United States
Environmental Protection
Agency
           Office of Air Quality
           Planning and Standards
           Research Triangle Park, NC 27711
EPA-453/R-94-005
February 1994
Emission Standards Division
Alternative Control
Techniques Document -
PM-10 Emissions from
Selected Processes at
Coke Ovens and Integrated
Iron  and Steel Mills

-------
                                  EPA-453/R-94-005
      ALTERNATIVE CONTROL
     TECHNIQUES  DOCUMENT --

         PM-10  EMISSIONS
    FROM  SELECTED  PROCESSES
         AT  COKE OVENS
AND  INTEGRATED   IRON AND  STEEL
               MILLS
          Emission Standards Division
  U.S. ENVIRONMENTAL PROTECTION AGENCY

         Office of Air and Radiation
   Office of Air Quality Planning and Standards
  Research Triangle Park, North Carolina 27711

             February 1994

-------
             ALTERNATIVE CONTROL TECHNIQUES DOCUMENT

     This report  is issued by  the Emission  Standards Division,
Office of Air Quality  Planning  and Standards,  U.S. Environmental
Protection Agency,  to  provide information  to  state and local air
quality management agencies.  Mention of trade names and commercial
products   is  not   intended   to   constitute  endorsement   or
recommendation for  use.  Copies  of this report are available, as
supplies permit,  from  the  Library Services Office (MD-35),  U.S.
Environmental Protection  Agency,  Research Triangle  Park,  North
Carolina  27711  (919-541-2777)  or, for a  fee,  from the National
Technical Information Service,  5285 Port Royal Road, Springfield,
Virginia  (800-533-NTIS).
                                11

-------
                        TABLE OF CONTENTS

Section                                                      Page

List of Figures	v
List of Tables	vi

1    INTRODUCTION 	  1-1

2    SOURCES AND POLLUTANT EMISSIONS  	  2-1
     2.1  INDUSTRY DESCRIPTION  	  2-1
     2.2  PROCESS DESCRIPTIONS  	  2-2
          2.2.1   Processes Included in This Document ...  2-2
          2.2.2   Process Characteristics and Feedstocks   .  2-2
          2.2.3   Process Summary and Overview of Mill
                     Operations	2-3
          2.2.4   Coking	2-4
          2.2.5   Sintering	2-9
          2.2.6   Iron Production	2-9
          2.2.7   Hot Metal Transfer and Desulfurization   .  2-11
          2.2.8   Steelmaking	2-11
          2.2.9   Ladle Metallurgy  	  2-14
          2.2.10  Casting 	  2-14
          2.2.11  Finishing 	  2-15
     2.3  PM-10 SOURCE DESCRIPTIONS 	  2-15
          2.3.1   Processes Described 	  2-15
          2.3.2   Coke Pushing	2-18
          2.3.3   Coke Quenching	2-18
          2.3.4   Coke Sizing and Screening	2-20
          2.3.5   Iron Production (Casthouse)	2-20
          2.3.6   Hot Metal Transfer	2-21
          2.3.7   Desulfurization	2-22
          2.3.8   Other PM-10 Sources 	  2-23
     2.4  MODEL PLANTS AND EMISSIONS  	  2-28
          2.4.1   Introduction	2-28
          2.4.2   Model Plant Potential PM-10 Emissions .   .  2-28
          2.4.3   Model Plant Baseline Emissions  	  2-30
     2.5  REFERENCES FOR CHAPTER 2	2-32

3    EMISSIONS CONTROL TECHNIQUES 	  3-1
     3.1  INTRODUCTION	3-1
     3.2  SOURCE REDUCTION IN IRON AND STEEL MILLS  ....  3-1
     3.3  CONTROL EQUIPMENT 	  3-4
          3.3.1   Primary Control Equipment 	  3-4
          3.3.2   Secondary Control Equipment 	  3-11
          3.3.3   Control System Performance  	  3-14
          3.3.4   Control Devices	3-14
     3.4  NEW CONSTRUCTION CONTROLS	3-24
     3.5  CONTROL METHODS FOR OPEN FUGITIVE DUST SOURCES   .  3-24
     3.6  REFERENCES FOR CHAPTER 3	3-25
                               ill

