EPA-600/2-77-110
June 1977
Environmental Protection Technology Series
            APPLICATION OF FOAM SCI
                     TO  FINE PARTICLE CONTROL,
                                              PHASE  II

                              industrial Environmental Research laboratory
                                    Office of Research and Development
                                   U.S. Environmental Protection Agency
                              Research Triangle Park, North Carolina 27711

-------
                    RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental Protec-
tion Agency, have been grouped into nine series. These nine broad categories were
established to facilitate further  development and application of environmental tech-
nology. Elimination of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The nine series are:

          1. Environmental Health Effects Research
          2. Environmental Protection Technology
          3. Ecological Research
          4. Environmental Monitoring
          5. Socioeconomic Environmental Studies
          6. Scientific and Technical Assessment Reports (STAR)
          7. Interagency Energy-Environment Research and Development
          8. "Special" Reports
          9. Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumen-
tation, equipment, and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the new or improved tech-
nology required for the control and treatment of pollution sources to meet environmental
quality standards.
                             REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved for
publication. Approval does not signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.

-------
                                              EPA-600/2-77-110
                                                     June 1977
APPLICATION  OF FOAM  SCRUBBING
     TO  FINE  PARTICLE  CONTROL,
                     PHASE  II
                           by

                     T.E. Ctvrtnicek, S.J. Rusek,
                    C.M. Moscowitz, and L.N. Cash

                    Monsanto Research Corporation
                      1515 Nicholas Road
                      Dayton, Ohio 45407
                     Contract No. 68-02-1453
                      ROAP No. 21ADL-029
                    Program Element No. 1AB012
                 EPA Project Officer: Geddes H. Ramsey

                Industrial Environmental Research Laboratory
                 Office of Energy, Minerals, and Industry
                Research Triangle Park, North Carolina 27711
                        Prepared for

                U.S. ENVIROMENTAL PROTECTION AGENCY
                  Office of Research and Development
                     Washington, D.C. 20460

-------
                           ABSTRACT
This report summarizes the knowledge, the experience, and the data
gained to date relative to the application of foam scrubbing to
collection of fine particles from gaseous streams.  Experimental
data obtained on a 0.236-m3/s  (500-cfm) pilot-scale foam scrubber
facility are presented.  Economic analysis indicates that a foam
scrubber can be competitive with other fine particle collection
devices.  Areas for further foam scrubber development are recom-
mended .

This report was submitted in fulfillment of Contract No. 68-02-
1453 by Monsanto. Research Corporation under the sponsorship of
the U.S. Environmental Protection Agency.  This report covers a
period from 1 July 1975 to 31 July 1976, and work was completed
as of 31 July 1976.
                               111

-------
                            CONTENTS
Abstract                                                      iii
Figures                                                        vi
Tables                                                       viii

1.   Summary                                                    1
2.   Conclusions and Recommendations                            2
3.   Introduction                                               4
4.   Experimental Apparatus                                     5
     4.1  Design Considerations                                 5
     4.2  The Pilot Foam Scrubber Facility                      9
5.   Experimental Results                                      12
     5.1  Foam Scrubber Scaleup                                12
     5.2  Scrubber Residence Time                              13
     5.3  Particle Collection Experiments                      15
6.   Economics                                                 22
     6.1  Accuracy of Estimates                                22
     6.2  Baseline Collection Efficiencies                     24
     6.3  Results                                              26

References                                                     36
Appendices
     A.   Description of Pilot Facility Subsystems             37
     B.   Experimental Data on Foam Destruction                45
     C.   Troubleshooting of the Pilot-Scale Foam Scrubber     47
     D.   Fine Particle Collection Data and Foam Scrubber      55
          Operating Conditions for the 20-Hour Experimental
          Run
                                v

-------
                             FIGURES
Number

   1  Typical number concentration of fine particles
      from a 12-MW coal-fired boiler unit at 92% load.        7

   2  Particle collection efficiencies applicable
      to collection of particles by foam scrubber.            8

   3  Experimental process block diagram of pilot-
      scale foam scrubber.                                   10

   4. Pilot-scale foam scrubber process layout.              11

   5  Approximate scrubber velocity profile at steady
      state.                                                 14

   6  Average particle collection efficiency data (12 runs).
      Percent collection shown inside the unshaded area.     17

   7  Concentration of solids in scrubbing liquor as a
      function of time.                                      18

   8  Extrapolated fine particulate control efficiencies.    25

   9  Capital cost for particulate control devices.           31

  10  Operating cost for particulate control devices.        33

 A-l  Jet mill modified for dispersing fly ash fines.        38

 A-2  Foam generation screen arranged in a zigzag
      fashion.                                               40

 A-3  A view of the foam destruction chamber.                41

 A-4  Electrical mobility analyzer used for foam scrubber
      particle collection efficiency evaluations.            43

 B-l  Power requirements for foam destruction.               46

 C-l  Aerosol stability.                                     52
                               VI

-------
                       FIGURES  (continued)

Number                                                       Page

 C-2  Average foam baseline counts  (11 runs).  Scrubber
      residence time:  20.5 seconds.  Surfactant concen-
      tration:  0.25% Tergitol.  Disk speed:  4,000-5,000
      rpm.                                                    53

 D-l  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  1) .                        56

 D-2  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  2).                        57

 D-3  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  3).                        58

 D-4  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  4).                        59

 D-5  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  5).                        60

 D-6  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  6).                        61

 D-7  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  7).                        62

 D-8  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  8).                        63

 D-9  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  9).                        64

D-10  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  10).                       65

D-ll  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  11).                       66

D-12  Particle inlet and outlet concentrations and
      collection efficiencies  (Run  12) .                       67
                               VII

-------
                             TABLES


Number                                                      Page

   1  Cost Assumptions                                       27

   2  Foam Scrubbing Assumptions                             28

   3  Electrostatic Precipitations Assumptions               29

   4  Fabric Filter Assumptions                              29

   5  High Energy Wet Scrubber Assumptions                   30

   6  Foam Scrubbing Capital Costs as a Function of
      Scrubber Residence Time, Expressed as Percent of
      Total Purchase Cost                                    32

   7  Foam Scrubbing Operating Costs for 40-Second
      Residence Time as a Function of Capacity,
      Expressed as Percent of Total Operating Cost           32

 D-l  Ranges and Averages for Important Parameters of the
      Foam Scrubber Operations, 20-Hour Demonstration        55
                               Vlll

-------
                           SECTION 1

                            SUMMARY
This report summarizes the results achieved while investigating
the application of foam for removal of fine particles from gaseous
phase.  In the first phase of the program, the investigation was
initiated by theoretical evaluations of collection mechanisms in-
volved in fine particle collection by foam.  Then, the theoretical
findings were verified on a bench-scale foam scrubber apparatus
(1.2 x 10~3 m3/s; 2.5 cfm).1  Utilizing the experience gained
during the bench-scale verification, a pilot scale (0.236-m3/s;
500-cfm) foam scrubber was erected and operated.  Significant im-
provements in the scrubber operation and economics were achieved
during the pilot-scale testing and demonstration.  These improve-
ments verify the foam scrubber as a viable device for removal of
fine particles from gas streams, and warrant its further develop-
ment.  Specific areas needing further investigation and testing
are outlined.

Section 2 summarizes the conclusions and recommendations made on
this program.  After the introduction in Section 3, the pilot-
scale foam scrubber facility is described in Section 4.  Section
5 summarizes the experimental results achieved during the pilot
demonstration.  Finally, the foam scrubber economics are com-
pared with the economics of conventional particle collectors in
Section 6.
^tvrtnicek, T. E., R. F. Walburg, C. M. Moscowitz, and H.H.S.
 Yu.  Application of Foam Scrubbing to Fine Particle Control—
 Phase I.  EPA-600/2-76-125, U.S. Environmental Protection Agency
 Research Triangle Park, North Carolina, May 1976.  152 pp.

-------
                            SECTION 2

                 CONCLUSIONS AND RECOMMENDATIONS
These conclusions and recommendations reflect the observations
made during both phases of this program.  The reader is en-
couraged to familiarize himself with the observations made during
the program's first phase involving theoretical evaluations of
the foam scrubber collection mechanisms, verification of these
mechanisms in a bench-scale (1.2 x 10~3 m3/s; 2.5 cfm) experiment,
and preliminary economics.1  Only then can the conclusions and
recommendations presented here be truly comprehended.

1.   After demonstration on a pilot scale (0.236 m3/s; 500 cfm),
     the foam scrubber remains a viable method for removing fine
     particulates from gaseous streams.  The method is simple,
     consisting of basic operations such as pumping the liquid
     and gas, and spinning a disk to destroy the foam.  The foam
     is generated by forcing aerosol gas through a screen sprayed
     with a surfactant liquid.  Particle collection is believed
     to take place mainly by diffusion and sedimentation, mechan-
     isms that are predictable and rather well understood.

2.   Collection efficiencies demonstrated on the pilot scale
     using industrial dust were between 50% and 75% by count for
     particulates in a size range between 0.056 ym and 1.0 ym.
     These collection efficiencies are averages from an experi-
     mental operation that covered 20 hours and used a scrubber
     residence time of 20.5 seconds.  At this residence time, the
     averages are somewhat better than the theoretically predicted
     collection efficiencies  (based on particle density of 2
     g/cm3, bubble diameter of 2 mm, and temperature of 23°C),
     ranging from 32% to 67% by count for particles in the size
     range of 0.1 ym to 1.0 ym.  A further increase in scrubber
     collection efficiency can be easily achieved by an increased
     scrubber residence time as demonstrated by the bench-scale
     tests.

3.   Significant improvements in the foam scrubber operation were
     achieved due to the scrubber scaleup and the experience
     gained in operating the scrubber for longer periods.  These
     improvements include the ability to generate satisfactory
     foam with (1) a surfactant concentration of 0.25% (1% to 2%
     on bench scale), (2) a scrubbing liquor feed rate of 6.83 x
     10-* m3/s (10.8 gpm) per 0.259 m3/s (550 cfm) of gas (this

-------
     is equivalent to 19.6 gallons per 1,000 cfm versus 38.4
     gallons per 1.000 cfm achieved on the bench scale), and
     (3) a scrubbing liquor which was continuously recycled and
     contained up to 2 kg/m3 of solids.  In addition, the pilot
     foam scrubber operated at a lower pressure drop of 1.84 kPa
     (7.4 in. H20; 9 in. H20 on bench scale), and the power con-
     sumption for the foam destruction was reduced by a factor
     over 400 (847 watts/m3 of foam on bench scale versus 2.1
     watts/m3 of foam on pilot scale).  The improvements in the
     foam scrubber operations, however significant, were realized
     without an organized and conscientious effort to optimize
     foam generation and foam destruction steps.  They are a
     result of an organized and conscientious effort to demon-
     strate the foam scrubber as a viable system on a pilot
     scale.  It is therefore believed that further improvements
     of the foam scrubber are yet possible.

4.   Foam and foam scrubber liquor can be restored to their
     original condition by filtration.

5.   Economic evaluations based on the results of this program
     indicate that the foam scrubber can be made competitive in
     terms of capital investment and operating costs with conven-
     tional particle collectors including fabric filter, high
     energy scrubber, and high efficiency electrostatic precipita-
     tor.

Based on the favorable results obtained during this two-phase pro-
gram, it is recommended that further studies of the foam scrubber
be made.  These studies should be performed on a pilot or larger
scale foam scrubber unit and should primarily concentrate on the
following areas:

   • Optimization of the foam generation step with respect to
     foam generation screen size, surfactant concentration,
     characteristics of particles in gas feed (size, solubility,
     surface properties), and scrubber liquor spray (surfactant
     liquor recycle under no-leak conditions, regeneration of the
     liquor through removal of collected solids, and liquor spray
     distribution and rate).

   • Optimization of the foam destruction step with respect to
     minimization of secondary aerosol formation," consumption of
     energy, and design for large-scale applications.

   • Verification of foam scrubber operations under conditions
     actually existing in the field such as elevated temperature
     and presence of gas contaminants (SO , NO , C02r CO, HC,
      i_   \                                '^    ^*
     etc.).

