The primary objective of investigating the feasibility of FF/C
scrubbing has been achieved with a clearly affirmative result. It has
been shown in the present program that FF/C scrubbing can remove fine
particles at high efficiency and that, within some limitations, it is
more economical than other means of particle collection. Mathematical
modeling of the many simultaneous phenomena taking place in an FF/C
scrubber has been accomplished for some important unit mechanisms,
although 'several coefficients remain to be fitted to experimental data.
"In accordance with the original objective of selecting and performing
a brief exploratory experiment of significant nature, our experimental
work extended over a narrow range of conditions. Within this range
the agreement between theoretical predictions and experimental results
for particle collection is fairly good, once the heat and mass trans-
fer coefficients are evaluated from experimental data. However, des-
pite the good agreement, theoretical predictions of collection '
44
-------
efficiency for plates are consistently higher than the experimental re-
sults. This, along with the discrepancy between computed and predicted
liquid-phase heat transfer coefficients, indicates that the mathematical
model for bubbles should be revised."
Recommendations
To achieve the objective of building a working FF/C scrubber, Calvert,
et al. suggested further developmental investigations.
They recommended completing the experimental evaluations and develop-
ment of the theoretical models for bubbles, sheets, and drops. In
follow-on work, they proposed that attention should be given to:
1. Better estimates of heat and mass transfer areas
2. Better estimates of liquid surface temperature
3. Possibly unusual sieve plate design and operating
conditions
4. Different bubble shapes and gas dynamics, causing different
transfer rates and particle deposition by centrifugal
force
5. Measurement problems that may cause errors.
New experiments for testing assumptions in the theory of bubbles and
sheets should provide:
1. Test particles having a range of surface properties, and
high vapor pressures to minimize vaporization up to 80°C
gas temperatures-
2. Wettable particles to study the effects of growth, at
relatively low saturations
3. Large gradients in water vapor concentrations simultan-
eously with low saturation (high gas and liquid tempera-
tures, with nonwettable particles) to avoid condensation
on the particles
45
-------
4. Sieve plates with customary dimensions for mass trans-
fer, at usual gas velocities
5. Packed columns to validate the "liquid sheets" model (to
provide a more definable transfer surface area and liquid
renewal time).
Calvert, et al. also suggested defining and developing an optimal FF/C
scrubber, and field testing a pilot-scale FF/C scrubber.
WET SCRUBBER DEVELOPMENT II (FLUX FORCE/CONDENSATION (FF/C)
SCRUBBERS) (APT, INC., 1973-1975)
This study brings the FF/C scrubber to the pilot plant development
stage after earlier theoretical and experimental analysis (also under
IERL-RTP contract) showed the feasibility of flux-force/condensation
mechanisms in controlling fine particle emissions. The task established
for APT, Inc. by the IERL-RTP project description was
"APT will design and fabricate two pilot scale FF/C scrubbers
large enough for the exploration of scale-up problems. These
devices will be used in a lab pilot experimental program to
determine applicability to fine particles, develop design
equations, determine optimum conditions, investigate poten-
tial operational and maintenance problems, and determine
the sensitivity to particle size. Based on data collected,
APT will prepare a revised engineering and cost analysis, and
will submit a detailed industrial pilot test program."
The methods used by APT, Inc. were similar to their earlier work.
Pilot plant studies confirmed the feasibility of FF/C scrubbing and
showed that the FF/C effect on multi-stage collectors was much greater
than on single-stage devices.
Results
Comparisons were made between FF/C scrubbing data obtained in the APT
investigation under various conditions and scrubbing data of Rozen and
46
-------
Kostin for particles of 1- pm aerodynamic diameter. Figure 6 and Ta-
ble 4, taken from a draft final report submitted by APT, Inc. to IERL-RTP,
portrays these comparisons. Figure 6 shows particle penetration versus
condensation ratio (or injection ratio) q?, defined as mass of water
vapor condensed per mass of dry air. As Table 4 shows, the various
curves in Figure 6 compare the effects of:
Four sieve plates versus five sieve plates,
e 0.3 ym versus 0.5 yra particles,
Titanium oxide versus black iron oxide particles,
Initial concentration, n in the range 3 x 105 to
7 x 109/cm3. x
35
The data of Rozen and Kostin, for 0.3 urn oil particles in alternate
coo
hot and cold sieve plates, and n = 10 - 10 /cm , are included for
comparison.
Conclusions
The following conclusions may be drawn:
It is apparent from Figure 6 that collection increases in
all cases with increasing q1.
Five plates are significantly more effective than four
(probably due mainly to particle enlargement effects).
Scrubber performance is insensitive to the L/G ratio.
(However, a lower L/G ratio with recirculated liquid re-
quires a larger cooling range in the liquid cooling
system.)
Scrubber performance is improved significantly when the
vapor condensation is distributed over two sections of
the multiple-plate scrubber. This_is also consistent
with results of Rozen and Kostin.
Collection decreases with increasing inlet concentra-
tion, n . (This shows the importance of condensation
enlargement, and of competition for the available vapor.)
47
-------
i.o
0.5
z
O
H
O
<
cc
u.
z
O
oc O.I
UJ
z
UJ
Q.
UJ
O
(T
2
0.05
0.01
i r
\ NOTE: NUMBERS ON CURVES
\ CORRESPOND TO
REFERENCES IN
TABLE 4
T-
O REFERENCE 7
0.01 ' 0.05 O.I 0.5 1.0 5.0
CONDENSATION OR INJECTION RATIO , q* (g woter vapor/g dry gas)
Figure 6. FF/C scrubber performance comparison (APT, Inc.)
-------
Table 4. FF/C SCRUBBER PERFORMANCE COMPARISON
Figure
ref.
1
2
3
4
5
6
7
Experimental
studies
Rozen and
Kostin35
Present
investigation0
Present c
investigation
Present
investigation0
Present
investigation
Present
investigation
Present
investigation0
Scrubber
type
Alternate hot .
and cold sieve
plates
Five sieve
plates
L = 0.64 A/sec
Four sieve
plates
L = 0.64 8,/sec
Four sieve
plates
L = 0.64 A/sec
Four sieve
plates
L = 0.38 A/sec
Five sieve
plates
L - 0.38 fc/sec
Five sieve
plates
L .= 0.38 A/sec
Steam intro-
duction
distributed
under two
plates
d a
Pg
(urn)
0.3
0.55
0.55
0.5
0.5
0.5
0.5
Particle
material
Oil
Black
Iron
Oxide
Black
Iron
Oxide
Titanium
Dioxide
Titanium
Dioxide
Titanium
Dioxide
Titanium
Dioxide
b
ni '
(cm'3)
105 - 108
3 x 105
3.6 x 105
3 x 106
-7.2 x 109
2.4 x 107
-8.8 x 108
5 x 105
-1.7 x JLO7
3 x 105
-3.4 x 106
Particle mass median diameter.
Initial particle number concentration.
"APT, Inc. "Wet Scrubber Development II."
49
-------
FINE PARTICLE SCRUBBER PERFORMANCE TESTS (APT, INC. 1974)
Introduction
Testing of the fine particle collection efficiency of actual scrubber
installations was undertaken by APT under contract with IERL-RTP. The
program was designed to fulfill the need for more reliable data on col-
lection efficiency than previously available, in order to design better
scrubbers and to predict their performance. Inadequate methodology,
undefined parameters, and insufficient quantity were cited by APT as
being responsible for the shortcomings of previous scrubber testing,
failings which their study attempted to rectify.
Goals
As presented in the Final Report, the major objectives of the study
were:
1. Obtain data on fine particle (< 2 to 3 urn) collection
efficiency as a function of particle size for scrubbers
operating on representative industrial emission sources.
Record pertinent data on scrubber design and operating
conditions.
2. Reconcile the performance data with existing mathemati-
cal models, such as those presented in the Scrubber
Handbook by Calvert, et al.^. Where necessary
and possible, develop better models and/or design
approaches.
3. Obtain data on scrubber system costs for investment,
operating, and maintenance.
4. Compile available information on scrubber operating char-
acteristics and problems (e.g., entrainment), maintenance
requirements, corrosion and erosion experience, and simi-
lar operational problems.
50
-------
Methods
The experimental procedures employed for measuring performance were:
1. Determine gas velocity distribution at scrubber inlet and
outlet by pitot tube multi-point traverse
2. Measure inlet and outlet temperatures with a thermocouple
or dial thermometer
3. Measure pressure at inlet and outlet with manometer and
static pressure tube
4. Obtain wet- and dry-bulb temperature measurements on sample
withdrawn from stack to measure humidity
5. Measure dry gas density with pycnometer
6. Measure liquid entrainment by measuring liquid collected
in precutter upstream of cascade impactor. Alternate
measurement technique: dye-treated paper for drop size
determination.
7. Determine size of submicron particles with diffusion
battery, occasionally for particles as small as 0.01 um.
For particle sampling and size analysis, three types of cascade impac-
tors were used:
1. Andersen "viable" sampler (500 cnrVss) for sizing samples
withdrawn from the stack, calibrated by polystyrene
latex particles and a Climet particle counter. Glass fiber
filter used after the impactor.
2. Brink cascade impactor (100 cm /s) sampling both in- and
ex-stack. Same calibration as above. Glass fiber filter
used after last stage.
3. University of Washington (Pilat) Mark III Cascade Impac-
tor (500 cm3/s) for in-stack sampling. Manufacturer's
calibration used. Glass fiber filter placed after last
stage.
Isokinetic sampling was employed throughout the test program. A
round jet impactor having a cut diameter larger than 5 um was used as
a precutter to remove large particles from inlet samples and to remove
51
-------
entrained liquid from outlet samples. The parameter used to define
the scrubber performance was the particle aerodynamic diameter for
which collection efficiency is 50 percent, the "cut diameter."
Results
The scrubbers tested and the experimentally measured cut diameters for
50 percent penetration were:
Approximate
aerodynamic cut
Scrubber type diameter, urn
Valve tray 1.2
Vaned centrifugal 1.2
impingement plate 1.0
Wetted fiber 0.8
Venturi 0.7
Mobile bed (TCA) 0.4
Venturi rod 0.3
Extensive graphical and tabular material is presented in the Final
0£
Report to indicate the raw and partially reduced data used to arrive
at these results.
