EPA-450/3-80-041
A Review of Standards.of Performance
     for New Stationary Sources  —
     Ferroalloy Production  Facilities
             Emission Standards and Engineering Division
             U.S. ENVIRONMENTAL PROTECTION AGENCY
                Office of Air, Noise, and Radiation
              Office of Air Quality Planning and Standards
             Research Triangle Park, North Carolina 27711
                     December 1980

-------
                                    n nm M» nil minimi
This report has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air, Noise,
and Radiation,  Environmental Protection Agency, and approved for publica-
tion.  Mention of company or product names does not constitute endorsement
by EPA.  Copies are available free of charge to Federal employees,  current
contractors and grantees, and non-profit organizations - as supplies permit
from the Library Services Office, MD-35, Environmental Protection Agency,
Research Triangle Park, NC 27711; or may be obtained, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
VA 22161.                                                   F   B.
                 Publication No. EPA- 45073-80-:041
                                                It11
                                                              I'   in
                                 11

-------
                                  CONTENTS
Figures
Tables
1.   Summary
2.   Introduction
     2.1  Purpose and scope
     2.2  The industry
     References for Section 2
3.   Current Emission Standards for Electric Submerged-Arc
      Ferroalloy Production Facilities
     3.1  Federal New Source Performance Standards
     3.2  State emission limitations for ferroalloy plants
     3.3  Emission regulations for ferroalloy facilities in
           other countries
     References for Section 3
4.   Industry Status
     4.1  Present capacity
     4.2  Plant location
     4.3  Industry growth
     References for Section 4
5.   Test Results
     5.1  Particulate emission test data
     5.2  Visible emissions
     5.3  Emissions of organic matter
     5.4  Other atmospheric emissions
     References for Section 5
Page
   v
  vi
   1
   8
   8
   8
  11

  12
  12
  13
  14
  20
  21
  21
  28
  30
  36
  38
  38
  41
  44
  50
  56
                                      iii

-------
                            CONTENTS (continued)
6.   Production Processes and Best Available Control Technology

     6.1  Ferroalloy production processes
     6.2  Control technology
     6.3  Control of tapping emissions
7.
8.
References for Section 6

Conclusions
7.1  Industry growth
7.2  Process changes
7.3  Control equipment
7.4  Emissions
7.5  Source testing method evaluation

References for Section 7

Recommendations

8.1  Changes in regulations
8.2  Areas of further study
APPENDIX
Page

   58

   58
   63
   68

   73
::,:t.:


   75
   75
   76
   76
   80

   83

   84

   84
   84

   85

-------
                                    FIGURES
Number
 4-1
 4-2

 6-1
 6-2
 6-3
Trends in Domestic Production and Imports
Summary of Ferroalloy Industry Capacity and Demand
  Estimates
Ferralloy Production Process
Ladle/EOT Crane Fugitive Emission Collection System
Air-curtain Fugitive Control System
Page
  22

  34
  60
  71
  72

-------
                                    TABLES
Number                                                                Page
 1-1      Electric Submerged-Arc Furnace Capacity for Ferroalloy
            and Calcium Carbide Production, by State                     2
 1-2      Estimated Organic Emission Rates After Particulate
            Control Device                                               5
 2-1      List of Major Ferroalloys According to Manufacturing
            Processes                                                   10
 3-1      State Air Pollution Control Regulations in States with
            Ferroalloy Operations                                       15
 3-2      Comparison of Allowable Emissions from Ferroalloy Plants
            Under NSPS and State Regulations                            17
 3-3      Factors for Converting to Process Weight Rate                 18
 3-4      Comparison of Emission Standards for Selected Facilities      19
 4-1      Reported Ferroalloy Consumption as a Function of Steel
            Production                                                  22
 4-2      Ferroalloy and Calcium Carbide Plants Using Electric
            Submerged-Arc Furnaces                                      25
 4-3      Ferroalloy Product Mix in 1979                                28
 4-4      Electric Submerged-Arc Furnace Capacity for Ferroalloy
            and Calcium Carbide Production, by State                    29
 4-5      Ferroalloy Plants Using the Metalothermic and Electrolytic
            Process                                                     31
 4-6      Two Ferroalloy Growth Scenarios for 1980-1985                 32
 4-7      Estimates of Ferroalloy Demand in 1985                        33
 5-1      Compliance Test Data                                          39
 5-2      Types of Furnaces Tested and Their Emission Controls          40
                                      VI

-------
TABLES (continued)
Number                                                                Page
 5-3      Partieulate Emission Data from EPA Environmental Assess-
            ment Studies                                                42
 5-3      Visible Emission Readings                                     43
 5-5      Organic Emission Data           v                             45
 5-6   .   Organic Emission Rates and Reduction Materials Charged to
            Furnaces                                                    47
 5^7      PoTyriuclear Organic Compound Emissions                        49
 5-8      Organic Concentrations in Scrubber Water Discharge Streams    51
 5-9      Inorganic Constituents in Furnace Exhaust Gas                 53
 5-10     Furnace Emission Rates of Selected Metals                     55
 6-1      Reported Design Data for Fabric Filter Systems                64
 6-2      Reported Design Data for Scrubbers                            67
 7-1      Estimated Organic Emission Rates After Control                77
 7-2      Estimated Ground Level Concentration of Organics from
            a 20-MW Furnace                                             78
 7-3      Estimated Emissions of Hazardous Trace Metals after Control   80

-------

-------
                                ''.SECTION 1
                        ;-.          SUMMARY

     The current New Source Performance Standard (NSPS) for selected ferro-
alloy production facilities, which was promulgated on May 4, 1976, applies to
all facilities built or modified after October 21, 1974 (the original proposal
date).  This NSPS limits particulate emissions from electric submerged-arc
furnaces to 0.45 kg per megawatt-hour (0.99 Ib/MW-h) for silicon-based alloys
and 0.23 kg/MW-h (0.51 Ib/MW-h) for ma'nganese and chrome-based alloys, and
calcium carbide.  It also regulates visible emissions from the furnace emis-
sion control device, the furnace tapping process, and any associated dust
handling equipment and specifies that carbon monoxide emissions cannot exceed
20 percent by volume.
     This report presents the results of a study to determine whether the
current NSPS should be revised.  The findings are based on information from
manufacturers, regulatory agencies, and the open literature.

1.1  THE INDUSTRY
     In 1971 about 2.1 Tg (2,331,000 tons) of ferroalloys (including scrap and
recycled materials) were produced in the United States and an additional 0.35
Tg (388,000 tons) were imported.  At that time, ferroalloys were produced at
44 locations, 145 electric furnaces were utilized, and average production was
increasing by 1.5 percent per year.  Another 13 furnaces were used to produce
calcium carbide.  Since 1972, however, domestic production has declined
drastically, to a level of 1.6 Tg (1,830,000 tons) in 1979, and imports have
increased to approximately 1.2 Tg (1,280,000 tons).  This 21.5 percent decline
in production has been accompanied by a reduction in the number of electric
submerged-arc furnaces, to 89 for ferroalloys and 7 for calcium carbide.
Table 1-1 shows the current electric submerged-arc furnace capacity for
ferroalloy and calcium carbide production by state.

-------
             TABLE-1-1 ;• '"ELECTRIC -SUBMERGED-ARC FURNACE' CAPACITY

           FOR FERROALLOY AND CALCIUM CARBIDE PRODUCTION, BY STATE
State
Alabama
Iowa
Kentucky
New York
Ohio
Oklahoma
Oregon
South Carolina
Tennessee
Washington
West Virginia
Total
Plants
6
2
3
1
6
1
4
1
3
2
2
31
Capacity, MW
189
50a
182b ,
45
554C
15
57
80
123
54
257
1606
Furnaces
Ferroalloy
10
2
6
, 2
32
0
5
2
12
6
12
89
Carbide
0
1
2
0
2
1
1
0
0
0
0
7
 15 MW for calcium carbide (CaC2).

385 MW for CaC?.
rt             ^

"A portion of capacity at the Ashtabula plant can also be used to make CaC?.
                                      2

-------
     Because this decline in domestic production,has,resulted in no new plants
being built since the NSPS was proposed, no plants are currently subject to
this standard.  Although imports have recently leveled off, no plans have been
made to construct any new facilities in the near future.
     Growth of the currently static domestic ferralloy industry will be af-
fected most by the level of imports and steel production.   If imports level off
at their current rate and the steel industry demand increases, an optimistic
growth picture would indicate the need for some new capacity by about 1985.  If
imports continue to supply about 50 percent of demand and growth in steel
production is slow, no new facilities will be required in the next 5 years.

1.2  FERROALLOY MANUFACTURING PROCESSES AND CONTROL.TECHNOLOGY
1.2.1  Manufacturing Processes
     The processes used to produce ferroalloys are electric submerged-arc
furnace (at 31 locations), metalothermic (at 8 locations), and electrolytic
(at 4 locations).  Vacuum and induction furnaces,  which are in limited use,
are alloy refining processes for the production of specialty metals.  The
widespread use of the electric submerged-arc furnace accounts for about 95
percent of production and the copious amount of fume it generates make it the
only process of significance as far as air pollution is concerned.
     Based on the configuration of the hood, electric submerged-arc furnaces
are categorized as open, semisealed or mix-sealed, and closed.  In the open
furnace, which accounts for 87 percent of production capacity, the electrodes
extend through a canopy hood located above the upper rim of the furnace.
A 2- to 2.7-m (6- to 8-ft) gap between the furnace and the hood allows large
amounts of ambient air to be drawn into the hood.   As the air combines with
the hot gases, the carbon monoxide and most of the organic compounds are
burned and the exhaust gases are cooled and diluted.  The large opening around
the hood also allows fumes to escape from the furnace during upset conditions
or when insufficient exhaust draft is maintained.
     In the semisealed furnace the hood fits tightly onto the furnace, and raw
materials are charged through annular gaps around each electrode.  The hood is
exhausted to an air pollution control device, but some fumes may leak out
through the mix at the electrodes.  No gap exists between the hood and furnace

-------
and no dilution of the exhaust occurs; therefore no fumes can escape at that
point.  Because the furnace is closed, it cannot be easily stoked from the
outside.  This is the reason it is not in widespread use and accounts for
only 12 percent of total capacity.
     The closed furnace, which is used at only one plant in the United States,
has a tightly fitted hood and the raw materials are fed through sealed chutes
to the furnace.  All fumes are exhausted to an air pollution control device.
From an air pollution control standpoint, the closed furnace is the most
desirable because all the fumes are exhausted through the control system.
Total gas volume is only 2 to 5 percent of that from an open furnace.  High
silicon alloys are more difficult to produce in a closed furnace, however,
because they tend to bridge over if not stoked and the closed nature of the
furnace makes stoking much more difficult.
     The only innovation in processing has been the introduction of a split-
ring rotating furnace in which the upper and lower part of the furnace rotate
at different slow speeds.  Because this rotation minimizes the need for
external stoking, this type of furnace would be a better process for the
production of higher silicon-bearing alloys.
1.2.2  Control Technology
     Particulate emissions are controlled by such conventional control systems
as fabric filters, high-pressure-drop scrubbers, and occasionally electro-
static precipitators.  Fabric filters are by far the most common, especially
on open furnaces where large volumes of exhaust gas must be treated.  They
also are frequently used to reduce the particulate collected by control  sys-
tems on tapping operations.
     Fabric filter systems with glass filter or Nomex bags have achieved par-
ticulate collection efficiencies in excess of 99 percent and visible emissions
with less than 10 percent opacity.  Bag life is on the order of 1 to 2 years,
and air-to-cloth ratios are about 18-38 m/h (1 to 2 ft/min).
     High-pressure-drop scrubbers with pressure drops of 13.7 to 23.9 kPa (55
to 96 in. of water) also achieve particulate collection efficiencies of
about 99 percent.  Scrubbers are in more common use on closed and semisealed
furnaces, where the volume of gas to be treated is much smaller than that from

-------
an open furnace.  Scrubbers also decrease emissions of organic matter by from
16 to 97 percent.
     One electrostatic precipitator is currently in use on an open-type ferro-
alloy furnace, in this country.  The flue gas is conditioned with ammonia to
achieve a high level of control efficiency at this installation.
     Flares are used on semi sealed and closed furnaces to reduce carbon
monoxide emissions.  Data on their efficiency in reducing organic emissions
(both gaseous and particulate) are not available, but some reduction can be
expected.
     Control of furnace tapping emissions requires extensive hopding around
the tapping and pouring operation, from Which the collected fumes can be
directed to a fabric filter control system.  Installation of hoods on existing
furnaces is a problem at times because of site-specific space restrictions,
whereas adequate hooding can be incorporated into the installation of new
facilities.  ... ,'

1.3  ATMOSPHERIC EMISSIONS
1.3.1  Organic Emissions
     Since the NSPS was proposed, limited additional measurements have been
made as part of EPA's environmental assessment studies to quantify other
pollutants emitted by the electric submerged-arc furnaces in the ferroalloy
industry.  In addition to measuring total organic emissions, one test on a
semi sealed furnace and two tests on a closed furnace also identified and
quantified individual polynuclear aromatic hydrocarbons.  Table 1-2 summarizes
the organic emission data from these studies.
             TABLE 1-2.  ESTIMATED ORGANIC EMISSION RATES AFTER
                         PARTICULATE CONTROL DEVICE
Furnace type
Open
Semi sealed
Closed
- ; -
Control device
Fabric filter
Venturi scrubber
Scrubber
Venturi Scrubber
Total organic emissions,
kg/MW-h (Ib/MW-h)
0.20 (0.44)
0.29 (0.63)
0.15 (0.33)
0.01 (0.022)

-------
      When these emission rates are multiplied by the respective  furnace
 capacities, estimated total  annual organic emissions from electric  submerged-
 arc furnaces are approximately 2220 Mg (2440 tons),  based on  70  percent
 utilization.  Data on polynuclear organic matter (POM)  show that these com-
 pounds  represented from 8.3  to 75 percent of the total  organic emissions  after
 a scrubber and before the flare.   The  single test on a  semisealed furnace
 After a scrubber showed that benzo-a-pyrene (BaP) accounted for  0.84 percent
 of the  organic emissions at  a concentration of 1.64  mg/m3.  The  tests on  the
 closed  furnace showed no BaP to be present.
      Ground-level  concentrations  of organics were estimated by use  of an
 atmospheric dispersion model  (PTDIS).   These estimates  showed that  the highest
 concentrations occurred 1  to 2 km from the stack and that the maximum 24-hour
 concentrations for a  20-MW open furnace with a scrubber were 1.0 to 1.6 yg/m3.
      Additional  quantitative data on specific organic emissions  are needed
 for a better determination of the magnitude  of this  potential problem.
 1.3.2  Particu|ate Emissions
      Data  from tests  to determine compliance of existing  plants  with state
 emission regulations  showed  particulate emission  rates  in  the range of 0.07 to
 0.20  kg/MW-h (0.16 to  0.44 Ib/MW-h), which means  they could also meet the
 NSPS.   Particulate emissions  were determined by use  of  EPA Methods 1 to 5.
 These methods  proved  acceptable but the safety problems inherent  in sampling
 high  concentrations of carbon  monoxide  in  semisealed and closed  furnaces must
 be  recognized.  During  the EPA environmental  assessment studies, particulate
 emissions  after the control system ranged from 0.016 to 0.77 kg/MW-h (0.035 to
 1.7 Ib/MW-h).  As  operated during these tests,  two of the six furnaces could
 not comply with the NSPS.
 1.3.3  Other Emissions
      Emissions of  trace metals were also measured in the EPA environmental
 assessment studies.  These tests were made on an open furnace, at the inlet
 to a  fabric filter control system, and on a closed furnace, before and after a
 venturi  scrubber.  The major metallic emissions were composed of the product
being smelted.  Based on limited data,  the following emissions were  estimated
after particulate control.

-------
            Copper
            Mercury
            Lead
Estimated emission rate,
       nig/MW-h
         62
        0.02
         17
        8.5
         4
        2.7
     Thus, on a 20-MW furnace, for example, the daily emission rate for arsenic
would amount to about 30 grams (0.066 Ib).
     Sulfur oxide emissions, though fairly dilute at 20 to 80 ppm, can amount
to 0.5 to 2.0 kg/MW-h (1.1 to 4.4 Ib/MW-h).  Correlations between the sulfur
in the feed materials and sulfur oxide emissions are not available.  Nitrogen
oxide emissions are very low because of the lack of oxygen in the reaction
zone of the furnace.                   ,

1.4  CONCLUSIONS AND RECOMMENDATIONS
     The ferralloy industry is currently characterized as an industry with no
production growth and no new construction or modifications.  Since the early
1970's, production has declined by about 33 percent and the number of active
electric submerged-arc furnaces has decreased by about 39 percent.  No changes
in process or emission control technology have occurred since proposal of the
NSPS in 1974.
     Existing plants largely comply with state regulations governing par-
ticulate arid visible emissions except during tapping and pouring operations,
when visible emission limits are sometimes exceeded.
     The measurement of organic emissions, including polynuclear organic matter
(POM), from ferroalloy furnaces shows that these compounds are being emitted
from the furnace.  Additional information on organic emissions should be ob-
tained for better quantification of these emissions and for determination of
the efficiency of particulate control devices for these compounds.
     Because the industry is not growing and process and control technologies
have remained the same, no change in the NSPS is recommended at this time.
               .                       7

-------
                                  SECTION 2

                                 INTRODUCTION
2.1   PURPOSE AND SCOPE
     On May 4,  1976, the Environmental Protection Agency  (EPA) promulgated New
Source Performance Standards  (NSPS) for selected ferroalloy production facili-
ties (41  FR 18497).   These standards establish limits for and require testing
and reporting of particulate  emissions from electric submerged-arc furnaces
(producing specified alloys)  that were built or modified  after October 21,
1974.  The Clean Air Act Amendments of 1977 require that  the EPA Administrator
review and, if  appropriate, revise such standards every 4 years [Section
     This report presents the results of a review of the NSPS for ferroalloy
facilities.  The review covers recent and projected growth of the ferroalloy
industry and describes changes in process and control technology since NSPS
promulgation.  Because no ferroalloy facilities are currently subject to NSPS,
inquiries concerned enforcement aspects of complying with state regulations,
and this review is based on compliance test results from various states,
recent EPA environmental assessment studies, information from the literature,
and discussions with representatives of industry, control equipment vendors,
EPA regional offices, and state agencies.  The information obtained from these
sources was then analyzed to determine whether revisions to the NSPS are
required at this time.
     Available information was also gathered on other atmospheric emissions
from ferroalloy furnaces to determine if other portions of the act, such as
Section lll(d) or 112, should be implemented to regulate emissions.