-------
                  TABLE OF CONTENTS  (continued)

Section                                                      Page

4    ENVIRONMENTAL IMPACTS  	  4-1
     4.1  INTRODUCTION	4-1
     4.2  PM-10 EMISSIONS IMPACT	4-1
          4.2.1   Alternative Control Techniques for Coke
                     Pushing,  Coke Sizing and Screening,
                     and Casthouse Emissions  	  4-2
     4.3  WATER POLLUTION IMPACT  	  4-5
     4.4  SOLID WASTE IMPACT  	  4-8
     4.5  ENERGY IMPACT	4-9
     4.6  REFERENCES FOR CHAPTER 4	4-13

5    CONTROL COST ANALYSIS  	  5-1
     5.1  INTRODUCTION	5-1
     5.2  DESIGN PARAMETERS AND PURCHASED EQUIPMENT COSTS  .  5-2
          5.2.1   General Equipment Assumptions 	  5-2
          5.2.2   Ductwork Cost Methodology	5-3
          5.2.3   Fan/Motor System Cost Methodology ....  5-5
          5.2.4   Baghouse Cost Methodology	5-6
          5.2.5   Process-specific Equipment Assumptions
                     and Equipment Costs	5-7
     5.3  BASIS FOR CAPITAL COSTS	5-15
     5.4  BASIS FOR ANNUAL COST ESTIMATES	5-17
     5.5  COST EFFECTIVENESS	5-24
     5.6  CONTROL OPTION COSTS FROM  INDUSTRY
             REPRESENTATIVES  	  5-31
     5.7  REFERENCES FOR CHAPTER 5	5-35

APPENDIX A INTEGRATED IRON AND STEEL MILLS AND BLAST
           FURNACE COKE OVENS	A-1

APPENDIX B SAMPLE EMISSION FACTOR CALCULATIONS  	  B-l
                                IV

-------
                         LIST OP FIGURES

Number                                                       Page

2-1  General Flow Diagram for the Iron and Steel Industry .   2-5
2-2  Byproduct Coke Oven Battery with Major Emission Points   2-7

3-1  General Flow Diagram for an Emission Control System  .   3-5
3-2  Casthouse with Baghouse Control  	   3-10
3-3  A Hot Metal Transfer and Skimming Station	3-13
3-4  Fabric Filter	3-17
3-5  Wet Venturi Scrubber	3-21
3-6  Electrostatic Precipitator 	   3-23

-------
                         LIST  OF  TABLES

Number                                                       Page

2-1  Additional References for Iron and Steel Mills  ....  2-3
2-2  Uncontrolled PM-10 Emission Factors  	  2-16
2-3  Baseline PM-10 Emission Factors  	  2-16
2-4  Model Plant Operating Parameters 	  2-29
2-5  Uncontrolled PM-10 Emissions from Model Plants for
        Selected Processes  	  2-29
2-6  Baseline PM-10 Emissions from Model Plants for
        Selected Processes  	  2-31

3-1  Typical Particulate Collection Efficiencies of Control
        Devices Used in Iron and Steel Mills	3-16

4-1  Controlled PM-10 Emission Factors  	  4-3
4-2  PM-10 Control System Efficiencies for Baseline and
        Alternative Control Technique (ACT)  Systems  ....  4-6
4-3  Controlled PM-10 Emissions from Model Plants for
        Selected Processes  	  4-7
4-4  Compounds Commonly Emitted at Iron and Steel Mills .  .  4-8
4-5  Energy Required by Implementation of Process Control
        Options	4-11
4-6  Airflows Required for Process Control Options   ....  4-12