Conclusions
1. Arrangement of scrubber tests requires judgment in deciding
on the best compromises which will yield valid data.
2. Uncertainty in penetration determinations were _>_+ 10 per-
cent at a given diameter.
3. Comparison of experimental results with mathematical models
was successful in all cases but the mobile bed scrubber
(TCA). Theoretical relationship between pressure drop (or
power consumption per unit flow rate) and cut diameter
showed good correlation with test data, except for the TCA.
-------
4. Experimental methods need to be streamlined for better
accuracy and convenience.
5. Data on capital investment, operating costs, and mainte-
nance generally were not available or in reliable form.
Recommendations
Additional scrubber tests were recommended for:
1. Mobile bed scrubbers on variety of sources
2. Preformed spray scrubbers
3. Venturi scrubbers at high pressure drop, with nonwettable
particles and on large gas flows
4. Plate-type scrubbers on systems without condensation
effects and with nom^ettable particles
5. Impingement and entrainment scrubbers
6. Wet fan scrubbers
Improved experimental methods were deemed necessary:
1. Better impactor catch weighing
2. More convenient and reliable diffusion battery
3. Real-time particle size and concentration analysis
4. Aerosol dilution system (for particle counters and
diffusion batteries)
5. Particle density measurement
6. Opacity measurement for saturated gas stream
Model development should be furthered with respect to:
1. Mobile bed
2. Particle growth prediction for soluble materials in
near-saturated gas
3. Effect of adsorbed gases on particle growth
.53
-------
ENTRAINMENT SEPARATORS FOR SCRUBBERS (APT, INC., SEPTEMBER 1972)
Introduction
Scrubbers of gases and particulates usually produce entrained water
droplets that have to be removed by entrainment separators if the net
emissions to the atmosphere are to be minimized. Spray-type particu-
late scrubbers will have the captured material contained on or in the
droplet, which clearly must be caught. The use of, for example, lime-
stone slurries to remove SO- introduces particulate matter into the
spray, making it imperative to use demisters, another word for entrain-
ment separators. Calvert, et al. at APT, Inc. undertook a detailed
examination of the state of the art in entrainment separators, the
- 24
interim report for which was issued in October 1974.
Goals
The scope of work defined by IERL-RTP for Contract 68-02-0637 was, in
essence:
1. Evaluation of demister technology, including the existing
data and theory with an emphasis on identifying problem
areas
2. Experimental investigation of the performance of the
various types of entrainment separators, studying the
efficiency, pressure drop, reentrainment, plugging,
face velocity, etc.
3. Development of improved methods and equations for the
selection and design of entrainment separators
4. Pilot test of promising new designs
5. Recommendations concerning research and development for
improving demisters .
54
-------
Methods
A literature survey was conducted, directed toward the subjects of pri-
mary collection efficiency, pressure drop, reentrainraent, problems of
entrainment separators, and detailed information on presently-used
separators. This included visits to libraries, a computer search in
the APT library, contacts with manufacturers, and visits to EPA and
TVA facilities.
In the theoretical part of the program, equations.: predicting primary
collection efficiency and pressure drop were acquired for various
types of entrainment separators, including cyclones, packed beds, tube
banks, knitted mesh, and sieve plates. Mathematical models for pri-
mary efficiency and pressure drop in a zigzag baffle separator were
also developed, and models for reentrainment in cyclones and in hori-
zontal and vertical baffles were discussed.
The experimental part of the program consisted of three parts. In one
part, a small-scale experiment was conducted to observe the transition
from separated flow to separated-entrained flow, with special attention
to the effects of duct dimensions, dependence of entrainment velocities-
on liquid flow rate, drop size distribution, and impingment angle
between air and water phases. For this purpose a two-phase flow system
was constructed, consisting of an inclined, fixed channel through which
a.' film of water flowed, and an adjustable nozzle through which air could
be blown to impinge at chosen angles on the water film. Visual obser-
vation and a filter paper technique were used to determine the transi-
tion from separated to separated-entrained flow.
In a second part of the experimental program, a pilot plant was de-
signed and built for the study of entrainment separators. Its purpose
was to obtain data to check design equations for efficiency of separa-
tors and to study effects of gas velocity, liquid drainage and flooding,
55
-------
and vertical and horizontal sections. The pilot plant capacity was
1.4 m3/s or 3 x 10'
tors were studied:
3 3
1.4 m /s or 3 x 10 cfm. Five different types of entrainment separa-
1. Knitted mesh (Model 4CA, layered, crimped in alternate
directions, 0.144 g/cm^ density, 0.028 cm wire diameter,
98.2 percent voids, 2.8 cnr/cm^ mesh surface area, 10 cm
thickness, and AISI 304 material)
o 3
2. Packed bed (2.5 cm Pall ring packing, 1.9 cm /cm specific
surface, 0.088 g/cnr density, 30 cm bed length, and poly-
propylene plastic material) - . .
3. Zigzag baffle section (7.5 cm x 61 cm x 0.16 cm dimensions,
six rows, 2.5 cm spacing between rows, 30-degree angle
between baffle surface and air flow direction, and 7.3 cm
spacing between baffles in a row)
4. Cyclones (61 cm diameter x 243 cm high, 30.5 cm long x.
15 cm wide inlet, 30 m/s maximum inlet velocity)
5. Tube bank (six rows, 1.9 cm external diameter, 61 cm
length, and eight equispaced tubes in a row)
The ranges of variables were:
Drop diameter = 82 to 1600 um
Air velocity = 100 to 750 cm/s
Liquid flow rate = 17 to 670 cm /s
Air flow rate = 0.13 to 1.4 m/s
Results
In the theoretical part of the program, design equations were given for
entrainment separators. Specifically, formulas for primary efficiency
are exhibited for cyclones, packed beds, knitted mesh, sieve plates,
and zigzag baffles; they have the form
E = 1 - exp [- f (K )] (5)
56
-------
where K is the droplet inertial impaction parameter, and f is a func-
tion of this parameter. .For packed beds, sieve plates, and zigzag
baffles, f is linear; for the other separators, f is nonlinear, in-
creasing with K . For a tube bank, the exponential term in Equation
(5) is replaced by
(i - c.n)n (6)
where n is the target efficiency of a rectangular aerosol jet, C is
a geometric factor, and n is the number of rows. As n tends to in-
finity, this expression tends toward the exponential form.
The design equations for primary collection efficiency were found to
give values higher than the observed overall separator efficiency.
This may have been due to reentrainment, one cause of which is high gas
velocity. Usually, entrainment separators are operated at'low gas vel-
ocity, leading to large sizes and costs. Rupture of bubbles at the
gas-liquid interface, x^ith subsequent drop formation, is the principal
cause of reentrainment in sieve plates, bubble-cap plates, packed beds,
and knitted mesh. Lower velocities inhibit such rupture.
Design equations were derived for reentrainment in cyclones. These
equations gave lower reentrainment velocities than those observed
experimentally.
Collection efficiency is plotted versus gas velocity in Figure 7 for
horizontal flow through tube banks, for various inlet drop-diameters.
Penetration due to less than 100 p'ercent primary efficiency was observed
for velocities under 3 m/s. The slight drop in efficiency at 7 m/s may
be due to reentrainment (not specifically alluded to by APT, Inc.).
As gas velocity increases, efficiency first increases as the impaction
parameter increases, then efficiency decreases as the gas velocity be-
comes high enough to reentrain the water.
57
-------
z
Ul
o
o:
UJ
o.
y 60
o
z
UJ
o
u.
u.
UJ
!-
o
UJ
o
o
100
80
40
20
I i 1 1 1
/' 0
0
v
INLET DROP DIAMETER , pm
V 84
O 380
Q 1,230
O >l,230
1
1
1
1
1
23456
GAS VELOCITY, m/s
8
Figure 7. Collection efficiency versus gas velocity
in tube bank.
58
-------
In general, a minimum was observed in the drop size distribution of
the reentrained drops. The minimum was generally in the range of 40
to 80 ym, with a few exceptionally high values up to 190 ym associated
with baffle separators. Observed reentrainment velocities were much
37
higher than those obtained by Kotov in similar experiments.
Conclusions
The theoretical models for primary collection efficiency and pressure
drop developed for baffles showed good agreement with experimental
data. However, predicted reentrainment velocities in a cyclone were
much lower than the experimental values.
The effect of liquid load on pressure drop was negligible in the
packed-bed, zigzag-baffle, tube-bank, and cyclone separators. Collec-
tion efficiency in the experimental range was nearly 100 percent,
with 0.5 to 1.0 percent reentrainment at high gas velocities. The
mass-median drop size for reentrainment varied between 250 pm and 650
ym. The minimum drop size present in the reentrainment was 40 ym.
A problem with entrainment separators is the fact that the gas veloci-
ties are limited to low values by the onset of reentrainment and flood-
ing conditions. Another problem involves solids deposition and conse-
quent plugging.
Recommendations
Calvert et al. suggested that a further study is needed of:
1. Onset of reentrainment conditions,
2. Rate of reentrainment,
3. Equilibrium constant between entrainment and liquid film,
4. Drop size distribution,
59
-------
5. Smooth and shock contact gas liquid, and
6. Effect of duct dimensions.
The new knowledge would be applicable toward improving design methods
to determine reentrainment under operating conditions, the effect of
high gas velocities, and improvements to reduce reentrainment.
It would be desirable to have insight into the mechanisms by which
suspended and dissolved solids become deposited, so that this deposi-
tion could be minimized by appropriate design. Since some deposits
are inevitable, research on methods of washing, etc. to remove them
with a minimum expenditure of liquid is desirable. (Some small-scale
experimental work was done as part of the third part of this program.
More work is planned under other programs.) .
Performance data on industrial entrainment separators are generally
not available. The ususal assumption in evaluating overall performance
of scrubbers is that the entrainment separators are 100 percent effi-
cient; the liquid introduced in the scrubbers is assumed to be entirely
removed by the separators; and the effects of sedimentation, bends in
the duct, etc. are neglected. Hence, attempts should be made to col-
lect performance data under industrial conditions; this will help in
the future design of practical separators.
