Environmental  Protection  Technology Series


                R. Nekervis, J. Pilcher,
           J. Varga Jr. , B. Gonser, and J. Hallowell

              Battelle, Columbus Laboratories
                  505 King Avenue
                 Columbus, Ohio 43201
                Contract No. 68-02-1323
                     Task 9
                 ROAP No. 21ADK-017
               Program Element No. 1AB012
             EPA Project Officer: G. J. Foley

               Control Systems Laboratory
           National Environmental Research Center
          Research Triangle Park, North Carolina 27711
                   Prepared for

               WASHINGTON, D.C. 20460

                   October 1974

This report has been reviewed by the Environmental Protection Agency
and approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.

          The report summarizes the state of process modifications relative
to the control of fine particulate emissions from stationary combustion
sources (electric utilities and industrial processes); municipal incinera-
tors; iron and steel plants; ferro-alloy plants; and nonferrous metal
smelters (zinc plants, copper smelters, aluminum reduction cells).  This
study is to uncover modifications to conventional practices or new uncon-
ventional practices which appear to improve the control of fine particulate
emissions in these five areas.    Modifications to conventional stationary
combustion sources considered include ash fluxing, SO  addition to flue gas,
staged combustion, use of fuel  additives, fly-ash agglomeration, solvent
refining, and flue-gas recirculation.  Unconventional systems studied include
fluidized bed, coal gasification, and submerged combustion.  For incinerators,
combined fuel-refuse firing, gas cooling, and pyrolysis methods are considered,
Emphasis for iron and steel plants is given to the bottom-blowing oxygen
process (Q-BOP).   Modification  of the conventional reverberatory smelting
procedure and the introduction  of hydrometallurgical methods are discussed
for copper, and the chloride electrolytic (ASP) process by ALCOA is considered
for aluminum.  Each process is  considered with respect to its stage of
development, availability  or acceptability by industry, efficiency in reduc-
ing emissions, and environmental impact.


                           TABLE OF  CONTENTS


 1.   Introduction	   1

 2.   Summary	   2

     2.1  Stationary Combustion  	   4

          2.1.1  Electric Utilities   	   4

          2.1.2  Industrial Processing and Steam Generation  ....   6

     2.2  Municipal Incinerators 	   6

     2.3  Iron and Steel Plants	   7

     2.4  Ferroally Furnaces  	   8

     2.5  The Primary Nonferrous Metals Industry 	 .....   9

          2.5.1  Zinc Roasting, Sintering, and Distillation  ....   9

          2.5.2  Copper Roasting, Matte Smelting and Converting  . .   9

          2.5.3  Aluminum Reduction Cells  	  10

3.  Stationary Combustion,  Particulate Control Combustion
      Modification 	  11

     3.0  Introduction	  11

     3.1  Electric Utilities 	  11

          3.1.0  Background; the Problem of Fine Particles	  11

       New Combustion System Modifications  ....  13

       Ash Fluxing	  13

       Fluidized-Bed Combustion 	  14

             Status  	  18

       Staged Combustion  	  18

       Recirculation of Flue Gas	   19

       Gasification 	   19

                          TABLE  OF CONTENTS


       Modification of Fly  Ash Resistivity  ....   21

       Agglomeration of  Ash Particles   	   22

       Advanced Power Cycles 	   22

       Modification of Particle Size  	   23

       Solvent  Refined Coal  	   23

       Low Excess Air	   23

       Electrochemical Oxidation of Coal  	   24

       Fuel Additives  	   24

       Improved Control  Methods for Fine
                          Particulates	   25

       Electrical Control of Particulates
                          From Flames   	   25

     3.2  Industrial Processing  and Steam Generation  	   27

          3.2.0  Introduction   	   27

          3.2.1  Process Modifications  	   27

       Burne'r Design   	   27

       Flue-Gas Recirculation  	   28

       Two-Stage Combustion  	   28

       Other Combustion  Modifications   	   29

       Improved Service  and Maintenance  	   30

       Modifications to  Reduce PNA 	   30

4.  Municipal Incinerators  	   32

     4.0  Introduction	   32

          4.1.0   Process Modifications  	   32

                          TABLE OF CONTENTS


          4.1.1  Combined Firing	   33

          4.1.2  The CPU-400	   33

          4.1.3  Water-Walled Incinerators 	   34

          4.1.4  Pyrolysis Process 	   35

          4.1.5  Advanced Concepts Involving Heat Recovery ...   37

          4.1.6  Electron-beam Irradiation 	   38

     4.2  Predictions for the Year 2000	   38

5.  Iron and Steel Plants	   39

     S.I  Open Hearth Furnace	   39

          5.1.1  Hydrocarbon Additive to Lancing Operation ...   ^0

       State of Development -
                          Hydrocarbon Addition 	   ^
       Availability to Industry 	  ^

       Degree of Effectiveness  	  ^
       Environmental Effects
       Use of Liquid Oxygen and Liquid
                          Hydrocarbon Injection  ..... • •  •

     5.2  EOF Furnace  ................. ....   45

     5.3  Q-BOP Process  ....................   ^5

          5.3.1  State of Development - Q-BOP Process  .....   45

          5.3.2  Degree of Effectiveness ............   ^'

          5.3.3  Environmental Effects .............   ^'

     5.4  Electric Arc Furnace .................   48

          5.4.1  Preheating and Melting with
                   Oxygen-Fuel Burners .............   ^9

                           TABLE OF CONTENTS
       State of Development 	  49

       Degree of Effectiveness  	  49

       Environmental Effects  	  50

          5.4.2  Electric-Arc Furnace--Scrap Charge
                   Compatibility 	  50

       State of Development 	  50

       Availability to Industry 	  50

       Acceptance by Industry 	  51

       Degree of Effectiveness  	  51

     5.5  Metallurgical Coke Ovens   	  51

          5.5.1  Particulates from Charging Coke Ovens	  51

          5.5.2  Particulates from Pushing Coke	  52

          5.5.3  Particulate Emissions During Quenching  ....  52

6.  Ferroalloy Furnaces  	  54

7.  Process Modifications For Particulate Control
      In the Primary Nonferrous Metallurgical Industry 	  55

     7.1  Zinc Roasting, Sintering, and Distillation 	  55

          7.1.1  Roasting	  55

          7.1.2  Sintering	  56

          7.1.3  Reduction and Distillation  	  56

     7.2  Copper Roasting, Matte Smelting, and Converting  ...  56

          7.2.1  Pyrometallurgical Modifications 	  57

       Flash Smelting 	  57

            Electric Furnaces 	  59

                            TABLE  OF  CONTENTS


         Continuous Smelting  	   60

         Miscellaneous  Pyrometallurgical
                          Developments   	   63

           7.2.2 Hydrometallurgical  Copper Recovery  	   68

         Stanford  LCPR  Process  	   68

         Anaconda's "Arbiter" Ammonia              70
                          Leach Process   	  .....   70

         Sherritt-Gordon  Process  	   72

         Cymet  Process    	   73

         Duval  Corporation and Other Processes  .   73

     7.3  Aluminum Reduction Cells   	   75

           7.3.1  New Aluminum  Reduction  Processes  	   79

8.  References	   82

                               APPENDIX A



                           LIST OF TABLES


Table 1.  Emissions from Primary Aluminum Industry--1970 ....  80

Table 2.  Emissions from an Uncontrolled Prebake                  80
            Potline	80

Table 3.  Fluoride Removal Efficiencies of Selected
            Primary and Secondary Controlled Systems 	  80

Table A-l Annual Copper Production of Smelters and
            Refineries in the U.S	A-2

Table A-2 Control Practice in Copper Smelting  	 A-10

Table A-3 Composition of Typical Emissions From
            Reverberatory Stacks   	 A-13

Table A-4 Compositions of Atmospheric Emissions From
            Converter Operations in Primary Copper Industry   .  . A-15
                            LIST OF FIGURES

Figure 1. Simplified Fluidized-Bed Combustion Boiler Concept  .  .  15

Figure 2. Atmospheric Fluidized-Bed Combustion Power Plant ...  16

Figure 3. Pressurized Fluidized-Bed Combustion Power Plant ...  17

Figure 4. Statistical Distributions of Dust Loadings During the
            Operation of a 225 net ton Openhearth Furnace
            Using Oxygen Lancing and Oxygen + Propane Lancing   .  42

Figure 5. Relationship Between Dust Loadings of a 225 net ton
            Openhearth Furnace Operating with Oxygen Lancing
            and Oxygen -f Propane Lancing ............  43
Figure 6. OEM Process Vessel, U.S. Patent 3,706,549  ......  46

Figure 7. Outokumpu Furnace  ..................  58

Figure 8. INCO Flash Furnace ..................  58

Figure 9. Noranda Process  ...................  61

                            LIST OF FIGURES
Figure 10.  Schematic view of Mitsubishi's
              Semicommercial Plant 	  64

Figure 11.  Bureau of Mines Autogenous Smelting  	  66

Figure 12.  Flowsheet for the lime-concentrate-pellet-roast
              Process with Direct Electrowinning 	  69

Figure 13.  Anaconda Arbiter Plant—Block Flow Diagram 	  71

Figure 14.  Schematic Flowsheet--Cymet Process   	  74

Figure 15.  Schematic Drawing of a Prebaked Anode Cell 	  76

Figure 16.  Schematic Drawing of a Horizontal Stud Soderberg
              Aluminum Reduction Cell	  77

Figure 17.  Schematic Drawing of a Vertical Stud Soderberg
              Aluminum Reduction Cell	  78

Figure A-l  Generalized Diagram of Conventional Smelting
              Flowsheet	A-3

Figure A-2  Cutaway View Showing Key Features of a
              Conventional Reverberatory Furnace 	  A-6

1.  Introduction
          The purpose of this study is to summarize in one report the
state of technology relating to the control of fine particulate emissions
resulting from process modifications in the following areas (1) stationary
combustion as it relates to electric utilities and industrial processing,
(2) municipal incinerators, (3) iron and steel plants, (4) ferroalloy
plants, and (5) nonferrous metal smelters (zinc plants, copper smelters,
and aluminum reduction cells).  The report is intended to serve as a brief-
ing document for persons who are not well versed in these areas.
          Fine particulate emissions may be defined as those particles
smaller than 2 microns diameter.  The sizes that will deposit in the lungs
(called "respirable dust") vary from submicron (O.Olp, and smaller) to about
8 microns.  A high percentage of the larger particles (2 to 8u) that enter
the bronchial tubes is deposited by the mechanism of inertial impaction,
and a large fraction of the smaller particles (less than 0.1 u) is deposited
by means of diffusion.  However, for the in-between sizes 0.1 to 2.0 y,),
only a small fraction is retained in the respiratory tract because neither
mechanism of deposition is efficient in this size range.  Hence, the very
fine particles (less than 0.1 LL) which are not readily removed by collectors
designed in the conventional fashion are of growing concern because of their
possible adverse effects on health and reduction of atmospheric visibility.
Also, our ability to measure and characterize very fine particulates is
          The literature reflects this state of affairs.  In our study on
the effect of process modifications in the five areas mentioned in the
first paragraph above, we found that quantitive data on changes in par-
ticulate emissions as a result of either a modification or even a replace-
ment of a conventional process was not only not reported, but appeared to
be largely ignored.  Accordingly only qualitative data are provided in
substantiation of claims of beneficial effects on the amount, size and
properties of fine particulate emissions.  In almost all cases the degree
of reduction of fine particulate emissions could not be ascertained with
any degree of precision.  This necessiated some changes in the format of
the report as originally conceived.

           In identifying process modifications that may have beneficial
 effect on the control of fine participates,  the approach was to determine
 the mechanism responsible—agglomeration or  growth of participates,  altera-
 tions of their properties,  or reduction in mass.   Insofar as possible,
 information on the stage of development, availability and acceptance by
 industry, and environmental side effects,  if any,  of both modifications
 and the more unconventional system replacements have been covered.
           There was an unusual number of new modifications in copper smelt-
 ing technology which necessitated including  background information on the
 current status of emission  control practices in conventional U.S. primary
 copper smelters.   This is provided in Appendix A.

 2.   Summary

           This report summarizes the  state of technology relating to bene-
 ficial effects on fine particulate emissions that  come about through new
 process modifications in five  major areas:
           (1)   stationary combustion
                (a)  electric  utilities
                (b)  industrial  processing
           (2)   municipal incinerators
           (3)   iron and  steel  plants
           (4)   ferroalloy plants
           (5)   nonferrous metal  smelters
                (a)  zinc  plants
                (b)  copper smelters
                (c)  aluminum  reduction  plants.
          Summarizing  results  that  relate  to  all five  areas, the 1-2 percent
of  fine  particles that escapes present day high efficiency collectors are
practically all smaller  than 1 micron.  Quantitative information on  the
size distribution,  chemical properties, and  the effects of these fine
particulates is lacking, probably because of  the difficulties in their
          While quantitative data on fine particulate emissions from con-
vaitional stationary combustion equipment, municipal incinerators,  iron

and steel plants, ferroally plants,  and nonferrous metal smelters are
very sparse, similar quantitive data relating to process modifications in
these areas are nonexistent. Even the qualitative data on the degree of
fine particulate emission from these current modifications in practice are
not definitive.
          Bag filters are the most effective method for removing particles
smaller than 2 microns.  Some types  of scrubbers and many electrostatic
precipitators can also be effective  for the collection of particles under
2 microns.  The technology for particulate emission control has changed
little over the past 20-40 years.  Characterization of very small particles
has been neglected because they are  difficult to sample, classify, and
analyze.  This difficulty is reflected in the dearth of quantitative data
on the effectiveness of process modification in reducing fine particulate
          Fine particulates are formed largely as a result of metals in the
coal ash being vaporized by the intense heat of combustion.  Upon cooling,
these metallic vapors condense to form a multitude of very fine spherical
particles.  A few small particles are formed also from polynuclear aromatic
compounds present in the coal which show up as particulate POM (Polycyclic
Organic Matter).  Process modifications to circumvent or minimize particle
formation by these mechanisms would be helpful.
          Our study indicates that generally submicron particle abatement
can be achieved more economically through either process modification  or
outright substitution rather than through the installation of new, larger
and more efficient precipitators.  To achieve an increment of improvement
from 99.0 to 99.5 percent efficiency in an electrostatic precipitator
requires a great increase in the size of the collecting apparatus with a
corresponding great increase in cost.  Wet scrubbers are being designed as
a means of capturing particulate matter as well as S02 but these are
still in the early stages of development.  Venturi scrubbers work well
down below 1.0  .

2.1  Stationary Combustion

2.1.1  Electric Utilities

          There are a number of modifications in conventional combustion
practices, as well as some totally new unconventional systems, being
considered by the electrical utility industry which appear to exert a
beneficial effect on the control of fine particulate emissions.  Some
of these may have questionable side effects.  None in the following
listing have as yet been studied in a quantitive way to determine the
amount, size distribution, and characteristics of fine particulate
emission.  They are as follows:
          (1) Ash fluxing to capture the ash in a molten slag,
              thereby decreasing the dust loading in the flue
          (2) Adding SO  in controlled amounts to achieve an
              optimum resistivity of the fly ash particulates
              so as to improve their collection in electro-
              static precipitation equipment; also altering
              the burning rate, amount of excess air,  coal
              particle size and rank to achieve the same
          (3) Utilizing staged combustion (that is burning
              fuel in a primary zone with less than stoichio-
              tnetric amounts of ai'r, partially cooling the
              combustion products,  then completing combustion
              in a second zone with additional air) to reduce
              the emission of NO  and particulates by  reducing
              temperature.   The degree of beneficial effect  on
              the control of particulates depends on the fusion
              characteristics of the ash.

(4)  Utilizing fuel additives to reduce total participate
     emissions (the possibility of toxicity of the newly
     created emissions makes their use questionable for
     the present).
(5)  Agglomerating the fly ash in the combustion zone by
     extending residence time to produce tackiness and
     inducing high turbulence or pulsating flows to
     promote agglomeration (this approach is still very
     much in the concept stage).
(6)  Solvent refining (reconstituing coal) to achieve low
     ash and sulfur contents and thus resolve the particulate
     problem (the cost of this is excessive).
(7)  Recirculating flue gas, i.e., diluting the fuel air
     mixture in the combustion zone with cooled flue gases,
     has been found to contribute to lower emissions of
     smoke when firing fuel oil.  Application of this
     technique is still in the early stages of investigation.
Unconventional systems being considered include
(1)  Fluidized-bed combustion systems which use relatively
     large particles of coal in comparison to the pulverized
     coal used in conventional combustion systems; while
     more particulates are generated, the larger size leads
     to greater collection efficiencies with existing
     collectors.  It appears that high-pressure fluidized
     combustion systems offer even greater collection
(2)  Gasification of coal can be considered as modified
     staged combustion.   While fluidized-bed combustors
     have not as yet been applied to boilers, there are
     a number under development for gasification of coal
     to produce pipeline gas or liquid fuels.
(3)  Carrying the gasification of coal a step further,
     there are the submerged combusion processes.   One
     aspect of this concept is to burn pulverized coal
     in a bath of molten salts or. coal-ash slag

               to capture sulfur and fly ash in the molten salt or
               slag.  In a related process, crushed coal is injected
               into molten iron which dissolves the fixed carbon and
               sulfur constituents, the volatile constituents of the
               coal are driven off and cracked to H? and CO.  Air
               introduced into the molten iron produces a "Bessemer"
               reaction which oxidizes the carbon to CO,  Ash of the
               coal becomes a supernatant slag which is fluxed with
               enough lime to extract sulfur from the iron and hold
               it in the slag which is drawn off continuously.
               (Development work on these processes is in progress,
               but no details have been publicly released.)

2.1.2  Industrial processing and Steam Generation
          The above conclusions on modifications of stationary combustion
units such as flue-gas recirculation and two-stage combustion apply to the
industrial processing area also.
          Poly-Nuclear-Aromatic (PNA) compound emissions into the atmosphere
are generated principally by commercial and residential coal-fired units.
Oxidation during the combustion process will destroy PNA and other hydro-
carbons.   There is a dearth of data identifying the conditions required
for the oxidation of PNA,  and modifications of the combustion processes
to reduce PNA emissions.

2.2  Municipal Incinerators

          (1)  Tests on combined firing of municipal refuse (20
               percent) and fossil fuels (80 percent)  in steam
               generating  boilers  indicate that the particulate
               emissions can be controlled to the  degree obtain-
               able when firing fuel alone if modifications in
               the electrostatic collection system are made to
               accommodate the higher stack volume.

           (2)  Water-walled  incinerator  furnaces permit  the cooling of
               gases  to a  temperature which permits  the  operation of
               flue-gas cleaning equipment.  While this  substantially
               reduces the particulate emissions over  that of conven-
               tional incinerators,  there  is a corrosion problem
               associated with deleterious salts and gases that are
               present in all municipal  incinerators.
           (3)  Pyrolysis processes have  the potential  of reducing
               particulate emissions; quantitative data  on this
               aspect of these processes have not been published.
2.3  Iron and Steel Plants

          A major change in the oxygen-blowing process in steel making
from blowing at the surface of a melt in a vessel  (the EOF or BOP process)
to bottom-blowing oxygen with a hydrocarbon through the melt (U.S. Steel
Corporation's Q-BOP and related processes) results in a quieter bath with
reduced fumes and larger particulate sizes in the emissions; dusting is
reported to be 1/3 to 1/5 that of the EOF process; in addition the Q-BOP
and related processes provides slightly higher rates of production and
slightly better ultization of scrap than does the EOF process,
          Installed or planned capacity of Q-BOP vessels in the U.S. was
about 9,000,000 tons annually as of March, 1973.  U.S. 1970 production
of steel was 150,000,000 net tons, of which the basic oxygen furnace (BOF)
accounted for 55 percent (Q-BOP production is included in these BOF
statistics), the open hearth accounted for 26 percent, and the electric
furnace, 18 percent.  There is potential for explosive growth in the
Q-BOP and related processes.  Of the 21 open hearth shops in operation,
only ten are modern plants with extensive air pollution control,  three are
scheduled to close, and eight will be replaced.   The Q-BOP would appear
to be a logical replacement for these, but present economic factors and a
muddied patent situation make this choice unclear.  Since the production

 rates  and  degree of  scrap  utilization of  the Q-BOP vessels are only
 marginally higher  than  that of  the EOF vessels,  there  is no pressure on
 the  EOF  shops  to convert.
           The  use  of hydrocarbons in combination with  oxygen during
 lancing  open-hearth  steel  has been found  to reduce the emission of fumes
 and  the  grain  loadings  of  exhaust gases when tested on a laboratory scale
 ten  years  ago.  Except  for a few plant trials since, the results of which
 were not reported, there has been no further interest  in the process mod-
 ification;  it  has  not been accepted by U.S. industry.  On the other hand,
 the  Soviets are experimenting with the use of liquid oxygen-liquid hydro-
 carbons  in  the lancing  of  open hearth steels.
          The  reduction of  particulate emissions from coke-oven operations
 by agglomeration of the particulates, or  their reduction, by process
 modification,  does not  appear to be imminent, owing to the hard, angular
 characteristics of the  carbon particulates.  New processing equipment, such
 as pipe-line charging and  other modifications of charging methods will
 probably be able to achieve the required reductions in emissions during
 charging.  New equipment,  presently under evaluation under production
 conditions, may be the answer to the reduction or even elimination of
 emissions during the pushing and quenching operations.
2.4  Ferroally Furnaces

          The only modification in ferroally furnace practice which
resulted in improvement in the control of particulate emission uncovered
in this study was a study on the use of pelletized ore concentrates.
These reduced particulate emission by 42 percent in comparison to that
of fine ore concentrates.

