EPA-450/3-88-006
   Cadmium Emissions from Primary
Lead and Primary Copper Smelting -
         Phase I Technical Report
                 Emission Standards Division
            U.S. ENVIRONMENTAL PROTECTION AGENCY
                 Office of Air and Radiation
             Office of Air Quality Planning and Standards
            Research Triangle Park, North Carolina 27711

                     June 1988

-------
This report has been reviewed by the Emissions Standards Div.s.on of the Office of Air Quality Planning and
Standards EPA and approved for publication. Mention of trade names or commercial products is not intended
to constitute endorsement or recommendation for use. Copies of this report are available through the Library
Services Office (MD-35), U.S. Environmental Protection Agency, Research  Triangle Park NC 2771 1, or ,rom
National Technical Information Services, 5285 Port Royal Road, Springfield VA 22161

-------
                            TABLE OF CONTENTS

                                                                 Page

     DEFINITION OF  SOURCE CATEGORIES  	      1

     A.    Primary Lead  Smelters  	      1

          1.   PI ants i n Operati on 	      1
          2.   Processes  	      1
          3.   Projections  of  Industry Growth 	      3
     B.    Primary  Copper  Smelters
          1.    Plants  in  Operation  	      3
          2.    Processes  	      3
          3.    Projections of  Industry Growth  	      5

II.   EMISSIONS AND  CONTROLS	      5

     A.    Primary Lead Smelters  	      5

     B.    Primary Copper  Smelters  	      9

III.  PUBLIC HEALTH  RISKS  	     15

     A.    Approach  	     15

          1.    Primary Lead  Smel ters  	     16
          2.    Primary Copper  Smelters  	     17

     B.    Results....	     17

          1.    Primary Lead  Smelters  	     17
          2.    Primary Copper  Smelters  	     19

IV.   POTENTIAL FOR  IMPROVED  CONTROL 	     19

     A.    Primary Lead Smelters  	     19
     8.    Primary Copper  Smelters  	     26

V.   REFERENCES 	     30
                                     11

-------
                             LIST OF FIGURES
                                                                 Page
Figure
1         The Lead Smelting Process  	       2
2         The Copper  Smelter Process 	       4
3         Cadmium Distribution  in the Lead Smelting
            Process		       6
4         Cadmium Distribution  in the Copper Smelter
            Process 	      12
                                     fv

-------
                               LIST OF TABLES
Table

1         Inventory  of Emissions Sources and Controls
            for Primary Lead  Smelters  	     10

2         Inventory  of Emissions Sources and Controls
            for Primary Copper  Smelters 	     13

3         Summary of Emissions  and  Risks for
            Primary  Lead Smel ters 	     18

4         Summary of High Risk  Sources at Lead
            Smelters 	     20

5         Summary of Emissions  and  Risks for
            Primary  Copper Smelters 	     22

6         Summary of High Risk  Sources at Copper
            Smelters 	     24

7         Costs of Improved Control  for High Risk
            Sources  at Lead Smel ters  	     27

8         Costs of Improved Control  for High Risk
            Sources  at Copper Smelters 	     28

-------
                             TECHNICAL REPORT:
                      PRIMARY LEAD AND COPPER SMELTING

   I.  DEFINITION OF SOURCE CATEGORIES
      Cadmium, as an  impurity  in many nonferrous ore concentrates,  is
emitted during the smelting process.  This report describes the  sources  of
cadmium emissions from primary lead and primary copper smelters, documents
the health risks attributable to these emission sources, and discusses the
potential for improvements in existing levels of control.
      A.  Primary Lead Smelters
      1-  Plants in operation.  There are currently five primary lead
smelters located in three States.  The following three plants are currently
operating:  the ASARCO smelter in East Helena, Montana; the ASARCO  smelter
in Glover, Missouri;  and the St. Joe smelter in Herculaneum, Missouri.   The
lead smelter in Boss, Missouri, previously owned by a partnership of AMAX
and Homestake Lead Tollers, was acquired in full by Homestake in June 1986.
This smelter, referred to as AMAX-Boss for this study,  was immediately
closed after Homestake's acquisition.   The ASARCO smelter in El  Paso,
Texas, has been closed since 1985.   Both ASARCO and Homestake have  stated
that no reopening date has been scheduled, and a decision to reopen these
smelters will be based on market conditions.1'2
      2.  Processes.   Figure 1 shows the lead smelting  process.   Processes
include sintering,  reduction,  and refining.   Ore concentrates,  that
typically contain 20  to 70 percent  lead and  13 to 19 percent sulfur, are
blended with return sinter, flue dusts,  and  fluxes and  are fed to a
sintering machine.3   Sintering results in the oxidation of lead  and sulfur
present in the charge to lead  oxide (PbO)  and sulfur dioxide (S02).
Simultaneously,  the charge material  is converted to a dense permeable
material called sinter.   The sinter is crushed,  mixed with metallurgical
coke,  and charged to   a blast furnace where carbon monoxide formed by
burning coke reduces   PbO to elemental  lead.   In addition,  impurities with
an affinity for oxygen,  such as iron,  react  with the fluxes and  are

                                      1

-------
                 ACID PLANT
CONCENTRATES
SECONDARIES
               SINTER MACHINE
                                  STACK
                           SLAG
BLAST FURNACE
                                             SU
                                               \ t
 FUMING FURNACE
  (OPTIONAL)
 DROSS ING
                                                                       STACK
                                                                        1
ZnO PRODUCT
   BH
                                          LEAD
                                          BULLION
                                  Figure 1.  The lead smelting process.

-------
 collected in a molten slag.  3oth the slag and lead bullion ars tapped
 continuously from the furnace.   The waste slag is either disposed of
 directly, or,  as is the case at two smelters,  it may be charged to a zinc
 fuming furnace where the zinc in the slag is vaporized to zinc oxide.  The
 residual  slag  is then discarded.
       The crude lead bullion is transferred to a series of refining kettles
 where impurities are selectively removed as drosses.  Typically,  the
 drosses are treated in a reverberatory furnace where additional lead is
 recovered and  metals, such as copper and arsenic,  are concentrated.   The
 dross is  skimmed and recycled while the product lead bullion is pumped from
 the dressing kettles into molds.
       3-   Projections of industry growth.   Federal  regulations on lead in
 gasoline, in existence since 1976,  have reduced the demand for lead.
 Depressed lead prices and foreign competition  have  resulted in the
 temporary closure of two of the five domestic  smelters.   No growth in lead
 demand is projected through the year 2000  and  no  new smelters  are expected
 to  be built.3
       B.   Primary Copper Smelters
       l-   Plants  in operation.   There  are  currently 14  primary copper
 smelters  located  in 7 States.   Nine  smelters are operating  and five  are
 closed.   Of  the  nine smelters which  remain  open, two  are  scheduled to  close
 permanently  in  1987.   The  five  smelters  which  are closed  now could reopen
 if  market conditions  improve.
       2.   Processes.   Figure 2  shows the copper smelting  process.  The
winning of copper  from ore  concentrates  involves the following  processes:
 (a) roasting the  ore  concentrates (optional),  (b) smelting the  roasted
calcine or unroasted  ore concentrates and fluxes to form  an intermediate
copper-bearing product called matte, and (c) converting (oxidizing) the
matte to  produce blister copper  (approximately 99 percent pure copper).
Six of the 14 smelters roast ore concentrates prior to smelting.  In this
step,  ore concentrates, which typically contain 15 to 30 percent copper,
are charged to  the roaster.  The concentrates are heated,  but not melted,
in the presence of oxygen so that much of the sulfur in the concentrate is
oxidized to S02.  Volatile impurities such as arsenic also are emitted
during this process.