A demonstation plant should be built, of larger capacity (about 35 x
33
10 cfm or 170 m /s as compared with the present pilot plant capacity
3 x 103 cfm or 1.4 m3/s).
Finally, combinations of entrainment separators should be studied with
a view toward achieving higher efficiencies and/or lower capital costs.
60
-------
WET SCRUBBER LIQUID UTILIZATION (Stanford Research Institute,
September 1973)
Introduction
As interest in condensation effects in scrubbing intensified, this
IERL-RTP program with Stanford Research Institute was modified to shift
the emphasis from the study of the effects of energy consumption
and water flow rate on efficiency to the effects of condensation.
The scope of work included: literature review on water and power
consumption as well as on condensation/evaporation effects; bench-
scale study of co-current spray scrubber efficiency; recommendations
for potential commercial uses of condensation in scrubbing, with
emphasis on venturi scrubbers; development of plans for pilot-scale
research.
Goals
The principal goal was the study of the effects of the following on
collection efficiency of a venturi-type scrubber:
1. Power consumption per unit volume flow rate
2. Venturi orifice size (equivalently, gas velocity)
3. Liquid-to-gas ratio
4. Condensation/evaporation
Secondary goals were the literature review, recommendations for
commercial applications, recommendations for pilot-scale research.
Methods
The experimental apparatus used by Semrau and Witham is shown in
their report.^ It allowed them to control the temperature and
humidity of the gas to be scrubbed, the temperature of the scrub-
bing water, the aerosol concentration, and the relevant gas and
61
-------
liquid flows. The aerosol was produced by an ultrasonic nebulizer
(Devilbiss Model 35 A) which atomized a solution of ammonium
fluorescein to produce a spherical, nonhygroscopic aerosol that
had a mass-median diameter (from electron microscopy) of about 0.6 um.
-3 3
Inlet concentrations were about 10 g/m . Both inlet and outlet
concentrations were measured by fluorimetric analysis of samples
captured on Nuclepore filters.
During the investigation, the scrubber was fitted with three different
orifices, having diameters 2.54, 3.81, and 4.45 cm. Air flows ranged
3 3
from about 0.008 m /s (18 cfm) to 0.06 m /s (130 cfm) for the tests
at ambient conditions done to investigate the impact of orifice size
(gas velocity) and liquid-to-gas ratio (L/G) on efficiency. Various
air temperatures and humidities were used to evaluate the influence
of condensation/evaporation. Air and water characteristics were chosen
to make four different sets of conditions:
1. Ambient conditions
2. Adiabatic saturation, with the water temperature at the
wet bulb temperature of the gas to be scrubbed
3. Condensing conditions, with the inlet gas wet bulb tempera-
tures well above the temperature of the scrubbing liquid,
to promote condensation
4. Vaporizing conditions, with the gas at temperatures near
those for the condensing conditions but quite dry to
keep the wet bulb temperature was well below the tempera-
ture of the scrubbing water, thus promoting evaporation.
Results
The data were plotted as the natural logarithm of the number of "transfer
units" (N ) versus fluid power consumption per volume flow rate, also
called "effective friction loss" or "contacting power" or "equivalent
pressure drop,"Ap (mmH-O = rnmWC). The penetration, Pt, is the ratio
of the outlet to inlet concentrations:
62
-------
(7)
The number of transfer units, N is the negative of natural logarithm
of the penetration:
N = - In Pt = IhCPt"1) (8)
If the penetration has an exponential dependence on effective friction
loss to the power a, then:
Pt = exp (- bApa), (9)
and the logarithm of the number of transfer units would be
In Nfc = a In Ap + b. (10)
An exponential dependence on the effective friction loss, Ap, would be
a straight line with slope a on a log-log plot.
Because the aerosol generator was not constant during the trials, des-
pite considerable efforts to abate the variations, it was found that
the variance of the data could be reduced if the aerosol generated at
the high rates of generation ("aerosol A") was distinguished from that
produced at low rates of generation ("aerosol B"), the boundary rate
being about 750 yg/inin. The aerosol (B) produced at the lower genera-
tion rates caused lower scrubber efficiencies at equivalent pressure
drops than the aerosol produced at the higher rate, making it likely
that aerosol B had a smaller mass mean diameter or was more polydisperse.
The report by Semrau and Witham showed results of tests of scrubber
performance as a function of effective friction loss (pressure drop
across the venturi) for different throat velocities and different
liquid-to-gas ratios. Figure 8 summarizes these. The power consumption
63
-------
10
12
z
c
ui
UL
§ '
o
£
co
0.1
I I II I
I I I I I l_
OPERATING
CONDITION
^ ^ Condensing
Condensing
' Adiabatic Saturation
-- Vaporizing
Ambient
I I I I I
GAS TEMPERATURES fC)
Inlet
80.0
130.0
128.9.
125.6
~30.0
Outlet
37.8
373 and 54.4
71.7
58.3
~30.0
100
1000
EFFECTIVE FRICTION LOSS mm WC
10.000
Figure 8. Summary of scrubber performance curves for adiabatlc
saturation, condensation, and vaporization scrubbing-
aerosol B
64
-------
was found to be much more important than the orifice (throat) velocity
or the L/G ratio. (One reason this had to be established was that the
later tests for condensation/evaporation effects had to be done at
different L/G ratios.) At pressure drops larger than about 40 cm WC,
the slope (a) is less than one, but below this pressure drop, the slope is
greater than one. This shows "diminishing returns" from the higher
pressure operation.
In certain cases, scrubbers in series may have a greater efficiency than
a single scrubber with the same pressure drop. The penetration of two
scrubbers in series, Pt , is the product of their individual penetrations,
_ s
Pt.. and Pto, but the power consumption is the sum of their individual
power consumptions :
Ap = AP;L + Ap2 (11)
Two-stage scrubbing would give
Pt2 = efi e2 (12)
,Pt = e-pi 2 (13)
s
For a single scrubber with the same pressure drop,
, . a -b(Ap, + Ap0)
Pt = e-bAp = e l 2 , (14)
but for a < 1
Ap2)a < AP]La + Ap2a (15)
so
Pt > Pt .
s
65
-------
In words, if a, the exponent of Ap, is less than one, two scrubbers in
series will give less penetration (higher efficiency) than one scrubber
with a pressure drop equal to the sum of their pressure drops. The
comparison is more difficult when the slope a is a function of pressure
drop, as it seems to be, and where the successive scrubber(s) in a
series arrangement may follow somewhat different equations because of
particle conditioning by the previous scrubber(s).
Figure 8 is adapted from Semrau and Witham. It shows the effects of
condensation and evaporation on the collection efficiency of the scrub-
ber for aerosol B. (We have added the ambient data from another figure.)
Table 5 summarizes the results. (Series III tests were run with the
inlet gas only 5 C or so above the adiabatic saturation temperature, wet
bulb temperature, "so that the tests approximated scrubbing of a pre-
saturated gas.") The collection efficiency'at a given pressure drop, or
effective friction loss, was clearly better for the condensing situa-
tions than for the ambient or vaporizing conditions.
Conclusions
We quote directly from Semrau and Witham:
« Over very wide ranges of gas velocity, liquid-to-gas ratio,
and gas pressure drop, the collection efficiency of the
scrubber was dependent only upon effective friction loss
(essentially equal to the gas pressure drop) and indepen-
dent of gas velocity and liquid-to-gas ratio (except as
these latter factors affected gas pressure drop). (This
conclusion did not apply in cases where vaporization or
condensation effects were involved.) There may still be
some independent effects of gas velocity or liquid-to-gas
ratio at very low values of these variables, but if so,
the operational region is not one of much practical
significance.
The size of the orifice gas/liquid contactor had no ef-
fect on collection efficiency, independent of effective
friction loss.
66
-------
Table 5. SELECTED COMPARISONS OF SCRUBBER PERFORMANCES AT DIFFERENT OPERATING CONDITIONS
Test Condition
Gas Temp (°C)
Inlet
Outlet
Aerosol
A
B
Effective
Friction
Loss
(mm WC)
250
1000
1500
250
1000
1500
Adiabatic
Saturation
129
71.7
Transfer
Units
0.685
2.52
3.68
0.583
2.30
3.43
Collection
Efficiency
49.6
91.95
97.48
44.2
90.0
96.76
Condensing
Series I and II
.
130
54.4 and 37.8
Transfer
Units
1.01
2.93
4.00
0.840
2.46
3.36
Collection
Efficiency
63.3
94.66
98.17
56.8
91.46
96.50
Condensing
Series III
80.0
37.8
Transfer
Units
1.14
3.12
4.18
Collection
Efficiency
67.8
95.55
98.. 47
Vaporizing
126
58.3
Transfer
Units
0.303
2.20
3.05
Collection
Efficiency
26.0
88.9
95.23
ON
-------
Condensation scrubbing of the test aerosol produced
significant but not very large increases in collection
efficiency at a given effective friction loss. The re-
results were qualitatively similar but quantitatively much
less than indicated by earlier pilot-plant studies of con-
densation scrubbing of Kraft recovery furnace fume under
similar operating conditions.
Cooling the outlet gas below 54°C did not increase the
condensation scrubbing effect measurably, presumably be-
cause only a small additional quantity of water vapor was
condensed.
Adjusting the inlet gas to a nearly saturated condition
before contacting with cold water markedly improved the
effectiveness of condensation scrubbing.
Vaporization of scrubbing water tended to reduce scrubber
collection efficiency, particularly in the lower range of
effective friction loss.
In practical cases, condensation collection is not an
alternative to conventional high-energy scrubbirig, but
is a supplement that can reduce the scrubbing energy
requirements to a greater or lesser degree. Promoting
diffusional mass transfer (such as vapor condensation)
in practical equipment also requires energy for gas/
liquid contacting.
Because of the requirements for water cooling and pos-
sible gas humidification, the net saving of energy in
condensation scrubbing may not often outweigh the other
associated costs.
Condensation scrubbing may be, and probably is, warranted
under the following conditions:
- Where the waste gas already has a high water
vapor content and there is use for low-level
heat in the associated process system (e.g.,
Kraft pulp mills).
- Where the gas must be cooled and dehumidified
anyway, for process reasons (e.g., cleaning of
blast furnace gas; purification and condition-
ing of feed gas to contact sulfuric acid plants).