2.5  The Primary Nonferrous Metals Industry

2.5.1  Zinc Roasting, Sintering, and Distillation

          Electrolytic zinc plants, present and being constructed, have
negligble loss of particulates other than in general handling of concen-
trates and calcines or sinter.  The two pyrometallurgical zinc plants use
good practice in keeping the escape of particulates to the atmosphere at
a minimum, and no modifications have been suggested for improvement in
the basic steps of their processes.

2.5.2  Copper Roasting, Matte Smelting, and Converting

          Progress in controlling the extent of emissions of fine particles
in the copper smelting industry appears to be substantial.  There are two
general approaches; one is to modify the conventional reverberatory
smelting procedure, the other is to replace the pyrometallurgical methods
with hydrometallurgical methods.  The goal in both cases is first to
reduce or eliminate the emission of sulfur oxide and concommitantly to
reduce stack losses of particulates.
          Modification of reverberatory converter procedures has involved
the introduction of continuous smelting, flash smelting, and flash roasting-
electric furnace combinations, all of which seek a reduction in gas flow and
an enrichment of S0» in the effluent gases to permit recovery of sulfur
as sulfuric acid, liquid SO , or elemental sulfur, which in turn necessi-
tates cleaner off-gases.  An exception to the new smelter modification
to recover SO. is a limestone scrubbing treatment to control sulfur.  It
is only moderately efficient in removing SCL, but in combination with a
preceding dry-gas cleaning system, it is excellent for removing par-
          Hydrometallurgical methods substantially eliminate the emission
of particulates to the atmosphere.  A number of new processes are being
developed; among these are pressurized ammonia leaches and chloride leaches.
One of the new processes, Anaconda's "Arbiter" Ammonia Leach Process, has

reached the commercial stage.  The only particulate control involved is
in handling concentrates and lime and in preventing tailings from becoming
wind blown.

2.5.3  Aluminum Reduction Cells

          Aluminum reduction cells which use prebaked anodes,and account
for 59 percent of total U.S. production of aluminum, produce less voliti-
zation of pitch and less fouling of the emission control system than do
Soderberg reduction cells which use baked-in-place anodes.  However, the
separate anode baking  furnace requies an emission control system.
          Alcoa's aluminum-chloride electrolytic process (the ASP Process)
has reached the pilot plant stage.  Press releases on the process indicate
less dusting than is the case with the conventional Hall aluminum reduction
cells.  No cryolite (Na Al Fe,) will be required.  Fluorides, and the
working of Al  0  into the bath will be eliminated.  Emphasis, however, is
             J  O
on the savings of energy (as much as 30 percent less energy than the con-
ventional Hall process) and on the fact that "scarce and costly cryolite"
will no longer be required.
          There are a number of new processes aimed at exploiting new
sources of alumina, such as alunite, the kaolin clays, and laterite.
These do not affect materially the amount or type of emissions from the
reduction cells.


3.  Stationary Combustion, Participate Control Combustion Modification
3.0  Introduction

          Theburning of coal, more than any other fuel, impacts on the
ambient air because of the emission of particulates:   fly ash, fine
particules of ash; small, but nevertheless important amounts, of organics
as vapor or fine droplets; and sulfuric acid mist, in the amount of 1 or 2
percent of the sulfur equivalent in the coal.  Thirty-two percent of the
industrial particulate emissions in the U.S. are from coal burning.
          In the burning of pulverized coal, the characteristics of the
particulate emissions depend on many factors.  Among these are the heat
liberation rate in the furnace, the composition of the coal (especially
the ash content), the degree of pulverization of the coal, and the amount
of excess air.  In addition, the basic parameters of time, temperature and
turbulence, plus the reactivity of the coal influence the quantity and
characteristics of particulate emissions.
          Polycyclic organic matter (POM) which is emitted in particulate
form from a vast number of stationary combustion sources will be considered
part of this study.  Particulate POM consists of a variety of chemical
entities; however, it is common practice to use benzo (a?) pyrene as an indi-
cator of other POM owing to the demonstrated carcinogenicity of benzo (o)
pyrene and the relatively large amount of published data on it.  ^ Relatively
high levels of POM have been measured in coal-burning equipment.
          This section of the report considers practical combustion process
modifications that are likely to improve the situation with regard to par-
ticulate emission.

3.1  Electric Utilities
3.1.0  Background; the Problem of Fine Particles

          Over 80 percent of the potential fly ash from U.S.  coal-fired
power plants is now being captured by means of large  electrostatic pre-
cipitators,  and most new power plants  now have improved collectors guaranteed

 to  remove  98  to  99  percent  of  the  fly ash  leaving the  furnace.  However, the
 small  amount  (only  1 or  2 percent) which escapes these high-efficiency collect-
 ors, is  practically all  smaller  than 1 micron.  The environmental impact
 of  these extremely  fine  particles has been  largely ignored because quantitative
 information on the  particles and their effects is so difficult to measure.
 However, their effect  in making  stack plumes highly visible because of the
 light-scattering effect  of  0.4 to 0.8 micron particles, though hard to
 quantify,  is  often  readily  apparent to the  community.  Furthermore, such
 plumes add to the haze so common in metropolitan areas.
           There is increasing concern about these fine particles which are
 not adequately controlled with existing equipment.  Often, their physical
 and chemical  properties  are not known.  However, it is recognized that these
 fine particles, by nature of their size, are the principal contributors to
 adverse health effects, visibility reduction, and soiling of surfaces.  A
 review panel  of the National Research Council has identified the importance
 of fine particles and reports that current practices for evaluating control
 techniques as "tonnage-collection figures and weight-removal efficiencies
 are inadequate to delineate the entire particle-emission problem".  They
 further state that.... "small particles....may continue to limit visibility
 and may affect health even when presently uncontrolled sources of particulate
matter are equipped with the best collection devices currently available".
However, to obtain improved efficiencies in electrostatic precipitation
 from 99 to 99.5 percent with today's technology involves a great increase
 in size with a corresponding great increase in cost.   Accordingly, alternative
 strategies for the control of fine particulates are being sought.
           One of the most obvious alternatives would be to keep the ash
 in the  furnace.   A characteristic of conventional pulverized coal-fired
dry-bottom furnaces is that 50  to 70 percent of the ash leaves the furnace
as fine particles suspended  in  the hot  (300 F)  exhaust gas.  Although some
of the  ash particles become  liquid during combustion they are cooled and
 solidified before they reach the  walls  of dry-bottom furnaces.  In cyclone
and conventional  slagging or wet-bottom furnaces some of these particles
are liquid or sticky when they  reach the walls  where  they combine to form


a viscous slag layer which slowly drains down to a slag tap.  The cyclone
furnace retains  about 90 percent of the ash as slag.  However, because of the
intense combustion in the cyclone resulting in high flame temperatures, some
of the ash is vaporized.  When this vapor condenses later in the system,
large numbers of extremely fine spherical particles are formed by the La Mer
effect which defy conventional collecting systems,  New Combustion System Modifications
           Other alternative modifications to existing combustion systems
are being considered.  For example, fluidized combustion systems use
relatively large particles of coal in comparison to the pulverized coal
used in conventional combustion systems and this use of larger coal particles
generates larger particulates in the combustion gas.  Although there are
more particulates generated by this system, the larger size leads to a
greater collection efficiency with existing collectors.  It appears that
operating fluidized-bed combustion systems at high pressures leads to even
greater collection efficiency.  Among other new developments that may
influence fine-particle generation are combustion-system modifications
like staged combustion and flue-gas recirculation that are intended for
NO  control, and submerged combustion in molten salts.  These and others
are discussed individually below:  Ash Fluxing
           As pointed out earlier, conventional slag-tap and cyclone
furnaces convert much more of the ash into molten slag than do the more
common dry-bottom pulverized-coal-fired boiler furnaces.  Decreasing the
viscosity of coal-ash slags will tend to increase the amount of ash con-
verted to slag, thereby decreasing the dust loading in the flue gas.
Fusibility of coal ash varies widely.   However, the ash from some coal will
not form slag in the cyclone.  The addition of inexpensive fluxes such as
limestone (CaO) and mill scale or iron ore (Fe20.) can lower slag viscosity

                                   14   Fluidized-Bed Combustion
           Burning crushed coal in a fluidized bed of limestone has
 demonstrated merit as a means of reducing gaseous pollutants (SO  and NO ).
                                                                X       X
 Figure 1  shows a simplified fluidized-bed combustion boiler concept.
 Research  is under way on atmospheric as well as on elevated pressure
 (10 atmospheres) fluidized-bed combustors.  Development work for OCR and
 EPA by Pope, Evans and Robbins has lead to the announcement ' ' of an OCR
 contract  for the design, construction and operation of a 30 megawatt
 coal-fired atmospheric-pressure, fluidized-bed boiler at the Rivesville
 Station of Monongahela Power at Fairmont, West Virginia.  Several organiza-
 tions are engaged in bench and pilot plant scale development of pressurized
 fluidized-bed boilers, including conceptual designs of 30 to 635 MW plants.
 These studies include sorbent regeneration and sulfur recovery, as well as
 reduction of SO , NO , trace materials, and particulate emissions.
               X    X
           In an atmospheric fluidized-bed combustion power plant, the
 electric  power is generated by a steam turbine as shown in Figure 2.  In
 a pressurized combustion system, additional power is generated by a gas
 turbine which is driven by the high pressure hot-combustion gas as shown
 in Figure 3.
           Fluidized-bed combustors have not as yet been applied commercially
 in boilers, but there are at least five fluidized-bed reactors currently
 under development for gasification of coal to produce pipeline gas and/or
 liquid fuels.  '    These activities include Battelle"s program for the
 Office of Coal Research employing the Union Carbide coal gasification
 process aimed toward production of synthesis gas; the self-agglomerating
 fluidized-bed concept employed involves production of a  clean gas from
 the combustor and a circulating agglomerated ash burden to supply thermal
 energy to the fluidized-bed gas generator.
           Pressurized fluid-bed combustion (the process that appears to
be preferred for electric utility use) appears to be less complex than
combustion through coal gasification.  Pressurized operation is preferred
because of the significant decrease in the size of the plant over atmospheric
operation.  For large utility plants now contemplated,  the multiplicity of
atmospheric boilers needed would appear to be prohibitive.

Raffle ,.--•	
                                                  Particulate Removal
                                      Hest Recovery
                     ;--r:--'.-7--    , /i.','.,
                     '~»"*-.r~""-~—— ••. . —fv ' v
                                                                       Ash; Per
                               Pra'iaaiei'. ^L-i^.i; hc-atcr
                               or tfehtvisr Soil
                                                     Distribulor Plw'
     Ccal Sue;
     Air Flow:
     Temper a lure:
1  - 25 at;n
pf - 1/4 in.
2-15 ft/sec
1400 - 1900°F
Surfauo:        Water Wiills, Horizontal, avid
               Vartica) Tubsi if Bc.3
Snlfui removal:  »';aO  t S02 + >iO2 -* CaSO4


Paniculate Removal — r-^


Air -«,-.» -^-•T"!1
Force D'iitt



._. .,

...!T»» 	 p

— — - ~, 	 ^~^-
. 	 _.J C
"V, f V"™"
Lime Coal

# r i
v rH V

* %,;ipi 	
f j

	 	 . —

l\ '•';
— «S7-
jier Pi
Induced Draft Fan — j i —
Lconomizer — >

joi r°r.i


r<: ewe-;')'
r 	 ,
/ i
Stiltons A •!<



, 	 ITS.-



-r-W'.w''**— *^\_/-<







                                     18  Status
           The fluidized-bed combustion system offers the potential for
burning high-sulfur and low rank coals having a wide range of properties.
Since fuel-bed temperatures are low, slagging problems are minimized.
Heat release and heat transfer rates to tubes immersed in the bed of
fluidized solids are considerably higher than those in conventional boilers.
Furthermore, it is possible that fluidized-bed combustion can be advanced
to a commercial reality with less effort and time than many of the gasi-
fication methods now under development.
           The major disadvantages of the fluidized-bed approach are en-
gineering rather than fundamental in nature.  Problems currently requiring
solution relate to solids and gas handling, feeding (especially in pressure
operations) and limitations on control of load changes.   Additional
problems are concerned with efficiency of utilization of sorbent, bed
particle attrition and elutriation, and incomplete combustion of coal.
However, in combined cycle operations, there is the fundamental problem of
developing an adequate cleaning process to remove particulates.  An elaborate
reinjection system to handle carry-over of unburned combustibles probably
will be required.
           A fluidized bed of noncombustible particles (not consisting of
lime or magnesia), arranged so that particles of coal are burned in close
contact with inert particles to which they are transferring thermal energy,
offers some promise for particulate control.  This is a highly complicated
system and much is lacking in an understanding of the mechanism of combustion
in fluidized beds that would be required to achieve optimum conditions in
large central-station power plants.  Staged Combustion
          Staged combustion involves burning fuel in a fuel rich primary
zone with less than stoichiometric amounts of air (90-95 percent), partially
cooling the combustion products, and finally supplying additional air and
completing combustion in a secondary zone.  The earliest downfired furnaces
for firing pulverized coal employed essentially a two-stage combustion
process which suggests that two-stage combustion with pulverized coal

 may be feasible.   Two-stage  combustion has been demonstrated, at power
  plant scale,  as  a  successful method  of reducing the  emission of NQX by
  reducing temperatures.   Depending  upon the fusion  characteristics of  the
  ash, this process  may also have  a  beneficial  effect  on  control of
  particulates.  Recirculation of Flue Gas

            The recirculation  of flue  gas  to reduce  the peak  temperature has
  been demonstrated  on a small scale and has been found to  decrease the forma-
  tion of NO .   However,  its effect  on particulate  emissions  is not known.
  Factual data  are lacking regarding the volume of  flue gas that  should be  re-
  circulated,  its  composition  and  temperature,  and  the points at which  it should
  be injected  into the furnace.  Gasification

            Gasification of coal can be considered  as  a modified  combustor
  in which the  first stage, the gasifier,  is operated  rich.  Babcock  and
  Wilcox are currently studying one  variation  in which ash- and  sulfur-free
  gas is produced  and burned  in a  combustion boiler and gas turbine cycle.
  Other unconventional systems which involve  submerged combustion/gasifica-
  tion are as  follows:
           The ATC Molten-Iron Gasification Process,  The concept of
Applied Technology Corporation's  JtBolten-iron coal gasification process is
designed to gasify high-sulfur coal with the use of a molten-iron bath in
which the sulfur is separated  and  recovered in usable form while producing
an essentially sulfur-free,  low-Btu gas,  suitable  for combustion under
steam boilers as a second stage of combustion.
           The heart of the  process is a molten-iron bath with a depth of
3-4 feet contained in a cylindrical, refractory-lined vessel  called  a
combustor.  '   Crushed coal  is injected into the iron bath through a
submerged lance to a depth of  about 25 inches.   This depth has been  found
sufficient to allow time for  the volatile  constituents of the coal to be
driven off and cracked to H^  and  CO and for the fixed carbon  to be dissolved
completely in the  iron.  The  sulfur contained  in the coal is  also dissolved
in the iron.

            Simultaneously,  preheated  air  is  injected  through air  lances
 which need  to  be  submerged  to  a  depth of  only  5-6  inches.  The oxygen of
 the air produces  a  Bessemer-type reaction which preferentially oxidizes
 the dissolved  carbon  to  CO.  By  balancing the  coal  injection rate with the
 air injection  rate, the  carbon content of the  iron  would be maintained in
 the range of 3-4  percent.
            The ash  of the coal,  predominantly  Al 0  and SiO , floats to
 the surface of the  iron  to  form  a  slag.   Lime, as  limestone, is injected
 with the coal  to  flux the coal ash and to provide enough CaO to extract
 the sulfur  from the iron and hold  it  in the  slag as CaS.  Sulfur-laden
 slag is drained continuously from  the combustor to make way for new slag.
 The slag is treated hot  with superheated  steam to produce elemental sulfur,
 H2S,  and SO .   The  latter gases  are sent  to  a  Claus reactor and converted
 to  elemental sulfur.   The desulfurized slag  is disposed of or part of it is
 recycled through  the  combustor if  more slag  volume  is needed.
            The gas  generated is  composed  mainly of  CO, H«, and N_ and has
                             3                          Li
 a heating value of  160 Btu/ft  .  It comes off  the combustor at about 2500 F
 and is  conducted  directly to a steam  boiler  where it  is burned with secondary
 air to  CO   and HO.   A portion of  the  off-gas  is burned in a stove to preheat
 the primary air injected into  the  combustor.   It is reported that, under
 proper  operation, the  SO  content  of  the  off-gas will be less than 60 ppm.
            Status.  Experimental and  development work, under the sponsor-
 ship of  EPA, has  been carried  on in the Pittsburgh Laboratories of ATC,
 since 1970.  It was recently reported     that ATC has received a contract
 to  develop  design criteria for a 50 to 100 MW power generating unit based
 on  their process.  ATC has reported that  preliminary cost studies have
 shown the process to be competitive with  conventional coal-fired plant.
 EPA is also considering  some engineering  studies for application to power plant
which include more detailed cost estimated.
            Submerged Combustion in Molten Salts.   The concept of submerged
combustion of  coal in a molten bath of basic salts is another variation on
 the  theme of simultaneously providing a sink for slag, sulfur,  and trace
metal contaminants of the coal at the  combustion site.  Currently the use
of molten salts for this purpose is being investigated by two  organizations:
 (1) Atomics International is investigating the partial combustion of coal


with production of low-Btu gas by submerged combustion   ; and (2) M. W.
Kellogg is studying the potential for production of high-Btu gas by a
similar process.   '
           The Atomics International (AI) process is an extension of their
previous work with the use of a molten-salt scrubber for the removal of
sulfur oxides from flue gas.     Both Al and Kellogg use sodium carbonate
as the liquid medium at about 1800 F.  The major difference between the
two processes is that the Kellogg process employs steam as a source of
hydrogen to produce a high Btu gas.  The Kellogg gasifier is run at 1200
psia to enhance the hydrogenation of the CO produced by the combustion
process.  The AI process on the other hand involves simple partial combustion
of the coal to carbon monoxide.  In both systems, the sulfur is retained in
the ash/molten-salt mix.  A portion of the contaiminated salt is withdrawn
and regenerated with the production of hydrogen sulfide or sulfur as a
           Status.  Both systems are still in the early investigative
stages.  Many problems, such as amounts of sodium carbonate make up
required, corrosion problems, and the percentage of sulfur retained in the
melt remain unanswered.
           The Superslagging Combustor.  The concept of burning crushed or
pulverized coal in a bath of molten coal-ash slag is another example of
submerged combustion of coal incorporating sulfur removal.    The two
aspects of the Superslagging Combustor that offer an incentive for consid-
eration of the concept as a coal combustion method are (1) capture of coal
sulfur by the slag to provide hot, low-sulfur combustion gases to a boiler,
and (2) capture of fly ash by the slag that could reduce emissions of fine
particulates and perhaps eliminate the need for an electrostatic precipita-
tor.  This concept, being investigated at Battelle-Columbus has not been
tested experimentally and has received only limited analysis to date. Modification of Fly Ash Resistivity
           The ability of an electrostatic precipitator to capture particles
of fly ash depends in a large measure on the electrical resistivity of the
fly ash.  When the electrical resistivity exceeds 2 x 10   ohm-centimeters,

 the normal voltage gradients in a precipitator are upset and the collection
 efficiency drops sharply.  Controlled amounts of SO  may be added to the
 flue gas to alter the resistivity of the fly ash particles so as to improve
 collection of the fine  particulates.  Also,  by modifying  the combustion  process
 by such procedures as changing the rate of burning, particle size, amount
 of excess air and rank of coal, the resistivity of the ash, and conse-
 quently the efficiency of electrostatic precipitators can be improved.  Agglomeration of Ash Particles

          Ash particles could be more easily removed from the combustion
 gases if agglomeration of the fly ash particles could be achieved.  In
 order to adhere on contact the particles must be sticky which imples some
 lower temperature limit at which agglomeration can occur.  High turbulence
 or pulsating flows may promote agglomeration of the stick particles.  Also,
 a longer residence time in the flame should encourage agglomeration,  Advanced Power Cycles

          Concepts other than those presently in central-station power
 plants undoubtedly will be developed over the next 30 years to convert
 the energy in fuels into electricity.  Magnetohydrodynamics and fuel cells
 offermost promise of the so-called direct-energy-conversion processes,
 but both have serious shortcomings.  More attractive at present are advanced
 power cycles wherein two or more energy-extraction systems are  combined in
 a single power plant.
          Typical of such systems is the combined  cycle  burning of  a gas-
 eous or liquid fuel  in a pressurized -boiler furnace followed  by expansion
Of the hot products  of combustion through a gas  turbine.   The gaseous  fuel
could be natural gas, but another likely fuel is  a hot  sulfur-free  producer
 gas made from coal,  or a fuel gas produced from  residual fuel oil.  A  suit-
able gasification scheme including a hot gas-desulfurizing step of  the type
 described above is required.   In addition to higher overall thermal efficiency,
which can be expected to exceed 50 percent, such cycles  have  the great
 advantage of essentially complete removal of sulfur and  particulates  from
 the exhaust gas.