5-1  Equipment Parameter Assumptions for all Fabric Filters  5-4
5-2  Control Equipment Parameters and Purchase Costs for
        Coke Pushing (Al)  - Shed with Baghouse	5-8
5-3  Control Equipment Parameters and Purchase Costs for
        Coke Sizing/Screening - Enclosure with Baghouse .  .  5-9
5-4  Control Equipment Parameters and Purchase Costs for
        Casthouse (Al)  - Evacuation to Baghouse 	  5-10
5-5  Control Equipment Parameters and Purchase Costs for
        Casthouse (A2)  - Local Hooding and Baghouse  ....  5-11
5-6  Control Equipment Parameters and Purchase Costs for
        Hot Metal Transfer - Canopy Hood with Baghouse  .  .  5-12
5-7  Control Equipment Parameters and Purchase Costs for
        Desulfurization -  Ladle Hood and Baghouse 	  5-13
5-8  Conversion Factors 	  5-14
5-9  General Costs and Cost Factors for Fabric Filters  .  .  5-16
5-10 Total Capital Costs for Coke Pushing (Al)  ($1,000) .  .  5-18
5-11 Total Capital Costs for Coke Sizing/Screening  ($1,000)  5-19
5-12 Total Capital Costs for Casthouse (Al)   ($1,000)   . .  .  5-20
5-13 Total Capital Costs for Casthouse (A2)   ($1,000)   ...  5-21
5-14 Total Capital Costs for Hot Metal Transfer ($1,000)   .  5-22
5-15 Total Capital Costs for Desulfurization  ($1,000)  . .  .  5-23
5-16 Total Annual Costs for Coke Pushing  (A2) ($1,000)  .  .  5-24
5-17 Total Annual Costs for Coke Pushing  (Al) ($1,000)  .  .  5-25
5-18 Total Annual Costs for Coke Sizing/Screening ($1,000)   5-26
5-19 Total Annual Costs for Casthouse (Al)  ($1,000)   ....  5-27
5-20 Total Annual Costs for Casthouse (A2)  ($1,000)   ....  5-28
5-21 Total Annual Costs for Hot Metal Transfer  ($1,000) .  .  5-29
                                VI

-------
                    LIST OF TABLES (continued)

Number                                                       Page

5-22 Total Annual Costs for Desulfurization  ($1,000)   .  .  .  5-30
5-23 Cost Effectiveness for Emissions Reductions from
     Uncontrolled Case	5-32
5-24 Capital and Annual Cost Comparisons from Industry
        Representatives 	  5-34

A-l  Integrated Iron and Steel Mills and Blast Furnace
     Coke Ovens	A-2
                               VII

-------
                            CHAPTER 1
                           INTRODUCTION
     The  Clean  Air Act  Amendments of  1990  (November  15,  1990)
authorize the Environmental  Protection  Agency (EPA)  to designate
areas  that  violate  the  national  ambient air  quality standards
(NAAQS) for particulate matter nominally 10 microns or smaller in
diameter  (PM-10) as nonattainment  areas.   [See Section 107(d) of
the Clean Air Act (Act).]   Section  188(a)  of the Act provides  that
every designated nonattainment area for PM-10 shall be classified
as  a  "moderate"  nonattainment  at the  time  of  designation by
operation of law.  A moderate area can subsequently be reclassif ied
as "serious" if  EPA determines that (1)  the area cannot practicably
attain the PM-10 NAAQS by the applicable attainment date or (2) the
attainment  date has already  passed and  the  area has  failed to
attain the standards.
     State implementation plans  (SIPs)  for moderate nonattainment
areas must, among other things, provide for the implementation of
all  reasonably  available control  measures,  including reasonably
available control technology (RACT)  to achieve emission reductions
from  existing  stationary  sources.    [See Sections   172(c)  and
189 (a) (1) (C) .]  In addition to the requirements for moderate areas,
SIPs for serious areas must include, among other things, provisions
to assure that the best available control measures, including  "the
application of best available control technology (BACT)  to existing
stationary sources"  [H.R. Rep. No.  490,  101st Congress Sess. 267
(1990)], are implemented no later than 4 years  after the areas are
reclassified as serious.   [See Section 189(b)  (1) (B).]
     In accordance with Section 190 of the Act,  EPA determined  that
information for use  in determining RACT and BACT was needed for the
ferrous metals industries.  Therefore, EPA prepared this guideline
document on alternative control techniques  (ACT)  to assist States
in  identifying  RACT and BACT alternatives for  selected process
sources of  PM-10  in the iron  and  steel industry.  Although ACT
documents  review existing  information and  data  concerning  the