63
-------
Recommendations
Semrau and Witham of Stanford Research Institute recommended that similar
work be extended to aerosols of other mean particle sizes to investigate
size dependence. They also suggested that the decrease in efficiency at
the low equivalent pressure drop range be studied further. More work
was urged regarding two other types of scrubbers: those for which the
scrubbing liquid supplied the power and those for which a mechanically
driven rotor supplied power to mix particles and drops. If a field demon-
stration of condensation scrubbing was to be done, they recommended that
it be done on a Kraft recovery furnace fume, a possible commercial
application.
PILOT SCALE DEMONSTRATION OF CHARGED DROPLET SCRUBBING
(TRW SYSTEMS, INC. 1972-1975)
In previous work, TRW demonstrated that charged droplet scrubbing
(CDS) was feasible to aid fine particle removal. The present study
is to construct and evaluate a pilot scale demonstration of charged
droplet scrubbing. TRW's responsibilities included site selection,
verification of hardware, and actual testing of the CDS device. The
IERL-RTP project description laid down the following goals and guide-
lines for the project:
"TEW shall demonstrate a charged droplet scrubber (CDS) with
a capacity of at least 25,000 cfm on an industrial source.
TRW shall collect particulate data to include particle size
distribution at inlet and outlet, fractional efficiency,
Ringleman readings at exit with and without CDS operation, and
particulate concentration at inlet and outlet. TRW shall
record all relevant process data, including gas humidity, tem-
perature, velocity, and water flow rates, temperature, pH,
and conductivity for inlet and outlet. The CDS shall be
operated continuously for at least 500 hours. The final re-
port shall contain a comprehensive summary of test procedures
and analysis of results, a discussion of fractional efficiency,
limitations, advantages, etc. and a cost estimate of a full-
size application."
69
-------
SYSTEMS OF CHARGE DROPLETS AND PARTICULATES (MIT)
o
Under a contract with IERL-RTP, Melcher and Sachar of MIT studies ex-
perimentally and theoretically the various combinations of charged
droplets and particles which might be used in electrostatically aug-
mented spray scrubbers. They defined characteristic times for certain
particle and droplet behavior, and from these characteristic times the
collection efficiency of such a system could be determined. They con-
firmed their theoretical predictions experimentally. Their group t /t
1T6S
(ratio of residence time to characteristic time for collection) is
essentially the same as our wA/Q. The characteristic times all have
the form:
t = l/4irq q B N
where q is charge, B is mobility, and N is number concentration. The
motion of particles (subscript p) with respect to themselves is char-
acterized by:
t = l/47rq 2 B N
P P P P
The motion of droplets (subscript d) with respect to themselves is char-
acterized by:
The motion of particles with respect to droplets is characterized by:
% 1/4% "d BP Nd
These times are the average distance between the objects in motion divi
ded by their migration velocity, as calculated from the Coulomb force
at the average distance. The characteristic collection time for mutual
70
-------
repulsion by particles is t . The characteristic collection time for
collection of the particles by the droplets (and for mutual repulsion
if polarities are the same) is t, . The characteristic time for mutual
dp
repulsion of the spray droplets is t ; it must be substantially longer
than the characteristic collection time if it is not to limit collection.
The approach of Melcher and Sachar unified the analysis of many seem-
ingly diverse systems which involved charged particles and charged
droplets.
DESIGN AND FABRICATION OF MOBILE WET SCRUBBER FACILITY
(NAVAL SURFACE WEAPONS CENTER, 1973 TO 1975)
Information was needed on the suitability of scrubbers to various par-
ticulate sources. Responding to this need, by an interagency agreement
IERL-RTP contracted with the Naval Surface Weapons Center (NSWC) to
design, construct, and initially test a mobile wet scrubber pilot facil-
ity. NSWC was charged with testing the mobile scrubber on selected
sites before turning over the unit to IERL-RTP. The specific require-
ments for the effort according to IERL-RTP project description were:
"NSWC will perform an evaluation of existing scrubber technology
to develop the design for a mobile wet scrubber unit to be mounted
on a flat bed trailer. NSWC will specify, procure and/or fabricate
all equipment required to construct the unit. A plan for testing
the unit will be prepared and tests will be carried out on specified
sources prior to releasing the unit to IERL-RTP."
FOAM SCRUBBING FOR FINE PARTICLE CONTROL (MONSANTO RESEARCH
CORPORATION, 1973 TO 1977)
The use of foam as the capture agent in a scrubber had long been discussed
and the unit-operations were presented in the Scrubber Handbook. IERL-RTP
contracted with Monsanto Research Corporation (MRC) to conduct a small-
scale feasibility demonstration of foam scrubbing. The mechanism of
bubble formation and destruction and identification of operating
71
-------
characteristics of the scrubber system were to be primary targets of
the study. The IERL-RTP project description "Scope of Work" to MRC read:
"MRC shall conduct a preliminary investigation of foam scrubbing
to identify and verify by experiment mechanisms for foam genera-
tion, particle capture, and foam destruction. The foam scrubber
will likely consist of three components: the foam generator,
particle collector, and foam destroyer. The preliminary study
will determine the minimum energy requirements for foam breaking
and will attempt an economic analysis of foam scrubbing. The
second phase of work will involve the verification of fine par-
ticle collection. This will include study of foam generation to
determine surfactant type and concentration, and bubble thickness
and diameter; mechanisms for foam destruction; to develop and
execution of a test to determine particle collection efficiency
for a 500 cfm (minimum) capacity scrubber; and to evaluate the
results of the testing. The final report shall contain sum-
maries of the collection efficiencies, the operation, and the
commercial applicability of the device."
GROWTH OF FINE PARTICLES BY CONDENSATION (SOUTHERN
RESEARCH INSTITUTE, 1973 TO 1974)
The increased efficiency noted in the first FF/C work led IERL-RTP to
contract with Southern Research Institute (SoRI) to "determine the feas-
ibility of fine particle control systems based on growing fine particles
by condensation, and collection of the grown particles- and to provide the
data necessary to design such fine particle control systems." A com-
plementary objective of the contracted study was to evaluate the use of
vapor condensation in conventional particulate control devices to im-
prove their collection efficiency. SoRI's objectives in the study were
specified in the IERL-RTP project description:
"SRI will perform experimental studies to determine the gas phase
conditions necessary to grow fine particles to collectable size
by water vapor condensation, the practicality of various methods
of producing the required conditions, and the collectability of
the grown particles by such conventional devices as ESP's and
low energy scrubbers. Research will include investigation of the
response by particle growth mechanisms to varying particle size
distributions and to the chemical composition of .the gas and
particulate."
72
-------
EVALUATION OF ARONETICS TWO-PHASE JET SCRUBBER (SOUTHERN
RESEARCH INSTITUTE, 1974)
Introduction
As part of a program to investigate novel fine particle collectors,
Southern Research Institute (SoRI) conducted for IERL-RTP tests of frac-
tional and overall mass collection efficiency on the two-phase jet scrub-
ber of Aronetics, Inc. Tests were performed on a full-scale scrubber
used for controlling particulate emissions from a ferro-alloy electric
arc furnace. The Aronetics device utilizes high pressure (~24 atm) hot
water (~200 C) to produce a high velocity two-phase flow in a venturi
nozzle (Figure 9). The scrubbing water may be heated by an economizer
connected with the heat source of the industrial process. Alternatively,
steam generated by high temperature off-gases (>650 C) may be used in
an intermediate boiler to produce the hot water.
Goals
The objectives of the SoRI evaluation were:
1. Determine collection efficiency as a function of particle
diameter.
2. Evaluate economics of the scrubber operation.
3. Determine optimum operating conditions.
Methods
Tests were performed with the Aronetics scrubber on a 7.5 megawatt sub-
merged arc ferroalloy furnace at Chroinasco, Inc.'s Memphis, Tennessee
facility. Furnace operation was 24 hours a day with tapping done at 2
hour intervals; stoking and charging occurred at irregular intervals be-
tween taps. The variable cycle was accompanied by a variable emission
rate, which complicated the testing.
73
-------
,TWO-PHASE
JET NOZZLE
HEAT EXCHANGER
HOT GAS
.MIXING
SECTION
SEWRATOR-
MAKE-UP^
WATER
-ITU!
STACK
PUMP
WASTE WATER
TREATMENT
Figure 9. Generalized two-phase jet scrubber system
-------
Particle measurement techniques used were:
1. Condensation nuclei counters and diffusion batteries for
concentration and size distribution by number for d < 0.2 um.
(Extensive dilution, 65:1, was required.) P
2. Optical techniques for concentrations and size distributions
for 0.3 ym < d < 1.5 urn.
p .
3. Standard mass train measurements for total inlet and
outlet mass loadings.
Results
Figure 10 indicates the fractional efficiency of the Aronetics scrubber
based on reduced optical, diffusional, and impactor data. Total mass
collection efficiencies of approximately 96 percent were determined by
Guardian Systems, Inc. (Anniston, Alabama) under subcontract with SoRI.
Conclusions
The conclusions of SoRI were stated concisely in their final report:
"The collection efficiency of the Aronetics Two Phase Scrubber
determined by conventional (Method 5) techniques on a source
producing particulate having a mass mean.diameter of about 3 ym
was 95.1 and 96.7 percent for two days of testing. Measured
fractional efficiencies were about 70 percent at 0.01 ym, about
35 percent at 0.05 ym, 35 percent at 0.1 um', 99 percent at
0.5 ym, 99 percent at 1 ym, and 99.4 percent at 5 ym. The
scrubber energy usage during the tests was approximately 635
joules/1000 scm (17,000 Btu/1000 scf) at a net pressure rise of
12-1/2 to 16 in. H20. This energy usage was a result of using
all the process waste heat available and may have been in excess
of the minimum amount required to achieve the efficiencies
obtained during these tests."