2.2  THE INDUSTRY
     When the NSPS were proposed, 44 ferroalloy production facilities were
operating in the United States.2  In 1971,  the industry produced 2.11 Tg
                                     8

-------
                          o
(2,331,000 tons) of alloy,  and the growth rate was about 1.5 percent per
year.   In 1967, electric submerged-arc furnaces emitted 90.7 Gg (100,000
tons) of particulates per year and were by far the major air pollution source
in this industry.
     Table 2-1 lists the major ferroalloys and their manufacturing processes.
Although calcium carbide is not a ferroalloy, it is included with this indus-
try category because it is made in submerged-arc furnaces, has similar emis-
s"on characteristics, and is sometimes produced at ferroalloy plants.
     As discussed later in Section 4.1, in 1979 the industry produced only
about 1.5 Tg (1,650,000 tons) of alloy at 42 locations, an indication that the
industry growth rate has declined sharply in recent years.  The industry has
been characterized by the lack of any new facilities and the large-scale
installation of air pollution control equipment on existing furnaces to comply
with state emission control regulations.

-------
            TABLE 2-1.  LIST OF MAJOR FERROALLOYS ACCORDING TO
                           MANUFACTURING PROCESSES
     Process
          Ferroalloys
Submerged-arc furnace
  (Carbothermic)
Exothermic or metallothermic




Electrolytic


Vacuum furnace

Induction furnace process
Silvery iron
50% Ferrosilicon
65-75% Ferrosilicon
Silicon metal
Calcium silicon
Silicomanganese zirconium (SMZ)
High-carbon (HC) ferromanganese
Silicomanganese
Ferromanganese silicon
Charge chrome and HC ferrochrome
Ferrochrome silicon
Calcium carbide

Low-carbon (LC) ferrochrome
LC ferromanganese
Medium-carbon (MC) ferromanganese
Chromium metal, FeTi, FeV, FeCb, and FeMo

Chromium metal
Manganese metal

LC ferrochrome

Ferrotitanium
                                     10

-------
                         REFERENCES FOR SECTION 2
1.



2.


3.
4.
U.S. Environmental Protection Agency.  Code of Federal Regulations, Title
40, Chapter I, Part 60.  Washington, D.C., Office of Federal Register.
May.4, 1976.
Fisher, F.L.  Ferroalloys.  In:  Minerals Yearbook, Volume I.
ington, D.C., Bureau of Mines.  1971.
Wash-
Dealy, J.O., and A.M. Kill in.  Engineering and Cost Study of the Ferro-
alloy Industry.  U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina.  Publication No. EPA-450/2-74-008.  May 1974, p.
II-l.

Vandergrift, A.E., et al.  Particulate Air Pollution in the United
States.  J. Air Pollution Control Assoc. 21:326.  June 1971.
                                     11

-------
                                 SECTION 3
             CURRENT EMISSION STANDARDS FOR ELECTRIC SUBMERGED-
                   ARC FERROALLOY PRODUCTION FACILITIES
3.1  FEDERAL NEW SOURCE PERFORMANCE STANDARDS
     The final emission standards for ferroalloy production facilities were
published in the Federal Register on May 4, 1976 (41 FR 18498)1, and applied
to all facilities on which construction or modification began after October
21, 1974 (the original proposal date; 39 FR 37465).
     The promulgated standards limit particulate matter and carbon monoxide
emissions from electric submerged-arc furnaces and limit particulate matter
emissions from dust-handling equipment.  Emissions of particulate matter,from
the control device cannot exceed 0.45 kg/MW-h (0.99 Ib/MW-h) on furnaces
producing high-silicon alloys (in general) and cannot exceed 0.23 kg/MW-h
(0.51 Ib/MW-h) on furnaces producing chrome, manganese, and low-silicon
alloys.  The opacity reading of emissions from the control device must be less
than 15 percent on either type of furnace, and any 6-minute period during
which the average opacity exceeds 15 percent must be reported.  The regulation
requires that collection hoods capture all emissions generated within the
furnace (no visible emissions except during furnace upset conditions) and
capture all tapping emissions during at least 60 percent of the tapping time.
The concentration of carbon monoxide in any gas stream discharged to the
atmosphere must not exceed 20 volume percent on a dry basis.  Opacity readings
of emissions from dust-handling equipment must be less than 10 percent.
Monitoring of the opacity of emissions from the furnace control system and the
gas flow rate through the particulate collection system must be continuous.
     Because the standard covers only specific alloys (those constituting the
major products in the ferroalloy industry), it affects only facilities using
electric submerged-arc furnaces.  The electric submerged-arc process is the
major production method in this industry.

                                     12

-------
     The alloys affected by the standard and the applicable regulation are as

follows:

     Alloys subject to 0.45-kg/Mw-h limit

          Silicon metal
          Ferrosilicon
          Calcium silicon
          Siliconmanganese zirconium        ,,.

     Alloys subject to 0.23-kg/MW-h limit

          Ferrochrome silicon
          Silvery iron (ferrosi1 icon with less than 30% silicon)
          High-carbon ferrochrome
          Charge chrome
          Ferromanganese
          Silicomanganese
          Ferromanganese silicon
          Calcium carbide
  '                    '          - ,  "                          •               '
     Other ferroalloys, such as those containing vanadium, titanium, nickel,

tungsten, beryllium, and aluminum, are made in relatively small quantities by

metalothermic processes.  Ferrophosphorpus  (FeP) is produced as a slag by-
product of the manufacture of phosphorous by electric arc furnaces, but the

nature of the emissions and the control system are very unlike those of ferro-
alloy processes; therefore this product is not included under this standard.

     Particulate emissions must be tested by EPA Method 5, and carbon monoxide
(CO) concentrations should be measured by an Orsat according to procedures in
Method  3.  The electrical input to the furnace must be continuously measured
during  the test periddV and the type of product must be identified.   In

addition, the opacity  of the control system exhaust and the volumetric flow  rate

of  the  furnaceexhaust must be monitored continuously.
 3.2   STATE  EMISSION  LIMITATIONS  FOR  FERROALLOY  PLANTS

      State  regulations  pertaining  to industrial processes were  reviewed for

 those states  in  which ferroalloy plants  are  located.  These  state  regulations

 are  largely based on the  weight  rate limitations  on similar  processes.

      The most stringent regulations  are:

       E = 3.59 P**52  for charge rates less  than  30 tons/h and

       E = 17.31  p'16 for  charge  rates greater than 30 tons/h


   •;: •      '   '         "'  •    •       13

-------
where E = allowable particulate emissions, Ib/h
      P - charge rate, tons/h.
The least stringent process weight regulations are:
      E = 4.10 P'   for charge rates less than 30 tons/h and
      E = 55.0 P   -40 for charge rates greater than 30 tons/h.
Table 3-1 summarizes the state regulations in states that have facilities with
submerged-arc ferroalloy processes.
     Allowable particulate emissions for the alloys subject to NSPS have been
calculated on the basis of a 30-MW furnace and compared with the most strin-
gent and the least stringent state limitations (Table 3-2).  It is apparent
that the type of product is not a determining factor for allowable emissions
in state regulations; they are based only on the process weight rate.
     The data in Table 3-3 were used to estimate allowable emissions for each
of the metals or alloys on a pounds-per-hour basis for easier implementation
of the process weight regulations.  None of the state regulations is more
stringent than the lower NSPS of 0.23 kg/MW-h (0.51 Ib/MW-h), but a number of
state regulations are more restrictive than the NSPS for high-silicon-based
products.  State opacity regulations are fairly consistent.  They generally
limit opacity of exhaust gases from the control device to 20 percent, which is
similar to the NSPS limitation of 15 percent.
     The states have no carbon monoxide regulations specifically applicable to
ferroalloy submerged-arc furnaces.  Some states require a specific percent
control on CO streams from certain refinery or metallurgical processes, such
as a blast furnace or basic oxygen furnace, and others require incineration of
streams from specific processes (usually cupolas, blast furnaces, and catalyst
regenerators).

3.3  EMISSION REGULATIONS FOR FERROALLOY FACILITIES IN OTHER COUNTRIES
     Several countries that have ferroalloy facilities have passed specific
regulations for these operations.   Table 3-4 presents a comparison of some of
these regulations (where emissions can be shown on a common basis) with the
NSPS of the United States.   This comparison shows that the NSPS are the most
restrictive for open furnaces, but are generally less restrictive than regu-
lations of other countries  for closed furnaces.  These low emission limits on
closed furnaces result from the use of a concentration-type standard multi-
plied by a small  exhaust gas volume.
                                     14

-------
                    TABLE 3-1.  STATE AIR POLLUTION CONTROL
               REGULATIONS IN STATES WITH FERROALLOY OPERATIONS
             State
          Alabama
          Iowa
          Kentucky
          New York
          Ohio
          Oklahoma
          Oregon
(continued)
    Particulate0
            \
E = 3.59 P-D^ Px30 tons/h
E = 17.31 P'16 P>30 tons/h
(Class 1 county)

E = 4.10 P'67 p<30 tons/h
E = 55.0 P-]1-40 P>30 tons/h
0.1 gr/dscf may be imposed

E = 3.59 P'52 P<30 tons/h
E = 17.31 P'16 P>30 tpns/h

E = 0.024 P'67 P<100,000 Ib/h
E = 0.030 gr/dscf P>100,000 Ib/h
E = 4.10 P
                                        .67
                                      .11
P<60,000 Ib/h
                             E = 55 P '   -40 P>60,000 Ib/h
E = 4.10 P
                                        .67
                                      .11
P<60,000 Ib/h
                             E = 55. P '"-40 P>60,000 Ib/h
Curve for P<60,000 Ib/h
E = .55 P '71-40 P>60,000 Ib/h
                  Opacity
                    20%
                    40%
                    20%
                    20%
20%
20%
                    20%
                                     15

-------
 TABLE 3-1 (continued)
              State
     Participate3
                                                                Opacity
           Pennsylvania
           South  Carolina
          Tennessee
          Washington
          West Virginia
Emission = 0.76[E]*42

Where

E = 0.3.P  and

P = charging rate in Ib/h


E = 4.10 P'67 P<60,000 Ib/h

E = 55P'1T-40 P>60,000 Ib/h


E = 3.59 P'62 p<60,000 Ib/h

E = 17.31 P'16 >>60,000 Ib/h


RACT

0.1 gr/dscf
Process weight table or at
least 99% eff. for dupli-
cate sources0 smaller
than 125 tons/h
                                    20%
                                                                 20%
                                                                 20%
                                    20%
                               No.  1  Ringelmann
                             or equivalent
                             opacity  and not
                             equal  or exceed
                               .  2  Ringelmann
                             or equivalent for
                               minutes per 60-
                             minute period
,P = charge rate in tons/h in equations; E = allowable emissions in Ib/h.

 Most states allow the stated opacity to be exceeded for 5 or 6 minutes per
 60-minute period.  The excursion is usually limited to 40 percent opacity.
o
 Any combination of two or more individual source operations that have the
 same nomenclature.
                                     16

-------
              TABLE 3-2.   COMPARISON OF ALLOWABLE EMISSIONS FROM.
              FERROALLOY  PLANTS UNDER NSPS AND STATE REGULATIONSC
Product
Si metal
SOlFeSi
65-75% FeSi
CaSi
SiMnZr
H-C FeCr
Chg Cr
FeMn
SiMn
CaC2
FeCrSi
FeMnSi
Silvery iron
NSPS limitation
Ib/MW-h
0.99
0.99
0.99
0.99
0.99
0.51 i
0.51
0.51
0.51
0.51
0.51
0.51
0.51
Ib/h
29.7
29.7
29.7
29.7 .
29.7
15.3
15.3
15.3
15.3
15.3
15.3
15.3
15.3
State regulations, Ib/h
Most stringent*5
15.43
; 19.24
19.48
14.87
19.48
28.65
28.65
30.91
23.81
21.88
18.27 ..
25.65
23.50
Least stringent^
19.81
25.16
25.50
19.05
25.50
38.68
38.68
41 .94
31.68
28.91
23.80
34.33
31.22
 Based on 30-MW furnace.
 Most stringent State allowable:
          E (Ib/h) = 3.59 P'62
                   =17.31 P'16
cLeast stringent State allowable:
          E (Ib/h) = 4.10 P*67
                   = 55.10 P'^-
 where P<30:tons/h
..where P>30 tons/h

 where P<30 tons/h
 where P>30 tons/h
                                     •1.7

-------
          TABLE 3-3.  FACTORS FOR CONVERTING TO PROCESS WEIGHT RATE
                         (based on a 30-MW furnace)
Product
Si metal
50% FeSi
65-75% FeSi
CaSi
SiMnZn
H-C FeCr
Chg Cr
FeMn
SiMn
CaC2
FeCrSi
FeMnSi
Silvery iron
Charge rate,
lb/MW-ha
700
1000
1020
660
1020
1900
1900
2500
1410
1230
920
1590
1380
Ib/h
21,000
30,000
30,600
19,800
30,600
57,000
57,000
75,000
42,300
36,900
27,600
47,700
41,400
tons/h
10.5
15.0
15.3
9.9
15.3
28.5
28.5
37.5
21.15
18.45
13.80
23.85
20.7
Based on information in Reference 2.
                                    18

-------
rO
 CO
 Ltl
 o
 UJ
 co
 CkT
 O
 CO
 O
 CO i—
 CO
 CO
. LU
 u_
 o
 O
 co
 O
 O
 co

 LU
 _J
 CQ

• o
to
JC O
i • 3 S-
O <4-
CO «C

'


JD
0)
-a
CD
2s
. CO


2>fc
E
£
s-

co

„
o to
OJ "O
X) (O
 2-:
."if
E '


O)
Nl
•i—
l/> O)
CL
0) >>
0 +->
(O
E -O
S- E
3 re
u_

o
3
"O
O
S-
Q-
CT>
CO
r—
•— •
CM

CO
^
^— «*
0>
CT»

'^^
^™
•
CT>

CM
CM
LO
^— •
,^.
•
CO
CM
. r—
LO
CM
^^^*

*^
•
|M-B
"

oo"

I*N,»
r~~
	 '
• , 	
•
oo


E
O)
Q.
O

A
*ra»
' ?r^

00






•IT-
CO
00
oo
r—
	 '
CM
•
LO
00
^ 	 ^.
O
r-^.
"

o
-• •
LO
co
CO
o

• — •
oV
•
r^>
CO
oo"
|S^
LO


CM

LD
CM

•J^
• '
CTi
CM

LO
•
CO


e~
O)
O.
o

ft
: 3
s:

0
co
CO
O)
u.

^^
LO
r>
o"