            Many types of advanced power cycles can be devised, including
  those in which volatile matter in the coal is distilled off and only
  the residual char is burned in the power plant.   Such systems have been
  investigated for many years, but without the present-day incentive of
  emission control.  The need for "clean"  power is  renewing that interest^13}
  Power plants based on such cycles should emit substantially less  pollutants
  than current plants.  Modification of  Particle Size
            The  size of the  ash  discharged  from a conventional  furnace  is
  affected primarily by the  size of the coal fired.  The  finer  the grind
  the  finer  the  ash.  Unfortunately, from the particle control  standpoint,
  fine grinding  promotes combustion which makes the resulting ash harder
  to catch.  Also, high  temperature flames increase ash vaporization and
  subsequent condensation into fine particles.   The full effect of the
 particle size of pulverized coal on particulate control requires more
 investigation.  Solvent Refined Coal
           Solvent refined  coal  (SRC)  is  a reconstituted coal,  low  in
 ash and sulfur'   .  The use of SRC in place  of high  sulfur and high ash
 fuel  is a marked modification of  conventional  methods  of burning coal.
 The particulate emission problem  would be  resolved, but  cost is a  factor
 that  would  present  a serious obstacle.   Low  Excess Air
          If  pulverized coal  could be burned with low excess air rather
 than with 15  percent excess  air, as is common  today in large boiler
 furnaces, the problem with particulate emissions would be eased.  Advan-
 tages are that  the volume of the flue gas would be reduced appreciably
and more effective control of particulates would be possible, and higher
efficiency would be achieved (meaning less thermal pollution).  To  accomplish
this may require improved burners, finer  sized coal,  and higher turbulence
levels in the furnace.


          In a recent investigation of particulate emissions from oil-
fired residential heating units, conducted by Battelle for the EPA, two
units (a warm-air furnace and a boiler) were fired in the laboratory
while smoke and filterable particulate emissions were measured at several
excess air levels for both cyclic and steady-state runs.  In addition,
particle-size distributions were measured during runs on the boiler to
determine if particle-size variations might help explain the lack of
correlation between smoke and particulate emissions, based on earlier field
          It was determined that particulate emissions varied with excess
air in the same manner as smoke varies with excess air.  Also, correlations
between smoke and particulate emissions appeared practical for given units
firing at specific operating conditions.  However, the data did not suggest
a general correlation between smoke and filterable particulate emissions.
Particle size distributions indicated that over 80 percent of the particles
were below 1.0 micron and the distributions were nearly uniform for all
runs, so that particle size did not explain differences between smoke and
particulate emissions.  Electrochemical Oxidation of Coal
          Fuel cells using coal as the fuel may offer a long-range solution
to the need for supplying large quantities of electrical energy without
excessive emissions.  The extent to which particulate emissions would be
generated is uncertain and would depend on details of the process which
are not now known.  Fuel Additives
          The U.S. Environmental Protection Agency    completed in 1971
a study of the use of fuel additives to control air pollution from distill-
ate oil-burning systems.  Each candidate was analyzed for elemental com-
position to provide a basis for testing.  A standard screening procedure
was established to test the effect of each additive on emissions from
fuel oil combustion.  Screening tests were carried out on all distillate
soluble additives and the most promising additives were subjected to a
rigorous examination.

            Fuel  additives were  found not  to be a promising way of reducing
  air  pollution from  distillate  oil combustion.  A majority of the additives
  tested had no beneficial effects on air  pollutant emissions; in fact, some
  additives  even  increased total particulate and NO  emissions.  Several of
  the  metal  containing additives, e.g., Ferrocene, Cl-2, and Fuelco SO
                                                                     3, re-
  duced total particulate emissions; however, the unknown toxicity of new
  emissions  they  create makes their use questionable.  Further, there is
  evidence that for distillate oils, burner modifications are a more suitable
  route to air pollution control.  Improved Control Methods for Fine Particulates
           The only control method now being tried  for subtnicron particle
 abatement is the application,  in new installations,  of larger and more
 efficient precipitators.   Experience thus far indicates that  efficiencies
 of 99.5 percent  and  above  will achieve invisible plumes except  for  the
 sulfuric  acid mist that appears almost immediately  as a bluish-white  haze
 when the  small amount  of  sulfur trioxide  in the  emerging gases  combines
 with the  moisture in the ambient  air to form  ^SO^ mist.  Assuming  that
 SO  abatement processes become  feasible so  that H SO   mist  is eliminated,
 then achievement of  fly ash emission levels low enough to produce invisible
 exhaust plumes will  be  feasible at a price.   Electrical  Control of  Particulates from Flames
          A radical  departure from conventional submicron particle
 abatement methods is the application of electric fields to charged dis-
 persions  so as to alter particle trajectories so that fully formed particles
may be caused to deposit in specific places or to burn up if they are
 combustible.   Also,  the rate of particle formation, size,  and concentra-
 tion may be modified by applying electric fields to the region in which
 they are formed
          It has  been demonstrated that the entire  process  of  burning sprays
and particulate dispersions can be controlled  electrically,  all the  way
from electrical dispersion  and  charging,  over mixing with air  inducted by
ion pumps, to burning controlled by electrical fields.

          The attachment of flame ions, which would otherwise recombine
unprofitably, to particulate pollutants to allow their subsequent manipu-
lation by electric fields appears possible for all particulates.  For
species whose boiling point lies below the flame temperature, it is
necessary to use a second flame as an ion source in the cooler part of
the stream.
          The concept has only recently been introduced, but already the
theory of predicting rates of charging and hence particle trajectories
in fields, in terms of their growth histories has been established, and
predictions of optimum conditions can be expected to follow

 3.2  Industrial Processing and Steam Generation
 3.2.0  Introduction
           This subsection is concerned with process modifications with
 the objective of controlling particulate emissions from industrial processes
 such as cement and lime production, glass melting and brick and ceramic
 production, and from industrial steam generation and space heating.
 Metallurgical processes are covered in Sections 5, 6, and 7.
           Most pollutants from industrial processing sources  are not
 subject to reduction by modification of the combustion process.  For
 example,  the particulates include those from thermal processing in which
 the finest fractions of the processed materials or ash escape with the
 combustion gases;  such particulate emission would  normally be reduced by
 means of  dust collectors rather than by combustion modifications.
           The amount of particulate POM formed  will vary  greatly.  Efficient,
 controlled combustion favors  very low POM emissions,  whereas  inefficient
 burning results  in  high emissions.   Hand-stoked residential furnaces
 account for most of the particulate POM emitted.
 3.2.1  Process Modifications
           Energy-conversion devices for industrial  processing and  steam
 generation are used primarily to  convert  the chemical  energy  in  fuel to
 thermal energy in the  form of steam, hot water, or warm air.  Possibili-
 ties  for reducing particulate emissions by process modifications are
 discussed  in  the following paragraphs. Burner Design
           Burner design is not a well developed scientific discipline
 but has been, and still is, largely an art.  Most burner designs are
 arrived at by trial and error and hardware-oriented development, rather
 than by direct application of a body of scientific knowledge of  the
 subject of combustion.  This is not too surprising, considering  the
complexity of the combustion process.  However,  as the combustion process
is understood more  fully and as complete mathematical models describing
both the physical and chemical aspects of the process are  constructed,

there should be improved understanding of the relationship between par-
ticulate emissions and combustion conditions.  As a result, burners and
furnaces may eventually be designed to assure minimum generation of fine
particulates.  Flue-Gas Recirculation
          Flue-gas recirculation involves diluting the fuel-air mixture
in the combustion zone with cooled flue gases  and,thereby, reducing flame
temperature rise and peak gas temperature.  Utilizing cooled flue gas
as the diluent does not adversely affect combustion efficiency as neither
the mass nor the temperature of exhaust products need increase.
          Studies ^  'have shown that recirculated flue gas can also
contribute to lower emission of smoke when firing fuel oil.
          Application of flue-gas recirculation to small and intermediate
size combustion equipment is limited by knowledge gaps in the following
          •  Effect on emission levels of variables such
             as temperature of the recirculated gas,  point at
             which the gas is injected, and quantity of re-
             circulated gas
          •  Corrosion and deposits associated with recirculation
          «  Control of recirculated flue gas quantity in equip-
             ment operating at variable firing rates
          •  Start-up problems.  Two-Stage Combustion
          Two-stage combustion involves burning fuel in a fuel-rich
primary zone, partially cooling the combustion products, and, finally,
supplying additional air and completing combustion in a secondary zone.
Thus, two-stage combustion makes possible low peak gas temperatures,
because the combustion products are cooled before combustion is completed
and oxygen levels are low in the highest temperature  region—the primary
combustion zone.   Beinstock, Amsler, and Bauer    'found that,  when
pulverized coal was burned in a power plant with 5 percent excess air in
the primary zone and 17 percent excess air added at the proper  location
downstream, NO  emissions did not increase above levels obtained when


  burning with 5 percent excess  air  and  no  downstream air  addition.   Two-
  stage  combustion has  not  been  demonstrated  as  practical  on  smaller
  combustion units.
            Application of  two-stage combustion  to  small-  and  intermediate-
  size combustion  equipment  is limited because of gaps in  knowledge of:
            •   Effect on emissions of important  combustion
               variables such as primary-zone fuel-air ratio}
               temperature drop between  the primary and the
               secondary zones, and  secondary zone fuel-air ratio.
            •   Possible  corrosion associated with reducing con-
               ditions  in the primary zone.
            *  How to design combustion equipment to achieve
              effective two-stage combustion (i.e., how  to achieve
              necessary mixing in the secondary zone, how  to achieve
              burnout of incompletely burned products formed in the
              primary zone, what the residence-time requirements are
              for primary and  secondary  zones,  and  how to  incorporate
              sufficient cooling between zones  without increasing
              equipment size and emissions  of smoke,  C0, and  HC).  Other Combustion  Modifications
           Other techniques  to achieve combustion at  lower temperatures
 and  possibly  reduce  particulate  emissions  are  (1)  catalytic-  or
 surface-combustion burners, (2)  radiant-heating devices,  and  (3) com-
 bustors  that  operate at high excess air.   These techniques have not been
 developed  to  the  point of practical application to heating boilers or
           Surface combustion has been used successfully on a number of
 gas-fired  infrared or  surface heating units.  These units operate with
 combustion  zone temperatures of about 1800 F to 2000 F and, although no
 specific data are available, these units are likely to have low NO
 emission levels.  Surface combustion of  gaseous fuels could be incorpor-
ated into residential heating  units  with little difficulty, although

 practical  combustion  of  fuel  oil has  not  been  demonstrated  for  long-term
 operation.   Potential problems with fuel  oil include  achieving  complete
 combustion,  achieving uniform fuel feed across the burner face,  and
 preventing deposits on the  burner surface.  Improved  Service and Maintenance
           Although only  scattered data on burner  servicing  and maintenance
 are  available     '    , it  is known  that  lack of proper burner adjustment
 and/or maintenace can result in poor combustion ^   "and higher levels of
 pollutant emissions.  This is especially true with  regard to pollutants
 associated with incomplete combustion of the fuel (combustible particulate,
 CO,  and HC) .
          Medium- to large- size combustion units are most likely to receive
 proper maintenance under service contracts or by on- site operators.  There-
 fore, reasonably good combustion performance of these units should be
 realized.  However, emissions such as particulate POM that might be over-
 looked by service personnel still may not receive proper attention.
 Education and proper instrumentation would assist in this area.  Modifications to Reduce PNA
          PNA (Poly-Nuclear Aromatic) compounds, which appear in particulate
 POM  (Polycyclic Organic Matter) comprise only a small fraction of total
 particulate emissions, but they are significant as some of them are
 carcinogenic .
          Over 90 percent of all PNA emissions from energy-conversion
 combustion processes is attributed to commercial and residential coal-fired
 sources utilizing fixed-bed combustion.   Although small coal- fired units
 are disappearing from use,  their combustion with regard to PNA emissions
 is significant.
          PNA compounds may be  present in the coal itself or may form
                              (21 22)
during the combustion process v  '   .   However, there appears to be no
data in the literature which identify PNA as being present in coal.   Diehl,
DuBruel,  and  Glenn x  ' observed variations in PNA emissions when firing
different coals on a pulsating-grate stoker.   However,  they  also  observed


 such variations when operating a  pulverized-coal-fired  boiler with  one
 fuel and  under essentially constant conditions.   Consequently,  they were
 not  able  to  relate  PNA emissions  to coal, and  the source of  PNA  remains
          The  destruction  of  PNA, as  for any other hydrocarbons,  should be
 by oxidation during  the combustion process.  The  literature  appears to
 contain no data which  would identify  any unique conditions required for
 oxidation of PNA '    .   However,  the  fact that PNA compounds are  emitted
 suggests incomplete  combustion of these compounds.
          PNA  emissions  decrease as unit size increases.  Larger  units
 provide longer  residence times in the flame and,  generally, higher average
 gas  temperatures--so that conditions contributing to more nearly complete
 combustion prevail.  Also, fixed-bed combustion units (with poorer mixing)
 tend to emit larger quantities of PNA than do spreader stokers or pulverized-
 coal-fired units.
          It is possible that, with a better knowledge of how PNA occurs
or forms and of the conditions promoting its oxidation,  modifications of
 the combustion process to reduce PNA emissions  could  be  identified.

4.  Municipal Incinerators
4.0  Introduction

          Approximately three percent of the 4.3 billion tons of solid waste
generated in the United States annually is municipal solid waste (MSW)v   .
This means that about 130 million tons of MSW, high in organic content, is
generated per year comprised mainly of residential, commercial, and some
industrial wastes.  This waste can be utilized as an economical and ecological
fuel to help relieve the "energy crisis" and, at the same time, achieve
considerable recovery of certain natural resources such as metals and glass,
provided particulate emissions can be controlled.
          In 1968, about 90 percent of MSW was disposed of by landfill, with
only 9 percent being incinerated    .  Waste requiring disposal is increasing
at the rate of 7 million tons per year, whereas suitable areas available for
economic disposal by landfill is rapidly dwindling.  The alternative is to
incinerate an increased percentage of this material.  A result would be more
air pollution unless improved burning processes are developed and adopted.
The fact that incinerators can be located near population centers and that
they require less land areas in comparison to landfills enhances their
atrractiveness to urban planners.  Incineration is expected to increase in
popularity as improved methods, especially for particulate control and resource
recovery, are developed.  The use of MSW as a source of energy should be a big
plus factor.

4.1.0   Process Modifications

          A number of process modifications have been proposed and several
are under investigation (some at the pilot plant stage) to control emissions
of particulate and,  at the same time,  utilize some of the heat generated
and recover certain natural resources  such as metals and glass.   These will
now be discussed individually.

  4.1.1  Combined Firing

            In a  current task  order  study  for  the Emission Standards and
  Engineering Division,  OAQPS  of EPA, Battelle-Columbus is compiling information
  on the combined firing of municipal refuse and fossil fuels in steam-
  generating  boilers.  The study includes a survey to determine the planning
  status of new and  retrofit combined- firing installations in the United
  States, an  engineering discussion of the technology of combined firing,
  estimation  of the  emissions anticipated from such sources, descriptions of
  control devices  for these emissions, and the impact of combined firing on
  fossil fuel usage  and  on solid waste disposal situations in the United States.
           The only extensive application of combined firing at present is
  the demonstration  project in St,  Louis  which is  sponsored by EPA,  Union
  Electric and  the City of St.  Louis.  Test data  on emissions are incomplete
 at this time.  However, it is expected  that,  for  a  nominal  firing  ratio of
  80 percent pulverized coal and 20 percent prepared  municipal refuse,  the
 particulate emissions  can be  controlled by  electrostatic  precipitators,
 Codified somewhat to accommodate  the  higher  stack gas  volume produced,  so
 that final particulate  emissions  will be  similar  to those produced when
 firing coal  alone.

 4.1.2   The CPU-400

          Combustion Power Company, Inc., developed  the CPU-400 system which
 recovers energy  from the combustible MSW in the form of electric power through
 the use of a gas  turbine powered  electric generator.  Simultaneously, the
 CPU-400 will concentrate, separate, and recover valuable noncombustible
 materials for  recycling.  Claims are that CPU-400 produces about five percent
 of  the  electric power needs of the community supplying the solid waste.
 Income  from  the sale of  electric power and recylced materials should result
 in a substantial reduction in the cost of solid waste disposal.
          The pilot plant of the CPU-400 was  developed in Menlo Park,  California,
This plant is a complete system capable  of consuming 80 tons per day of


combustible solid waste while producing 1,000 kilowatts of electric power
and separating steel, aluminum, glass, and other inerts from the incoming
material.  The CFU-400 system will be composed of three identical fluid bed
combustor/gas turbine modules.  Each 3,000 kilowatts module will have a
capacity on the order of 150 tons per day.  The modular design of the
CPU-400 will allow plants processing from 150 to 1,500 tons per day to be
constructed and will provide redundancy to assure reliable disposal
capability.  During a visit to Menlo Park by a Battelle staff member on
                / n f \
February 6, 1974    , he observed that, when burning sawdust, the system
works well, the stack is clear, and the dust loading to the turbine and
to the stack is extremely low.  Turbine erosion and corrosion do not appear
to be problems.  However, there is a problem of deposition of A.l?0~ on
turbine blades.  This is attributed to the melting of aluminum foil which
enters with the garbage.  A particle bed collector is being considered as
a final stage for removal of these Al 0_ particles.  In cleaning the
combustion products do not meet turbine requirements, environmental
standards are met.

4.1.3  Water-Walled Incinerators

          The use of pollution-control devices requires that the incinerator
furnace gases be cooled to permit operation of the flue-gas cleaning equipment.
An attractive way to do this is to absorb much of the heat by water contained
in furnace wall tubes and in convection-pass tubes.
          In addition to greatly decreasing the size requirements for pollution-
control equipment and fans,  the water-wall incinerator has several attractive
features.  First,  there is a gainful use of the heat energy available in the
refuse.  Second,  the furnace throughput can be increased because of the rapid
absorption of heat.   Third,  wall slagging problems often encountered in
refractory-type construction are absent.
          The technology of the water-wall incinerator was developed first  in
Europe where it is in fairly extensive use.   Corrosion problems  have been
reported, however, in some instances.   The first operating water-wall unit  on

  this continent was at the Navy Public Works Center in Norfolk,  Virginia,
  beginning in 1969.  Units are now operating in Montreal,  Braintree,
  Massachusetts, Harrisburg, and Northwest Chicago.  In all locations  except
  Braintree, the steam is being wasted for lack of local demand.   It  is
  anticipated that this type of construction will be used more and more  in  the
           In March, 1969, research was started at Battelle-Columbus  on a
  grant program supported by the Solid Waste Management Office, EPA, which was
 aimed at determining the cause and extent of fireside metal  wastage  in
  incinerators and devising methods of alleviation.   In March,  1971, work was
 started on a supplemental program aimed at obtaining  a better understanding
 °f incinerator-gas scrubber corrosion and also of  metal wastage  of grates.
           Field  and laboratory studiesv    demonstrated that  the wastage
 in water-wall refuse boilers can  be more severe  than  that normally encountered
 in fossil-fuel-fired boilers.   The complex nature  of  the refuse used as the
 fuel  and the relatively  poorer control  of burning  in an incinerator combine
 fco increase  the  possibility  for corrosion.   The  contributors to the attack
 are corrosive gases  and  low-melting chloride and sulfur-containing salts
 which exert  a fluxing action on the protective films on the metal surface.
 These low-melting  salts  primarily  contain  compounds such as zinc and lead
 chlorides along with potassium bisulfate and potassium pyrosulfate.   The
 data developed reveal that the gases S02, S03, HC1, and C12 are also playing
 a major  role  in  the wastage  processes.
          Analyses of tube deposits and furnace gases confirm the belief
 that sufficient quantities of  the deleterious salts and gases are present
 in all municipal incinerators  to warrant careful consideration from  a
 corrosion standpoint.  Particulate emissions should be substantial,  more
 controlable in water-walled incinerators.