It should be noted that the thermal energy consumption, '17,000 Btu/1000
scf (= 6.3 x 105 J/m3), is 400 hp/1000 hp/1000 cfm or 6300 cmWC equivalent
pressure drop. In addition, SoRI evaluated the costs attending the scrub-
ber operation based on 1967 factor and fuel costs and found annual direct
operating costs per scfm:
75
-------
I
+ IMPACTORS
D OPTICAL
O DIFFUSIONAL
I
I
O.I 1.0
PARTICLE DIAMETER ,
10.0
Figure 10. Fractional efficiency of the Aronetics Scrubber based oh optical,
diffusional, and impactor data
-------
o
$1.016/scfra - year ($2,085/(m /s)-year) when waste heat can be used.
o
$5.59/scfm - year, ($ll,843/(m /s)-year) when fuel for heating water
must be purchased.
Excluding electric power used for fans and blowers, if the jet flow in-
duced by the hot water suffices, the above costs became $0.042 and $4.44,
respectively. These figures can be changed to approximate 1975 values
by multiplying them by the ratio of the Marshall and Stevens price
indices for the years 1975 (first quarter) and 1967: (437)/(263) = 1.66.
EVALUATION OF THE LONE STAR STEEL STEAM-HYDRO AIR CLEANING
SYSTEM (SOUTHERN RESEARCH INSTITUTE, 1974)
Introduction
In another novel device evaluation, IERL-RTP contracted with Southern
Research Institute (SoRI) to evaluate the Lone Star Steel Steam-Hydro
air cleaning system designed to control particulate emissions from an
open hearth furnace. The Lone Star device incorporates a high speed
steam drive with injected water to accomplish scrubbing. The no moving-
parts device consists of a steam nozzle, water injector, mixing tube,
and twin cyclones.
Goals
Determine fractional and overall mass-efficiency of the Lone Star scrubber.
Methods
SoRI tested the performance "of seven Lone Star scrubber modules fed by ef-
fluent from four open hearth steel furnaces, each producing three 300 ton
o
(273 x 10 kg) batches of steel per day. Particulate measurements were
made over a 1 week period.at scrubber inlet and outlet ports using the
following instruments and techniques:
77
-------
1. Condensation nuclei counters and diffusion batteries for
number concentration of d < 0.2 ym particles.
2. Multi-channel optical counter for concentration and size
distribution for 0.3 ym <_ d < 1.5 ym particles.
3. Cascade impactors for concentration and size distribution
by mass for 0.25 ym £ d _< 5 ym particles.
4. EPA Method 5 mass train measurements for total mass loading.
Sample dilutions of 500:1 at inlet and 65:1 at outlet were required for
the optical and condensation nuclei instruments. Tests were conducted at
different steam pressures, gas flow rates, and cyclone positions. SoRI
indicated that direct comparison of data between specific tests would not
be meaningful due to the variable cycle of the open hearth process.
Results
A summary of the averaged fractional efficiencies was provided in SoRI's
final report and is reproduced here (Figure 11). The particulate source
had a mass mean diameter about 1 ym. The overall mass collection ef-
ficiencies for each of the two days of tests were 99.90 percent and 99.84
percent, respectively. Lone Star's estimated energy requirements for
scrubber operation at the latter efficiencies were 8250 to 12,750 Btu/
1000 scf for pressure drop of 1 to 6 inches of water. This power con-
sumption is 200 to 300 hp/1000 cfm or about 4000 cmWC equivalent pressure
drop, 4 x 105 J/ra3.
Conclusions and Recommendations .
No conclusions regarding optimum scrubber operating conditions were made
by SoRI, and no recommendations for further testing and development were
put forth.
78
-------
VO
0.01
O.I
5
10
o
<
cc
5 30
a.
90
99
99.9
99.99
O OIFF
D OPT
+ IMPACT
1 i 111
I
i i i i i
99.99
99.9
99
95
90 .
o
z
UJ
o
UJ
50
o
u
o
o
10
3
O.I
0.01
0.01 O.I 1.0 10.0
PARTICLE DIAMETER, jjm
Figure 11. Fractional efficiency of the Lone Star Steel Steam-Hydro scrubber
-------
DYNACTOR SCRUBBER EVALUATION (GCA, 1974) .
Introduction
As part of lERL-RTP's program for evaluation of novel types of scrubbers,
GCA tested a 1000 cfm (0.47 m/s) version of RP Industries' Dynactor,
a scrubber design based on aspirative pumping. Figure 12 is a cross-
section of a single-stage Dynactor diffusion contactor. The reaction
column is typically 6 feet (2m) long, and is 12 inches (0.3m) in diam-
eter. The liquid is atomized under high pressure, into a shower of
high-velocity droplets, using a proprietary nozzle design. This shower
of droplets educts large volumes of gas while simultaneously scrubbing
it. The gas enters at ambient pressure and low velocity. Contact
between gas and liquid occurs throughout the reaction chamber.
Goals
As in other evaluations, the objective of this study was to test a novel
device purportedly capable of high-efficiency collection of fine parti-
cles. The testing included measurements of power consumption and frac-
tional collection efficiency:
At three flow rates
At two temperatures
At two dust loading levels
For two different kinds of dust
Methods
The test methods were state-of-the-art techniques for measuring collection
efficiency as a function of particle size using cascade impactors upstream
and downstream from the scrubber. The data were analyzed using an F-test
analysis of variance to determine what factors have a significant
influence on collection efficiency in the various aerodynamic size
80
-------
f
B»
,
m»
f
«
H
u
LIQUID-INPUT, UO TO 200psl
AIR INPUT, LOW VELOCITY, AMBIENT PRESSURE
HIGH VELOCITY, SUB-AMBIENT PRESSURE
SHOWER OF THIN FILMS AND PARTICLES
REACTION COLUMN
'/S
'// fi\
MM*
TURBULENT
MIXED
FLUID
GAS OUTPUT
BAFFLE
RESERVOIR/SEPARATOR
(LIQUID)
LIQUID LEVEL DETER-
MINING TRAP
LIQUID OUTPUT
Figure 12.
Single-stage Dynactor diffusion system
cross-sectional view
81
-------
fractions and to estimat'e experimental uncertainty. Pitot tubes were
used to measure flow rates, and Magnehelic pressure gauges to measure
pressure gain or drop.
The following measurements were made:
Air flow and pressure gain versus spray nozzle pressure
Electrical power consumption versus spray nozzle pressure
Mass collection efficiency versus aerodynamic particle
diameter for fly ash and iron oxide dusts at two values
each of flow rate, temperature, and concentration
Results
Some of the principal results are shown in Figure 13 and 14, where col-
lection efficiency is plotted versus particle aerodynamic diameter.
Figure 13 shows the effects of loading and dust type, and Figure 14
shows the effects of flow rate and inlet temperature.
Conclusions
Efficiency is found to be greater for:
Lower flow rate
Lower temperature
Higher dust concentration
Fly ash than for iron oxide
Except for the temperature variation measurements, the differences were
statistically significant for most size fractions.
From its aerodynamic cut diameter versus power per unit volume flow
rate, it is found that the Dynactor Scrubber is essentially equivalent
in power consumption to a well-designed venturi scrubber of the same
collection efficiency.
82
-------
99
98
z
Ul
o
_""
o
ui 95
o
u.
u.
Ul
z
o
p 90
0
Ul
_l
J
o
o
80
50
10
r\
-
-
"*
-
_
-
-
_
-
-
-
.+
**
-J
~s>
-
I GRAND
I MEAN
+ ~lgr/ft3
Mg/m3)
Q -0.1 gr/f»3
(-0.1 g/m3)
.
^
S
O
. .
*
6
£
1 1 1
+
5
0
,
-
-
^
-
_
-
-
_
-
,
>
0
+
1+
^!
+'
5
0
i
5
O
J GRAND MEAN
+ FLY ASH
O IRON OXIDE
1 II
^*"
-
-
~
1.0
2.0
z
Ul
5.0 cc
Ul
Q.
Z
0
p
^
10 p
Ul
Ul
a.
20
50
100
1234
DIAMETER, urn
I 2 34
DIAMETER, pm
Figure 13. Dynactor scrubber collection efficiency versus particle
aerodynamic diameter, effects of loading and dust type
83
-------
99
98
z
UJ
o
on
UJ
a.
"
> 95
z
UJ
o
u.
u.
UJ
z 90
o
h-
o
UJ
_J
o
o
80
50
10
r\
-
*
I
-
'
m
"
1
"I
J GRAND MEAN
500 cf m
+ {0.24m3/s)
QlOOOcfm
(0.47 m3/s)
,
T
i
«fc
o
+
i
o
1
OY+
it i
+
5
0
i
-
-
**
^
*
.
~
^
.
_
.j.
5
-
1
1
|
O
.
X,
o
J GRAND MEAN
+ 20°C(70°F)
O 95°C(200°F)
1 1 1
-
"
-
-
-
1.0
2.0
t^
UJ
5.0 o
cc
UJ
0.
z
o
H
10 2
UJ
z
UJ
0.
20
50
tf\f\
1234
DIAMETER, pm
1234
DIAMETER, pm
Figure 14. Dynactor scrubber collection efficiency versus particle
aerodynamic diameter, effects of flow rate and inlet
temperature
-------
TM
PENTAPURE IMPINGER EVALUATION (GCA, AUGUST 1974)
Introduction
Sponsored by IERL-RTP, an evaluation was performed by GCA/Technology
TM
Division on the "Pentapure " scrubber of Purity Corporation. A gray
iron foundry having a particulate effluent with a mass median aero-
dynamic diameter of 0.5 urn was the site of GCA's Pentapure performance
tests. In the Pentapure scrubber dust-laden gas mixed with water spray
is accelerated in a converging tube, after which the jet is impacted on
a plate mounted perpendicular to the stream.
Goals
Evaluation goals were:
Determine the collection efficiency as a function of particle
aerodynamic diameter.
Determine total mass collection efficiency.
Methods
The Pentapure Scrubber was installed after the spray-cooling section for
the effluent gases in a grey iron foundry. Collection efficiency was
determined by measuring particle mass concentrations at inlet and outlet
of the scrubber. For total mass collection efficiency GCA used Method 5
total mass sampling traverses at inlet and outlet. Collection efficiency
as a function of aerodynamic diameter was obtained with Andersen In-Stack
Impactors up- and down-stream of the scrubber. A preimpactor for removing
droplets larger than 15 ym diameter and heating tape to dry remaining drops
were employed with each Andersen impactor. Comparison between experimental
results and theoretical collection efficiency was carried out using formulae
4
from the Scrubber Handbook.