i**^
—
CM
•
CO

^ 	 	
^
co

~~~* .
r*^

f—

VXD
CM

• — *
OJ

p_

•p
^
CM
. **—*

CM
•
f—
'""

oo"

f^.
r_
— ,
, 	
•
00
•o
(D
CO
0

13

. , «v
.3


LO
co



c~
^*
•r—
co
     O)


     03


     O)
     I/)
     


     O)





     LO
                                                  •  en
                                                co  .—
a>
o
O)

I
a>
                                                      O)

                                                      O

                                                      (!)
                                                     J3
                                                      CO
                                                      4-»   ' ?
                                                      C CO
                                                      «O   E
 O
4-
                                                 O

                                                •a
                                                 0)
                                                 to
                                                      S-
                                                      O
                                                          O
                                                          O
                                                          CM
                                                      
-------
                     REFERENCES FOR SECTION 3
U.S. Environmental Protection Agency.  Code of Federal Regulations, Title
40, Chapter 1, Part 60.  Washington, D.C., Office of Federal Register.
May 4, 1976.

Dealy, J.O., and A.M. Killin.  Engineering and Cost Study of the Ferro-
alloy Industry.  U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina.  Publication No. EPA-450/2-74-008.  May 1974.
p. VIII-4.

Denizeau, J., and H.D. Goodfellow.  Environmental Legislation Approaches
and Engineering Design Considerations for Ferroalloy Plants.  In:  Pro-
ceedings of the Fourth International Clean Air Congress (Paper VI-26).
Japan.  May 16-20, 1977.  5 p.
                                 20

-------
                                  SECTION 4
      j   -'•..-.;   V '       '          - "                    .         '    ,
                               INDUSTRY. STATUS

     This section presents information on the current status and changes in
the industry's demographic features since background information was assembled
in 1974 for the new source performance standard.
4.1  PRESENT CAPACITY
     Ferroalloy production in the United States has been declining over the
past 9 years, primarily because imports have increased and demands have re-
mained fairly constant.  Figure 4-1, which illustrates the trends in imports
and domestic production since 1970, shows that imports have more than doubled
in this decade and now account for approximately 45 percent of demand.
                                                                      1,2,3
Domestic consumption is essentially keyed to the production of steel.  Table
4-1 shows that the consumption of ferroalloys per ton of steel has remained
                                          14
relatively constant for the past 10 years. '   This ratio is not expected to
change in the near future.  The 1979 increase in domestic production shown in
Figure 4-1 resulted not from increased steel production, but mainly from
increases in stock levels and unreported consumption.  As discussed later,
this increase is not believed to represent an actual reversal in the overall
downward trend in domestic production.  Exports of ferroalloys have recently
averaged only about 0.043 Tg (47,000 tons/yr) and consequently are not a
significant factor in assessing industry capacity.
     The sale of domestic plants to foreign companies is another .recent
development in the industry.  For example, Airco has sold its plants in
Calvert City, Kentucky; Charleston, South Carolina; and Niagara Falls, New
     5    '         '        '            "
York;  and Union Carbide recently announced the sale of all of its ferroalloy
operations.   Although this trend does not directly influence industry growth
because all the plants involved are still operating, it may tend to stem the
growth in imports and precipitate some plant rehabilitation.  This is highly
              i' '•-'•/•'     .         '      •
           •   '    •    •        •        21      .       '     :   •

-------
     3,ooor
     2,500 *
 in

 o
o
o
O
to
UJ
cc
o
0.
o


-------
CO
•ar-
CL,
so
:=? »-i
oo ^—
•z. o
CD •=>
O O
   O
>- oe.
o o.
eC UJ

O UJ
C£ I—
C£ CO
UJ
U_ U_
   O
O
UJ Z
I— O
C£ 1-4
O I—
Q_ O
CO
1
r— -O
O 0) E
S- CO •»->
^, >
M- °c:
00,
i^i
)
. •*
 4-> (O
CO E-l->
>> O-4J
0 i-
i— O)<4-
•— Q.O
•=c
o
• *l *'
E.
o .- ;- .
•r— ^™N
"£. 00
•Ji-.-O'';-.:
,. ovo
r— S-> 0
« Q-rrr
w^' V™^
•Or—1' '•"-
- ".. "

o .
H3 •
S- •>
. O
4J E ,
CO 3
§CO O
E r™-
•oo
. O)
•O >>J—
•4->,°.
P re
O.



CO
i^
CO
CO
CO

r—
•~

«*•
O
Cvl
CO
CT>
O


00
O



CM
r-
CO
CO
•
CO

r<
r-^
i—

CM
CO
co
o>
o
CM


r~
«*
r-»
•*
CM
in
CM
CM
CM
r>~
en


en
en
co
o
•
0

o
CM
r—

CO
o
LO
en
CO
CO


«*
§
CD
co
co
(•>..
CM
co
r-»
en


r—
O
«*
•
0

r— •
CO
r— •

,r^.
in
^i-
co
CXI
ro


CO

en
CM
in
co
CM
<•
r^
en


o
vo
CO
CD
CO .

«b
«*

CO
co
en
in
o


p^.
r-».
en
0
' CM
CD
en
• r— ••''
in
r^
en


in
LO
co
CO
•
r^.

in
CM

o
CO
CM
CM
CO


CO
en
CO
CM
CM
§
CM
CO
!>-.
en


«*
r>.
co
r-
CO

r~.
CO

CO
in
CM
r»~
co


r-.
o
•*
co
CM
CO
CM
f-.
r^
en


o
r>.
CO
in
CO

i^
Sf
i—

o
r-^
co
•*
«d-

E
CO
O
s-
cu
o.

CD
!-*•

4->
3
O
J3
(O

>>
JD

•o
CO
4-J
re
4->
CO
S-
0)
-o
E
3

CO
-Q

O
•+->

t3
CO
+J
re
£
•r—
4J
CO
 CO
•"- S-

E
E 3
O CO
•i— E
4-> O
Q. 0
E •
3 >,
co n
E
O cn
O E
+-> 4J
E 'S-
0) 0
0 CX.
(U CO
D; s-
QJ
• f >
r1^ Q}
m-^
CO CX
0 E
E o

O)
S- E
CLO
 S-
•i- Q.
<4- X
CO
CO
•r- E
J= O
1— •<-
4->


i— 3
CO CO .
QJ E S-
+J O CO
CO 0 5
0
S f
re •
S-F- 4->
CO d
1- CO CO
O 4-> O
CO S-
*-^ . 4JCO
\ Or^-
-Q +J
•— 4->
«f- 3
CO O O
J3
co 4-> re
CO E
	 0) CO
0 T-
0) S-
s: o E
"->. 0- 0
o> •*-
*i to -M
O
CO co 3
. ib TJ
CO O
r- i— S-
CO Q-
CO O)
r- 4-> >,
re co s-
3 -o
0- >, E
CO O 3
i— O
cn r— M—
r^ to
en co
i — 4-3
1 O O
in s-

cn  OJ
» CO CO CO
^t" E 4^
O >J CO
CO O r—
O CO •—
E co > re
CO OVl- 4->
s- re 4J o
(U s- re 4->
(f_ CD r—
oi =t s- o
o -o
                                                        23

-------
speculative, however.  In the past five years, plants at Brilliant, Ohio;
Sheffield, Alabama; and Houston, Texas, have discontinued production.
     The total estimated capacity of plants in the United States for production
of ferromanganese, ferrochrome, ferrosilicon, and other major ferroalloys is
estimated to be 1.82 to 2.09 Tg (2.0 to 2.3 X 106 tons).  The higher estimate
is based on the furnace megawatt (MW) ratings shown in Table 4-27"9 and the
energy requirement for the products (given in Reference 10).  An annual
operating time of 7884 h/yr was used, which corresponds to 90 percent utiliza-
tion.  Capacity for a given furnace was computed as follows:
             - (MH rating of furnace) X (7884 h/yr)
                    MW-h/ton of product
The resultant estimate is considered to be a maximum value because a 90
percent utilization rate probably could not be sustained on all  furnaces over
the long term.  Capacity also varies with the desired product and its power
requirement.
     Another capacity estimate of 1.82 Tg (2 X 106 tons) is given in Reference
11, and an indirect capacity estimate in Reference 12 states that the industry
was running "flat out" in 1979, a year in which 1.63 Tg (1.79 x  106 tons) was
produced.  The exact industry capacity is difficult to determine because many
furnaces operate only part of the time, others are in cold standby condition,
and still others have been dismantled.   Based on the cited references,  how-
ever, it appears that the industry could not satisfy a total demand of  about
2.3 Tg (2.5 x 10  tons) without importing approximately 0.2 to 0.3 Tg (200,000
to 300,000 tons) per year.
     The capacity for specific ferroalloys is difficult to estimate because
most furnaces can produce several  different products, depending  on market
conditions.  Table 4-3 shows the product mix in 1979.  During that year 25 to
55 percent ferrosilicon accounted for 30.7 percent of production, ferroman-
ganese accounted for 17.4 percent, and chrome-based alloys accounted for 16.0
percent.  Import penetration is highest in chrome-based alloys and manganese
alloys.   It is likely that silicon products will  continue to comprise  the
major share of domestic production.
     The estimated calcium carbide production capacity of U.S. plants is 0.34
Tg (0.37 x 10  tons).   This estimate was calculated in the same  manner  as
                                     24

-------
OO
l-cn
Q CJ
>-< <:
ca 2:
o; Q:
se
<_>. i
__! O
eC LU
O O
   Q;
a LU
2: s:
- co
C3
-J O
O H-
OC o
CU LU
<\J
UJ


CO
4J
I
a.
a-
«
-S
C
O


'• I

cu
N
CO

' u»
Q> S-
U 0)
lO O
II

1


t/l

u
I


o
•o
C
o

*•£
u
o


rO

O.







1

2
A.



•o
* £j»
cu 01 fc-
£* oS
•j— >r- -Q
*- S.
•r- W
t- -C

"fs-?OJ
Ll_ f S


S



'—


C
01
n,
0






CM
C_J
O



w

?



3
O






^

O Q
C

.si
t_ «
•«- o




S- 1.
£ £

T '£

<•*


in i*^
CO



r~ r-


C 1C
a> cu
a. ex
o o



CO
ai

CM
O . tt^
cj in
^

M fO


o »
&. ra
ai x


3 i






I

5

i
2
rO
5



1


•£

(O
U.
in
CM

•a
c
IO
0
CM

CM


1
°


CO
«
in

CO



p

t/i




S
3 •:
•i— O
S- C

5
>i

°.i
C)


C -C
f"" t. >
o ••-
•*



i-
 t.
m -r-j O
Lt_ LiJ V»
in tn
CO r-^
CO
to to
r— CO
C C
"S


a. ex
0 0
CO

u_
o
in


1 • t-
£
O C
OJ
» (—
•o

U- , O
Cf» •«->
C W>

C o
a. o
CO 3

-o

03
O O O>
0 C • C
'.K 5i

cxm S - _
t- 1 -r- ' O £
O .r- 1_ (J 3
<_>(/» as £=
•M O) 3 E O
o s-o o u
, J3 (0^ 1- .E
U 0







































>

Q
- „

£•
o

CT
C


V
u
in
|j
**f
o
j_


O .
V)

O

to
1




01
4_
CJ
s_.

LJ_


CM


J3



1
O



CO

u.
o
in



en
0)

°
cu

T3

£




.
O

=*
C

£
i
ra

C
10
*/>
cu

4— c/>
0 0
•r- ro
i- c
XI t-
(O 3

CM CO
-C
CJ


O)
a.
O


CO
CO

La-
ir)
f-»


m
2

I
0,
4-1

C



0

C7»
c •
•— >
CU -r-
E O
CO
C
r- O
a> u
:z in

                                      25

-------
•a
 cu
CM
                     S    S
                                                                                 trt      ••—
                                                                         O (U
                                                                         13     H-
                                                                 r— CM   ,—
                                           s.
g.
                                                                                <§•
                         S t—
                                                                                •r-»—    er
                                                         26

-------
-a
 
c



3
o*
OJ
'o
i_
4->
. . c
CJ











I


QJ
N
to


t/1

Q.
ff






Ul
4-J
.U
3,
•o
O
i-
• a.

' s-












§
5
IS
|

+J
C
RJ
ex













•*-»
O
3
T3
O.
•S_
. . CL,










£
S
£
•if
S-
J3
(T3
U_







*O •
C '
«J

m *s- -cr
r- r- U
«*>S

ex,- 3-





CO CM r—











• I 2.1
o oo








f
c
^
V)

•> *(/>


0) OJ OJ









£>
„
j?
CJ

4->
S_
(U
>
10
cj












o
c



>»
o



£

s_
.O
-
2:

to
fO
Lt.

«o
'1
IS




























t_
(U
•M
C

11
JD
«J
u_









QJ
J=.
O










c
c
QJ
h-

O
o
2
.^
U
s





a> o;
i/l O
 u> i— -a
.COR) S- «J C 13
.--=-*= jgfc * -S
3 CX (U J3 3 S~ I- 3
.ICO 3 QQ -  3 t, 3 !_
tn»— S-r— eri- XiS. .0 OJ a)
 (J (O CJ) *O C t-
OO*»-_J<4- >(/> LutO LL.O 03
T3
O • R3
LO VI
i/> U)
13 CO
C O «3
^ '5 £
* i — (O 3
in  OJ
no a. cj
* .G .C 4-> O - t.
CM • T3 (J CJ O *O
CM CTSflfOCO -T- |
fl a» a> i t-- ^ "o
" IT) r-~ QJ OJ
o in CM oo *o .r— - — u_ ,^»
*" '"^ I_ fe. .
•i
~O ' W
cu
wi O
•* CMf-r— CMOt. 'ZJ C

O .  i —
 w> H vi C
fQ -r- QJ •(— QJ -r- QJ 4_) (J »— -i— -r- -r—
CJ (/lU-tOLUl/lU- WQJOJi —
•»- f— ^^ 3: >>
i — QJ (/) OJ E r—
•r- t. S-
4J c o o  *
 O (U  irfa Qj3 o
* , O QJ i- 4- * t—
ftJ • * • -CE C, aJrtJr-JCCOJO
.a +-» m * o>oj-^.ogr-T-o
ra OJ t— >>T? O» t) -»-> M-
+J -r- 4-> O -r- 3 S- QJ r— M- C
J= S- t- r— S- E ' O QJOJfO
'to flj O ' — 'jQ 3S.CQJ
 t-
c , ta • — • •*-> o- L«J
•-- 3 fO X)
V) E " QJ fO
vi i- in o 3
O) • . >
- O 4J r— U (U ••- iO
4- , fO i— - — - *O 1 4-»
C QJ C C C "O
a. s_ E . - 1 4J ' O
^3 CQJO-C(J-r->,
I- QJ i_ S_ Qj QJ 4-> i —
fO - Q-OflCLr-CJ+J
CJ . O 1 O QJ 3 C

d • O3QJOQJOCJ
0 SOQ-SCJ-O)
••- , h-u_oH-oa.a=
c
:^. «3J3O'OQ)I*-O1
                                       27

-------
 described earlier for ferroalloys.
 0.22 Tg (0.2
 consumption.
0.22 Tg (0.24 x 106 tons) per year.2'3
Production has been relatively constant at
    Production rates were based on lime
                 TABLE 4-3.  FERROALLOY PRODUCT MIX IN 1979a
                Alloy
           High-carbon FeCr
           Low-carbon FeCr
           FeCrSi
           Other Cr
           FeMn
           SiMn
           Electrolytic Mn
           FeSi 25-55%
           FeSi 56-96%
           Si
           Other silicon
           Other5
                                             Percent'of total
                                                  11.6
                                                   1.8
                                                 ,  1.4
                                                   1.2
                                                  17.4
                                                   9.0
                                                   1.5
                                                  30.7
                                                   9.6
                                                   8.0
                                                   5.3
                                                   2.5
            Product mix derived  from data  given  in  Reference  1,  production
            data  for FeAl,  FeMo,  and "other"  alloys from Reference  13.
           Include FeAl,  FeMo,  FeCb,  FeW,  FeV,  FeB.

4.2   PLANT LOCATION
      Historically,  the ferroalloy  industry has  been  located  in  areas of rela-
tively  low electrical  power cost and  near  steel manufacturing facilities.
Table 4-2,  presented earlier, lists the major submerged-arc  ferroalloy and
calcium carbide  facilities in operation early in i960.  The  table  gives com-
pany  name;  plant location; major products; the  number,  type, and approximate
size  of the furnaces;  and the air  pollution  control  equipment applied.  This
listing shows a  total  capacity of  1606 MW  at 31 locations in 11 states.  Of
this  total  capacity, 85 percent  is supplied  by  open  type furnaces, 12 percent
by semisealed furnaces, and 3 percent by closed furnaces.  In addition to
these submerged-arc furnaces, two. open-arc furnaces  operated by the Foote
Mineral Company  in Cambridge, Ohio, produce  specialty ferroalloys, and five
open-arc furnaces operated by various other companies smelt ore/lime mixtures.
Table 4-4 summarizes, by state,  the ferroalloy production capacity of the
                                      28

-------
               TABLE 4-4.  ELECTRIC SUBMERGED-ARC FURNACE CAPACITY
             FOR FERROALLOY AND CALCIUM CARBIDE  PRODUCTION, BY  STATE
State
Alabama
Iowa
Kentucky
New York
Ohio
Oklahoma
Oregon
South Carolina
Tennessee
Washington
West Virginia
Total
Plants
6
2
3
1
6
1
4
1
3
2
2
31
Capacity, MW
189
50a
182b
45
554C
15
57
80
123
54
257
1606
Furnaces
Ferroalloy
10
2
6
2
32
0
5
2
12
6
12
,89
rior.:rr*i.!benQ if
Carbide
0
1
2
0
2
1
1
0
0
0
0
snwe^-'-
• > • nor.iDubo'iu y/ioimui, euujv ivri , •
                    ••.
15 MW for  calcium carbide  (CaC2)

85 MW for  CaC2.                                            afi3

A portion  of catDacity at the  Ashtabola can also be .used to make
                                                «,-.-,-.''. ;i . -4-T-t, ; — »-. ^ f-,{ :^ s- -,-




                                               l^yn^rr..  v? C^*V ; i' ^'%r *"•':
                                       wp.f'-'bfi&"
                                        29

-------
 electric submerged-arc furnace  processes.   Ohio,  in  EPA  Region V, accounts for
 almost one-third of the production  capacity, with 34 furnaces at 6 plants.
 West Virginia,  Alabama, and  Kentucky are also major  production areas.
      Two other  processes are also used  to  produce ferroalloys and related
 pure metals:  the metalothermic or  exothermic process and the electrolytic
 process.  These methods are  used to make specialty alloys or metals in rela-
 tively small  quantities.  Table 4-5 lists  eight plants using the metalothermic
 process and four using the electrolytic process.   Blast  furnaces are not
 currently used  in the  production of ferroalloys.

 4.3   INDUSTRY GROWTH
      Discussions with  various industry  representatives and observers revealed
 no plans  for  new capacity or major  expansions in  the ferroalloy industry.14"18
      Increased  imports have  caused  a  continual decline in the domestic produc-
 tion rate, which has resulted in plant  closures.   According to 1979 production
 figures,  the  ferroalloy industry was  operating at  80 to  90 percent of capac-
 ity,  based on the capacity estimates  cited earlier.  The production of calcium
 carbide has ranged from 0.20 to  0.23  Tg (0.225 to  0.256  x 106 tons) in recent
       23
years,  »  which  corresponds  to  a capacity utilization of 60 to 70 percent.
 Some production  growth is indicated for this product, which is used in the
 steel,  foundry,  and  metalworking industry.  For example, Airco has announced a
 rehabilitation program for a calcium  carbide furnace in  its Calvert City,
 Kentucky, facility.19
      The  growth  of ferroalloy demand  is dependent  on the following factors:
          Raw steel  production
          Ferrous  foundry production
          Alloy  and  stainless steel  production
          Level  of imports
          Level  of exports
      For prediction  of the demand for ferroalloys  in the period 1980-1985, two
scenarios are examined:
     Scenario A:  A  high-growth  scenario,  possible but not likely
     Scenario B:  A more likely  growth scenario
     In Scenario A, all factors reflect a  level  that would normally lead  to
high growth, e.g., large steel growth and  low imports.   In Scenario B,  the

                                      30

-------
          TABLE 4-5.  FERROALLOY PLANTS USING THE METALOTHERMIC
                        AND ELECTROLYTIC PROCESS7,9
                                   Location
                         Products
Metalothermic Process
  Amax Inc.
    Climax Molybdenum.Co. Div.
  Cabot Corp.
    Penn Rare Metals Div.
  Duval Div,., Pennzoit Corp.
  Engelhard Minerals and
    Chemicals Corp.
  Metal!burg, Inc.
  Molycorp, Inc.
  Reading Alloys, Inc.
  Teledyne, Inc.

Electrolytic Process
  Foote Mineral
  Kerr McGee
  Sedema S.A., Chemetals, Corp.
  Union Carbide
Langeloth, Pa.

Revere, Pa.
Sahuarita, Ariz.

Strasburg, Va.
Newfield, N.J.
Washington, Pa.
Robesonia, Pa.
Albany, Ore.

New Johnsonvilie, Tenn.
Hannibal, Mo.
Kingwood, W. Va.
Marietta, Oh.
FeMo

FeCb
FeMo

FeV
Cr, FeCb,
FeTi, FeV
FeB, FeMo, FeW
FeCb, FeV
FeCb
Mn
Mn
FeMn
Mn, Cr
                                       31

-------
        TABLE 4-6.  TWO FERROALLOY GROWTH SCENARIOS FOR 1980-1985
Factor
 Scenario A
 Scenario B
Raw steel growth

Alloy and stainless steel growth

Ferrous foundry products growth

Exports + unaccounted for
  consumption

Import levels in 19851
Ferroalloy usage rate"3
 2.0%/yeara

 4.0%/yearc

 5.0%/yeare

 359,000 Mg/yrg
(386,000 tons/yr)

 Use 1979 level
 of 1.16 Tg
 (1,280,000 tons)
 17.5 kg/Mg
 (35 Ib/ton)
 0.9%/yearD

 3.0%/yeard

 4.0%/yearf

 239,000 Mg/yrh
(263,000 tons/yr)

 50% of total
 domestic demand
 (1.23 Tg or 1,350,000
 tons)

 16 kg/Mg
 (32 Ib/ton)
aSee Reference 21.
 See Reference 22.
°Value from Reference 23+1 percent.
 See Reference 23.
6Value from Reference 24 + 1 percent.
fSee Reference 24.
^Derived from data in Reference 1, highest value in last 5 years.
 Derived from data in Reference 1, average for 5 years (1975-1979).
11mport levels are assumed values.

