July 1975
Environmental Protection Technology Series

                    OF          .1.LOY  FURNACE


            A  STUDY

       C. E. Mob ley and A. O. Hoffman

        Battelle Columbus Laboratories
             505 King Avenue
           Columbus, Ohio 43201
      Contract No. 68-02-1323 (Task 18)
           ROAP No. 21AUY-011
        Program Element No. 1AB015
      EPA Project Officer: R.D.Rovang

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

         WASHINGTON, D. C, 20460

               July 1975

                       EPA REVIEW NOTICE

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

Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series.  These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields.  These series are:






          9.  MISCELLANEOUS

This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series.  This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution.  This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 2Z161.

                Publication No. EPA-650/2-75-063

                           TABLE OF CONTENTS



      Subtask 1:  Describe, in General,  Ferroalloy Furnace
      Designs and How These Designs Vary According to the
      Product Being Produced	    6

      Subtask 2:  Define The Parameters  Which Must Be
      Considered When Switching Products in Both Open
      and Sealed Ferroalloy Furnaces	12
      Subtask 3:  Identify and Compare General Procedures for
      Changing From One Product to Another in Open and Sealed
      Ferroalloy Furnaces 	   26
      Subtask 4:  Identify Deviations from Optimum Conditions
      and Other Penalties (Including Economics) Which Are In-
      volved in Switching Products in Open and Sealed Furnaces.  .   29

      Subtask 5:  Do Case Studies of the Product Families
      Which are Most Frequently Interchanged Based on Typical
      Open and Sealed Furnaces	33

      Subtask 6:  Draw Conclusions, Where Possible, on the
      Limitations to Flexibility Which Would Be Encountered
      in the Domestic Ferroalloys Industry if all Future
      Capacity Expansion Were to Consist of Sealed Furnaces ...   35

      Subtask 7:  Make Recommendations for Additional
      Research, Development, and Possible Demonstration
      Programs Which Could be Undertaken by EPA to Aid in
      the Solution of the Product Flexibility Problem
      Identified in Subtask 6	40
            Battelle's Recommendations Relative to Open and
            Sealed Ferroalloy Furnaces	41

            Battelle's Suggestions for Possible Research
            Projects	43
            Possible Research Projects	44

      References	51

                             LIST OF TABLES

Table 1.  Furnace Design and Operation Data for
          Selected Ferroalloy	   18
                            LIST OF FIGURES

Figure 1.  Schematic Illustration of Electric-Arc Ferroalloy
Figure 2.  K-Factor Versus Electrode Powder Density for
           Selected Ferroalloys	   13

Figure 3.  Curves Relating Furnace Loads, Voltages, "K"
           Factors, and Electrode Currents 	   15

Figure 4.  Furnace Load Versus Electrode Diameter	   24


          This program was undertaken to provide the U. S.  Environmental
Protection Agency (EPA) with insight into (1) the extent to which the U. S.
ferroalloy industry uses furnace-product flexibility, (2) the effect of
ferroalloy-furnace design (specifically a comparison of open versus totally
sealed furnaces) on product flexibility, and (3) research and development
areas which could enhance product flexibility in sealed furnaces and/or
make sealed ferroalloy furnaces more practical in the U. S. industry.
          With one exception, all ferroalloy furnaces are of the "open"
type in the United States.  Elsewhere, particularly in Japan, many ferro-
alloy furnaces are of the "sealed" type.  An EPA study has indicated that
the sealed ferroalloy furnaces offer the potential to be operated with
lower pollution of the atmosphere than the other types of furnaces.  In
light of this advantage, the question has been raised as to why sealed
ferroalloy furnaces are not presently utilized to a greater degree in the
United States.  Reservations about the "flexibility" of sealed furnaces
relative to open furnaces have been one of the primary deterrents cited
concerning the lack of adoption of sealed furnaces in  the United States.
          Flexibility of a ferroalloy furnace is the term selected to
describe the ability to change the furnace from  the production of one type
or  family of ferroalloys  to another family requiring different smelting
conditions.  Furnaces with high flexibility  can  be used  to manufacture a
wide variety of  ferroalloys, usually with only short changeover periods
and/or low loss  of production  occurring during product  switches.  Furnaces
with low flexibility are  generally designed  to produce  ferroalloys under
conditions that  are optimum for a specific  type  or family.   Low flexibility
furnaces do not  readily  lend themselves  to  product conversions, as product
changes in these  furnaces may  incur significant  penalties  in technology,
changeover time,  and costs.
          Based  upon an  extensive literature review  and  numerous  interviews
with  representatives of  ferroalloy  companies, it is  concluded  that  the
U.  S.  ferroalloy industry has  used  and  is  likely to  continue utilizing
high  furnace-product flexibility to remain competitive in  a  changing
domestic  and  international marketplace.
           In  comparing the  product  flexibility  of open and  sealed furnaces,
 it is  concluded  that  sealed  ferroalloy furnaces are  less flexible than

comparable, open-hooded furnaces, because  (1) it has not yet been demon-
strated feasible to smelt all the common ferroalloys in sealed furnaces,
and (2) a sealed-furnace operation generally requires greater charge
preparation and attention to operating details than a comparable, open-
hooded furnace on the same product.  U. S. ferroalloy producers are
reluctant to install sealed furnaces until either (1) a long-term market
demand is assured for that particular sealed-furnace product and/or  (2)
it has been demonstrated feasible to smelt essentially all the ferro-
alloys, specifically the high-silicon-content ferroalloys, under sealed-
furnace conditions.
          Battelle investigators recommend that EPA (1) conduct a cross-
media assessment of the overall (i.e., air, water, solid waste, etc.) en-
vironmental characteristics of open-hooded and sealed ferroalloy furnaces
to define future policy statements, and (2) assist in evaluating the design
and performance of baghouses used with ferroalloy furnaces with the ob-
jective of securing methods for lowering the annual total emissions from
existing and future ferroalloy furnaces.
          The investigators also formulated two suggestions for additional
programs aimed at increased utilization of sealed ferroalloy furnaces and/or
lower furnace emissions from U. S. operations.  These suggested programs
involve (1) determination of whether the substitution of iron-ore pellets
(as used in Japan) for ferrous scrap (as used in the U. S.) decreases
the need for stoking and improves the ease of operation in the smelting
of 50 and 75 percent ferrosilicon alloys, and (2) research directed at
insuring that sealed furnaces are safe operations through the development
of improved furnace-monitoring techniques and devices.



          There are three basic types of electric-arc ferroalloy furnaces
with respect to the design and operation of the furnace offgas system.
These three types are the open, the semiclosed, and the totally closed
furnace.  The totally closed furnace is herein referred to as the "sealed"
furnace.  In the operation of the open ferroalloy furnaces, the carbon
monoxide by-product released during smelting is burned above the mix and
the products of combustion (and the large volume of diluting/cooling ex-
cess air) are then cleaned to remove particulate matter.  In the semi-
closed and sealed furnace operations, a cover is placed over the fur-
naces to collect the carbon monoxide-rich offgas, without combustion.
This is done to clean a minimum amount of gas per ton of product and/or
to collect the carbon monoxide for its fuel value.  While both the semi-
closed and sealed furnaces are equipped with covers to exclude air from
the furnace interior, they differ in the design of the seals between the
furnace cover and the movable-electrode columns.  In the semiclosed fur-
nace, the mix being charged is used to achieve a quasi-seal between the
furnace cover and the electrodes.   More positive  seals are utilized
with the sealed furnaces.  The sealing performance of mix seals on semi-
closed furnaces is not considered satisfactory by furnace operators or
the EPA^l' and it is-, therefore, anticipated that new U. S. ferroalloy
furnaces will be either of the open or sealed design.
          Based upon U. S. Environmental Protection Agency (EPA) studies
of the emissions associated with ferroalloy furnaces, it has been re-
ported that sealed furnaces equipped with appropriate air-pollution^cpntrpl	
devices release significantly less particulate emissions than open-hooded fur-
                                              (1  2}
naces with their air-pollution-control devices^ '  '.  EPA has emphasized
the advantages (from air-pollution and energy standpoints) of sealed
furnaces compared with open-hooded furnaces and has considered setting air stan-
dards which would encourage the use of sealed furnaces'2' •*).  u. S.
ferroalloy producers have expressed reluctance to install sealed fur-
naces, indicating that the use of sealed furnaces would lower the producers'
product flexibility (i.e., their ability to change from one ferroalloy
type to another in one furnace).
(1)  References are listed on page 51

          To give a greater insight into the issue of furnace-product
flexibility, EPA, under a basic ordering agreement, requested that Battelle's
Columbus Laboratories (BCL) conduct a study of ferroalloy furnace-product
flexibility.  The specific objectives of the BCL study were
          (1)  To compare the flexibility of open and totally
               closed (sealed) ferroalloy furnaces as it re-
               lates and applies to the U. S. ferroalloy in-
               dustry, so that the significance of product
               flexibility can be ascertained
          (2)  To identify existing technology and applied
               engineering practices which can be used to
               achieve additional flexibility in sealed fur-
               naces without further research and development
          (3)  To develop, where warranted, recommendations for an
               additional program which could be undertaken by
               EPA to (a) make sealed furnaces practical in the
               domestic industry if they are not at the present
               time and/or  (b) develop or demonstrate techniques
               or systems  to allow both ferroalloy product flexi-
               bility and  lower  furnace emissions  from U. S.
               ferroalloy  furnaces.
          To achieve  the  above objectives,  seven subtasks  were
formulated  by EPA for which responses were  to be prepared by BCL.
          The subtask statements, as presented in  the Work Statement,
          (1)  "Describe,  in general, ferroalloy furnace designs
               and how these designs vary according to the product
               being  produced.
          (2)  Define the  parameters vhich must  be considered when
               switching products in both open and totally enclosed
               ferroalloy  furnaces.
          (3)  Identify and compare general  procedures for changing
               from  one product  family  to another  in  open and  totally
               enclosed  ferroalloy  furnaces  (and semienclosed  fur-
               naces, if  these appear relevant).

          (4)  Identify deviations from optimum conditions and other
               penalties (including economics) which are involved in
               switching products in open and totally enclosed
          (5)  Do case studies of the product families which are most
               frequently interchanged based on typical open and
               totally enclosed (sealed) ferroalloy furnaces.
          (6)  Draw conclusions, where possible, on the limitations
               to flexibility which would be encountered in the do-
               mestic ferroalloy industry if all future capacity ex-
               pansion were to consist of totally enclosed furnaces.
          (7)  Make recommendations for additional research, develop-
               ment, and possible demonstration programs which could
               be undertaken by EPA to aid in the solution of the
               product flexibility problem identified in Subtask 6. "
The Battelle-prepared responses for each subtask inquiry are given sequentially
in the following text.  Following the last subtask response (No. 7), a summary
and conclusion section is presented*
          As part of  this study, Mitsubishi Research Institute, Inc.  (MRI),
on a subcontract  basis  conducted a  study of sealed  ferroalloy  furnace
operjitions in Japan.  A series of questions concerning Japanese sealed
ferroalloy furn.ice operations was formulated by BCL personnel and pre-
sented to MRI for answering.  The complete MRI report (with the question-
answer format) is presented in the Apnendix.  Because most of the sealed
furnaces in the world are in Japan, the MRI report was used to give BCL
personnel additional  insight on the flexibility of  sealed furnaces.
          Prior  to  addressing  the Subtask questions and  responses,  it is
appropriate  to define what  is  meant in  this report  by "ferroalloys" and
to clarify the intended meaning of  furnace  flexibility.  Listings  of  ferro-
alloys and their  corresponding chemical analyses are available  in  the
literature^1 ' ^'.  For  the  purposes of  this study,  primary atten-
tion is directed  at  those ferroalloys produced in the greatest quantities,
that is the manganese,  si]icon, and chromium families of alloys, including
standard ferromannanese  (Std FcMn) , silicomanganeue (SiMn) , ferrosilicon

(50 FeSi and 75 FeSi) , silicon metal (Si), ferrochronrium (FcCr) , and
ferrochrome-silicon  (FeCrSi)*.  Other  ferroalloys, such as  the rare-earth
silicides and calcium silicide, etc.,  are introduced and included as
specific entries in the presentation of the flexibility issue.
          Flexibility is defined as the "capability of responding
or conforming to changing or new situations"^.  Furnace flexibility
herein refers to the capability for changing or adjusting particular
ferroalloy furnaces to produce different ferroalloy products.  It is
important to note that a time-frame is not inherently defined as part
of the "flexibility" statement; no qualifying statement as  to the
length of time required to respond to  the changing situation is included
in the flexibility definition.
          Based upon our review of the literature and on interviews with
ferroalloy-industry representatives, It appears advantageous to consider
at least two types of flexibility, that is, short-term and  long-term,
when discussing ferroalloy furnace flexibility.  Short-term flexibility
as presented here, refers to the capability of carrying out product
changes in a given furnace within or over a period of several months.  Long-
term flexibility relates to the capability and/or practice  of effect-
ing a product change in a furnace on an annual or longer-term basis.
These time-frame generalizations to the flexibility statement are not
absolute, but are introduced as relative guidelines to simplify the dis-
cussion of ferroalloy-furnace flexibility.
          The U. S. ferroalloy industry utilizes both short-term and lonjt-
term furnace flexibility.  For example, one ferroalloy plant visited
during this study produces a variety of speciality ferroalloy products
and typically produces about ten different alloys in a given
furnace per year.  While this example  of short-term furnace flexi-
bility may represent the most product  changes in a given furnace, other
*  The  cited group of ferroalloys accounts for about  75 percent of  the
   aporoximately 2 million  tons of ferroalloys produced in the U. S.
   ferroalloy  furnaces annually.

ferroalloy plants also utilize relatively small open furnaces to produce a
variety of specialty products on a similar short-term flexibility basis.
          As indicated in several communications between EPA and the
U. S. ferroalloy industry, practically all (if not all)  of the U. S.
ferroalloy companies have used and expect to continue utilizing long-
term furnace fJexibility to remain economically competitive in a
changing product-demand marketplace.   Examples of U. S. ferro-
nlloy company's long-term furnace flexibility practice are presented
in the "Statement on Proposed Air Quality Standards for High Carbon
Ferromangane.se" prepared by the Ferroalloys Association^'.

                    SUB TASK  STATEMENTS  AND  RljSPONSKS

    Subtask  1;  Describe,  in General. Ferroalloy  Furnace  Do si RUS  atid
     How These Designs Vary  According to  the  Product  being Produced

          Design details for ferroalloy furnaces  are  presented  in several
                                                           f 6— Q ^
texts  on the  theory and practices  of electric-arc  smelting   ~   .
While  there  are numerous minor design variations,  most  3-electrode, 3-
phase  electric-arc furnaces  are quite similar in  their  basic design.  Thu
basic  smelting furnace consists of a cylindrical  furnace  shell, the interior
of which is  lined with insulating  bricks  and  a carbon hearth.   Electrical
energy is continuously supplied to the  reaction area  through three
vertically suspended prebaked carbon or self-baking  (i.e., Soderberg)
electrodes arranged in a delta formation.   The mix material  is  batch fed  to
the top of the descending  bed of charge material  and  the  reaction products
(i.e.,  ferroalloys and slags) are  periodically or, in some cases,  con-
tinuously tapped from the  side of  the furnace.  A  schematic  illustration
of a submerged-arc ferroalloy furnace is  shown in  Figure  1.
          The upper portion  of the furnace-charge  materials may be
thought of as a particle bed, with the  hot  carbon  monoxide offgas  as-
cending through and heating  the descending mix components.   The basic
reaction which occurs within the furnace  reaction  zone  Is the high-
temperature reduction of an  oxide with  carbon to yield  the ferroalloy plus
carbon monoxide.  Large quantities of carbon  monoxide are formed  as part
of the  smelting reaction for all the ferroalloy smelting  operations.  The
thermochemistry of the various ferroalloy smelting operations is des-
cribed by Elyutin^9).
          While generally  quite similar in design, there  are several furnace-
design  features which are  associated with specific ferroalloys.   For example,
while  the circular (or near circular) furnace shell with  three electrodes in
a triangular configuration is the predominate furnace of  the industry, some
companies prefer the so-called "packet" furnace with  a  rectangular crucible
with in-line electrodes for  the production  of silicon metal  .  Silicon
metal  is also produced in  the 3-electrode,  circular furnace common to
other  alloys.  Because the comparison of packet versus  circular or in-line versus
   Various in-line electrode furnaces have been proposed for smelting
   particular ferroalloys(10).