^•1.4  Pyrolysis  Processes

          Pyrolysis is a  process by which carbonaceous materials  break  down
into simpler compounds and elements by means of heating in  the absence  of  oxygen.


A  breadown  of  the organic portion of the refuse into oils, gases and chars
occurs.  The Bureau of Mines has a pyrolysis unit which consists of a furnace
heated with nickel-chromium resistors and a recovery train to trap products.
Dependent on the reaction conditions, one ton of municipal refuse yields
154-230 pounds of char residue, 0.5-5 gallons of tar and pitch, 1.2-2 gallons
of light oil, and 11,000 - 17,000 cubic feet of gas, in addition to 80-133
gallons of  aqueous liquor, and 18-25 pounds of ammonium sulfate.  Typical
thermal values for the products are: char, 8-13,00 Btu/lb; oil, 150,000 Btu/gal;
and gas, 500 Btu/cu ft.
          Enviro-Chem has developed a Landgard pyrolysis process which
emphasizes  disposing of the solid waste, rather than recovery.  The refuse
is shredded and reduced to a fairly homogeneous mixture with particle size
                            ( 28}
about 3-4 inches in diameter    .  It is then dumped into a kiln lined with
refractory  material.  The kiln is slightly inclined from exit to entrance
so charred  material falls by gravity into the water-quenching unit.  The
residue is water-cooled before passing to a magnetic separator which removes
ferrous metals.  The wet residue, reduced from the original trash volume by
94 percent, is then loaded into trucks and taken to landfill.
          Gases driven off from the kiln during pyrolysis are passed through
a combustion chamber where hydrocarbons are oxidized.  Complete combustion
is assured  by an afterburner module.  Product gases--carbon dioxide, nitrogen,
etc.--are passed through an adiabatic spray-scrubber and then released to
the atmosphere.  Steam plumes, which might be objectionable from a public
relations standpoint, are elmininated by heating the stack gases.  Since the
gases are clean, very low stacks can be used except where otherwise required
by local ordinance.   The Landgard system accepts municipal refuse as it is
and no hand separation procedures are necessary.  The company guarantees to
meet all existing air pollution standards in effect when the contract was
signed.   Cooling water is diverted to sedimentation basins for recirculation.
However, the Landgard process  has not yet been demonstrated commercially.
          The Garrett process  is designed to recover salable heating fuels,
glass and magnetic metals.   The organic portion of these wastes is  converted
to low sulfur oil, char and gas using a flash pyrolysis process.   The process

 is designed to be  expanded  into an  integrated series of processing stages for
 the recovery of over  90  percent of  the raw materials contained in municipal
 refuse.   Incoming  solid wastes are  shredded, dried, and passed through an
 air classifier which  separates most of the metals, glass, and other inorganic
 materials.   The overhead  stream from the air classifier is then subjected
 to a two-stage screening  to improve separation of inorganics.  The remaining
 refuse  is shredded a  second time and then pyrolized, where it is broken down
 into smaller molecules through the  application of heat in the absence of
          Laboratory  studies of the pyrolysis process resulted in the production
 of approximately one  barrel of good quality oil per ton of as-received refuse.
 Such refuse usually will also yield about 140 pounds of magnetic metals,
 120 pounds  of glass,  and  160 pounds of char.
          The flash pyrolysis operation which is the heart of the Garrett
 process was  researched for  over a year  in a continuous laboratory reactor.
 Product yields, quality, and the initial favorable economic projections have
 since been  confirmed  at a 4 TPD pilot plant during an 18-month period of
 operations  at LaVerne, California.
          The  quantity of solid material going to landfill from a 200 ton/day
 plant is only  16 tons/day.  The solid debris contains unrecovered glass,
 aluminum  copper,  zinc, nickel, and other sterile material.  GE&D is currently
 investigating  the  economics of reclamation of nonferrous metals,  and if a
 system can be developed, the tonnage of solid rejects going to landfill
will be halved.
          Other pyrolysis processes which could reduce particulate emissions
are described in a catalog of resource recovery processes  prepared by Midwest
Research Institute for the Council on  Enviornmental  Quality in February,  1973^29'

4.1.5  Advanced Concepts Involving Heat Recovery

          Other methods of incineration involving  energy conversion are
described in the NCRR Bulletin for the Summer of 1973V   '.   They  include:

           •   Burning  refuse  in  existing heat exchangers
           •   Hydrogenation,  in  presence of CO and steam
           •   Anaerobic  digestion  to  produce methane
           •   Cubetting/Briquetting.
           Energy recovery and the productive reutilization of materials,
 combined with effective control of particulate emissions should result in
 the commercial development of some of these concepts over the next five years.

 4.1.6   Electron-beam  Irradiation

           The Japan Atomic Energy Research Institute (JAERI) and Ebara Mfg.
 Co., are jointly grooming a  project at jAERI's Takasaki laboratories, whereby
 such treatment causes both sulfur dioxide and nitrogen oxides to drop out.
           In  one test,  exposure to a "few megarads" of irradiation produced
 90  percent SO- removal and  virtually 100 percent NO  removal from a 10-cu m/hr
              fc                                     X
 flow of flue  gas containing  about 1,000 ppm of SO. and 80 to 100 ppm NO .
                                                 £*                     X
           Full process  details  have not been disclosed, but the Japanese
 indicate that the two impurities become captured in aerosol form by electro-
 static  dust collectors.

 4.2 Predictions for  the Year 2000

           Potential incinerator particulate emissions for both uncontrolled
 (furnace emissions) and for  those abated by control devices (stack emissions)
were estimated as follows:

                                   Thousands of tons per year
                                   1968                  2000
           Furnace particulates      182                  1064
           Stack particulates        142                   391

These data reflect a  large increase  in disposal  of MSW  by  incineration, but also
a great improvement  in the control of particulates.

5.  Iron and Steel Plants

          Particulates generated during the production of steel can originate
from several sources.  These sources are rust, scale or dirt on the steel
scrap used as part of the furnace charge, the degrees of size and thickness
of the individual pieces of scrap (small, thin pieces of scrap will oxidize
so rapidly that they approach an actual burning situation), fine parti-
culates associated with flux additions to the charge, refractories used for
patching, and the oxidation of metallics during the period which high-
purity oxygen is blown into the  molten metal bath to accelerate the
refining reactions.
          Particulates of interest in this study are those that originate
from the flux additions and the oxidized metallics formed during the
oxygen lancing periods.  These particulates are associated with the three
major processes for making steel; open hearth, basic oxygen (EOF or BOP),
and electric-arc furnace processes.   Considerable work has been done in
the United States and Europe in an attempt to determine the mechanism or
                                                          f *\C\ *3 1 *%. *) ^ ^ ^ ZL T ^ ^ A \
mechanisms for the formation of fume during oxygen lancing^  '  '  '  '  '  '  .
Information on the actual mechanism of formation, that could possibly lead
to process modifications and improved particulate control have not been
forthcoming.  However, the factors that affect the formation of fume
have been found to be: (1) height of the oxygen lance above the molten
metal bath, (2) the carbon content of the molten metal bath, (3) molten
bath temperature, (4) concentration of oxygen in the lance gas (i.e.,
purity of oxygen), and (5) size of orifice and number of orifices in the •
oxygen lance

5.1  Open Hearth Furnace
          The open hearth furnace is a shallow hearth furnace that can
be alternately fired from either end.  Refractory brick work regenerators
are located at each end of the furnace which serve to recover heat from
the products of combustion and in turn preheat the incoming air for fuel
combustion.  Briefly the process consists of charging steel and iron
scrap into the furnace, melting or partially melting the scrap, charging

molten pig  iron  to  the furnace, and refining the steel by blowing high
purity oxygen  into  the molten bath.  The greatest particulate problem
occurs during  the time high-purity oxygen is used to remove carbon,
manganese and  silicon from the molten bath.  During the oxygen lancing
period very  turbulent conditions are created in the bath, with the result
that  significatnt quantities of iron, manganese and silicon oxides are
formed and carried  into the exhaust system of the furnace.

5.1.1  Hydrocarbon Additive to Lancing Operation
          Laboratory research work started in 1959 resulted in a report
in  1963 concerned with the use of natural gas in combination with oxygen,
as  a means for reducing the amount of emissions during oxygen lancing of
steel     .  Various factors were investigated  as  to their affect on fume
formation.   These factors were: (1) amount of methane, (2) carbon content
of  the bath, and (3) temperature of the bath.  The use of methane in this
manner was shown to be beneficial in reducing the amount of fume, on
laboratory scale experiments.  State of Development - Hydrocarbon Addition
          Development work was not actually undertaken since the applica-
tion of the modification requires only a source of hydrocarbons and the
valving with controls to regulate the flow of hydrocarbons to the oxygen
lance.  Industrial trials of tMs process modification were reported in
1966    .  This work was done with propane-oxygen combinations on 225
net ton open-hearth furnaces in Canada.  Results of this work were
similar to those obtained with natural gas in the earlier laboratory
studies.  Availability to Industry
          There are no apparent restrictions on the availability of this
process modification for use by the industry.
          Acceptance by Industry.   The iron and steel industry has not
accepted this process modification, if the lack of  further industrial
work can be taken as an indication of acceptance.   Shortly after the final
laboratory work reported in 1963,  plant trials were held,  but the results
were inconclusive and were not reported.   This was  followed by the work

reported in 1966.  No additional work of record has been reported.  Lack
of acceptance may be attributed to the following: (1) more stringent air
pollution control requirements (hydrocarbon additions, reduced but did
not eliminate emissions), (2) possible additional safety hazards associated
with such a modification, and (3) increased cost of natural gas and similar
hydrocarbons.  Degree of Effectiveness
          The use of hydrocarbons in combination with oxygen during the
lancing of open-hearth steel did reduce the grain loadings per cubit foot
of exhaust gases.   No data were avaialble pertaining to the agglomerating
characteristics of the particulates.
          Particulate loading for the work reported on the 225 net ton
open-hearths is shown by the statistical distribution in Figure 4.  A
relationship between dust loadings with the use of oxygen-propane mixtures
and oxygen alone, is shown in Figure 5, for the same data.  The work con-
ducted with the oxygen-propane mixtures did not consider many of the
operating variables that would be of interest to an open-hearth furnace
steelmaker.  The report did state the desireability of further tests to
evaluate the process modification in greater detail.  Environmental Effects
          There were no detrimental effects mentioned in the work reported.
While the reported results did show a reduction in the grain loadings,
the work was done without the use of any type of pollution control equip-
ment.  Therefore, the reported results do not lend themselves to a con-
clusion on the overall effects of collection efficiency and improvements
in the ease of collection.


01 ~
w ' - .:.
c : " •
ra — 11"
c ~-~-
60 ~0— fr
•H ~
w -


> 1
— Log
. ._
' - .  ^^

1 i ' :
\ ,


g (O^+Propan
— —

s -
X^ • 1
" 1
:9.-6 0-7 0.8
0.9 1.0
Dust Loading

i .
• • •
e)) =
. . ._ _.


-- •
• -- '

> \
1 \ i

• i
; :.. .-.- ~r-*
172 + 0.3221
— ~
. 	 -

— 	 	 1 \
" 1
L. 	 :_
. . _ . 	 _
. - — i

1 ! i

i i i 1

43 (Du
\ :_ :

\ :J
— ] — \ — ' — '
i j
— ; : i ;
', '

1 i 1

• - •
t Load
• ! . i
— ~ 	 [—

— -^— —

1 . '
I 1
I ' '
! i
ing 02)^— j+
— • — i 	 '-
__ — ._, ,
— 	 , 	 ._
	 — — ^ 	 	
— — '. 	 1... . . ~~
	 . 	 1 — 	
— — • —
	 , — -


| - ^- — 	 	
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.
With Oxygen Only, grains per SCF
5.   Relationship Between Dust Loadings of a 225 net  ton Openhearth
    Operating With Oxygen Lancing and Oxygen + Propane Lancing
                                                                       Furnace  Use of Liquid Oxygen and Liquid Hydrocarbon Injection
          A search of the foreign literature resulted in three items origin-
ating in the USSR with respect to the use of liquid oxygen and liquid
oxygen-liquid hydrocarbons in the lancing of open hearth steels.   A con-
tinuing search is in progress to obtain any information concerning reports
on the work.  The three items are quoted from abstracts in the literature
and are given for information and possible future use.

          Iron and Steel Refining, USSR Patent No. 236499v   .  Iron and
steel refining is carried out by blowing with liquid oxygen and this reduces
the smoke formation and increases the amount of refined product.  Liquid
oxygen in quantities corresponding to the gas is fed through the insulated
pipes to the 100 ton ladle containing iron.  The heat is spent on bringing
the metal to the boil, for evaporation and for maintaining the temperature
of the reaction.  Due to the low temperature of the reaction, the forma-
tion of smoke is reduced.  The oxygen is diffused in the melt ensuring
fast oxidation of the impurities.
          Treatment of M>ital. USSR Patent No. 276116    .  In ferrous
metallurgy, a molten metal is treated with liquid hydrocarbons.  To speed
up the process and reduce smoke formation, the liquid hydrocarbons are
fed into the molten metal in a mixture with liquid oxygen.
          Liquid Oxygen Blastv  '.  First pilot operations with the use
of liquid oxygen blast were carried out in 600-ton furnace No. 10 at the
Kommunarsk Metallurgical Plant.  During a two-month period, 32 heats
were melted producing 20,000 tons of steel.  The use of liquid oxygen
reduced iron losses to 90-92 kilograms per one ton of steel, reduced the
amount of particulates in waste gases by 30 to 35 percent and increased
the steel output by 1-2 percent.
          V. Pereloma revealed plans to put in operation the first
commercial installation during the second half of 1973 when special
cryogenic equipment will be available.

 5.2  EOF Furnace

           The EOF (Basic Oxygen Furnace)  process,  or BOP (Basic Oxygen
 Process), as it is sometimes referred to, utilizes a pear-shaped steel
 shell lined with refractories to contain  the molten iron and scrap used
 in making steel.  The usual charge consists of scrap, molten pig iron
 and flux.  After the materials are charged into the EOF furnace, a water-
 cooled oxygen lance is lowered into the mouth of the vessel and high-
 purity oxygen blown into the charge.   The high velocity of  the oxygen
 stream (approaching super-sonic speeds) impinging  on the surface of the
 tnolten metal, produces a violent agitation and intimate mixing of the
 oxygen with the molten iron.  Rapid oxidation of carbon,  silicon and
 manganese in the iron, and  reduction of sulfur content by the chemical
 action of the flux produces a heat of steel.   Violent agitation of the
 molten bath during the flowing process results in  the oxidation of very
 fine  particles of iron,  and the ultimate  formation of the fume associa-
 ted with EOF steel making.   The first EOF was placed in operation in the
 United States in 1957.   Since that time the EOF process has  become the
 major steel producing process in the  United States,  accounting for
 83,260,000  net tons in 1973,  for 55.3 percent of the total  150,431,000
 net tons of steel produced.   Research on   the various factors affecting
 the formation of fume are the same as those described in  the  section on
                                                              /OQ O 1  OO
 open  hearth furnaces  and  reported  in  the  published  literature   '*'
 5.3   Q-BOP  Process
          Work has not been  reported  concerning  any  efforts to  improve  the
 collection  characteristics of  BOF  steelmaking  fume.   However,  the "Q-BOP
 Process" developed  in Germany  to improve production  capabilities has re-
 sulted in a secondary benefit of lesser fume evolution during the steel-
making process.
 p,                i
5-3.1  State of Development - Q-BOP Process
          The modified bottom-blowing process  was  developed and placed
into commercial operation by the Maximillianshuette Iron and Steel Company

in West Germany, during 1967-1968    .   The process was developed primarily
to produce steels from high-phosphorus  pig irons which are predominant in
Europe.  It is called the OEM (Oxygen Bottom-blown Maxhuette) process and
                                                   (43 441
is covered by United States Patent Number 3,706,549       .   Process
modification consists of shrouding high-purity oxygen, blown into the
vessel through tuyeres in the bottom of the furnace, with a  hydrocarbon.
The vessel is shown schematically in Figure 6.  In 1971,  officials and
technologists of the United States Steel Company visited  Maxhuette,
undertook development work of their own directed toward the  production of
steel from low-phosphorus pig irons, which are predominant in the United
States    .  In December 1971, the U.S. Steel Corporation announced th^
construction of a Q-BOP plant at their  Fairfield, Alabama works    ,
which was followed by an announcement in early 1972, that the conventional
BOP shop under construction at Gary, Indiana, was to be  converted  to  a Q-BOP
A very good description of U.S. Steel Corporation's development
work and plant installations is given in the published literature
                                               Steel Shell
                                     Plug Bottom

                                                       or Shielding
           Figure  6.  OEM  Process Vessel, U.S. Patent 3,706,549

          Availability  to  industry.  The availability of  the Q-BOP, and
 a  few  similar  processes, to  the  industry is  somewhat clouded by the
 patent situation of  the  several  steel companies involved in the various
 process  developments^    .  This  is especially  true with extensive patent
 litigation which started in  1966 and is still  in progress^51'52'53»54>55>56)
 with the conventional EOF  process.
          Acceptance by  Industry.  The acceptance by industry is of
 necessity influenced by  the  patent situation mentioned above.  However, if
 the patent situation were  to become clarified, it is possible that within
 5  years, eight of  the remaining  21 open-hearth steelmaking  shops could be
 replaced by Q-BOP  type steel plants.   Three will be  closed.  The  remaining
 10 open-hearth shops have  furnaces that are of rather recent construction
 and have had considerable  air pollution contol equipment installed,  would
 probably still be  operated until replacement was required.  This replacement
 would  probably take  place  over a 5 to  15-year  period from now.

5.3.2  Degree  of Effectiveness
          No published reports have been made of the Q-BOP process that
Provides quantative information on the reduction in particulate emissions.
Reports indicate that the bath is quieter,  the particle content of the
waste gases less,  but coarser, and that quantities are only 1/3 to 1/5
that of the BOF furnace.
5.3.3  Environmental Effects.
          More definitive information on off-gases and  particulates  from
the Q-BOP steelmaking process must be made  available  before  an  accurate
evaluation of  the  effects on  the  environment can  be made.  However,  if the
reported reductions in particulate emissions bear out early  indications,
the only effect on  the environment can be  that of an  improvement.

 5.4   Electric Arc Furnace

          The electric-arc  furnace  is a  short cylindrical-shaped  furnace,
 having  a  rather  shallow hearth.  Three carbon, or  graphite,  electrodes
 project through  the roof into  the furnace.  Electric energy  passing
 through the  electrodes and  into  the charge create  the heat required to
 melt  the  charge.  Furnaces  are constructed with  fixed or moveable roofs,
 with  the  vast amjority in the  United States having moveable  roofs.
          Charge materials  usually consist of 100  percent steel scrap,
 with  the  exception of one major  steel plant that uses molten pig  iron as
 part  of the  charge.  Prepared  scrap is loaded into charging buckets in
 advance of the melting operation and charged as required.  The charging
 operation consists of opening  the top of the furnace, lowering the
 charging  bucket part way into  the furnace and dropping the scrap  into
 the furnace.  The roof is moved back into position and the electrodes
 lowered through the roof for the start of the melting operation.
          The melting operation is started by turning on the electric
 power to  the electrodes.  Arcing occurs between the electrodes and scrap
 as the  electric current passes into and through the scrap.  When the
 scrap is  almost completely melted a second scrap charge is added, followed
 by a third or forth scrap charge.  The number of scrap charges depend on
 the apparent density of the scrap as it is charged to the furnace.  Re-
 fining  is accomplished by blowing high-purity oxygen into the molten steel
 to remove carbon and silicon.  This, combined with the refining action of
 the slag,  brings the heat of steel close to its required composition.
Ferroalloys are added to achieve the final composition,  power is shut off,
and the heat poured into a ladle for transfer to the ingot pouring area
or continuous casting machine
          There has been no reported work concerning the increase in
particle size or the reduction of emissions as applied  to electric-arc
furnace operation.   One process modification reportedly  does  result  in
the apparent reduction in fume, that of  using an oxygen-fuel  burner  for
preheating a more dense scrap charge to  eliminate successive  opening of
the furnace  to charge the less-dense scrap charges.