-------
                                                                     MAO
cm r
ACID PI AH I
-j



£SP


KOASII!*
(OI'IIUIlAl)



CALCINE
STACK
1
ESP



SHU TING
FURNACE
I
<>i i
/
\

MATTE
(OPTIONAL)
ACID PLANT 1
;

CSP
	 	 3

CONVERTER

BLISTER
COPPER
	 -SI AS
                             Figure 2.  The copper smelting process.

-------
      The roaster product  (calcine), or, if the roasting process  is not
used, ore concentrates mixed with flux, are then charged to a smelting
furnace.  Here, copper and iron combine with sulfur to form the molten
matte.  Lighter impurities, such as zinc and oxides of iron, silicon, and
calcium, float to the top  of the furnace and are drawn off as slag which is
then either recycled to the smelting furnace or discarded.  The copper-
rich matte, a mixture of copper sulfides and iron sulfides, is tapped from
the bottom of the furnace  and is transferred to the converters by ladle.
In the converter, fluxes are added and air is blown through the matte.
First, the iron sulfide is oxidized to various iron oxides which  are
skimmed off as slag and discarded or recycled.  This slag skimming
operation leaves impure copper sulfides.  Next, the copper sulfides are
oxidized to copper and SC^.  The converter operation is terminated at this
point to prevent further oxidation of the copper to copper oxide.  The SOg
laden off-gases are ducted to control devices for particulate removal and
then are either used to produce byproduct sulfuric acid (at 11 smelters) or
emitted directly to the atmosphere (at 3 smelters).  Typically, the product
blister copper is further  refined in an anode furnace prior to casting of
copper anodes for electrolytic refining.  A more detailed discussion on the
copper smelting process is contained in Reference 4.
     3-  Projections of industry growth.  Demand for primary copper is
expected to grow at 1.7 percent annually through the year 2000.5  A world
surplus of mine, smelter,  and refinery capacity is expected to exist
through 1990 leading to a  permanent shift in smelter and refinery capacity
away from the U.S.5  No additional  copper smelters are anticipated to begin
operating in the near future.   Two of the currently operating smelters are
scheduled for permanent shutdown in 1987,  and only 9 of the 14 smelters are
currently operating.

II.  EMISSIONS AND CONTROLS
     A.   Primary Lead Smelters
     As an impurity in lead ore  concentrates,  cadmium is  emitted  in  varying
amounts at each stage of the  smelting process.   The typical  lead  ore
concentrate contains  0.02 percent  cadmium.   The distribution  of cadmium
through the lead smelting process  is  shown  in  Figure 3.   Approximately half
                                      5

-------
                              STACK
             ACID PLANT
                 Bit
100 UNITS
CADMIUM
50 VOLATILIZED
           SINTER MACHINE
                              SO
                       SLAG
                                          STACK
                                           BH
                                            48 VOLATILIZED
                                      RIAST  FURNACF
                                          T
                                         SLAfi
                                     MCE
                                                                   STACK
                                                                         TRACE
GROSSING
                                                                  STACK
-ZERQ-M
                                                                                LEAD
                                                                                BULLION
FUMING FURNACE
(OPTIONAL)
\
>
ZnO PRODUCT
BH
            Figure 3.  Cadmium distribution  1n  the  lead  smelting  process.

-------
  of  the  cadmium  input  to  the  smelter  is  volatilized  in  the  sintering  step
  and half  is volatilized  in the blast furnace with small  amounts  remaining
  in  the  slag.  Other point sources of cadmium emissions at  primary  lead
  smelters  include dross furnaces, slag fuming furnaces, and, at one plant, a
  cadmium oxide roaster.   These sources are discussed in more detail below.
      At three of the  five smelters, the sinter machine produces  two off-gas
  streams.  One stream  is  taken from the charging end where the majority of
  the sulfur combustion takes place.   This stream has a high S02 content
  (typically 4 to 7 percent) and is referred to as the strong off-gas
  stream.6  The second off-gas stream,  taken from the discharge end of the
 machine, has a low S02 content (typically 0.5 percent) and is referred to
 as the weak off-gas stream.6  The S02 in the strong off-gas stream can be
 recovered economically in a byproduct sulfuric acid plant.   The gas is
 passed through a series of particulate  control  devices prior to acid
 conversion.   A baghouse and/or an ESP is first used to reduce the
 particulate  matter concentration.  Particulate loadings are reduced further
 by wet treatment devices-scrubbers  and  spray towers.   This extensive
 pretreatment is  required  to  keep  the  particulate matter from poisoning the
 acid catalyst  conversion  beds.   Thus, virtually no  particulate matter or
 cadmium emissions  are  discharged  from the  acid  plant stack.   Sinter machine
 weak stream  off-gases  contain too little S02  to be  converted to  sulfuric
 acid economically.   Therefore,  they are  emitted to  the  atmosphere after
 baghouse control.
     The sinter  machine at the ASARCO smelter  in  El  Paso  recirculates  the
 weak stream  off-gases  with the strong off-gases  to produce  a single off-
 gas  stream (from the charging end of the sinter  machine)  that  contains
 approximately 6  percent S02.  This stream is processed through an acid
 plant.   The ASARCO-Glover smelter has no S02 control.  Consequently, there
 is no concern to maintain a strong stream for an economical acid  conver-
 sion.  Thus,  the sinter machine off-gas  is a single stream containing gas
 from che enure length of the sinier machine.   It is vented through a
baghouse for particulate control,  then to the atmosphere.
     Process  gases from blast furnaces at all  five smelters are vented to  a
baghouse for  particulate matter removal  and  discharged  directly to the

-------
 atmosphere through  stacks.  The  S02  concentration  in the  blast  furnace
 off-gas  is too  low  to  be  recovered economically  in  an  acid  plant.
      At four of the five  smelters,  the blast  furnace  baghouse  dust  is
 recycled to the sinter machine.  At  the ASARCO smelter  in El  Paso, Texas,
 however, the blast  furnace dust, which can contain  up  to  18  percent
 cadmium, is roasted in a  Godfrey roaster to produce a  crude  cadmium  oxide
 byproduct (66 percent  cadmium).  Cadmium oxide product  from  the  roaster  is
 collected in a  baghouse which  is vented through  a  louvered  cupola.
      Two smelters  have slag fuming  furnaces for zinc  recovery:  ASARCO-
 £ast Helena, and ASARCO-E1 Paso.  Neither has operated  the  furnace since
 1982.  A decision to restart the furnaces will be  based on the zinc  market.
 Because of the  possibility that these facilities may be restarted, they
 were included in the source assessment.  However,  since slag  contains very
 small quantities of cadmium, the slag fuming furnace is a relatively minor
 source of cadmium emissions.
      Process fugitive emissions result from leakage in and  around process
 equipment and from the handling and  transferring of materials.   Control
 techniques include measures to preclude their occurrence and  measures to
 confine and capture emissions.  Once captured, emissions may  be  vented
 directly to a particulate control device or combined with process gases
 prior to control.
      Major process fugitive sources include the handling and transfer of
 sinter machine  and blast furnace flue dusts, sinter machine  leakage, sinter
 crushing and handling, blast furnace charging, blast furnace  upsets, and
 hot metal tapping and transfer operations.   Process fugitive  emissions from
 sinter charging and product handling and transfer operations  can be  limited
 oy the use of dust suppressants (e.g., water sprays) and the  application of
 local hooding and ventilation followed by a control device.    Local hooding
 and ventilation techniques are also  applied to the control of tapping
 emissions from blast furnaces and dross reverberatory furnaces,  and  the
 control  of drossing kettles.   Captured emissions  are generally combined
with the blast furnace process gases  and treated  in the blast furnace
 baghouse.  Process fugitive emissions- from  sinter machine and furnace
 leakage  can  be minimized or eliminated by proper  operation and maintenance.
Control  of process fugitives  from flue dust handling includes enclosing
                                      8