                3       (in Delta  Formation)
            ELECTRODES  "
                                S~T- TAP HOLE
       (Third electrode is not shown.)

delta electrode configuration is not a major issue in the comparison of
open versus sealed furnaces, no further consideration will
be given to the in-line electrodes or packet-type furnaces.
          The taphole configuration is another furnace-design feature
which depends somewhat on the particular type of ferroalloy produced.
In smelting some ferroalloys, a slag is produced as a process by-product
(e.g., in the smelting of Std. FeMn and FeCr), while little or no slag
occurs in smelting other ferroalloys (e.g., silicon and ferrosilicon) .
For those cases wherein a slag by-product occurs, the tapholes may be
situated in such a manner that the ferroalloy and the slag components
can be tapped from separate sites to allow separation of the two portions,
Such a tapping arrangement  is not required for "dry" operations.  Like
the packet versus cylindrical furnaces, the issue of taphole configura-
tion has only little to do with the issue of product flexibility wifch
open and sealed ferroalloy  furnaces.
          A furnace-design  feature which does have direct bearing on
the flexibility issue is the need for and incorporation of auxiliary
means of controlling charge descent and flow into the reaction zones.
One means of affecting the  descent of charge material into the reaction
zones is by stoking (i.e.,  rabbling the charge).  Another is by rotating
the furnace shell.  With many ferroalloys (e.g., FeMn, SiMn, FeCr, 50%
FeSi), the use of sized and controlled-property mix components is
adequate to insure that the charge materials will descend uniformly
under reasonably controlled smelting conditions.  However, this in not
the case with high-silicon  ferroalloys.  As  the silicon content
of the alloy increases, the extent  to which  silicon monoxide (S10) vapor
is generated also increases.  This  SiO vapor condenses throughout the
descending portion of the charge and can form an impervious crust or
"bridge" in the mix.  This  crusting in turn  causes a buildup of the hot
by-product gases beneath the crust.  When sufficient internal pressure
*  These smelting processes  in which no slag by-product occurs are fre-
   quently referred  to as "dry" processes.

has developed, the gases either flow along the electrode
sides (by-passing the crust) or erupt through the charge.  Eruption
can throw hot mix for distances of tens of feet*.  The  Jetting of
gases along the electrodes represents a significant loss
nl" energy Us contrasted to uniform gas flow up through the charge) and a
significant increase  in both silicon loss and dust generation.  The
Jons nf hot  SIO vapor along the electrodes results in the release of
larBe quantities of super-fine fume; i.e., this SiO is not con-
densed on the charge and is not returned to the reaction zone.  Stoking
is the primary measure used to break up the crusts once they are formed,
or as a preventive practice to keep thorn from forming initially.
          Rotating the furnace shell is another means utilized  to pro-
vide more uniform control of the mix descent and to minimize crust
formation and "blowing"^" .
          While stoking can be and has  been incorporated in sealed-furnace
operations, stukLnp,  Ls more.- easily practiced with open  furnaces where a
furnace  operator lias  access to and a complete view of the mix  surface.
Likewise, while a rotating  shell can be designed with a  totally closed
cover,  the rotation  of  the  shell with an  open  top is simpler  and more
amenable to engineering design and operation.  As previously  stated,
the need for  stoking and/or shell rotation is related to the  silicon con-
tent  and/or  the content of  other chemical  species which are volatile
at high  temperatures  but which condense out in the descending charge,  forming
viscous  and impervious crusts.  The tendency for bridging and  crusting
and,  in  turn,  the need  to stoke the charge, coupled with the  extremely
high  temperatures of the  offgas, are  the  reasons why such ferroalloys
as  siJLeon and FeSi  with  silicon content  greater  than 75 percent have  not
to  date  been  produced in  scaled furnaces.  This  is also true  for the
rare-earth silicides.  Those  alloys for which  the uniform flow of  charge
can be  controlled by sizing and preparation of the mix  components  (with-
out the  neyd  for extensive  stoking  and/or shell  rotation) arc amenable
 *  The erupt.Lon of the gas from the mix Is appropriately
    cnlled a "blow".

 to either open  or  sealed  furnace  production.   Thus,  standard FeMn,  SiMn,
 FeCr,  and 50  percent  FeSi can he  produced with only  minor  stoking if
 the  charge materials  are  properly sized  and prepared prior to charging.
 These  ferroalloys  are presently made  in  both  open  and sealed furnaces.
          Another  particular furnace-design feature  associated with
 sealed furnaces producing ferromanganese is  (he use  of drying and pre-
 reducing procedures to prepare the manganese  mix prior to  smelting.   It
 should be noted in this instance  that the emphasis is shifted to  the
 complete mix  drying and prereduction  system rather than only a
smelting-furnace operation.  Drying and prereduction  of the manganese-
bearing mix lowers the probability of a  furnace blow  resulting from
mix cave-ins and rapid reduction  of the  large quantities of  the high-
manganese oxide contained in the  untreated ore.  While  drying and pre-
reduction are not  an  absolute requirement for smelting  ferrotnanganese In
sealed  furnaces, utilization of this practice significantly  increases
the operator's confidence that the sealed-furnace operation will be smooth.
 In turn, a smooth  operation is less prone to  dangerous  furnace explosions  and
 requires less stoking. The use of drying and  prereduction  of  manganese charge
materials could be (and sometimes is)  used as part of an open furrotnanganese
 furnace operation, and should contribute the  same improvement in  operating
 conditions on the  open furnaces as with sealed furnaces. However,  this more complex
mix preparation is less needed on the open furnaces,  because furnace blows  are
 generally less harmful on open furnaces  and the  open-furnace  operation can
 counteract to a greater extent the occurence  of blows by stoking  and  rabbling  the mLx

          In summary, there are notable  minor variations in  furnace design
 throughout the industry.  Here the word  "design" is  used In  a general
 sense.  Specific parameters of "design"  that  ncrtain  to particular  ferro-
 alloys  are discussed  in the following suhtasks.
          In the general  sense of the word design, the  seeking requirement
 is a major factor  in  assessing the ease with  which an ouc-n  furnace opera-
tion can be converted to  a sealed operation.   The other general furnace

desi-gn variations have little direct effect on the issue of open versus
sealed furnaces and furnace flexibility.
          Furnace size is an element of significance with regard to the
flexibility issue.  This issue will be addressed in Subtasks 2 and 4.

     Subtask 2;  Define The Parameters Which Must Be Considered When
      Switching Products in Both Open and Sealed Ferroalloy Furnaces

          The parameters which affect the successful smelting of the
various ferroalloys are described in texts on electric-arc furnace
          ( ft — f^
operations       .   As presented by Robiette(6), the boundaries
within which a given elertrlc-arc furnace can be utilized are establish-
ed by
          (1)  the maximum electrodp current (and/or current density),
          (2)  the range of transformer secondary voltage taps,
          (3)  the transformer capacity,
          (A)  the acceptable natural power factor,
and       (5)  the maximum operating resistance.
          The operating resistance is, to a .large extent, determined by
the specific resistivity of the mix materials.  The interrelations be-
tween the charge resistance and the other electrical and geometrical
parameters of the furnace (i.e., electrode-to-hcarth potential, electrode
currentiand diameter) were empirically established and set forth as K-
factor versus Electrode Power Density curves by Andreas'*-^ and Kt'lly l^'.
The K-factor, or peripheral resistance of the charge, is defined as:
                K = (E/I) irD,                       [1]
where           E = electrode to hearth voltage (volts)
                I = electrode current (amperes)
and             D = electrode diameter (cm).
Curves for K-factor versus electrode power density  for selected ferro-
alloys are shown in Figure 2
          As Indicated in Figure 2, a characteristic K-Cactor - electrode
power density relation is associated with the smelting of a given ferro-
alloy   .   By combining the K-factor - electrode power density curves
with the electrical formulations for the furnace circuit (i.e., the
electrode critical current as a function of the electrode diameter); the
  *  The electrode power density is the electrical power carried in each
     electrode (kw), divided by the electrode cross-sectional area (cm2).
 **  In metric units.
***  Persson(13) indicates that the elcctrode-to hearth potential divided
     by the square root of the electrode diameter (i.e., E/^D~)is another
     equivalent unique characterizing parameter for each ferroalloy.

    1.0  „
§   0.8
    0.6  „
    0.2  -
                                                      Hi. C. Ferrochrome
                oa :
O.|2    i    043       0.4       .0,5
I  j     !   •  i          |    i

                          FIGURE 2.  K- FACTOR  VERSUS  ELECTRODE POWER

                           I   ;       DENSITY FOR SELECTED FERROALLOYS
                                                   !     I

furnace power as a function of einctrode current, tun voltage, and power
factor), a set of technically acceptable electric smeLting conditions
can be established for a given ferroalloy.  Curves of the type shown Ln
Figure 3, relating furnace load, voltages, K-factors, and electrode
currents may be used as furnace"operating charts^^^ lor smelting a
given alloy.  Knowledge and control of the mix conductivity and its K-
factor, and the maximum allowable electrode current allows one to estab-
lish the maximum furnace load for given available tap voltages .  For ex-
ample, referring to Figure 3, if the electrode current must not exceed
60,000 amperes and the smelting of the alloy is recommended over a range
of 0.50 to 0.55 ohm-cm, then the maximum furnace load will range from
about 14.2 to 15.9 MW, with voltages between 161 and 175.
          To obtain maximum transformer utilization at a given voltage,
the furnace is generally operated with the greatest ;i.l.lowable current
that will also satisfy the other conditions of charge resistance and
power factor.  It is desirable to have the current-conduction paths re-
stricted to the "reaction" zones between the tips of the electrodes nnd
the hearth.  To achieve this design condition, the voltage is
adjusted or prescribed such that there is little or no current flow
through the charge between the electrodes.  If a maximum voltage Is ex-
ceeded in smelting a given ferroalloy, thf current flow between the
electrodes increases and becomes more difficult to regulate.  Creatfjr
electrode motion is required when there is excessive current flow between
the electrodes than when the predominate current flow is between the
electrode tips and the hearth.  Under extremely high-voltage-input con-
ditions, the electrode penetration is shallow  (i.e., the electrodes arc
high in the mix) and there may be large open areas or zones under each
          In the 3-electrode furnaces there are three interconnected reaction
zones near the hearth level .  These reaction zones are essentially flxt'd
relative to the furnace structure, but have a  contitm.-i] flow-through of
materials.  The material in the reaction zones is continuously under-
going reduction and melting.   While the concept of "reaction" zones ha.-.
*  There is more art  and  empirical know-how  behind  this  statement  than
   science.  Science, however,  is making  inroads  in this

                                      Electrode Current
                                         65,000  amperes
                                                 60,000 amperes
                                     'K" FACTOR, Ohm-Cm
                 FIGURE 3.


been successfully applied  to  furnace  design  and  operation,  the  current-
flow patterns  and the  precise size  and shape of  the  zones  have  not been
experimentally  determined.
          The  current-carrying capacity  of particular diameters of electrodes  is
an intrinsic furnace-design limitation.  As  the  electrode-current  density
(and,  in  turn,  the electrode-power  density)  increases, the  surface tem-
perature  of the electrode  increases.  The temperature of the electrode
portion immersed within  the charge  is relatively unimportant , because  the  heat
developed in that section  is  dissipated  directly to  the charge.  However,
the exposed electrode  surface  (i.e.,  the portion above the  charge)  is
critical, because it oxidizes  and/or deteriorates rapidly as its surface tempera-
ature  rises.  Thus, the  limiting electrode-current density  is determined
by the physical properties of  the electrode  material and the amount of
exposed surface area.  The maximum  allowable electrode current  is  a
principal design feature,  in  conjunction with the K-factor  - electrode
power  density relations.
         Thus,  the electrical-design parameters for  the ferroalloy furnaces
are essentially contained within the K-factor - electrode power density
and the electrode diameter - critical current relations.   However, there
are also geometrical parameters to be considered in designing and  operating
the furnace.  The electrode diameter is a design parameter of extreme
significance as will be  discussed later  in this section.  Practically all
of the geometrical features of the  furnace are derived from or  scaled to the
electrode diameter. -The electrode  diameter  is inherent in  the  K-factor -
electrode power density  relation and  in  considerations of  the critical cur-
rent density for available electrode  sizes.
          Other geometrical factors of importance for smelting a particu-
lar ferroalloy are the electrode spacing (herein specified  as  the distance
between electrode centers)  and the  crucible and shell dimensions.   The
preferred electrode spacing for smelting a given ferroalloy is  based
primarily on experience.   The ratio of electrode spacing  to electrode
diameter (ES/D) varies from slightly less than  2 to somewhat less than
3 for most of the common ferroalloys.   Generally, the higher the silicon
content of the alloy,  the closer the electrodes are grouped.  Thus,

silicon metal is smelted in furnaces with electrode~spacing-to-diameter
ratios of 1.9 to 2.1, while 50 percent ferrosilicon operations utilize
ratios of 2.2 to 2.7.
          Kelly(  ' suggests that the crucible and shell diameters should
scale with the electrode spacing.  Data taken from the literature indi-
cate  that the ratio of  shell diameter to electrode diameter  (S/D) falls
within the range of 6 to 8 for the common ferroalloys.
          Furnace depth is one of the more difficult design parameters to
reduce to a formal relation with the other design parameters.  It has
been suggested     that the ratio of electrode penetration (i.e., the dis-
tance the electrode is submerged in the mix) to the distance between the
electrode tips and the hearth could provide a design parameter for pre-
dicting desirable furnace depths.  Typically, this ratio varies from
3.0 to 0.3 for the standard ferroalloys.
          Furnace designers have developed today's ferroalloy furnaces for
particular alloys by collecting design and performance data from previous
furnaces and attempting to improve upon earlier designs.  Ferroalloy-
furnace designers point out that this is an ongoing process, and that the
"optimum" furnace remains to be defined for many ferroalloys.
          To more fully develop  the interrelation of the design variables
of ferroalloy furnaces for given products, ferroalloy-furnace design and
operating data were collected and tabulated  from  the  open literature
and from field-trip data.  These data are presented in Table 1.  The data
entries include furnace rating,  furnace load, electrode diameter, ratio of
electrode spacing to electrode diameter, ratio of shell diameter to electrode
diameter, crucible depth, electrode penetration, K-factor, electrode-power
density, Persson's parameter^13), tap voltage, electrode-current density,
electrical energy consumed (kwhr) per tonne of product, the percent operating
time, the recovery percentage for the element, and the furnace type (i.e.,
open or sealed), and reference numbers identifying the source of the data.
Entries are grouped according to the ferroalloy product.
          The ranges of K-factors and electrode-power densities, as well as
Persson's electrode-to-nearth potential factor presented  in the table are in
keeping with  the values depicted in Figure 2.  Other Table 1 design and operating


15% FeSi

50% FeSl

50.6* SI

Standard FeMn

78% Mn
14. 8% MB

78-82% Mn

High Carbon FeCr

41% Ct. 41% Si
37% Ci. 47% Si
36% Ct. 40% Si
36% Ci. 42. 3% Si


16. 5






10. 5


















2 31





7. SO











2 33





1. 5-2. 5




High SI

0 43-0.30



0. 50-0. 83
(1)  ES/D = Electrode Spacing (centc:-to-cemer)/Elecuodc Diameter
(2)  S/D - Shell Diamcier/Llertrode Diameter.
(3)  "K" = K Factor = Peripheral Resistance.
                                                                                    •Estimated Values.