 5.4.1  Preheating and Melting with Oxygen-Fuel  Burners
           Most of the work on preheating  has  been  done with  respect  to
 preheating scrap prior to charging it  into  an electric furnace.   This
 type of operation is  considered  more favorable  since  it  reduces  the
 actual time required  to complete a heat of  steel in the  furnace.   Develop-
 ment work has been done on the use of  oxygen-fuel  burners  to preheat and
 melt scrap in the electric furnace prior  to the  application of electric
 power(57>58>59>60).  State of Development
           The required  equipment and operating practices have been developed
 to  the point  where they can be used in the  routine production of steel.
           Availability  to Industry.  There  are no  apparent restrictions
 on  the availability of  this technology, unless there may be  some  licensing
 fees connected with use of some  of the specialty burners designed  for
 the  process.
           Acceptance  by Industry.  One of the principle deterrents to
 the  use of oxygen fuel,  preheating and melting technology has  been the
 increase  in price  of  hydrocarbons.  A  second deterrent has been the high
 noise  level created by  the burners which are designed to obtain maximum
 efficiency of  heat output.  Degree of Effectiveness
           No  reports have  been published which contain data pertaining
 to measurements of particulates,  which would permit a good evaluation to
 be made of  the effects of oxygen-fuel burners on the reduction of
 particulates.  In  the discussion of a technical paper given on oxygen-
 fuel preheating, when the  speaker was questioned with respect to fume
occurence, made the following reply, "The amount of fume  appears to be
 less than normal.  There is, of course, an increase in waste gases.  In
one case where the burners were used before  the  power  was turned on,
the exhaust gas coming out of the electrode  ports was  almost clear.1
Two reports on the technology contained photographs to illustrate the
before and after situations   '    .


5.A.1.3  Envrionmental Effects
          From the meager amount of information reported it does not
appear that there would be any particular detrimental effects to the
environment from emissions.  However, as stated earlier there is a
problem in the noise created by the high-efficiency oxygen-fuel burners.
The gas and oxygen emerging from the tip of the burner are traveling at
supersonic velocity.  Although no reports have been made of the problem,
the high noise level situation has been verified in conversations with
operating plant personnel in past visits to electric-arc furnace steel

5.4.2  Electric-Arc Furnace--Scrap Charge Compatibility
          One of the factors tha t causes an increased amount of particulate
emissions is the number of times the furnace must be opened to charge the
amount of scrap required to meet the total weight requirements.  If a
furnace could be charged with a consistent, high-density scrap charge,
the furnace would need to be charged only once.  A method for producing
such high-density scrap charges for electric-arc furnaces has been
                   (f\"\ "4
developed in Japan  State of Development
          The technology involves the use of a special press to compact
low-density scrap into a single high-density scrap charge tha t conforms
to the control of the inner volume of the furnace.   Scrap charges from
5 to 60 tons have reportedly been produced.'

5.4,2.2  Availability to Industry
          The special presses for providing the customized scrap charges
are being marketed in the United  States by ai U.S.  equipment broker.
Unless there are some specialized design concepts  involved, similar
equipment could probably be designed and constructed by U.S.  press  builders.

                                  51  Acceptance by Industry
          There should be no hinderance to the acceptance of this technology,
unless the charging cranes in the electric furnace shop have insufficient
capacity to handle a complete furnace charge.  In such a situation, a
scrap crane of large capacity would have to be installed and the support-
ing crane structure would have to be strengthened.  The costs for such
modification of the crane and structures could rule out the use of special-
ized scrap compacting procedures.  Degree of Effectiveness
          No information has been published pertaining to the air pollution
aspects of this technology.

5.5  Metallurgical Coke Ovens
          Particulates originate in coke oven operations from three sources,
(1) charging of the ovens with coal, (2) pushing incandescent coke from
the oven, and (3) the quenching operation.  There are no published reports
on technology or process modification that pertain to the agglomeration
of particulate emissions from coke ovens, or a reduction in particulates.
There is only one report of record pertaining to the characterization of
particulates from coke ovens    .  Another report on the collection and
analysis of particulate emissions during the coke-oven charging operation,
is due for release in one and a half to two months^   .   Efforts to obtain
any information from the contractor were unsuccessful.

5.5.1  Particulates from Charging Coke Ovens
          Moist, pulverized coal is charged into coke ovens from larry
cars.  The ovens are at an incandescent heat and the initial amounts of
coal dropped into the ovens are heated very rapidly with a resulting
evolution of steam and volatile hydrocarbons.  The hot expanding gases
rise very rapidly and exit the oven through the charging hole at the
start of the charge, are suppressed during the flow of the bulk of the
coal and then may again exit through the charging port at the end of the

charging operation, before the lid can be replaced.  However, if the
coke-oven battery has double collecting mains with sufficient suction
in the steam aspirator system, the gases and particulates do not reach
the atmosphere and particulate emissions are minimized.  Larry cars with
integral scrubbers and other process modifications serve to improve the
efficiency of particulate collection, but they in no way cause an
agglomeration of particulates or a reduction in particulates from the
charging operation itself.

5.5.2  Particulates from Pushing Coke
          Particulates generated during the pushing operation occur
primarily from the abrasion of the coke on the refractory of the oven.
The particles are hard, angular carbon.  The characteristics of these
particles do not permit any agglomeration to occur.  When materials of
this nature are bonded together, such as in the manufacture of briquettes,
a petroleum base material or tar is used as the bonding agent.  Some
times very fine particulates may occur during the pushing operation.
These particulates originate from coal that is insufficiently coked.
It is often referred to as "green coke".  Extended coking time is a
suggested remedy for eliminating the fine particulates.  The extension
of coking time is not a remedy for these emissions.  Incomplete coking
is caused by insufficient heat transfer which is caused by warping of
the oven walls, in the case of older ovens, incomplete combustion in the
flues, and many other structural changes that may occur in an oven with
age, that affect the heat transfer characteristics of the ovens.   Main-
tenance procedures may reduce the amount of emissions, but this is
questionable.  Equipment is presently being installed on coke-oven
batteries to collect the emissions that occur during pushing.   It will be
some months before definite results will be available from these  install-
5.5.3  Particulate Emissions During Quenching
          The emissions occuring during the quenching operation are
essentially the same type of particulates generated during the pushing

operations.  They are carried into the atmosphere by the heavy clouds of
steam generated by the quench water spraying onto the incandescent coke.
These particulates can be minimized by the use of baffles in quench
towers, which trap the particulates and carry them into sumps for
recovery from the water.  Special quench cars, incorporating self-
contained water spray and emissions control devices, are in the process
of evaluation on full scale, operational equipment.

6.  Ferroalloy Furnaces

          Ferroalloys are used to make additions of alloying elements to
iron and steels, in order to achieve desired mechanical properties.  At
one time ferrosilicon and ferromanganese were both made in blast furnaces,
but today only ferromanganese is made in blast furnaces, and that at
only two plants in the United States.  All the other ferroalloys are
produced in submerged electric-arc furnaces or by alumino-thermic
          Considerable work has been reported in the past five years with
             .  .    -    £     ,,   f       (66,67,68,69,70,71,72,73,
respect to emissions from ferroalloy furnaces   >>>>>»>»
  '  '  '  '   .  Although the reports contain a great deal of informa-
tion pertaining to the chemical analysis of the particulates, some data
on size, and material describing the operation of the various types of
air pollution control equipment, there is an absence of data concerning
the properties of the emissions that may lead to the evolution of fewer
particulates or possibilities of agglomeration that would make collection
easier or more efficient.  One study did make a comparison between the
type of ore charged, i.e., lump ore, fine ore and pellets    .   Using
pelletized ore concentrates in comparison to fine ore concentrates the
amounts of particulate emissions was reduced approximately 42 percent,
while the use of lump ores reduced the amount of emission only 22
percent.  The extent of pelletizing ores in the United States is not
known; however, an economic evaluation would have to be made to compare
the economics of crushing, grinding and pelletizing ores, against the
cost of air pollution control without pelletizing.

7.  Process Modifications For Particulate Control
In The Primary Nonferrous Metallurgical Industry

          This technical review and discussion is limited to consideration
of the primary zinc, copper, and aluminum, industries.  It is concerned
only with possible changes in present procedures that will aid in control-
ling fine particulate emissions.

7.1  Zinc Roasting, Sintering, and Distillation

          Of the primary zinc plants in the United States, the only hori-
zontal retort plants presently operating (Amarillo, Texas, and Bartlesville,
Oklahoma) are doing so on variance for a limited time.  Since no changes in
equipment or operation are contemplated or likely for these, they will not
be considered in this review.  There are five plants remaining, then, for
consideration—three electrolytic and two pyrometallurgical.
          The trend in American primary zinc plant construction has been,
and is expected to continue to be, toward the hydrometallurgical or electro-
lytic process rather than toward pyrolytic processes.  Two new plants
definitely planned by American Smelting and Refining Company and by National
Zinc Company, respectively, will be electrolytic.  Also announcement has
been made recently by New Jersey Zinc Company of plans to build an electro-
lytic zinc plant.  This will give a preponderance of zinc production to
hydrometallurgical plants where control of fine particulates is a problem
only in roasting and handling.

7.1.1  Roasting

          Particulates in roasting zinc concentrates are bourne with the
gas stream through flues and in some cases over waste heat boilers to dust
collection and scrubbing systems.   In all plants under consideration the
gases go to an acid plant and are  well cleaned before passing to the atmos-
phere.   As a consequence, the roasting operation presents a very minor
factor in fine particulate control.


7.1.2  Sintering

          The two pyrometallurgical zinc plants sinter roasted calcine
and there is some stack loss of fine gas-bourne participates.  This is
minimized by water sprays and a dust collection system.  No new process
modification has been suggested or is presently being investigated to our
knowledge to reduce this small loss still further.

7.1.3  Reduction and Distillation

          In reducing and distilling zinc from the two pyrometallurgical
plants there is very little production of fine particulates that can escape
to the atmosphere.  Gases from the retorts carrying zinc metal vapor pass
through molten zinc or through a chamber filled with a rain of molten zinc
droplets to condense the zinc.  This is followed by a scrubbing system to
collect zinc oxide that may be formed.  Since there is no passage of gases
through the retorts, other than vaporizing of reduced zinc, there is
extremely small carry-over of charge; hence, about all of the particulates
that emerge from the retort-condensation system is zinc oxide which is too
valuable to lose.  Present methods of controlling loss of particulates
from the pyrometallurgical reduction furnaces are effective and no better
system for improvement appears to be emerging other than greater attention
to good plant housekeeping in handling dry concentrates and calcines.

7.2  Copper Roasting,. Matte Smelting, and. Converting

          There are two general methods under consideration for modifica-
tion of emitted solid particles in copper plants.  One is to modify the
usual reverberatory-converter procedure and equipment; the other is to
eliminate pyrometallurgical treatment entirely.  A brief review of conven-
tional U. S. primary smelter practice with data on the degree of efficiencies
realized with current control practices is given in Appendix A.

 7.2.1   Pyrometallurgical Modifications

          The  chief  smelter modifications presently  being used  to  reduce
 emission  of  particulates are  the  result  of  efforts  to  reduce or largely
 eliminate emission of  sulfur  oxides.  In copper  smelting elimination  of
 sulfur  gases is  of far greater  importance than reducing stack  losses  of
 particulates.  Consequently,  the  trend that  largely affects emission  of
 particulates is  that of controlling sulfur  gases.   All copper  smelters now
 have electrostatic precipitators  and/or  baghouse units, as well as  settling
 flues,  to recover solids,  since it is to their economic benefit to  avoid
 such losses.   However, to  capture  the sulfur, as by passing the gases
 through an acid  plant, greater  attention is  needed  to  clean the gas stream
 and fewer particulates escape to  the atmosphere.  This has meant  that
 efforts to raise the S02 content  of the  gases evolved  to permit effective
 acid recovery, or combining the reverberatory-converter operations, also
 aids in controlling particulate emissions.  Flash Smelting(79'8°'83)

          The  type of flash smelting developed at the  Outokumpu smelter
 in Finland (Fig. 7) or the variation developed by International Nickel
 Company in Canada (Fig. 8) is a favored  replacement method over the present
 reverberatory.  Flotation  concentrates,  with flux and  preheated air, are
 injected  into a hot chamber where the flash  burning of sulfides in suspension
 furnishes the energy needed for smelting.  This eliminates the need for
 coal, gas or oil burners with attendant  large volumes  of products of combus-
 tion.   In fact, with the Outokumpu flash furnace a  14  percent SO'  gas can
 be produced  (8 to 12 percent is more common which is 15 to 25 times more
 than a  reverberatory furnace using air-fuel combustion) which is excellent
 for making sulfuric acid.   The Inco variation is to use 95 percent oxygen
 in place of air which makes a still more concentrated S02 gas,  i.e., 75 to
80 percent.   This is used  in Canada to make liquid S02 for pulp and paper
plants  in the general vicinity.  'Aside from excellent utilization  of heat
 and diminished  gas  volume,  flash smelting is considered to give a  higher
grade matte  than reverberatory smelting.   This requires minimum oxidation

                                     Concentrate   Concentrate burner
                         OUTOKUMPU  FURNACE  produces 14%  SO2 otfgas, an
                         ideal gas grade for sulluric acid plants

                                            FIGURE  7.
          Sand  Chalcopyrite
            I    concentrate
   weight feeder
         Oxygen ^- — —
                                     Pyrrhotite, chalcopyrite
                                     concentrates and sand
                                                                                           	^  Oxygen
INCO FLASH FURNACE utilizes 95% oxygen for combustion of concentrates, produces 75% to 80% SO, offgas
                                             FIGURE 8.

in the converter.   Consequently,  a stationary converter can be used with
a more manageable  gas containment system.   Where there are outlets for
sulfuric acid or for liquid S02,  flash smelting is deservedly receiving
considerable attention.   A disadvantage of flash smelting is that both the
reverberatory and  converter slags must be  slow cooled, ground, and treated
by ore concentrations methods for recovery of copper values.
          An Outokumpu type flash smelting furnace has been estimated to
cost $1,600,000 (1973) and an electric furnace for slag treatment, $1,100,000.
The cost for an entire plant with waste heat boilers, dust collection
system, site preparation, etc., would be on the order of $9,000,000.
Generally, the capital cost for a flash smelting unit is considered to be
higher than for a conventional reverberatory furnace.  An Outokumpu flash
furnace is planned for the new Phelps Dodge smelter in Hidalgo County, New

                             (79,80)  Electric. Furnaces

          Where the cost of electric power is not excessive, the use of
electric furnaces is also a recognized and well-developed method of getting
a more concentrated SO™ gas stream suitable for making sulfuric acid.
Since there are no products of combustion for gas dilution, an S02 workable
range of 2 to 4 percent can be attained.  Usually this can be combined
with converter gases for the acid plant.
          Electric furnaces for copper smelting were  pioneered in Norway
and Sweden, and there is an excellent  installation  in Uganda.  In  the
United States the first commercial electric furnace for copper smelting
was started in early 1973 at Copperhill, Tennessee.   Another, by Inspiration
Consolidated Copper Company, was  started at their Arizona  plant, adjacent
to  their conventional reverberatory  smelter,  in November,  1973.  This unit
is  a 51,000 KVA furnace  (operating on  unroasted but thoroughly dried con-
centrates), five  large Hoboken-type  converters  and  a  Lurgi double  contact
acid plant.  Construction  is expected  to begin  in early 1974 at Anaconda,
Montana, on a fluid-bed roaster and  electric  furnace  installation  to
eventually replace the present reverberatory  furnace  operating on  unroasted

          An electric furnace that is  smelting roasted concentrates  is
considered  to have some advantages over a reverberatory  in  that  a higher
concentration of  S0_ is secured by the flash roaster--electric furnace
combination; an electric  furnace operating on raw dry concentrate also has
some advantages,  as a lower capital cost for the simpler construction of one
unit, but many factors affect the choice.  Both methods  tend  to  reduce
the amount  of particulates evolved to  the atmosphere by  producing less
volume of gas flow and a  gas that can  be utilized in an  acid  plant.  Continuous Smelting

          Combining the present reverberatory-converter  operations into
a single unit  has attracted much experimental work in the  past  few  years.
Such simplified equipment, making only one stream of gases  carrying  parti-
culates has the potential of greater control over particulate losses.  A
number of processes in this category have progressed to the pilot plant
stage and a few to initial commercial or semi-commercial installations.
          The Noranda Process.   '    Developed by Noranda Mines, Ltd. in
Canada, this process uses a single long (about 70 feet) combined smelting-
converting unit.  It is a horizontal, cylindrical, furnace having a central
depressed area for copper collection and a raised hearth at one  end as
shown in Fig. 9.  A burner heats the smelting end where concentrates and
flux are charged, but air or an air-oxygen mix is introduced  through sub-
merged tuyeres along the furnace base to oxidize the matte that  is formed.
The furnace can be tilted for access to the tuyeres.  Gases with accompanying
particulates may run 5 percent S0« or higher as oxygen enrichment above
25 percent is used.  Thus, there is only one unit to control, without the
fume and trouble of conveying molten material from one furnace to another,
and the exit gas is of sufficient S02 concentration for use in making acid
with attendant thorough cleaning of particulates.  After thorough testing
on a 100-ton per day pilot plant,  Noranda has built a $19,000,000 initial
commercial plant of 800 tons/day capacity which was started in 1973.

and flux
                                      Side view
                                                   1 -,     C
                           Nitrogen and SO
                                           	1  -^     I—
            Reducing gas
                   and converting
                                         White metal  '  Copper
                                     I    converting   I  settling
                                     I              I

              Slag reducing
              and settling
                                                                                • Reducing gas
     PROCESS produces rich off-gases to facilitate SO2 recovery-
                                             FIGURE 9.

Kennecot't has  also announced plans to  install a Noranda  type continuous
                             (85 )
converting  and  smelting unit.
                             /Q I p £\
          The  WORCRA Process.   *  '   Developed by Howard Worner  and Con-
ZincRio Tinto  of Australia, this process has also undergone considerable
pilot testing  over the past few years.  It combines  smelting, converting,
and  slag cleaning in one operation, making metal directly and continuously
in a long stationary furnace.   Slag is removed at one end; molten copper
at the other end.  Off gases in a single stream are  said to range from
9 to 12 percent SCL which can be efficiently handled in  an acid plant.
By utilizing the exothermic reactions  efficiently in burning the  sulfides
in the charge, a minimum of energy is  required.  No  tuyeres are used as
in the Noranda process, but air lances from the top  are used for  converting.
Although the process has been tested in pilot plants and semicommercial
plants of up to 72-80 tons of concentrates per day, no large commercial
plant has been constructed so far.  Some disadvantages may be the rather
low-grade blister copper produced which requires considerable fire refining,
and the probability of short refractory life because of the vigorous bath
          The Q-S Process.      Named  after the inventors, Paul Queneau
of Dartmouth and Reinhardt Schumann of Purdue, this  is also a multistage,
progressive converter operation that combines continuous smelting and
converting into one furnace.  Sulfide concentrates are flash smelted by
oxygen, and oxygen is also used through submerged tuyeres to effect conver-
ting to copper.  Dust production is said to be minimized by this  procedure,
but as the SCL rich gas is thoroughly cleaned anyway for recovery of sulfur
in some form, this is of minor importance from the standpoint of  particulate
          A slag scavenging operation is part of the process.   This involves
bottom blowing the slag with coal, oxygen,  and sulfur dioxide to  recover
residual copper, presumably as a matte that can be recirculated.  The
process can be applied to the recovery of nickel and lead also;  in fact,
the St.  Joe Minerals Corporation is said to have obtained an option for the
exclusive use of the process for replacing  sintering and  blast furnace
operations in smelting lead concentrates.   Pilot plant investigations of
the process are planned.