^Usage expressed as kg ferroalloy per Mg of steel plus ferrous foundry pro-
 duction (see Table 4-1).
                                     32

-------
factors reflect the growth level  that appears to be more likely.   Because only
a high demand calling for new source construction (or major modifications) is
of interest, a lowest possible growth scenario is not considered.   Table 4-6
presents the values of the factors for each scenario.
     The most difficult factor to assess is the trend in imports.   There is no
substantial reason to expect imports to decrease.  Two factors may result in a
slowing of the growth of imports:  (1) Japan has cut back on ferroalloy
                              20
production to conserve energy,   and (2) worldwide steel demand is expected to
     21 22
grow.  '    Scenario A is based on the very optimistic assumptions that the
total quantity of imports remains constant through 1985 and that exports plus
unaccounted for consumption remain at the high level of 1979.   Scenario B is
based on the assumption that imports remain at a constant percentage of
domestic demand, which is somewhat optimistic.  Growth in the  markets using
ferroalloys is moderate and not a major factor.
     Based on the data in Table 4-6, the domestic demand for ferroalloy prod-
ucts in 1985 is expected to range from 2.0 Tg (2.25 x 106 tons) in Scenario A
to 1.45 Tg (1.61 x 10  tons) in Scenario B.  Table 4-7 shows the composition
of these estimates.
                TABLE 4-7.  ESTIMATES OF FERROALLOY DEMAND IN 1985
                                Tg/yr (106 tons/yr)
Component
Demand based on steel and
growth and usage rate
Exports plus unaccounted
Less imports
Total estimated domestic

ferrous foundry
for consumption3

demand (rounded)
Scenario
A
2.85 (3.14)
0.35 (0.39)
-1.16 (1.28)
2.0 (2.25)
Scenario
B
2.45 (2.70)
0.24 (0.26)
-1.23 (1.35)
1.46 (1.61)
aAverage exports for the period 1970-1977 were 0.043 Tg (0.047 x 10  tons/yr).

     Figure 4-2 shows this range of domestic demand (converted to a production
basis using 90 percent product yield) relative to the capacity estimates
derived previously.  Although the figure indicates that some expansion is
                                      33

-------
 2.6
2.4
2.2
2.0
o.


5   1.8


^
1.6
1.4
1.2
                   ACTUAL PRODUCTION
            1965
                          1970
                                        AREA OF POSSIBLE,


                                         BUT UNLIKELY, 	


                                     NEED FOR NEW CAPACITY
                                                AVERAGE PRODUCTION


                                                  LAST 5 YEARS
                                                                     2,500
                                                                     2,000
                                                                    1,500
1975

  YEARS
                                            1980
                                                           1985
                                                                             o
                                                                             •—I
                                                                             fr—

                                                                             o
                                                                                I
                                                                                o;
                                                                                LU
  Figure  4-2.   Summary of ferroalloy  industry capacity  and

                         demand estimates.
                                 34

-------
conceivable (a most optimistic view),, the weight of data'1 cited in earlier
references and discussion supports the conclusion that no expansion or major
modifications are expected in the industry through 1985.  In fact, the pessi-
mistic view would be to assume a continuation of present trends and the
resulting continued contraction of the industry.  The lack of chromium and
manganese ores in this country will add to the decline in production of these
metals.  Silicon production growth will be more favorable due to abundant
ore resources and stable electrical power rates.  The replacement of some
existing capacity by new furnaces is possible, but no such plans by the pro-
ducers are known to exist at this time.14"17
                                     35

-------
                          REFERENCES FOR SECTION 4
1.   The Ferroalloy Association.  Statistical Yearbook.   Washington, D.C.,
     1979.

2.   Department of the Interior.  Minerals Yearbook, Vol. I, Metals, Minerals,
     and Fuels.  Washington, D.C.  1970 to 1977.

3.   Department of the Interior.  Mineral Industry Surveys:   Lime.  Washing-
     ton, D.C.  1978 to 1979.

4.   American Iron and Steel Institute.  Annual Statistical  Report.   Washing-
     ton, D.C., 1979.

5.   Ferroalloys, Supplement to American Metal Market.  Fairchild Publica-
     tions, New York, May 23, 1979.  p. 17A.

6.   Chemical and Engineering News, June 9, 1980.  pp. 6-7.

7.   Pietrucha, W.E., and R.L. Deily.  Steel Industry In Brief:  Databook,
     U.S.A.  Institute for Iron and Steel Studies.  Greenbrook, New Jersey.
     1979-1980.  pp. 39-41.

8.   Letter and attachments from G. Watson, Ferroalloy Association,  to R.
     Gerstle, PEDCo Environmental, Inc., May 19, 1980.

9.   Draft preprint from F.J. Schottman, U.S. Department of the Interior, to
     PEDCo Environmental, Inc., April 7, 1980.

10.  Dealy, J. 0., and A. M. Killin.  Engineering and Cost Study of the Ferro-
     alloy Industry.  U.S. Environmental Protection Agency, Research Triangle
     Park, North Carolina.  EPA-450/2-74-008, May 1974.   p.  VI-15.

11.  Heine, H. J.  Using Ferroalloys Effectively Part I.  Foundry Management
     and Technology, February 1980.  p. 28.

12.  Reference 5, p. 16A.

13.  U.S. Department of Interior.  Preprint for Minerals Yearbook.  Ferro-
     alloys.  Washington, D.C., 1977.  pp. 2-3.

14.  Trip Report by PEDCo Environmental, Inc., to Union Carbide Plant, Alloy,
     West Virginia.  June 5, 1980.
                                     36

-------
15.  Trip Report by PEDCo Environmental, Inc., to Union Carbide, Marietta,
     Ohio.  June 5, 1980.

16.  Trip Report by PEDCo Environmental, Inc., to Foote Mineral Company
     Plant, Keokuk, Iowa.  May 16, 1980.

17.  Telcon.  R. W. Gerstle, PEDCo Environmental, Inc., and 6. Watson, Ferro-
     alloy Association, Washington, D.C., April 3, 1980.  Industry Growth.

18.  Telcon.  W. F. Kemner, PEDCo Environmental, Inc., and B. Reddy, Charles
     River Associates, Boston, June 9, 1980.  Industry Growth Rates.

19.  33 Metal Producing.  18:16.   February 1980.

20.  Reference 5, pp.  10A-11A.

21.  U.S. Department of Commerce, Office of Industrial Economics.  Industrial
     Economics Review, Vol. I.  Washington, D.C., May 1979.  p. 11.

22.  American Iron and Steel Institute.  Steel at the Crossroads, The American
     Steel Industry in the 1980's.  January 1980.  pp. 28-29.

23.  American Metal Market, 87_(200):4  October 12, 1979.

24.  Foundry Management and Technology.  Foundryme'n See 4.4$ Growth in 198;0.
     January 1980.  pp. 28-34.
                                     37

-------
                                  SECTION 5
                                TEST RESULTS

     Within the past few years several emission tests have been made on ferro-
alloy furnaces to determine their compliance with state regulations.  Because
such compliance is frequently based on visible emissions, manual tests Often
are not conducted.  This lack of manual test data is due partly to the diffi-
culty of testing open-type fabric filter systems and partly to the explosion
danger inherent in closed and semisealed furnaces.  No tests have been made to
determine compliance with NSPS because no plants are subject to these standards,

5.1  PARTICULATE EMISSION TEST DATA            .
     During this study, available emission test data were obtained from a
number of state pollution control agencies.     These data, summarized in Table
5-1, represent information on emissions from control systems.  No data were
available on fugitive emissions, but the results of one test on a tapping hood
system were obtained.  The emission test data show that these systems could
meet the NSPS, usually by a comfortable margin.  Furnace emissions ranged from
0.073 to 0.20.kg/MW-h (0.16 to 0.44 Ib/MW-h).  No single control system pro-
vided significantly better control than any other; however, the data are
insufficient for any firm conclusions to be drawn.  It should also be noted
that most of these data represent compliance tests on new or modified emission
control systems, which means the systems were operating under essentially ideal
conditions.
     The EPA Environmental Assessment Studies also provide recent particulate
                   78
emission test data*-     Although the test methods used in these studies are not
compliance test techniques, the results are useful as indications of the actual
emission rate.  Table 5-2 shows the types of furnaces studied and their emis-
sion controls; except on Furnaces B-l and D-l, measurements represent
                                      38

-------
g
o
CO
LU
I—

UJ
O
o
o
in

UJ
_i
OQ

UJ












J=:
1 ,

^T"
*^^
XI
^


IC*^
1
"^!
y
^^
.o>


.c:
*^^
X)

*^"*


-Si'
en




M- t—
o o E
S- CU
CU +J -P
Q. C to
>> 0 >>
f— 0 l/>

(O
en
£=

^" >4-J
(O
S-
o
"7
•o
o
• s-
Q.

t(— CU
0 0
to
cu c
Q- S-
>» =>
1— M-



r*—





in
co
•
o





to
r—
- * -
o



CM
CM
•
in
. ^" *

r~.
- *
CM





j^
u_
u_





in
r—




CM
0
to
0




C
cu
ex
0



CM .





vo

•
CO
*^^



co

0

o



CM

•
CM



> •
f—^






U_
u_





in
1— t




CM
o
(0

o

cu
1 r—
•r- tO
E CU
cu to
^



CO




o
1 >

CO
co
•
o

o


in
l«—
•
o

o
-l_>

f"*^
•
«^_>
o

sT

•
CM


•r-
i-
1 *
c
cu
CM



in
I—


>,
cu
>
r-^-
•I—
co


-o
O)
to
o
r—
CJ)








o
^^^ 4-> ^~^
« • 01
in

o
in r—
• in *
CO '• 
• •
co m



10 r*«
• •
CO CM





u_ u_
U- U-





o in
CM CM




•r-
•i- CO
to cu
U-




c:
cu
EX
o



in




o
1 ^ ^-««
T*^*
in CM
0 •
• o
o
o


CM CM
CM i—
o •
• o
o

o
.f-3 **— >
^~
m •
• o^
^

o
, 	
1^^ *
* ^J"
o



cu
Q,.
co
UJ

co
• •
r—
co
Ntvri*

o


i.
CJ
cu
u_




c:
cu
ex
CD












10


1

•'
^-^. '
ID
co
•
o




to
^~
•
o



*~-^
^j*
' •
CO



,_
«
vo





Q.
CO
UJ

9-^
CO
. •
f-x^
co
v«x

o


s_
(_2



•
CO
•
^J-

,
to
s_
cu

^
•1—
o



f^
CM




c
s:
•r—
co



. 1
"i
cu
co
































S-
cu
i Q
J^)
s-
o
to
s-
0















"O
0)
r~~
cu
 CO














en
c
•r- -a
CX 0
CX 0
to -c
1— .
TD
 rO
 O
                                                                                                  T3

                                                                                                   (O
 O
 3
-a
 o

 a.
                                                                                                   CJ)
                                                                                                  -a

                                                                                                   o
                                                                                                   o
                                                                                                   o
                                                                                                   cu
                                                                                                   i-
                                                                                                   n)
      S-
      cu

4J   r-
 3   •!-
 CX  M—

•i-    O



I   1
                                                                                                            T3
                                                                                                             CU
                                                                                                            •o
                                                                                                            -a
                                                                                                             to

                                                                                                             to
                                                                                                             ro
                                                                                                             I
                                                                                                             s-
                                                                                                             cu
           o
          •r-
           S_


          I

          r—
           (O

           o
          T3
          •a
           to
           to

          T3
           C
           (O
           cu
           o
           (O
           E:
           c.
           cu
           Q.
           O
                                                                                                            TD
                                                                                                             CU
                                                                                                            •!->

                                                                                                             CU


                                                                                                             O
                                                                                                             o
                to
                cu
                -p

                en
                s
                3
                XJ
                                                                                                   O
                                                                                                        II
           cu
           o
     LL.   CU
     U-  C£
    XI   O
                                                                                                                   

 a.

 o
                     to
                     o
                     s-

                     o
                     cu

                     cu

                     n
                4->  Cu
                O  CO
                                                                                                                      CU
                                                          39

-------
                  TABLE 5-2.   TYPES OF FURNACES TESTED AND
                           THEIR EMISSION CONTROLS
Furnace
A-l
A-2
B.-1
B-2
C-lc
C-2C
D-1&2
Type
Semisealed9
Open
Open
Semisealed
Semisealed
Semisealed
Sealed
Emission
control
Scrubber — High-energy,
AP = 13.8 kPa (55.6 .in, H20)
Scrubber — medium-energy,
AP = 11.6 kPa;(46.5;..in. H20)
Fabric filters
Scrubber--hi gh-energy ,
AP = 20.1 kPa (81 in. H20)
Scrubber — low-energy,
disintegrator- type
Scrubber — 1 ow-energy ,
disintegrator- type
Scrubber — high-energy,
AP = 22.4 kPa (90 in. H20)
L/6, liters/m3
(gal /1 000 ft3)
6.5 (49)
0.61 (4.6)
-
4.0 (30)
8.0 (59.6)
8.5 (63.9)
2.0 to 3,2
(15 to 24)
 Semisealed furnaces vary in the degree of undercover combustion.  Combustion
 was essentially complete in Furance A-l during the tests, but it was
 substantially less than complete in the other semisealed furnaces tested.
 Designed for high energy, but operating at medium energy during test.
cFurnaces are now shut down.
                                     -.40

-------
controlled emissions.  Tests on Furnace D were made under two different opera-
ting conditions (D-l and D-2) and the two sets of data on this furnace cannot
be compared with each other.  Table 5-3  summarizes the particulate test data
obtained in these studies.  In all cases where a scrubber was used, tests were
made at the outlet (except on Furnace D-l), and the inlet data were then
calculated from the particulate removed by scrubbers.  In all cases, tests were
made prior to any flares.
     The particulate emissions data in Table 5-3 show that most furnaces would
comply with the NSPS (even though they were not subject to these standards).
Furnace A-2, an open furnace producing FeMn, exceeded the NSPS limit of 0.23
kg/MW-h (0,51 Ib/MW-h) by 50 percent.  Furnace C-2 exceeded the NSPS limitation
of 0.45 kg/MW-h (0.99 Ib/MW-h) for ferrosilicon production by about 70 percent.
This was a semisealed furnace equipped with a low-pressure-drop, disintegrator-
type scrubber.  The Tow emission rate of 0.016 kg/MW-h (0.035 Ib/MW-h) measured
on Furnace D-2 was obtained after a venturi scrubber with a pressure drop bf
22.4 kPa (90 in. of water).  No fabric filter outlet data were obtained during
these assessment studies, probably because it is difficult to sample open-type
fabric filters.

5.2  VISIBLE EMISSIONS
     Visible emission readings are made by enforcement personnel from state and
EPA Regional Offices.  These readings, which are made according to EPA Method
9, are made on emissions from control equipment vents and roof openings to
determine compliance with state regulations.  Apparently no visible emission
data have been taken directly at the tapping station.
     Table 5-4 summarizes visible emission readings taken at a plant whose open
furnaces are equipped with fabric filter control systems.9  This is an older
plant and is not subject to NSPS.  During normal operations opacity of visible
emissions from the fabric filter vent system was consistently under 10 percent
(well under the NSPS limit of 15 percent); however, when the system operated
with a broken bag, opacity readings averaged 22.5 percent over the maximum 6-
minute period.  Tapping operations on the two 40-MW furnaces caused opacity of
visible emissions at the roof monitor to be much higher.   "Normal" taps pro-
duced opacities averaging approximately 30 percent over the worst 6-minute
                                      41

-------
00
  CO
  UJ
  co
  UJ

  co
  CO
  UJ
  co
  CO
  «=c
  UJ

  £
  UJ
  «=c

  o
  CO
  CO
  CJ
   o:
   •a:
   o_
  co
    i
  LO
   CQ
   <£.
                          •*-> C E
                          u*>i- tt-
                          ^E o
                            (O £-


                            II
                            O d.
                                                       in     .—     .—
                                  i—     r-.
                                                              S      F:
                     CO     t—
                            CO     —
                                   i—     CM
                                  f—     O
                                                                      r—      CD
                                         r—      cn
              i—     en
 .r—,  in cn
                                            r—     o
                                            I	     CNJ
•—     in
                            f—     f—     •—     CVJ
                                                                      .<«      rr      r:
                                                       42

-------
TABLE 5-4.  VISIBLE EMISSION READINGS'
Emission point
Fabric filter
serving two
open furnaces


Roof monitor
over two 40-MW
open furnaces



Roof monitor
over one open
furnace .




Operation
Normal
Normal and
tapping
Normal
Broken bag
Normal
Normal
Casting
Tapping
Tapping
Tapping
Normal
Normal
tapping
Tapping
Ladle mixing
Ladle mixing
Ladle mixing
Average opacity readings, %
Highest
6-minute reading
2.7
0.4

5.6
22.5
19.2
8.8
17.3
29.8
31.7
18.1
0
1.5
0
0
47.3
40.4
36.7
Next highest
6-minute reading
1.5
0.2

5.0
8.3
le.o :
8.3
12.5
27.1
12.1
5.2
0
0
0
0
29.6
22.7
34.6