85
-------
Results
Total mass, efficiency was 10 ± 2.5 percent.
Particle aerodynamic diameter for 50 percent collection
efficiency was between 2 and 4 pm. This agreed with
theoretical predictions.
Pressure drop across scrubber was 6 in. H«0 = 15 cmWC.
Conclusions
GCA noted that, "Measured efficiencies correspond to those expected from
venturi scrubbers having somewhat less pressure drop. The Pentapure is not
an efficient fine particle collector."
Recommendations
In its final report GCA stated:
"The Pentapure scrubber does not give high efficiency for col-
lecting fine particles and should not be investigated further
for that purpose. In those situations in which a low-energy
venturi scrubber would be suitable, the use of the Pentapure
scrubber might be considered as well."
EVALUATION OF ELECTROSTATICALLY AUGMENTED PARTICULATE CONTROL
DEVICES (GCA, OCTOBER 1974)
Nearing completion is a study by GCA/Technology Division of various applica-
tions of electrostatics to the control of fine particulates. The CSL-
supported study encompasses electrostatic precipitators, filters, and
scrubbers. It includes a review of the electrostatic force mechanisms, an
analysis of programs in electrostatic augmentation supported by CSL, a check-
list for electrostatic augmentation experimental work, and a method of
setting control research priorities. A simplified method for predicting col-
lection efficiencies in devices such as electrostatic spray scrubbers is
presented, as well.
86
-------
THEORETICAL EVALUATION OF THE CENTRIFIELD1" SCRUBBER
(MIDWEST RESEARCH INSTITUTE 1974)
Introduction
The Centrifield Scrubber developed by the Entoleter Corporation was
evaluated theoretically based on an idealized model by Midwest Research
Institute (MRI) under IERL-RTP contract. The Centrifield scrubber
achieves particle collection with a centrifugal vortex containing water
droplets through which passes contaminated gas. The vortex is established
by stationary vanes at the periphery which admit tangentially the gas and
water spray. Large droplets are radially accelerated to pass through slots
in the vanes, while smaller droplets exit at the core and must be deentrained.
Goals
The objective established by MRI was to evaluate the Centrifield scrubber
with respect to its operating efficiency and economic feasibility.
Methods
The evaluation of the Centrifield scrubber by MRI proceeded as follows:
An idealized scrubber model was formulated, assuming ideal, incompressible
gas flow with a radial velocity that increased toward the core. Energy
consumption by the radial vanes was calculated, but that used to accelerate
the scrubbing droplets was assumed negligible. Certain assumptions had to
be made about such system parameters as maximum water concentration in the
vortex. Collection efficiency versus particle size was then calculated for
several different pressure drops.
Results
Particle collection efficiency based on MRI's theoretical model of the
Centrifield Scrubber is shown in Figure 15 as a function of diameter for
87
-------
VANE ANGLE
00
oo
AP - PRESSURE DROP
0.001
0.01
0.10
PARTICIPATE DIAMETER
1.00
10.00
TM
Figure 15. Fractional efficiency of idealized Centrifield scrubber (9 =65)
-------
the reasonable vane angle- (with respect to tangential direction, 65 .
The MRI report points out that the model is tentative because energy
requirements for droplet formation were not considered and the droplet
concentration is difficult to estimate. Since this is so, MRI indicates
that Figure 15 must be viewed with a degree of skepticism.
Conclusions
1. Lower gas velocities and water feed rates compared to
venturi scrubbers or other scrubbers may be possible.
2. It is not certain that the present device can operate at
conditions necessary for the efficiencies of Figure 15.
3. It is impossible to make an economic evaluation of the
device at present due to the uncertainty in collection
efficiency.
Recommendations
The MRI report makes the following recommendations concerning further
investigation of the Centrifield Scrubber:
"1. Additional operational data including the water concen-
tration in the vortex, be obtained by the manufacturer.
The drop size distribution would also be desirable;
"2. Operational data obtained in Item 1 be used as input
into a simple model such as that presented in this
report and estimated collection efficiencies determined;
and
"3. If the estimated collection efficiencies obtained with
actual equipment parameters are comparable to or exceed
those for venturi scrubbers, the Centrifield Scrubbers
should be tested under field conditions using approved
EPA methods."
89
-------
TM
EXPERIMENTAL EVALUATION OF THE CENTRIFIELD SCRUBBER
(SOUTHERN RESEARCH INSTITUTE, 1975)
Introduction
In an effort to get more definitive information on the performance of the
TM
Centrifield Scrubber, IERL-RTP contracted with Southern Research Insti-
tute (SoRI) to experimentally evaluate the performance of the device on an
asphalt batching process. (This is included at this point, rather than.
in strict chronological order, for clarity.)
Goals
The goals of the study were to determine the total mass collection
TM
efficiency and fractional collection efficiency of the Centrifield
Scrubber at a single plant installation.
Methods
The scrubber was attached to the asphalt plant following the plant's
cyclone. Inlet and outlet particle concentrations and distributions
of the scrubber were monitored using a diffusion battery, a multi-
channel optical counter, an electrical sizing device, a cascade impactor>
and the standard EPA Method 5 measuring technique for total mass loading.
The diffusional and optical measurements were made consecutively rather
than similtaneously at inlet and outlet because only one set of instru-
ments was available.
Results
Test averages of the size-dependent collection efficiencies were made
and appear in the SoRI final report as Figure 16. To reduce the cascade
3
impactor results a particle density of 2.5 g/cm was assumed. Accord-
ing to SoRI, the scrubber energy usage during the tests was about
90
-------
99.99
99.9
99.8
99
98
95
90
80
o
I 40
.o
t 20
10
5
2
I
O.I
0.01
1 ' ' I '"I
o -ELECTRICAL
0-DIFFUSIONAL
A -OPTICAL
o-INERTIAL
I I I
I I I
I I I I I I I I
I I I I I I I I I I I I I I I
I I I
I I .1
0.01
0.1 1.0
PARTICLE DIAMETER,
10
Figure 16. Fractional efficiencies as determined by the four methods used in
the test program. The particle sizes shown for the impactor data
are Stokes Diameters based on a particle density of 2.5 grams/cm^
-------
3 joules/1000 cm-* at a pressure drop of 28 cm WC. The pressure drop
33 33
28 cmWC actually corresponds to 2.7 x 10 J/m or 2.7 x 10 W/(m /s) or
1.7 hp/1000 cfra; presumably the SoRI figure of 3 J/1000 scm is a mistake.
Overall mass collection efficiency for the ~ 100 urn mass mean diameter
source was 99.50 percent and 99.73 percent for the 2 days of testing.
In addition SO collection efficiencies of 84 percent, 68 percent, and
Ji.
88 percent on three tests were determined using H?0~ reagent in impingers.
Conclusions and Recommendations
SoRI offered no conclusions or recommendations beyond the test data.
FINE PARTICLE COLLECTION WITH UNIVERSITY OF WASHINGTON ELECTROSTATIC
SCRUBBER (UNIVERSITY OF WASHINGTON, 1974 TO 1975)
The IERL-RTP studies done by TRW established the feasibility of charged
droplet augmentation to improve the collection efficiency of scrubbers.
It was then desirable to have a trailer-mounted mobile electrostatic
wet scrubber for testing at coal-fired electric generating plants.
IERL-RTP contracted with the University of Washington to build and
field test such a portable scrubber. The project description read:
"UW proposes to expand the technology gained from prior wet
scrubber experience by building and testing a 1000 acfm
portable scrubber. The UW wet electrostatic scrubber oper-
ates in the 120 to 200°F temperature range and is designed
for the following characteristics: gas AP= -0.5 in H20,
liquid to gas flow rate ratio = 1-5 gal/1000 acf, liquid
AP= 40-80 psig (spray nozzles). Analysis of the data col-
lected will be made to determine the design requirements
for a larger (perhaps full-sized) electrostatic wet scrub-
ber system. Simultaneous inlet and outlet measurements
will be made of the gas properties (velocity, temperature,
moisture, pressure) and the particle properties (mass con-
centration, size distribution, charge/mass ratio). The
scrubber operating parameters also will be recorded for
each test."
92
-------
FLUX FORCE/CONDENSATION (FF/C) SCRUBBER PILOT
PLANT (APT, INC., 1974 TO 1975)
Following APT's previous theoretical and bench-scale FF/C scrubber
studies, it was decided to have them conduct a 5000 to 10,000 cfm pilot
demonstration of an FF/C system. APT was to test the pilot scrubber on
an industrial source of particulate emissions selected jointly by APT
and IERL-RTP. The scope of APT's responsibilities was defined in the
IERL-RTP project description:
"APT shall conduct a literature survey to determine alterna-
tive FF/C scrubber designs and shall design the pilot scrub-
ber based upon the alternatives considered. APT shall prepare
a test plan to document the performance of the scrubber system
on a selected industrial source. The testing will deter-
mine the particle collection efficiency of the system
over the size range of 0.01 to 10 um, the overall mass
efficiency, the operational reliability, and the econom-
ics of the system. From the results of the pilot scale
tests, APT will prepare an estimate of the cost of an
optimum, full-size FF/C scrubber system."
EVALUATION OF WET SCRUBBERS FOR CONTROL OF PARTICULATE EMISSIONS FROM
UTILITY BOILERS (METEOROLOGY RESEARCH, INC., 1974 TO 1975)
Complete evaluations of several existing scrubber systems at coal-fired
utility boilers was the objective of this study. In addition to utilities
companies' interest in scrubbers as S02 cleaners, they require performance
and economic data on scrubbers applied to remove fly ash. To fill this
need, IERL-RTP contracted with Meteorology Research, Inc. to perform the
testing and evaluation of scrubbers at several selected sites. The
IERL-RTP project description was:
"Meteorology Research shall test a wet scrubber system on
a coal-fired utility boiler to determine scrubber overall
mass efficiency and fractional efficiency, ash composition,
etc., and shall collect all performance, system operating,
system design and cost data necessary to evaluate the
performance and economics of the system."