                                  (81  82  88
           The  Mitsubishi  Process.    *   '   '   '   Developed  in  Japan  by
 Mitsubishi Metals  Corporation,  this differs  from the Noranda  and  WORCRA
 continuous processes  in that  it is a  continuous  process  but the work is  not
 all  done  in a  single  unit.  Rather, the  concentrates are smelted  in one
 furnace,  the slag  and matte flow continuously through  a  slag  cleaning
 furnace to a converting furnace having overhead  air  lances, as shown in
 Fig.  10.   Since  these are  interconnected,  there  need not be the loss of
 fumes  that usually exist  in transferring from reverberatory to converter
 and  the continuous converting unit is not  cooled intermittently,  which
 favors better  refractory  life than with  the  usual  converter blows.   Out-
 standing  features  are said to be economy in  capital  investment (compared
 to a reverberatory-converter unit)  and operating costs,  and exit  gases of
 over 10 percent  SO- which  indicates a concentrated or  low-volume  gas stream.
 The  dusting rate is said  to be  low because of liquid particle entrapment
 in the furnace bath.  This should  further decrease loss  of particulates  in
 the  gas stream.  Charge preparation in drying to 1 percent moisture and
 instrumentation  control are considered to be  complex.  The process,  after
 thorough  testing in a pilot plant, has been  further demonstrated  in a
 1,500-ton blister  copper per month semi-commercial plant that was started
 in November, 1971.   Miscellaneous Pyrometallurgical Developments

           Although the trend in  building new  pyrometallurgical copper plants
 throughout  the world  has been toward flash smelting, electric furnaces
 and continuous smelting systems, no discussion of the  subject would  be
 complete  without at least mention  of other methods which may be considered
minor  or  presently not definitely proven to the  commercial stage.

          The Momoda  Blast Furnace.v  J  Successfully developed by  Sumitomo
Metal Mining Company, the Momoda blast furnace is being  used in two  copper
 smelters  in Japan,   It appears to be adaptable to, and have advantages for

    Returning by moving bucket system or air lifting
                                                                      Revert slag
                                              Converting furnace
                               Slag granulation
Schematic view of  Mitsubishi's semicommercial plant.
              FIGURE  10.

 small-scale operation, such as a capacity of 500 tons of charge per day
 per  furnace.  Since the energy requirement for smelting a ton of charge
 in a Momoda furnace is said to be  only about 28 percent that for reverbera-
 tory smelting with a wet charge, there is economy in operation, at least
 under Japanese conditions.  A feature of the process is that the concentrates
 are charged with other copper bearing materials as a stiff plasticized mass
 containing 10 to 15 percent water, rather than being sintered or briquetted.
 The process has been proven successful abroad but has not been tested in
 this country and would appear to have no direct benefit from the standpoint
 of greater particulate control.
          The U. S. Bureau of Mines.   ' Autogenous smelting work has been
 done on a laboratory scale and no  commercial development of the process
 appears to be pending, but this procedure has received considerable favorable
 attention.  Like the Noranda and WORCRA processes, it is a continuous
 smelting method in a single unit to produce copper directly from concentrates.
 Oxygen is used in place of air which gives a concentrated, low-volume, high
 SO- gas.  This should be favorable for minimum particulate loss since the
 gas would be thoroughly cleaned before going to a plant for sulfur recovery.
 The furnace combines flash smelting with converting by means of an oxygen
 lance immersed through the slag into the matte from the top.  A simplified
 sketch of this furnace is shown in Fig. 11.
                             /o i n I \
          The Kircet Process.   '     The Kircet Process is a development
 in the U.S.S.R. which has been demonstrated there on a pilot-plant scale.
 It consists of charging a dry copper concentrate into a cyclone type furnace
where an air-oxygen mixture (of up to 100 percent oxygen) affects oxidation
 and smelting similar to flash smelting.  An electric furnace attached to the
 cyclone smelting furnace receives  the molten product to complete the treat-
ment.  A second electric furnace receives matte from the main unit and
 produces blister copper and slag.  Since the process is designed to treat
combined lead-zinc-copper ores,  liquid zinc is recovered by condensation
 in the gas stream.   Incomplete descriptions give few details of how this is
done or how lead is recovered.   Although a saving is claimed in being able
to float lead-zinc-copper values from the complex ores,  the complexity of
recovering all three metals in a single furnace would seem to be disadvanta-
geous.  Off-gases from the system are said to run 70 to 85 percent S02

Bureau of Mines Autogenous Smelting
              Outer furnace  FeS3
         Slag overflow
                            White metal  Copper

                    Verticle.elevation section through center
                                           Copper overflow
                      FIGURE 11.

 (using oxygen in place  of air)  and only  a  small  amount  of  reducing material
 is needed.
          Other Process Modifications.   Among  general improvements of present
 installations,  the use  of oxygen  enriched  air  to reduce the volume of gas
 emissions,  and  attention  to  hooding converters to control  more efficiently
 gas and  dust  emissions, have  been common and need not be detailed.
          The top-blown rotary  converter as developed by International
                                                         (79 81 82)
 Nickel Company  has followed  oxygen steel-making  practice.   '  '  '  Although
 currently a commercial  installation is being made at Copper Cliff, Ontario,
 to treat a  mixed  nickel-copper  charge, this procedure is considered to be
 of potential  benefit  for  handling copper matte alone where low-cost oxygen
 is available.   There  are  metallurgical advantages to using such a converter,
 such  as  excellent mixing,  less  refractory wear,  and a higher temperature
 in final conversion to blister  copper, but for particulate control the only
 advantage is  in the reduced volume  of gases from using  oxygen.
          Improvements  in removal of sulfur from the gas stream in copper
 plants indirectly aids removal  of particulates since most  procedures require
 cleaning the  gas  stream more  thoroughly  than when it merely goes up the
 stack.   Thus, Monsanto1s  Cat-Ox process  which  has been  tested to remove S00
 from  power  plant  gases may be adaptable  to dilute  SO- gases presently
 evolving from copper  smelters.  The Wellman-Power Gas procedure uses "sodium
 sulfite  solution  to absorb S02» forming  sodium bisulfite,  then reversing
 the reaction  to release concentrated SO  '.'  As  with Monsanto1 s Cat-Ox process,
 the gas  stream  must be thoroughly cleaned before  reaching  the absorber;
 hence, particulates of all sizes  would be removed.  Allied Chemical Corpora-
 tion has been developing  another  procedure which  reduces a comparatively
 rich  S02 stream with natural gas  to produce elemental sulfur.    '  This
 could apply to  a  12 percent or over S0_ gas as produced by other methods,
 including the Wellman-Power or Monsanto Cat-Ox methods,  where  production
 of elemental  sulfur was much to be desired over producing acid.   Also,  the
Bureau of Mines has been operating a small pilot plant,  and extending their
 operations on a larger scale to a lead plant and coal-fired power plant on
removal of S07 from low-concentration  gas streams by absorption  in a sodium
citrate  solution.      Here again, of  course,  the gas  stream is  thoroughly
cleansed of particulates by passing through electric precipitators or

baghouses and being water scrubbed before reaching the absorber.  American
Smelting and Refining Company has had considerable experience  in recovering
liquid SC>2 from copper smelter fumes, and is building a  sizeable plant to
do this.W
          In place of absorbing SO- from dilute gases, limestone scrubbing
has received considerable attention.  The Smelter Control Research Assoc-
iation in pilot plant tests, have found the wet limestone scrubbing process
to be unreliable and have put emphasis instead on the ammonia  double-alkali
process for the removal of S02 from copper reverberatory furnace gas.

7.2.2  Hydrometallurgical Copper  Recovery

          The effort to avoid air contamination with S0_, and, to a lesser
extent, with dust, has led many copper companies to consider hydrometallur-
gical treatment.  Many different procedures have been investigated and
some have reached the commercial plant stage.  This is in addition to the
well-developed present technology of leaching oxidized ores or roasted
concentrates with sulfuric acid and recovering the copper by electrolysis
or precipitation on iron scrap.  Leaching fractured ores in place, heap
leaching, and treatment of mine waters with iron scrap to recover copper
are important procedures but beyond the scope of particulate control.
Improvements in using ion exchange or solvent extraction to isolate and
concentrate a desired metal from solution has enhanced the possibilities
of useful and economic recovery of metals by hydrometallurgy but such means
have only indirect effect on control of particulates.

                              (95)  Stanford LCPR Processv  '

          A lime-concentrate-pellet-roast procedure investigated by the
Process Metallurgy Group at Stanford University is noteworthy  in that it
combines a pyrometallurgical advancement with hydrometallurgical extraction
and recovery of copper.   The general plan,  as shown in the flowsheet of
Fig.  12, is to tie up the sulfur with lime during roasting, then leach the
calcine with spent sulfuric acid electrolyte for copper extraction and

                                  COPPER STRIPPING

                                  OR CEMENTATION
           Flowsheet (or the lime-conccntratc-pellet-roast
             process with direct electrowinning
                                            LIME OR
                            FIGURE 12.


electrowtnnlng.  Copper concentrates  are palletized with lime and roasted
at 400-600 C whereby at least 90 percent and in laboratory tests, about
98 percent, of the sulfur is retained as anhydrite (CaSO,).   Little fuel is
required, as the process is semi-autogenous, and the capital cost and
operating cost of a plant has been calculated to be favorable.  The reaction
of lime to retain sulfur in roasting is not new but the pelletizing and
roasting conditions to obtain maximum results are an advancement.  However,
this method has not yet been investigated on more than a laboratory scale.
          Since most of the sulfur (95+ percent claimed) is retained in
roasting and is discarded as a tailings residue after leaching, and possibly
cyaniding, there is no need to clean the gases from roasting.  Using a
traveling grate roasting operation as advocated, a large amount of fine
particles, including lime and anhydrite, would be expected to accompany
the gas stream.  A bag house or electrostatic precipitation unit undoubtedly
would be used.  This would collect most of the dust for recycling, but the
stack loss of  parttculates obviously would be greater than in a strictly
hydrometallurgical plant but probably not as much  as from a reverberatory-
converter plant that does not have an acid plant.  Anaconda's "Arbiter" Ammonia Leach  Process^96'

          This is an entirely hydrometallurgical  process  that has  gone
through  the  pilot plant  stage.  The  first commercial plant being built at
Anaconda, Montana,  is  scheduled to go into  production  in  September,  1974.
A general  flow sheet of  this plant is shown  in  Fig. 13.  The  only  particulate
control  involved  is in handling concentrates and  lime,  and in preventing
tailings  from  becoming windblown.  These  items  are expected  to be  under
good  control by recognized  operating methods.
          The  Arbiter  process,  as  it is often  called,  is  essentially a
continuous  ammonia  leach with oxygen under  only slight  pressure  for control,
followed by  solvent extraction  to  isolate and  concentrate the copper, electro*
lytic recovery of copper through  a closed cycle of stripping  the  loaded
organic  solvent with spent  electrolyte, and  disposal of ammonium sulfate.

rogen NH3


=5— -

r —

Feed Slurry .
""^ ""^ ** j


.«,. CCD 	 p. fV-TA-riQM 	 ,

it Gas To £
Y To Disposal
Strong Solution
1 ,
.LIQUID ION 	 ^ prTRnwiMWiMr 	 1
* i *
     Gypsum Slurry
      To Disposal
                         Spent Electrolyte
Copper Cathode
               FIGURE 13.

In the commercial plant provision  is made for a "lime boil" or hot treat-
ment of the ammonium sulfate solution after copper removal by addition of
lime with subsequent regeneration  of anmonia and production of a calcium
sulfate residue.  This can be effected also by adding lime directly to the
leaching tanks to regenerate amnonia directly and pass the calcium sulfate
into the concentrate tailings.  Alternate methods of ammonium sulfate
disposal *re to dry the solution and sell the crystals, to thermally reduce
the arrmonium sulfate to sulfur and recover ammonia, or to obtain bacterial
decomposition of the pmmonium sulfate solutions to produce ammonia and
elemental sulfur.  The process is  flexible in that comparatively low capa-
city plants can still be efficient and variations in operation can be made
readily to accommodate different concentrates or objectives.  Ammonia
leaching has definite advantages,  such as no solution of pyrite or iron
oxides and inexpensive materials of construction.  Sherritt:-Gordon Process

          From its long and successful experience in recovering nickel
end cobalt by pressure-leaching with amnonia, the entrance of Sherritt
Gordon technology into the copper  field would be expected.  However, the
direct application to copper concentrates of ammonia leaching as practiced
at the Fort Saskatchewan, Alberta, plant has not been considered to be
economically attractive.  Likewise, their procedure for precipitating metal
powders from solution by hydrogen under pressure has not been entirely
successful commercially as applied to copper.
          Announcement was made in 1971 that Sherritt Gordon and Cominco
would jointly explore and develop a hydrometallurgical process for recovery
of copper and other metals, and elemental sulfur, from sulfide concentrates.
A procedure mentioned,  one that has been patented by Sherritt Gordon,  is
pressure leaching with sulfuric acid.  This would appear to have an advantage
over the Arbiter process in recovering elemental sulfur directly,  but  pres-
sure leaching with acid would seem to be less desirable than near-atmosphere
leaching with ammonia.

7V2,2.4  Cymet Process

          The Cyprus Metallurgical Processes Corporation has developed a
ferric chloride leaching process for sulfide copper ores which is considered
to be essentially pollution free.  The only step that may involve loss of
particulates, other than general handling of concentrates, is the possible
need for grinding the concentrates to facilitate leaching; otherwise, all
operations are hydrometallurgical.  A flow sheet of the general process is
shown in Fig. 14.
          This process is rather complicated as it involves leaching with
ferric chloride anolyte recycled from diaphragm iron-chloride cells,
production of electrolytic iron, recovery of elemental sulfur, and recovery
of copper electrolytically as a powder which is subsequently electrorefined.
The many interdependent operations which add complexity make development
difficult, but a large scale $9,000,000 pilot plant for demonstration of
practicality and cost has been built.  Recovery of elemental sulfur and
of high grade iron are features of the process, and their value is expected
to aid substantially in reducing operating costs.  Duval Corporation and Other Processes

          Typical of other entirely hydrometallurgical processes for copper
recovery, the Duval Corporation announced several years ago the development
of a pollution-free process for treating copper sulfide concentrates.
This also involves a chloride leaching,  closed-cycle system with recovery
of elemental sulfur and electrolytic grade copper, but recovery of iron
oxide in place of electrolytic iron as done by Cymet.  A pilot plant was
planned and presumably built but details of results have not been noted.
          Other suggested procedures in recovering coppy hydrometallurgically
have been mentioned or briefly discussed in the technical literature in the
past few years.  These have largely been variations of the processes out-
lined.   °*      In all cases, since they depend entirely on hydrometallurgy,
their effect has been the same--little or no air pollution from particulates.

trVty^l   | Q- ET-0
                                                                       Schematic Ffowsheet"
                                                                           Cymef Process
                                                                  C>prus McUllurgical Process  CorpO'at°
                                                                         Los Angeles, California
                                   FIGURE 14.

 7.3  Aluminum Reduction Cells
           At present there are 14 U.S. producers operating 32 reduction
 plants which, in 1970, produced nearly 4 million tons of aluminum,
 46.6 percent of the world total'    .
           The processing of primary aluminum is essentially one of
 materials handling, the difference in the process modifications is in the
 type of reduction cells used.  There are three types, prebake (PB) cells
 which use prebaked carbon anodes, and two types of Soderberg cells which
 use single, baked-in-place anodes.  One type of Soderberg anode is held
 in place with horizontal studs (HSS) and the other with vertical studs
 (VSS).   Schematic drawings of each type are shown in Figures 15, 16 and 17.
           In plants using prebaked (PB) cells,  the anode baking furnace
 used in manufacture the anode can be a source of particulate and gaseous
 emissions.   However,  these may be controlled by electrostatic precipitators
 or more commonly by wet scrubbers.
           In operation,  the prebake  anodes in  the reduction cell produce
 less volitization of  pitch and fouling of collection system scrubbers
 than do the Soderberg types.   Another  advantage of  the  prebaked   pot is  that
 it is more  readily  enclosed without  interferring with process  operations.
 Prebaked   (PB) potlines accounted  for  59  percent of  U.S.  aluminum production.
          In the  course of  operation with horizontal stud Soderberg cells
 (HSS),  the  necessity  for  frequently changing  the studs  in turn makes
 necessary the frequent  removal  of  parts of  the  hoods with the consequent
 escape  of emissions.  There is  a further  disadvantage in  that the hooding
 system  does  not allow for  the burning of  organics or condensed tars;
 these unburned organics can cause  fouling of  the emission control system.
 HSS Soderberg units account for 25 percent of U.S. production.
          In the vertical Soderberg stud  (VSS)  pot, the anode compartment
 is stationary, so it permits the installation of a "skirt" around the
 base of the anode, and since the volume of gas  is small, carbon monoxide
 and hydrocarbons can be burned in integral gas burners as shown in Figure
 1? •  Since the skirt does not completely enclose the pot, additional
hooding is necessary to minimize fluoride emissions into the pot room.
v-S.s. Soderberg production units  account for 13 percent of  U.S.  aluminum

             Alumina Hopper
       Molten Crvolite
Segmented Doors


\\\ •   • -v '^
                                                                                        Control Svste:
                        Carbon Anode
                        Molten Aluminum

Alumina Hopper
 Carbon Anode
                        V-^L-V^-''-     ' •y//'   /.•/.'
                                                                                        To  Primary
                                                                                        Control System
                                                                                   Kood Door
                                                                                   Anode Studs
                                                                                  Molten Aluminum
                                        Molten Cryolite

Carbon Anode
 Cell Surface
 Molten Cryolite
   Molten Aluminum
                          Anode Studs
                                                                                        Control  Systen
                                                                                              Gas  and
                                                                                             Tar Burning


          There  is a continuous evolution of gaseous reaction products
 and  fume  from  the reduction cells consisting chiefly of carbon dioxide,
 carbon monoxide, sulfur oxides, gaseous and particulate fluorides,
 alumina and carbonaceous materials.  Hydrocarbon emissions from hot
 HSS  and VSS anodes can also result in a visible haze problem.  Table 1
 shows total industry-wide emissions in 1970, Table 2, the emissions
 from a typical uncontrolled pre-bake potline, and Table 3, the removal
 efficiency of  selected primary and secondary control systems.
 Primary control  systems involve the hooding and ducting for the indivi-
 dual cells; secondary control systems involve pot room collection
 systems located  in the roof of the cell room to remove fluorides which
 escape the primary system.
          In 1970, in the U.S., 75 percent of the plants had primary
 controls only, 15 percent had secondary controls only, 7 percent had
 both, and 3 percent had none at all(   '.
          Collection efficiencies approaching 100 percent are not
 possible under the present state of the technology (see Table 3).
 The best achievable appears to be 95 percent delivered to the primary
 system, with 5 percent going to the secondary roof monitors.    '

 7.3.1  New Aluminum Reduction Processes
          There are a number of new aluminum reduction processes under-
 going development.   Among these are Alcoa's chloride electrolysis
process (the ASP process), Alcan1s sub-halide process, and the Applied
Aluminum Research Corporation^ (manganese reduction of aluminum chloride)
process^   '     .  Furthest along is Alcoa's chloride electrolysis
process.     .   In this,  alumina is reacted with chlorine in a reactor
to produce aluminum chloride which in turn is electrolyzed in a closed
cell separating the compound into aluminum and  chlorine.   The separated
chlorine is recycled continuously to the  reactor in a closed loop
system.   The process has the advantages of up to 30 percent less energy
consumption, and presumably less dusting  since  there are no fluorides,
the chief effluent in the conventional Hall process which causes the

             1970  (1,2)

          Total Huorino
          Gaseous fluorides
          Fluorine in particulars
          Total solids
                               Emissions (Ions)
 Pot-  Bake
rooms  plants
 650  23,850
 600  10,800
  50  13,050
4,200  57,000
                 PREBAKE POTLINE  (1,2)

         "F" as gaseous fluorides
         F as solid fluoride's
         Total F
         Total solids
         Ibslton Al
              CONTROLLED SYSTEMS  (1,2)
              Control System

           Coated filter dry scrubber
           Fluid bed dry scrubber
           Inieetod alumina dry
           Wet scrubber • wet ESP"
           Dry ESP • wet scrubber
           Floating bed
           Spray screen
           Bubbler scrubber  r
             wel ESP
Fluoride removal
Dlliciencles. %

99 >
             1 Electrostatic precipit.itor.

industries a most serious air pollution problem, and no working of
finely powdered Al-0  into the bath.  Also, the requirement for cryolite
(Na.Al F,) is circumvented at a time when this mineral has become
   3    o
scarce and costly.  There have been no claims as to the degree of re-
duction in dusting except for that implied in the Alcoa statement that
the shift from the Hall process to the ASP process, eliminates both
the need for cryolite and the expense of containing cryolite emissions"
          A number of the other new processes have as their goal, the
utilization of alternative mineral sources of aluminum such as laterite,
kaolin clays, and alunite      .   The implementation of these processes
does not affect materially the amount and type of emissions from the reduction

 (1)  "Particulate Polycyclic Organic Matter, National Academy of Sciences,
      Committee on Biologic Effcts of Atmospheric Pollutants (1972).
 (2)  "Abatement of Particulate Emissions From Stationary Sources",
      National Research Council - National Academy of Engineering, COPAC-5,
      Washington, D. C. (1972).
 (3)  Office of Coal Research, Press Release, July 3, 1973.
 (4)  "Evaluation of Coal-Gasification Technology: Part I--Pipeline-
      Quality Gas", National Academy of Sciences, National Adademy of
      Engineering (December, 1972).
 (5)  "Evaluation of Coal-Gasification Technology:  Part II--Low-BTU
      Fuel Gas", National Academy of Science, National Academy of Engineer-
      ing (to be published in 1973).