-------
  dust transfer points and conveyors and/or application of dust suppressants.
  Many of these controls have been put in place to comply with State
  implementation plans (SIP's)  for fugitive lead emissions.
        The Texas Air Control  Board (TACB)  has  required the  ASARCO smelter in
  El  Paso to enclose  all  transfer and  handling  of dusts containing greater
  than 1  percent lead.   If ASARCO decides to  reopen  the lead plant it  must
  comply  with  this  mandate. ?> 8   The enclosures  put in  place  tQ control  ^
  emissions also will  significantly reduce  cadmium emissions.
       The Montana lead  SIP  has  required the ASARCO smelter in  East Helena
  to  install a  number  of  control  measures that  have reduced  lead and cadmium
  emissions.  These include improved ventilation  and covers  for  the dross
  kettles,  an air pressure control  system to reduce blast furnace  upsets,  and
  a blast furnace secondary hooding system vented  to a  baghouse.   Montana
  estimates the  particulate matter  capture efficiency to be  95 percent for
  the blast furnace and 50 percent  of the drossing operations.9*^
       Open fugitive sources  include ore concentrate storage piles, slag
 storage piles, and plant roadways and yard areas.  In the case of ore
 concentrates and slag,  the cadmium content is  low (approximately 0.02
 percent for ore and  0.005 percent for slag).   As a  result,  potential
 cadmium emissions from  these sources  are relatively minor.   Similarly,
 emissions from plant roadways  and yard areas are also minor,  due  to both
 low cadmium content  and measures taken to  mitigate  lead emissions.   Control
 for open fugitives consists  of wet suppression or enclosing the storage
 piles.
       An inventory of cadmium  emission sources at primary lead  smelters and
 the  existing  controls is presented in  Table  1.   In general,  the SIP's  for
 lead  have  forced  or  driven the  application of  best available  technology
 (BAT)  for  most  of  the potentially  significant  sources  of cadmium  emissions
 If the ASARCO  plant  at  El Paso  is  reopened then  the SIP's for lead would
 require  enclosures for dust handling processes  significantly  reducing
 cadmium  emissions  from this operation.
         B.  Primary Copper
      As an impurity in copper ore concentrates,  cadmium is emitted in
varying quantities at each stage of the smelting  process.   The typical
copper ore contains 0.01 percent cadmium.   The distribution of cadmium

-------
 TABLE 1.   INVENTORY OF EMISSIONS SOURCES AND CONTROLS
                FOR  PRIMARY  LEAD SMELTERS
Plant/Location
Anax,
Boss, MO














ASARCO, East
Helena, MT









ASARCO
Glover, MO






St. Joe,
Herculaneum,
Mo.







ASARCO,
El Paso, TX


Source
Main stack; blast furnace
weak sinter
Sinter/flast furn. fugitives
Sinter prep/ returns
Sinter prep/returns
Blast furnace charge preo
Blast furnace charge prep
Blast furnace cnarge oreo
31 ast furnace charge orep
Combined area fugitives

Combined process fugitives



TOTAL
Sinter machine
Zinc fum* stack
Blast/reverb furnaces
Combined area fugitives

Combined process fugitives




TOTAL
Sinter machine (main stack)
Blast furnace
Sinter returns fugitives
N1x1ng/pellet1z1ng cones.
Combined area fugitives

Combined process fugitives
TOTAL
Main stack: add plant (AP)
sinter machine blast furnace
Sinter discharge
Drums/conveyors
Rolls/conveying
Vib/conveyor belts
Returns conveyor
Sinter plant fugitives
Combined area fugitives
Combined process fugitives
TOTAL

Sinter machine (lead stack)
Blast furnace stacks
Zinc fume bagnous* stack
Pb unloading stack
fb bedding Bldg. stack
Cadmium oxide production
Combined area fugitives
Combined nrocess fugitives -
dust handl ing
TOTAL
Type*
H

H
H
H
H
H
H
H
F

F




H
H
H
F

F





H
H
H
H
F

F

H
H
H
H
H
H
H
F
F


H
H
H
H
H
H
F
e

Emissions
kg/yr
1,150.0

18.0
40.0
36.0
111.0
108.0
18.0
11.0
19.7

498.0



2,010.0
240.0
75.0
2,502.0
105.5

146.0




3,070.0
598.0
247.0
3.5
10. 0'
39.8

40.3
939.0
3,470.0
15.0
1.3
50.0
36.0
0.2
3.0
7.9
82.8
3,660.0

300.0
1,810.0
91.0
13. S
104.0
2,720.0
3.3
1,510.0
6,660.0
Control^
Saghouse

Baghouse
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
• Enclosures, *et
suppression
Enclosed conveyer.
dust storage.
8F air control ,
wet suppression

Sagnouse
Baghouse
Baghouse
Wet suppression,
sweeping
BF fugitive
hoods, 8F air
control , dross
charging covers
and ventilation

Baghouse
Bagnouse
Baghouse
Scrubber
Met suppression.
sweeping


Baghouse
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Enclosure, wet
suppression


Baghouse
Baghouse
Sagfiouse
Baghouse
Saghouse
Baghouse
Enclosure, wet
suppression
«et suppression
of dust

aH * point source;  F * fugitive source.

DBF » Slast furnace;  AP • add plant.
                                   10

-------
 through the copper smelting process  is shown  in Figure 4.   If the  smelter
 uses a roaster, approximately 15 percent of the incoming cadmium will  be
 volatilized in the roasting step.  The remainder stays with the calcine and
 is charged to the smelting furnace.  If there is no roaster, concentrates
 are charged directly to the smelting furnace.  About one-third of  the
'cadmium input to the smelting furnace is volatilized and two-thirds is
 retained in the matte.  The matte passes from the smelting furnace to  the
 converters where approximately 80 percent of the remaining cadmium is
 volatilized.   The remaining 20 percent of the converter input is  entrained
 in the converter slag which is returned to the smelting furnace or is
 discarded.
      Since copper ore concentrates contain a significant amount of sulfur,
 most copper smelters  also have acid plants to convert the S02-containing
 off-gas streams to sulfuric acid.   Off-gases from  the converters,  and  less
 often  from the smelting furnace  if a  flash or electric smelting furnace is
 used,  are  pretreated  for particulate  matter removal  and vented  to  an  acid
 plant.  At  ASARCO-E1  Paso  multiple  hearth  roasters  are used  with the  off-
 gases  treated  at  an acid plant.   The  acid  plant  tailgas is virtually  free
 of particulate matter and  cadmium.  Pretreatment is  usually  accomplished by
 an ESP  followed by wet  treatment devices.
      A review of  the cadmium  emission  sources and  their  corresponding
 controls currently applied  is  presented in  Table 2.   Electrostatic
 precipitators  are  the control  most  often used  on process  off-gases  for
particulate control.   Eleven smelters have  an  acid plant  that employs
additional wet treatment devices to reduce  particulate  from process off-
gases from roasters and  converters.   The two plants  that  stand out  as high
emitters are Phelps Dodge-Douglas and Kennecott-McGill.  Neither facility
has acid plant control on the converter off-gases.    In  the case of  Phelps
Dodge-Douglas, the converter off-gas stream is vented directly to the
atmosphere after ESP  control, and for Kennecott-McGill  the converter off-
gas stream is vented  to the atmospnere after muiticlone controi.  These cwo
plants are presently not operating.  If the plants  are restarted,  then acid
plants would have to be installed in order to meet  S02 standards.   The
cadmium emissions  from the two  plants  would then  be sharply reduced.
                                     11

