United States
                    Environmental Protection
                    Agency
 Industrial Environmental Research
 Laboratory
 Cincinnati OH 45268
                     Research and Development
 EPA-600/S2-84-042  Apr. 1984
c/ERA          Project Summary
                    Evaluation  of an  Air  Curtain
                    Hooding  System  for  a  Primary
                    Copper  Converter

                    Chuck Bruffey, Paul Clarke, Thomas Clark, Mark Phillips, and John 0. Burckle
                      This report presents the results of
                    tests conducted to evaluate the effec-
                    tiveness of a full-scale air curtain cap-
                    ture system installed on a primary
                    copper smelter converter for capture of
                    low-level fugitive particulate, including
                    trace metals and sulfur dioxide. The test
                    work was  performed  onsite at
                    ASARCO's Tacoma Smelter on the first
                    domestic  full-scale prototype system,
                    resulting  in the  first  published
                    evaluation  of  a  full-scale  fugitive
                    capture system  based  upon the air
                    curtain approach as applied to a primary
                    copper converter.
                      The installation of the air curtain
                    hooding  system  has  permitted  a
                    quantitative approach to the direct
                    measurement of the fugitive emissions
                    from a primary copper converter for the
                    first time. In this program, the fugitives
                    captured  by  the  air curtain  were
                    measured at a downstream sampling
                    point in the exhaust side of the air
                    curtain system  during the  various
                    portions  of the converter  cycle.
                    Emission factors were established for
                    sulfur dioxide, filterable  particulate
                    (Method 5), inhalable particulate, and
                    selected trace elements.
                      This Project Summary was developed
                    by  EPA's Industrial  Environmental
                    Research Laboratory. Cincinnati.  OH.
                    to announce  key  findings  of  the
                    research project that is fully document-
                    ed in a separate report of the same title
                    (see Project Report ordering informa-
                    tion at back).

                    Introduction
                      Copper converting is a batch operation
                    conducted   in two stages  to convert
copper matte  produced in  a  smelting
furnace into blister copper. The Peirce-
Smith converter, used in all but one U.S.
smelter, is acknowledged to be the major
source of fugitive  emissions in  the
smelter. These fugitive emissions first
enter  the workplace and, because they
are   present  in  relatively  high
concentrations, are  considered
hazardous to worker health. They are
emitted from the smelter at relatively low
elevations through roof  monitors  and
other  openings in the  building. These
emissions cause deterioration of the air
quality and are believed to pose adverse
health risks to the general population
suffering  prolonged exposure. While
some  dispersion  and  dilution of  the
fugitive emissions occur upon leaving the
smelter workplace, the resulting ambient
concentrations are high relative to a well-
dispersed emission from a tall stack.

  A number of approaches to controlling
these emissions have been attempted by
industry with unsatisfactory results. The
major  barrier to the development of an
acceptable secondary hood has  been the
inability to design  a  system capable of
permitting crane and ladle access while
simultaneously providing for reasonably
effective capture of fugitive emissions.
  The  air curtain (Figure 1) is formed by
blowing air from a supply plenum or a row
of nozzles  which is especially designed to
form an air sheet, or curtain, with as little
turbulence as possible. This curtain is
directed over the open space, well above
the converter, which  permits crane
access. On the opposite side of the space,
the curtain and entrained air are captured
by an exhaust system. Fumes which rise

-------
Figure  1.   Air curtain operating concept.

from the  source are directed into the
suction plenum by the curtain. Air is also
pulled  into the curtain from both above
and below. Since all air flow is inward,
into the curtain, a high capture efficiency
is  achievable with a properly designed
and operated curtain.
  In  the past it  has been  possible to
estimate only very approximately, the
quantity and composition of the converter
fugitives for defining  control strategies
and  needs, making  actual  design and
selection of equipment somewhat risky.
  Because the  fugitive  emissions are
effectively captured by the air curtain and
are  collected  by  ducting,  it  becomes
possible to completely characterize these
emissions with a much higher degree of
confidence in order to  provide actual
engineering data for design.
  The  tests described in the full report
were  conducted  jointly  by  the  U.S.
Environmental   Protection Agency's
(EPA's) Office of Air Quality Planning and
Standards and the Office of Research and
Development   in  cooperation  with
ASARCO, the Puget Sound Air Pollution
Control Agency, and the EPA Region X
Office.
Test Program
  The  test  program  was  designed to
achieve two major objectives: to estimate
the  effectiveness   of  capture  of  the
converter fugitives not controlled by the
primary  hood and  to characterize the
captured fugitives  by the "quasi-stack"
method.