                               1-1

-------
technology  and cost  of  various  control  techniques  to  reduce
emissions, they are,  of  necessity,  general in nature  and  do not
fully account  for unique variations  within  a stationary  source
category.  Consequently,  the purpose of ACT  documents is to provide
State and  local  air pollution  control  agencies  with  an  initial
information base for proceeding with their own analysis of RACT and
BACT for specific  new and existing stationary  sources.
                               1-2

-------
                            CHAPTER 2
                 SOURCES AND POLLUTANT EMISSIONS

2.1  INDUSTRY DESCRIPTION

     The United States  iron  and steel  industry includes the U.S.
Government Standard  Industrial Classification  (SIC)  codes 3312,
3315, 3316, and 3317.   In  1990,  domestic production of steel was
estimated to be around 78 million tons,  four percent less than the
previous year.  In 1989, the domestic  steel  industry operated at
approximately  88  percent of its  capacity.1   Domestic  demand for
steel is expected to remain constant through the 1990s at a level
of 90 to  100  million tons  per year.   If domestic demand exceeds
domestic  supply,  foreign  imports will  be used  as  supplements.
There  will  probably   be   increased  demands  for  high-quality
corrosion-resistant products  such as coated  steel  and stainless
steel.
     The SIC coding system has several categories that comprise the
domestic  iron  and steel industry,  including the steel  wire and
related products category  (SIC 3315), the cold finishing of steel
shapes category (SIC 3316), and the steel pipe and tubes category
(SIC  3317).2    This  document  is  directed   for coke  producing
facilities and integrated  mills.   An  integrated iron  and steel
plant is one in which coke, iron ore, and other raw materials are
converted into a  finished  or semifinished steel  product.   These
facilities are contained primarily in SIC 3312, which also includes
facilities whose  primary function is hot  rolling  iron  and steel
into basic products  like plates, sheets, bars,  and  tubing.   The
majority of products produced  by the domestic steel industry are
converted to final products at  other facilities within SIC 3312 or
in other industries.3  A list  of integrated iron and steel mills and
furnace coke  ovens  operating  in 1991  is  given in Table  A-l  in
Appendix A.
     The majority of  these iron and steel facilities are located in
states having  heavy  manufacturing.   Fully integrated  mills are
concentrated in areas that  have access to coal and iron ore.

                               2-1

-------
     In 1984,  there were 36  operational coke plants containing 134
batteries (rows of ovens) in the United States.  About 92 percent
of their coke  production was used in manufacturing iron and steel.4
     The largest  customers  of steel products  are  the automobile
industry and the construction  industry.  Manufacturing appliances,
containers,    electrical  equipment,  and   machinery  also   use
significant amounts  of  steel.   The  use of  steel  has  declined
somewhat in  recent years  due to  the  increased substitution  of
materials such  as  aluminum,  plastics,  glass,   and  ceramics.   The
automobile  industry  also  has increasingly  replaced  steel  with
aluminum and plastics.  Despite these smaller markets, demand for
steel will probably remain constant in the near future.5

2.2  PROCESS DESCRIPTIONS

2.2.1  Processes Included In This Document

     This document does not discuss all the possible sources  of
PM-10 emissions from integrated iron and steel plants but focuses
on emissions  from two areas:   the  transfer  of  coke from the coke
oven battery  to its uses at  the  blast  furnace and  EOF,  and the
transfer  of  molten  iron  from the blast  furnace  to the  EOF.
Specific sources  include coke pushing (see Section  2.3.2),  coke
quenching  (see  Section  2.3.3),  coke sizing  and  screening  (see
Section 2.3.4),  casthouse emissions (see Section 2.3.5), hot metal
transfer   (see   Section   2.3.6),   and   desulfurization   (see
Section 2.3.7).  All of these sources are not regulated under the
New Source Performance  Standards  (NSPS).  Other  sources  of PM-10
emissions are mentioned briefly.  Agencies may consult references
such as  those listed in  Table 2-1 for  more information  on NSPS
processes and other processes not included in this document.

2.2.2  Process Characteristics and Feedstocks

     The primary  materials  used in an integrated  iron and steel
mill are coal, iron ore, gaseous oxygen, steel  scrap,  and alkaline

                               2-2

-------
    TABLE 2-1.  ADDITIONAL REFERENCES  FOR IRON AND STEEL MILLS
  U.S.  Environmental Protection Agency.   Electric Arc Furnaces and Argon-
  Oxygen  Decarburization Vessels in Steel Indus try-Background Information
  for Proposed Revisions to Standards. EPA-450/3-82-020a.  Office of Air
  Quality Planning and Standards.   Research Triangle  Park, NC.  July  1983.