-------
'.IACK
                                                      STACK
    NIL
I

100 UN US CADMIUM
MCONUAItU'j
ACID J
J
4
'IAN I

>l>

VOUIIimD
KOASIIfi
(01* II 1)11 At )


8',
CALCINE '
STACK
t
ESP

2U VOLAIIUHO
SC.ELTING
FURNACE
•r
X"
i;
MATTE J
(OPTIONAL)
AC 10 PLANT
i

ESP
j
4/

1/OlAdl I/ID
CONVERTER
- teru-
8LISTER*
COPPER
1
 Figure 4.  Cadmium  distribution in the copper smelting process.

-------
 TABLE  2.   INVENTORY  OF EMISSIONS SOURCES  AND  CONTROLS
                  FOR  PRIMARY  COPPER  SMELTERS
Plant/Location
ASARCO, £1 ?aso, TX






ASARCO, Hayden, AZ



Cities Service
Coppernill, TN

Copper Range,
White Pine, MI




Inspiration Copper,
Miami, AZ



Kennecott Copper,
Garfleld, UT


Kennecott Copper,
Hayden, AZ




Kennecott Copper,
Hurley, NM


Source
Main stack 'smelting furnace)
Calcine slag matte tapping
Combined area fugitives
Converter fugitives

Cu unloading stack
Cu transfer stack
Cu bedding bldg. stack
TOTAL
Main stack (all processes)
Matte and slag tapping
Comoined area fugitives
Converter fugitives
TOTAL
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Main stack (all processes)
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Furnace/converter/AP stack
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Main stack (all processes)
Matte and slag tapping
Combi-ned area fugitives
Converter fugitives
TOTAL
Smelting furnace stack
Roaster/converter/AP stack
Caldne/matte/slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Main stack (all processes)
Matte/ slag tapping
Comoined area fuqitlves
Converter fugitives
Type
y
F
F
F

H
rl
H

H
F
F
F

F
F
F

H
F
c
F

H
F
F
F

H
F
F
F

H
H
F
F
F

H

-
r
Emissions
kg/yr
998.0
53.0
54.2
54.0

204.0
10.2
104.0
1,270.0
775.0
42.0
54.5
33.0
954.0
9.0
46.3
14.0
69.3
102.4
1S.O
3.1
42.0
164.0
30.2
30.0
1.4
84.0
146.0
640.0
49 .0
27.5
137.0
854.0
57.0
25.5
48.0
3.7
77.0
211.0
827.0
36 .0
36 .0
98.0
Control
ESP
3agnouse 'matte)
£SP (calcine)
--
2°hoods,
bagnouse
Baghouse
Sagnouse
Sagnouse

ESP, AP
ESP
wet suppression
ESP

--
--

Reverb: ESP.
Converter:
baloon flue
--
•-


ESP, AP
--
•-
m _

ESP, AP
--
""

ESP
Cyclone,
scrubber, AP
--

*~

ESP. AP
"


Kennecott Copper,
  McGIll, MY
                               TOTAL

                    Main stack (all processes)
                    Matte and slag tapping
                    Combined area  'ugitives
                    Converter fugitives

                                'OTAL
  997.0

4,850.0
   33.0
   0.1
   91.0

4,980.0
Reverb:  ESP.
Converter:
  cyclone

-------
TABLE.2  (cont.
.   INVENTORY OF  EMISSIONS  SOURCES  AND
 FOR PRIMARY COPPER  SMELTERS
                                                           :ONTROLS
   Plant/location
                    Source
Magma Copper,
San Manuel , AZ




Phelps Dodge,
Ajo, AZ



Phelps Oodge,
Douglas, AZ







Pnelps Dodge,
Hidalgo, MM





Pnelps Dodge,
Morencl, AZ



Smelting furnace stack
Converter/ AP I stack
Converter/AP 2 stack
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Main stack (all processes)
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Main stack (roaster,
smelting)
Converter stack
Calcine/matte/slag tapping


Contained area fugitives
Converter fugitives
TOTAL

AP stack
Slag furnace stack
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
Converter/AP stack
Matte and slag tapping
Combined area fugitives
Converter fugitives
TOTAL
H
H
rf
F
F
F

H
F
F
F

H

H
F


F
F


H
H
F
F
F

H
F
F
F

«9'vr
378.0
19.3
19.3
37.0
5.3
240.0
749.0
1,990.0
12.0
0.2
32.0
2,030.0
3,540.0

12,000.0
101.0


2.5
160,0
15,800.0

1,640.0
16.3
54.0
31.1
147.0
1,880.0
42.0
31.0
36.7
85.0
195.0
tontroi
ESP
ESP, AP
ESP, AP

	
--

ESP, AP

--

ESP

ESP
Sagfiouse
(calcine only) ,
ESP (other)

--


ESP, AP
Scrubber

...
--

ESP, AP

„_
--

    AH - point source;  F « fugitive source.

-------
       Process fugitive emissions for each copper  smelter  also  are  presented
 in Table 2.  The largest source of fugitive emissions  is  the converters.
 An emission factor for process fugitive cadmium emissions from converters
 was developed from a test performed in 1983.n  Emission  factors for matte
 and slag tapping fugitive emissions from reverberatory furnaces were taken
 from the previous source survey (GCA, 1981).12  These  factors  indicate that
 process fugitive emissions are about three times greater  from converters
 than from reverberatory furnaces-roughly the same proportion as the point
 source emissions discussed previously.   Process fugitive emissions from
 roasters were considerably lower than fugitive emissions from the
 reverberatory furnace and the converter.   Process fugitive emissions from
 roasters and smelting furnace tapping operations (matte and slag) are
 generally captured by local  ventilation systems.  Roaster emissions are
 typically controlled in a baghouse  or ESP,  while the furnace tapping
 emissions are typically discharged  to the  atmosphere uncontrolled.   Six
 smelters have some degree of secondary  or  fugitive emissions capture.   In
 only  two cases,  the ASARCO  El  Paso  and  Hayden  smelters, are  the capture
 systems  considered to be  effective  and  the  captured emissions  controlled.
 The El  Paso  smelter uses  a  baghouse  for control  and the Hayden  smelter  uses
 an ESP.   Except  for these plants no  primary  copper smelters  employ  process
 fugitive controls  on  roasters,  smelting furnaces,  or  converters.