-------
                           SECTION 3

                         INTRODUCTION
A variety of air pollution equipment is available to industry for
containing solid particulate emissions  (e.g., electrostatic pre-
cipitators, baghouses, inertial separators, wet scrubbers, set-
tling chambers, impingement separators, and panel filters).  In
terms of mass removal efficiency, a majority of the commercially
available particulate control devices are adequate, with mass re-
moval efficiencies up to +99%.  However, submicrometer particles
are not captured by this equipment; they pass readily into the
atmosphere, creating hazards to human health, as well as visibil-
ity and smog problems.

New concepts and technology for the control of fine particles are
needed to help remedy this situation.  In this category is foam
scrubbing.  Prior investigations, performed under the first phase
of this contract, established theoretical and technical feasibil-
ity for the foam scrubber.1.  The mechanisms involved in collecting
particles via a foam scrubber were defined and quantified.  Theo-
retical collection efficiencies were then verified experimentally
on a 1.2 x 10~3 m3/s  (2.5 cfm) bench-scale apparatus.  Experimen-
tal results agreed well with those theoretically calculated.

Preliminary and limited economic evaluations indicated that the
foam scrubber could be competitive in terms of capital investment
costs with conventional particle collection devices, including
fabric filter, high energy scrubber, and high efficiency electro-
static precipitator.  However, the operating costs for the foam
scrubber appeared substantially higher than those required for
the aforementioned conventional devices, mainly due to the cost
of surfactant.  Based on the encouraging bench-scale experimental
results and the lack of data on larger scale, it was recommended
that foam scrubbing be evaluated on a pilot scale.  This evalua-
tion was done during the second phase of this contract, and the
results achieved are presented in this report.

-------
                           SECTION 4

                    EXPERIMENTAL APPARATUS
This phase of the program involved design, erection, and opera-
tion of a pilot-scale foam scrubber, scaled up by a factor of 200
from the bench-scale device developed in Phase I.  The three
major objectives of this phase were to:  1) investigate the feasi-
bility of improving the process economics by recycling foam liquor
solution continuously, 2) determine scrubber performance with a
real-life solid fine particulate aerosol, and 3) determine the
effects of scale up on scrubber performance and economics.

4.1  DESIGN CONSIDERATIONS

The primary scaleup requirement was a gas-handling capacity of
0.236 m3/s (500 cfm).  Other important criteria considered early
in the design stages of this equipment were inlet gas conditions,
and aerosol type and loading.  One major objective of the second
phase was to evaluate scrubber performance on solid particulate
aerosols from an existing source (e.g., powerplant, foundry, pulp
and paper mill, etc.), simulating as closely as possible an in-
dustrial stack containing fine particulates.  This departed from
bench-scale experiments where "ideal" liquid droplet aerosols
were generated and used.  To increase the chances for successful
simulation of an industrial effluent, several practical attributes
were required of the particulate dust that was to be used in pilot-
scale demonstration experiments.  The dust had to: 1) contain a
large mass concentration of fine particles  (<3 ym), 2) be repre-
sentative of an actual emission source, 3) be nonmagnetic so as
to not to interfere with particle sampling instruments, 4) be uni-
form in composition, and 5) have desirable feeding and handling
qualities (nontoxic, noncaking, easily dispersed, insoluble,
etc.) .

After considerable searching, a suitable dust was found.  The
source of this dust is a stoker and coal-fired powerplant, in
Nucla, Colorado, operated by the Colorado Ute Electric Associa-
tion (Nucla Plant).  Details concerning the origin and properties

-------
of this dust were previously described.2'3  In the Sem2 study,
the Nucla Plant was sampled for submicrometer particles before
and after the baghouse collector on boiler unit 2.  Figure 1
shows typical loadings of fine particles between about 0.05 ym
and 1.0 ym in diameter upstream of the baghouse particle collec-
tor.  These data are averages of 19 runs taken during 35 minutes
of boiler operation at near full load on two separate days.

The data in Reference 2 were taken during relatively constant
boiler operation.  Even so, the concentrations of fine particles
fluctuated by as much as ± 93% in the particle size range of the
lowest particle loading (0.562-1.0 ym) to ± 31% in the smallest
particle size range.  The concentration of total particles for
all size ranges  (0.0562-1.0 ym) fluctuated by ± 45%.  It should
be noted that although nearly 90% of the particles by number in
the stack are less than 0.18 ym in diameter.  These same particles
constitute only 18% of the total mass loading.

The last important design criterion was scrubber residence time.
As previously established, particle collection efficiency is a
strong function of the time the particles reside within the foam.1
Figure 2 shows theoretical trends for one set of experimental con-
ditions (bubble diameter,  2 mm: particle density, 2 g/cm3; atmo-
spheric pressure; and temperature, 23°C).  Curves for other sets
of conditions were also calculated.1  As shown in Figure 2, the
particle collection efficiencies follow curves that steadily in-
crease with time and asymptotically approach 100%.  After a cer-
tain time, the gain in particle collection efficiency that would
result with a further increase of residence time becomes rather
negligible, and gets within the range of experimental error.  To
operate the pilot scrubber in this region would greatly reduce
the possibilities to observe the effect of influential variables
on the scrubber collection efficiency.  Consequently, the pilot
scrubber was designed for a shorter residence time of between 20
and 30 seconds where the particle collection curve is relatively
steep and changes in collection efficiency can be observed more
readily.  Operating the pilot foam scrubber at these residence
times meant that high scrubber collection efficiencies obtainable
with longer scrubber residence times were sacrificed for the
2Sem, G. J.  Submicron Particle Sizing Experience on a Smoke
 Stack Using the Electrical Aerosol Size Analyzer.  Presented
 at the Seminar on In-Stack Particle Sizing for Particulate
 Control Device Evaluations, U.S. Environmental Protection
 Agency, Research Triangle Park, North Carolina, December 3-4,
 1975.  10 pp.

3Bradway, R. M., and R. W. Cass.  Fractional Efficiency of a
 Utility Boiler Baghouse—Nucla Generating Plant.  EPA-600/2-
 75-013-a,, U.S. Environmental Protection Agency, Research
 Triangle Park, North Carolina, August 1975.  148 pp.

-------
             E


            CO
            Q£


            £
            *t
            O
            o
            O
            HH
            I—
            ce:


            a.
40U
380
360
340
320
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
-
-
-
















45.9%










• ii»


41.5%











Z PARTICLES / cm3
= 7.15xl06±45%











10.3%
U7%, ,043%
. . .I/,
                 0.05   0.08 0.1
0.2   0.3  0.4   0.6  0.8 1.0
                          PARTICLE DIAMETER, d  \m
Figure  1.   Typical  number concentration  of  fine particles from

            a 12-MW  coal-fired  boiler unit at 92% load.2

-------
oo
                                                  BUBBLE DIAMETER -2.0mm
                                                  PARTICLE DENSITY-2 g/cm3
                                                  PRESSURE-ATMOSPHERIC
                                                  TEMPERATURE-23°C
                                                          DIFFUSION AND
                                                          SEDIMENTATION
                                                   	DIFFUSIONONLY
                                                  * CURVE IDENTICAL WITH CURVE FOR
                                                   DIFFUSION AND SEDIMENTATION
                             20
30       40       50
  RESIDENCE TIME, s
60
                 Figure 2.   Particle collection efficiencies applicable
                            to  collection of particles  by  foam scrubber.1

-------
purpose of having a more feasible pilot test facility.  This
should be borne in mind when pilot scrubber collection efficiency
data are reviewed.

4.2  THE PILOT FOAM SCRUBBER FACILITY

The pilot-scale foam scrubber process consists of the operational
steps diagrammatically shown in Figure 3.  Filtered room air is
intimately mixed with aerosolized fly ash fines and this mixture
proceeds to the foam generation section.  In this section, foam
is made by passing the air laden with particlates through a 250-
mesh stainless steel screen onto which the surfactant solution is
uniformly sprayed.  As the foam is formed, it encapsulates the
dust particles and proceeds toward the foam destruction chamber,
flowing in essentially plug-flow fashion.  The ductwork between
the foam generation section and the destruction chamber is large
enough to provide a foam scrubber residence time between 20 and
26 seconds.  The foam holdup period allows the encapsulated
particles to migrate to the bubble walls for capture.

The foam then enters the destruction chamber where it is mechani-
cally destroyed by rapidly rotating disks.  The foam liquid pro-
duced from foam destruction is returned to the foam generator,
and the cleaned air containing some residual particles and second-
ary aerosol from the foam destruction operation proceeds from the
scrubber into the induced draft fan where it is finally exhausted
outdoors.  Due to some liquor recycle pump leaks and mist losses
up the stack, a fresh surfactant solution is supplied using a
makeup liquor.

During the course of an experimental run, the inlet and outlets
of the scrubber are sampled periodically for particulate concen-
trations.

A more detailed process layout is shown schematically in Figure 4,
The process is broken down into the following subsystems:

          - dust aerosolization
          - bulk air supply
          - foam generation
          - foam destruction
          - scrubber liquor supply and recycle
          - fine particle instrumentation
          - scrubber operation

These subsystems are described in Appendix A.

-------
                                              FOAM SCRUBBER
                                                RESIDENCE
                                                 SPACE
Figure  3.   Experimental  process block diagram of pilot-scale  foam scrubber.

-------
                                      SCRUBBER EXHAUST
                                                                   ~L FLOWMCTER
Figure  4.  Pilot-scale  foam scrubber process layout.

-------
                            SECTION 5

                      EXPERIMENTAL RESULTS


In view of the conclusions reached during Phase I of this program1
and the objectives for this second phase  (see Section 4), the most
critical variable for a successful foam scrubber was the process
operating cost.  Even with 99% recycle of the scrubbing  liquor,
the operating cost of the foam scrubber was an order of magnitude
greater than that of the most expensive conventional control method
(high energy wet scrubbing).  Due to limited capabilities of the
bench-scale experiments, no experimental data were available to
demonstrate that 99%, or perhaps even higher, recycle of the scrub-
bing liquor solution is feasible.

The initial efforts of Phase II, therefore, concentrated on two
important factors:

(1)  Use of the best engineering judgment and experience to scale
     up the foam scrubber from the bench scale (1.2 x 10~3 m3/s;
     2.5 cfm gas-handling capacity) to the pilot scale  (0.236 m3/s;
     500 cfm), and

(2)  Use of this scaled up facility to investigate and demonstrate
     high recycle of the surfactant liquor while maintaining a
     scrubber particle collection efficiency close to that theore-
     tically predicted  (see Section 4.1).

Based on the results presented below, both of these factors were
implemented with a great degree of success.

5.1  FOAM SCRUBBER SCALEUP

The most important items in the scaleup of the foam scrubber were
the foam generation step, flow of the foam through ducts of large
diameters, and the foam destruction step.  All three of these are
unique with the foam scrubber and are not known to be used in in-
dustry.  Consequently, no data could be found which would suggest
and provide direction for correct scaleup.

Additional data were developed on the effectiveness of foam destruc-
tion by high-speed disks using the bench-scale apparatus.  Disks of
different shapes and sizes were tested for energy consumption at
the point where foam of constant qualities was being effectively
destroyed.  These data were then used to specify and design the

                                12

-------
foam destruction section of the pilot-scale foam scrubber.  A
brief summary of the foam destruction data is presented in
Appendix B.

After thorough engineering analysis of the foam generation
phenomena previously discussed and quantified,1 it was concluded
that a proportional scale up of the foam generation section
relative to air throughput would be proper.

Flow of the foam through ducts of large diameters was a serious
concern.  The actual scale up was based on keeping constant the
velocity experimented in the bench-scale scrubber.  Velocity was
chosen over Reynolds number because air velocity is a critical
parameter for foam generation.  An additional problem with scal-
ing up through, for example, constant Reynolds number would be the
large physical dimensions of a pilot-scale unit [diameter 19.5 m
(64 ft), height 4.8 cm  (2 in.)].  Scale up through velocity in-
volved some amount of risk since the Reynolds number for the
pilot-scale scrubber in comparison to bench scale increased about
fourteenfold.  This higher turbulence could cause flow channel-
ing and foam collapse due to frictional forces.  Some such tur-
bulence was observed during the bench-scale experiments whenever
there was an increase in foam flow velocity.  This was probably
due to peculiar characteristics of the foam; however, no signi-
ficant increase in turbulence of foam flowing through the pilot-
scale facility was observed.