  O.S.  Environmental Protection Agency.   Electric Arc Furnaces in  Ferrous
  Foundries-Background    Information    for   Proposed    Standards.
  EPA-450/3-80-020a.    Office of  Air Quality  Planning  and  Standards.
  Research Triangle Park, NC.  May 1980.

  U.S.  Environmental Protection Agency. Control Techniques for Particulate
  Emissions from Stationary Sources-Volume 2.  EPA-450/3-81-005b.  Office of
  Air  Quality  Planning  and  Standards.    Research  Triangle  Park,  NC.
  September 1982.

  U.S.  Environmental Protection Agency. Revised Standards for Basic Oxygen
  Process  Furnaces-Background  Information   for   Proposed   Standards.
  EPA-450/2-82-005a.    Office of  Air Quality  Planning  and  Standards.
  Research Triangle Park, NC.  December 1982.

  U.S.  Environmental  Protection  Agency.    Control of Open Fugitive Dust
  Sources.  EPA-450/3-88-008.  Office of Air Quality Planning and Standards.
  Research Triangle Park, NC.  September  1988.

  Buonicore, A.J., and W.T. Davis,  eds.   Air Pollution Engineering  Manual.
  Van Nostrand Reinhold.  New York,  NY.   1992.

  Lankford, W.T., et al., eds. The Making,  Shaping,  and  Treating of Steel.
  Tenth Edition.   Association of Iron and  Steel Engineers.  Pittsburgh, PA.
  1985.
fluxes such as limestone.  Secondary  feedstocks  include  water  for

coke quenching, desulfurization compounds such as calcium carbonate
(CaCO3) ,  calcium  carbide  (CaC2) ,   or  salt-covered  magnesium,   and

alloying agents such  as nickel,  silicon,  and manganese.


2.2.3   Process Summary and Overview of  Mill Operations


     An integrated iron and steel plant is one  in which  iron ore,

coal,  and  other  raw  materials are  converted  into  a finished  or

semifinished  steel product.   First,  coal is converted to coke  for

use  as  a  fuel  in  the production  of  molten  iron.    Fine  coke

particles  (breeze),  which are not  directly usable, may  then  be

burned,  in the  presence  of   iron ore  fines and flux,   forming  a


                                   2-3

-------
compound known as  sinter.   The ore, coke,  sinter,  and fluxes are
charged into  the blast furnace.  In the blast  furnace,  the iron
oxide in the ore  is reduced to molten iron.  Afterwards, the molten
iron is usually treated with reagents to remove any excess sulfur.
The molten  iron,  along with  varying  amounts of steel  scrap,  is
charged into the steel-producing  furnace.   In the  steel furnace,
oxygen is blown into the molten iron to remove excess carbon and to
help  the fluxes  remove  impurities  in the  steel.    Additional
alloying steps may be conducted at a metallurgical station; then,
the molten steel is poured into ingots (teemed) or solidified into
a  steel  "casting"  in  a continuous caster.   Finally,  the steel
undergoes semifinishing  and  finishing operations  such  as rolling,
chipping, grinding, and scarfing.   Figure  2-1 shows a general flow
diagram for the processes in a typical integrated mill.
     The production processes in an integrated iron and steel mill
are  generally  batch  processes  with  distinct  cycles   in  which
materials  are  combined at  the  process   station,  processed  or
converted,  and emptied from the  station.   The  exception  is the
casting stage,  where most casting currently performed is continuous
casting.
     The conversion of  coal to coke is a  form of  distillation.
Heat applied to the coal in an oxygen-deprived environment drives
off  volatile  compounds  from the  coal,  leaving relatively pure
carbon.   Melting in  the  blast furnace and in  the  basic oxygen
process furnace  (EOF)  involves two-phase reactions.  In the blast
furnace, oxygen  reacts  with coke to form  carbon monoxide.   In  a
reduction reaction, the carbon monoxide converts iron oxide in the
ore to metallic  iron.   In the basic oxygen furnace,  an oxidation
reaction is used to remove  excess carbon and fluxes  are used to
remove  other   impurities  from the molten  iron.    The oxidation
reaction also provides heat to keep the mixture molten.