                 43

-------
period.  Tapping periods for these furnaces lasted about 36 minutes.  The NSPS
limits the occurrence of visible emissions at the tapping station to no more
than 14.4 minutes (40% x 36 min.).  The roof monitor over a single smaller open
furnace showed no visible emissions during normal operation and two tapping
cycles.  Uncontrolled hot metal mixing performed in open ladles in this build-
ing caused visible emissions averaging about 40 percent at the roof monitor.
Fabric filter systems used on crushers and a screening operation produced no
visible emissions at the plant.
     Information received from air pollution control offices in Ohio (the State
with the most ferroalloy plants) showed that most plants complied with a 20
percent opacity regulation; only two furnaces had violations during tapping.
Data on another Ohio plant showed zero opacity except during tapping, when roof
monitor emissions exceeded 20 percent opacity throughout almost all of the
tapping period.    Maximum readings of 100 percent opacity occurred several
times during some tapping periods at this plant.
     Region III of the EPA issued a notice of violation to a plant in West
Virginia for opacity violations of the State regulation during tapping and
                   12
casting operations.

5.3  EMISSIONS OF ORGANIC MATTER
     Organic compounds have been identified in both the furnace emissions and
scrubber water discharges from submerged-arc ferroalloy furnace systems.
Organic emissions data were also obtained in the two recent EPA environmental
                   7 8
assessment studies,  '  but quantitative data on specific compounds are ex-
tremely limited.   In addition, all available organic concentration data repre-
sent measurements taken prior to the flare on the vent to the atmosphere.
5.3.1  Gaseous Organic Emissions
     Table 5-5 summarizes the organic emission data obtained in the two EPA-
                  7  8
sponsored studies."0  These data indicate that the furnaces emitted a wide
range of organics,  i.e., before any control  system.   Concentrations of organics
before any control  are much higher from the semisealed and closed furnaces than
the open type because less dilution occurs and no oxygen is available for
                                      44

-------
CO
CO
in

m
"o °
i. «

<-» tl
of-
O (U






-g
1 ^
C
O
U
s-

CQiEE
fez
F^.
Is
0»3S
' C 2!
•r—
1 \ _
«•
1 CO S—
S- 0)

CM
*
O
oo
in
O
• •
O




^t"
CO
CM

CTl
00
•
• f*">*




o
10
r—


cn
CM
00
in


o
CO
«*



oo
cu
u_

^^
o
in
•a
cu
"co
0)
to
•rr
CO



CM
1
CQ


f>».
|>^




O
«C}"
•

oo
o
CM


un
m
f— •



•r-
OO
CU
u_

^^
in
l^-
•a
cu
cu
in
•r—
0)
CO



'^_
• 1
0


cn
fX



cn
•53-
1"*^,
•
o
CM
CM

•
0




UD
cn
1

in
10
• •
oo



CM
cn
in
o


CO
in
cn



CO
10
r—



•I—
CO
cu
U-

fe^
o
in
13
cu
cb
o
to
•r—
CO



CM
1
0


1






I



1







1



^o
vo
*
o




CD
OO
0


CM
O
CM
r—


oo
rx
r—




c
ST*
CU
u_




-a
cu
in
o
CJ



,_
i
o


i




i*^»
^^
o
•
o
CM
r—
O
'•
o




o
in



i













i



in
CM
CM




c:
s:
•r—
CO




-a
CU
to
O



CM
1
O
                                                                                   s-
                                                                                   3


                                                                                   O
                                                                              OO

                                                                              •o
                                                                              ' C
                                                                               '
                                                                               to   CO
                                                                               C/)  I—

                                                                              
              CU


              CO
              CO


              O
S-
O

Z.

O-


T3


CO


CU
O
                                                                                           CU
                                                                                           •a
                                                                                            o
 o
 o

 cu

 CB

 3
 O


 t
 CO
 a.

 co

 s-
 01
               CO


               E
               ret
                                                                                    O)
                                                                                       CO
                                                                                    •


                                                                                    O
                                                                                           •a
                                                                                            c
                                                                                            CO
                                                                                             1
                                                                                            CQ
               CU
               in
               o

               •P
                                                                                            a.
                                                                                            cu
                                                                                            o
                                                                                            X.
                                                                                            O)
 M-   
                                                                                             cu  to
                                                                                                    to
                                                                                                    a
                                                                                                    CO
                                                                              CO
                                                                                           O
                                                          45

-------
 combustion at the top of the furnace.   The  average  concentration was  31.65
      o                                     «'
 mg/Nm  (0.013 gr/scf) from the  two  open furnaces, compared with 2518  mg/Nm3
 (1.1  gr/scf)  from the three semisealed  and  one  closed  furnace  (not  including
 A-l).   The lower organic concentration  in Furnace A-l  is  due to combustion that
 reportedly occurred  even though it  was  a semisealed furnace.
      The  range of emissions on  a weight basis is 0.063 to 1.6  kg/MW-h (0.14 to
 3.5 Ib/MW-h).   Before the particulate control system,  the quantity  of organic
 emissions  was  relatively small  compared with the particulate emissions  (mate-
 rial  collected in  the probe,  cyclones,  and  on the filter, as shown  in  Table
 5-3).   On  a basis  of kg/MW-h, organic emission  rates were 1.5  to 17.1  percent
 of the particulate emission rates.   In  contrast, organic  compounds  after the
 particulate scrubber systems  represented a  much larger percentage of  emissions
 (15.8  to 113.7 percent)  compared with particulate emissions.   In all  cases,
 however, organic emissions  were reduced by  passage  through a scrubber.  The
 nonfilterable  portion  of the  organic emissions  (i.e.,  the portion that passed
 through the sampling train  filter)  ranged from 67 to 92 percent at  the scrubber
 outlets.   On Furnace B-l  the  nonfilterable  portion  of  organic  emissions was 77
 percent of the total  before the fabric  filter.  The remaining  23 percent were
 trapped in the probe or  on  the  filter of the sampling  train.   Only  1  percent of
 the organic emissions  were  collected on  the sampling train filter and probe in
 the scrubber inlet test  on  Furnace  D-l.
     To estimate the reduction  in particulate by passage  through a  flare, one
 testing company heated samples  from closed furnace emission tests in an oven at
 approximately  482°C  (900°F), the temperature at which  particulates  ignite.3
 When the samples were  reweighed, they had lost from 68 to 87 percent of their
 original dry weight.   These fired samples were then returned to the oven and
 held at 697°C  (1290°F) for  15 minutes; a further weight loss of approximately
 11 percent was obtained.  This  overall weight loss of  79 to 98 percent was
 probably due largely to  the combustible content of the particulates.  Of
 course, other  volatile compounds or metals would have been part of this weight
 loss.
     The use of coal and, to a  lesser extent,  coke and wood chips and the lack
 of oxygen in the furnace can contribute to the formation of organic compounds.
Data in Table 5-6 summarize the percent of coal, coke,  and wood feed materials
and the organic emissions from the furnaces  tested in EPA's  environmental

                                      46

-------
          TABLE 5-6.  ORGANIC EMISSION RATES AND
          REDUCTION MATERIALS CHARGED TO FURNACES3
Furnace
A-l
A-2
B-l
B-2
C-l
C-2
D-l
D-2
Organic
emissions
before control ,
kg/MW-h
0.063
0.347
0.247
1 .600
1.270
0.592
0.300
Reduction material in charge,
% by weight
Coal
0
0
22
17.5
21
21.5
3.0
2.6
Coke
18.7
16.8
0
7.8
0
1.7
11.9
11.8
Wood
0
0
14.3
0
29.8
9
0
0
a Information for A, B, and C furnaces is based on Reference 7;
  information for D furnaces, on Reference 8.
                           47

-------
assessment studies.
                   7,8
Based on these limited data,  correlations between  feed
materials and organic emissions apparently do not exist, and the use of coal
does not have a major impact on the amount or the characteristics of these
emissions.
     Compounds in the polycyclic organic matter (POM) classification, also
commonly referred to as polycyclic or polynuclear aromatic hydrocarbons (PAH),
have been identified in the organic fraction of the furnace emissions.  Data on
individual POM compounds that were identified in three tests are summarized in
Table 5-7.  These emissions include data from a single semisealed furnace
producing 50 percent FeSi and one closed furnace producing SiMn and FeMn.  All
of the concentrations were measured before a flare, and in the case of the FeMn
product, also before a scrubber.  Higher POM concentrations were found in the
lighter compounds with a mass of less than about 228 (chrysene and benz(a)an-
thracene).
     Total POM compounds after the scrubber on the closed furnace amounted to
1.0 g/MW-h (2.2 x 10~3 Ib/MW-h) and were much lower than those from the semi-
sealed furnace, which totaled 91 g/MW-h (200 x 10"3 Ib/MW-h).  This lower
emission rate is probably due more to the high-energy venturi scrubber on the
closed furnace [22.4 kPa (90 in. of water)] than to any major process variable.
Emissions from the semisealed furnace were controlled by a low-energy scrubber.
Based on the closed furnace data, it appears that the high-pressure-drop
venturi scrubber is a very effective means of reducing the higher-molecular-
weight, less volatile POM compounds.
     Several of the measured POM compounds are suspected carcinogens.  These
include benz(a)anthracene, chrysene, indo(l,2,3-cd)pyrene, benzo(j)fluoran-
thene, and benzo(a)pyrene.  Of these, benzo(a)pyrene amounted to 1.0 g/MW-h
          o
(2.2 x 10   Ib/MW-h) from the semisealed sealed furnace before the flare.  This
compound was not detected in emissions from the closed furnace.
5.3.2  Aqueous and Solid Waste Organic Emissions
     Limited data are available on the organic content of both liquid and solid
wastes, but specific compounds are not identified.  The only contact-type
process-water discharge is found in plants equipped with scrubbers for partic-
ulate control.  Boiler blowdown end cooling tower water blowdown are other

-------









oo
A
r-.
CO
2:
0
i— «
CO
CO
1— 1
III
III

Cj
"Zi
=>
o
Q_
o
!-<-> .•
t—i
ZT
CO
o
•
LU
_l
1
s
CL.

•
^*
LO

LU
_I
00

*~.














: *
2?S
•c.°.
SO
•03
•O

JO '
= 4J c:
•• ens:
3 0) 0)
4-H-U.

13
0)
V)
3
j cnc
,c.°
'C+J
3 (J
T3
CM5-
1 O.
0

(/)r-
IB C*>
ETS
3 3T3
t_-o p
.L>
ID CM
njovo
III
%*~

























I
^



^O
E

Ol
s



i
^
en



V)

s
1
x:
i
|
cr


5
1














t
.c
c
c

c.
o







o
fx. «3- !>• OO in CM CM CO CO
Of CT> o o> in in o cor— o r^ co •— in co CM o i— r-
^C\3«a- IO «*C30T Cvjr-C5CO omm
vocn ci •* o CM •* os ino r- ro •— o 10 r— oooo
r— CM CM CM i— «* f
CM CM


. CM 01
«j ,— r— in in o o
co mo ooo o
o o o ooo o




O in 10

in •— o CM CM o o
r— CM O OOO O .
CM *^ m os o y3 o i** ^^ *~"
Os ' CM CM in os r> ^j* o ^ f^ os ^t~ *x *— • »"• f™* CM o *o 01 ui ^
l^ ^-r-lO  CMOCMOt— Or-O O


i^«j- c— «aco«a- oro
o'«ocor».io. «*moosoji— in eo in •— tn «* 10 r-cMOO o
in^oocoo r- r-^coc3oon-odc3 o coocooi— •— co.o •—
r^i — r— * — CMCM " — • -•





01
OJ C
CO)
0) O
Q) t. (O
c o s-
4-»»
^-, OJ +J 1_-O»— -Mt-O.
^_tjc O-O uc oj-ctdisw s c«a -^

^ "^ ti C Q) w j^1*". u *^ w* ^t ?[ 5 S S Si) ef J»"Q ,, ^ £• f— M^ i
 CS CL£ ' ^ *f~ *2 -C CD O. O *f- **-» *O D- CO" •*— t. HJ O *O

C O O > C Q. OT rt3 C *^ -q -^> "T" •^•^T-"^!- flJ*wT— ^**O>*^-' C N N N fl



^" *"* ^ y. rt ^ t m S S ^ S flu '""«' S S 5 ai dJ flj a3 ' "S T c 
-------
wastewater sources.  Solid wastes, however, are generated by all  plants in  the
form of collected parti culates, slag, broken fabric filters, and  sludge (when
scrubbers are used).
     Table 5-8 summarizes data on the organic content of scrubber water outlet
streams obtained in the recent EPA Environmental  Assessment Study.7  Individual
organic compounds were not quantified, but a number of high-molecular-weight
compounds were identified in the scrubber water of Furnace C-2.   The more
predominant identified compounds were:
                                      l *
     Fluoranthene and/or pyrene
     Benzo(a)pyrene and/or perylene and or 10,  11-benzofluoranthene
     C15/16 benz°Pyrene> Possibly with a naphthalene group.
     Analyses of grab samples of particulate collected in fabric filters
                                                                         on
open furnaces showed concentrations of organics in the range of 65 to 384 ppm
by weight.    In one case, POM compounds accounted for 1 to 3 percent of the
total organics in the particulate.  Concentrations of 7031  ppm of organics were
detected in the particulate matter from a fabric filter controlling the emis-
sions leaking from the top of a mix-sealed furnace, and POM accounted for 15 to
20 percent of these organic compounds.
     Other studies on organic concentrations in liquid streams include data on
the oil content of wastewater streams, which indicate a concentration of less
than 2 ppm by weight.

5.4  OTHER ATMOSPHERIC EMISSIONS
     Limited data have been obtained on other atmospheric emissions from
ferroalloy furnaces.  These include data on sulfur acids, gaseous hydrocarbons,
and metals.
5.4.1  Sulfur Oxides
     A joint EPA-Ferroalloy Association report on ferroalloys stated that
sulfur oxide emissions were less than 20 parts per million  (ppm)  and never
exceeded 3.2 kg/h (7 Ib/h).    For this reason, sulfur oxides are rarely
included in an emission test program.  The results of one set of  recent tests
showed concentrations of 69 to 74 ppm for a 20-MW silicon open furnace and 83
                                      50

-------
  TABLE  5-8,  ORGANIC CONCENTRATIONS  IN SCRUBBER WATER
                   DISCHARGE  STREAMS
Furnace
A-l
A-2
B-2
C-T
C-2
mg/ liter
11. 2a
14.1b
551
134
71
kg/MW-h
0.11
0.12
1 .55
0.97
0.48
*8.2 percent adsorbed  on  solids.
344 percent adsorbed on solids.
                           51

-------
ppm from a 25-MW ferrosilicon open furnace.   These concentrations amount to
approximately 45.5 kg/h (100 Ib/h) for each furnace, or 1.8 to 2.1 kg/MW-h (4
to 4.6 Ib/MW-h).
     This value is considerably higher than the previously reported data,
possibly because of the higher sulfur content of the petroleum coke used in the
electrodes.  Because the sulfur content of the feed materials was not measured
during the emission test, however, no direct correlation between feed compo-
sition and emissions can be made.
5.4.2  Nitrogen Oxides
     No measurements have apparently been made for these compounds, but con-
centrations are expected to be very low because of the lack of oxygen in the
reaction zone.
5.4.3  Inorganic Constituents
     Tests for the inorganic constituents of the furnace exhaust gases were
conducted on three furnaces as part of EPA's environmental assessment stud-
    17 18
1es.  *    Table 5-9 summarizes the results of these tests.  The tests on
Furnaces D-l and B-l (both open furnaces) were performed on the gas stream
preceding particulate control systems, and the test on Furnace D-2 (a closed
furnace) was performed on the gas stream after it had passed through a venturi
scrubber with a high pressure drop.  Analyses of these samples were performed
largely by spark source mass spectrometry techniques on each fraction of the
sample collected with a source assessment sampling system.  The fractions were
then added to obtain the values in Table 5-9.  Arsenic, mercury, and antimony
were analyzed by atomic absorption.
     Although data from the tests on Furnaces D-l and D-2 cannot be directly
compared (because they made different products), the much lower values for all
elements obtained in the test on  Furnace D-2 after the scrubber still show that
the scrubber was very effective in reducing the emissions.  The composition of
these components varies widely, depending on the material charged.  Major com-
ponents were not quantified because they could not be analyzed by these tech-
niques.
                                      52

-------
(C
 rj

 x
 o
 

O* Q
a=. *~

i E
2 .2
-o -o
^ c



^- , CM
lO CM
CO r—




S S
01 O
^•'S

"i s
O CM



C? 0
g **-
5 1
e s-
•s «c
- *•
CM CM
| g


tn

r— O



3 3
i 1
, to



*


0
«>

o
o

1 '1
T3 *i—
0 S-


in
in or»
in r—
U3 O




£ °

E °
o



1
E r-
3 r—




o
o


g


,

-o


m
0
CM"




0
s


CM



d>
|
s-

i*.





CJ





•a
o
V)



5

o
o
CO
r-T

£

J=
4->



in
CM




§
f*.

g





1

<->
^

1
CM


O




.3
4J
1
4->
V)



g


in


CO
o

E
3
•M



£




«

<_>
Z





*0

0
• ' . to' r«- •
Oe»j^S2 S1**^!^. **
coocvi o r*. ^- o CM. f*> •* »— r— o CM ac r-*
o o o .
o oooo o o o o o o o o
r*. CM -*— co * * r- co f— E in
. CO r*. r-

CM CM .

O O O r— CM O CO OOJ O


E E E'
i -2 e' 1 E e ' 1 § E 1 •? e 2
S- f-1 (- '3 -r- 3 3 •*- *»' 3 •»- ^ 3 C
3 (d 3 -r- r— -r- -r- C « •»- "O t. *t- O
^- -C r- S- m O3 C4-»C « C4-»4J C U
3 <0 0) 4) .C . .C .C >r- n- 3 S- (O +J -4-» -^ -r-



cor^tn-cocMvoo' O1 O esjr—
« • — o in «s- in « —


£ 2 S | S § g i" ^ S
co" *" «^

fO
O O f*«- in • r% <_> ^o O* O* e_> o «— ^
z s: — . . S . z M z sc • •
U> . CM CM t\J 0

=J 3> C 3 C 1 U E § >>
'£c£>;S'c lOTEE^^i'wEi
• 4) > (U 'i— O 3 > TD J^ JQ S-T-CTl,— tft+J B3 (A C
O>C S- r— O E >t r— O(O4->fQ (U



o^ vn CM o co • cj ^ in o> co PO • CM
ZECM • **» C O* Z • .lO-r-^O .
' •—" CO f~" CO ^" O O CT> »"• "^ f~~ O
CM



O CO O O
o o o co co ' ' - o o
O-r—COOCMOOOOOOOOCO
r- CO CO

QJ ^f O CO I-. O CM
O*OmjJCJCMOroCMr-C_JCMCMCMOCO
"*^- •— OCOO^.Of^f-OCO



(U E 3 E t-
ja •»-•(- o o ITS O.O.T- p o o^- g~ *o c:

UOOOOOUOUJUIU-CSOOU)^




, .
i-* .N!
QJ
•s %
3 '
b r-T
(/> 1
3 0

O) C
« 1
S- J.
(0 t/)
TJ O
if '£
3 n
o c
I- O
OJ u

(d C
•o in
csl
'V o'
0 CM
4-> 0
8 1
o particulate
Furnaces D-1 i
t |
fc- ' en
Q. C
T3 -2
0) 4J
i 2

s s
I 0
e 1
? _>,
»— a.
i i
CO
^- .c
s t
IB
E S
3 "
U. +J
i- 0) •
*l =
2 e <
wP ^!~ ,!
*D 43 U
O
•M

'i
                                                                                                        r-     c
                                                                                                              s i