Capture Effectiveness
  The   effectiveness  of  capture  was
evaluated using  three techniques.

  •  Mass balance  using hexafluoride as
     a tracer
  • Opacity  of  emissions   escaping
    through the slot

  • Observation of visible emissions

Tracer Experiments
  Sulfur  hexafluoride was injected  into
various points  within the  air  curtain
control volume, defined by the top, sides
and front of the air curtain structure and
the converter and primary hood which
formed the back of the structure.  The
tracer  experiments  were  of two types,
those in  which the tracer was injected
into  the  air  curtain volume  above the
converter (the upper portion of the air
curtain  control  volume)  and those in
which the tracer was injected below the
plane of the top of the converter and near
the front  of the air curtain  side walls (the
lower portion of the air curtain control
volume).
  The recovery efficiencies measured in
the first test for individual tracer releases
above the converter varied from 69 to 119
percent, and the overall average efficien-
cy for the 45 tests was 94 percent. The
port  through which the releases of the
tracer were made did not have any effect
on the average collection efficiency. The
average  collection   efficiency  of  all
releases  made  through  a  given  port
ranged from 93.0 percent  for Port C-6 to
95.4 percent for Port C-1. This difference
was  not  statistically  significant.  The
variability between the average collection
efficiency of the replicates  made at a
given position (between the jet  side and
the exhaust side) was statistically signif-
icant. The greatest difference occurred at
Port  D-1, where the average collection
efficiency  ranged from 83.3 to 105.7
percent. The average collection efficien-
cies  for  Positions  1  and 2 (near the
exhaust side) were approximately 96.6
percent and were generally higher than
those for Positions 3 and 4 (near the jet
side)  which  were  approximately 91.6
percent.
  The tracer recovery efficiencies for the
various converter operating modes were
also measured. With the exception of cold
additions, the average recovery efficiency
was not affected by the operating mode of
the converter; averages varied from 92.8
percent during  blowing to 95.0 percent
during slag skimming.
  For the second experiment above the
converter,  the  overall average  tracer
recovery  efficiency  was  96.0  percent.
Again, the  port through which the tracer
releases were made had no effect on the
average tracer recovery efficiency of the
air curtain  hood. The average efficiency
varied from  94.5 percent at Port C-6 to
98.0 percent at  Port B-2 (Figure 2). For
positions within  the matrix, the average
collection efficiency varied from  80.7
percent at Position 4,  Port D-1, to 106
percent at Position 2, Port D-1. As in the
first test series, the recovery efficiencies
were  consistently higher for  positions
near the exhaust side than for positions
near the jet side (Figure 3). Again, the
operating mode had no adverse effect on
the recovery efficiency measured.
  Two  special  tests  were  conducted
during slag  skimming  where the tracer
was injected just above the top of the
converter at the front of the jet side baffle
wall.  The average collection  efficiency
measured was  94.5 percent, which  is
comparable   to  that  reported for the
releases on the three-dimensional matrix
in the space above the converter.
  For the third experiment, several series
of tests involving the release of the tracer
into the lower portion  of the air curtain
control  volume near the front of the air
curtain  side  walls were conducted. The
first series   (three tests)  involved the
release of the tracer material at a location
slightly above the ladle near the jet side of
the hood. The average recovery efficiency
was 64.3 percent.
  For the second series (six tests), the
tracer was released at a location slightly
above the ladle and very close to the wall
on   the  exhaust  side.   During   slag
skimming,   the  recovery  efficiency
measured (four tests) ranged from 52 to
79 percent for an average of 63.5 percent.
During   matte   charging,  the average
recovery was 68.5 percent.
  In the third series of tests, the tracer
material was also released at a location
slightly above the ladle, but farther from
the  wall on  the exhaust side.  The
collection efficiency measured for the
seven tests ranged from 30 to 89 percent,
with an overall average of 58.7 percent. It
should be noted that the samples for tests
conducted during the  operation  in the
blowing  mode   yielded  the  lowest
recovery efficiencies, i.e., 32, 33, and 33
percent. These values would be expected
because the hooding system was in the
low flow mode and there was no thermal
lift which causes air to be drawn into the
air curtain control volume from the front
and carried to the upper control volume, a
phenomenon  which  enhances the
collection efficiency.
  In the final series of tests, the tracer
was released very near the ladle on the
exhaust side of the  hooding system.
Recovery efficiencies  were determined
for 53 releases of the tracer material and