93
-------
OPERATION OF EPA-OWNED MOBILE UNITS, AERODYNAMIC TEST CHAMBER, AND
PILOT SCRUBBERS (MONSANTO RESEARCH CORPORATION, 1974 TO 1975)
IERL-RTP contracted with Monsanto Research Corporation (MRC) to provide
personnel, materials, services, and auxiliary equipment to conduct test-
ing programs with EPA-owned equipment. MRC undertook responsibility for
all phases of field testing on industrial sources using this mobile equip-
v
ment. The scope of work which MRC is to provide was defined by IERL-RTP.
''MRC will operate the following EPA-owned equipment: Mobile
Fabric Filter Unit, Mobile Wet Scrubber Unit, Mobile ESP,
Mobile Cyclone Facility, Aerodynamic Test Facility, and 862
Model Pilot Scrubber. MRC will provide maintenance for
the mobile facilities as well as the associated process
equipment and vehicles. MRC will use the Aerodynamic
Test Facility to conduct tests on various devices and
dusts. MRC will prepare a test plan for each mobile unit."
ROCKET MOTOR EMISSIONS CONTROL (APT, INC. AUGUST TO OCTOBER 1974)
As part of an inter-agency agreement between EPA and the U.S. Air Force,
APT, Inc., studied current systems for removing toxic gases and partic-.
ulate matter from rocket exhausts during ground-based static test firing-.
The scope of work assigned by IERL-RTP was to:
"... survey the literature and contact installations
where high-temperature gas stream control systems are
in effect, to gather data and develop mathematical
models for fine particulate removal from high-speed,
high-temperature gas streams. An analysis of each con-
trol system will be made to determine its operating
characteristic efficiency, and applicability to the
USAF facilities. Variables to be investigated include
gas velocity, gas temperature, gas composition, type
of fuel used, test-stand geometry, and particle
collectability."
94
-------
ENTRAINMENT CHARACTERISTICS OF MOBILE BED SCRUBBERS
(APT, INC. SEPTEMBER 1974 TO DECEMBER 1974)
The quantity, drop size distribution, and flow pattern of liquid entrain-
ment leaving a scrubber is important information for designing entrain-
ment separators, as pointed out in APT's earlier work. IERL-RTP contracted
with them to study experimentally and theoretically the entrainment prop-
erties of mobile-bed (turbulent contacting absorber, TCA) scrubbers, TCA
support grids without balls, and plate-type scrubbers in power plants.
Using the EPA-owned mobile-bed scrubber (developed by APT), APT was to
conduct tests in conjunction with selected sources. IERL-RTP defined
APT's task as follows:
"APT shall assemble the pilot scale facility, collect
experimental data, and analyze the results. The entrain-
ment is dependent upon the scrubber geometry, gas flow
rate, liquid flow rate, suspended solid concentration,
and height above the gas/liquid contact zone. The basic
pilot plant will consist of a filter, blower, scrubber,
observation section, liquid tanks, pumps, etc. APT shall
design and construct an entrainment measurement system
using techniques developed previously under contract no.
68-02-0637. These techniques involve the use of treated
filter papers for droplets over 15 urn in diameter and
impactors for sizes below 15 urn."
EVALUATION OF HORIZONTAL SCRUBBER (APT, INC., JANUARY 1975 TO MAY 1975)
A relatively new concept in scrubbing technology is the horizontal spray
scrubber. APT, Inc. was assigned the task by IERL-RTP of theoretically
and experimentally, on a small scale, studying the collection efficiency
of horizontal spray scrubbers. IERL-RTP's task outlined was for APT to:
"Critically evaluate the available literature from 1969 to
the present on horizontal scrubbers.
"Assess the importance of inertial impaction, condensation
effects, and Brownian diffusion for particle collection in
horizontal scrubbers.
95
-------
"Develop mathematical models describing particle collection
in horizontal scrubbers which include parameters describing
the effects of scrubber operating and design factors and
important particle collection mechanisms."
EVALUATION OF SYSTEMS FOR CONTROL OF ROCKET MOTOR TEST PAD EMISSIONS
(APT, INC., FEBRUARY 1975 TO AUGUST 1975)
Primary emissions from rocket motor test firings needing control are the
gases CO, HCl, and HF and A^O- particulate. The Air Force Rocket
Propulsion Laboratory (AFRPL) requested that a study be made of existing
scrubbing systems to determine what equipment should be used to clean
rocket exhausts at test firing. AFRPL will conduct scrubbing tests on
its firing pads according to APT's recommendations. IERL-RTP specified
that:
"The contractor will contact the Air Force Rocket Propulsion
Laboratory at Edwards AFB, California, for additional inform-
ation on the emissions from rocket motors. The contractor
will design experiments to be conducted by the Air Force
Rocket Propulsion Laboratory to obtain data needed to prepare
preliminary design of the scrubbing system. The contractor
will evaluate the experimental data. The contractor will then
relate his findings to the problems of scrubbing the exhaust
produced by firing solid fuel rocket motors ranging in thrust
from 5,000 pounds to 450,000 pounds. The contractor will pay
particular attention to the problems of entrainment separation."
EVALUATION OF ELECTROSTATIC SCRUBBER (APT, INC., APRIL 1975 TO MAY 1975)
Air Pollution Systems, Inc. developed an electrostatic scrubber which
IERL-RTP studied briefly via APT, Inc. to determine whether further
investigation was merited. The evaluation requested by IERL-RTP. had
this scope:
"Complete a limited literature search to determine if similar
devices have been .developed or proposed by others. Determine
if the device is truly novel and define the mechanisms which
are responsible for particulate capture. Determine the reli-
ability and significance of any experimental data submitted.
Assess the practicality of the device for collection of fine
particulate. Estimate the capital and operating costs for
96
-------
the device evaluated.. Determine which sources might be con-
trolled by the new system and estimate the probability of
successful application.
"This evaluation is a paper study only and does not require
testing by the contractor.
"A recommendation will be made as to whether the device
should be tested by EPA. If the recommendation is to test
the device, a recommendation for the testing procedures to
be used will be provided."
EFFECTS OF INTERFACIAL PROPERTIES OF FINE-PARTICLE SCRUBBING
(METEOROLOGY RESEARCH INSTITUTE, JUNE 1975)
Introduction
In the development of engineering design methods for the evaluation of
scrubber systems, mathematical models to date (the Scrubber Handbook,
for example) have tended to ignore the effects of particle-liquid
interfacial properties. That is, collection efficiency is typically
equated with "target" or collision efficiency, tantamount to assuming
that all particles contacting a scrubber drop are retained by the drop.
The interfacial properties may be such as to inhibit the retention of
nonwettable particles, however. There are theoretical results showing
that a nonwettable particle may bounce off the drop under certain con-
33
ditions, such as low impingement velocity. However, experimental
investigations have been subject to dispute. (See References 26-31).
In view of the dearth of data and its possible significance for improve-
ment of scrubber design for finer particle collection, IERL-RTP is ini-
tiating a program to study this problem. This will supplement the work
being done by.Southern Research Institute on the growth of hygroscopic
aerosols in scrubber environments.
97
-------
Goals
The scope of work defined by IERL-RTP includes effectively the following;
1. Literature review and evaluation, with .identification of
weak areas, regarding the nature of particle-liquid inter-
face interactions and how they might affect particle
retention in scrubbing. The literature is likely to
extend into other fields such as cloud physics and pre-
cipitation scavenging.
2. Development of theoretical models of particle retention
as functions of particle size and their properties gov-
erning the surface interaction.
3. Well-defined experiments to test the theoretical models,
using fine, relatively monodisperse aerosol particles.
4. Assessment of the results of the theory and experiments
with regard to the relative importance of interfacial
effects in scrubbing.
5. Identification of remaining problem areas and the nature
of further analysis and measurements to be done.
6. Recommendations for applying the results to enhance the
design of scrubbers for particulate pollution control.
MOBILE BED FLUX FORCE/CONDENSATION (FF/C) SCRUBBERS FOR
COLLECTION OF FINE PARTICLES (CONTRACT IN NEGOTIATION)
Introduction
Previous contract efforts showed that FF/C scrubbing was a feasible con-
cept. In earlier work, mathematical models and design equations were
developed for a sieve plate scrubber and a spray scrubber which make use
of diffusiophoresis and particle growth due to water vapor, condensation.
Significant advantages would accrue if it were possible to combine com-
monly used mobile bed scrubbers for SO with particulate collection by
X
the same, device. IERL-RTP is therefore in the process of negotiating
a contract for an in-depth study of this project.
98
-------
Goals
The objectives of the proposed study would be to:
1. Determine feasibility of using mobile bed scrubbers as
Flux Force/Condensation Scrubbers.
2. Compare mobile bed FF/C scrubber with the spray and
sieve plate FF/C scrubber.
Methods
The recommended approach to achieving the study objectives is:
1. Conduct preliminary tests of collection efficiency with
.a mobile bed scrubber.
2. Design and construct an entrainment separater for the pilot
mobile bed scrubber (gas flow rate > 1000 acfm, 0.5 m3/s).
3. Undertake an experimental study of FF/C scrubbing in
mobile bed scrubbers. Parameters of interest are:
number of scrubber stages, condensation ratio, liquid-
to-gas flow rate ratio, bed height, pressure drop,
particle diameter, particle concentration, gas velocity,
liquid entrainment. Run tests with no condensation.
4. Use above results to develop mathematical models and
design equations. Compare predictions with the
reported performance of the mobile-bed TCA Scrubber in
EPA's Alkali Scrubbing Test Facility at TVA Shawnee
Power Plant.
5. Recommend an FF/C scrubber configuration for pilot
demonstration based on critical technical and economic
comparison of studied scrubbers.
99
-------
EFFECTS OF ENERGY CONSUMPTION AND ENERGY SUPPLY MODE ON SCRUBBER
OPERATION (STANFORD RESEARCH INSTITUTE, JUNE 1975)
Introduction
Wet scrubbers may have energy supplied to them in a variety of ways. The
energy consumption of a scrubber per unit volume of gas treated and
whether some of this energy may be taken at no cost from process heat
will have great bearing on the economic justification of a given scrubber,
Consequently, IERL-RTP has negotiated a contact to study the effect of
energy consumption and supply mode on scrubber collection efficiency
and economy.
Goals
The goals of the study will be to:
1. Determine effects of energy utilization and the varied
means of supplying energy on collection efficiency of
wet scrubbers.
2. Recommend a research and development program to achieve
scrubbers which make optimum use of energy.