Using these three scale up principles, the foam scrubber was de-
signed and erected ready for further experimental work.  Addi-
tional comments on troubleshooting this apparatus and making it
acceptable for pilot demonstration tests are presented in Appendix
C.  The subjects discussed include:

          - foam quality
          - scrubber dust feed
          - particle collection of surfactant spray, and
          - foam baseline aerosol

5.2  SCRUBBER RESIDENCE TIME

As the foam is generated, it flows horizontally into an elbow-
shaped ductwork and then proceeds vertically to the foam destruc-
tion chamber.  During the pilot-scale experiments, it appeared
that the physical configuration of the scrubber and the peculiar
foam flow characteristics caused distortion of the normal plug-
flow velocity profile of the foam, thus making it necessary to
measure residence time experimentally.  Figure 5 attempts to illu-
strate this problem.  Although somewhat exaggerated, the figure
shows that the foam travels in the path of least resistance in the
horizontal and elbow sections.  Foam, unlike gas,  exerts a signi-
ficant weight force on the lower portions of its flow, causing
these portions to remain stagnant.  Once the foam weight becomes
uniform in the vertical section, a full plug flow develops.

                               13

-------
          OUTLET
                 FOAM
             DESTRUCTION


            *- FULLY
            DEVELOPED
            PLUG FLOW
STAGNANT FOAM
    REGION
             INLET
FOAM GENERATION
    Figure  5.  Approximate scrubber velocity
              profile  at steady state.
                       14

-------
The actual residence or space time was measured for 0025% and 1%
Tergitol foam by using a dye injection technique„  A concentrated
solution of methylene blue in 0«25% or 1% soap solution was placed
in a pressurized bomb»  After steady conditions were attained in
the scrubber, the dye was injected into the inlet of the foam liq-
uor feed pump.  The time required for the dye to travel from the
foam generation step to the destruction chamber was defined as the
foam or scrubber residence time,,  A series of experiments indicated
that the actual residence time is around 20-25 seconds„  This value
is far from the time required to observe high scrubber collection
efficiency (refer to Section 4ol), and it is about 62-78% of the
calculated space time  (volume/air flow rate).  Actual scrubber
residence times observed during specific particle collection ex-
periments are presented in Section 5o3<,

5,3  PARTICLE COLLECTION EXPERIMENTS

Once the pilot scrubber was operational, it was tested for fine
particle collection capabilities„  The first crude collection
efficiency measurements showed efficiencies ranging from 32% to
57% on a count basis  (i0e,,, 32% for 0,,78-ym particles).  Effi-
ciencies were measured at a dust loading of 0.8 g/m3  (Oo4 gr/acf).

Theoretical calculation of the collection efficiency involves
three important parametersi  foam bubble size, particle density,
and scrubber residence time,,  The foam bubble size could be only
very crudely estimated by visually observing the foam flowing
around the viewing ports,° its bubble size was estimated to be
about 3 mm in diameter.  The size of the bubbles in the bulk of
the foam flow could not be estimated,,  The second important para-
meter, particle density, also was not precisely determined,  But
scrubber residence time, the third parameter, was actually meas-
ured and determined to be 25 seconds for this experiment.

Considering the limited knowledge of accuracy of these important
parameters and referring to the calculated collection efficiencies
in Figure 2  (bubble size 2 mm, particle density 2 g/cm3) for the
25-second residence time, the collection efficiencies of particles
in the size range between 0*1 ym and 1 pm should be between 38%
and 76%o  This means that the collection efficiencies experiment-
ally determined are in agreement with those predicted by the
theoryo  In order to increase the collection efficiency, longer
residence times are requiredo   (Refer to Section 6 to see how the
residence time influences the scrubber cost,)

The pilot scrubber was purposely designed for shorter residence
times, as was explained in Section 4„1o

After the first collection experiments were completed, an attempt
was made to demonstrate the scrubber operation for a longer period
of time and establish whether or not the recirculation of surfac-
tant liquor affects two important variables, the foam characteris=
tics (stability, bubble size) and the scrubber collection effi-
ciency.  Of course, during the recycle, the collected particulates

                                15

-------
would also accumulate in the surfactant liquor, and it was neces-
sary to determine if this factor also has an influence on the two
variables mentioned above.  This was done successfully in an ex-
periment lasting for about 20 hours.

During this 20-hour run, the particle collection efficiency of the
scrubber was periodically measured.  A summary of the 12 collec-
tion efficiency measurements is presented in Figure 6.  The re-
sults of the 12 individual measurements, as well as a summary of
scrubber operating conditions while obtaining these results, are
presented in Appendix D.  Specifically, the scrubber air through-
put was 0.259 m3/s ± 1.8% (550 acfm) , the scrubber liquor feed was
6.83 x 10~4 m3/s ± 1.2%  (10.8 gpm), and the scrubber AP was 1.84
kPa ± 7% (7.4 in. H20).

Operating the foam scrubber for longer periods provided the opera-
tors with more experience and the  ability to improve the scrubber
performance.  This is illustrated  by the collection data.  While
the initial collection efficiencies for particles in the range of
0.14 ym to 0.78 ym were between 32% and 57% by count, later effi-
ciencies covering the broader particle size range of 0.056 ym to
1.00 ym were between 50.0% and 75% by count, Figure 6.  In addi-
tion, the latter collection efficiencies were obtained at a scrub-
ber residence time of 20.5 seconds  (versus 25 seconds in the first
experiment).  Corresponding theoretical collection efficiencies
(Figure 2)  for particles in the size range between 0.1 ym and
1.0 ym and for this shorter residence time are 32% to 67% by
count.

This means that the improved average collection efficiencies over
the period of 20 hours agree with  those predicted theoretically
even better than the data of the first experiment.  The improve-
ment is primarily due to the ability to generate foam with better
characteristics (small bubbles and more stability) and maintain
these characteristics with time.   The surfactant concentration
during the 20-hour experiment was  kept at 0.25% by weight.  This
is another significant improvement over the bench-scale phase
where about 2% surfactant solutions were needed to generate satis-
factory foams.  In addition, the surfactant liquor requirements
were reduced from 5.17 x 10~3 m3 per m3 of gas (38.4 gpm per 1,000
cfm) to 2.64 x 10~3 m3 per m3 of gas (19.6 gpm per 1,000 cfm) .
These observations are very important.   They indicate the limited
nature and applicability of the results generated in the bench-
scale experiments.   More importantly, however, they indicate that
the costs for surfactant, which are the major factor in the foam
scrubber operating cost, can be significantly reduced (refer to
Section 6).

During the 20-hour run, the surfactant liquor was recirculated,
suggesting that its consumption can be reduced even further.
As a consequence of the recirculation,  the concentration of
particles in the scrubbing liquor  steadily increased as indi-
cated in Figure 7.   This concentration should not be considered

                                16

-------
    200


    190


    180


    170


    160


    150
m

5  140
CO
UJ

g  130


1  120
(T!
O

 x  no
 2  100
 O

 g
 O
 O
     90


     80

     70


     60


     50


     40


     30


     20


     10

      0
      0
          SHADED AREA • OUTLET CONCENTRATIONS
          SHADED + UNSHADED AREAS • INLET CONCENTRATIONS
          75.0%
                   61.4%
                           58.9%
                                    58.9 %
                                            ,50.0 %
       05
0.08 0.1        0.2    0.3  0.4    0.6  0.8 1.0
    PARTICLE DIAMETER, dp, \im
Figure 6.   Average particle collection efficiency
             data  (12 runs).   Percent  collection shown
             inside  the unshaded  area.
                         17

-------
00
           9000
           8000
           7000
           6000
       CO
       £   5000
       p   4000
       o

       o
       o
3000
           2000
           1000
                     100
                       300
400
500
600     700
800
900
1000    1100    1200
                                                  TIME, MINUTES

                                              ( 1 MINUTE =60 SECONDS)
           Figure 7.   Concentration   of solids in scrubbing liquor as  a function  of time.

-------
absolute since there was some accumulation of solids in the
scrubber sumps.  Ideally, under steady state conditions (stable
particle concentration in the scrubber gas feed and stable
particle collection efficiency), the increase in particle concen-
tration with time should follow a straight line.  This was not
the case in the 20-hour run  (see Figure 7) due to some pump leak-
age of the scrubber liquor.  To compensate for this leakage, as
well as the losses of the scrubber liquor through the scrubber
exhaust, a steady makeup stream of surfactant liquor was needed
[about 1.23 x 10~5 m3/s  (0.2 gpm)].  In the 20 hours, the con-
centration in the scrubber just about reached the steady state
as Figure 7 illustrates.  At that time, the amount of particu-
lates collected in the scrubber liquor became equal to the amount
of particulates lost through the leakage and accumulation so that
no further increase in solids concentration in the scrubber liquor
was observed.

The particle collection efficiencies in Figure 6 are based on
particle count rather than mass.  Expressing the particle collec-
tion efficiencies on a mass basis has some limitations, since
collection of a few larger particles can account for high mass
collection while no significant collection of particles in the
fine size range is realized.  In this respect, the foam scrubber
has demonstrated a strong capability to collect fine particles.

The collection mechanisms utilized in foam scrubbing are those of
diffusion and sedimentation.1  These mechanisms are rather well
understood and are not subject to human error.  As such, they
will function in any environment with gravitation field and
concentration gradient.  Proper operation of the foam scrubber is
thus reduced to control of foam characteristics and creation of a
practically acceptable environment to permit these mechnisms to
function.  For these reasons, it is important that future develop-
ment of the foam scrubber include further investigation of the
foam generation step.  This step has not been fully explored and
optimized.

Even though significant improvements in foam generation and foam
characteristics were achieved during the pilot-scale demonstration,
they only suggest that additional efforts in the area of foam
generation on a large scale are warranted.  These efforts should
primarily concentrate on  (1) optimization of the foam generation
screen mesh  (refer also to Appendix C, Section 1, for further
comments), (2) optimization of the scrubbing liquor spray  (spray
pattern, liquor spray rate, and surfactant concentration), and
(3) the flow of the foam through the scrubber (foam stability,
turbulence, and channeling).

Also, the collection efficiencies in Figure 6 represent collec-
tion of the injected dust.  The background liquid aerosol levels
produced in foam destruction were substracted from the particle
levels escaping from the foam scrubber.  Additional data on foam


                               19

-------
are presented in Appendix C, Section 4.  The formation of the
secondary aerosol in the foam destruction step is another factor
which needs further investigation.  Just as the other important
areas of the foam scrubber operation reported above enhanced
particle collection efficiency, foam generation with lower sur-
factant concentration, ability to recycle the scrubber liquor),
it is believed that the area of foam destruction can also be
greatly improved.  Appendix C, Section 4 indicates the scope of
such improvements.  The potential improvements should concentrate
on minimization of secondary aerosol formation and minimization
of energy consumption.  Both of these variables were observed  to
be a strong function of disk design and speed, surfactant concen-
tration and overall geometry of the foam destruction chamber.
The effect of geometry on the secondary aerosol formation, par-
ticularly, is not adequately understood.

While operating the foam scrubber for 20 hours, no deterioration
in foam characteristics was observed.  It is believed that the
length of the run could have been extended with no significant
change in scrubber collection efficiency.  Of course, as already
mentioned, some leaks of the surfactant liquor accompanied this
experimental run, requiring a steady makeup of fresh surfactant.
This led to some difficulty in determining the maximum number  of
surfactant liquor recycles.  Nevertheless, being able to reduce
the surfactant liquor concentration from 2% to 0.25% while main-
taining satisfactory foam characteristics lowered the cost of
surfactant eightfold.  Through recycling the surfactant liquor,
it was possible to reduce the consumption of 0.25% surfactant
liquor to about 1,23 x 10~5 m3/s  (0.2 gpm).  The effect of these
factors on the foam scrubber economics will be further discussed
in the following section.