-------
           Jet
                                                          Exhaust
  110
  100
 i  90
   80
   70
   60'
                                                             D-7
                                    Position
 Figure 2.    Effect of injection port on recovery efficiency.
ranged from 27 to 128 percent, with an
overall average of 70 percent. Recovery
varied from 38 percent for the 6 tests
performed during blowing to 84 percent
for the  28 tests performed during slag
skimming.   The   difference   between
average  collection efficiencies for  the
several operating  modes is statistically
significant.

Opacity Measurements
  An opacity monitor was mounted on
the top of the air curtain below the crane
rail  in order to obtain information on
emissions escaping  capture  by the air
curtain  and passing through the slot. A
total  of  86 discrete  observations  was
made with results ranging from 2 to 54
percent opacity for the major converter
operations.  During   slag  and  finish
blowing, no attenuation of the monitor's
light beam was observed resulting in zero
percent opacity. The instrument output
range was  0 to  20 milliamps which
corresponds  with 0 to 98.4  percent
opacity.  The relationship of the instru-
ment output to opacity was exponential,
with 5 milliamps corresponding with 50
percent  opacity.  Therefore,  emissions
during the test program were in the lower
end of instrument response. No correla-
tion   between  opacity  and  capture
effectiveness could be made because of
emissions from the front of the air curtain
system.


Visual Emissions
  Two observers visually monitored the
air  curtain  capture  effectiveness  by
noting the location, approximate opacity,
duration,  and  significance  of  visible
emissions.  Their estimates of  capture
efficiency were within 5 to 10 percent
with  only   a  few  exceptions.  Most
variability in  the estimates  occurred for
those  operations  involving rapid
evolution  of  emissions  over  a  short
period,  such  as roll-in, roll-out,  and
pouring. The average of the observations
for  the  various converter  operating
conditions displayed the same trends as
the tracer experiments  and indicates a
reasonably effective capture of fugitives.

Conclusions
  In summary, the visual observation and
tracer  recovery data indicated that the
fugitive emissions capture effectiveness
of the secondary  hood  is greater than
90 percent,  averaging about 94 percent
overall. The  capture effectiveness during
converter  roll-in,  roll-out,  and  slag
skimming operations is more  variable
than   other  converter   modes,  since
fugitive  emissions  generated  during
these  events  are  dependent  upon
converter and crane operations. It is also
evident that capture efficiencies  of 90
percent or better are achievable for these
events  under  the proper  crane  and
converter   operating   conditions  to
minimize  fume   "spillage"  into  the
converter aisle.
  Thermal lift plays a significant role in
increased collection efficiencies for fume
generated in the  lower portion of the
control area. Also,  the  lower tracer
recovery efficiencies for  the  various
converter roll-out modes are indicative of
fume  "spillage" outside of the control
area.
  It is believed that no practical correla-
tion can be made between opacities
recorded by the observers and the trans-
missometer.  The  transmissometer was
mounted perpendicular to the longitudinal
axis of the slot, whereas the position of
the visual observers was such that their
view was parallel to the longitudinal axis
of the slot, which resulted in a consider-
ably longer  path  length through  the
escaping emissions. The apparent opacity
increases as the path length through the
emissions increases. Also, when posi-
tioned in front of the  converter,  the
overhead crane interfered with visual
observations above the slot area.

Emission  Characterization
  The capture of the fugitive emissions by
the  air  curtain permitted  their
characterization  by  the "quasi-stack"
method  using  standard  EPA  stack
sampling techniques in the exhaust duct.
The  converter is a  batch  operation
comprised of a number of steps requiring
the roll-out, charge or pour, and roll-in of
the converter. The generation of fugitive
emissions occurs  primarily during these
operations because the  primary hood is
raised and the draft to the primary hood is
closed off to prevent dilution of the strong
sulfur dioxide gases processed in the acid
or liquid S02 plants. During the blowing
phase  of the  operation, some  small

-------
  110
  100
.S
.5  90
   SO
   70
          Exhaust
                                                 Position 2
                                                 Position 1

                                                 Position 3
Jet
                                                         .V Position 4
                               _L
                 8-2
                Primary Hood
                   D-1           C-6

                         Ports
                                                         C-1
                                           •*• Front
 Figure 3.    Effect of injection position on recovery.
quantities of fugitive emissions are seen
to occasionally escape the primary hood.
Because of the large number of different
operations --  i.e., roll-in  and roll-out;
charging of  matte, anode  slag, various
reverts, and scrap; slag skims; copper
pours;  and  blowing and  holding—we
recognized that we could not character-
ize  emissions  for each  individual
condition.  Therefore,  the  test  was
structured to provide composite data for
selected operations.