 (6)  "Clean Energy from Coal—A National Priority", Amnual Report for
      Calendar Year 1972,  Office of Coal Research, U.S. Department of
 (7)  Karnavas, J. A., LaRosa, P. J., and Pelczarski, E. A., "Ttoo Stage
      Coal Combustion Process", Chemical Engineering Progress,  Vol. 69,
      No. 3, p 54 (March,  1973).

 (8)  Chementator, Chemical Engineering, August 6, 1973, page 34.
 (9)  Personal communication with representatives of Atomics International
      Division,North American Rockwell.

(10)  Cover, A. E., Schreiner, W. C., and Skaperdas, G. T., "Kellogg1s Coal
      Gasification Process", Chemical Engineering Progress, Vol. 69, No. 3,
      p 31-36 (March, 1973).
(11)  Oldenkamp, R. D.,'and McKenzie, D. E., "The Molten Carbonate Process
      for Control of Sulfur Oxide Emissions", APCA paper 68-183 (1968).

(12)  Reid,  W. T., "Superslagging Combustor", Battelle-Columbus internal
      memorandum to D. W.  Locklin confirming disclosure of concept
      (April 2, 1973).
(13)  Squires, A.  M., "Clean Power from Coal", Science, 169 (3948)
      (August 28,  1970), p 821.
(14)  Jimeson, R.  M., and  Shaver, R. G., "Credits Applicable to Solvent
      Refined Coal for Pollution Control Evaluations",  p 25C,  presented
      at the Symposium on  Synthetic Hydrocarbon Fuels from Western Coals,
      AIChE Meeting,  Denver, Colorado,  August 30, 1970.

(15)  Martin, G.  B.,  Pershing,  D. W.,  and Berkau, E.  E., "Effects of Fuel
      Additives on Air Pollutant Emissions from Distillate-Oil-Fired
      Furnaces",  U.  S. Environmental Protection Agency  (June,  1971).

 (16)  Andrews,  R. L.,  Siegmund, C. W., and Levine, D. G., "Effect of Flue
      Gas  Recirculatlon on Emissions  from Heating Oil Combustion",
      presented at  61st Annual Meeting, APCA, St. Paul  (June 23-27, 1968),
      41 pp.

 (17)  Bienstock, D., Amsler, R. L., and Bauer, E. R., Jr., "Formation of
      Oxides of Nitrogen in Pulverized Coal Combustion", Jour. APCA, 16^
      (8),  (August,  1966), p 442-445.

 (18)  Burroughs, L.  C., "Air Pollution by Oil Burners Measurable but
      Insignificant", Fuel Oil and Oil Heat, 22  (6), (June, 1963), p 43-46.

 (19)  "HEW Issues Guidelines to Control Pollution; Tests Show Distillate
      Within Limits", Fuel Oil and Oil Heat, 28  (3), (March, 1969), p 57.

 (20)  Kroshel, C. F., "Improving Performance of Domestic Heating Oil Equip-
      ment  Througti  Installation and Servicing Procedures", presented
      at the 40th Annual Convention, NOFI, Chicago,  p 26, (April 12,
 (21)  Ornig, A. A.,  Schwartz, C. H., and Smith, J. F., "A Study of the
      Minor Products of Coal Combustion", Paper No,  64-PWR-4, presented,at
      National Power Conference, Tulsa, Oklahoma (September 27-October 1,
      1964), 8 pp.

 (22)  Ornig, A. A., Smith,  J. F.,  and Schwartz, C. H., "Minor Products of
      Combustion in Large Coal-Fired Steam Generators", Paper No. 64-WA/FU-2,
      presented at Winter Annual Meeting, ASME, New York (November 29-
      December 4, 1964), .12 pp.

 (23)  Diehl, E. K., DuBreil,  F., and Glenn,  R.  A., "Polynuclear Hydrocarbon
      Emission from Coal-Fired Installations",  Trans. ASME,  Series A,
      Jour.  Engineering for Power, 89 (2),  (April, 1967),  p 276-282.

(24)  NCRR Bulletin; (Summer,  1973),  Vol.  Ill,  No. 3, National Center  for
      Resource Recovery, Inc.,  1211 Connecticut Avenue,  N.W.,  Washington,
      D. C., 20036.

(25)  Niessen,  W.  R., "Systems  Study of Air  Pollution from Municipal Incin-
      erators,  Vol.  1,  Report prepared by Arthur D.  Little,  Inc.  (March,
      1970), under  Contract CPA-22-69-23.

(26)  Hazard,  H. R., Trip report covering visit to Combustion Power Company,
      Inc., on February 6,  1974.

(27)  Miller,  P. D., et al.,  "Corrosion Studies in Municipal  Incinerators",
      SW-72-3-3, EPA Solid  Waste Research Laboratory, National  Environmental
      Research Center (1972).

(28)  Environmental  Science and Technology (April,  1971), Pyrolysis  of
      Refuse Gains  Ground,  p 310-312.

(29)  Franklin,  W.  E.,  Bendersky, D.,  Shannon, L.  J., and Park, W.  R.,
      "Resources Recovery,  Catalog  of  Processes,"   Final Report,  Project
      No. 3634-D, Prepared  by Midwest  Research  Institute for the  Council
      on Environmental  Quality  (February,  1973).

      Two  Battelle  reports  that served as  general references  for  the stationary
      combustion and municipal incineration sections of this report are:

      A.  The Federal R&D Plan for Air-Pollution Control by Combustion-
          Process Modification, PB-198-066, prepared for EPA (January 11,
      B.  Working Paper on Factors Affecting the Future of the Coal Industry
          in the United States (April 1, 1973).  Not for Distribution
          Outside Battelle.
(30)  Rengstorff, G. W. P., "Formation and Suppression of Emissions From
      Steelmaking Processes, " AIME Open Hearth Proceedings, 44. 120-143 (1961).

(31)  Rengstorff, G. W. P., "Factors Controlling Emissions From Steelmaking
      Processes," AIME Open Hearth Proceedings, £5, 204-219 (1962).
(32)  Turkdogan, E.  T., Grieveson, P., and Darken, L. S., "Mechanism of the
      Formation of Iron Oxide Fumes", AIME Open Hearth Proceedings, 45.
      470-490 (1962).
(33)  Bates, R. E.,  "Fume Formation", Journal of the Iron and Steel Institute
      201 (9), 747-751, (September, 1963).
(34)  Morris, J. P., Riott, J. P., and Illig, E. G., "A New Look at the
      Cause of Fuming", Journal of Metals, J.8 (7), 803-810 (July, 1966).
(35)  Rossi, G., and Perin, A., "Some Notes on Brown Fume Powders",
      Journal of the Iron and Steel Institute, 207 (10), 1365-1368, (October,
(36)  Ellis, A. F.,  and Glover, J., "Mechanism of Fume Formation in Oxygen
      Steelmaking",  Journal of the Iron and Steel Institute, 209 (8),
      593-599, (August, 1971).
(37)  Rengstorff, G. W. P., "Role of Methane and Other Factors in Controlling
      Emissions from Steelmaking Processes", AIME Open Hearth Proceedings,
      4j6, 438-452, (1963).
(38)  Savard, G., and Campbell, J. C., "Suppression of Iron Oxide Fumes
      in the Open Hearth, Furnace", 33/The Magazine of Metals Producing, J>
      (2), 61-70, (February, 1967).
(39)  Kocho, et. al., "Iron and Steel Refining—USSR Patent No. 236499",
      Soviet Bulletin of Discoveries, Inventions, Models and Trade Marks
      (March 2, 1969).
(40)  Naydek, V. L., et. al., "Treatment of Metal—USSR Patent No.
      276116", Soviet Bulletin of Discoveries, Inventions, Models and Trade
      Marks, (July 14, 1970).
(41)  Kiyanskiy, D., "Liquid Oxygen Blast", Rabochaya Gazeta,  45, (4912),
      3  (February 22, 1973).
(42)  "Q-BOP;  From  Blow to Go in 90 Days", Journal of Metals, 2A  (3),
      31-37, (March, 1972).
(43)  Chatterjee, A., "The  New Oxygen Steelmaking Process,"  Iron and Steel
      International, .46  (5), 440-448 (October, 1973).

 (44)  Knuppel, H.,  et. al., "Method of Refining Pig-iron  Into  Steel",
      U.S.  Patent Number 3,706,549, Patent Gazette, 905.  567  (1972).

 (45)  "U.S. Steel to Substitute Q-BOP for Open Hearths at Fairfield",
      Metalworking News, 8  (December 20, 1971).

 (46)  "Q-BOPs for Gary", Journal of Metals, 2A (4), 8  (April,  1972).
 (47)  "U.S. Steel Managers  Rave Over Q-BOP", American Metal Market, 1
      (May  25, 1973).
 (48)  "Development  in the Iron and Steel Industry During  1973,"  Iron
      and Steel Engineer,..51  (1), D24 (January, 1974).

 (49)  Hubbard, H. N., Jr., and Lankford, W. T., Jr., "Development and
      Operation of the Q-BOP Process in the U. S. Steel Corporation",
      Iron  and Steel Engineer, j>0 (10), 37-43 (October, 1973).
 (50)  "Bottom-blown Steel Processes Now Number Three: Q-BOP, LWS, and
      SIP", 33 Magazine, 12 (9), 34-38, (September, 1972).

 (51)  "Kaiser Suing J&L Over L-D Rights", Metalworking News, 12
      (November 27, 1967).

 (52)  "Oxygen Furnace Patent Fight to Continue",  American Metal Market,
      2 (March 10, 1969).
 (53)  "High Court Bars Oxygen Steel Patent", Metalworking News, 16
      (March 10, 1969).
 (54)  Butler, P., "EOF Patent Fight Flames in Court Today", Metalworking
      News, 11 (September 27, 1971).

 (55)  "J&L  to Appeal Court Decision Charging Patent Infringement",
      American Metal Market, 81 (23),  4 (February 1, 1974).
 (56)  "Q-BOP: Year II",  Journal of Metals,  25 (3), 37 (March, 1973).
 (57)  Hinds, G.  W.,  and  Hodge, G.  W.,  "Use of Oxygen-Fuel Gas Burners for
      Scrap Meltdown in  Electric Furnace",  AIME Electric Furnace Proceedings,
      J.7 290-298 (1959).

 (58)  Howard, V. J., "Auxiliary Meltdown Toarch",  AIME Electric Furnace
      Proceedings, .17, *19 (1959).
 (59)  Howard,  V.  J., "Use of the Oxygen Gas Burner for Scrap Meltdown in
      the  Small  Arc  Furnaces", AIME Electric Furnace Proceedings,  18.  398-405
(60)  Spenceley,  G.  D.,  and  Williams,  D.  I.  T., "Fumeless" Refining with
      Oxy-fuel Burners",  Steel Times,  150-158,  (July,  1966).

(61)  Grobel,  E.  A.,  and Maselle,  A.  J., "Blowing with Oxygen-Natural Gas
      for  Decarburization and  Fume  Suppression", AIME Electric Furnace
      Conference,  .27,  67-70  (1969).
(62)  Discussion to: Hinds,  G. W.,  and  Hodge,  G. W., "Use  of  Oxygen-Fuel
      Gas  Burners  for  Scrap  Meltdown  in  Electric Furnace",  AIME Electric
      Furnace Proceedings, JJ, 300  (1959).

 (63)   "New  Scrap Press  Reduces  Electric  Furnace  Operating Cost",  Iron and
       Steel Engineer, 43  (8),  146,  (August,  1966).

 (64)   Herrick,  R.  A., and  Benedict,  L. G., "A Microscopic Gassification  of
       Settled  Particulates Found  in  the  Vicinity of  a  Coke-Making Operation",
       Journal  of the Air  Pollution Control Association, .19 (5), 325-328
       (May,  1969).

 (65)   Telephone Conversation with Mr.  Robert Bee, Mitre Corporation in
       reference to EPA  Contract Number 68-02-0290.

 (66)   Ferrari,  R.,  "Experiences in Developing an Effective Pollution Control
       Systems  for  a Submerged Arc Ferroalloy Furnace Operation",  Journal
       of Metals, 20, 95-104, (April,  1968).

 (67)   Scott, J.  W., "Design of  a  35,000  KW High  Carbon Ferrochrome  Furnace
       Equipped  with an  Electrostatic  Precipitator", AIME  Electric Furnace
       Proceedings, .29,  80-82,  (1971).

 (68)   Young, J.  H., and Singer, D. H., "Manufacture of Low  Carbon Ferro-
       chromium  at  the Steubenville Plant, Foote  Mineral Company", AIME
       Electric  Furnace  Proceedings, .29,  83-87 (1971).

 (69)   "Ferroalloy Fume  Collection From Submerged Arc Electric Furnace
       Operation",  Industrial Heating, 38. (1), 86 (January, 1971).

 (70)   Fegan, G.  J., "Cleaning Ferroalloy Furnace Fume  with  High Energy
       Scrubbers", AIME Electric Furnace  Proceedings, ,30,  65-68 (1972).

 (71)  Meredith, W.  R., "Operation of a Baghouse  Collecting  Silica Fume",
      AIME Electric Furnace Proceedings, .30, 69-71 (1972).

 (72)   Sherman,  P. R., and  Springman, E.  R., "Operating Problems with High
      Energy Wet Scrubbers on Submerged Arc Furnaces", AIME Electric Furnace
      Proceedings .30, 72-76 (1972).

 (73)  Rentz, 0., Siebert,  G.,  and Stracke, R.,  "Reducing Fume Emissions by
      Improving Furnace Operation, by Feed Pretreatment",  AIME Electric
      Furnace Proceedings, 30,  77-83 (1972).

 (74)  Lopuszynski, T.  W.,  Trunzo,  J.  P.,  and Wllbern, W. L., "Design and
      Operation of a 45 MW 50  Percent Ferrosilicon Furnace", AIME Electric
      Furnace Proceedings, 30,  89-93 (1972).

 (75)  "Development Document for Proposed Effluent Limitations Guidelines
      and New Source Performance Standards for the Smelting and Slag
      Processing Segment of the Ferroalloy Industry", U. S. Environmental
      Protection Agency, Contract  No. EPA 440/1-73/008 (August,  1973).

(76)  Jenkins,  R. D.,  "Potential Utilization and  Disposal  of Particulate
      Materials Captured From a Silicon Metal Furance", Preprint  AIME
      Electric  Furnace  Conference, 10 pp  (December,  1973).

(77)  Killin, A. M., "Progress  Report Air Pollution  Control Study  of the
      Ferroalloy Industry", Preprint1 AIME Electric Furnace Conference,
      9 pp (December,  1973).

(78)  McClure,  R. N., "Disposal of Submerged  Arc  Furnace Fume", Preprint
      AIME Electric Furnace Conference,  6 pp  (December, 1973).

(79)  "Copper Smelting Today: The State of the Art", Chemical Engineering,
      Engineering and Mining J., Special Section, pp. p-z, (March, 1973).

(80)  Lane White, "The Newer Technology: Where it is Used and Why",
      Chemical Engineering, E&MJ, Special Section pp AA-CC, (March, 1973).

(81)  Price, F.- C.,"Copper Technology on the Move", Chemical Engineering,
      E&MJ, Special Section, pp RR-DDD, (March, 1973).

(82)  Shoemaker, R. S,,"Minerals Processing in 1973", Mining Congress J.,
      pp 24-29, (February, 1974).

(83)  Holderreed, F. L., "Copper Smelting", Mining Engineering, p 45,
      (September, 1971).
(84)  Themelis, N. J., McKerrow, G.  C., Tarassoff, P., and Hallett, G. D.,
      "The Noranda Process", J. of Metals, pp 25-32, (April, 1972).

(85)  American Metal Market, p. 19,  (March 26, 1974).

(86)  "What's Happening in Copper Metallurgy", E&MJ, pp 75-79, (February,
(87)  "Form Consortium to Exploit New QS Process", J. of Metals, p. 12,
      (March, 1974).
(88)  "Mitsubishi's Continuous Copper Smelting Process Goes on Stream",
      E&MJ, pp 66-68,  (August, 1972).

(89)  Suzuki, T., and Nagano, T., "Development of New Continuous  Copper
      Smelting Process", Paper given at Tokyo Meeting of AIME, May 27,
(90)  Worthington, R.  B., "Autogenous Smelting of Copper Sulfide Concen-
      trate", U. S.  Bureau of Mines, Report of Investigation 7705 (1973).

(91)  Quarm, T. A. A., "Copper Smelting with a Cyclone Furnace", E&MJ,
      pp 80-82, (October, 1969).

(92)  Hunter, William D., and Michener, A.  W., "New Elemental Sulfur
      Recovery System Establishes Ability to Handle Roaster Gases",
      E&MJ, pp 117-120 (June, 1973).

(93)  George, D. R., Crocker, L., and Rosenbaum,  J. B.,  "The Recovery of
      Elemental Sulfur from Base Metal Smelters", Mining Engineering,
      PP 75-77 (January, 1970).

(94)  "ASARCO Building $16 Million Liquid S02 Plant", E&MJ, p 26,
      (September,  1972).
(95)  Bartlett, R. W., and Haung, H. H., "The Lime-Concentrate-Pellet
      Roast Process  for Treating Copper Sulfide Concentrates",  J.  of
      Metals,  pp 28-34,  (December,  1973).
(96)  Arbiter, N., "Anaconda's Ammonia Leach Process",  Paper presented
      at the Dallas  Meeting of AIME, February, 1974.


 (97)  Krusi, P. R., Allen, E. S., and Lake, J. L., "Cymet Process--
       Hydrometallurgical Conversion of Base Metal Sulfides to Pure
       Metals", CIM Transactions, Volume 76, pp 93-99 (1973).

 (98)  Kruesi, P. R., Allen, E. S., and Lake,  J. L., "Inventor,
       Developers Explain How New Cymet Process Works", Pay Dirt, pp 4-10,
       (October 23, 1972).

 (99)  "Duval Claims Development of 'Polution Free1  Hydrometallurgical
       Copper Refining Process", E&MJ, p 171, (September, 1970).

(100)  Malouf, E. E., "Current Copper Leaching Practices", Mining
       Engineering, pp 58-60 (August, 1972).
(101)  Beal, John, "Copper in the U.S.--A Position Survey", Mining Engineering,
       pp 35-47, (April, 1973).

(102)  Iversen, R, E.,"Air Pollution in the Aluminum Industry, Journal of
       Metals, V. 25, No. 1, pp 19-23 (January, 1973).

(103)  Gerard, Gary, Editor; "Extractive Metallurgy of Aluminum, Volume 2,
       Aluminum", Interscience Publishers,  New York, p 572, (1963).

(104)  Anon., "New Smelting Process, Alcoa  Prepares the Site for a Pilot
       Plant", American Metal Market,  p.  18, (September 18, 1973).

(105)  Piccolo, L., Chirga, M., and Calcagno, B., "Gas-solid Reactor:
       Effects of Chemical and Fluodynamic  Parameters on Alumina Chlorination
       Process", Chimie ET Industrie-Genie  Chimique, Vol. 101, No. 19,
       pp 2485-2489, (November, 1971).