                                                                                                              o '£•
                                                                                                              "O '-C

                                                                                                              u>  a-
                                                                                                            .  C 4J
                                                                                                          a-5 a
                                                             53

-------
     Of special concern are those metals the EPA has designated as priority
pollutants.  These are summarized in Table 5-10 on a mg/MW-h (10~  Ib/MW-h)
basis.
     No detailed data on particle size were obtained in these assessment
studies, but the cyclone catch in the particulate sampling trains confirms
previous data indicating that the particulates are largely less than 10 micro-
meters and frequently less than 1 micrometer in diameter.
                                       54

-------
              TABLE 5-10.  FURNACE EMISSION RATES OF SELECTED METALS9
                             [mg/MW-h (10-6 lb/MW-h)]
Metal
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Furnace—product
D-2— SiMn
0.13 (0.29)
62.5 (138)
0.017 (0.037)
MC
MC
MC
2.75 (6.06)
4.25 (9.36)
MC .
0.21 (0.46)
0.14 (0.31)
0.30 (0.66)
MC
D-1-- FeMn
475 (1,046)
12,000 (26,432)
2.5 (5.5)
1,675 (3,689)
13,000 (28,630)
8,500 (18,720)
MC
127 (280)
1,000 (2,200)
375 (826)
250 (551)
750 (1,650)
MC
B-l--50% FeSi
2,584 (5,692)
866 (1,907)
1.3 (2.9)
1,811 (3,990)
5,751 (12,667)
MC
MC
3.8 (8.4)
3,990 (8,790)
342 (753)
6,901 (15,200)
54 (119)
MC
 MC = M,ajor component.

aRates for D-l  and B-l represent uncontrolled emissions; rates for D-2 represent
 measured  emissions after a venturi  scrubber.
                                      55

-------
                           REFERENCES  FOR  SECTION  5
     Memo anc)  attachments  from  C.  Harvey,  PEDCo  Environmental,  Inc.,  to  R. W.
     Gerstle,  PEDCo  Environmental,  Inc., April 25,  1980,  transmitting informa-
     tion from EPA Region  VI.

     Letter  from M.  Hayward, State  of  Iowa  Department of  Environmental Quality,
     to R. W.  Gerstle,  PEDCo Environmental,  Inc., June 3,  1980.

     Letter  and attachments from M. Hayward, State  of Iowa Department of
     Environmental Quality, to  R. W. Gerstle, PEDCo Environmental,  Inc., June
     3, 1980.

     Letter  and attachments from P. A. Nelson, State of Washington, Department
     of Ecology, to  J.  Zieleniewski, PEDCo  Environmental,  Inc., April  18, 1980.

     Copy of Emission Test Report from State of  South Carolina.  Sent  to R. W.
     Gerstle,  PEDCo  Environmental,  Inc., May 15, 1980.

     Copy of Emission Test Report from R. Gore,  State of Alabama to R. W.
     Gerstle,  PEDCo  Environmental,  Inc., June 16, 1980.

     Westbrook, C. W., and D. P. Daugherty.  Draft Copy of Environmental
     Assessment of Electric Submerged-Arc Furnaces for Production of  Ferro-
     alloys.   Research Triangle Institute.  EPA  Contract 68-02-2630.  March
     1980.   206 pp.

     Rudolph,  J. L.,  et al.  Ferroalloy Process  Emissions Measurement.  U.S.
     Environmental Protection Agency, Research Triangle Park, North Carolina.
     EPA-600/2-79-045, February 1979.  185 pp.

     Yerino, L., and H. Belknap.  Determination of Compliance Status and
     Evaluation of Baghouse Rebuilding Efforts SKW (Airco) Alloys Plant.   PEDCo
     Environmental,  Inc.  EPA Contract No. 68-01-4147, Task 83.  Auqust 1979
     pp.  10 &  11.

10.  Memos from L.  Gibbs of PEDCo Environmental,  Inc., to project file re-
     garding contacts with regulatory personnel  in Ohio,  April 3 and 14,  1980.

11.  Information from C. Mikoy,  Ohio Environmental  Protection Agency,  North-
     east Office,  to  R. W.  Gerstle, PEDCo Environmental,  Inc., April 17,
     1980.
1.



2.


3.



4.


5.


6.


7.




8.



9.
                                     56

-------
12.  Telecori.  P. McManus, U.S. Environmental  Protection Agency,  Region  III,  to
     R. W. Gerstle, PEDCo Environmental, Inc., April  22, 1980,  regarding
     compliance status of ferroalloy plants.

13.  Reference 7, p. 187.

14.  Reference 7, p. 67*

15.  Cywin, A., and P. W. Diercks.   Development Document for Effluent Limita-
     tions Guidelines and New Source Performance Standards for  the Smelting and
     Slag Processing Category.  U.S. Environmental  Protection Agency.  Wash-
     ington, D.C.  Publication No.  EPA-440/l-74-008a, February  1974.   169 p.

16.  Dealy, J. 0., and A. M. Killin.  Engineering and Cost Study  of the  Ferro-
     alloy Industry.  U.S. Environmental Protection Agency, Research Triangle
     Park, North Carolina.  EPA-450/2-74-008,  May 1974.   p. VI-48.

17.  Reference 7, pp. 119 and 120.

18.  Reference 8, pp. 47 and 48.                        .
                                      57

-------
                                  SECTION 6
         PRODUCTION PROCESSES AND BEST AVAILABLE CONTROL TECHNOLOGY

     Because no ferroalloy facilities in the United States are subject to NSPS,
this section deals with process changes, the best control technology currently
available, and the operational problems of this technology.  It also presents
cost data for fabric filter systems, by far the predominant method of emission
control.
     The type of particulate emission control used varies with the type of
furnace.  Scrubbers predominate on closed and semiclosed furnaces, and fabric
filters are by far the most widely used control devices on open furnaces.
Tapping fumes are generally vented to fabric filter systems (either separate
systems or the furnace's own control system).

6.1  FERROALLOY PRODUCTION PROCESSES
     As described in Section 4, ferroalloys are produced by the following
processes:   electric submerged-a"c furnace at 31 locations, metalothermic at
8 locations, and electrolytic at 4 locations.  The single blast furnace, in
operation until a few years ago, is no longer in use.   Vacuum and induction
furnaces, which are in limited use, are alloy refining processes for the
production of specialty metals.  Only the electric submerged-arc furnace is
significant as far as air pollution is concerned, partially because of its
widespread use and partially because of the copious amount of fume it gen-
erates.  This is the only ferroalloy production furnace subject to the NSPS,
and it is still the major method for producing ferroalloys.
     Ferrophosphorous is a byproduct of manufacturing phosphorous by the
electric arc furnace process.  The emissions from this process are unique
                                      58

-------
because the product (phosphorous) is condensed and collected from the furnace
exhaust gas stream.  The ferrophosphorous is a slag byproduct that is peri-
odically tapped from the furnace.
6.1.1  Electric Submerged-Arc Furnaces
     Descriptions of the electric submerged-arc furnace can be found in the ,
                                                         90
open literature, and they are reviewed only briefly here. '
     Figure 6-1 is a flow diagram of a typical ferroalloy production facility.
The electric submerged-arc furnace in which the smelting takes place consists
of a hearth lined with carbon blocks.  Openings in the hearth permit tapping
(or draining) of metal and slag.  The steel furnace shell and its hood or cover
components are water-cooled to protect them from the heat of the process.
Car'bon electrodes are vertically suspended in a triangular formation above the
hearth.  Normally there are three (sometimes more), and they may be prebaked or
of the self-baking, Soderberg type.   These electrodes extend 1 to 1.5 m (3 to 5
ft) into the charge materials.  Three-phase current arcs through the charge
materials from electrode to electrode, and the charge is smelted as the elec-
trical energy is converted to heat.   Coke and other reducing materials that are
added to the furnace react chemically with the oxygen in the metal oxides to
form carbon monoxide and reduce the ores to base metal.  The furnace emits
byproduct carbon monoxide along with entrained particulate matter and metal
vapors.
     Power is applied to the furnace on a continuous basis, and feed materials
may be charged continuously or intermittently.  Molten ferroalloy and slag are
intermittently tapped into ladles from tap holes in the lower furnace wall.
(Furnaces producing calcium carbide may be intermittently or continuously
tapped.)  The melt is poured from the ladles into molds or casting machines.
After the product cools and solidifies, it is crushed, sized, and loaded into
rail cars for shipment.  Slag may be disposed of in landfills, but most is sold
for road ballast.
     For reduction of atmospheric emissions, the furnaces and tapping stations
are hooded and the off-gases are ducted to a particulate control device (scrub-
ber, fabric filter, or electrostatic precipitator).  The configuration of the
hood and/or furnace roof determines whether the furnace is categorized as open,
semisealed, or closed.
                                      59

-------
                                                 ro
                                                   
                                                   U  V)
                                                   0  >,
                                                   S-  CO
                                                   O  S-
                                                  •r- •)->
                                                  •P  C
                                                   O  O
                                                   =S  O
                                                      Q.
                                                  CU  (O
                                                 IO •!-
                                                  C71  O
60

-------
6.1.1.1  Open Furnaces--
     In the open furnace, a canopy hood through which the electrodes extend is
located 2 to 2.7 m (6 to 8 ft) above the furnace's upper rim.  This opening
between the furnace and hood allows large amounts of ambient air to enter the
hood and exhaust system.  As the air combines with the hot gases, the carbon
monoxide and most of the organic compounds are burned and the furnace emissions
are diluted and cooled by the ambient air.
     This type of furnace is by far the most popular in the United States
because of its product flexibility.  However, the large opening around the hood
allows fumes to escape if sufficient draft is not provided.  Control equipment
must, of course, be designed to handle the large volumes of gas [8,520 to
18,720 Nm3/MW-h (30,000 to 660,000 scf/MW-h)] inherent in an open furnace
design.  Many open furnaces are partially hooded to minimize air intake, and
still allow complete combustion of furnace gases.
6.1.1.2  Semisealed Furnaces—
     The semisealed (or mix-sealed) furnace has a water-cooled hood that fits
tightly around the top of the furnace and is vented to an air pollution control
system.  The electrodes extejrid down through the hood, and raw materials are
charged through annular gaps around each electrode.  Because the seal provided
by the raw material mixture around each electrode is not airtight, fumes may
leak out unless sufficient draft is provided by the air pollution control
system.
     Much less outside air is drawn into a semisealed furnace than into an open
furnace, and pollutarlt concentrations are therefore much higher.  The resulting
gases are also rich in carbon monoxide.  Only four U.S. plants currently use
the semisealed furnace.  This type of furnace is not used in the United States
to produce silicon metal or alloys with more than about 75 percent silicon
because it cannot be readily stoked from the outside.  -If the high silicon
mixes are not stoked, bridging and resulting pressure buildup from entrapped
gases may occur in the furnace.  This condition leads to "blows" (or possibly
explosions) when the gas breaks through the mix or the bridged material col-
lapses.                                        ,
                                       61

-------
6.1.1.3  Closed or Sealed Furnaces--
     This type of furnace utilizes a tight-fitting, water-cooled hood on top of
the furnace, which is vented to an air pollution control system.  Raw materials
are fed through separate sealed chutes, and the electrodes penetrate the hood
through seals.  The furnace is thus completely sealed and operates under a
slight positive pressure regulated by the fume exhaust system.  No outside air
enters the furnace system, and high concentrations of CO (80 to 90 percent) and
particulates are emitted.  Exhaust gas volumes reportedly range from 200 to 260
Nm3/MW-h (7060 to 9180 scf/MW-hj, and uncontrolled particulate concentrations
range from 11.5 to 70 g/Nm3 (5.0 to 30.6 gr/scf).4
     From the standpoint of air pollution, a closed furnace is the most de-
sirable because all its fumes exhaust through an emission control system and
the total volume of exhaust gas is only 2 to 5 percent of that from an open
furnace.  Only two closed furnaces are currently in operation in the United
States; both at one plant, they pfoduce silvery pig iron containing less than
20 percent silicon and smaller amounts of other alloys.  Ferroalloys with
higher silicon contents are more difficult to produce in a closed furnace
because they tend to bridge over in the furnace if they are not stoked, and the
closed nature of the furnace makes stoking from the outside much more dif-
ficult.  Lack of stoking can lead to explosions from trapped gas.
6.1.2  Process Modifications
     In the United States little change has occurred in the process technology
of this industry over the last 5 years.  Investigation of beneficiation of feed
materials, operating practices, and possible mechanical modifications to
furnaces continues in an effort to improve operations and minimize emissions.
Generalizations cannot be made regarding design and operation, however, because
a specific evaluation of each furnace type, raw material, and product mix is
required.
     The split furnace is an innovative development in furnace design.   In
this design the furnace is divided into two separate parts.  The upper part is
a relatively narrow ring with flat interior surfaces.  This upper ring rotates
more rapidly than the lower furnace portion (e.g., in one design the ring
rotates  at 0.1 revolution per hour (rph) while the furnace rotates at 0.01
rph.)  This rotation around the stationary electrode has a mixing effect on
                                       62

-------
the furnace contents and reduces bridging and crust formation problems.   A
small (8.5^MW) closed split furnace producing 75 percent FeSi has been operat-
ing in Norway for several years.
     A study by Battelle reports that sealed furnaces in Japan are producing
FeMn (high-, medium-, and low-carbon), SiMn, FeSi, high-carbon FeCr, and SiCr.
In these furnaces stoking devices are inserted through seals in the furnace
walls.                       :  .  .  '
6.2  CONTROL TECHNOLOGY
     No basic changes in control technology have occurred in this industry
since promulgation of the NSPS in 1976; however, some changes in control
device design arid operating practices have evolved and have resulted in im-
proved reliability of these devices on ferroalloy furnaces.
6.2.1  Fabric Filter Control System
     The fabric filter control system is generally the method of choice for
controlling particulate emissions from the open submerged-arc furnace.  This
system is used at 27 of the 31 plants in the United States.  The following
discussion is based on generalized information from fabric filter manufacturers
and users contacted for this study. * "'    Site-specific conditions will
vary.
     The predominant fabric filter system is the pressure type, in which the
fan is on the inlet or dirty side of the filter.  These systems exhaust di-
rectly from the top of the baghouse and have no final stack.  Cleaning is
accomplished by a reverse-air system, a mechanical shaking system, or a com-
bination of both.  Pulse-air cleaning systems are also used at a few instal-
lations.  During the cleaning cycle, which is either timed or triggered by
pressure drop and lasts for 1 to 2 minutes, the compartment is isolated from
the gas stream.  Because temperatures are in the 177°C (350°F) and higher
range, glass fiber and Nomex fabrics are the most popular, and bag life is on
the order of 1.5 to 2 years.  In systems with reverse-air cleaning and glass
fiber bags, air-to-cloth (A/C)  ratios are approximately 37 m/h (2 ft/min) as
shown in Table 6-1.  In systems with mechanical shakers and Nomex bags, A/C
ratios are a slightly higher 55 m/h (3 ft/min).  In systems with pulse-jet
cleaning, A/C ratios are 92 to  130 m/h (5 to 7 ft/min).
              •      •    •:  .  '         63

-------
to



to

l/>

D;
LU
H-

t—i
U.

o
1—4
QC
CQ
o;
e

S
o

or
LU
O
LU

fe
O
a.
LJJ
a:
 i
uo
LU
_1
CQ
i
(U  a
QJ CJ
(— o
i, _
4-> C
1%
i E"
||
. o

m
|

Product
O)
ce
at
u
£o
a.
o






S
^^.
-is-
Reverse-air
Reverse-air

in x
D» $- Z
QJ
CD r-. t- 00
.p. QJ
(/I U_

ID CO
CO •—
CO • — •
Reverse-air




o> s
CJ O
^
o
CO
I
o
CO






o
o
in
CO
ID
CO
o
^
cn r—
Reverse-air

0)
JD
CO
CD

^
in
oo
I
-






I
i —
o
s
in
r-.
OJ
10 ^~
cn oo
CO* — •


I
tO
CD
11
CJ O
^
s:
oo
(=
01
Q.
CD
,-


CO
ro
CO


OJ
S
O
in
oo
§"~T §
•I** CO
« (0 »
i — Q. r-.
CO O
CM J= ^~
nJ
O 
rd ro
CD CD
CJ
to *
C'tJ T-
CJ _J (/)
«c
CM r—
EC C
O) OJ 0}
Q.O. CL
00 0
OOOJ r—





I
in
o
*3-
CO
~
CD
O
CD"
ro
ID CO
ro i —
Reverse-air

(U
V) X
CD 3=
t/)
O)

I
1
O
—


in . — * .— ^ oj , —
• CO **" CO * Cn
^J- i — • . OJ t
' i ro ro 'CO
CD i£) r— C\J k_y
CD ^J"
CD
CD CO f~^
O ID CT>
in • oo
f— CD OJ
* — O O O CD
in CD \o CM ID CD
O O r- CO OJ in
OJ <3- I | ||
i i o «*• r-v o
• — m *a- co r-» in
cn r^. •— oj r— . ro
I ~i
CO O O
« o *
VO CD O
OJ "CD
i — O OJ
0 *""" 0
o  >
tn
(O
OJ cn
X X O J= 0) &-
e e u ro ra .Q
Z Z 0 t3 S 'C
» OJ
b. »«^— tn
•r- OJ M •»— CJ -r- iQ OJ
•r-s-'n- c; i_oi-u_-i-a» 
-------
     Fabric weights vary from 170 to 475 gm/m  (5 to 14 oz/yd2), and a
typical bag is 28.75 cm (11.5 in.) in diameter and 9.2 m (30 ft) long.
Pressure drops reach about 3.2 kPa (13 in.  of water) prior to a cleaning
cycle.  When a system is operating properly, it achieves collection efficien-
cies in excess of 99 percent.  Broken bags are discarded in landfill areas.
     Table 6-1, which presents fabric filter design data obtained from the
recent literature, shows the range of fabric filter applications.  Glass
fiber is the fabric cited most frequently.   Air-to-cloth ratios usually ran
slightly less than 37.8 m/h (2.0 ft/min).  One installation reported that a
higher pressure drop of 3.7 to 4.5 kPa (16 to 18 in. of water) occurred at an
air-to-cloth ratio of 51.0 m/h (2.75 ft/min).  The limited data reported on
particulate loading before and after the filter systems showed efficiencies
in excess of 99 percent.
     In Europe, pretreatment of the gas stream has been accomplished by the
use of a perforated rotating drum dust agglomerator filled with ceramic
      18
balls.    The gas stream is first cooled and then passed through the agglom-
erator, where the dust impinges on the ceramic balls.  The resulting large
particles are then collected in a fabric filter system, which uses polyester
or acrylic bags and has an air-to-cloth ratio of about 91.3 m/h (5 to 1
ft/min).
6.2.1.1  Fabric Filter System Costs-*
     Costs of fabric filter systems depend mainly on the gas flow, type of
fabric (which is related to the operating temperature), and the number of
bags requ" red (related to air-to-cloth ratio).  Based on discussions with
                   p Q          -   '  .
equipment suppliers '  and generally used engineering factors, the cost of a
fabric filter system can be calculated by the following equation:

          Cost> •* =  -d      ^5'64 + 1<95 x bag cost^ + 47>160
          (Jan. 1980)
where:    acfm = actual cubic feet per minute
           A:C = air-to-cloth ratio, ft/min
          Bag cost is expressed in $/ft2 as follows:
               Polyester                0.65
               Acrylic                  0.79
               Nomex                    1.59
               Coated glass fiber       0.74
                                       65

-------
The entire installed cost (including both direct and indirect items) would be
approximately 2.4 times the fabric filter system cost (see Appendix).
     A fabric filter system treating 343,000 m3/h (200,000 acfm) and having
an air-to-cloth ratio of 36.6 m/h (2 ft/min) with Nomex bags would cost
$921,000, and the total installed cost would be approximately $2,200,000.
6.2.2  Scrubbers
     High-pressure-drop venturi scrubbers have been applied successfully to
a number of ferroalloy furnaces, especially closed and semi sealed furnaces.
         19-22
Table 6-2      summarizes reported scrubber data.  Pressure drops in the range
of 13.7 to 22.4 kPa (55 to 90 in. of water) make the use of these scrubbers
very energy-intensive, especially when large volumes of gas must be treated.
On the order of 5 to 10 percent of the furnace power requirement may be used
                                                             0-3
by the fan motor to draw the gas through the scrubber system.
6.2.3  Other Control Systems
     One installation of a sand-bed filter (gravel bed) was reported on a
closed ferrosilicon furnace in Sweden.  The furnace utilized a water-cooled
                                                              o
hood, and filter system emissions ranged from 400 to 500 mg/Nm  (0.17 to 0.22
gr/scf).  This concentration would result in a visible plume, but because of
the fairly small quantity of gas exhausted from a closed furnace, this
installation complied with an emission limit of 15 kg/1000 kg of product (30
lb/ton).12
     One electrostatic precipitator is in operation on an open furnace in the
United States.  This system utilizes gas conditioning with ammonia to enhance
particulate resistivity and increase collection efficiency.
6.2.4  Flares
     Flares are used on closed and semisealed furnaces to reduce carbon
monoxide emissions.  A flare, which is essentially an open afterburner,
should also reduce combustible particulate and organic matter.  Because
actual test data on flares are not available, an approximation of a flare's
ability to reduce organic or particulate matter must be based on afterburner
data.
                                       66

-------
               •On
               -se
UI
CD
CQ
Z>
D;
o
o
LL,
03
Q
o
a.
LU
a:

                                                     S«>  ^^
                                                     •^ o. ^m
                                                     CM (
                                                           ^i£ <*
            «*£T  "l^T
             .; in  ^.in
                                                       kO
                                                       CO
                                                                 CO
                                                                          CT»
^*-CM     \o  ro o  in r-
O» O  CO C3  II    • O
 . o  ^r f-  a* o  ro  •
tn  -  CM  •   - r—  «3- O
  O     O  CM O    •—-
  ^—^    «—•  CM  •
                                                                              o o t
                                                                         f O   *• *•
                                                           ._-.  _.     j-  •  oor-
                                                     mcM  r^*a-  CM*— CM i— .o in
                                        O CJ
                                        O CTi
                                        CVJ CO
                                                           o o  o o en o
                                                           \O ^  VO O *3" CO
                                                           CM in  CM in CM ^-
                            O O  Ci O
                              SCO  O »—
                              iD  O ID

                            o CM  ur> at
             ,
 »co  ^OCM.  ^cs  oo
ir>  . r-r-  CT.O  00
f— co  *3-ro  C7>tr>  otui
 ||    A  *   r. »   r. »
OO  C\JCT»  CO CM  OO CM
0*0  r»-o  om  oco
^^  f—<—  CMr—  CMi—
                                                     CMO  CM1O  i— CO  r—CO'
                                                       CM     r-     i—   'i—
                                        1
                                              1     s
                                                                        g.    2
                                                                       O     O
                                              67

-------
     The combustion efficiency of a flare is primarily a function of tempera-
ture, turbulence, oxygen content, and residence time.  Because of the rapid
cooling and dilution inherent in flares, residence time is short (about 0.1
second).*'  Temperatures of 581° to 994°C (1078° to 1822°F) have been measured
experimentally in a small pilot plant equipped with flares burning natural
    24
gas. ^  Combustion of liquid droplets requires vaporization, and the rate at
which a droplet burns is dependent on its,.s4ze. and,,,thfi temperature.  This
relationship is expressed by:
                       2

          .  = 29,800 md
where:
t
P
m
d
T
              time required for combustion, seconds
              partial pressure of oxygen, atmospheres
              molecular weight
              droplet diameter, cm
              temperature, °K
     For a 200-molecular-weight hydrocarbon, this formula gives a residence
time of 0.0084 second at 1000 K (1340°F) and an oxygen content of 0.1 atm for
a 50-micrometer droplet.  This relationship indicates that more than enough
residence time exists in a flare to vaporize and burn small droplets.
     For solids, the combustion relationship is much more complex and cannot
be readily predicted.  For a 1 -mi crometer particle, estimated combustion
times range from 0.043 second at 793°C (1460°F) for a coal char to 175 seconds
at 827°C (1520°F) for a soot particle.26  Based on this minimum amount of
data, one can only assume that most solids composed of volatile organic
matter probably would be burned in a flare, whereas inorganic and carbon
particles would not be burned.

6.3  CONTROL OF TAPPING EMISSIONS
     Molten ferroalloy is removed from the electric submerged-arc furnace
through a tap hole that is flush with the floor of the carbon hearth.  The
tap hole is closed by the manual insertion of a carbon graphite plug after
completion of tapping or by means of a hydraulic or pneumatic mud gun.
 Based on an assumed 16.6 m/s (50 ft/s) gas velocity and a 1.7 m (5 ft)
 long flame.
                                      68

-------
     When the molten metal is ready to be tapped, the tap hole is pierced by
drilling, by use of a single shot pellet fired from a gun-like piece of
equipment, or by oxygen lancing.
     The molten ferroalloy passes through the open tap hole to a trough ar-
rangement fixed to the furnace shell and then to runners that direct it to a
ladle or, in some instances, to a runner going directly to a casting bed.
When the ladle is filled with the molten metal, an electric overhead trav-
eling (EOT) crane transports it to the point of deposit, where the metal  is
then poured into a bull ladle, a reaction ladle, a pigging machine, or a
shallow cast bed.
     Emissions occur during the following:
     o
     o
     o
     o
          Piercing of the tap hole, if drilling or lancing is used
          Tapping at the furnace proper, the runners,  and ladle
          Transporting in the ladle
          Pouring into the bull ladle, reaction ladle, pigging unit,  or the
          cast bed.
     Problems encountered in the control of emissions  from retrofit installa-
tions include lack of space to install hooding and still  have sufficient room
for operators to do their work; ducting and hooding interference with the
operation of the EOT crane at the tap hole area; and hooding difficulties at
the ladle, especially during pouring, because of floor layout, design, and
existing equipment.  Proper planning could prevent most of these problems in
new furnaces.
     Since furnace tapping takes 10 to 15 percent of the total cycle  time,23
significant emissions can occur.  When the molten metal is poured into a bull
ladle, emissions are very high but of a short duration; whereas, when the
metal is poured into a reaction ladle, emissions are both significant and of
rather long duration.  Emissions are also high but of a short duration when
the metal is poured from a ladle to a casting bed; however, they are  both
higher and prolonged when the molten metal goes directly from the furnace to
an adjacent cast bed during tapping.
                                      69

-------
     Several possible measures are available for reducing emissions during
tapping in new furnaces.

     0    Hooding could be designed to minimize emissionsi  instead of in
          arrangement in which an EOT crane transports the  ladle,  a ladle car
          on tracks or a rubber-tired unit with a ladle tilting device
          operating under a canopy tunnel  enclosure could be used  to trans-
          port the molten metal to the bull ladle or casting bed;  casting
          beds could be designed to eliminate crosscurrents and emissions
          could be collected overhead; and reaction ladle operations could be
          designed to take place in an enclosed building and emissions could
          be collected overhead.   In each  instance, the control  equipment
          would be a fabric filter system.

     0    It may be possible to enclose and vent the tap area around each
          furnace and to use pendant-operated EOT cranes to service the  area,
          independent of the main EOT cranes.   Airflow would be  70.7 to  94
          m3/s (150,000 to 200,000 cfm).

     0    Another possible control  method  is the use of a telescopic emission
          capture device (see Figure 6-2).

     0    The use of an air curtain in the  tapping area.would permit the
          crane access to the tap hole area, runners,  and ladle  (see Figure
          6-3).   Airflow would be 16.5 to 21.2  m3/s (35,000 to 45,000 cfm).

     0    Ladles could be equipped  with covers,  but this would increase
          handling time and maintenance.

     0    The use of a vermiculite  slag blanket  would  minimize emissions  in
          transportation,  but it  would also create pouring  problems.

     0    A  verticle takeoff on the hood over the  tapping area could help  to
          alleviate fugitive emissions.

     0    A  total  building enclosure with an overhead  collection system would
          also  minimize fugitive  emissions  to the  atmosphere,  but would not
          achieve compliance with the NSPS  since visible  emissions  are
          determined at the hood.
                                     70

-------
                              FAR SIDE- ID FAN
                                                  AUX HOIST DRUM
       TRANSFER
         DUCT
                                                          SPLIT RUBBE
                                                           COVER OVER
                                                         TRANSFER DUCT
TELESCOPIC CAPTURE
 DUCT (SIMILAR TO
STEEL HILL SOAKING
   PIT CRANE)
 TRANSFER DUCT
    TO FAN
 AND BAGHOUSE
(CAN BE MOUNTED
 OVERHEAD UNDER
THE ROOF TRUSSES)
                                                                           BLDG
                                                                          COLUMN
             CAPTURE HOOD
              FIXED TO
            SPREADER BEAM
                                                                     HOIST
                                                                     DRUM
                                                  TRANSFER
                                                    DUCT
                                                                       TROLLEY
WALKWAY 1











•

B0
n
TELESCOPIC
CAPTURE
DUCT

E.O.T.
GIRDER

    Figure 6-2.   Ladle/EOT  crane .fugitive emission collection system.
                                                71

-------
ELECTRODES-
                                           EXHAUST TO
                                       CONTROL EQUIPMENT
             ELECTRIC SUBMERGED ARC FURNACE
                      OPENING FOR
                      CRANE  CABLES
          AIR IN
               PICKUP
              AIR AND
AIR FLOW  -FUGITIVES  OUT
 PATTERN
 Figure  6-3.  Air-curtain  fugitive control system.
                         72

-------
                          REFERENCES  FOR  SECTION 6
 1.   Draft  Preprint  from  F. J.  Schottman,  U.S.  Department of  the  Interior,  to
     PEDCo  Environmental,  Inc., April  7,  1980.

 2.   Dealy,  J.  0., arid.A.  M.  Killin.   Engineering  and  Cost  Study  of  the
     Ferroalloy Industry.   U.S. Environmental  Protection Agency,  Research
     Triangle  Park,  North Carolina.   EPA-450/2-74-008,  May  1974.   p.  VI-48.

 3.   Background Information for Standards of Performance:   Electric  Sub-
     merged-Arc Furnaces  for  Production  of Ferroalloys, Volume  1:  Proposed
     Standards.  U.S.  Environmental  Protection Agency,  Research Triangle
     Park,  North Carolina.  EPA-450/2-74-018a,  October 1974.  p.  7.

 4.   Denizeau,  J., and H.  D.  Goodfellow.   Environmental Legislation  Approaches
     and Engineering Design Considerations for Ferroalloy Plants.  In:
     Proceedings of  the Fourth  International  Clean Air Congress (Paper,
     VI-20).  Japan.  May 17-20,  1977.

 5.   Engineering/Mining Journal.   October 1978.  pp. 45, 51.

 6.   Mobley, C.'£.,  and A. 0. Hoffman.  A Study of Ferroalloy Furnace Product
     Flexibility. Battelie Columbus Laboratories.  NTIS No.  PB-247-273/657,
     July 1975.  83  pp.

 7.   Telecon.   Hoffman, C., W.  W. Criswell Co,s Inc.,  with  M. Giordano,  PEDCo
     Environmental,  Inc.   Fabric  Materials for Baghouses.   May  9, 1980.

 8.   Telecon.   R. F. Morand,  Wheelabrator-Frye, Inc.,  with  L. Yerino, PEDCo
     Environmental,  Inc., May 29, 1980.   Fabric Filter Control  Systems.

 9.   Telecon.   G. Applewhite, American Air Filter Co., with L.  Yerino, PEDCo
     Environmental,  Inc., May 30, 1980.   Fabric Filter Control  Systems.

10.   Telecon.   R. Scherrer, Midwest Carbide Co., with  R. Gerstle, PEDCo
     Environmental,  Inc., June  6, 1980.   Emission Control  Systems.

11.   Boegman, N.  Ferro Alloy Furnace Emission Control in  South Africa -
     Policy, Progress, Problems,  and Cost.  Department of  Health, Government
     of South Africa, Pretoria, South Africa.  4pp.  May  16-20,  1977.

12.   Lomo, A.   Pollution Problems with Electric Reduction  Furnaces in the
     Ferroalloy Industry.  In:   Proceedings INFACON.  1974.  pp.  251-257.
                                      73

-------
 13.   Payton, R. N.  Innovations in Ferroalloy Baghouse System Design.
      Journal of the Air Pollution Control  Association, 26:18-22.   January


 14.   Bettanini, C.  Four Years'  Experience with Bag Filters for Ferrosilicon
      Fumes.   Filtration and Separation.   July/August 1977.   pp.  398-402.

 15.   Pollution Control, Materials and Gas  Reuse and Features of Big New
      Quebec  Plant.  Modern Power and Engineering,  May 1974.   4 pp.

 16.   Meredith, W.  R.  Operation of a Baghouse Collecting Silica Fumes.   In-
      Electric Furnace Proceedings,  Air Pollution and Environmental  Control
      Equipment and Processes.   1972.   pp.  69-71.

 17.   The  Fabric Filter Newsletter.   No.  17.   March 1977.  pp.  9 &  10.

 18.   Telecon.   Bentner, H.  P.   Interal  Corp.t  with L.  Yerino,  PEDCo Environ-
      mental,  Inc.   Luhr EKU System.   May 23,  1980.

 19.   Ratzlaff, R.  6.   Construction  and  Operation of a  New Ferromanganese
      Facility.  Union Carbide.   4 pp.   1974.

 20.   Field Test of a Venturi Scrubber in Russia.   Presented  at Second Fine
      Particle  Scrubber Symposium,  New Orleans,  May 2,  3,  1977.  8 pp.

 21.   Sherman,  P. R.,  and  E.  R. Springman.  Operating  Problems  with  High-
      Energy Wet Scrubbers  on Submerged-Arc Furnaces.   Union  Carbide Corp  ,
      Niagara  Falls,  New York; Marietta,  Ohio.   Presented  at  AIME Furnace
      Conference, Chicago,  Illinois,  December  1972.  20 pp.

 22.   Horibe, K.  A Completely Closed  Electric  Furnace  for the  Production of
      75 Percent Ferrosilicon.  In:   Proceedings  INFACON.  1975.  Johannes-
      burg, South Africa,  pp. 91-98.  1975.

 23.   Trip Report.   To  Union Carbide Corporation Plant, Marietta, Ohio.  By
      PEDCo Environmental, Inc.   June  5,  1980.

 24.   Straitz,  J. F.   Flaring for Gaseous Control in the Petroleum Industry
      Paper 78-58.8, presented at the  1978 Annual Meeting of  the Air  Pollution
      Control  Association.  June  1978.  p. 8.

 25.   Rolke, R. W., et  al.  Afterburner Systems Study.  National Technical
      Information Service.  PB212560, August 1972.  p. 196.

26.  Reference 25, pp.  197-198.
                                     74

-------
                                  SECTION 7

                                 CONCLUSIONS
7.1  INDUSTRY GROWTH
     Total domestic production of ferroalloys has remained fairly static
since promulgation of the NSPS in 1976.   Because no new furnaces have been
built or modified, no furnaces are currently subject to the NSPS.  The
industry's annual growth rate has declined from 1.5 percent to zero.  Two
large plants (one at Brilliant, Ohio, and the other at Sheffield, Alabama)
have been shut down and furnaces at many plants have been shut down and some
have been dismantled.  In 1971, 158 electric arc furnaces were in operation,
145 for ferroalloy production and 13 for calcium carbide production.
Currently, approximately 89 are producing ferralloys and 7 are producing
calcium carbide.  No new furnaces are expected to begin operation in this
country in the next few years.  This decline in the industry has resulted
from a rapid increase in imports ancl the fairly static condition of the
domestic steel industry.
     Another trend in this industry is the purchase of U.S. plants by foreign-
based companies.  Approximately 50 percent of the capacity of the plants in
the United States are now owned by or being sold to foreign companies.

7.2  PROCESS CHANGES
     No new process technology has been introduced to this industry, and the
basic technology has remained unchanged.  The electric submerged-arc furnace
is still  by far  the predominant method for producing ferroalloy.  The metal-
othermic  and electrolytic processes are still used for speciality alloy
production, but  they account  for only about 5 percent: of total alloy produc-
tion.
                                       75

-------
     Advances  in  furnace  design  include  the  split  rotating furnace, which
minimizes  buildups  and  bridging  of the mix and thereby  reduces the amount of
external stoking  required.    This  design allows the furnace to be more
enclosed than  an  open furnace and  thus decreases fugitive emissions around
the hood.  No  split rotating  furnaces are currently used in the United
States.

7.3  CONTROL EQUIPMENT
     No innovations in  control equipment or  control systems have been devel-
oped, but  operation of  the existing equipment has  improved as experience has
been gained.   Fabric filter systems are  the  most common means of particulate
control.   When properly maintained, they can reduce emissions to a level that
complies with  applicable  state regulations.  In this country, high-pressure-