It should be possible to reduce the consumption of the surfactant
liquor even further.  The recycling basically produces a buildup
of collected particulates in the scrubber liquor which eventually
should influence the foam generation and perhaps the foam charac-
teristics.  During the 20-hour experiment, the solids built up to
about 2 kg/m3 of the liquor, and with this concentration of solids,
no signs were observed which would suggest worsening of foam
generation conditions or foam characteristics.

Temperature is another very important variable for the foam scrub-
ber development, and it was not experimentally verified.  Due  to
a proportional relationship of the particle diffusion coefficient
and the temperature, an increase in collection efficiency with an
increased temperature may be expected.  Changing the gas tempera-
ture, however, may influence other factors involved in the foam
scrubber operation.   A discussion of these factors was presented
in the Phase I report1 (Section 4.3.2 on Thermophoresis and Dif-
fusiophoresis).   Furthermore, temperature may influence the sur-
face properties of foam.   This can subsequently affect foam flow
characteristics, stability, bubble size, and the like.  Most of


                                20

-------
the factors which may be influenced by the changes in gas tem-
perature are not easy to quantify.  Consequently, it is recom-
mended that the foam scrubber be operated at various temperatures
and the effect of these temperatures on foam scrubber operation
and collection efficiency be properly evaluated.

Before completing the 20-hour experiment, an attempt was made to
restore the scrubber liquor to its original condition via filtra-
tion.  This was accomplished by in-line filtration of the scrubber
liquor containing about 2 kg/m^ of suspended solids through a
pressure filter while maintaining foam generation.  No difficul-
ties were observed during the filtration.  Removal of the sus-
pended material from the scrubber liquor restored the foam to its
original white appearance.

The improvements in the foam scrubber operation, however signifi-
cant, were realized without an organized and conscientious effort
to optimize the foam generation and foam destruction steps.  They
are a result of an organized and conscientious effort to demon-
strate the foam scrubber as a viable system on a pilot scale.
It is therefore strongly believed that further improvements of
the foam scrubber performance and economics are yet ahead and
may be discovered when additional efforts are undertaken as
recommended above.
                                21

-------
                             SECTION 6

                             ECONOMICS
Preliminary data on foam scrubber economics were previously re-
ported.1  These data were based on results achieved during the
foam scrubber bench-scale testing.  In comparison with those
results, significant improvements were noted while testing the
foam scrubber on a pilot scale as discussed in the previous
sections of this report.  All these improvements in the foam
scrubber operation strongly affected the foam scrubber economics.
Consequently, this section revises and updates the foam scrubber
economics to reflect the findings from the pilot-scale tests.

As before, the costs generated were compared with costs for three
conventional particulate control techniques.  The cost estimates
for the conventional devices were obtained from the open litera-
ture and it should be recognized that the comparison presented
here is only as good as the available estimates.  In addition,
even though the foam scrubber has been tested on a pilot scale
(0.236 m//s; 500 cfm), its scaleup to a full size facility still
remains to be demonstrated.  As such, there are no accurate
scaleup data from which to verify the reliability of the foam
scrubber economics presented in this report.

The assumptions used in preparation of all cost estimates, their
sources, and the deficiencies and limitations of the cost esti-
mates presented are discussed and reported below.  The three con-
ventional particle collection techniques are high efficiency
electrostatic precipitation, fabric filtration, and high energy
wet scrubbing.

6.1  ACCURACY OF ESTIMATES

The accuracy of the foam scrubbing cost estimates is limited by
the amount of information gained during the pilot testing.  Pilot
test results were the only source of data used to estimate the
foam scrubber economics on the scale several magnitudes larger
than the pilot-scale facility.  Hence, the accuracy of the esti-
mates for the foam scrubber is judged to be about ± 50%.  Addi-
tional information used to determine the foam scrubber costs was
                                22

-------
taken from Modern Cost-Engineering Techniques edited by H. Popper1*
and Plant Design and Economics for Chemical Engineers by M. S.
Peters and K. D. Timmerhaus.5

The costs for conventional control devices were obtained from
three journal references:  Dust Collection Equipment by G. D.
Sargent,6 A Systematic Procedure for Determining the Cost of
Controlling Particulate Emissions from Industrial Sources by
N. G. Edmisten and F. L. Bunyard,7 and Estimating the Costs of
Gas-Cleaning Plants by J. R. F. Alonso.8

Even though the economic data for the conventional devices should
be representative of fabric filtration, high energy scrubbing,
and high efficiency electrostatic precipitation, the corresponding
collection efficiencies of these devices in the fine particle
range are questionable.  They all are defined as high efficiency
collectors based on total particle mass collection.  Since fine
particles in most industrial gases represent only a small frac-
tion of the total particle mass (e.g., an estimated 68% of fine
particles in the range between 0.01 ym and 3 ym from pulverized-
coal-fired powerplants are in the 1 ym to 3 ym range on a mass
basis, while the same size range represents less than 1% of
particles on a number basis),9 these devices may not remove fine
particles even if a very high overall mass collection efficiency
(99%+) is obtainable as claimed.  As a result, the economics of
a foam scrubber device collecting fine particles might be com-
pared with a conventional collector whose collection efficiency
in the fine particle range is not very significant.  Nevertheless,
the economic data for the three conventional control techniques
^Modern Cost-Engineering Techniques.  H. Popper, ed.  McGraw-Hill
 Book Company, New York, New York, 1970.  539 pp.

5Peters, M. S., and K_ D, Timmerhaus.   Plant Design and Economics
 for Chemical Engineers, Second Edition.  McGraw-Hill Book Company,
 New York, New York, 1968.  850 pp.

6Sargent, G. D.  Dust Collection Equipment.  Chemical Engineering,
 76(2):130-150, 1969.

7Edmisten, N. G.,  and F. L. Bunyard.  A Systematic Procedure for
 Determining the Cost of Controlling Particulate Emissions from
 Industrial Sources.  Journal of the Air Pollution Control
 Association, 20 (7):446-452,  1970.

8Alonso, J. R. F.   Estimating the Costs of Gas-Cleaning Plants.
 Chemical Engineering, 78 (28):86-96, 1971.

9Shannon, L. J., P. G. Gorman, and M.  Reichel.  Particulate
 Pollutant System Study.  Volume II-Fine Particle Emissions.
 APTD-0744, U.S. Environmental Protection Agency, Durham, North
 Carolina, 1 August 1971.  348 pp.


                               23

-------
are based on extensive past experience in design, construction,
and operation and are believed to be accurate to within 25% to
30%.

6.2  BASELINE COLLECTION EFFICIENCIES

The following discussion substantiates the questionability of
collection efficiencies for the three conventional devices in the
fine particle range.  In a study of fine particulate emissions
performed by Midwest Research Institute,9 collection efficiency
data for the conventional devices to the submicrometer particle
size range were extrapolated.  These are presented in Figure 8.
For convenience, the collection data from bench-scale experiments
for the foam scrubber are also included.

The collection efficiencies for wet scrubbers and fabric filters
in the particle size range between 0.1 ym and 1 ym can be suffi-
ciently supported by experimental data.  Data from at least 10
scrubbers and 6 fabric filters of different designs were compiled
to produce the extrapolated curves.  The curve for fabric filters
was produced using data with 0.1-ym particle penetrations between
about 0.51 and 65%.  Thus, this curve represents a questionable
average with a relatively large spread rather than a typical
fabric filter operation.  Wet scrubbers show a fairly steep
curve with collection efficiencies in the particle size range
quickly decreasing with the decreasing particle size.

The collection curve of electrostatic precipitation extrapolated
to the particle size range between 0.1 ym and 1 ym can be sup-
ported by only one experimental datum measured for particles
of about 0.7 ym.  The rest of the experimental data were .taken
in the range above 1 ym.

The collection curves for the foam scrubber do not extend below
the 0.1-ym particle size, but as indicated by the theoretically
calculated collection efficiencies1 (and the theoretical collec-
tion efficiencies were found to agree with those determined
experimentally), reasonably high collection of particles in this
range may be expected.

In conclusion, the collection of fine particles by the three
conventional devices is questionable and will need further
experimental verification.  This may lead to changes in the cost
estimates if these devices should collect significant amounts of
particles in the fine particle size range.  The changes are more
likely to increase than to decrease the costs presented here.
The costs for the foam scrubber are based on pilot-scale tests
and can increase or decrease in the future.  The bench and pilot
tests were designed to demonstrate technical feasibility, not to
optimize operating conditions and costs.  Regardless of this, a
significant reduction in foam scrubber costs was achieved as a
result of scaleup from bench to pilot scale.


                                24

-------
U1
           99.99


            99.9
            99.8

            99.0


            95.0

            90.0
     A - POLYPROPYLENE GLYCOL 425 -60 SECOND RESIDENCE TIME (BEST COLLECTION)
     B - DOP AEROSOL - 60-SECOND RESIDENCE TIME (BEST COLLECTION)
    "* - COLLECTION EFFICIENCIES FOR FOAM SCRUBBER ARE ON A NUMBER BASIS  ~
50.0



20.0

10.0

 5.0


 1.0'
 0.5
 0.2

0.05
                                   FOAM SCRUBBER B
                   FABRIC FILTER
                                                                  1
              0.01
                            0.1                        1.0
                              PARTICLE DIAMETER, jjm
                                                                  I	I
0.01

0.05

0.2
0.5
1.0


5.0
10.0   *

20.0   o
                                                                                      50.0
      o

      o
90.0

95.0


99.0

99.8
99.9

99.99
                 Figure 8.  Extrapolated fine particulate control efficiencies.

-------
To obtain a meaningful comyarison of capital and operating cost
estimates for particulate control alternatives, the estimates
should be based upon consistent parameters and have similar accu-
racy.  This was attempted here insofar as the available informa-
tion permitted.  Because it allows order-of-magnitude comparisons
with conventional control techniques and identifies critical cost-
sensitive areas, an economic analysis of foam scrubbing provides
direction for further experimental work.  This was strongly demon-
strated by the previous cost estimate1 which identified the
significance of surfactant cost.  But the previously mentioned
limitations of any of the cost estimates and comparisons presented
here should not be overlooked.

The economic analysis of foam scrubbing and comparison with the
three conventional control devices follow.  General assumptions
applicable to all cost estimates are listed in Table 1, followed
by the assumptions for each specific type of control equipment in
Tables 2 through 5.

6.3  RESULTS

The results of the capital cost estimates are presented in Table 6
and Figure 9.  Table 6 presents percentages of foam scrubbing
capital costs as a function of residence time for a 26.3-m3/s
(50,000-acfm) unit, and Figure 9 depicts the capital costs for
foam scrubbing as well as for conventional controls as a function
of unit capacity.  Table 6 shows that the largest capital cost is
incurred for the scrubber itself (from 30% to 56% of total capital
requirements).  Figure 9 indicates that capital costs for foam
scrubbing are similar to the costs for conventional devices.
Depending on foam scrubber residence time, the capital cost for
foam scrubbing is generally above that for fabric filtration or
high energy wet scrubbing and below that for eletrostatic precip-
itation.

Table 7 and Figure 19 present the results of the operating cost
analysis.  Foam scrubbing operating costs as a percentage of total
operating costs are presented in Table 7 as a function of scrubber
capacity.  The operating cost is relatively independent of foam
scrubber residence time since the operating costs resulting from
the installed equipment cost are insignificant compared to the
cost of the surfactant.
                               26

-------
                 TABLE 1.  COST ASSUMPTIONS
                           General

Capital and operating costs do not include waste treatment
and disposal.

Gas temperature, gas composition, grain loading, and control
efficiency were not considered as variables due to insuffi-
cient information concerning their influence on cost.

Labor costs for both operation and maintenance are contained
within the estimates of maintenance labor and materials.

Time reference 1976     Marshall and Stevens index 470

                        Chemical Engineering facribated
                          equipment index 200

                           Capital

Cost scaling exponent 0.6

Installation charge includes field installation, start-up
cost, working capital, and interest on construction loan.