                                              TABLE!  1.   (Continued)

8. 625



10. 9








7. 0-1. 75

per Tonne.
13, 780
5. 600-6. 000
- 4, 500
3. 1(4. 4 Max) 2.645





125- ISO


160- 180



-A 4
4. 850
4. 850
- 4. 400
7 .358
7. 500-8. 600
Ope wing Element
Time, Recovery,
percent pcnem

89.9 58. 4 Si
92 77-81 SI

85.1 SI
86. 6 Si
86. 2 SI
85. 1 SI

93.2 Si
93 SI
96. 3 Si
93. 5 SI
- 95 97- 99 Si


95 73 Mn
93. 5 71 Mn

94 85 Mn

92.9 85.3Cr

93-95 88-90 Ci
90-96 82-91 Cr
93 84-92 Cr



















Field Data
1.8. 3. 7
Field Data
Field Data
Field Data
Field Data
Field Data
Field Data
(4)  k/V5 = I'urssnn Elcclruh: in llc.inh I'niciiiul Factor
(S)  EluuroJe Ty|« Code
       P = rith,il«:
       S = Socluiburi;.
(6)  FuniiHL Type CouV
       (1 -Open
       S = Scaled (CMally tUvidl
       Scmi-C = Sciiii-CnviTL'd
       R = Itotaiuig.
"Prereduced Charge Material).


                           References  - Table  1
 (1)   Kelly,  W.  M.,  "Design  and  Construction  of  the  Submerged  Arc  Furnace",
      Metal Progress,  Vol.  73, May,  1958.

 (2)   Persson,  J.  H.,  "The  Significance of  Electroclc-to-Hearth Voltage
      in Electric  Smelting  Furnaces",  ATME  Electric  Furnace  Proceedings
      1970, Vol.  28, pp.  168-169.

 (3)   Engineering  and  Cost  Study of  the Ferroalloy  Industry, U.  S.  En-
      vironmental  Protection Agency, Office of Air  and  Wastu Management,
      EPA 450/2-74-008, May, 1974.

 (4)   Fairchild, W.  T., Journal  of Metals,  Vol.  22,  1970,  pp.  55 to 58.

 (5)   Ascik,  A.  L.,  "Manufacture of  Silicon Metal in a  Three-Phase  Sta-
      tionary Submerged Arc  Furnace",  ATME  Electric  Furnace  Proceedings,
      1964, Vol.  21, pp.  330 to  336.

 (6)   Wolfe,  J.  and  Pait, R. ,  Interchangeable Sheds on a  25,000 KVA
      Electric-Smelting Furnace, AIME  Electric Furnace  Proceedings, .1964
      Vol.  22,  pp.  153 to 156.

 (7)   Elyuten,  V.  P.,  et  al, Production of  Ferroalloys  Electrometallurgy,
      Second  Edition,  Translated from  Russian, Published by  NatJon.il
      Technical Information  Service, U. S.  Department of Commerce,

 (8)   Wise, W.  H.,  "Manufacture  of  75  Percent Ferrosilicon i.n  Large
      Submierged-Arc Furnaces",  AIME Electric Furnace Proceedings,

 (9)   Dann, T.  E.,  and Wise, W.  H.,  "Comparative Operating Characteristics
      of Large Versus  Small  Ferroalloy Units", AIME  Electric Furnace Pro-
      ceedings , Vol. 22,  1964, pp.  150-152

(10)   Zherdev,  I.  T.,  et  al, Critical  Rate  of Rotation  for Ferrosilicon
      Baths,  Stal  (English)  5, 1968, pp,  406 to  408.

(11)   Horibe, K. ,  Totally-Closed Electric Furnace for 75 Percent Ferro-
      silicon Production.

(12)   Semenco, V.  E.,  "Melting  75 and  65 Percent Ferrosilicon  in Air-
      Enclosed Furnace",  Stal (in English)  1969, pp. 563 to  565.

(13)   Schneider, W.  R., and  Eyrich,  J. F.,   "Operation  of  a  75 MVA, 75
      Percent Ferrosilicon  Furnace  with Heat Recovery Equipment",  AIHE
      Electric Furnace Proceedings,  Preprint, 1974,  Meeting, Pittsburgh,

(14)   Tada, Y., et al, Development of  50% Si Fe-Si  Smelting  by the Ore
      Process with a Large  Closed-type ElccLrIc  Furnace.


                      References - Table 1  (continued)

(15)   Lopuszynski,  T.  W.,  et  al, "Design and Operation  of  a  45 megawatt,
      50 Percent Ferrosilicon Furnace", AIMS Electric Furnace Proceedings.
      Vol.  30,  1972,  pp.  89 to 93.

(16)   Harmon,  C. N.,  "Production of  High Carbon  Ferromanganese with  Dis-
      card  Slag Practice", AIMS Electric Furnace Proceedings, Vol. 28,  1970,
      pp.  112  to 116.

(17)   McClure,  R. N.,  "Manufacture of  Standard Ferromanganese in  Electric
      Furnaces", AIME  Electric Furnace Proceedings.  Vol. 19, 1961, pp.  307
      to 314.

(18)   Hopper,  R. T.,  "The Production of Ferromanganese", AIME Electric
      Furnace  Proceedings, Vol. 25,  1967,  pp. 141 to 145.

(19)   Chisaki,  T.,  and Takeuchi, K., "Electric Smelting of High Carbon
      Ferromanganese with Preheated-Prereduced Materials at  Kashima  Works",
      AIME  Electric Furnace Proceedings, Preprint,  1974 Meeting,  Pittsburgh,

(20)   Rossemyr, L., "Manganese Alloy Production  in  a Large Submerged Arc
      Furnace", AIME  Electric Furnace  Proceedings.  Vol. 28,  1970, pp. 121
      r.o 123.

(21)   Oxaai, J., "Manufacture of Silicomanganesc",  AIME Electric  Furnace
      Proceedings.  Vol. 19, 1961, pp.  315  to  344.

(22)   Gamroth,  R. R.,  "Operation of  30,000-kw Submerged Air  Furnace  on
      Silicomanganese", AIME  Electric  Furnace Proceedings^ Vol. 27,  1969,
      pp.  164  to 166.

(23)   Leeper,  R. A.,  and Dyrdek, T.  J., "Smelting of High-Carbon  Ferro-
      chromium in a Three -Phase Electric  Furnace",  AIME Electric Furnace
      Proceedings,  Vol. 23, 1965, pp.  110  to  114.

(24)   Scott, J. W., "Design of a  75-kw High Carbon Ferrochrome  Furnace
      Equipped with an Electrostatic Precipitator",  AIME Electric Furnace
      Proceedings.  Vol. 29, 1971, pp.  80  to 82.

(25)   Madronic, I., "Use of Chromium Ore  Briquetts  in the  Manufacture of
      Ferrochrome Silicon", AIME Electric  Furnace Proceedings,  Vol.  28,
      1970, pp. 174 to 178.

(26)   Nagasawa, S., "Ferrochromium  Silicon Refining in  Closed-Type  Furnace",
      AIME Electric Furnace Proceedings,  Vol.  31, 1973, pp.  125 to  130.

data of particular interest include  (1) ranges of electrode spacing used
with each alloy, (2) use of prebaked versus Soderberg electrodes, and (3)
the electrode diameter versus furnace load for the various alloys.
          As indicated in Table 1,  the electrode spacing  (center  to
center) typically is 2.2 to 2.6 times the electrode diameter for  all
the ferroalloys considered, with  the exception of silicon metal.  Silicon-
metal operations utilize electrode  spacings of about 2.0  times  the
electrode diameter .  It appears  that with the exception  of silicon, each
and all of the ferroalloys considered can be successfully produced with
electrode spacings in the above-cited range (i.e., 2.2 to 2.6 times
electrode diameter).  While individual companies and particular furnace
operators have defined the optimum  or preferred electrode spacing for
smelting particular alloys, it does  not appear that the electrode spacing
in itself is a restricting parameter to furnace flexibility.  That is,
in general, the electrode spacing need not be changed in carrying out a
product change.  However, some furnace operators do adjust electrode
spacing (on some open furnaces) when making a product change.
          Prebaked electrodes, which are capable of carrying a  greater
current density than the Soderberg  (or self-baking) type  of electrode
are used predominately in the smelting of the high-silicon-content alloys
(e.g., silicon metal and 751 FcSi).  However, Soderberg electrodes are
available in larger diameters than  the prebaked types and are,  therefore,
used predominately in today's larger ferroalloy furnaces.  As presented
in Table 1, Soderberg electrodes  are widely used in smelting of standard
FeMn and silicomanganese.  While  a  given type of electrode may  be pre-
ferred for smelting any of the given ferroalloys, furnace flexibility is
affected primarily through considerations of the power and current-
carrying capacity associated with the particular electrode used.
 *  The  one  silicon  furnace entry  in  Table  1 wherein an  ES/o  ratio of 2.43
    was  used was  an  inefficient  operation and was corrected,  in part, by
    reducing  the  ES/D  ratio to 2.23.

          As indicated in Table 1, the electrode diameters (and, in turn,
all the parameters which scale with electrode diameter,  such as electrode
spacing and shell diameter, etc.)  increase with the furnace load.   The
furnace load - electrode diameter  data for the silicon alloys (i.e.,
silicon metal, 75% and 50% ferrosilicon),  silicomanganese, and ferro-
manganese are plotted in Figure 4.  As shown in Figure 4, most of the
furnace load - electrode diameter  data for the given ferroalloys indi-
cate that the furnace load is proportional to the square of the electrode
diameter (i.e., W a D2) .  Such a load - electrode, diameter dependence
is anticipated if the maximum electrode current-carrying capacity (i.e.,
current density) is proportional  to the reciprocal of the
square root of the electrode diameter.
          The specific point of interest to be derived from Figure 4 is
that the furnace load - electrode diameter relations not only differ for
the  ferrosilicon  alloys  and  for ferromanganese  furnaces, but  that the extent
of difference  increases  with  increasing furnace  load  and/or  electrode
diameter (i.e., with furnace size).
          Thus, as indicated by Figure 4, .\ furnace with 90 cm-diameter
electrodes  (specifically Soderberg electrodes) would probably operate
at a load of about 10 megawatts for smelting 50  and 75% ferrosilicon
and about 6 megawatts for smelting ferromanganese.  Such a difference
in furnace  load (i.e., 6 to 10 megawatts) is not particularly difficult
to obtain or achieve with a variable-tap  transformer with an  acceptable
power rating and  operating power  factor.  However, a  furnace which
utilized 180-cm-diameter electrodes would probably be designed  to operate
at a load of 37 or more megawatts  for smelting  the ferrosilicon alloys
and  25 megawatts  for  the  ferromanganese operation.  While a  transformer
and  associated  electrical  system  with a variable power  output  over  the
above range (i.e., 25  to  37 megawatts) is certainly possible,  the cost
of such a system  is quite high.   In general,  the voltage range  of the
transformer and the total  range of allowable  power input,  are determined
by capital-cost considerations.   The  transformer costs  increr.se rapidly
with both  the  number  of  voltage taps  and  the size  (i.e., power



                       	I _  Si Alloys-

                       i  !    W =  11.4  D"1
     20 L
                                                    W  =»_7!. 7!

                                                    I   '   !
                —f- — r—i—r--~
                ;  .! :  I  I  '     :
                i ..-i	!—i—r
                                                                            jg  Silicon!
                                                                                      rir r~r
                           IGURE 4. __ FURNACE^ LOAD, VERSUS! ELEQTRpI)| DIAMEtER!   ;  !   !  j.

                            ([Arrows indicate  overlapping data  po|ints.j)      ;          1
                             .___,.              .   ._„. ._ ..    ._i_^..       _.._

120        140      160

  Electrode Diameter (cm)

rating) of the transformer.  In addition, terms of the ferroalloy com-
pany's contract with the electric power company regarding firm, inter-
rupted and dump power also influence the design versatility of any in-
stalled transformer system.
          As a result of capital and operating cost considerations,
large ferroalloy furnaces and, in turn, the required large transformer
systems and electrode columns are generally designed specifically for
the production of one ferroalloy or a limited group of ferroalloys re-
quiring similar smelting conditions.  Based on the Figure 4 data and
economic considerations, it appears that furnace size is a major para-
meter relative to the issue of furnace flexibility.
          Based upon this review of the theory and practice for ferro-
alloy furnace design and operation, the major operating parameters which
must be adjusted to effect any product change in a given furnace are
          (1)  the mix composition
and       (2)  the voltage and power to secure proper smelting
          As will be discussed in Subtask 3, many of today's ferroalloy
furnaces are equipped with on-line voltage changers which allow the
furnace operator to stepwise change the transformer tap voltage without
a major interruption of the power supplied to the furnace and charge.
The ability to change tap voltage (and, in turn, input power) as the
charge resistance changes in conjunction with the change in mix composi-
tion, allows the furnace operator to effect a continuous, gradual product
change in a given furnace.
          While generally not a major prerequisite to achieving a product
change, some furnace operators also adjust electrode spacings when carry-
ing out a product change.  It is the opinion of the Battelle researchers
that, in general, the electrode spacing need not be changed in carrying
out most product changes.
          BCL personnel realize that the foregoing presentation of opera-
tion parameter relations are not simple or even "straight forward".   The
relationships required  to understand and parameterize a  total furnace  opera-
tion are couched in mathematical terms.  A mathematical  formulation was
not presented herein, but rather a  simplified overview of each parameter
and the effects of parameter changes.


    Subtask 3;  Identify and Compare General Procedures for Changing
 From One Product Family to Another in Open and Sealed Ferroalloy Furnaces

          Prior to describing the general procedures involved in carry-
ing out a change of product in a given furnace, it is advantageous to
describe what is meant by a given furnace.  In this report, a given
furnace is taken to indicate that the following furnace components or
parameters are prescribed and fixed (i.e., they are not subject to
          (1)  the installed transformer and associated
               electrical system (e.g., the bus bars),
          (2)  the electrode diameter,
and       (3)  the furnace crucible and shell.
          To change any of the above furnace components (or parameters)
requires major alteration to the unit.  The above components are generally
only changed in a complete retrofit and/or overhaul situation , resulting
essentially in a new "piven" furnace.  While the above components can be
changed (i.e., essentially a new furnace can be constructed in place of a
previous one), such major alterations are long-term events and
do not affect short-term furnace flexibility.
          Now, let us turn our attention to those procedures and/or
conditions which are involved in carrying out a change of product in a
given ferroalloy furnace.
          In effecting a change of product in a furnace, several basic
operating conditions must be adjusted or changed.  These basic or funda-
mental changes include:
          (1)  the change of mix composition
and       (2)  adjustment of voltage and power for proper
               smelting conditions.
That the mix composition must be changed to effect a change in product
is obvious.  For example, in changing from a 50% ferrosilicon to a 75%
ferrosilicon operation, less scrap and more silica and
reductant (e.g. coke) are charged per unit of mix fed to the furnace.
Mix compositions for smelting the various ferroalloys arc available in
in the literature(6» 9» 16>.