drop scrubbers are  used mainly on  semisealed and closed furnaces.  These
devices can reduce  particulate emissions by more than 99 percent and can
achieve emission  levels that  meet  applicable regulations.  In addition,
scrubbers  reduce  organic  emissions  by 16 to  97 percent.
     Flares are used on closed and  semisealed furnaces to reduce carbon
monoxide emissions.  They probably  also  reduce organic and combustible
particulate emissions,  but quantitative  data on their efficiency are not
available.
     An electrostatic precipitator  is used effectively on one open furnace in
this country.
     Control of fugitive  particulate emissions generated by furnace tapping,
ladle transfer, and  casting continues to be a problem on existing furnaces
because it is difficult to retrofit adequate hooding systems.   This is a
site-specific problem that does not require new technology.   Such control
could be designed into a  new  installation more easily than it can be retro-
fitted into an existing one.

7.4  EMISSIONS
     New data on emissions from ferroalloy furnaces have been  gathered as
part of EPA's environmental  assessment studies.   These studies included tests
to determine the amount of particulates,  organics  (including polynuclear
                                      76

-------
aromatic hydrocarbons)> and trace metals in the emissions.   Also,  a number of
tests were conducted on existing furnaces to determine whether they were in
compliance with state regulations.
7.4.1  Organic Emissions
     Of prime interest among the new data are those on organic emissions,
particularly POM.  Quantitative data obtained to date are very limited, and
only rough estimates of industrywide organic emissions can be made.  These
are shown in Table 7-1.
          TABLE 7-1.  ESTIMATED ORGANIC EMISSION RATES AFTER CONTROL
Furnace type
Open
Semi sealed
Closed
Control device
Fabric filter
Scrubber
Scrubber
• • , •
Scrubber
Total organic
kg/MW-h
0.20
0.29
0.15
0.01
emission,3
(Ib/MW-h)
(0.44)
(.0.63)
(0.33)
(0.022)
 Before any further reduction by flares.

When these emission rates are multiplied by the respective furnace capacities,
the estimate of total annual organic emissions from submerged-arc furnaces is
2220 Mg (2440 tons), based on 70 percent utilization.  Data presented in
Section 5 on polynuclear organic matter show that POM compounds represented
from 8.3 to 75 percent of the total organic emissions after a scrubber and
before the flare.  The single test on a semisealed furnace after a scrubber
showed that BaP accounted for 0.84 percent of the organic emissions.
     The ground-level ambient air impact can be estimated by an atmospheric
dispersion model.  Table 7-2 presents ground-level concentrations of organics
obtained by utilizing EPA's PTDIS dispersion model for selected 20-MW plants
at a windspeed of  4.4 m/s  (10 miles/hour) and a Class C atmospheric stability.
These estimates show that the highest levels occur 1 to 2 km from the stack
and that an open furnace with a scrubber represents the worst case with 24-
      '         '        •               3                      •
hour concentrations of 1.1 to 1.6 ug/m  .  This is because the scrubber water
lowers the stack gas temperature, which in turn decreases dispersion.
                                       77

-------
 U-

 oo

 1—1
 2:
 •=c
 cu

 o

 u_
 o

 2E
 O
 UJ LU
 O O
    D;
_]  I

a CM
cs

Q
LU
oo
UJ
CM
uj
_ i
CO


10 3.
c-*:
o u
rd 
>
i
> 0)
- u
) >r«
1 QJ
> -a
a)
CL
cu
o
re
c:
i-
u_
r-oo
«JCM
o o
0 0
•— o
o o
CO CO
00
IOCO
o o
0 0
§§
0 0
CO
0
CO
o
~
o
LO
CO
o
S.
u 
IT) r—
o a
0 0
LO
•d- -a-
•— 0
o o
CM C3
•a- 1—
CO i —
CO O
CTi CO
00
"~ CM
LO
CD
CO
CO
CO
CD
C3
CO
CO
o
S,

-------
     Ground-level concentrations of other compounds, can be estimated by a
direct proportion of the values in Table 7-2.   Thus, if BaP comprises 0.8
percent of the total organic emissions from a  semisealed furnace, the average
24-hour ground-level concentration would be:  (0.8% x 0.64 yg/m ) = 0.005
    "3          '••','        '                  '
yg/m  at a point about 2 km from the stack.
     None of the ferroalloy plants practices direct control of organic
emissions.  Scrubbers used for particulate control also reduce organic emis-
sions and would be expected to reduce POM compounds by a similar amount.  The
POM compounds thus end up in the wastewater stream, where they could cause a
disposal problem.  Flares used for CO control  would also be expected to
reduce organic and POM emissions; however, data pertaining to the extent of
control are lacking.  The use of open furnaces, in which combustion of CO and
other combustible matter takes place in the furnace, minimizes organic emis-
sions.
7.4.2  Trace Metal Emissions
     Section 5 presented limited data on emissions of trace metals.  The major
metallic emissions are composed of oxides of the  product alloys produced in the
furnace.  The trace element emissions, which are  related to impurities in the
ores, may Vary widely.  Ferroalloy plants with high efficiency particulate
controls as required by most state and local regulations are not generally
major sources of lead.  Lead emissions are  subject to the national ambient air
         ...    -   - 3     '     •  - • •         '  -                   .
standard of 1.5  yg/m .
     Cadmium, arsenic, and copper compounds are also emitted by ferroalloy
furnaces.  Again emission data are sparse  and highly variable.  Table 7-3
presents estimated  emissions of  trace metals that are or may be designated as
hazardous pollutants under Section 112 of  the Clean Air Act.
                                      79

-------
              TABLE 7-3.   ESTIMATED EMISSIONS OF HAZARDOUS  TRACE
                             METALS. AFTER CONTROL

Metal9
Arsenic
Beryllium
Cadmi urn
Copper
Mercury
Leadd
Estimated Emissions
mg/MW-h
62
0.02
17C
. 8.5C
4
2.7
g/24 hoursb
30
0.01
8.2
4.1
1.9
1.3
       Beryllium* mercury,and  arsenic  are_ currently  listed  under Section  112 of the
       Clean Air Act.  The other metals are being considered for possible
       designation as hazardous pollutants.
       For a 20-MW furnace.
       Based on uncontrolled emission measurement and assumed 99% reduction
       in a scrubber.
     A
       Designated as a criteria pollutant, not a hazardous pollutant.

These data show that mercury and beryllium emissions are approximately 1.9
and 0.01 g/day respectively.  These are fairly low levels compared with the
allowable emission of 10 g/day for beryllium and 2300 g/day for mercury for
other processes as designated in Section 112 of the Clean Air Act.   Lead
emissions are also relatively low.

7.5  SOURCE TESTING METHOD EVALUATION
     The Standards of Performance for new ferroalloy facilities specify the
testing methods for measurement of the emissions.  The testing methods
described (specifically Method 5) have been used successfully on wet scrubber
and enclosed fabric filters.
     Several  precautions have been taken by various testing teams to ensure
accurate data collection and safety of sampling personnel.   For example,  for
sampling at the inlet to a control device, the use of a cyclone of  large
volume (200 ml)  is recommended to minimize the need for changing plugged
filters during the test run.   This allows  uninterrupted sampling of integral
                                      80

-------
furnace cycles..  If a filter  change is necessary, it should be made at the
point in the furnace cycle where the run began.
     High carbon monoxide content is inherent in the production of ferro-
alloys in sealed and semi sealed furnaces.  The testing method states that
when the CO content is greater than 10 percent by volume, the heating systems
specified in Method 5 should not be used.  Electrical-resistance heating
systems can create an explosion hazard in high CO content gases.  Although
the absence of probe and filter heating systems may not bias the results of
tests on fabric filter control devices, it will  bias the results of tests at
the outlet of wet scrubbing systems because of condensation.  Use of a steam
         , •.       '.•••.'  ,1.:1   '    .   •.....••.           •          4
heating system is recommended when sampling the exit from wet collectors.
The probe and filter portions of the sampling system should be heated to 120°
+14°C (250d + 25°F), as specified in Method 5.
     As a safety precaution, the vacuum pump should be properly grounded to
                     '           3                   '     '                     •
minimize the chance of sparking.   It is also recommended that a length of
large diameter tubing be attached, to the exit end of the meter box orifice.
This tubing should be vented away from al1 personnel, to minimize the health
                                                          45
hazards associated with high CO content in the gas stream.  '   Before sam-
pling begins, the orifice should be calibrated with the length of tubing
attached, to insure isokinetic sampling conditions.
     In stacks with high CO contents, high negative static  pressures, and
high temperatures [540°C  (1000°F)], an air-tight  seal should be provided
around the, probe.  In one case, an improper seal  reportedly caused air to
leak into the stack and  ignite the gas.-
     In addition to EPA Method 5„ Level  1  Source  Assessment Sampling Systems
(SASS trains) have been  used to characterize emissions from ferroalloy fur-
naces.4'6  Qas chromatography and electron capture have been used  successfully
for the determination and quantification of reduced and oxidized sulfur  com-
pounds4, and gas chromatography by flame ionization has been used for the
         ' -•  •       -••••'    '    '     .     ' •  ..     '       4
determination and quantification of hydrocarbon compounds.
     The CO content  of the  gas stream  associated  with the production of
ferroalloys has been measured by EPA Method 3 procedures, with minor diffi-
culties.  A standard Orsat  apparatus with  the volumetric capacity to measure
CO quantities in the 30  to  80 percent  range should be used  for  testing sealed
furnaces.
                                      81

-------
     Testing open fabric filters presents difficulties because an enclosed
duct is not available and a complete traverse of the open area is impossible.
A Specific test plan must be developed for each site.  Erection of a  tem-
porary stack on one module of the fabric filter system (to facilitate mea-
surement of particulate concentrations) and measuring total gas flow  at  the
fabric filter inlet have been used to obtain emission rates:
                                                            .7,8
the amount
of ambient air induced into a fabric filter system .has beeji. estimated  by
                                    "   7 : •'.."• •':•     '•'-• v    $''• ."  ?';'
making heat balances around the system.   Measurements after flares  have not
been attempted because of inaccessibility and the unconfined nature  of a
flare.
                                       82

-------
                          REFERENCES FOR SECTION 7
1.   Dealy, J. 0., and A. M. Killin.  Engineering and Cost Study of the
     Ferroalloy Industry.  U.S. Environmental Protection Agency, Research
     Triangle Park, North Carolina.  EPA-450/2-74-00&, May 1974.  p. IV-3.

2.   Lomo, A.  Pollution Problems with Electric Furnaces in the Ferroalloy
     Industry.  In:  Proceedings INFACON.  pp. 251-257.  1974.

3.   Telecon.  M. Bockstiegel, PEDCo Environmental, Inc., to J. Gluts, Mogul
     Corporation, May 2, 1980.  Source Testing of Ferroalloy Production
     Facilities.

4.   Telecon.  M. Bockstiegel, PEDCo Environmental, Inc., to D. Harris,
     Monsanto, May 13, 1980.  Source Testing at Ferroalloy Facility in
     Canada.

5.   Telecon.  M. Bockstiegel, PEDCo Environmental, Inc., to R. Schab, Beling
     Engineering Consultants, May 16, 1980.  Source Testing of Ferroalloy
     Production Facilities.

6.   Westbrook, C. W., and D. P. Daugherty.  Draft Environmental Assessment
     of Electric Submerged Arc Furnaces for Production of Ferroalloys.  U.S.
     Environmental Protection Agency, Contract No. 68-02-2630.  March 1980.

7.   Telecon.  M. Bockstiegel, PEDCo Environmental, Inc., to J. Fox, Frederick-
     sen Engineering Co, May 15, 1980.  Source Testing of Open Baghouses.

8.   Telecon.  M. Bockstiegel, PEDCo Environmental, Inc., to H. Rogers,
     Environmental Management, May 15, 1980.  Source Testing of Ferroalloy
     Production Facilities.
                                     83

-------
                                  SECTIC-N 8
                               RECOMMENDATIONS
8.1  CHANGES IN REGULATIONS
     No changes in particirtate emissions regulations are recommended at this
time.  Limited additional test data for determination of compliance with State
emission regulations have indicated particulate emissions also comply with the
NSPS.  Regulations in other countries are generally less restrictive than U.S.
regulations, except those for closed furnaces.
     Visible emissions from control equipment vents ate usually in compliance
with the NSPS requirement of 15 percent or less opacity, except during
malfunctions.  Most plants are also in compliance with existing state regu-
lations.  Visible emissions from tapping are a problem on existing furnaces,
but technology available for hooding and control of new furnaces should
enable them to comply with the NSPS.  No change in the visible emission
levels is recommended.
     Flares are used on streams with high CO content, and there is no need to
change the CO emissions limit.

8.2  AREAS OF FURTHER STUDY
                                                                          I
     Although emissions of many organic compounds have been identified from
open, closed, and semisealed furnaces, data are currently insufficient to
determine the magnitude and impact of these emissions; especially after
control devices.  It is therefore recommended that tests be performed on open
furnaces equipped with fabric filter control systems to determine the quantity
of the organic and especially POM compounds in the! exit gas stream and how
effectively they are controlled by fabric filtration systems.   A determination
of how much a flare reduces organic and POM concentrations will be necessary
for an assessment of emissions from closed and semiclosed furnaces.
                                     84

-------
                             APPENDIX

                  COSTS FOR FABRIC FILTER SYSTEMS
Data required:
     Gas flow and gas temperature to baghouse
     Type of ferroalloy produced
     A/C ratio
     AP = 13 in. of water
     Reverse air and insulated baghouse
Baghouse cost:

          x acfm to baghousex

-------
4.  Installed costs:
     Direct

     Fabric filters, fans, and motor
     Instrumentation and controls
     Electrical (no transformers)
     Piping
     Insulation

     Painting
     Site preparation
     Foundations
     Structural work
     Indirect

     Engineering
     Contractor's  fee
     Field  expenses
     Freight
     Offsites
     Taxes  (if not exempt)
     Allowance for startup,  etc.
     Spares
     Contingency
     Escalation
     Interest during construction  (12%/yr)
Material
  and
equipment^

     x
  O.Olx
  0.05x
  O.Olx
Supplied/
baghouse
  O.OOlx
  O.OOSx
  O.Olx
  O.Olx

  1.096x
  0.1 Ox
  O.lSx
  0.05x
  0.04x
  0.02x
  0.025x
  0.015x
  0.02x
  0.06x
  0.075x
  0.162x

  0.717X
Installation

    0.4x
    O.OOSx
    0.05x
    O.Olx
    0.05x

    0.009x
    O.OOBx
    O.Olx
    O.Olx

    0.584x
         Total installed cost = 2.397x.  [Total installed cost does not
         include ducting to the baghouse or to the stack nor the stack cost.
         (Note:  Most of ducting and stack costs would be part of the furnace
         costs).  Dust removal would be accomplished by plant's vehicle   No
         costs for disposal area.]
                                     86

-------
5.  Annual Operating Costs:
    Utilities                                       ,
    a.    Electricity:
          I.D. fan:   0.00314 x 0.746 x baghouse acfm x 8760  x  0.035
                     =  0.6823 x acfm                              =  $
          Reverse:   0.0614 x 0.00314 x 0.746 x acfm x, 8760 x 0.95
                    x  0.2 x 0.035 =0.0084 x acfm                =  $
                        11 x 0.746 x acfm x 0.25 x 0.035  x 8760
          Screw conv.:
                        6.00143 x acfm
     b.   Operations
          1.  Labor
          2.  Supervision
     c.   Maintenance
             440,000
    1  man/shift x $8.50/h x 8760  = $ 74,500
    15% of labor                  =   11,200
          1 .   Labor
     d.   Overhead
          1.  Plant
          2.  Payroll
     e.   Fixed costs:
          Depreciation
    1 man/shift x 9.00 x 8760     =   78,840
                               P
acfm (baghouse) Y bag cost $/ft   _
   A/C ratio    *      2          ~
   :   0.1 (b.l + b.2 + c.l)       =   16,500
    0.5 (b.l + b.2 + c.l)         =   82,270
    0.2 (b.l + b.2 + c.l)         =   32,908
    100
                       equip. 1ife (15 yr)
          Interim replacement
          Insurance
          Taxes
          Capital cost
=  6.67%
=   .35%
    .30%
=  2.00%
=12.00%
                                               21.32%
                    0.2132 x total installed cost
                                     87

-------
f.   Dust removal
     1 truck per day - 2 h/day - captive truck
          2 (8.50) x 1.9 x 365 x .95

     Total annual operating cost
          is total of all items
= $11,200


  $

-------
TECHNICAL REPORT DATA
(I'lcau- rcaJ Instructions on ill? mem be fun i am/ilt linn i
1 (REPORT NO.
EPA-450/3-80-41
2.
4. TITLE ANDSUBTITLE
Review of New Source Performance Standards for
Ferroalloy Production Facilities
7 AUTHOR(S) ..'.,„•
R. w. Gerstle, W. F. Kemrier, and L. V. Yerino
9. PERFORMING ORGANIZATION NAME At>
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
JD ADDRESS
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, M.C. 27711
3 RECIPIENT'S ACCESSION NO.
5 REPORT DATE
December 1
980
6 PERFORMING ORGANIZATION CODE
e. PERFORMING ORGANIZATION REPORT NO.
P/N 3450-10
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract No. 68-02-3173
Assignment 10
13. TYPE OF REPORT AND PERIOD COVERED
Draft Final
14. SPONSORING AGENCY CODE
EPA/200/04
IE SUPPLEMENTARY NOTES
OAQPS Project Officer: George B. Crane, MD-13, (919)541-5301
16 ABSTRACT
The purpose of this study was to determine if any revisions are needed in
the NSPS for Ferroalloy Production Facilities. Information was obtained from
manufacturers, regulatory agencies, and the literature.
In 1979, ferroalloy production was 1.6 Tg (1,830,000 tons), which represents
a 21.5 percent decline since 1971. The number of electric submerged-arc furnaces
has also decreased from a total of 158 to 89 for ferroalloys and 7 for calcium
carbide. No new furnaces have been built since 1974, and none are subject to
the NSPS. Tests made to determine compliance with state regulations showed
particulate emission rates in the range of 0.07 to 0.2 kg/MW-h (0.10 to 0.44
lb/MW-h). Fabric filter control systems are widely used on open type furnaces,
and high-pressure-drop scrubbers are used on semisealed and closed furnaces.
Flares are used to reduce CO emissions. Limited organic emissions data showed
a range of 0.15 to 0.29 kg/MW-h (0.33 to 0.63 Ib/MW-h) prior to the flare.
Because of lack of growth and an absence of new technology, no changes in
the NSPS are recommended.
17.
a. DESCRIPTORS
Air pollution, regulations
Ferroalloys
18. DISTRIBUTION STATEMENT
Unlimited
KEY WORDS AND DOCUMENT ANALYSIS
b.lDENTIFIERS/OPEN ENDED TERMS
New Source Performance
Standards (NSPS)
Iron and Steel Industry
19. SECURITY CLASS f This Report)
Unclassified
2O. SECURITY CLASS (This page)
Unclassified

c. COSATI Field/Group
13B
11F
21. NO. OF PAGES
96
22. PRICE
EPA Form 2220-1 (t-73)
                                                                89

-------

-------