                          Operating

Stream time 8,000 hours/year

Pump and air mover efficiency 50%

Utility costs

  electricity $0.015/kWh

  water $0.079/m3 ($0.30/103 gal)

Depreciation - 7% installed cost

Capital charges - 10% installed cost (includes interest,
taxes, insurance, overhead, general and administration, etc.)
                              27

-------
             TABLE 2.  FOAM SCRUBBING ASSUMPTIONS
                            Capital

Minimize surface area of scrubber  (length = diameter)
with maximum diameter of about 9m  (30 ft) - if larger add
another scrubber train

Scrubber - 6.35 mm  (1/4 in.) carbon steel with cost as function
of weight  (spray nozzle 45 kg, 100 Ib)

Screen - 96 m2/m3 (29 ft2/103 ft3) 316 stainless
  250 mesh screen - $65/m2 ($6/ft2)

Foam destruction system (10% of scrubber cost)

Residence time 10-60 seconds

Surfactant makeup vessel or surge vessel with recycle 10 minute
capacity

Pump requirement 2,6 m3/103 m3 (20 gal/103 ft3)

Installation charge - 100% purchase cost

                           Operating

Pressure drop 1.8 kPa (7.4 in. H20)

Solution nozzle pressure 359 kPa  (52 psig)

Maintenance labor and materials - 4% installed cost

Surfactant solution (0.25%) consumption - 0.05 m3/103 m3
  (0.4 gal/103 ft3); and none

Energy required for foam destruction - 74 watt/m3(2.1 watt/ft3)

Surfactant - 0.25% solution utilized for Tergitol

     Cost on 100% basis

          Tergitol $1.43/liter ($5.40/gal)
          Aerosol $1.27/liter ($4.80/gal)
          Sterox $0.53/liter ($2.0/gal)
          Alkanol $3.17/liter ($12.00/gal)
                               28

-------
   TABLE 3.  ELECTROSTATIC PRECIPITATION ASSUMPTIONS







                        Capital




References listed in text



Installation charge - 70% purchase cost



                       Operating




Maintenance labor and materials - $64/m3 s"1  ($0.03/acfm)



Pressure drop 249 Pa (1.0" H20)



Contact power 9 watts/m3  (0.00034 hp/acfm)



For costs based on installed cost, use high end of range
          TABLE 4.  FABRIC FILTER ASSUMPTIONS







                        Capital



References listed in text



Installation charge - 75% purchase cost



                       Operating



Maintenance labor and materials - $170/m3 s"1  ($0.08/acfm)



Pressure drop 1.99 kPa (8.0" H20)




For costs based on installed cost, use high end of range
                          29

-------
    TABLE 5.  HIGH ENERGY WET SCRUBBER ASSUMPTIONS







                        Capital



References listed in text



Installation charge - 200% purchase cost



                       Operating



Maintenance labor and materials - $127/m3 s"1  (0.06/acfm)



Pressure drop 14.9 kPa (60" H2O)



Liquid head 1.79 x 105 Pa  (26 psig)



Liquid circulation - 2.67 m3/l,000 m3 (20 gal/103 acf)



Makeup water - 6.68 x 10~5 m3/m3  (0.03 gal/acfm hr)



For costs based on installed cost, use high end of range
                          30

-------
106
                                     CAPACITY, ACFM
                        10
                   10
                   10
O
•o

-------
TABLE 6.  FOAM SCRUBBING CAPITAL COSTS AS A FUNCTION OF
          SCRUBBER RESIDENCE TIME, EXPRESSED AS PERCENT
          OF TOTAL PURCHASE COST
          Capacity - 26.3 m3/s  (50,000 acfm)

Residence time, s
Component
Scrubber
Screen
Destruction system
Mixing/storage vessel
Feed/recycle pump
Total purchase cost
10
30
19
3
40
8
100
20
38
16
4
35
7
100
30
47
14
4
29
6
100
40
51
13
5
26
5
100
50
52
12
5
26
5
100
60
56
11
6
23
4
100

TABLE 7.  FOAM SCRUBBING OPERATING COSTS FOR 40-SECOND
          RESIDENCE TIME AS A FUNCTION OF CAPACITY,
          EXPRESSED AS PERCENT OF TOTAL OPERATING COST

Capacity, mVs

Component
Surfactant
Utilities
Maintenance
Depreciation
Capital charges
Total operating cost
2.63
(5,000)
53
14
6
11
16
100
26.3
(50,000)
66
17
3
6
8
100
(cfm)
263
(500,000)
74
19
1
2
4
100
                          32

-------
U)
U)
                                   CAPACITY,  ACFM
           105,r
              10
10T
10
       CO
           10
        Q_
        O
                      I    I   I
       I    I   I  I  I  1 I T
                                                                           10.0
I FABRIC
- FILTER
              FOAM SCRUBBER
                   (40 s )
                                                              CASE A
                                   HIGH ENERGY WET SCRUBBER
                                         FOAM SCRUBBER -NO
                        '  SURFACTANT MAKEUP
                                        (40s)
                                                                                 o
                              CO
                              o
                              o
                              o
                                                                   
-------
Surfactant costs for this estimate, based on using Tergitol,
amount to 53% to 74% of the total foam scrubber operating costs.
Costs for other surfactants were listed in Table 2.  Figure 10
presents a comparison of operating costs for foam scrubbing with
the conventional collection methods.  A 40-second residence time
was arbitrarily chosen since operating costs are not extremely
sensitive to residence time.

There are two curves indicating foam scrubber operating cost in
Figure 10, Cases A and B.  Case A represents the operating cost
based on scaling up the pilot-scale scrubber to a unit of identi-
fied full capacity.  The full-capacity unit would then operate at
conditions identical to those experienced on the pilot scale
except for the scrubber residence time which would be 40 seconds.
The pilot scale scrubber operated at 20.5 second residence time.
Identical operating conditions include scrubber pressure drop,
gas throughput per unit square area of foam generation screen,
liquid-to-gas ratio, surfactant concentration, energy consumption
of foam destruction disks per unit foam volume, and scrubbing
liquor makeup per unit of scrubber liquor feed rate.

The scrubber liquor makeup during the pilot-scale testing was
1.23 x 10~5 m3/s (0.2 gpm) which still represents a significant
operating expense  (53% to 74% of the total operating cost; see
Table 7).  As discussed in the previous section, it should be
possible to reduce the consumption  (makeup) of the surfactant
liquor further.  The makeup of 1.23 x 10~5 m3/s observed during
the pilot test was almost completely the result of pump leakage.
Some efforts were made to eliminate this leakage (the pump
packing v/as tightened and replaced) .  However, due to the strong
ability of Tergitol surfactant solutions to dissolve oils, greases,
and lubricants, the leakage was not completely stopped. Using a
pump with a mechanical seal could have eliminated the leakage.
However, this type of pump was not available at the time of
testing.

At the leakage and scrubber surfactant makeup rate of 1.23 x 10~5
m3/s (0.2 gpm), the concentration of solids in the scrubber
liquor built up to about 2 kg/m3 of the liquor.  With this con-
centration, no influence on foam scrubber operation (deterioria-
tion of foam quality) was observed.  This indicates that lower
surfactant makeup rates and subsequent lower scrubber operating
costs should be attainable.   With less surfactant makeup and
multiple liquor recycles the concentration of collected parti-
culate would increase and could negatively influence the foam
quality.  Should the foam quality deteriorate, surfactant solu-
ion could be restored to its original condition and recovered for
further usage by filtration as mentioned earlier.
                               34

-------
Case B represents the foam scrubber operating cost with no sur-
factant makeup.  At present, the minimum surfactant consumption
required for a troublefree foam scrubber operation is not known.
Since it is believed that the surfactant consumption may be
reduced significantly the actual foam scrubber operating cost
should be lower than that represented by the Case A curve. At the
same time, some surfactant losses may be expected. Consequently,
the actual operating costs will be higher than those represented
by the Case B curve.

Once surfactant consumption is at its minimum, the utilities be-
come the next highest operating expense (see Table 7).   With the
successes attained during the pilot-scale testing, it is strongly
believed that future experimentation may also demonstrate addi-
tional reduction in foam scrubber utilities requirements and lead
to further lowering of the Case B curve.

The foam scrubber capital investment cost is competitive with
that for existing conventional particulate collectors.   Consid-
ering the potential for further improvement, the foam scrubber
operating cost is also well within the competitive range.  Con-
sequently, further development of the foam scrubber is highly
recommended.
                              35

-------
                           REFERENCES
 1.  Ctvrtnicek, T. E., R. F. Walburg, C. M. Moscowitz, and
     H. H. S. Yu.  Application of Foam Scrubbing to Fine Particle
     Control—Phase I.  EPA-600/2-76-125, U.S. Environmental Pro-
     tection Agency, Research Triangle Park, North Carolina, May
     1976.  152 pp.

 2.  Sem, G. J.  Submicron Particle Sizing Experience on a Smoke
     Stack Using the Electrical Aerosol Size Analyzer.  Presented
     at the Seminar on In-Stack Particle Sizing for Particulate
     Conti'ol Device Evaluations, U.S. Environmental Protection
     Agency, Research Triangle Park, North Carolina, December 3-4,
     1975.  18 pp.

 3.  Bradway, R. M., and R. W. Cass.  Fractional Efficiency of a
     Utility Boiler Baghouse—Nucla Generating Plant.  EPA-600/2-
     75-013-a, U.S. Environmental Protection Agency, Research
     Triangle Park, North Carolina, August 1975.  148 pp.

 4.  Modern Cost-Engineering Techniques.  H. Popper, ed.  McGraw-
     Hill Book Company, New York, New York, 1970.  538 pp.

 5.  Peters, M. S., and K. D. Timmerhaus.  Plant Design and
     Economics for Chemical Engineers, Second Edition.  McGraw-
     Hill Book Company, New York, New York, 1968.  850 pp.

 6.  Sargent, G. D.  Dust Collection Equipment.  Chemical Engineer-
     ing, 76(2):130-150, 1969.

 7.  Edmir.ten, N. G. , and F. L. Bunyard.  A Systematic Procedure
     for Determining the Cost of Controlling Particulate Emissions
     from Industrial Sources.  Journal of the Air Pollution Control
     Association, 20 (7):446-452, 1970.

 8.  Alonso, J. R. F.  Estimating the Costs of Gas-Cleaning Plants.
     Chemical Engineering, 78 (28) :86-96, 1971.

 9.  Shannon, L. J., P. G. Gorman, and M. Reichel.  Particulate
     Pollutant System Study.  Volume II—Fine Particle Emissions.
     APTD--0744, U.S. Environmental Protection Agency, Durham,
     North Carolina, 1 August 1971.  348 pp.

10.  Strbn,  L.  Transmission Efficiency of Aerosol Sampling Lines.
     Atmospheric Environment, 6:133-142, 1972.


                                36

-------
                           APPENDIX A

            DESCRIPTION OF PILOT FACILITY SUBSYSTEMS
The entire scrubber system is housed in a three-story building
having a 6.4 mx 6.1m floor plan.  The scrubber geometry was
altered to meet the physical constraints of this building.  The
scrubber is arranged in a "U" shaped configuration  (Figure 4).
The right leg of the "U" contains the dust feeding, gas condi-
tioning, mixing, and foam generation operations, while the left
leg contains the foam residence chamber, foam destruction chamber,
air blower, and feed and recycle »liquor control loops.

Except for the exhaust duct, the scrubber is square.  The inlet
air enters a 30 cm square stainless steel mixing duct 91 cm long.
The mixing duct is connected to a diverging duct section which
slows the air velocity and changes the flow direction 90° into
the 91 cm square foam generation section.  As foam is generated,
it diverges into a 120 cm square ductwork elbow which changes the
foam flow direction 90° upward.  The foam continues to flow up-
ward in a 120 cm square duct about 2.5 m long.  Several viewing
ports installed in the scrubber walls permit the foam flow through
the scrubber to be observed.