 HI.  PUBLIC HEALTH  RISKS
      A.  Approach
      Risk assessment is the process used by EPA to develop quantitative
 estimates of public health risks associated with individual and population
 exposure to a hazardous or toxic air pollutant.  The resultant estimates
 are considered by EPA to be rough but plausible upper-bound approximations
 of the risks.  Two measures of risk are calculated.  One is maximum
 individual risk and the other is aggregate  risk.  Maximum individual risk
 (MIR)  is an estimate of the probabincy  of  contracting cancer experienced
by the person or persons exposed to  the  highest predicted annual average
concentration of the pollutant.   Aggregate  risk is an estimate of the
increased number of cancer cases for the  entire population exposed.   It  is
expressed in  terms  of annual  incidence or number of cancer cases per year.
                                     15

-------
      The estimates are calculated by coupling a numerical constant that
defines the statistical exposure-risk relationship for a particular
hazardous pollutant with estimates of public exposure to the pollutant.
The numerical constant used by EPA in its analysis of carcinogens is called
a unit risk factor.  It represents an estimate of the increase in lifetime
cancer risk occurring to a hypothetical  individual exposed continuously
over a lifetime (70 years) to a concentration of 1 microgram per cubic
meter (ug/m3) of the pollutant in the air the individual breathes.  For
cadmium, the unit risk factor is estimated to be 1.8x10-3 or 1.8 chances in
1,000.13
      Estimates of public exposure are derived using an atmospheric
dispersion model and census data contained in EPA's Human Exposure Model
(HEM).14  Dispersion models are used to predict concentrations of a
pollutant in the ambient air at varying distances and directions from a
stationary emission source.  By combining the predicted ambient
concentrations with population distribution data, both the number of people
exposed and their levels of exposure can be calculated.  Needed inputs to
the model include plant location (latitude and longitude) and source
specific data including the height and area of the release point,
temperature and velocity of exit gases,  and emissions rate for each source
to be modeled.14  Details of the methods used to develop the health risk
estimates for both source categories are described below.
      !•  Primary lead smelters.  The inputs to the HEM, including emission
estimates and stack parameters, for primary lead smelters were generated
from Section 114 request responses, test reports, trip reports,  State
implementation plans (SIP's)  for lead emissions,  and the previous source
survey (Radian, 1985).   Although two smelters are currently closed,  plans
are for both smelters to reopen when the market for lead improves.
Therefore, all five primary lead smelters were included in the risk
assessment.
      All stack sources ac lead smeiters were modeled individually.
Fugitive emission estimates were modeled in separate groups.   Process
fugitive emissions were modeled in several  different ways based  on how the
States had estimated the emissions.  At  ASARCO-East Helena and ASARCO-
                                     16

-------
  Glover,.process  fugitive  emissions  tests  had  been  conducted  by MRI  in
  1977.15  At  these  smelters,  the  buildings  from  which  the  process  fugitives
  escaped  (the  sinter  building,  the blast furnace building,  and  the dressing
  building) were modeled  as  individual  sources.   For  the  other smelters the
  lead  SIP estimates for  lead  process  fugitives were  multiplied  by  a
  cadmium-to-lead  ratio and  modeled as  a group.   Open fugitives  were  modeled
  as one source consisting of  ore  storage and slag pile emissions.
       2-  Primary copper smelters.  The HEM inputs  for  primary  copper
  smelters were generated from Section  114 request responses, EPA-sponsored
  and State-sponsored tests, trip  reports, the Background Information
  Document for Arsenic emissions (Reference 4), and the previous  cadmium
  source survey for primary copper smelters (GCA,  1981).  All 14  copper
  smelters, whether operating, closed, or scheduled for closure,  were
  included in the risk assessment.   This was done because the modeling  of all
  14 smelters was manageable and because of uncertainties regarding whether
 closures and reopenings  will actually occur.
       The emissions from copper smelters were grouped  by source types for
 the HEM modeling.  All  stacks were modeled as individual sources.   In
 addition,  each smelter was modeled with  three types of fugitive emissions:
 area  fugitive emissions, combined process  fugitive  emissions,  and  converter
 fugitive  emissions.   Area  fugitives  are  comprised of emissions  from  ore
 storage  and  slag  piles.   Combined process  fugitives include emissions from
 matte  and  slag tapping of  the smelting furnace,  and, if  roasters are used,
 calcine transfer  emissions.   Converter fugitive  emissions,  the  major
 fugitive  source,  were modeled separately.
       8.  Results
       1.  Primary lead smPltPrs.   Taole 3 shows  the  emissions,  maximum
 individual risk,  and annual  incidence  for each plant.  Three smelters  had
 maximum individual risk  exceeding or approaching IxKT4:  ASARCO-E1  Paso
ASARCO-East Helena,  and AMAX-Boss.  ASARCO-E1  Paso clearly has the highest
maximum individual risk  (2.03xlQ-3),  and El Paso's annual incidence of
0.056 is 78 percent  of the total incidence  for the source category   Table
3 also shows that  the  HEM model  predicts  that  there  is  less than one  person
located at the point of maximum risk  meaning,  in  general, that  the
                                     17

-------
                                 TABLE 3.  SUMMARY OF EMISSIONS  AND  RISKS FOR PRIMARY LEAD SMELTERS
00
Plant/Location
ASARCO, El Paso, TX
ASARCO. fcast Helena, MT
ASARCO, Glover, MO
AMAX, BOiS, MO
St. Joe, Herculaneum, MO
Operating
Status
Closed
Open
Open
Closed
Open
Cadml urn
Emissions,
kg/yr
6,820
3,070
940
2,010
8,660
21,500
Maximum
Individual
Risk
2.00xlO-3
1.30xlO-4a
7.16xlO-7
9.58xlO'5a
3.47x10-5
Incidence
case/yr
0.056
0.0005
0.0002
0.0004
0.015
0.072
People
Maximum
concen-
tration
<1
<1
123
<1
3
exposed to:
Total
493,000
48,600
97,300
52.800
1,500,000
        a  Review of USGS maps Indicates It Is  unlikely  that anyone resides at location of maximum risk predicted by HEM.
           More plausible MIR estimates are 5.97x10-5  for ASARCO-East Helena and 4.5x10-5 for AMAX-Boss.

-------
population  i'mmediatsly  surrounding the. smelter  is sparse.3   Subsequently,
USGS topographic maps were analyzed to determine the location of the  point
of maximum  concentration predicted by the HEM.  The maximum  concentration
and point of maximum risk was predicted to be 200 meters from the center of
each smelter.  For the  ASARCO-East Helena smelter, this point appeared  to
be on plant property.   A more plausible point of maximum risk lies at 500
meters from the center  of the plant.  Using this point, a more plausible
value for maximum individual risk is 5.96xlO"5.  The AMAX-Boss smelter  is
located in  a national forest.  The more plausible point of maximum risk is
1,000 meters from the smelter.  The more plausible maximum individual risk
that corresponds to this point is 4.5x10-5.  For the ASARCO-E1 Paso smelter
it was concluded that people could reside at the point of maximum
concentration.
      2.  Primary copper smelters.  Table 4 shows the emission rate and
public health risks for each smelter.  All smelters have less than 1x10-4
MIR and incidence below 0.01 case per year.  The total annual incidence for
the source  category is  0.008 case.  The highest risks and incidences  in the
copper category were for fugitive emissions from copper converters and
smelting furnaces.  However, these sources had lower risks than most  of the
sources in  the lead category.  The converter fugitives with  the highest MIR
all had risks considerably lower than 1x10-4.  The highest risk for any one
source (3.9x10-5) was for converter fugitives at Magma Copper, San Manuel,
Arizona.