Other objectives to be reached are specified in the IERL-RTP scope
of work:
1. "Identify the mechanisms by which the energy supplied
is utilized to collect particles.
2. "Identify other energy consuming mechanisms, e.g.,
friction.
3. "Develop quantitative relationships between the energy
supplied to the scrubber and the particle collection
efficiency.
4. "Explain any differences in the energy-particle col-
lection efficiency relationships that may exist for
different types of scrubbers."
100
-------
Methods
The study will include a literature survey of the energy consumption and
supply problem in scrubbers. A bench scale experimental effort should
be made to evaluate the following scrubber types within the above frame-
work :
1. Scrubbers which derive contacting power from the avail-
able gas stream.
2. Scrubbers which utilize mechanical rotors.
3. Scrubbers which derive contacting power from a
liquid stream.
Controlled, monodisperse, nondeliquescent test aerosols (at least three
different sizes below d = 3 ym'
P
scrubbing will be investigated.
different sizes below d = 3 ym) will be employed. The use of series
101
-------
SECTION VI
REFERENCES
1. Mergenthaler, H. and D. Keller. Developments in Wet Scrubbers.
Staub (English) 26(6):1-4. June 1966.
2. Particulate Pollutant System Study. Vol. II. Fine Particulate
Emissions. Midwest Research Institute. Prepared for U.S. EPA.
August 1971.
3. Calvert, S. Engineering Design of Wet Scrubbers. J APCA 24:929-934.
1974.
4. Calvert, S., J. Goldshmid, D. Leith, and D. Mehta. Scrubber Hand-
book. Vol. I. APT, Inc. Riverside, California. Report Numbers
NTIS PB-213016, EPA-R2-72-1182. Environmental Protection Agency.
July 1972.
5. Semrau, K. and C. L. Witham. Wet Scrubber Liquid Utilization.
Stanford Research Institute. Menlo Park, California. Report
Number EPA-650/2-74-108. Environmental Protection Agency.
October 1974. 116 p.
6. Fuchs, N. The Mechanics of Aerosols. New York, Pergamon Press.
1964. 408 p. .
7. Cooper, D. W. Fine Particle Control by Electrostatic Augmentation
of Existing Methods. Paper #75-02.1. Presented at the 68th Annual
Meeting of the Air Pollution Control Association. Boston. June 1975.
8. Melcher, J. R. and K. S. Sachar. Charged Droplet Scrubbing of Sub-
micron Particulate. Draft Final Report to EPA for Contract No.
68-02-0250. July 1974.
9. Lear, C. W., W. F. Krieve, and E. Cohen. Charged Droplet Scrubbing
for Fine Particle Control. J APCA 25(2):184-189. February 1975.
10. Pilat, M. J. Collection of Aerosol Particles by Electrostatic Droplet
Spray Scrubbers. J APCA 25(2):176-178. February 1975.
11. Sparks, L. E. and M. J. Pilat. Effect of Diffusiophoresis on Par-
ticle Collection by Wet Scrubbers. Atmos Environ 4:651-660. 1970.
102
-------
12. Lancaster, B. W. and W. Strauss. A Study of Steam Injection Into
Wet Scrubbers. Ind Eng Chem Fundamentals. 10(3):362-369.
March 1971.
13. Fuchs, N. and A. Kirsch. The Effect of Condensation of a Vapor
on the Grains and of Evaporation From Their Surface on the Depo-
sition of Aerosols in Granular Beds. Chem Eng Sci 20:181-185.
1965.
14. Calvert, S., J. Goldshmid, D. Leith, and N. C. Jhaveri. Feasibility
of Flux Force/Condensation Scrubbing for Fine Particulate Collection.
APT, Inc. Riverside, California. Report Number EPA-650/2-73-036.
Environmental Protection Agency. 1973.
15. Semrau, K. T. and C. L. Witham. Condensation and Evaporation
Effects in Particulate Scrubbing. Paper # 75-30.1. Presented
at 68th Annual Meeting of the APCA. Boston, Mass. 1975.
16. Schauer, P. J. Removal of Submicron Aerosol Particles From a
Moving Gas Stream. Ind Eng Chem 43(7):1532-1538. July 1951.
17. Waldmann, L. and K. H. Schmitt. Thermophoresis and Diffusiophoresis
of Aerosols. In: Aerosol Science, Davies, C. N. (ed,.) . New York,
Academic Press, 1966. p. 137-161.
18. Lapple, C. E. and H. J. Kamack. Performance of Wet Dust Scrubbers.
Chem Eng Prog. 51(3):110-121. March 1955.
19. Semrau, K. T., C. W. Marynowski, K. E. Lunde, and C. E. Lapple.
Influence of Power Input on Efficiency of Dust Scrubber. Ind
Eng Chem 50(11):1615-1620. November 1958.
20. Slinn, W. G. N. and J. M. Hales. A Reevaluation of the Role of
Thermophoresis as a Mechanism of In- and Below-Cloud Scavenging.
J Atmos Sci 28:1465. 1971.
21. Goldsmith, P. and F. G. May. Diffusiophoresis and Thermophoresis
in Water Vapor Systems. In: Aerosol Science. Davies, C. N. (ed.).
New York, Academic Press. 1966. p. 163-194.
22. Calvert, S. and N. C. Jhaveri. Flux Force/Condensation Scrubbing.
J APCA 24(10):947-952. October 1974.
23. Nannen, L. W., R. E. West, and F. Kreith. Removal of S02 From Low
Sulfur Coal Combustion Gases by Limestone Scrubbing. J APCA 24:29-39.
1974.
103.
-------
24. Calvert, S., I. L. Jashnani, S. Yung, and S. Stalberg. Entrainment
Separators for Scrubbers - Initial Report. APT, Inc. Riverside,
California. Report No. EPA-650/2-74-119-a. Environmental Protection
Agency. October 1974.
25. Statnick, R. M., and D. C. Drehmel. Fine Particle Control Using
Sulfur Oxide Scrubbers. J APCA. 25:605-609. June 1975.
26. Berg, T. G. 0., T. A. Gaukler, and L. A. Squier. Kinetics of Wet-
ting in Washout of Dust. Aerojet - General Corp., Downey, California.
In: Radioactive Fallout From Nuclear Weapons Tests. U. S. Atomic
Energy Commission. 1965. 281-300 p.
27. Engelmann, R. J. A Surface Energy Model in the Raindrop Wetting of
Scavenged Particles. Abstracted from: Report No. HW 79382.
U.S. Atomic Energy Commission. Hanford, Washington. 1963.
28. Rayleigh, Lord. The Influence of Electricity on Colliding Water
Drops. Proc Roy Soc 28:406. 1879.
29. Fuchs, N. A., F. I. Murashkevich, and A. D. Raikhin. Preliminary
Report on the Efficiency of Collisions Between Dust Particles
and Water Droplets. Staub 33(4):182-183. April 1973.
30. Adam, N. K. The Physics and Chemistry of Surfaces. London,
Oxford Univ Press. 1930. 402 p.
31. Berg, T. G. 0. The Mechanism of Washout. Aerojet - General Corp.
Downey, California. Report No. 0780-01(02)SP. 1963. 44 p.
32. McCully, C. R., M. Fisher, G. Langer, J. Rosinski, H. Glaess, and
D. Werle. Scavenging Action of Rain on Airborne Particulate Matter.
Ind'Eng Chem 48:1512-1516. 1956.
33. Pemberton, C. S. Scavenging Action of Rain on Nonwettable Particu-
late Matter Suspended in the Atmosphere. In: Aerodynamic Capture
of Particles, Richardson, E. G. (ed.). London, Pergamon Press.
1960. ' 200 p.
34. McDonald, J. E. Rain Washout of Partially Wettable Insoluble
Particles. J Geophys Res 68(17) :4993-5003. September 1963.
35. Rozen, A. M., and V. M. Kostin. Intern Chem Eng 7:464. 1967.
36. Calvert, S., N. C. Jhaveri, and S. Yung. Fine Particle Scrubber
Performance Tests. APT, Inc. Riverside, California. Report No.
EPA-650/2-74-093. EPA. October 1974.
37. Kotov, N. A. Influence of Spray Entrainmeat on Gas Scrubber Effi-
ciency. Translated by EPA from Vodosnabzheniye: Sanitaria
Tekhnika. 7:31-34. 1972.
104
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
1. REPORT NO.
EPA-600/2-75-054
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Overview of EPA/IERL-RTP Scrubber Programs
5. REPORT DATE
September 1975
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Douglas W. Cooper, Lee W. Parker, and
Eugene Mallove
8. PERFORMING ORGANIZATION REPORT NO,
GCA-TR-75-22-G
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA Corporation
GCA/Technology Division
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO.
1AB012; ROAP 21ADL-002
11. CONTRACT/GRANT NO.
68-02-1316, Task 10
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
Task Final; 3-7/75
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT Tne rep0r{ gjves. an introduction to fine particle control scrubber programs
supported by EPA's Industrial Environmental Research Laboratory at Research
Triangle Park in North Carolina (IERL-RTP); and topical and chronological over-
views of the various projects which IERL-RTP has sponsored relating to scrubber
technology. Having reviewed the state-of-the-art which culminated in the Scrubber
Handbook, IERL-RTP sponsored scrubber tests which generally validated the models
presented in the Handbook. Possible exceptions (mobile-bed scrubbers and series
scrubbing with low-energy devices) became objects of further study. Two major
mechanisms for increasing scrubber efficiency have been studied by IERL-RTP con-
tractors: flux-force/condensation scrubbing (use of diffusiophoresis and condensation
growth of particles), and electrostatic augmentation (scrubbing with charged drop-
lets); both have shown efficiencies which are higher than those achieved by conven-
tional means at equal power consumption levels. Problems with entrained water
droplets have caused IERL-RTP to fund theoretical and experimental research in
entrainment separator technology, some of the results of which are presented. The
hypothesis that particle wettability is important in determining scrubbing efficiency
is to be tested by an IERL-RTP sponsored research program.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Scrubbers
Electrostatics
Tests
Air Pollution Control
Stationary Sources
Fine Particulate
13 B
07A
20C
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport!
Unclassified
21. NO. OF PAGES
112
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
105
------- |