1.   DUST AEROSOLIZATION

Before the dust is aerosolized, the fly ash fines are loaded into
the high-capacity hopper of a vibratory feeder.  A second feeder
(not shown in Figure 4) links the main hopper feeder to a modified
jet mill, Figure A-l, which serves to disperse the feed material
into a stream of ionized air.  Ordinarily, a jet mill serves as a
particle classifier, breaking up coarser material with the oppos-
ing high-pressure air jets designated "O" and "P" in Figure A-l.
For the foam scrubber experiments, however, the jet mill was modi-
fied to disperse the dust into an air stream rather than to
intensely pulverize it.

The feed dust material in the vibratory feeders is air aspirated
into the impact chamber and through the upstack into the classi-
fication chamber where the air flow pattern swirls the mixture
onto the discharge port.  Large particles or agglomerates are
reentrained at the impact chamber entrance.

Prior to entering the jet mill, the "P" jet air flows through a
"static eliminator" radioactive source (Po-210, 20 mCi) which

                                37

-------
                     LARGE PARTICLES
                                  DISCHARGE
                                    PORT
                CLASSIFICATION
                  CHAMBER
               PJET
              IONIZED AIR
                         IMPACT
                         CHAMBER
                                      \
                                     OTUBE
                                       DISPERSED
                                        DUST
  OJET
( BLOCKED OFF)
          Figure A-l.
                  Jet mill  modified for dispersing
                  fly ash fines.
serves to  lower  static attractive forces between dust particles
arising from particle  motion and helps to lower particle
agglomeration.

To prevent large particles from entering the scrubber,  a metal
dust cyclone was installed downstream of the jet mill discharge
port to cat the  coarse material fraction out of the  scrubber  in-
let stream.  The coarse gritty material and highly charged  agglom-
erates are deposited into a hopper installed at the  bottom  of the
cyclone while  the fines are sent to the scrubber inlet.  At the
scrubber inlet,  the suspended dust is mixed with the bulk scrubber
air.  A mixing orifice is located in the scrubber duct  just down-
stream of  the  scrubber air inlet and the dust tube exit.  The
orifice mixes  the relatively slow-moving bulk air with  the  high-
speed dust jet.

Further mixing occurs  in a 30 cm square, 90 cm long  stainless
steel scrubber entrance duct.  A thin stainless steel perforated
plate oriented normal  to the air flow at the end of  this duct
serves to  diffuse and  homogenize the bulk dust mixture.
2.
BULK AIR SUPPLY
The bulk air for  the  scrubber is sucked from the room by  means of
the induced draft fan mounted on the roof of the building.   The
inlet room air is filtered by four absolute filters prior to
                                 38

-------
mixing with the test dust.  This ensures proper control of the
particulate concentration entering the scrubbers.  Penetration
tests made on the absolute filters indicated that when no dust is
fed, the scrubber air at the scrubber inlet is perfectly clean.
The particle count was below the detectable limits of the electri-
cal mobility analyzer.  The total flow rate through the scrubber
is monitored by a venturi meter installed in the round scrubber
exhaust duct.

3.   FOAM GENERATION

As the air/particulate mixture proceeds through  the scrubber,  it
enters the foam generation section.  The bulk scrubber air feed
loaded with the particulate generates the foam by passing through
a 250-mesh screen continuously sprayed with surfactant solution.
The screen is arranged in a zigzag fashion normal to the air flow,
Figure A-2.

4.   FOAM DESTRUCTION

After the foam-encapsulated and particulate-laden gas has flowed
through the main scrubber residence chamber, the foam is destroyed
and the clean gas is exhausted through the stack.  The foam is de-
stroyed with high-speed disks.  This method of foam destruction
was selected, used and investigated during the bench-scale phase
of this program.1  Practically, the foam is funneled upward to-
ward the destruction disks driven by electrical motors.  The
motors are on the outside of the destruction chamber, with shafts
extending into the chamber.  High shear force exerted on the foam
by the rotating disks causes the foam to collapse, forming a liq-
uid which carries the collected particulate matter and drains  into
the lower portion of the destruction chamber.  Under normal opera-
tion, the scrubber is under vacuum.  Therefore,  some liquid level
is maintained in the destruction chamber to provide a seal and a
head for draining the liquid into the scrubber feed tank for re-
cycle.  A view of the foam destruction chamber is shown in
Figure A-3.

5.   SCRUBBER LIQUOR SUPPLY AND RECYCLE

Altogether, the scrubber liquor supply and recycle subsystem
consists of three pumps and three tanks (shown in Figure 4).   The
scrubber feed pump delivers the scrubber liquor  from the scrubber
feed tank through a flowmeter and a flow control valve to the
foam generation section.  Here the liquor is evenly sprayed on
the foam generation screen (see Section 3 of this appendix).
During foam generation, a portion of the liquor drains from the
screen and flows by gravity to a lower sump tank.  The tank is
equipped with a liquid level control to maintain liquid in the
tank and assure a permanent liquid seal between  the scrubber under
vacuum and ambient conditions.  As the liquid in the lower sump
                               39

-------
Figure A-2.  Foam generation screen
             arranged in a zigzag
             fashion.
                  40

-------
Figure A-3.  A view of the foam destruction chamber,
                         41

-------
tank accumulates, the level control activates the sump pump which
transfers the excess liquid from the sump tank into the feed tank.

The liquor drained from the scrubber foam destruction section also
accumulates in the feed tank.  Finally, a fresh surfactant solution
is supplied from the fresh feed makeup tank into the feed tank as
needed.  A transfer pump is used for this purpose.

6.   FINE PARTICLE INSTRUMENTATION

Particle counters available for use on this program were described
in the previous report.1  They include the model 3030 electrical
mobility analyzer (EM counter) manufactured by Thermal-Systems,
Inc., model 220 optical counter manufactured by Royco, Inc., and
model 100 condensation nuclei counter  (CN counter) manufactured
by Rich, Inc.  The EM counter, Figure A-4, counts the number of
0.025, 0.044, 0.078, 0.139, 0.247, 0.44 and 0.78 micrometer
particles in the sample stream.  The optical counter counts the
number of 0.3, 0.5, 0.7, 1.0 and 2.0 micrometer particles in the
sample stream, and the CN counter gives the total count of
particles.

The aerosol dilution and sampling system was optimized for the
inlet and outlet conditions of the foam scrubber.  Aerosol sam-
pling of both inlet and outlet streams was performed using 9.5 mm
bore Tygon tubing at a Reynolds number of approximately 2,300.
This Reynolds number is optimum, according to L. Strom, for mini-
mizing sample deposition on tubing walls, as well as for minimiz-
ing the breakup of larger particles (>2 pm).10  In addition, both
inlet and outlet sampling lines are of equal length.  The sampling
probes are made from 9.5 mm bore stainless tubing with uniform
radii.  Both the inlet and outlet probes are pointed upstream and
located in the center of the main air mass.

7.   SCRUBBER OPERATION

Due to its simplicity, the operation of the scrubber is straight-
forward.  With the equipment at rest, the sump of the foam de-
struction chamber is filled with surfactant solution to the level
that both permits the liquid level control to function properly
and maintains gravity siphoning of the foam liquor produced in
foam destruction.

The lower sump tank is then filled to the level where its liquid
level control functions properly and where a seal between the
scrubber and ambient conditions is established.  The scrubber
feed tank and fresh feed makeup tank are filled.  The dust hopper
10Strom, L.  Transmission Efficiency of Aerosol Sampling Lines
  Atmospheric Environment, 6:133-142, 1972.
                               42

-------
                                                1


Figure A-4.  Electrical mobility analyzer used for
             foam scrubber particle collection
             efficiency evaluations.
                         43

-------
is then filled with solid particulate dust which is to be fed in-
to the scrubber, and the scrubber facility is ready for operation.

Before each experimental run, the induced draft fan is started up
and the dry air flow through the scrubber is allowed to stabilize.
Room air is sucked through the absolute filters on the scrubber
inlet.  These filters are very efficient; and provide clean air
to the scrubber at or near the detectable limits of the fine
particle instrumentation.  If necessary, the scrubber air flow is
calibrated.

The scrubber feed pump is then switched on and adjusted to pro-
vide the required solution spray rate and spray nozzle pressure.
At this point, the total scrubber pressure drop starts to in-
crease.  The air flow is adjusted upwards to compensate for the
increased pressure drop created by the wet spray on the foam
generation screen.  Foam generation commences immediately upon
wetting of the screen.  Air flow characteristics through the
scrubber stabilize when foam reaches the destruction chamber.
The foara destruction disks are then turned on, dust feeding is
initiated, and the scrubber is ready for testing and experimental
data collection.
                               44

-------
                           APPENDIX B

              EXPERIMENTAL DATA ON FOAM DESTRUCTION
A three-point experiment was designed and performed to establish
trends useful for scaling up the foam destruction operation.  In
this experiment, the degree of foam destruction was determined
visually.  Three disk diameter sizes were employed in the experi-
ment:  10, 14.6, and 19.7 cm.  For purposes of control, the foam
impingement velocity was maintained constant at a nominal value
of 0.125 m/s for all sizes of disks used.  Impingement velocity
is the foam velocity normal to the blade prior to destruction.

The experimental runs measured power input to the high-speed uni-
versal motor, speed of the spinning disk, and foam flow rate.
Power consumption was measured with a direct reading watt meter
accounting for the motor's power factor  (E-I-cos £). Speed of the
disk was measured stroboscopically, and air and surfactant solu-
tion flow rates were determined from calibrated rotameter and
pump curves, respectively.  Foam used for the experiment was made
from Tergitol® TMN surfactant and water, mixed as a 2% surfactant
solution.  Bubble size for all runs appeared consistent and com-
parable to the 0.8-mm-diameter foam obtained in previous bench-
scale experiments.

The results of the experiment are shown in Figure B-l.  Data
accuracy is approximately ± 20% because during defoaming, power
input can fluctuate by this amount.  Motor size for a full-scale
defoamer can be extrapolated from a plot of power versus disk
diameter in Figure B-l.  As no distinct difference in power con-
sumption with speed were noticed for the same size disk, the
values of power were averaged.  The curve holds for a range of
speed between 5,000 rpm and 10,000 rpm.
                               45

-------
     200
     150
     100
 o
 Q_
     50
      0
        0
              5,000
-------
                           APPENDIX C

        TROUBLESHOOTING OF THE PILOT-SCALE FOAM SCRUBBER
1.   FOAM QUALITY

Initial experiments performed with the pilot-scale scrubber re-
vealed the problem of poor foam quality.  Foam could be made, but
it was of large bubble size  (>10 mm), and it contained many air
pockets and striations.  The problem was attributed to poor cover-
age of the screen with the scrubber liquor spray and to insuffi-
cient and nonuniform air velocity through the screen.  It appeared
that the foam liquid spray was not covering the edges of the
screen and that it was not of a uniform density.  It was further
observed that the liquid was draining off of sections of the
screen  (runoff) without penetrating the screen mesh.  Installation
of multiple spray nozzles, which gave better and more uniform
coverage of the foam generation screen, solved this part of the
problem.

The reason for insufficient and nonuniform air flow through the
foam generating screen was readily observed and defined as air
leaks in the scrubber system.  As discussed previously, the air
feed to the scrubber is supplied by an induction fan connected to
the scrubber outlet.  Consequently, the scrubber operates below
atmospheric pressure.  The total air flow is measured with a Dall
tube flowmeter connected between the fan and the scrubber outlet.
During each experimental run, the air flowmeter was set at a
specific flow rate, e.g., 0.236 m3/s (500 acfm).  Thus, when foam
was being generated, it was difficult to determine whether the en-
tire amount of air measured at the scrubber outlet passed through
the foam generation screen.  Any air leakage downstream from the
foam generation screen would cause a reduction in the absolute
flow of air through the screen.