(106)  Bureau of Mines, U. S. Department of Interior, "Aluminum from
       Domestic Sources, a Miniplant to Evaluate Alumina Recovery Processes",
       pp.  85.
(107)  Hardesty, D. R., and Weinberg,  F.  J., "Electrical Control of Particulate
       Pollutants from Flames" 14th International Symposium on Combustion,
       Combustion Institute,  p 907, (1973)


                              APPENDIX A
                       COPPER SMELTING PRACTICE

          U.S. copper ore reserves are predominately sulfides rather
 than oxides with a low overall copper content (1% or less).  Concentra-
 tion and  separation from other mineral values and gangue by grinding and
 flotation yields concentrates containing 15 to 35 percent copper. Con-
 centrates with a high sulfur content, those that are relatively high in
 iron sulfide, and those containing certain volatile impurities such as
 arsenic,  antimony, and selenium require a preliminary roasting step.
 Where roasters can be by-passed, air drying, kiln drying, or drying in
 a roaster may be practiced.  In other respects, smelting practice at the
 15 U.S. primary copper smelters (Table A-l) until very recently, has
 followed  the universal technology of matte smelting, converting, and
 refining  shown in Figure A-l.  By the way of introducing new modifications
 in pyrometallurgical practice and new process schemes  which by-pass the
 smelting  step entirely, we shall describe briefly what has been, up to
 the present, conventional smelting practice.

          Conventional U.S. Primary Copper-Smelting Practice

          The three principal processing steps in conventional U.S.
 smelting  practice are roasting,  matte-smelting,  and converting.   We
 shall describe each in turn.
          As explained above, the need for roasting depends on the sulfur
content of the copper concentrates.   The conventional  oxidizing  roast
oxidizes  some of the sulfur content  of the concentrate and  converts part
of the iron sulfide to iron and  sulfur oxides.

                                      IN THE  U.S.
State and Location
San Manuel
White Pine
Great Falls
New Jersey
Perth Amboy
Perth Amboy
New Mexico
New York
El Paso

Ins piration
Fhelps Dodge
Phelpi Dodge
Phelps Dodge
White Pine
Cities Service
1972 Annual
103 kkg
— —
1C3 Short
White Pine(0
Pbelps Dodge
1 Phelps Dodge
I Phelps Dodgelfl
103 kkg
10* Short
(a) Data are (or 1971 from Yearbook of the American Bureau of Metal Statistics published in 1972.
(b) Data supplied by company.
(c) Capacity of production.
(d) Reference (IV.I).
(e) Reported as fire refined production.
(f) Fire refinery.

                Copper bearing material
     (Blended concentrate* + direct •melting ore, etc. )
         (Sand, gravel, low grade siliceous ore)


                   Furaace charge

           _ I
Multiple hearth roaster.
(partial roast)
It air'
                                                       (3 to 5% SO2)
          -i    «- i  •      I Flue   i Cold slaa
          If   tC*lc""«  Tdust   n matte     Gas
              	T	•	1   rift
    Reverberate ry
                                                      . |  Cottrell
                                                       I  precipltator.
      (1%SO2)  J	— — — — -.-        ;
     	I .J  Wa.te heat  1
             ~H  boil*"
                            W Slag to dump
                              (37% SiO2 0. 5% Cu)       r —-—•*•"
               Matte                                  J  Cottrell      J_
               (30% copper)                            j_^precipitator.  t
                         iwoia   .
                         'lV«   V
Converter .lag
26% Si02
J     "i-__*J   "H
                  Scrubber   j
                                                         Hot        I
                                                     *    cottrelU    I

                                                     .     Plant
                  Fire refining
                                             H2SO4   I    Wet
                                                     I    cottrell
           Anode, to electrolytic refinery
     (gold, silver, selenium, tellurium recovery)

           In current conventional practice roasting may be carried out
 either  in multiple-hearth or fluid-bed roasters.
           The multiple-hearth roaster is a cylindrical, brick-lined
 vessel  divided from top to bottom by horizontal brick hearths.  Each
 hearth  has either one or several drop holes located alternately on the
 inner and outer peripheries of successive hearths.  A central, rotating,
 brick-lined steel column extends vertically through the center of the
 roaster.   On each hearth, arms equipped with rabbles are fixed to the
 rotating central shaft.  Feed is dropped onto the top drying hearth near
 the central shaft and rabbled to the outside of the hearth, where it
 falls through the drop holes to the hearth below.  The rabbles on this
 hearth  push the feed toward the central column, the feed drops to the
 next hearth, and so forth until feed exists at the bottom hearth.
          Air introduced at the bottom of the roaster passes up through
 the heated chambers, and the oxygen in the air stream reacts with the
 iron and sulfur in the feed to liberate heat, which sustains the roaster's
 hearth  temperature.  The gases leaving the top of the roaster contain from
 2 percent to 6 percent SO --rather low for sulfuric acid manufacture—and
 carry away about 6 percent of the roaster calcine product.  The coarser
 calcine product rabbled from the bottom hearth is comprised of roasted
 copper  concentrates and flux and totals 94 percent of total calcine
 produced.  Sulfur elimination in the roaster is governed by regulation of
 air flows and charge retention time on the hearths.
          The fluid-bed roasting process is characterized by a gas-solid
 reaction in a dense suspension of solids maintained in a turbulent mass
 by the upward flow of gases that affect the reaction.   The roaster is
 essentially a cylindrical refractory-lined steel shell used to contain
 the suspended solids.
          Air is forced into the roaster through tuyeres in a refractory-
 lined steel distribution plate that is placed at the bottom of the shell.
 The two best-known types of fluid-bed roasters,  the Lurgi and the Dorr-
Oliver,  are characterized by different tuyere design.

           The use of fluid-bed  roasting installations  is  not  yet  as
 common in copper smelting as  it is  in zinc,  nickel,  iron  calcine,  and
 sulfuric acid production units.   This is partly because the copper
 industry is reluctant to wipe out the large  investment already made
 in existing multiple-hearth roasters  and partly because of difficulties
 experienced in fluid-bed operation, shut downs  caused  by  sintering of  the
 roaster bed, and heavy sulfate  carryover to  the electrostatic precipi-
           Another possible drawback is  the excess  calcine carryover of
 80 percent in outlet gases, as  compared  with 6  percent for multiple-
 hearth roasters.   A  more elaborate dust-handling system is therefore
 required with fluid-bed  units.
           There  are  also some advantageous features  to fluid-bed  roasting.
 There  are no moving  parts in  the  combustion  chamber, and  maintenance is
 simplified  accordingly.   The vigorous gas currents existing in the fluid
 bed maintain very uniform bed temperature and composition.  And the
 roasting action  is so  vigorous  that little excess air  is  required,
 permitting  S02 contents  of 12 percent to  14  percent  to be obtained.

 Matte  Smelting in Conventional
 Reverberating Furnaces
          Matte  smelting is done  in a reverberatory  furnace (Figure A-2).
 These  are  large  shallow-hearthed  structures of up to 40-meters (130 feet)
 long,  12-meters  (38  feet) wide, with the capability of  treating up  to
 1450-metric  tons  (1600-short tons) of charge per day.  The objective of
 this treatment is to collect virtually all of the copper  in a molten
 copper-iron-sulfide  "matte" layer which  is tapped from the furnace for
 subsequent treatment in converters.
          The reverberatory furnace  has remained the workhorse of the
copper smelting industry for 80  to 90  years and is  only now being
challenged by newer furnaces  that emit a more concentrated exhaust gas.
          In conventional operation,  calcine  from the roaster  hoppers,
flues, and electrostatic precipitators is gravity charged  into hoppers
staggered along each side of  the furnace and  located  above drop holes
in the roof (Figure A-2).  Heat  is supplied by coal or  gas burners
located at one end of the furnace with about  (4,000,000 Btu) required per

      i" " .i-• »»«•"••» -.    Fettling drag "•          Fettling pipet  .'™t,':~ •,'
      •:  '..'I" --'V-'  *  conv-yor v -•-X-':..^,. ; .  ,  , lrf~*:^^....:'>^
               \-**Hfe  •                               ——^y***

    Air and
                  'Nrj^^SS       ^-.c,     N-—-rr?^
                  /   jT*f^\              Slag''
„  .  ii^v.-.'^i        ,      .  v    ••;-*"•??" '                 ^     ••••-•,
           BumenX         ^Matte
     ::.J5^i^«;^iaft^^»:rv ^•-i:f^?. ^- v.1. ;«_,> " ' :f^Slii8^::''vl•'•
                                                                                   '  $jji

                                                                                             ' -•'•


 ton of charge  smelted when  the  charge  is hot  roasted material.   Thermal
 efficiency of  the  reverberatory furnace is  low; however,  furnace
 installations  are  usually equipped with waste-heat  boilers, which
 recover much of  the  heat of the combustion  gases  in the form of  super-
 heated steam.
           Anywhere from 11  percent to  16 percent  of the sulfur content
 in a calcine charge, and up to  20 percent of  the  sulfur content  in an
 unroasted  charge,  is liberated  during  smelting in a reverberatory
 furnace and is mixed with the products of combustion to give an  outlet
 gas usually containing between  0.5 percent  and 1.0  percent SO  .  Gases
 pass through waste-heat boilers and electrostatic precipitators  and are
 then vented to the atmosphere.   Considerable  experimental work is being
 done using limestone scrubbers  to remove small percentages of S0« from
 stack gases, but the problems involved in treating  upward of 3-million cubic
 meters (100-million cubic feet)  of gas per  furnace  per day are formidable.
           Generally, slag is side-tapped near the exhaust end of the
 furnace  into locomotive-drawn slag pots of  about  6-cubic meters  (200-cubic
 feet)  capacity.  Slag may be either granulated or dumped molten.
           Matte is tapped into  ladles, each containing about 16  metric
 tons  (18 short tons), which are  drawn down  the matte tunnel to the
 converter  aisle by winch and then charged by cranes into one of  the
 waiting  Peirce-Smith Converters.

           Converting is a batch  process and is done in large, horizontal
 cylindrical Peirce-Smith side-blown converter furnaces up to 9 meters
 (30  feet)  in length and about 4 meters (13 feet)  in diameter, with a
 centrally  located aperture for charging, unloading and,.exit of gases.
 They are also fitted with numerous tuyeres along  their length near the
 bottom through which the air necessary for converting is blown.  The
 converters are mounted on trunnions so that they may be tilted.  During
 the converting operation,  the furnace is run in an upright position with
 the aperture directly beneath a hood  through which the converting gases
are exhausted.   The furnace is tilted for  charging,  discharging,  and

           In  operation,  the  converter converts matte  from the reverberatory
 furnace  to an impure  form of copper called "blister copper"  (so called
 from the appearance of  its cast  surface which shows evidences of gas
           It  is  a  two-stage oxidation process involving high through-puts
 of  air,  850 cubic meters (30,000 cubic ft) per minute, through the molten
           In  the first  stage, the iron sulfide component of the matte
 is  oxidized to sulfur dioxide and iron oxide, leaving nearly all of the
 copper as  molten copper sulfide or "white metal".  Sulfur dioxide
 leaves the converter as a gas and may be processed to sulfuric acid.
 Fumes and  dusts which may contain lead, bismuth, arsenic, etc.,
 produced by converting are collected and sent to lead smelters.  The
 iron oxide formed during converting reacts with silica, added as a flux
 to  the converting operation, to form a molten iron silicate slag which is
 removed  and returned to the reverberatory furnace for reprocessing since
 it  contains significant concentrations of copper.
           In  the second stage, after removal of the slag, blowing is
 continued  until virtually all of the remaining sulfur is oxidized and
 removed as sulfur dioxide.  The converted copper is normally transferred
while still molten to the refining furnace.  Blister copper at this stage
 is  still relatively impure; it contains varying amounts of heavy metals,
arsenic, some sulfur, and all of the precious metal constituents originally
present  in the concentrates.
          A feature of the Peirce-Smith converters is that they can be
modified into a kind of smelting furnace; oxygen is mixed with blowing
air to increase the heat available from exothermic converting reactions.
          The chief disadvantage of the Peirce-Smith converter is the
relatively low concentration, 2 percent to 6 percent SO  outlet gas
produced when excessive air is allowed to infiltrate into the off-take
hood over  the converter mouth.
          In conventional primary copper smelter practice,  converters
 produce more than half of  the  SO.  emissions, the converter  can produce

 S02 gas emissions containing in excess of 6 percent SCK.  Gas concentra-
 tions are not steady, since the conversion of matte to blister copper
 is a batch process.   However, with several converters, these irregular
 streams can be averaged out for sulfuric acid manufacture by regulating
 blowing schedules, with good results.   Most of the sulfuric acid from
 smelters is made from converter gas.
 Current Emission Control Practice in Conventional
 U.S.  Primary Copper Roasting,  Matte-Smelting and Converting Operations
            The type of emission control devices used in U.S.
 primary copper smelter operations,  with their control  efficiencies,  are
 shown in Table A-2.
           Seven of the 15  U.S.  smelters have roasters;  all are equipped
 with  emission control devices.   Smelters equipped with  fluid-bed roasters
 are controlled with acid plants,  those with multihearth roasters are
 controlled  with electrostatic  precipitators (ESP's).  Control  efficiencies
 of  the  acid plants were  99.5 percent,  the  ESP's  handling emissions from
 the multihearth roasters had efficiencies  of 90-90.5 percent.   Particulate
 emissions  from the roaster  are  in the  form of fine dust and  fume, in
 many  cases,  with  a particle size  of less than 1  micron.   Roaster particu-
 late  emissions will  contain, in addition to  10 percent  copper  and a
 large amount of  silica and  alumina from the  gangue, quantities of the
more volatile  concentrate constituents  such  as zinc, cadmium,  antimony,
 and arsenic  in about  the same amounts  as that  in  the concentrate.
          Most of  the reverberatory furnaces are  equipped with either
 ESP's or cyclones  to clean the reverberatory  smelter gases.  As of 1972,
 four were not  equipped with emission control devices of any kind.  Particle
 sizes of the dust and fume emissions from the reverberatory furnace  varied
 from less than 0.1 micron to greater than 5 microns. "In an ESP controlled
reverberatory, 70 percent of the stack emissions were less than 0,5  micron,
with 34 percent between 0.1 and 0.3 micron.  Compositions of typical
emissions from reverberatory stacks are given in Table  A-3.
          All except three  of the converters listed in  Table A-2 have
some type of particulate emission  control, these may be  ESP, acid plant,
scrubber or multicyclone, alone or in combination.  Raw  emission factors

                     TABLE A-2  CONTROL PRACTICE IN  COPPER SMELTING (1972)
JJew Mexico
Size of
Feed, T/da
(19% Cu)

Control Device 7.
Balloon flues-long
brick flue-multi-
stack " 70-85
Cyclone —
ESP-stack 95
Multicyclone-stack 85
Balloon flues-ESP-'
stack*** 45
ESP-wet scrubber-
ESP-drying tower
acid plant —
Cyclones-gas cooler-



300      Reverberatory

         (F-B, D-0)
450      Roasting  (Mh)

  ESP-acid plant

Gas cooling-balloon
Balloon flue-ESP-
  acid plant               —
None                       0
None                       0

Brink's mist eliminator
Acid plant

ESP-stack             •    98
                                                          ESP-stack                 98
                                                             phere or acid plant

                      (Continued )
Size of Operation
; Copper
Concentrate Produced,
,state Feed, T/da T/da
^ington(h) 300



Control Device
Cyclone-water spray-
acid plant

Refining (electro-

ill«essee(b) 300 -57

'chigan & 220 (c)

2°na 60,000 T
(ore) -450

*°na 350

!**s(e) 350
(T Cu)

(fluid bed)

_ id )
furnace ...
(d }


Scrubber-acid plant

Settling chamber
Scrubber-acid plant

ESP or cyclone










— —






Size of Operation
Concentrate Produced,
Feed, T/da T/da Operation
670 -200 Reverberatory
Refining and
anode furnace
2800 Roasters

Control Device



None _


j 	 ,



— —

(a)   Estimated (EPA)  particulate emissions/day of 1.2 T/da; 2000 T (concentrate) capacity
     (29-31% Cu).

(b)   New copper system planned.

(c)   Stack losses  "1.5 T/day of particulate dust; single stack for plant.

(d)   Particulate emissions 14.8 T/day at 35 percent Cu.

(e)   Plan to build acid plant to treat converter gases.

(f)   0.7-T dust emitted/day; additional emission controls being constructed.

(g)   Ore mined at  a rate of 230,000 T/day.

(h)   Total stack emissions—253 Ibs/hr.

*    ESP - Electrostatic precipitator.
***   New system planned to increase efficiency to 95+%.  ESP-wet scrubber-packed tower
     system has given 99.5% efficiency.

Element, %
Reverberatorv Stack
._ 0.0102
(1)  New Mexico.

(2)  Arizona.

(3)  Nevada.

for the uncontrolled converters have been estimated to be between 22-24 kg
per metric ton (45-90 pounds/short ton).   Typical reported compositions
of emissions from converter operations are given in Table A-4.   Character-
istics of the emissions from the converter stack of the second  smelter in
the listing in Table A-2 are given in Table A-5.  This converter was
controlled with a multicyclone.
          In addition to the emissions noted above, there is an additional
problem involved in transfer operations in the batch operations of a
conventional smelter.  The new continuous process described in the
accompanying report would eliminate this.

            Composition of Converter Emissions,
                   Percent, Metal Basis


Gas  stream characteristics

--volume flow rate:  160, 000 SCFM @60°F and 750 mm Hg
  (upper limit) and 120, 000 SCFM @60°F and 750 mm Hg
  (lower limit)
  Condition assumes three converters in operation, two in
  the stack and one out.
--temperature:  550°F
--pressure drop in system:  1. 75" w. c.
--gas color: pale white--transparent
--odor: pungent,  irritating
--composition volumetric:

      S02    :    3. 0%
      H20    :    3. 0%
      CO2    :    0. 5%
      02     :   19,4%
      N2     :   74. 1%

Particle characteristics

--grain loading:  0. 7 grains/ft
--size analysis of dust:

      Size microns^           Percent

         0-5                    7
         5-10                   5
        10-15                   3
        15-20                   2
        20-25                   1
        25-30                   1
        30--                  81

                                 TECHNICAL REPORT DATA
                           (Please read JnXructions on the reverse before completing)
                                3. RECIPIENT'S ACCESSION>NO.
 Process Modifications for Control of Particulate
    Emissions from Stationary Combustion,
    Incineration. and Metals
                                5. REPORT DATE
                                October 1974
                                6. PERFORMING ORGANIZATION CODE
 	rfril *-* -*~J"rf A MiM AWAJI •  MiiiM JHW ilfciilfcrf            	^^^_^^_
 7.AUTHOR(S)R> Nekervis, J.  Pilcher, J. Varga Jr. ,

 B. Gonser, and J. Hallowell
                                8. PERFORMING ORGANIZATION REPORT NO
 Battelle, Columbus Laboratories
 505 King Avenue
 Columbus, Ohio 43201
                                10. PROGRAM ELEMENT NO.
                                1ABQ12; ROAP 21ADK-017
                                11. CONTRACT/GRANT NO.
                                 68-02-1323 (Task 9)
 EPA, Office of Research and Development
 NERC-RTP, Control Systems Laboratory
 Research Triangle Park, NC 27711
                                13. TYPE OF REPORT AND PERIOD COVERED
                                Final; 3/74-7/74 	
                                14. SPONSORING AGENCY CODE
              rep0rf summarizes the state of process modifications relative to the
 control of fine particulate emissions from stationary combustion sources (electric
 utilities and industrial processes); municipal incinerators; iron and steel plants;
 ferro-alloy plants; and non-ferrous metal smelters (zinc plants, copper smelters,
 and aluminum reduction cells). This study is to uncover modifications to conventional
 or new unconventional practices which appear to improve the control of fine particu-
 late emissions in these five areas.  Modifications to conventional stationary combus-
 tion sources  considered include ash-fluxing, SOS addition to flue gas, staged combus-
 tion, use of fuel additives, fly ash agglomeration, solvent refining, and flue gas
 recirculation. Unconventional systems studied include fluid-bed, coal gasification,
 and submerged combustion. For incinerators, combined fuel/refuse firing, gas cool-
 ing, and pyrolysis methods are considered.  For iron and steel plants, emphasis is
 given to the bottom-blowing oxygen process  (Q-BOP).  Modification of the conventional
 reverberatory smelting procedure and the introduction of hydrometallurgical methods
 are discussed for copper; the AI-CI electrolytic (ASP) process is considered for alu-
 minum. The  stage of development,  availability or acceptability by  industry, emis-
 sion reduction efficiency, and environmental impact of each process is considered.
                              KEY WORDS AND DOCUMENT ANALYSIS
                    b.lDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Group
 Air Pollution
 Electric Utilities
 Industrial Processes
 Iron and Steel Industry
Zinc Industry
Copper Converters
Aluminum Industry
Air Pollution Control
Stationary Sources
Fine Particulate
Process Modifications


                    19. SECURITY CLAS
         CLASS (This Report)
                                                                    21. NO. OF PAGES
                    20. SECURITY CLASS (Thispage)
                                             22. PRICE
EPA Form 2220-1 (9-73).