IV.   POTENTIAL FOR IMPROVED CONTROL
      A.  Primary Lead Smelters
      The potential  for improved control  was evaluated at the sources that
drive the health risks at the three high-risk primary lead smelters
     aThe HEM model uses population data from the 1980 census.  Population is
divided into block enumeration groups wnicn are defined rougnly oy the
numoer_of people a census taker can cover in 1 day.   Each enumeration group
is assigned a location.   If an enumeration group is  within 3.5 km of the
source, the enumeration  group's population is apportioned among 96 receptor
sites at which the HEM model  predicts pollutant concentration.  If there is
only one block enumeration group within 3.5 km and its population is less
than 36 persons,  fewer than one person is assigned to each receptor site.
                                      19

-------
                                  TABLE 4.  SUMMARY OF EMISSIONS AND RISKS FOR PRIMARY  COPPER SMELTERS
10
O
Plant/Location
ASARCO, El Paso, TX
ASARCO. Hay den, AZ
Cities Service,
Copperhlll. TN
Inspiration Copper,
Miami, AZ
Kennecott Copper,
Garfleld. UT
ASARCO (Ray Smelter)
Hayden, AZ
Kennecott Copper,
Hurley. M
Kennecott Capper,
McGIll, AZ
Copper Range,
White Pine, MI
Magma Copper,
San Manuel , AZ
Phelps Dodge, Douglas, AZ
Phelps Dodge, AJo, AZ
Phelps Dodge, Hidalgo, NM
Phelps Dodge, Morencl, AZ
Operating
Status
Open
Open
Closed

Open

Closed

Closed

Open

Closed

Open

Open

Closed
Closed
Open
Closed
Cadmium
Eml sslons.
kg/yr
1,270
950
70

150

850

210

1,000

4,980

160

750

15,790
2,030
1,880
190
Maximum
Individual
Risk
3.90xlO-5
5.35xlO-5
3.56xlO-5

2.14x10-5

2.50X10'7

3.93x10-5

5.42x10-5

4.61x10-5

9.43x10-6

5.84x10-5

5.41x10-5
1.37x10-5
6.37xlO-6
4.60x10-5
People exposed to:
Incidence
case/yr
0.0042
0.0001
0.0001

0.0001

0.0005

0.0001

0.0001

<0.0001

<0.0001

0.0002

0.0017
0.0003
0.0001
0.0003
Maximum
concen-
tration
<1
1
<1

<1

2

1

<1

<1

<1

<1

1
114
909
2
Total
493,000
46,800
164,000

35,700

810,000.

46,800

26,300

7,350

16,900

211,000

31,100
6^600
2,*560
25,500
                                                      30,300
0.0079

-------
described previously.  These high  risk  sources  are  shown  in Table  5.   Three
sources  at ASARCO-E1  Paso  are  responsible for more  than 95 percent  of  the
incidence at this  smelter:  the cadmium oxide product collection baghouse,
combined process fugitives  (70 percent  of which  is  due to blast furnace
baghouse dust handling), and the blast  furnace  baghouse stacks.
      Cadmium oxide,  produced  from roasting cadmium-bearing blast  furnace
baghouse dust,  is  blown to  a product collection  baghouse.  Emissions from
the baghouse are vented to  a large louvered cupola.  Although the  outlet
particulate matter concentration is low, 0.0125  gr/dscf, 66 percent of the
particulate matter is cadmium.  Furthermore, the emissions are very
poorly dispersed due  to the low stack velocity  (0.1 m/s) and stack height
(14 m).  The poor dispersion, plus the high emission rate of 2,720 kg/yr,
combine to boost the  MIR for this source.
      Another high risk source at the El Paso smelter is process fugitive
emissions.  Seventy percent of the process fugitive emissions rate is due
to the blast furnace  baghouse dust handling.  The TACB has estimated the
lead emissions  from this source for the SIP.  The cadmium emissions were
estimated to be 1,610 kg/yr using a cadmium-to-lead ratio supplied by
ASARCO.  The MIR from fugitive emissions is 3.54x10-4, and the annual
incidence is 0.014 case/yr.  However, if this plant is reopened then the
lead SIP would  require that these process fugitives be controlled.   The
blast furnace baghouse at El Paso has six separate cellars where dust
catches are stored.   To remove this dust, which  is up to 18 percent
cadmium, plant  personnel open the cellar doors and transfer the dust by
bucket loader to a railcar for transport to the cadmium plant.   Water is
sprayed on the  bucket loader's path and on the dust in the railcar during
dumping to minimize fugitive emissions.
      The third significant source at the El Paso smelter is the blast
furnace.  These emissions are controlled by a large baghouse and are vented
through six stacks.  Test reports have shown that the outlet grain  loading
is 0.0046 gr/dscf of pamculate matter and 0.00039 gr/dscf of cadmium.16
      At ASARCO-East Helena and AMAX-Boss the process fugitive  emissions
controlled the majority of the risk although both of these sources  had
MIR's less than  IxlO'4.   At ASARCO-East Helena the process fugitives were
primarily from the  blast furnace  and  dross  furnace.   Controls  for  this
                                     21

-------
10
t-0
                               TABLE 5.   SUMMARY  OF HIGH RISK SOURCES AT  LEAD SMELTERS
Plant/Location
ASARCO. £1 Paso. Tex.
ASARCO, last Helena. Mont.
AMAX, Boss. Mo.
Source
CdO bag house
Process fugitives
Blast furnace stacks
Process fugitives
Process fugitives
Cadmium
emissions.
kg/yr
2.722
1.613
1.814
146
498
People
493.000
493.000
493.000
48.600
52.800
Maximum risk
1.57x10"'
3.54x10""
9.71x10
1.11x10-"*
8.26xlO"Sfl
Annual
Incidence.
case/yr
0.025
0.014
0.012
0.0001
0.0004
*Rev1ew of USGS maps Indicates It 1s unlikely that anyone resides at the location of maximum risk     5
 predicted by the HEM.  More plausible MIR estimates are 4.06xlO~  for ASARCO-East Helena and 3.90x10
 for AMAX-Boss.

-------
sourca have been  installed  as  a result of the requirements  of  the  Montana
lead SIP.  These  controls include:  blast furnace air controls,  blast
furnace  secondary hooding vented to a baghouse, covers  for  the dross
kettles, and  improved ventilation for the dross kettles.  At AMAX-Boss  the
primary  emission  source was baghouse dust handling.  The controls  present
for this source include a covered conveyor for transfer of  recycled dusts,
an enclosed dust  storage building, and pug mill wetting of  baghouse dusts
prior to storage.
      All high risk sources in Table 5 were investigated for possible
improvements  in control.  For  three of the five sources best available
control  is considered to be in place.  The blast furnace emissions at
ASARCO-E1 Paso are controlled  by a baghouse, and the outlet grain  loading
have been tested  to be 0.0046  gr/dscf total particulate.13  Further
emission reductions are not likely.
     For ASARCO-East Helena, a number of control devices have been required
by the lead SIP to control process fugitive emissions from the blast
furnace and dross furnace (see Section III.B).  These controls are
considered to be best available technology.  Process fugitive control  at
AMAX-Boss also is considered to be best available technology.  To control
process fugitives from dust handling, AMAX has enclosed storage buildings
and installed a covered conveyor for dust transfer.
      For the two remaining high-risk sources, the cadmium oxide baghouse
and process fugitives from dust handling at ASARCO-E1 Paso,  improved
controls are technically feasible.   Table 6 shows the possible control
improvements for these high-risk sources at ASARCO-E1 Paso along with the
associated costs and cost effectiveness.
      For the cadmium oxide baghouse at ASARCO-E1  Paso,  the emissions have
been estimated by mass balance to be approximately 2.7 Mg/yr.  This
estimate corresponds to an outlet grain loading of 0.012 gr/dscf.b
Baghouse experts consulted believe that 0.005 gr/dscf may be achievable,
     DThis grain loading is based on the emissions estimate of 2,722 kg/yr of
cadmium and the gas flow rate.  Both were provided on ASARCO's Section 114
information request response.   The stack has a 5.5 meter equivalent
diameter with an exit velocity of approximately 0.1 m/s.  The configuration
and placement of this stack preclude-emission testing for this source.  The
emission estimate is based on  a TACB mass balance or the cadmium plant.
                                     23

-------
M
                          TABLE 6.   COSTS  OF  IMPROVED  CONTROL FOR HIGH-RISK  SOURCES  AT  LEAD  SMELTERS
PI«it/Loc*tlon
ASAftCO. El Peso. I««.