A system was devised to quantify the scrubber leakage rate.  A hot
wire anemometer with probe accessory was installed in the scrubber
inlet.  The induced draft fan was then started, and the outlet air
flow was set at 0.236 m3/s (500 acfm) on the Dall tube flowmeter.
The screen was kept dry so that the negative pressure in the
scrubber would be negligible.  Under these conditions, the inlet
and outlet air flows were nearly identical.  The inlet hot wire
velocity reading was noted to indicate an anemometer reading of
0.236 m3/s (500 cfm).
                               47

-------
The scrubber operating pressure was next simulated by carefully
blocking off a portion of the inlet air suction ports while main-
taining constant  (0.236 m3/s) air flow at the outlet.  Next, the
total air flow at the outlet was increased until the anemometer
read the velocity equal to that measured at the negligible scrub-
ber negative pressure (0.236 m3/s at the scrubber inlet).  This
was done while maintaining the scrubber operating (negative)
pressure constant.  The air leakage rate was estimated to be the
difference between the two total air flow readings on the Ball
flowmeter.

Our initial experiment showed a leakage rate of 0.090 m3/s  (38%)
at 0.236 m3/s  (500 acfm) of air flowing through the screen.  This
leakage appeared to be one of the reasons for poor foam quality.
The air leakage rate was reduced considerably by sealing the de-
tectable air leaks to a value less than 10% of the total flow.

As a result of stopping the majority of scrubber air leaks, pro-
viding proper screen air velocity, and changing to the multiple
spray nozzle configuration, acceptable foams were generated with
surfactant liquor concentrations ranging between 0.25% and 2% by
weight,,  Foam was judged acceptable by visual observation of uni-
form bubble size and minimal air pocketing.

Another reason for poor foam quality was plugging of the foam
generation screen with large dust particles.  The Nucla dust,
although highly concentrated in fines, contains a small fraction
of large size particles, roughly 4% by weight >20ym.  Foam quality
degradation and increased screen pressure drop with loss of
scrubber collection efficiency were some results observed as the
screen began to plug.  To solve this problem, a cyclone to remove
large size particulates was installed in the dust feed line.  One
may argue that the same problem may exist with all industrial
streams containing particulate matter.  This warrants further
discussion.

Based on our observations, it is important that the foam genera-
tion screen be kept clean.  Otherwise, the foam generation con-
ditions are altered, resulting in poor foam quality and lowered
scrubber particulate collection efficiency.  These are associated
with a reduced open area of the screen due to physical plugging,
changed gas velocity through the screen, and, ultimately, non-
optimum foam generation conditions.   (It is not expected that
this problem would exist with soluble particulate.)
                              48

-------
 a.    Selection  of  Foam Generation  Screen Mesh  to  Eliminate
      Plugging

 The  screen  plugging  was observed during  the  pilot-scale  experi-
 ments only  when working with  the Nucla fly ash and with  a 250-
 mesh screen.  Feeding  fine  particles  only into the scrubber
 eliminated  the  problem on the pilot scale to the  extent  that  in
 the  20-hour experimental run,  no deterioration of foam charac-
 teristics was observed.  It is quite  likely  that  a simple solu-
 tion such as properly  selected screen mesh for dusts with differ-
 ent  particulate size distributions may completely prevent the
 screen plugging.

 The  selection of the optimum  screen mesh will  be  a tradeoff be-
 tween the scrubber residence  time and the screen  plugging for
 each specific dust stream.  When a screen with very fine mesh is
 used,  plugging  may become severe depending on  dust particle dis-
 tribution.  At  the same  time,  fine screen mesh produces  foams of
 small bubble size  which  reduces the time required to obtain high
 particle collection  efficiency.

 When the dust contains a high  number  of  large  particulates, a
 screen with rather large openings will be needed  to prevent plug-
 ging.   It was determined during the bench-scale phase that such
 screens produce foams  with  relatively large  bubbles.1  This may
 lead to longer  scrubber  residence times  in order  to achieve high
 particle collection  efficiencies.  Presently no specific guide-
 lines  can be given as  to selection of the proper  screen mesh  as a
 function of particle size distribution.   It  is highly recommended
 that this problem  be investigated further.

 It is  very  important to  recognize that the main purpose of the
 foam scrubber is to  collect particles in the fine particle size
 range  (<3 ym).   It was demonstrated that the collection of these
 particles by the foam  scrubber can be successful, and that such
 collection  is technically and  economically feasible.

 The  foam scrubber  was  not intended as a  universal collection  de-
 vice for collection  of all particulate of any  size.  Since all
 existing particle  collectors operate  best in certain specific
 types  of applications, the foam scrubber will  most likely be  no
 exception.  At  present,  the best application conditions for the
 foam scrubber are  not  fully defined,  nor was it the objective of
 this program to define them.  Additional efforts  could easily
 develop the necessary  data in  this area.

 b.   Removal of Large  Particle Sizes  from the  Dust Stream Prior
     to Entering the Foam Scrubber

Many particle collectors function rather well  for large particle
 sizes.  They utilize various principles  of operation and range
widely  in collection efficiency, initial cost, operating and

                                49

-------
maintenance costs, space requirements, arrangement, and materials
of construction.  Examples of these devices include inertial sep-
arators, settling chambers, impingement separators, and panel
filters.  More expensive types of collectors are wet collection
devices, baghouses, and electrostatic precipitators.

It is not expected that the use of the more expensive collectors
would be an attractive alternative for removal of large size
particles in combination with the foam scrubber.  The cost of
such an alternative would most likely be prohibitive.

The use of the other, more economically attractive devices, how-
ever, may be feasible.  A cyclone inertial separator successfully
removed large-size particles from the Nucla dust during the pilot-
scale experiments.  It should be recognized that the selection of
a particle collection train for a specific application is a func-
tion of many important variables including dust characteristics,
carrier gas characteristics, space availability, influence on
production cost, and the like.

The question of what is feasible economics for a fine particle
control technology has not yet been answered  (refer to the Phase
I report for further information1).  Based on the economic evalu-
ations, the foam scrubber has been shown to be economically com-
petitive with conventional particle collectors  (see Section 6).
But further investigation is needed to determine when the foam
scrubber has to be combined with a "precleaner," and what type
of "precleaner" is necessary.  Then, the potential of the foam
scrubber, whether with or without a "precleaner," can be deter-
mined more accurately.  It is therefore recommended that, after
the data on foam generation screen selection are available, a
study be made to determine both the necessity for gas precleaning,
and its impact on foam scrubber economics.

c.   Mechanical Cleaning of the Foam Generation Screen

It was observed during the pilot-scale experiments that the
screen can be mechanically cleaned should it get plugged with
large-size particles.  This solution, then would involve develop-
ment of a continuous mechanical cleaning concept for the foam
generation screen.  When this is done, the effect of the mechan-
ical cleaning on the foam scrubber economics should be evaluated.

2.   SCRUBBER DUST FEED

As already mentioned, aerosol consisting of the Nucla fly ash
fines was dispersed into the scrubber inlet with staged vibrat-
ing feeders and a jet mill.  Figure C-l gives a typical stability
trace of the inlet aerosol fed to the foam scrubber.  This is a
high sensitivity trace made with the EM counter, and it shows
that the dust feeding system is reasonably stable in performing
collection efficiency measurements.

                                50

-------
  3.   PARTICLE  COLLECTION OF  SURFACTANT  SPRAY

Prior to  running  collection  efficiency  experiments with  foam,  an.
attempt was made  to measure  the collection efficiency of the wet
screen only.   It  had been observed during bench-scale testing  that
some  particles had been collected on  a  water-wetted  screen.  Pure
tap water at the  rate of 6.33 x I0~k  m3/s sprayed on the clean
screen.   The fan  was set to  deliver 0.258 mvs air flow.  An in-
complete  series of particle  data was  taken because,  with the nega-
tive  pressure  in  the scrubber, the limits of scrubber design were
reached.  The  pressure drop  across the  screen and the negative
scrubber  pressure increased  markedly  when tap water  was  sprayed
on the screen.  One percent  surfactant  solution produces a pres-
sure  drop about three times  lower than  when pure water is used.
The higher viscosity and surface tension of water explains this
phenomenon.  As a result of  the excessive pressure drop,  the wet
screen particle collection effect for the pilot scrubber could
not be determined.  Based on the data generated on the bench scale,
however,  collection by the foam generation spray alone should  be
insignificant.l

4.    FOAM BASELINE AEROSOL

The amount of baseline aerosol generated from the foam destruc-
tion  was  measured.  The particle baseline is measured at the
scrubber  outlet with foam being generated and destroyed,  but with
no dust feed.  The baseline  aerosol originates from  two  sources,
the first being the foam destruction  disks.  The high-speed disk
blades break the  foam into liquid which is returned  to the feed
tank.  The shearing and chopping of the blades may result in a
net generation of aerosol.   In addition, the leakage of  room air
into  the  scrubber resulted in a net count of fine particles in the
outlet stream.

The average foam  baseline counts for  the 20-hour experiment are
shown  in  Figure C-2.  The counts made when running 1% Tergitol
solution  and approximately 6,500-rpm  disk speeds indicated 106
particles/cm3 in  the particle size range between 0.025 ym and
0.78  ym.  Attempts were made to reduce  baseline counts.   It was
found  that the baseline could be reduced nearly 10 times by
switching to a lower concentration of surfactant% (0.25%  versus 1%).
and decreasing disk speed from 6,500  rpm to 4,000 rpm.   Under
these  conditions, foam was sufficiently stable and at the same
time was  easily liquefied, even at the  lower destruction disk
rpm's.

The above observation is very important.  It indicates the poten-
tial  improvements that can be realized  due to the scaleup.  Based
on bench-scale experiments,  0.25% surfactant solutions did not
generate  acceptable foams.  But reduction of the surfactant con-
centration during the pilot experiments not only had a dramatic
influence on the operating cost of the  foam scrubber, it also  pro-
duced  foams which are much more readily destroyed, as indicated

                                51

-------
                                  CONCENTRATION
01


i
3 10 20





























1
i
f
30 ^>4


VjJ
ON
g
L_
—1
m
CO





^
f"
j
\
k
f
l>
-f



0 5









^»


0 6
D

r
i
j
i
/





-'^
5/24/76
# iQn9d'
?
7r l/Ut-Ht.
0 70 8
ILUTIONR
10.4 / 1
JUCLA DUS
.2.6 g/Mlf
5.5M3/MI
MR FLOW





ATIO

T
d
N
N





                             Figure C-l.   Aerosol stability.

-------
            a:

            o_
            CO
            O
190

180

170

160

150

140

130

120

110
            2  100

            o:
            o
            o
                      SHADED AREA - OUTLET CONCENTRATIONS
                      SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
 90

 80

 70

 60

 50

 40

 30

 20

 10

 0
 0.05    0.08 0.1
                      36.5%
                             32.2 %
                                     21.5%
                                   0.2    0.3  0.4   0.6  0.8 1.0
                           PARTICLE DIAMETER, dp> \im

Figure C-2.   Average foam baseline counts  (11 runs).   Scrubber
               residence  time:   20.5 seconds.   Surfactant concen-
               tration:   0.25%  Tergitol.   Disk  speed:  4,000-5,000
               rpm.
                                   53

-------
by lower destruction disk rpm requirments.  The lower speed also
had a significant influence on the secondary  (baseline) aerosol
formation.  Ultimately the power consumption by the destruction
disks was reduced.

It is strongly believed that even further improvements of foam
destruction  (reduction in secondary aerosol formation and power
consumption) are achievable once the foam destruction step is
more thoroughtly understood.  It is therefore highly recommended
that the foam destruction step be investigated further.
                               54

-------
                           APPENDIX D

    FINE PARTICLE COLLECTION DATA AND FOAM SCRUBBER OPERATING

           CONDITIONS FOR THE 20-HOUR EXPERIMENTAL RUN
During the 20-hour experiment, the scrubber particle collection
efficiency was measured 12 times.  The results of these 12 mea-
surements are presented in Figures D-l through D-12.  A summary
of the data was presented in Figure 6.