*  InterlakeUS) has reported on the use of interchangeable crucible
   and shell units for increasing the flexibility of one of their

          As described in the Subtask 2 section on parameters which must
be considered in performing a product change,  the voltage and,  in  turn,
the furnace load are adjusted (using the K-factor-elcctrode power  density
data) in any change of product.  The particular adjustment of tap  voltage
and furnace load will depend on both the product change and the allow-
able range of voltage (and power) for the given transformer.
          Most of today's transformer-furnace  systems are equipped with
on-line voltage-changing facilities which enable the furnace operator to
incrementally raise or lower the tap voltage with a simple switch  (a tap-
voltage selector).  As the composition of the mix charged to the furnace
is altered to secure a product change, the charge resistance is also
altered.  For a fixed voltage, the current flow through the electrodes
will change with the change in mix resistance.
          The furnace operator may monitor both furnace voltage and
current and can adjust the working voltage via the on-line voltage selector
to maintain the current at a prescribed value within the allowed working
range.  The maximum current allowed is determined by the current-carrying
capacity of the electrical system components (e.g., transformer,
and electrodes).  In general, the furnace operator will attempt to
utilize a current slightly less than the allowed maximum, in order to
obtain a near-maximum total power input under  all smelting conditions.
          The above changes are required in any product change, inde-
pendent of the furnace being open or sealed.
          There are two basic methods for securing a change of product;
(1) a gradual change of product by continuing  to feed the new mix  com-
position to the operating furnace; and  (2) to  shut down the furnace, dig
out and remove the old product mix from the furnace, and recharge  and
start again with the new product.  The decision to utilize one  or  the
other of these approaches to a product change depends on a number  of
factors, such as (1) feasibility of a continuous (i.e., furnace-on) product
change, (2) the time required to effect the product change from Alloy No.
1 to on-grade Alloy No.  2, and (3) the marketability of the off-grade pro-
duct (s) produced during the product-change period.
          The gradual and continuous change of product is the most fre-
quently utilized procedure.  Ferroalloy-company representatives estimated

that better than 95 percent of the U. S. industry's product changes were
performed as continuous operations, without shutdown of the furnaces.
Those product changes wherein the furnace is shut down and subsequently
restarted on a new product are only used when the product changeover
can be concurrent with a shutdown of the furnace for major repairs and
          During a gradual product changeover period in the switching of
products with significantly different smelting conditions, the furnace
operation may not be so "smooth" nor so well controlled as during "steady-
state" operations.  In the anticipation of trouble, sealed furnaces may
be opened during changeover to observe closely the furnace top and to
be in a position to stoke, if necessary.  In these instances, the emission-
control equipment would have difficulty handling the increased flow of
gases and the offgases may be released without cleaning.  If this "safety"
practice were followed (i.e., sealed furnaces were operated in an open mode
during product transition), the air-pollution control and energy advantages
of sealed furnaces might be lowered.
          Some furnace operators also adjust electrode spacings when
carrying out a product change in a given furnace.  An adjustment of
electrode spacing is relatively easy on some open furnaces.  However,
with a sealed furnace, the furnace cover is designed to seal against
the electrodes at a particular design spacing.  Thus, the electrode
spacing on a sealed furnace is essentially fixed and constant.  While the
data in Table 1 indicate that most of the ferroalloys can be smelted
with an electrode-spacing-to-diameter ratio between 2.2 and 2.6, the
ability to adjust the electrode spacing in conjunction with a product
change gives the furnace operator additional freedom to improve the smelt-
ing operation.
          The general procedures for changing from one ferroalloy product
to another in open or sealed furnaces should be quite similar (i.e.,
change in mix composition, voltage, etc.).  However, because of the
accessibility of the charge in an open furnace (and, in turn, the
ability to stoke and rabble the charge), an open furnace is inherently
more flexible (i.e., more  product  changes can be carried  out) than  a
sealed furnace.

      Subtask A;  Identify Deviations from Optimum Conditions and
             Other Penalties (Including Economics) Which are
       Involved in Switching Products in Open and Sealed Furnaces
          The incentive for changing the ferroalloy produced in a given
furnace is purely one of economics.  In general, a product change has
been made for one of two basic reasons—(1) the raw materials which were
used are no longer available at an acceptable price level or (2) the
market demand and/or profitability associated with a new product suf-
ficiently exceeds that of the old product to warrant a furnace change-
over.  U. S. ferroalloy companies have practiced furnace flexibility
for both of the above-stated reasons and have indicated that long-term
furnace flexibility is a desirable consideration in any new furnace
designs, because of the possible reduced availability of manganese and
chromium ores to U. S. smelters.  As noted in the Introduction, many
speciality ferroalloy producers use small (^10 MW) furnaces to produce
several ferroalloys per year in a given furnace.  In this case, the
market demand is limited and the output of a year-long campaign on one
of these specialty ferroalloys would exceed the total marketable
quantity several times over.  These companies practice short-term fur-
nace flexibility to remain profitable and competitive.
          Thus, while there may be operating "penalties" associated
with a product change in a given furnace, there should be no economic
penalty, assuming that the company management has adequately and accurately
analyzed the business market.  That is, even though the furnace operation
may be  "rough" and require greater operating attention, the product switch
is designed to be profitable.

          When a product change is made in a given furnace, be it open
or sealed, there are several factors which influence the overall profit-
ability of the change.  Several of these factors to be considered are:
          (1)  The required down-time to achieve the product
               switch and/or the loss associated with the
               production of off-grade or nonsalable alloys,,

          (2)  the required power consumption per ton product,
and       (3)  the change in furnace productivity (i.e.,  tons
               of ferroalloy produced per month).
          The time required to achieve a product change  is,  in  part,  a
function of  the specific alloys involved, the size  of the  furnace,  and  the
furnace design (e.g., whether the furnace is open or sealed).   A minimum
conversion time is associated with a gradual product change  between ferro-
alloys of similar smelting conditions  (i.e., 50% *•  75% ferrosilicon)  in
small  open furnaces  (of  the order of  10 megawatt  load).   A conversion time
of about several days  to a week is anticipated  for  the above case  and,
also,  the product associated with the conversion period  is often  salable
as off-grade material.  As  the furnace size  increases and  smelting con-
ditions differ,  the  time required for conversion  increases.   Typically,
4 to 8 weeks may be  required  to convert a reasonably large (about 30 MW)
open furnace from one  product  to another.   In  some  cases,  this  time
period is  downtime  during which ao ferroalloy is produced.
          There is very little field experience to date  to define precisely
the time required for a product change in a largo sealed  furnace.   However,
for products  with a major change In smelting conditions,  the conversion
time is expected to exceed that of a compariable open furnace because of
the additional time  required to adjust and stabilize the operation prior
to the replacement of the cover seal.    It is anticipated  that  the conver-
sion time required for a large sealed furnace could exceed that of an open
furnace by several weeks.
           To date,  the procedures  for carrying out a product change be-
 tween  ferroalloys requiring significantly different smelting conditions
 in a large  sealed  furnace  have not been  presented in the open literature.
 Based  on the MRI study (see the Appendix) and  discussions with the U. S.
 ferroalloy companies,  the large sealed furnaces operated  to date are es-
 sentially single-product furnaces  and are likely to remain single-product
 operations for the  life of the furnace.   Because there are no field data as to
 how a  product change would be performed  within a large sealed furnace,
 our presentation is based upon our best  estimate of this  technology.  We

anticipate that a large sealed furnace may be operated in the open mode
(l.e, without the sealed cover) during any major product change.  Thus,
whereas a product change with an open furnace would essentially utilize
the same air-pollution-control devices as during its steady-state opera-
tion, additional air-pollution-control systems and/or permission to operate
the furnace with increased emissions may be required in the changeover of
a sealed furnace.  This premise is based on the belief that the successful
sealed-furnace operation requires greater control of the process than the
corresponding open-furnace operation.  The product change with a sealed
furnace is anticipated to involve (1) opening the furnace (bypassing the
air-pollution-control system) and (2) stabilizing the operation on the
new product prior to resealing.  Such a procedure may have a direct bear-
ing on the overall air cleanliness rating of the operation.
           As  stated  in  the Sub task  2 and  3  sections,  the  voltage  and  fur-
nace  load  are generally changed when carrying  out  a product  change  in a
given furnace.   These changes, in turn, may causu  an  electrical energy
consumption penalty.  For example,  if the  furnace  and  transformer system
were  designed for optimum power  input for  smelting one  ferroalloy,  a  shift
from  that  alloy  to another may cause a  change  in power  factor  and  in  energy
usage.  Thus, the operator may have  to  pay  for more electrical energy than
would be required in a  system specifically  designed  (and  optimized) for
smelting the  new product.
           There  is an  anticipated change in furnace productivity with
any change in furnace product.   As  indicated in Table  1,  the electrical
energy consumption per  tonne  of  product (Kwhr/tonne)  varies  with  the
particular ferroalloys.  For  example, approximately  15,000 Kwhr are
consumed to produce  a  tonne of silicon  metal in  the U.  S.  practice,
while a tonne of standard ferromanganese  requires  about 2600 Kwhr.  The
anticipated productivity of a furnace is  the furnace  load divided by  the
specific energy  consumption per  tonne of  alloy times  the  operating time
factor.  This is:
           Q = (P/EC  ) x Operating Time Factor,                      [2]
where     Q = quantity  (tonne) of ferroalloy produced  per hour,
           P = furnace  load  in kilowatts
 and      EC = required  energy consumption  in kilowatt-hour per tonne of alloy.

          Based on the estimated furnace output (i.e., tonnes of product
per unit time), the market demand and the profit margin (i.e., the
sales price minus the production cost), the ferroalloy company manage-
ment can formulate their policies and decisions regarding furnace
          Throughout this discussion of the penalties (and gains) as-
sociated with product changes in a given furnace, it has been assumed
that the product change is feasible and possible within the particular
furnace.  There have been cases where a product change was initiated
and the ultimate penalty of zero production occurred; that is, the fur-
nace could not be tapped under the prescribed smelting conditions.
While this condition of an untappable furnace has occurred, it is
rare in today's operations.


     Subtask 5;   Do Case Studies of the Product Families Which are
 Most Frequently Interchanged Based on Typical Open and Sealed Furnaces

          The examples of product changes cited in this section are based
on open literature data and information provided by individual U. S.
ferroalloy companies.   Because of the longer history of usage and the
inherent greater product flexibility associated with open furnaces, there
are significantly more examples of product changes with the open-type
furnaces than with sealed furnaces.
          The most frequent and/or common changes of product within given
open furnaces are those involving similar smelting conditions.  For ex-
ample, product changes of silicomanganese to standard ferromanganese or
vice versa (hereafter represented as SiHn +• FeMn) or 50% ferrosilicon to
 75% FeSi  to  silicon metal or vice  versa  (i.e.,  50  FeSi <- 75  FeSi «- Si)
 are  fairly  common.  Other  reasonably frequent product  changes in open  fur-
 nace  include the sequence  50%  ferrosilicon  to ferrochrome-silicon and  then
 to ferrochrome (i.e. ,  50 FeSi «- FeCrSi «- FeCr) .  Changes of alloy grade within
 a family  are also common (e.g., FeCr «- blocking chrome «- charging chrome).  The
 product switch  between  ferrosilicon alloys  and ferromanganese is less
 frequent, but this  change has  been performed.   As  a  generalization, one
 can state that  all  the  permutations of product changes  have  been carried
 out with  small  open furnaces.   The time  required for a  particular product
 change has  typically  been  1  to 2 weeks,  depending  on the extent  of dif-
 ference in  required smelting conditions  for the particular ferroalloys
          The versatility of a small open ferroalloy furnace is  evidenced
 by the operation of what may be termed the  "universal"  furnace,  in which
 essentially  all  the common ferroalloys have been successfully smelted.
 The basic "universal" furnace  consisted  of  three 89-cm-diameter electrodes,
 with  a 2.44-meter center-to-center-electrode  spacing (i.e.,    /   = 2.74)
 and a  7.92-meter diameter shell (i.e.,   /  «  8.9).  The open furnace  was
 equipped with a  variable-tap transformer to allow  the smelting of  all  the
 common ferroalloys in one furnace.   These basic furnaces were used for
 many years to smelt all  the  common  ferroalloys.  The only major  parameters

changed (besides mix composition) in performing a product charge were
voltage and furnace load.  For example, ferromanganese was smelted with
a furnace load of 6 to 7 megawatts, silicomanganese with about  8 megawatts,
ferrochromium with about  10 megawatts  and  50% ferrosilicon with 12 to 14 mega-
watts.  Stoking was practiced with all these  operations.
          The examples of short-term furnace  flexibility, wherein up to
ten specialty ferroalloys (including the rare-earth silicides,  calcium
silicon, and the ferrosilicons) are produced  per year in a single ferro-
alloy furnace are also achieved with relatively small (up to about 10-
megawatt) open  furnaces.
          Some longer term product flexibility is also associated with
larger (about 30-megawatt) open furnaces.  u- s« ferro-
alloy companies cited numerous examples of product changes (e.g->
FeSi •*• FeCr) carried out on large open ferroalloy furnaces.  These pro-
duct changes typically required several weeks for completion and were
dictated by a major shift in market demands.
          To date, sealed furnaces are designed and operated essentially
as single-product furnaces.  Because  the majority of  the world's sealed  ferro-
alloy furnaces are located in Japan, a subcontracted study through
Mitsubishi Research Institute, Inc. was made  of the Japanese ferroalloy
industry.  The full text of the M*I study  of  the Japanese ferroalloy sealed
furnace operations is presented in  the Appendix.   The MRI  study supports
the previous statement that the Japanese sealed furnaces are currently de-
signed and operated as one-product furnaces.  The greatest number of
sealed ferroalloy furnaces  is  producing  standard  ferromanganese.   The
ferromanganese sealed-furnace operation can and has been converted to the
production of silicomanganese and vice versa; however, no other product
changes are known of in sealed furnaces.   The lack of demonstrated flexi-
bility for sealed furnaces may be due in part to (1) the relatively large
size of most of the new sealed furnaces and (2) the lack of operating ex-
perience associated with the sealed-furnace operations.  In either case,
no case histories baaed on foreign or domestic field experience is  avail-
able relating to product changes (other than FeMn «- SiMn) in sealed ferro-
alloy furnaces.

         Subtask 6:  Draw Conclusions. Where Possible, on the
         Limitations to Flexibility Which Would Be Encountered
          in the Domestic Ferroalloys Industry if All Future
         Capacity Expansion Were to Consist of Sealed Furnaces
          Recognizing that sealed-furnace operations have not been demon-
strated feasible for smelting some ferroalloys, the installation of only
sealed furnaces in the future would limit the production of
the new furnaces (by comparison with open furnaces) to only those products
amenable to sealed-furnace production.   The literature and field data
indicate that the current sealed-furnace product capability is as shown
     Demonstrated                                 Not
     Operational       	Marginal      	Feasible	
         FeMn               75 FeSi                Si
         SiMn          FeCrSi (1 stage)           CaSi
         FeCr                             Rare-Earth Silicides
     50 FeSi
FeMn, SiMn, and FeCr are currently smelted in large sealed furnaces
throughout the world.  Fifty percent FeSi is smelted in relatively large
sealed furnaces in Japan.  However, the  large U. S. 50% FeSi  furnaces are
all  open.
           Japanese sealed-furnace smelting of 50 percent ferrosilicon
differs from the U. S. open furnace practice in that the Japanese use
siliceous  ore as their source of iron.   Steel scrap is the iron-bearing
charge material used throughout  the U. S. industry to produce ferrosilicon.
While a greater electrical energy should be required per tonne of ferro-
silicon product with the ore practice relative to the scrap method, the
Japanese ferroalloy representatives indicate'the ore usage is justified
in terms of allowing better mix sizing and feed control and yielding a
smoother smelting operation (Refer to Yakagi Interview, pages A5.1 and 5.2
of the Appendix).   U. S.  ferroalloy industry representatives have indicated
that, while the sealed furnace smelting of 50 FeSi is clearly demonstrated
with  the use of ore, the sealed  furnace  production of ferrosilicon
with a steel scrap charge has not yet been proven feasible.