'
AS AMD.
(*tt Helen*. Mont.
MAX. Boss. Mo.
Source
CdO b*ghouse
Process fugitives
(dust hendllng)



BUst fumtct bug-
house itftcks
Process fugltlxs

Process fugitives
Source
category
Pb
Pb




Pb

Pb

Pb
Ealsston Na*l«a
reduction, risk «fter
(•proved control Mg/yr control
Switch to better b*g 1.U3 t.2falO~4
Enclose transfer points 1.210 i.SS.10"5




Best Av*ll4ble lechnology
(BAI) In pUce
BAI In pi ice

BAT In pUce
Annuel Cost
Incidnce Costs. ( **f«ct JKCfili
(reduction) Cepltel Amutl VMg V\\tt
0.01 (O.OU) 31.900 e.700 4.100 44e.OOO
0.001 (0.013) Hone* Hone* Mindeted by feus
for lead SIP.
Will put In
piece befoi e
pUnt reopens.





       Sl*nd«ted by TAlB lor leed SIP.  !e«>rov«
-------
 and,  based  on  the  limited  data  available,  one  possible  means  for  reducing
 the  emissions  is changing  the type  of  bag  material  used.c   One  bag
 manufacturer would  guarantee to  meet 0.005 gr/dscf  or less.17  The  capital
 cost  in Table  6 reflects the cost to purchase  and install  135 felt-type  or
 Goretex-type bags.   The annual ized  cost  reflects the incremental  cost
 difference  between  use of  the current  bags and use  of the  new,  more
 expensive bags.  The cost/benefit ratio  for this source is  far  lower  than
 the  cost of other  options,  such  as  installing  a new control system.
 Goretex-type bags.   The annualized  cost  reflects the incremental  cost
 difference  between  use of  the current  bags and use  of the  new,  more
 expensive bags.  The cost/benefit ratio  for this source is  far  lower  than
 the  cost of other  options,  such  as  installing  a new control system.
 However, this  cost/benefit  ratio is an uncertain figure because of  the
 uncertainty of the  emission estimate and because the diagnosis  of the
 problem is  based on  limited information.   Further onsite investigation is
 needed before  the true emission  reduction  can  be assessed.
      The second high-risk  source at El  Paso,  process fugitives from blast
 furnace baghouse dust handling,  can also be  controlled.  The  State  of Texas
 has mandated control  for this source in  the  SIP for lead emission
 reductions.  The future handling of any dust containing greater than 1
 percent lead must be done  in an  enclosed space ventilated to  a  fabric
 filter.  Should ASARCO decide to reopen the  lead smelter, they  plan to
 construct an enclosure for blast furnace dust  handling.   Thus,  although
 control is  not now in place, it will be  installed before the  lead smelter
 reopens.  Based on a 75 percent control efficiency,  the MIR after control
 will fall  from 3.54xlCT4 to 8.85xlO'5.
          baghouse seems to be well operated and maintained with an air-to-cloth
      ratio of 1.04:1.  Since the A/C ratio is already low, reducing ;his
ratio would not significantly reduce emissions.  Therefore, additional
filtering area or a new baghouse are not plausible solutions.   Further, the
mean particle diameter for cadmium oxide is approximately 1 micrometer.
With a small particle size such as this, particle bleed-through is a
potential problem particularly with the woven bag currently in use.  A
plausible solution is changing to a falt-type or Goretex -type coated bag.
                                     25

-------
      B.  Primary Copper Smelters
      A summary of the highest converter and reverberatory furnace risks
are presented in Table 7.  The sources in this table represent the sources
of highest risk within the category, but MIR's for all  sources are less
than IxlO'4.  In addition, these data show that cancer incidence is
extremely low for the copper source category.
      The other emission source that was investigated further was the
converter stack at Phelps Dodge in Douglas, Arizona.  This source had the
highest cadmium emission rate in either source category, 12 Mg/yr.
However, because of the extensive dispersion and the low population
density, the incidence and MIR are relatively low, 0.0011 case/yr and
6.5xlO'4, respectively.  Converter process emissions result from the off-
gases of the molten matte undergoing conversion; the off-gas stream is
drawn through an ESP.  Most copper smelters vent these emissions to an
acid plant for further treatment.  The Douglas smelter vents it to the
atmosphere because they do not have an acid plant.  This not only results
in high cadmium and particulate matter emissions, but also a high S02
emission rate.  The Douglas plant is scheduled for closure in January 1987
by agreement between Phelps Dodge and EPA because of the high S02 levels in
the area.
      Improved control is technically feasible for copper smelters;
fugitive arsenic emissions from converters and reverberatory furnaces have
been identified and evaluated, and control costs are presented in the •
Arsenic BID.4  Table 8 summarizes the costs of the improved control.
     Improved control for converters, consisting of air curtain secondary
hoods vented to a baghouse, was evaluated to have a 90 percent capture
efficiency.4  Likewise, conventional hoods vented to a baghouse were rated
at 90 percent capture efficiency for controlling the reverberatory
fugitives.4  Thus, use of these controls could reduce cadmium emissions and
risk approximately ten-fold.  However, if these controls were implemented,
the costs (as shown in Table 8) would be high.  The copper converters and
reverberatory furnaces for which improved control was assessed were chosen
because they had the highest risks and incidences in the copper category
and because it was possible to use the technical data and costs in the
                                      26

-------
TABLE 7.  SUMMARY OF HIGH RISK SOURCES AT COPPER SMELTERS
Plant/Location
Magma, San Manuel , AZ
Phelps Dodge, Douglas, AZ
ASARCO, Hay den, AZ
Kennecott, Garfield, UT
ASARCO, tl Paso, TX
Source
Converter fugitives
Converter fugitives
Matte/slag tapping
Converter stack
Converter fugitives
Converter fugitives
Matte/slag tapping
Cadmium
Emissions
Icg/yr
240
160
3,500
12.000
83
137
53
People
211,000
31,000
31,000
3 1 ,000
46,800
810,000
493,000
Maximum Risk
3.90x10-5
2.31x10-5
3.07x10-5
6.59xlO~6
1.40x10-5
1. 31xlO-7
1.04x10-5
Annual
Incidence
case/yr
0.00013
0.00018
0.00016
0. 00111
0.00004
0.00022
0.00045

-------
TABLE 8.   COSTS OF IMPROVED CONTROL FOR HIGH-RISK  SOURCES AT COPPER SMELTERS
Plant/Location
Hagma Copper,
San Manuel. AZ

Kennecott,
airfield, UT

("helps Dodge,
oo Douglas. AZ




Source
Source Ca tegory
Converter Cu
fugitives

Converter Cu
fugitives

Converter Cu
fugitives

Reverb furnace Cu

Converter stack Cu
[Mission
Reduction
Improved Control Hg/yr
Air curtain secondary 0.216
hoods vented to
baghouse
Air curtain secondary 0.123
hoods vented to
baghouse
Air curtain secondary 0.144
hoods vented to
baghouse
Hooding for Mtte and 0.091
slag tapping
99 percent efficient 9.68
	 HailM
Risk After
Control
3.9xlO-6


1.3U10-8


2.3U10-6


3.07x10-6

1.26xlO-6
Annual ~ ~ ~~
Incidence Costs
(reduction) Capital
0.00001 (0.00012) 1.3x10?