Several variables important for foam scrubber operation were
periodically monitored.  The most important of these are summa-
rized in Table D-l.  The table indicates the ranges as well as
the averages for each important variable.  As can be seen, the
operation of the scrubber during the 20 hours was very stable
with rather minor fluctuations in the scrubber gas flow through-
put, scrubber AP, and scrubber particle feed.
 TABLE D-l.
RANGES AND AVERAGES FOR IMPORTANT PARAMETERS OF THE

FOAM SCRUBBER OPERATIONS, 20-HOUR DEMONSTRATION

Parameter
Barometric pressure, kPa
(in. Hg)
Wet bulb temperature, K
(°F)
Dry bulb temperature, K
(°F)
Relative humidity, %
Air throughput, m3/s
(cfm)
Liquor feed, lO"4 m3/s
(gpm)
Scrubber AP, kPa
(in. H20)
Range
98.37-99.00
(29.13-29.32)
291-294
(65-70)
298-303
(76-86)
41-56
0.255-0.261
(540-555)
6.75-6.90
(10.7-11.0)
1.74-1.97
(7.0-7.9)
Average
98.74
(29.24)
293
(68)
301
(81)
47
0.259
(550)
6.83
(10.8)
1.84
(7.4)
                               55

-------
         f"N
          E
          o
         LU
         d
   200

   190

   180

   170

   160

   150

   140

   130

   120
Q-

x  110

2  100
i—

£   90

    80
         O

         O

         3  70h
             60

             50

             40

             30

             20

             10
                  SHADED AREA - OUTLET CONCENTRATIONS
                  SHADED + UNSHADED AREAS • INLET CONCENTRATIONS
              0
              0.05
          68.7 %
                          55.4%
                                 43.0%
                       VVX///XX//
42.7%
                                                X25.0%
                                               33ZZ3ZZEE
           0.08 0.1       0.2   0.3  0.4    0.6 0.8 1.0
              PARTICLE DIAMETER, dp, \im
Figure D-l.   Particle inlet  and outlet concentrations and
               collection efficiencies (Run  1).   Percent
               collection shown inside the unshaded  area.
               Scrubber residence time:  20.5  seconds.
               Surfactant concentration:  0.25%  Tergitol,
                              56

-------
        oo
        LU
        —I
        O
        HH

        ct:


        a.

        o
        r—i

        X



        O
        HH


        <
        o
        o
        ex
        <
        Q_
200


190



180


170


160


150


140


130


120


110


100


 90


 80


 70


 60


 50


 40


 30


 20


 10


 0
                 SHADED AREA = OUTLET CONCENTRATIONS

                 SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
Figure D-2.
  0.05   0.08 0.1       0.2   0.3  0.4   0.6  0.8 1.0

           PARTICLE DIAMETER, dp, pm



   Particle  inlet and  outlet concentrations and

   collection  efficiencies  (Run  2).  Percent

   collection  shown inside the unshaded  area.

   Scrubber  residence  time:  20.5 seconds.

   Surfactant  concentration:  0.25% Tergitol.
                               57

-------
   200

   190

   180

   170

   160

   150

5 WO
to
LJJ
d 130
I—I
or
         o.
        m
         o
            120
            110
         2  100
         <
         t:   90
         o
         O
         o
    80

    70

    60

    50

    40

    30

    20

    10
                 SHADED AREA - OUTLET CONCENTRATIONS
                 SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
                          61.5%
             0
              0.05   0.08 0.1
                                 53.2%
                                        57.6 %
                                                ^59.1 %
                       0.2   0.3  0.4    0.6  0.8 1.0
                       PARTICLE DIAMETER, dp>
Figure D-3.
      Particle inlet  and outlet concentrations and
      collection efficiencies  (Run  3).   Percent
      collection shown inside  the unshaded area.
      Scrubber residence time;   20.5  seconds.
      Surfactant concentration:   0.25%  Tergitol.
                              58

-------
             200

             190

             180

             170

             160

             150
          S 140
130

120

110
          Q_
         f*
          s
          X
          § 100
          I— I

          jl  90
          o
          o
         o
         I—I
         I—
         on
         <
         O-
 70

 60

 50

 40

 30

 20

 10
              0
              0.05
     SHADED AREA • OUTLET CONCENTRATIONS
     SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
       0.08 0.1       0.2   0.3  0.4
           PARTICLE DIAMETER, dp, \u\
0.6 0.8 1.0
Figure D-4.   Particle  inlet  and outlet concentrations  and
               collection efficiencies  (Run 4).   Percent
               collection shown inside  the unshaded area.
               Scrubber  residence time:   20.5  seconds.
               Surfactant concentration:  0.25%  Tergitol.
                               59

-------
                 SHADED AREA • OUTLET CONCENTRATIONS
                 SHADED * UNSHADED AREAS • INLET CONCENTRATIONS
         E
         o

        co
        U-l
        O
        P
        o;


        o
        r—I
         X
        o
        g
        o
        o:
        <
        Q.
  100%
   Possible
   data
   bias)
                                        0.4   0.6  0.8 1.0
                      PARTICLE DIAMETER, dp,
Figure  D-5.
Particle inlet  and outlet  concentrations and
collection efficiencies  (Run 5).   Percent
collection shown  inside  the  unshaded  area.
Scrubber residence time:   20.5 seconds.
Surfactant concentration:   0.25% Tergitol.
                              60

-------
    200

    190


    180

    170

    160


    150
p"t

5  140
CO
LLJ
o  130
o:
£  120
r"-\
O
 x  HO
        2  100
        o
        O
        o
        K—
        0£.
        <
        Q.
    90

    80

    70

    60

    50

    40

    30

    20

    10

     0
                 SHADED AREA • OUTLET CONCENTRATIONS
                 SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
     0.05   0.08 0.1      0.2    0.3  0.4
               PARTICLE DIAMETER, dp, pr
                                              0.6 0.8 1.0
Figure D-6.
       Particle inlet and  outlet concentrations and
       collection  efficiencies  (Run 6).  Percent
       collection  shown inside the  unshaded area.
       Scrubber residence  time:  20.5 seconds.
       Surfactant  concentration:   0.25% Tergitol.
                              61

-------
            200

            190

            180

            170

            160

            150
         0  140
         y  13°
         P
            120
         O-
         r^
         O
            110
         2  100
         <
         £   90
         o

         o
         LJJ
         o
    80

    70
SE   60
a.
    50

    40

    30

    20

    10

    0
                   61.1 %
                 58.0%
                       SHADED AREA •
                         OUTLET CONCENTRATIONS

                       SHADED + UNSHADED AREAS
                         INLET CONCENTRATIONS
                                 61.1 %
                                         63.7 %
                                                r61.8 %
              0.05    0.08 0.1
                       0.2   0.3  0.4    0.6 0.8 1.0
Figure D-7.
              PARTICLE DIAMETER, dp< urn

      Particle  inlet and outlet  concentrations and
      collection efficiencies  (Run 7).   Percent
      collection shown inside  the unshaded area.
      Scrubber  residence time:  20.5  seconds.
      Surfactant concentration:   0.25%  Tergitol.
                              62

-------
    200



    190




    180



    170



    160



    150



    140



 o  130
 i—i



JP  12°
 O

 X  110



 g  100

 <

 1  90
 LU
 O
        GO
        UJ
        o
        o
        o
        HH
        H—
        a:
    70



    60



    50



    40



    30



    20



    10
             0
             0.05
                 SHADED AREA - OUTLET CONCENTRATIONS

                 SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
                  73.3%
                         58.6%
                                61.4%
                                        59.6%
                                                -50.0%
           0.08 0.1       0.2   0.3  0.4   0.6  0.8 1.0

               PARTICLE DIAMETER, dp, \im
Figure D-8.
       Particle inlet  and outlet concentrations  and

       collection efficiencies  (Run 8).   Percent

       collection shown inside  the unshaded area.

       Scrubber residence time:   20.5  seconds.

       Surfactant concentration:  0.25%  Tergitol.
                              63

-------
                 SHADED AREA • OUTLET CONCENTRATIONS
                 SHADED + UNSHADED AREAS = INLET CONCENTRATIONS
                  0.08 0.1      0.2   0.3  0.4    0.6  0.8 1.0
                      PARTICLE DIAMETER, dp, \im

Figure  D-9.  Particle inlet and outlet concentrations and
              collection efficiencies  (Run 9).   Percent
              collection shown  inside the  unshaded  area.
              Scrubber residence time:  20.5  seconds.
              Surfactant concentration:   0.25% Tergitol.

                             64

-------
    200



    190




    180



    170



    160



    150
f»
 e

-£  140

LLJ


S  130
i—



j£  12°
O

 x  no



2  100



1  90
           |  80
               70



               60



               50



               40



               30



               20



               10
                    SHADED AREA • OUTLET CONCENTRATIONS

                    SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
                    75.0%
                            61.4%
                                   58.9 %
                       /////////
                                58.9%
                                      X50.0%
     0.05   0.080.1       0.2    0.3  0.4

               PARTICLE DIAMETER, dp, nn
                                                 0.6 0.8 1.0
Figure D-10.
     Particulate inlet and  outlet concentrations and

     collection efficiencies (Run 10).   Percent

     collection shown inside the unshaded  area.

     Scrubber residence time:  20.5  seconds.

     Surfactant concentration:  0.25%  Tergitol.
                                  65

-------
                    SHADED AREA • OUTLET CONCENTRATIONS
                    SHADED + UNSHADED AREAS « INLET CONCENTRATIONS
Figure  D-ll.
       0.08 0.1       0.2   0.3  0.4   0.6 0.8 1.0
          PARTICLE DIAMETER, dp, \nm

Particulate inlet  and outlet  concentrations and
collection efficiencies (Run  11).   Percent
collection shown  inside the unshaded area.
Scrubber residence time:  20.5  seconds.
Surfactant concentration:   0.25% Tergitol.
                                66

-------
                 .  SHADED AREA - OUTLET CONCENTRATIONS
                 .  SHADED + UNSHADED AREAS - INLET CONCENTRATIONS
               0.05   0.08 0.1       0.2   0.3  0.4   0.6  0.8 1.0
                        PARTICLE DIAMETER, dp, \nm

Figure  D-12.  Particulate  inlet and  outlet concentrations and
               collection efficiencies  (Run 12).   Percent
               collection shown inside  the unshaded area.
               Scrubber residence time:   20.5 seconds.
               Surfactant concentration:   0.25% Tergitol.

                                 67

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA-600/2-77-110
                           2.
                                                       3. RECIPIENT'S ACCESSION NO.
4. T.TLE AND SUBTITLE
                           Qf
 Particle Control, Phase II
            5. REPORT DATE
              June 1977
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
T.E. Ctvrttnicek, S.J. Rusek,  C.M.  Moscowitz,
and L.N. Cash
            8. PERFORMING ORGANIZATION REPORT NO.

              MRC-DA-682
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio  45407
                                                       10. PROGRAM ELEMENT NO.
             1AB012: ROAP 21ADL-029
            11. CONTRACT/GRANT NO.
             68-02-1453
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
             Phase Final; 7/75-7/76
            14. SPONSORING AGENCY CODE
              EPA/600/13
15. SUPPLEMENTARY NOTES jERL-RTP project officer for this report is Geddes H.  Ramsey,
Mail Drop 61, 919/549-8411 Ext 2298.  EPA-600/2-76-125 was the Phase I report.
16. ABSTRACT
          The report summarizes the knowledge, experience, and data gained to date,
relative to the application of foam scrubbing to collecting fine particles from gaseous
streams.  Experimental data are presented obtained on a 0.236-cu m/s (500-cfm)
pilot-scale foam scrubber facility. Economic analysis  indicates that a foam scrubber
can be competitive with other fine particle collection devices. Areas for further foam
scrubber development are recommended.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                           b.lDENTIFIERS/OPEN ENDED TERMS
                         c. COSATI Field/Group
Air Pollution
Scrubbers
Foam
Dust
Gases
Air Pollution Control
Stationary Sources
Particulate
Fine Particles
13B
07A
11G

07D
13. DISTRIBUTION STATEMENT
 Unlimited
                                           19. SECURITY CLASS (This Report)
                                           Unclassified
                         21. NO. OF PAGES
                              74
20. SECURITY CLASS (Thispage)
Unclassified
                         22. PRICE
EPA Form 2220-1 (9-7£)
                                        68

-------