          The smelting of 75 FeSi and single-stage FeCrSi in large sealed
furnaces is presented herein as marginal; that is, these alloys have been
smelted in relatively small (by today's standards) sealed furnaces, but
further development is required before large sealed-furnace operations are
proven.  From both the theoretical and practical views, the smelting of
silicon metal and the rare earth silicides in sealed furnaces is not yet
          As previously stated, the average size of the new installed fur-
naces has shown a steady increase over the last several decades    .  Thus,
the significance of furnace size relative to the ability to seal the furnace
on a given product is extremely important.   In  general,  a  large  open  furnace
is more difficult to operate smoothly than a small open furnace, particularly
if stoking is required.  This is, in part, the reason  that U. S. ferroalloy
producers do not accept the operation of a 11 MW sealed-furnace operation
with 75 FeSi as confirmation or demonstration of feasibility for smelting
75 FeSi in a large (e.g., 45 MW) furnace.
          As implied in the Mitsubishi Research Institute report on the
Japanese sealed-furnace operations, a sealed ferroalloy furnace is less flexible
than an open furnace.  The Japanese essentially present the sealed furnace as
a single- product furnace.  However, product flexibility, particularly long-
term flexibility, is a matter of definition and degree.
          It is of some interest to herein speculate on the differing need
for furnace-product flexibility in foreign countries relative to the United
States.  As noted in the Introduction, most of  the world's sealed ferro-
alloy furnaces are in Japan.  Japan currently has 39 totally sealed ferro-
alloy furnace operations  (Refer to Page A.2.1.  of the  Appendix).  There is
only one sealed furnace currently producing a major ferroalloy product
(silicomanganese) in the United States .
*  The particular U. S.  scaled  furnace was  originally designed  and  operated
   for the production  of pig  iron.  Operational difficulties were encountered
   in converting this  sealed  furnace  to  the production  of  ferrosilicon  and
   ferromanganese.  The  furnace is  currently utilized to produce silico-
   manganese.   Because this particular furnace was  not  specifically designed
   for the smelting of ferroalloys, it is  the position  of  BCL personnel
   that  this  particular  sealed  furnace should not be cited as a pro or
   con in judging the  attributes of sealed-furnace  technology.

          In the United States, producers of ferroalloys are concerned that
the ability of sealed furnaces to switch from product to product is or may
be less than the ability of open furnaces to make such switches.  In Japan,
the ferroalloy producers generally do not have the same concern about
furnace "flexibility".  Thus, sealed furnaces generally face a more receptive
attitude in Japan than in the United States.  One can question the reason
for this difference in attitude.  The answer involves some speculation,
but basically appears to the present BCL investigators to be the result of
(1) differences in the relationships between the Government and the ferro-
alloy producers and (2) differences in the relationships between competing
producers of ferroalloys.
          As indicated in a previous EPA-prepared document^ ', differences
do exist between Japan and the United States with respect to Government
involvement and interactions with industry with regard to taxation, sub-
sidy payments for promotion of industrial growth, export policies, and
other aspects.  Although the specific mechanisms are obscure when viewed
by outsiders, it is a fact that Government - industry relationships in
Japan are such that the Government is in a strong position to influence
the output and operating practices of even private-sector manufacturers.
The Japanese government has at its disposal methods for applying pressure  on
industrial manufacturers to a degree not practiced in the United States.
One form of this pressure appears to be as follows.
          A Government planning group (frequently MITI, The
          Ministry of International Trade and Industry) devises
          a national plan for the production of some commodity,
          say, for example, ferroalloys.  The plan takes into
          account existing facilities, domestic demand, export
          and import potentials, costs of production, availability
          of raw materials, and even social factors.  One pos-
          sible conclusion from such an analysis  is that it
          would be better overall for the national economy for
          Plant 1 to  specialize in Product A, Plant 2 to specialize
          in Product B, etc.
Although the organization of the Government-industry relationship does not
necessarily permit "control" of the manufacturers to follow the plan,  in

 practice  the  types  of  pressure  available  to the  Government generally re-
 sult  in acceptance  of  the  plan  by the  individual manufacturers.
           The  Japanese system,  therefore,  in essence  evolves  to  a  type  of
 allocation of  tonnage  and  products  among  various manufacturers.  The
 tendency  is for  each manufacturer to be responsible for  the production  of
 a narrower range of products, but at a higher rate of output  than  if a
 larger number  of his competitors  were  also making  the same products.  To
 the degree that  sealed ferroalloy furnaces have  limited  flexibility  rela-
 tive  to open furnces,  this Government-industry relationship makes  sealed
 furnaces  more  attractive to the  Limited-product  companies  that tend  to
 evolve more extensively in Japan  than  in  the  United States.
           The  Government-industry relationships  in Japan also has  an in-
 fluence on the relationships that develop  between competing ferroalloy
 companies in the private sector.  Somewhat assured of a  large market for
 their specific products, and somewhat  insulated  from  the threat  of un-
 expected  foreign intrusion of another  producer into their  prime  market,
 individual ferroalloy  companies  operate in an environment  that encourges
 them  to exchange technical and research information through industrial
           Less there be misunderstanding,  the  Japanese system as described
 here according to our  understanding of the  system, involves little in the
 form of formal "controls".  An individual manufacturer can  on his  own
 initiative deviate  from the plan  and take  his  chances  in open competition.
 This has  been done  from time to time,  but  the  resulting  experiences  have
 been such that such deviations involve more risks than does compliance
 with the  plan.   Furthermore, it should be  recognized  that  the Japanese
 system of  "planning and pressure" is not limited to the  ferroalloy in-
dustry, but extends also into other industries,  of which the steel in-
dustry is another example.
           In the  United States, there  is no equivalent of  the Japanese
national plans,  and the pressures available to the Government to influence
 the tonnage or type of product made by any manufacturer  usually are  in-
 significant.  The result is that  U. S.  producers of ferroalloys compete

vigorously with each other over a wide range of products, often changing
product mix or tonnage based on each producer's evaluation of the total
market, sources of raw materials, import and export situations, actions
of his competitors, and relative profitability of various products.  In
this environment, flexibility of ferroalloy furnaces takes on great im-
portance.  Large sealed furnaces, especially in their present stage of
development are not so acceptable in the U. S. system as in the Japanese
system and, in the final analysis, this difference arises from difference
in business modes between the United States and Japanese ferroalloy
          The U.  S. producers are concerned about the availability of
chromium and manganese ores to the U. S. for the future smelting of
ferrochrome, ferromanganese, and related products.  Because of this
long-range concern, U. S.  producers would prefer the installation of
large open furnaces.  Thus if a forced product switch were made
(i.e., manganese and/or chrome products to silicon products), the
adjustment could be carried out more easily in existing furnaces.
          In conclusion, if all new furnace installations were sealed
furnaces,  flexibility would be curtailed.   Short-term product
flexibility, which is essentially associated with the smelting of
specialty products in small furnaces, would be severely affected, if not
eliminated.  Long-term flexibility would be possible.  However, the
difficulty, as measured by the time and cost, associated with a product
change in a large sealed furnace would be greater than the corresponding
product change in a large open furnace.

             Subtask 7;  Make Recommendations for Additional
                   Research, Development, and Possible
           Demonstration Programs Which Could be Undertaken by
                EPA to Aid in the Solution of the Product
               Flexibility Problem Identified in Subtask 6

          As has been shown, a movement cowards large, sealed ferro-
alloy furnaces is a movement towards furnace inflexibility.  The develop-
ment and acceptance of sealed ferroalloy furnace operations is in its
early stages*.  As the state of the art evolves, it is considered probable
that large ferroalloy companies in the United States having large resources
(capital and technology), and an assured market, will in the future install
sealed furnaces dedicated to one product.  Battelle has no recommendations
for research and development that could make a sealed furnace more flexible
in terms of product shifts.
          As indicated in this report, some ferroalloys cannot as yet be pro-
duced in sealed furnaces.  Also, to retain some elements of product flexi-
bility, some ferroalloy producers want to install only open furnaces in the future.
On large furnaces this flexibility aspect may be more theoretical than practical,
but the flexibility aspect is more available with open furnaces than with sealed
furnaces.  It is judged, therefore, that there will continue to be many open
furnaces even in the future.  Battells has, therefore, included recommendations
for (1) reassessment and comparison of the relative environmental impact of open
and sealed furnaces via a cross media impact type study, and (2) approaches to be
considered for the improvement of the air-pollution control of open furnaces .
          The EPA is interested in aiding a movement towards sealed ferroalloy
furnaces for new installations in the United States.  As stated by Japanese
ferroalloy company representatives, the normal development of sealed furnace
operations is to improve the smelting performance in open furnaces to the
point where sealing of the operation can be technically justified.  Battelle
researchers have, therefore, included in this Subtask some suggestions for
possible research projects that could improve smelting operations in general.
*  Semiclosed (i.e. , mix seal) ferroalloy furnaces and totally sealed pig
   iron and calcium carbide furnace operations predate the introduction of
   sealed ferroalloy furnaces.  Sealed ferroalloy furnaces came into use
   in the mid to late 1960's.

Battalia's Recommendations Relative to
Open and Sealed Ferroalloy Furnaces

          (A)  It is recommended that the EPA study and evaluate
               the comparative environmental-impact level of both
               sealed and open furnaces using the criteria of all
               of the expected environmental guidelines of the
               future.  These criteria should include water- and
               air-pollution guidelines and waste-control standards,
               and energy considerations.  Also included in this
               evaluation should be the expected improvements in
               baghouse performance (if any) and expected improve-
               ments in water-treatment methods to achieve zero-
               discharge standards.
          It is appreciated that EPA has stated that  "sealed" furnaces are
inherently superior from an air-pollution-control aspect  C1).  While  this
superiority in one environmental area is significant,  the need exists to
define  that furnace system which is best from the viewpoint of  total  environ-
mental  impact.  To quote a recent  paper entitled  "Assessing Cross-Media
          "Certain control technologies aimed at  achieving  specific
          limits  generate new waste streams  which, in turn, require
          controls.  Without a  systematic  assessment  of  these  im-
          pacts,  there  is a substantial  risk that pollution control
          strategies  for controlling  pollution  in one medium  will
          exacerbate  problems in another medium".
Such may be  the case  with sealed ferroalloy  furnaces.
          In  sealed  ferroalloy  furnaces  coupled to wet-scrubber systems,
there is no combustion  of the organics released from the operation prior
to the  delivery of  these  gases  to  the wet-scrubber system.   All of the
volatiles from the  coal,  coke,  electrodes, and  wood  chips then enter the
scrubbers.   Under these conditions it is probable that some of these
organics report  both to the  cleaned  carbon monoxide  gas (which is later
combusted under  varying conditions)  and  to the  recycled scrubber water.
 This point  is recognized in  the EPA waste-water-treatment guidelines for
 the ferroalloy industry.   In these guidelines,  cyanide and phenols are

expected in the waste water from wet air-pollution-control devices on
sealed ferroalloy furnaces.  These pollutants are not encountered in the
waste water from open ferroalloy furnaces.
          There is also concern over the buildup  (and elimination) of
soluble elements and compounds (organic and inorganic) in totally recycled
scrubber water (from sealed furnaces) over time.  There is a need to
evaluate ferroalloy furnaces of the sealed and open variety from a view-
point of total environment impact.
          These potential water problems have already been evaluated in
part by the water-pollution-confrol sections of EPA.  From one viewpoint,
a sealed ferroalloy furnace can be regarded as a coal gasifier.  The over-
all environmental impact of coal gasification projects is being evaluated
and the knowledge developed on these projects can be used to evaluate the
air- and water-pollution control aspects of both sealed and closed furnaces.
          Battelle researchers also recommend that the air-pollution-control
aspects of sealed ferroalloy furnaces be reviewed taking into considera-
tion the emissions from sintering and drying operations often associated
with the use of upgraded material in sealed furnaces.  Sintering and dry-
ing operations are often difficult to control adequately from an air-
pollution point of view and, if required as part of the ferroalloy-furnace
operation, those sources of air pollution should be examined in any
evaluation of the total system.
          (B)  It is recommended that EPA evaluate the designs, op-
               eration, and performance of the baghouses associated with
               ferroalloy operations.  This evaluation should be from
               the viewpoint of aiding in decreasing the annual average
               emission weight (per ton of a particular alloy) in
               future baghouse installations for both old and new
               furnaces.  At this time, it is not technically
               possible to produce silicon metal  (and several other
               ferroalloys) in sealed furnaces.  There is, therefore,
               a need to improve baghouse performance with open
               furnaces to lower the overall emission rate and to
               approach the particulate emission rate of sealed

          It appears that there will be a need to operate some new and
old ferroalloy furnaces in the open mode with dry dust collection.
Ferroalloy producers are, in part, dissatisfied with the efficiency and
operating-time performance of baghouses.  It is recommended that EPA
sponsor research and evaluation studies to improve the performance of
baghouses on ferroalloy furnaces.  This study should take into consider-
ation the special or different characteristics of the ferroalloy furnace
dust and develop new approaches to improving baghouse performance
collecting these materials.  The possibility exists that the industry may
lend support (possibly as a joint sponsor) to such a program.
          In the evaluation portion of this baghouse study, information
should be obtained on the factors influencing the operating-time perfor-
mance of existing baghouses.  Some baghouse designs or equipment require
considerable "down-time" for maintenance and repairs.  Some of the problems
arc (1) difficulty in locating and eliminating individual broken or torn
bags and (2) maintenance of the fans and blowers required to operate these
units.  Baghouses are most often purchased on a competitive-bidding basis
and only some purchasers are in the technical position to specify desired
design improvements that relate to their type of particulate dust.  These
design improvements either remain proprietary or they become incorporated
into units sold to other companies, provided that the price remains com-
petitive.  There is thus only a slow improvement in baghouse performance,
most of which (say the buyers) indirectly results from efforts by individual
ferroalloy companies.  There is a need for an overall evaluation from the
user's viewpoint and an acceleration in the improvement of baghouse per-
formance as the result of a concentrated and innovative technical input.

Battelle's Suggestions for Possible Research Projects

          Because of EPA's interest in sealed furnaces, Battelle researchers
have considered research and development projects that would promote the
development (and acceptance) of sealed furnaces.

          Overall, it is judged that additional conferences and dis-
cussions will be required between  EPA and ferroalloy-industry execu-
tives before specific research programs are formulated.  A problem
to be resolved is to avoid duplication of confidential in-house research
and development projects being carried out at competing ferroalloy com-
panies.  As an example, Battelle researchers until recently were expect-
ing to suggest research in the fields of  (1) development of computer con-
trols on electric smelting furnaces to smooth the operations in prepara-
tion for eventual operation in sealed furnaces and (2) salvaging the
heat generated from combustion in open furnaces (using close-coupled
waste-heat boilers).  Further consideration of these suggestions was
abandoned when two papers on these subjects were given at a recent meeting
                                                                   (19   20^
of the AIME Electric Furnace Conference (December 11 and 12, 1974)   '  "'•
It is possible that some or even .all of the following research projects
may already be underway at individual ferroalloy companies.
          Battelie's suggested research programs relate mainly to  im-
proving the technology of smelting operations for 75% ferrosilicon.  This
particular alloy is selected as a target  for production improvement be-
cause  (1) it is expected that there may be a future emphasis on silicon
alloys in the United States, (2) this is  probably the next major ferro-
alloy to be "tamed" by advancing aealed-furnace technology, and (3)
improvements in the production technology of this alloy can be trans-
ferred to other alloys  (including Si metal).

Possible Research Prelects

           (A)  Determine whether the  substitution of  iron-ore
               pellets  for ferrous scrap  in  the production of
               50 and  75% ferrosilicon alloys decreases the need
               for  stoking these operations  and results in a smoother
          U. S. producers of ferrosilicon alloys use  ferrous scrap in  the
charge to supply  the ferrous component of the alloy.   Japanese manufacturers,
on the other hand,  use  iron ore or siliceous iron ores  to provide  the
ferrous component of the alloy.  U. S. producers prefer the use of scrap
 because,  in their evaluation,  it  is  a more  economical source  of  iron.   A
Japanese producer  (see  Yahagi  Iron Company  information in the Appendix)

claims that (1") the use of ore permits a "low refining temperature (1220 C)
to be used that makes it convenient to totally seal 50% FeSi furnaces";
(2) the use of ore permits proper charge sizing, minimizes segregation,
and promotes a smooth operation, and (3) the use of ore prevents the con-
tact of quartzite and  improves  the movement of  the charge because it avoids
"the softening and sticking to each other" of the quartzite pieces.
          The subject of iron ore usage versus scrap usage in ferro-
silicon production is controversial in the United States with most (if
not all) of the producers having negative opinions on the use of iron
ores in this application.  However, Battelle researchers suggest that this sub-
ject be studied in a demonstration program.  It is suggested that this
topic be studied in some small experimental electric-smelting furnace
even though there are some negative sentiments in the industry relevant
to the validity of the conclusions that can be drawn from small-scale
          As part of the above experimental program it is also suggested
that lumps of coal char incorporating silica fines be used as part of
the charge to the test furnaces.  The intimate mixture of the major raw
materials to a ferrosilicon operation nay improve the operation of these
furnaces.   The Russians are experimenting with the approach^  '.  It is
suggested  that this coal-char-plus-silica-fines agglomerate could be pro-
duced in the United States with traveling-grate coker units'
          Overall, all ferroalloy companies are interested in upgrading
their reducing agents.   Further study and/or information from the industry
is required before it is possible to judge whether improved reductants
could significantly improve the smoothness of ferrosilicon (and other)
          (B)  At one time, the EPA was considering a requirement
               that the U. S.  ferroalloy industry construct only
               sealed furnaces for those ferroalloy products that
               can be produced in sealed furnaces, as a requisite
               part of the best-demonstrated air-pollution-control
               technology'  .   If this requirement may develop in
               the future, it is suggested that EPA sponsor re-
               search directed at insuring that sealed furnaces are
               safe operations.
*  One such unit is the Peabody Coal Company's traveling-grate coker
   in Tennessee.