Sulfur Dioxide Emissions
  The concentration of  sulfur dioxide in
the air curtain exhaust was monitored by
a  continuous  emission monitor.  More
than 470 individual data points were
utilized to characterize the  converter
emissions, resulting in an emission factor
of 3.0 kg/Mg of blister copper for the total
converter cycle and 0.1  kg/Mg when the
converter was in the blowing or standby
mode.

Paniculate Emissions
  Total filterable particulate was sampled
using  EPA Method 5.  For each of the
three converter cycles, a sample was
taken compositing all emissions over the
                             total converter cycle by traversing the
                             exhaust duct. Single point sampling was
                             used  to  obtain  a  composite  sample
                             representing the emissions during those
                             converter operations where the primary
                             hood was open, i.e., charging or discharg-
                             ing. The emission factor for the total cycle
                             was calculated as 0.45 kg/Mg of blister
                             copper for the total cycle and 0.43 kg/Mg
                             for those operations where the converter
                             was rolled out.

                             Particle Size
                               Particle  size samples  were taken by
                             impactors to define the particle size distri-
                             bution within the inhalable particulate
                             range of 10//m and less by aerodynamic
                             size. The  tests were conducted at points
                             of  average  velocity   simultaneously
                             with,but at points different from, those at
                             which  the  particulate  samples  were
                             taken. The sampling was conducted in
                             such  a manner  so as  to provide  a
                             composite over a converter cycle for each
                             major converter operating condition.

                               • Charging mode which consists of all
                                 additions to  the converter such as
                                 matte, anode slag, and cold additions
                                 such as scrap
  • Skimming mode, which consists of
    slag skimming and pouring of blister
    copper

  • Blowing mode, which consists of all
    operating  conditions during which
    the primary hood is closed, including
    the slag, cleanup, and finish blows

  The  average particle size distribution
for each mode indicates that: 1) the bulk
of the particulate  (88 to 98 percent) is
above  10  /t/m  during  blowing; 2)  the
particulate is composed of both fine and
coarse particulate (70 to 84 percent less
than 10 //m) during charging; and 3) the
particulate during skimming and pouring
is predominantly (86 to 92  percent less
than 10/um) in the inhalable range (Figure
4).

Trace Metal Emissions
  Arsenic, emitted in the form of arsenic
trioxide, was measured to determine both
the filterable particulate and gaseous
emission rates (Table 1). The  filterable
arsenic fraction represents material col-
lected  in the sample probe and on  the
filter,  both  of  which  were  heated to
approximately  121°C  (250°F).  The
gaseous arsenic  fraction  represents
material that passed through the heated
filter and condensed or was trapped in the
impinger section  of  the sample train,
which was maintained at a temperature
of 20°C (68°F) or less. In  retrospect,  the
sampling  train  should have  been
operated at the temperature of the stack
gas,  i.e.,   15° to 30°C,  to  prevent
revolatilization of  arsenic trioxide and
passage  through  the   filter.  Should
revolatilization  occur,  the   amount  of
arsenic reporting  as  gaseous would  be
increased, which, could lead to a false
conclusion regarding  the amount that
could be removed by dry collectors.
   During  Test 2,  the  gaseous arsenic
concentration was considerably greater
than in the other tests. During this test,
the  loss of draft  in the primary hood
caused by operational problems at  the
chemical  plant  resulted  in  frequent
releases of smoke and fumes from  the
primary hood. Significant quantities of
heavy  smoke escaped the primary hood
systems, and much of these emissions
were captured by the secondary hood.
Sampling continued throughout these
intermediate  upsets,  but was
finally  stopped  when the  air  curtain
control system became overwhelmed by
continuous  and  heavy  emission
discharge   from   the   primary  hood.
Therefore, it is reasonable to conclude
                                    4

-------
 53.3
  30.0
  50.0
  10.O
   J.O
   0.1
                                                          / Skimming
                                                         Charging
                                                           Blowing
                          i i 11
                                             i  i  i 111
                            1.0                    10.O

                               Particle Size, micrometers
                                                                         100.0
'igure 4.    Average particle size distribution for converter operating modes.
hat fugitive emissions generated by the
nalfunctioning primary hood draft con-
ributed to the higher arsenic concentra-
ions observed during this second cycle
est.