0.00002 (0.0002) 5.2x10*


0.00002 (0.00016) 9.8X106


0.00002 (0.00014) 1.8x106

0.00021 (0.0009) 7-SxlO6
Annual
4.0x106


I.3xl06


2.8x106


5.1x105

6.0x10*
Cost Effectiveness
t/Hg
1.35x10'


1.1 xlO7


1.97xl07


5.6x106

6.23x10*
J/L 1 f e
3. 45x1010


6.6xl09


1.71xlO>°


3.5x10^

6.7x108
              ESP (baseline 1s
              94.8 percent)

-------
Arsenic BID to evaluate the cost effectiveness for reducing cadmium
emissions.  Also, by choosing the sources of highest risk for the
cost-effectiveness evaluation,  the most optimistic estimates for cost
effectiveness were obtained because risk and incidence reductions were
maximized.•
      The one process point source for which improved control costs were
calculated was the converter stack at Phelps Dodge-Douglas.  The current
ESP control  is 94.8 percent efficient as calculated from the emission
estimate.  Improved control evaluated for this source is a 99-percent
efficient ESP.  The major benefit derived from additional control is a
reduction in the incidence from 0.001 case/yr to 0.0002 case/yr.  The MIR
would also be reduced.
                                     29

-------
  V.   REFERENCES
 1.  Telecon.  B. Nicholson, MRI, with J.  Richardson, Senior Environmental
     Scientist,  ASARCO, Inc.  Salt Lake City, Utah, August 18, 1986.

 2.  Telecon:  B. Nicholson, MRI, with M.  Kearney,  Director of Safety,
     Health, and Environmental  Services,  Homestake Lead Tollers.   Boss,
     Missouri.   September 15, 1986.

 3.  Minerals Facts and Problems: Chapter on Lead.   Bureau of Mines, 1985
     edition,  p. 18.

 4.  Inorganic  Arsenic Emissions From Low Arsenic Primary Copper
     Smelters — Background Information for Proposed Standards.  U. S.
     Environmental  Protection Agency Publication No. EPA-450/3-83-010a.
     April 1983.

 5.  Minerals Facts and Problems: Chapter on Copper.  Bureau of Mines,
     1985 edition,   p. 23.

 6.  Background  Information For New Source Performance Standards:  Primary
     Copper, Zinc,  and Lead Smelters.  Volume I:  Proposed Standards.
     U. S. Environmental  Protection Agency Publication
     No. EPA-450/2-74-002a.  October 1974.

 7.  Proposed Revisions to the State Implementation Plan for the Control
     of Lead Air Pollution - El Paso County.  October 11, 1983.

 8.  Telecon.  B. Nicholson, MRI, with J.  Richardson, ASARCO, Inc.
     August 7,  1986.

 9.  State of Montana Air Quality Implementation Plan:  Lead.
     September  16,  1983.   pp. 5-72 - 5-92.

10.  Letter from J. Richardson, ASARCO, Inc.  Concerning cadmium NESHAP
     supplementary  information.  August 6, 1986.

11.  Evaluation  of  an Air Curtain Hooding System for a Primary Copper
     Converter  - ASARCO,  Inc.,  Tacoma,  Washington.   PEDCo Environmental,
     Inc.  December 1983.

12.  Uodate of  Cadmium Emissions Estimates ror Monferrous Smeltars.   Memo
     from Elizabeth Anderson, GCA,  to John Copeland, EPA-ESED-ISB, during
     cadmium source survey.  May 26,  1981.

13.  Notice of  Intent to  List Cadmium Under Section 112 of the Clean Air
     Act published  in Federal Register, 50 FR 42000.  October 16, 1985.
                                     30

-------
14.   User's Manual  for the Human Exposure Model (HEM).  Pollutant
     Assessment Branch, Office of Air Quality Planning and Standards,
     U.  S. Environmental  Protection Agency.  October 1985.

15.   Sample Fugitive Lead Emissions From Two Primary Lead Smelters.  U. S.
     Environmental  Protection Agency Publication No. EPA-450/3-77-031.
     October 1977.

16.   Telecon.   B.  Nicholson,  MRI, with 0. Richardson, ASARCO, Inc.
     August 11, 1986.

17.   Telecon.   B.  Nicholson,  MRI, with T. Fisher,  W. L. Gore and
     Associates.  August  11,  1986.
                                     31

-------
                                     TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse before completing;
 1  "EPORT NO
        EPA-450/3-88-006
                                                           3 RECIPIENT'S ACCESSION MO
TITLE AND SUBTITLE
 Cadmium Emissions  From Primary  Lead and Primary
 Copper Smelting  -  Phase I Technical Report
                                                              5. REPORT DATE

                                                                      LQi^rt
                                                              6. PERFORMING ORGANiZAT'CN CODE
 7 AUTHOR(S)
                                                              8. PERFORMING ORGANIZATION REPORT NJO
 9 PERFORMING ORGANIZATION NAME AND ADDRESS
    Office  of Air Quality  Planning and Standards
    U.S.  Environmental Protection Agency
    Research Triangle Park,  N.C.  27711
                                                           10. PROGRAM ELEMENT NO
                                                           1 1. CONTRACT. GRANT r^J

                                                              68-02-3817
  2 SPONSORING AGENCY NAME AND ADDRESS
     DAA  for  Air Quality Planning and Standards
     Office of Air and Radiation
     U.S.  Environmental Protection Agency
     Research Triangle Park,  North Carolina   27711
                                                           13 TYPE OF REPORT AND PERIOD COVERED
                                                           14. SPONSORING AGENCY CODE
                                                                  EPA/200/04
 15. SUPPLEMENTARY NOTES
           A  technical report  on cadmium emissions  from primary  lead and primary
     copper smelting.  Descriptions of these  industries and .associated air pollution
     control  equipment are presented.   Cadmium  emission estimates  for all U.S.  plants
     in these  two  source categories and health  risks  from exposure to these emissions
     from each plant are discussed.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
     Air Pollution
     Pollution Control
     Cadmium Emissions
     Primary Lead Refining
     Primary Copper Refining
18 DISTRI8UTION STATEMENT


         Unlimited
                                               b.IDENTIFIERS/OPEN ENDED TERMS
                                              Air Pollution Control
19 SEC'JRtTY CLASS (This Report/
  Unclassified
                                            20 SECURITY CLASS iThis pagei

                                           j	Unclassified
                                                                           c. COSATI Held/Group
                                  13B
                                                                        21  NO. OF PAGES
                                                                              36
                                                                           22. PRICE
= PA Form 2220-1 (Sev. 4-77)   PREVIOUS eoi TION \ s OBSOLE TE
                                                                                           _J

-------