          For a ferroalloy company which has learned to produce 50% and
75% ferrosilicon quietly and smoothly in open furnaces (probably with
computer controls), without a need for excessive stoking, it is not
difficult to foresee a future safe operation in sealed furnaces.  However,
if all future ferrosilicon operations are to be conducted in sealed fur-
naces, it is suggested that EPA should be interested in aiding  the
development of controls that will improve the safety of operation of
all sealed ferrosilicon operations.
           Japanese producers of ferroalloys in sealed furnaces maintain
 that operations in sealed furnaces are safe provided that the procedures are
 developed on open furnaces and provided that there are sufficient controls on the
 follow-on sealed furnaces (see Appendix:  MRI Study on Sealed Ferro-
 alloy Furnace Operations in Japan).
           A serious problem is the development of pressure in the high-
 temperature cavity(s) deep in the charge resulting from crusting or
 sealing of the top mix.   At the present time there is no direct signal
 or indication of the contained pressure inside ferrosilicon or
 silicon-metal operations.  Excessive pressure can result either in ex-
 cessive "blowing" (with attendent overheating of the furnace roof and
 high loss of silica fume) or in heaving of mix out of the crucible
 and, in extreme cases, furnace explosions.  All present control methods
 are indirect or secondary in nature and there is no direct method of
 knowing or anticipating when a furnace will erupt.  It is to be expected
 that some U. S. furnace operators will, in the interest of lowering costs,
 be using low-grade local charpe materials.  In these instances, it would
 be particularly important that improved controls be developed  for sealed-
 furnace operations.

          No attempt has been made on this project to develop concepts
for the improved control of silicon smelting operations.   On first evalua-
tion, it would appear that the internal pressure could be monitored by the
use of hollow electrodes and pressure measuring equipment.  However, one
company which has pioneered the use of hollow electrodes  for feeding fines
to electric furnaces is not enthusiastic about this approach to improve
controls.  Here again, if EPA is interested in improved furnace controls,
it is suggested that conferences be held with representatives of the U. S.
ferroalloy companies.  It is considered probable by the Battalia personnel
that research on this subject may be in progress at some  companies.

                         SUMMARY AND CONCLUSIONS

          The major topic of this study is ferroalloy furnace flexibility.
Flexibility is a relative term which has been discussed,  in this report,
in terms of (1) open versus sealed furnaces, (2) small versus large fur-
naces, and (3) by ferroalloy type.  The subject is both complex and con-
troversial and it can be misleading to summarize via generalizations.
However, six general conclusions are presented here as a summary concern-
ing the flexibility issue.

(1)  The U. S. ferroalloy industry has used and will
     probably continue to use furnace flexibility
     (on both a short-term and long-term basis) to
     remain competitive in a changing product-demand
     market in the United States.
(2)  Sealed furnaces are less flexible (i.e., it would
     be more difficult, time consuming, and costly to
     effect a product change) than comparable,
     open-hooded furnaces.
(3)  It has not yet been demonstrated feasible to smelt
     all the common ferroalloys in sealed furnaces.   Sealed
     furnaces have been successfully used (mainly in
     Japan) to produce ferromanganese, si lie manganese,
     ferrochrome, ferrochromesilicon and the ferro-
     silicon alloys with silicon contents of
     75% or less.  The high-silicon-content (75 Si)
     ferrosilicons, silicon metal, and the rare-earth
     silicides have not been commercially prepared under
     sealed-furnace conditions.
     A sealed-furnace operation requires greater charge
     preparation and attention to operating details  than
     a comparable open-hooded furnace on the same
     product.   However, a sealed furnace often exhibits
     a greater operating time percentage and a good  ef-
     ficiency rating (e.g.,  kwhr/tonne,  element recovery
     percentage,  etc.) as a result of the additional
     attention given to the operating practice (see
     Table  1).
(4)   In general,  the newer  and,  in  turn,  larger ferroalloy
      furnaces  are currently used to produce one product
      (or,  as with FeMn and  SiMn,  to make very similar
      products).   However, U.  S.  companies desire  the ability
      to change products  (even in the newer furnaces)  if
     market  conditions require  or warrant such  a  change.
(5)  A large (open or sealed)  furnace is less flexible
      than  a  small furnace.   The time and coat associated

     with  a product change  in a large  furnace is  signifi-
     cantly greater than for the corresponding change in
     a small furnace.
(5)   Combining Conclusions (2) and (5) above, a graphic
     presentation of the degree of flexibility associated
     with  ferroalloy furnaces is as follows:
      Small Opcm  Small Sealed  Large Open  Large Sealed
       Furnace      Furnace      Furnace      Furnace



-* A... .

• - - >

 (6)  A meaningful  comparison between open and  sealed
     ferroalloy  furnaces  In carrying out a product  change
     is  extremely  difficult to  formulate, as  there  are
     multiple differences in furnace  designs and
     operating procedures from  plant  to plant.   Based  up-
     on  the  furnace  design and  operating data presented
     in  Table 1, it  anpears  that  a Riven ferroalloy product
     can be  successfully  made with n  range of operating
     conditions  (such  as  electrode spacing to electrode
     diameter ratio, electrode  penetration,  etc.).

     While  the ranges  of  operating conditions for various  products
     overlap, the  ability to  switch products is presently  dependent
     on  the  operator's ability  to stoke and  observe the  electrode
     penetration and the  charge-level  condition. Thus,  it may be
     that to change  products  in a sealed furnace, the  furnace would
     have to be  opened, the  product change carried  out,  and then
      the furnace returned to operation in  the sealed  condition.   As
     sealed  ferroalloy furnace  operations  are a relatively recent
     development,  the  data base for carrying out a  product change
      (e.g.,  from standard ferromanganese  (FeMn) to  silicomanganese

                (SiMn)  or vice  versa),  Is not simply a matter of changing
                the  charge materials.   A product change requiring a major
                adjustment of smelting  conditions may require that the
                furnace be opened  to make the product changeover.
           These conclusions  are essentially in accord with  the  statements
 concerning ferroalloy  furnace  flexibility presented in a  recent EPA report
 on Standards  of Performance  for Electric Arc Furnace for  the Production of
           Relative  to  the second  specific objective of this  study,  BCL
 personnel  could  not identify any  existing technology and  applied  engineer-
 ing practices which could be applied simply to existing furnace  facilities
 to allow the  achievement of  additional  furnace flexibility in sealed fur-
 naces without further  long-term research and development  efforts.
           However,  BCL personnel  recommend that EPA (1) study and evaluate
 the overall (i.e.,  air,  water, solid waste,  etc.) environmental character-
 istics of  open-hooded  and sealed  ferroalloy furnaces  to establish future
 policy statements,  and  (2) evaluate the  design and  operation and  performance
 of  baghouse associated with  ferroalloy  furnaces  with  the  objective  of de-
 fining methods  of decreasing the  annual  total  emissions from existing and
 future furnaces.
           The BCL-formulated suggestions  for additional programs  which
could be undertaken  by EPA (possibly jointly with the  U. S.  ferroalloy
 industry)  to enhance the  utilization of  sealed  furnaces and/or lower
furnace emission from U.  S. furnaces include:
           (A)  Determine  whether  the substitution of iron-ore pellets
                (as used  in Japan) for ferrous  scrap  (as used  in the
               U. S.) decreases the need  for stoking and improves
               the ease  of operation in  the  smelting of 50 and 75 per-
               cent  ferrosilicon alloys.
and        (B)  Sponsor research directed at  insuring that sealed
               furnaces  are safe operations  through the develop-
               ment of improved furnace-monitoring  techniques and


 (1)   Engineering and Cost Study of the Ferroalloy Industry, James 0.
      Dealy  and Arthur M. Killin, U. S. Environmental Protection Agency,
      EPA  -  450/2-74-008, May, 1974.

 (2)   Background Information  for Standards of Performances:  Electric
      Submerged Arc Furnaces  for Production of Ferroalloys, Vol. 1,
      U. S.  Environmental Protection Agency, EPA-450/2-74-018A.

 (3)   Feik-ral  Register, Vol.  39, Nn. 204, Monday, October 21, 1974.

 (4)   Webster's Seventh New Collegiate Dictionary, G & C Merriam Company,
      Publishers, 1965.

 (5)   Statement on Proposed Air Quality Standards for High Carbon Ferro-
      manganese, Silcomanganese and Calcium Carbide, The Ferroalloy
      Association, Washington, T). C., 1973.

 (6)   A.G.E. Robiette, Electric Smelting Processes, John Wiley & Sons,
      Hew  York, 1973.

 (7)   C. L.  Mantell, Electrochemical Engineering, Fourth Edition, MaGraw-
      Hill Book Company, Hew  York, 1960.

 (8)   V. Paschkis and J. Persson, Industrial Electric Furnaces and Ap-
      pliances, Second Edition, Interscience Publishers, Inc., New York,

 (9)   V. P.  Elyutin, et al, Production of Ferroalloys Electrometallurgy,
      Second Edition, Published for the National Science Foundation, Wash-
      ington,  D. C. and the Department of the Interior, by the Israel
      Program  for Scientific  Translations, 1957.

(10)   J. A.  Persson, "Six-Electrode-In-Llne Smelting Furnaces", Electric
      Furnace  Proceedings, AIME, Vol. 31, 1973.

(11)   F. Andreas, "Design  and Control of Ferroalloy Furnaces", AIEE
      Transactions, Vol. 69,  1950.

(12)   W. M.  Kelly, "Design and Construction of the Submerged Arc Furnaces",
      Metal  Progress, Vol  73, 1958.

(13)   J. A.  Persson, "The  Significance of Electrode-to-Hearth Voltage in
      Electric Smelting Furnaces", Electric Furnace Conference Proceedings,
      AIME,  Vol.  28, 1970.

0/0   Z. B.  Wowk, "Characteristics of a Submerped-Arc Furnace", Electric
      Furnace  Conference Proceedings, AIME, Vol. 22, 1964.

(15)   J.  C.  Wolfe and R.  R.  Pait,  "Interchangeable Shells  on a 29,000  KVA
      Electric Smelting Furnace",  Electric Furnace Conference Proceedings,
      AIME,  Vol.  22,  1964.

(16)   Development Document  for Proposed Effluent Limitations Guidelines
      and New Source  Performance Standards for the Smelting  and Slag Pro-
      cessing Segment of the Ferroalloy Manufacturing Point  Source  Cate-
      gory,  U. S. Environmental Protection Agency, Effluent  Guidelines
      Division, 1973.

(17)   C.  H.  Schanche, N.  Bugge, and 0.  Braaten,  "Today's Trend Toward
      Larger Electric Smelting Furnaces - Some Features in Design and
      Operation", Electric  Furnace Conference Proceedings, AIME,  Vol.  24,
(18)  H. Reiquam, N.  Dee, and P.  Choi,  "Assessing Cross-Media Impacts",
      Environmental Science and Technology.  Vol.  9,  No.  2,  pp.  118-120,
      February, 1975.

(19)  W. R.  Schneider and J. F. Eyrich,  "Operation of  a  75  MVA,  75% Ferro-
      silicon Furnace with Heat Recovery Equipment", Electric Furnace
      Conference Proceedings, AIME,  Vol.  32, 1975.

(20)  W. L.  Wilbern,  "Computer Control  of Submerged  Arc  Ferroalloy Fur-
      nace Operations",  Electric  Furnace  Conference  Proceedings,  AIME,
      Vol. 32, 1974.

(21)  S. I.  Khetrick,  et. al, "Production of 45% Ferrosilicon Using a
      Briquetted Charge", Stal. December, 1966.



                           TABLE OF CONTENTS

                 MRI REPORT COVER LETTER
  TELEPHONE 03/214-1331
  TELEX 222-2287
                                        December 27, 1974
Dr. C. E. Moblcy
Primary Operations Section
Battelle Columbus Laboratories
SOS King Avenue
Columbus, Ohio 43201
Dear Dr. Mobley:

     I am sending under the separate cover the report on
sealed ferro-alloy furnace operations in Japan requested
by your purchase order No. J-827S.

     Our conclusions for this research are as follows.

     1.  It is not technically feasible at present for
         every company to convert the production of 75%
         FeSi to totally sealed furnace.  Judging from
         experiences in Japan, introduction of sealed
         furnaces should be in paralled with the operation
         of open furnaces, through which technical develop-
         ment be achieved and experiences be accumulated.

     2.  50t FcSi can be safely produced with sealed
         furnaces without any pollution problems.
         Especially, the process using ore as raw material
         has been technically established.

     3.  Those interviewees in Japan arc ready to sell
         their know-hows to foreign companies.

     4.  A scaled furnace builder in Japan has actual experi-
         ences in supplying furnaces to foreign companies over-
         seas except for the U.S.  The company considers that
         the export directly to the U.S. ferro-alloy producers
         is rather difficult, since they seem to have their
         own accumulation of technology.  Therefore, the company
         wishes to find an appropriate furnace builder as the
         partner in the U.S. to make its know-hows available
         in the U.S.

                               Very truly yours,
                             /Research Director



Q. 1.   How many totally sealed ferroalloy  furnaces are  there in operation
        in Japan?  What percentage of the total Japanese furnace capacity?
        Is it possible to obtain the total  number (of sealed furnaces), the
        alloy type per furnace, and the rated and actual average power in-
        put and ton of product per year per furnace and  age of furnace?

A. 1-1  At the present time,  39 units of totally sealed  ferroalloy furnaces
        are in operation in Japan.

A. 1-2  We refrain from revealing the production capacity.   However, the
        following table shows the production  during the  period from April
        1 to June 30:

        (Ton)                      Insiders*               Outsiders*

                          Total    666,500 Ton

        The above figures show the highest record in the past history.

A. 1-3  Power rating of furnaces in June, 1973

                        Total furnaces           Sealed furnaces
                     Unit     Rating (KVA)     Unit     Rating (KVA)
He Fe-Mn
He Fe-Cr
M & LC




 A. 1-4  We are unable to give answers to other questions.  It is necessary
        for MRI to conduct an additional market research to obtain proper

 Q.  2.   Of the sealed furnaces, what is the listing of gas cleaning
        equipment?  (bag house, scrubbers, scrubbers plus wet

 A.  2.  We can not reply to this question.  An additional market research
        is required.  However, the characteristics in Japan are;

             1.  Baghouses are used for open tyce furnaces.
             2.  Scrubbers or scrubber + Wet electrostatic facilities
                 are used for sealed furnaces.
 *  We believe that  these  terms are used to designate members and nonmembers
    of the  Japanese  Ferroalloy Association, respectively*

Q. 3.   As  of  1073 it  was  understood  that  there  was  one  sealed  furnace
        (Joethu Electric Furnace Industry  Company) in operation in
        Japan  making 75  percent  FeSi..   This  we understand  is  a  small
        furnace.   Is this  furnace st::ll in operation? Are there other
        sealed furnaces  making 75 percent  FeSi now?   Are there  larger
        furnaces  in operation  or going to  be put into operation on
        this difficult alloy?