Conclusions
 The bulk of the fugitive sulfur dioxide
missions (over one-half) was emitted
uring  those converter operations
ivolving  the rolling  in or  out of the
onverter and  the  charging  of  cold
dditions, particularly  anode slag. The
emaining  operations in  order  of
ignificance were  matte charging; slag
kimming; and   copper pouring  and
lowing, including standby and idle. The
results of the total paniculate sampling
indicated  that the bulk of the fugitive
paniculate emissions occur during those
operations occurring when the converter
is charging or discharging. This suggest
that the  primary hood system is quite
effective   in preventing the  escape of
emissions during blowing.
  The particulate size information leads
to the conclusion that the bulk (90%+) of
the fugitive particulate emitted during
blowing is greater than 10 //m, while that
emitted   during  the  charging  or
discharging  is predominately (70 to 90
percent)  in  the  inhalable   range.
Emissions which occurred during a
process  upset   in the blowing  mode
                                         exhibited an increase in the proportion of
                                         emissions in the fine (less than 2.5 j/m)
                                         particle  range,   in  addition   to  the
                                         increase in total loading.
                                           The trace  metals contained in the
                                         particulate  emissions   comprise  an
                                         appreciable portion of the total, some 12
                                         to 40 percent by weight for charging and
                                         pouring, but only a small part (5 percent)
                                         during  blowing.  The bulk of the trace
                                         elements emitted in the fugitives tends to
                                         occur in the inhalable range with a very
                                         considerable contribution from the fine
                                         (less than  2.5 fjm) range.
                                           The  trace   metal  emissions  were
                                         dominated  by arsenic and lead that are
                                         present   in  high  quantities   in  the
                                         concentrate and  carry through to the
                                         matte  which   is   processed   in  the
                                         converter.  The potential  for trace metal
                                         fugitive emissions is then greatest during
                                         the  charging  of matte followed by he
                                         charging of reverts, scrap, matte skim-
                                         ming, copper pouring and finally blowing.
                                         Because of the greater variability of the
                                         trace element content of the feed mate-
                                         rials, the content of the emissions during
                                         charging is the most variable, followed by
                                         slag skimming and then copper pouring.
                                         then copper pouring.
'able  1.    Summary of Arsenic Emission Data
Converter
cycle No. Test
1 TC
SM
2 TC
SM
3 TC
SM
Mass emission
rate, kg/h (Ib/h)
Filter Gas
0. 33 f 0.73)
0.99(2.18)
0.61 (1.35)
1.97 (4.35)
0.21 (0.471
1.48 (3.26)
0.04 (0.08)
0.20 (0.44)
0.72(1.59)
0.99 (2. 18)
0.07(0.16)
0.05 (0. 1 1)
Emission factor
kg/Mg (Ib/tonl
0.03 (0.06)
0.03 (0.07)
0.09 (0.20)
0.05 (0. 12)
0.01 (0.03)
0,01 (0.02)

-------
      This Project Summary  was authored by staff of PEDCo Environmental, Inc..
       Cincinnati, OH 45246.
      John O. Burckle is the EPA Project Officer (see below).
      The complete report consists of two volumes:
       "Evaluation of an Air Curtain Hooding System for a Primary Copper Converter:
       Volume I" (Order No. PB 84-160 514; Cost: $ 17.50, subject to change).
       "Evaluation of an Air Curtain Hooding System for a Primary Copper Converter:
       Volume II. Appendices" (Order No. PB 84-160 522; Cost: $53.50, subject to
       change).
      The above reports will be available only from:
              National Technical Information Service
              5285 Port Royal Road
              Springfield, VA 22161
              Telephone: 703-487-4650
      The EPA Project Officer can be contacted at:
              Industrial Environmental Research Laboratory
              U.S. Environmental Protection Agency
              Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
                                      H S    w 0 u u ^ d -i
                                      U  o  LAivin   r'KuftClluN
                                      KhblC-i  b  Li>KAr
-------