A. 3-1   Yes, the  furnace is  in operation at  the  present  time.

A. 3-2   Only 75%-Fe-Si is  produced.

A. 3-3   No  larger furnaces are scheduled to  be put in operation.

Q. 4.   Is  the small furnace (making  75 percent  FeSi)  kept on this
        product or is  it necessary to  switch to  50 percent FeSi (or
        other  products)  for  periods to clean out the silicon  carbide
        that may  have  formed in  the furnace?

A.  A .   Only 75%-Fe-Si is  produced.

Q. 5.   In  all ferrosilicon  and  silicon operations there is an  input
        quartzite to reductant ratio with  a  variable loss  of  silicon
        monoxide.   Are there any attempts  being  made to  measure the
        rate of loss of  silicon  (as &  compound)  in order to adjust
        the input quartzite  to reductant ratio?   Is  it possible to
        control a ferrosilicon or silicon  furnace accurately without
        knowing the silicon  losses hour by hour? (This  question is
        asked  because  in the U.  S. high-silicon  alloys represent a
        problem in manufacture because of  bridging,  high-temperature
        blows, and serious control problems.  On the other hand,  EPA
        descriptions of  high-silicon  alloy production in sealed
        furnaces  in Japan  indicate that there are no particular
        problems,  difficulties,  or system  "upsets".)

A.  5.   There  are no particular  problems,  difficulties or  system
        "upsets",  associated with  quartzite addition.  As you pointed
        out, the  input is  adjusted by  the  weight and no  other ways are

Q. 6-   The U. S.  EPA  visitors to Japan were impressed by  the dust-
        collection efficiency  of various scrubber and wet  electro-
        static precipitator  arrangements on  sealed furnaces in  Japan.
        The use of a wet electrostatic precipitator  following two
        scrubbers  in series  represents a new approach to collection
        equipment  to Battelle  researchers.   In this  regard we have the
        following questions:

        (a)  In the series arrangement of:   scrubber-scrubber-wet
            electrostatic precipitator, what percent of dust col-
            lection (of the total) occurs at the wet. electrostatic


       (b)   Does  this  arrangement  control and  collect  super-
            fine  silica  (Si02)  fume?

       (c)   Is  this  arrangement  more  efficient than  a  well-
            maintained bag  house on  a sealed furnace?  Why?

       (d)   On  ferrosilicon operations is there any  official
            data  on  dust  release to  the atmosphere by  (1)  type
            of  collection system,  (2)  type of  alloy,  (3) weight
            per MW-hour?

       (e)   Are there  any bag house  installations in Japan
            connected  to  ferrosilicon operations both  on
            sealed and open furnaces.   If the  answer is yes,
            what  is  the  relative performance (in terms of
            emissions  to  the atmosphere in grams per M7-hr)?

A. 6.  Since the  manufacturers  concerned have  directly submitted  the  data
       to EPA,  this  Association  knows nothing  about  the contents  of data.

Q. 7.  Do the safety records (or injury data)  of the Japanese Ferroalloy
       Association indicate that sealed furnaces are as safe  as  open

A. 7.  No accidents  have  been recorded.  (Refer Yahagi's  counter

Q. 8.  Are there  any sealed ferrosilicon furnaces  connected  to suction
       bag houses?   If yes, what is  the reported emissions rate  to the

A. 8.  There is one  (1)  unit at  Showa Denko. ^ is  used.   No  data

Q. 9.  Is there any  information  on the emissions of  organic  volatile
       compounds  (by weight and  by compounds)  from  sealed  furnaces?
       It is appreciated that all  the gas from sealed furnaces is com-
       pletely  burned either as  fuel or in flaring,  but what  amount
       of the volatile organics  end  up as input to  the water-control
       system?  U.  S.  ferroalloy companies are concerned  about using
       scrubbers  on  sealed  furnaces  because they occasionally use a
       rather high  percentage of coal and/or wood  chips  in the charge
       both for economic and operational considerations.   The probable
       condensation  of coal and  wood volatiles would appear to have
       many problems in  common  with  coal coking and  tar  processing.
       Does the Japanese Ferroalloy  Association have any  information
       on the amount of  organics that may appear in  the  scrubber water
       on sealed  furnaces?   Does the industry  have  a position or pre-
       ference  for wet versus dry  collection techniques?   To what ex-
       tent are water effluents  considered a pollution problem and
       how is the industry  dealing with the water  effluents?

                                      A-2. A
A. 9.    No data available.
Q. 10.   It is appreciated that of the many advances made in ferroalloy
         operations in Japan, two improvements concern themselves with
         (a) well distributed unsegregated feed to each electrode, and
         (b) monitoring the charge consumed at each electrode.  Given
         that the operator has information of when bridging starts at
         one electrode in a sealed furnace, what action does he take to
         return the consumption of charge material to steady flow conditions?

A. 10.   Since the quality of electrode is excellent and the facilities used
         for charging materials into the furnace are superior, no troubles
         have been experienced.

Q. 11.'  In general, what actions are taken (and what actions are possible)
         in a sealed furnace to improve any "rough" operation that may
         develop?  It is appreciated that drying the mix, sizing the mix,
         and preventing segregation on charging improves the operation of
         any ferroalloy furnace.  However, in the judgment of many, no ferro-
         alloy furnace is actually in close hour-by-hour control.  An er-
         roneous mix change, for example, can result in furnace difficulties
         2 to 4 hours later.  In instances when this does happen, what cor-
         rective actions arc possible?

A. 11.   No particular mix-preparation is made for the materials.  The
         operations used for the closed furnaces are equal to those used
         for the open furnaces.  (Refer Yahagi's counter opinion.)

Q. 12.   In sealed furnaces, are electrodes ever raised to determine the
         length submerged in the charge:?  If this is not possible, what
         method aside from estimated eJectrode consumption is used to es-
         timate electrode length?

A. 12.   We depend on experiment.

Q. 13.   In selecting quartzite for use in the production of silicon-bearing
         alloys is there any qualification other than chemical analysis?  Is
         the resistance to thermal shattering determined?  What is the source
         of quartzite in Japan and are there any local quartzites that are
         not used because of the problems of bridging that may be related
         to this particular quartzite?

A. 13-1  Chemical analysis is used for qualification.

A. 13-2  Since the specifications of supplyers are trusted, no special test-
         ing is carried out and there has been no troubles.

A. 13-3  In Japan, quartzites is produced in Fukushima, Mikawa, Fukui, etc.
         As a matter of fact, quartzites is abundantly produced elsewhere
         in Japan.  In order to save the transportation cost, the locally
         produced quartzites is selected by each suitable ferro alloy manu-
         facturer who is located at different area.


A.  13-4   There  has been  no  problems  of bridging  related  to any particular
          quartzitcs.   (Refer  Yahagi's  answer.)

Q. 14.    It  is  judged, by Battelle-Columbus,  that  the  operation of  a sealed
          furnace  requires more  knowledge,  expertise, and understanding  than
          the operation of an  "open"  furnace making the same  alloy.  Is  this
          correct? If  this  is indeed correct,  it is difficult to  convert an
          operator used to working  an "open" furnace to operating  a  closed
          furnace, making the  same  alloy?   Is  special training required  to
          operate  a totally  sealed  furnace  over and above that given for
          open furnace  operation?   In the United  States,  many operators  are
          very reluctant  to  "give up" their control of  the operation based
          on  the visual observation of  the  top of the charge.  Is  this
          also true in  Japan?

A.  14-1   Yes.

A.  14-2   It  is  easy  to convert  an  operator used  to working an "open"  fur-
          nace to  operating  a  close furnace, making the same  alloy.

A.  14-3   Special  training is  required  to operate closed  furnaces.

A.  14-4   Most of  operations and actions  are controlled by instruments
          which  are observed in  the control room, but the furnace  operating
          conditions  are  determined through the inspection shutter by  visual
          observation,  if necessary.

Q. 15.    Has any  method  been  developed to  measure  or obtain  direct  warning
          of  either (a) deep bridging down  in  the mix and/or  (b) pressure
          build-up in the scaled furnace.   Again  it is  appreciated that  a
          sealed furnace  may require extensive mxi  preparation  (like a
          modern Japanese blast  furnace)  and mix  preparation  improves  opera*-
          tions  in general.   However, in  a  ferroalloy  furnace there  is no
          indirect indication  of internal pressure, or  has this  been develop-
          ed?  If  not,  would this be a  worthwhile topic of research?

A.  15-1   The pressure  within the cover of  sealed furnace is  measured.

A.  15-2   Since the relative position of  the electrode  and the material
          charge chute  is designed  excellently, there has been no  trouble.

A.  15-3   The method  of removing furnace  gases and  dust is very  excellent—
          this is  another reason which  contributes  to  eliminating  troubles.

Q. 16.    Mix preparation in blast  furnace  operation had the  effect of in-
          creasing productivity.  It is appreciated that  ferroalloy furnaces
          do  not work on  the same principle (blowing air into carbon to
          produce  heat  and  reducing agent)  but are  there any  indications
          that mix preparation improves the productivity of  ferroalloy
          furnaces?  We would be interested in the  productivity  figures  on
          two furnaces  (open and sealed,  ordinary mix and prepared mix)  hav-
          ing about the same transformer rating.   Is information of this
          type available  Ln  Japan?


A.  16.  We do not think productivity can be  improved,  but  the working
         conditions can be improved and  pollution  can be  reduced.

Q. 17.   Is the collection and  effective utilization of  the
         carbon monoxide off-gas  from scaled furnaces  the
         economic justification for  sealing  ferroalloy furnaces
         and preparing  the mix  for  these furnaces?  If this is
         correct, would it be possible  to  elaborate on the

         [In this instance, the Battelle researchers appreciate
         the gains made by the Japanese  in installing  the  OG
         system on Japanese LD vessels.]

A. 17.   With most of sealed  furnaces, gases  from  the furnaces are
         collected and used for  boilers.   Some plants sell  such  gases
         to outsiders.  Nippon Kokan plans  to use  this  off-gas from
         sealed furnace for pre-reduction of  raw materials  or chemical

Q. 18.   If absolutely necessary,  at what cost and at what  time
         loss  can a sealed fcrromanganese furnace be converted
         to a  furnace for producing 75 percent ferrosilicon?
         Here  the Battelle investigators appreciate that this
         is not an efficient procedure,  but for the moment  if
         this  had  to be done,  how flexible are sealed furnaces?
         Please elaborate.   (In  this question we are assuming
         that  the  off-grade product made during the switchover
         can be sold somewhere at  a discount).  Our interest is
         mainly in the expense  involved  in adjusting electrode
         diameter  and electrode  working  diameter with a scaled
         furnace. )

A. 18-1  Fe-Mn furnaces cannot be  converted to Fe-Si furnaces.
         This  kind of conversion is  not  practiced  in Japan,  be-
         cause each furnace  is highly specialized.  Even if  con-
         verted,  the depth  of furnace is  different (technically
         speaking).   A large amount  of slag is formed in case
         of Mn operation.

A. 18-2  For the  data of electrode,  refer  to MRI note on  Electrode Diameter

Ferro Alloy
Std. Ferromanganese
75 Percent
Electrode Diameter, cm
Kelley Aono
200 160
150 130
151 132
207 163
195 161
150 137
166 145
210 173
137 125
124 120
127 120
168 145
190 160
173 150


Q. 1.  The EPA representatives have reported that this company is operating a
       12 MW furnace to produce 75 percent FeSi, without special mix preparation.
       It is also reported that this company has determined that the silica fume
       collected at this furnace is an order of magnitude larger in size than
       similar fume from open furnaces.  Our general questions are as follows:

       (a)  Is this small furnace still producing 75% FeSi?

       (b)  To establish operating reliability, will this company
            report for one year:

            (1)  The tons of various alloys made in this sealed furnace?

            (2)  Average MW-hrs per ton of each product.

            (3)  Number of operating hours per one year (power on time).

            (4)  Average MW hour input per ton of product.

            (5)  Range of MW input (high and low).

            (6)  Average number of repair or maintenance hours per week.

       It is appreciated that Joetsu may be varying the power input to this
       sealed furnace depending on power availability.  If this is the case,
       we would appreciate knowing the power availability schedule.

A. 1.  (a)  Yes, as of now, this furnace produces 75% FeSi.

       (b)-l  Only 75% FeSi was produced.

       (b)-2  8100 kWh/tapped metal ton.

       (b)-3  We refrain from answering.

       (b)-4  Refer to (b)-2.

       (b)-5  6000 - 12,000 kw.

       (b)-6  The furnace has not been repaired since 1971.  It is, however,
              scheduled to repair it in 1975.

              Note:  The charging chutes and contact chutes are maintained
                     at all times to prevent abrasion caused by raw material.

       (b)-7  Electric power availability.

              10,000 - 11,000 kw load is the target value.  However,
              it drops down to 6000 kw in the dry seasons in winter
              and summer.

Q. 2.   Several countries have reported difficulties with attempting to
        produce 75 percent FeSi in sealed furnaces.  Besides the expected
        dust build-up problems, the major operating problem was the crusting
        of the mix.  In some instances special cokes were used in an attempt
        to correct this problem.  What action is taken at this company to
        overcome or prevent the typical bridging problems with 75 percent FeSi

A. 2-1  No special cokes are used.

A. 2-2  The problem has been solved by improving the facilities concerned.
        That means that the sealed furnace is provided with the fixed and
        the mobile poking devices and, therefore, there is no trouble of
        "dust build-up".

Q. 3.   Does this company plan a larger furnace?  If they do not plan to
        utilize a large furnace due to current economic or market reasons,
        do they consider a larger 75 FeSi furnace feasible?

A. 3.   Yes, this company plans to build a sealed furnace of 26,000 - 30,000 kva,
        but as of now, this plan is pending due to the increasing power rate
        and the problems in the FeSi market.  Technically speaking, however,
        there is no problem to build a large scale sealed furnace.

Q. 4.   Japanese technical descriptions of this 75% FeSi operation have
        not indicated any special technological factors that would appear
        to have solved the production problems with this alloy whether
        using an open or a closed furnace.  Are there special considerations?
        (A satisfactory answer would be that there is proprietary technical
        information other than engineering modifications.)

A. 4.   There is a difference between the closed furnace and the open furnace
        in regard to the engineering of design itself.  Especially, with the
        experience to operate open furnaces, a highly efficient operation of
        closed furnaces is possible.  However, in order to learn the operating
        procedure in a short period of time, a "know-how" should be used for
        the operation.

Q. 5.   Of particular interest  (to the world) is the reported growth in
        size of the fume from this furnace making collection a relatively
        simple matter.  It would be interesting to know if this fume-
        growth factor has been confirmed  particularly in the fume collected
        routinely at the dust collector?  It is possible to visualize
        growth in particle size of deposits in and on the furnace structure.
        Growth in transit or flight to the collector could, however, represent
        new technical information that would rapidly encourage the use of
        sealed furnaces on 75% FeSi with Japanese know-how.

A. 5.   The characteristics of  sealed furnace is such that the size and
        shape of dust are fundamentally different.  The fact that the
        size of dust is large makes the closed furnace fundametally
        different  from the open furnace in regard to the know-how of
        operation, structure and mechanism of furnace, etc.

A. 5-1  For the shape and size of dust,  refer to Joetsu Report.
        Note:  This report was made by checking the dry dust (before
        soaked in water).

A. 5-2  Can be seen with naked eyes.

Q. 6.   Most high-silicon operations build up a deposit of silicon
        carbide in the bottom of the furnace after some period
        forcing switching to 50% silicon operation for a period.
        Others keep operating by rotating the furnace.  What is
        the practice of Joetsu Electric?

A. 6.   The furnace has not been repaired.  The 50% FeSi will not
        be produced.

Q. 7.   What is the electrode spacing to electrode diameter ratio
        for the sealed 75 FeSi furnace?

A. 7.   Electrode ratio is 1.52 which is used formally for producing

(1)   U.S.S.R.  is using closed furnaces  to produce 75% FeSi.   Please refer
     to the book; "Steel in USSR",  November,  1972,  page 881.   Closed furnace
     melting 75% FeSi in USSR.

(2)   The know-how possessed by Joetsu is  "How to  dispose of  the dust peculiar
     to the closed furnace operation".

     (a)  How to prevent the dust from being accumulated in  the furnace
          cover,  i.e.,  how effectively ventilate  the gas in  the furnace

     (b)  To ensure a stabilized operation by detecting a signal from the
          furnace (related to the proper  operation).

(3)   The closed furnace of Joetsu is manufactured by Tanabe  Kakoki Co.
     (refer visiting report of Tanabe Kakoki Co.).  Tanabe and Joetsu are
     under a special agreement for their  closed furnaces. Whenever
     Tanabe sells a furnace to other company, Joetsu receives a "fee".
     from Tanabe.  Tanabe has already supplied a  FeMn furnace to UCC in
     USA and,  as of now, received about 25 inquiries from companies of
     different countries in the world.*

(4)   A dry type baghouse is used for the  dust collector of closed furnace
     of Showa Denko (Toyama Works).

(5)   With closed furnaces, gases are not  released into the atmosphere and,
     therefore, the environment is kept clean, but special care is required
     in operation because incompletely combusted  gases and CO gas are

(6)   In case of Joetsu, The speed of scrubber is  so slow that it should
     rather be called "washer".

(7)   Since the electric resistance of Japanese cokes is low at a high
     temperature, coal is used partially.  By doing so, tar from coal is
     a problem for the water system.

(8)   Joetsu makes electrodes at their own factory.  No bridging is caused
     by the paste of electrode by controlling the quality and temperature
     of electrodes as well as the quality of paste.

(9)   The raw material is charged to the surrounding area of one (1) electrode
     through four (4) charging chutes.  Also, there is one chute common at
     the center.  Furthermore, since there is a fixed and mobile poking
     device, no abnormal condition occurs in the furnace.
  This is somewhat incorrect as the sealed ferromanganese furnace supplied to
  Union Carbide Corporation by Tanabe is part of a Canadian ferroalloy plant
  operated by Union Carbide, Canada.

(10)   We understand that,  in the USA,  the electrode consumption is
      measured by means of the load cell, but we are doubtful of the
      effects of this method.   In Japan,  we rely on the operator's
      experience and, so far,  have no troubles.

(11)   Joetsu buys quartzites from Fukui and Ashikaga.

(12)   When the melting point of quartzities is low, banking occurs.
      However, there has been no operational accident or trouble by
      using the purchased quartzites.

(13)   In case of Joetsu, the exhaust gas is used for boilers.
      1500 m^/FeSi ton is produced.

(14)   Some companies in Japan are usi.ng wood chips.  Rate of wood
      chips consumed; 350 kg/ton.



          The production results on 50% FeSi on a large sealed furnace are
          impressive.  Possible questions for this company are as follows.

Q. 1.   Is this company considering making 75% FeSi in their furnace?

A. 1.   Yes.  The company is considering making 75% FeSi in the closed furnace.

        Reasons:  Since the FeSi of this company is produced by the ore
        method in this company, the amount of gases (02, H2> generated
        from the product is less than those generated by scrap process.
        Since the FeSi of this company has excellent quality, it has been
        enjoying a good reputation.

        Furthermore, in case of the closed furnace operation, a low
        refining temperature (1,220 C) is used in the ore method of
        this company and, therefore, it is very convenient to totally
        seal the furnace.

Q. 2.   If not, what are the problems to be overcome?

A. 2.	

Q. 3.   It is understood that this operation (50% FeSi) is carried
        out using siliceous iron ore which keeps the fraction of
        quartzite charged from touching and fusing.  Could this company
        produce 50% FeSi routinely using only quartzite and coke?

A. 3.   This company uses special coal.

        Reasons:  If special coal is not used, proper reducing reaction
        can not be ensured.

Q. 4.   Are there any indications of large sized silica fume  in the off-
        takes of this furnace?

A. 4.   Only very fine dry Si02 is obtained.

        Reasons:  Differences  from other types of closed  furnace in
        obtaining Si02 are  (a) refining method,  (b) raw materials,
        and  (c) temperature.

        The offtakes are not clogged with  the Si02-

Q. 5.   Would this company  consider  an operation using  quartzite •  coke,
        coal, and iron scrap?  We would be particularly interested  in
        any information on  the effect on the smoothness of  operation
        using iron or steel scrap.

 A.  5-1  Iron scrap has never been used.

 A.  5-2  Even if the price of the iron scrap becomes  low, we  are most
         certain that the low temperature  refining  of FeSi would
         continue to be more attractive and  beneficial.

 A.  5-3  In order to ensure the smoothness of operation,  the  "preparation
         of material mix" is very important.   In  the  operation  using
         the scrap,  the "mix preparation"  is  made for the sizes of  raw
         materials such as scale,  chips, etc,  and the size of the quartzite,
         beforehand,  but a segregation is  unavoidable when the  materials are
         charged,  thus  the smooth operation being impaired.   This is the
         reason  why this company uses  "Ore".   In  case of  "Ore",  sizing can
         be done properly against the  size of quartzite and other charging

 Q.  6.    It is probable that hanging is encountered in the smelting zones
         at times.   What action is  taken when this occurs.  What danger
         exists  if a hanging condition is  not  corrected?

 A.  6-1   If the  experience,  controlling, etc  are  not  enough and "hanging"
         is encountered,  the operation of  closed  furnace  can  not be performed,
         Therefore,  a company (operators,  workers, etc) must  obtain enough
         experience  and acquire the know-hows  by  operating open furnaces and
         learn how to prevent hanging  before  proceeding into  operating
         closed  furnaces.
        MRI Note:  A large amount of Si02 fume is especially produced in
        a high temperature operation for which scrap, quartzite and coke
        are used as raw material.

        Rate of fume produced is; 300 g/M3 gas x 1500 M3/ton
                                  FeSi = 450 Kg/ton FeSi.

A. 6-2  Around the year of 1970, there was an accident caused by hanging
        of closed furnace (not in Yahagi) .  In this accident, a furnace
        cover was blown up and two operators in the control room were
        killed by the explosion

Q. 7.   Is the silica fume collected at the collection equipment
        monitored to serve as an adjustment on the silica to carbon
        ratio of the mix?

A. 7.   The operation cannot be performed only by changing the carbon
        ratio by monitoring the fume generating rate.  The following
        matters should be taken into consideration.

        (a)  Temperature of gas (in offtakes and furnace cover) .

        (b)  Tapping temperature.


A.  7.   (Continued)
         (c)  Analyzed values of CO,  C02,  and H2  in  the gas.

         ^d)  Contents of Al and Ca  in the metal.

        Also,  the problems should be solved from the standpoint of
        facilities concerned, because the electric  unbalance should
        be prevented by checking the balance of  consumption of raw
        material with reference to  each electrode.

         (Patent No. US 3,499,970)

Q.  8.   If this company uses coal as part of the charge material,
        what effect is noted in the  water treatment system, if any?

A.  8.   Cyanogen and heavy metals are contained  in water and, even when
        the water is passed through  a thickener  and a pre-filter, it
        cannot be reconverted to pure (fresh) water.

        At the present time there is no problem, because a perfectly
        closed cycle is used for the water system.  There is, however,
        a minor problem which stems  from  excessive bubbles produced in
        the water system due to the  effects of tar  from coal.

Q.  9.   If possible, please ask this company for performance figures
        (particularly for the sealed 45 and 60 kva  furnaces) over an
        entire year as follows:

        (a)  Number of operating hours per year

        (b)  Tonnage of alloys (by type produced)

        (c)  Average MW hours per ton of product

        (d)  Average MW during operating period

        (e)  Range in MW (high and low)

        (f)  Electrical energy supply conditions?

A.  9.   Please refer to the Yahagi Report.

        The closed furnace should not be evaluated on the basis of
        the operating hours,  because there is a restriction in the electric
        power supply.


        (1)  The restriction in electric power supply at the
             peak time of day.

        (2)  Restriction in the power consumption in winter and summer.

        (3)  Night factor (38% of total consumption should be alloted
             during 8 hours at  night);  12 hours/month is always used
             for periodical maintenance of closed furnaces.

Q. 10.   What is electrode spacing to electrode diameter ratio for  the
         two sealed 50 FeSi furnaces?

A. 10.   This is the specific "know-how", and the answer to this
         question cannot be obtained .

Q. 11.   Is this company, or the equipment supplier, ready to export
         the know-how and furnace design for producing 50% FeSi under
         other operating conditions?  What are the restrictions?  What
         are the requirements?

A. 11.   Yes, this company is ready to export the know-how, patent,
         and furnace design.  However, the decision shall be made
         according to the possibility of territorial conflict with
         the would be licensee in marketing the product.  That is,  the
         decision shall be made on t.he case-by-case basis.

Q. 12.   Is the quartzite given any qualification other than chemical
         analysis?  Is spalling resistance determined?

A. 12-1  The softening point of quartzite within the range of 1,600 C -
         1700 C is measured.

         Any quartzite which is softened quickly can not be used for
         Si refining.

A. 12-2  Thermal shocks  are also tested.

         Other comments:

         (1)  It is said that, in general, furnaces for FeSi and for FeMn
              cannot be  substituted for  each other.

         (2)  The  (electric) power density of refining zone for FeMn is
              35W/cm2, and 85W/cm2 for FeSi.

              We believe that  if the  furnace capacity and  structure are
              appropriately designed, a  mutually substitutable furnace
              can be constructed, but not economical.

General Consideration design
for Closed Furnace

          When a closed furnace is used, a technical bottle neck is how to
handle CO gas produced from the carbon contained in the raw material.   On the
other hand, since the air flow-in rate is low, the absolute amount of disposed
gas is small and the amount of produced gas is only a few times more than
that of the theoretically estimated gases.  Because of these reasons,  the
dust collecting system is much smaller when compared with that of the open
electric furnaces.  However, since the concentration of CO gas is 50 - 70 percent,
it is unavoidable to use a wet type dust collector as long as the technical level
remains in the present condition.  The reasons why the wet type is used are:
When a dry or an electric method is used, air may flow-in, air lock may occur
when dust is taken out, the rotating units may be overheated, mechanical shock
may occur, gas explosion may be caused by an arc, etc.  Now, the matters
concerning the plan, operation and maintenance of dust collecting system shall
be discussed with reference to the above matters.  First, "dust" is produced
as follows:  the raw material is sent from the mixing tank to the plural
number of raw material throwing tanks installed on the top of furnaces.  Then,
a fixed quantity of raw material is alternately thrown into the furnace through
the air locked gate provided on the lower part of the above tanks.  Generally
speaking, th« raw material in the furnace consists of (from the top to the
bottom) the layer drying, preheated layer, the slag and molten metal on the

          The exhaust gas rises in the reverse direction of the above described
sequence of layers and the heat of heat-gns is exchanged with that of the
layers of raw materials and the lower raw material in the raw material throwing-
tanks.  Then, it filters the coarse grains of dust in the raw material layer
and sends the gas and fine grains to the dust collecting system provided
outside.  In this case, if the pretreatment of raw material is insufficient,
the water content in the raw material moves into the gas, condenses on the
inner walls of pipes and may cause a clogging in piping.  In order to prevent
this trouble, the temperature of inner wall of pipe and dew point' of gas
should be calculated with reference to the gas temperature, gas calories,
ambient temperature, etc and necessary measures such as heat-insulation, heating,
etc of external walls of piping should be taken.

          Besides the above matters, there is another effective method.  Let
water flow through the piping to protect the inner wall of pipe with a film
of water, preventing the dust from sticking.

          In order to ensure reliability and effects of operation of the wet
type dust collecting system, it is necessary to combine different types of
dust collectors of different functions, thus making it possible to cope with
troubles of every possible cause.  For example, a combination of Taizen washer,
Venturi scrubber, mist catcher, etc or a combination of a low pressure Venturi
and a high pressure Venturi-cyclime scrubber.

          When  combining  a  low pressure Venturi, Taizen washer,  filling  tower,
 etc,  it  is necessary  to take  into  consideration  the characteristics  of each
 element  referring  to  the  limitation  of the  amount  of weter  used  by  the factory,
 the capacity  of waste water treatment, etc.  When  designing a  system and  if
 water and dust  should be  removed from each  dust  collector,  the diameter  of
 grain, physical properties  of dust at each  dust  collecting  stage, etc should
 be predicted  properly and the system should be provided with such a  function
 which can fully cope  with the sinking speed, condensing property, PH values
 of drained water,  etc.  Generally, in case  of the  wet type  dust  collector,
 the ferroalloy  dust is often  hardened while sinking and cannot be discharged
 to the outside  and, when  water is  used for  recycling, the substances dissolved
 in the liquid are  educed  by the chaige in the density and humidity,  thus
 often resulting in clogging the syscem.  From this point of view, the amount
 of water to be  added  should be decided by taking into consideration  the
 economy  of factory water  as well as  the safety in  operation.  Another important
 matter is that  each dust  collector should be provided with  a bypass  route so that,
 when  repairing  or  replacing the equipment,  it is not necessary to stop the
 entire system.  It is quite natural  that the density of discharged "dust"
 should be kept  lower  than the value  prescribed by  laws concerned.  In case of
 a closed furnace,  CO  gas  is combusted in the flare-stack before  it is
 released into the  atmosphere.  It  is possible that the aforementioned heavy
 metals,  other solid substances, dissolved substances, harmful substances
 such  as  cyanogen may be mixed in the drained water and, therefore, this is
 another  problem that  should be taken into consideration when designing the
 drained water disposing facilities.

          Cyanogen was detected by a public research organization a  few
 years  ago.  Cyanogen  gas  is produced as the carbonic substance reacts by
 heat  in  the closed furnace.   This phenomenon is not observed in  open  furnaces.
When  designing  a wet  type dust collecting system,  it is inevitable to take
 necessary measures that can remove smoothly the drops of liquid  from the inside
 of equipment.   This is because the driving  source  is a gaseous body of small
 density which is the characteristics of this system.

 Characteristics of Dust and System for
 Ferro-Cr and  Ferro-SiCr Open  Furnace

          In  case  of Cr,  the  temperature is at an  intermediate level of FeSi
 and FeSi and  FeSiMn but the dust is comparatively coarse and large -- for
example, with the  centrifugal circulation type multi-cyclone with Z = > 150,
 the exhaust gas density was 0.1 - 0.1 G/Nm3.  In case of an open electric
 furnace, damage on electrodes, calcination of paste, condensation point
 of carbon, tar volatilization, etc are taken into consideration and in case
of the actual example, the  load of electrode calcination (approximately
 1/3 of power consumption required foir melting) and a back-filter used for
filtering only  the amount of  occasionally produced disposed gas should be
provided.  In case of Cr,  there are many large coarse particles at the
preduster and when Z = > 150,  it is just as  described above, but with silicon
steel plate  multicyclone,  the  conical section was damaged  in approximately
6 months.

          As a general theory, Z = 30 of the common idea is to use a
gravity settling type preduster and provide a back-filter on the rear stage.
For the dust discharging system of the first stage dust collection, it is
better to use a double damper type air lock method rather than using a rotary
type.   However, if the effects of preduster is insufficient, the runners
of blower will wear excessively, thus unabling to perform operation and,
therefore, it is necessary to keep spare runners at all times.  Depending
upon the temperature at the inlet of preduster, it may be better to use a
multicyclone and apply a rubber lining is force-fitted into the back-filter.
When FeSiCr Is compared with FeCr, FeSiCr has finer particles.  In case of
a wet type, the filtering thickness is approximately < 2m/m by Oliver Test,
thus making it difficult to apply it for industrial use.  Water can be
separated from dust very quickly but is condensed and becomes very hard.
It was, therefore, difficult to dispose of it mechanically.

          From the above point of view, the dry method is the best way to
dispose of FeCr and FeSiCr.

(Pit osc read liiilnictiviH on the rci t r