2.2.4  Coking

     Coke,  a   material  that  is  primarily  elemental  carbon,  is
produced by distillation of  coal.   The coal is heated in an  oxygen-

                               2-4

-------












b
a




cc
UJ
M
CO







OL
1-




UJ
-0
C
UJ
O kJ


a:
UJ
UJ
UJ
Q;
? \
** ^^
CO



ulO
QO
Z



UJ
o
o

QUENCH
TOWER




CO
UJ
z
u.
Ul
0
o
0^
xo
UJtt
ot
05

m

o
2;
to
to
UJ
o
o
EC
UJ
ce









tjO '• O
o02 <,
3 Jo














i-
IRON-BEARING
PLANT BYPRODUC



"• '


a; to "
UJ UJ
z ^
CO
z
	 tj
o ^
E
H

Z
CO






X
0?








UJ
cz
O
o





                                                      4J
                                                      W
                                                      H
                                                      0
                                                      o
                                                      •0
                                                      c
                                                      rt

                                                      C
                                                      O
or
UJ
      *
      0
                                                      (0
                                                      •H
                                                      •0


                                                      I
                                                      (0
                                                      Vl
                                                      
-------
free environment in order to remove its more volatile components.
Coke, along with flux,  iron ore,  and/or scrap steel, is one of the
primary raw materials in the manufacture of steel.  Integrated iron
and steel mills have coke-producing facilities on-site or nearby;
non-integrated mills are supplied with coke by rail or water.6
     In 1991,  the byproduct coke making process was used to produce
nearly all  coke produced in  the  United States.   In this process,
exhaust gases  from  the coking oven are collected and treated to
recover usable  byproducts  such  as  tar,  ammonia,  and  light  oil.
Figure  2-2  outlines  a  typical  byproduct  coke  oven  process
operation.4   Byproduct coke  ovens  are constructed  in  rows,  or
batteries,  that  may contain  between  19  and  102  ovens, with  an
average of about 58 ovens  per  battery.7   Coking chambers  in  a
battery are heated on both  sides.  Typically, each individual oven
is 3.0 to 6.7  meters high  (9.8  to 22 feet), 11 to 16.8 meters long
(36 to 55  feet),  and  0.35  to 0.5 meters  wide  (1.2  to 1.7 feet),
though larger  ovens  may be used.  The top of each oven has charging
holes and at least one additional opening,  equipped with an offtake
pipe, through  which volatilized byproducts are collected while the
coal is heated.  The oven also has  long,  thin doors  at each end.
When the conversion  of  coal to coke is complete, the coke is pushed
out of the coke oven by a powered ram which is inserted through the
"push-side" door.   Coke exits  the other side  of  the  oven through
the  "coke-side" door and  falls into a special rail  car called a
quench car.
     Overhead coal bins at  the battery are used to load a measured
amount of coal, which is fed into a larry car that transports the
coal to each oven.  The larry car,  which moves on rails along the
top of the battery,  loads coal to each oven via the charging holes.
Coal is charged by  gravity,  assisted  by a mechanical system that
employs either  a screw conveyor or a  revolving  table discharge
arrangement.
     The composition of the coal  charged to the ovens is important
for several reasons.  Coke in the blast furnace is used both as a
fuel and as a  structural matrix that allows air to  flow through the
blast furnace.  Because the coke chunks (around one or two inches

                               2-6

-------
                           C3

                           £
                    t  \®®d)
                                     n

                                     S
                                     •H

                                     a


                                     §
                                     •rl
                                     n
                                     0)
                                     •H
                                     M
                                     O
                                     •n


                                     3

                                     .d
                                     ^j
                                     •H
                                     0)

                                     4J
                                     4J

                                     n)
                                     I
                                     O


                                     0
                                     X
                                     o
                                     o
                                     o
                                     3
                                     •o
                                     o
                                     n
                                     cu
                                     >1
                                     n
CM

CM

a
M

g.
•H
2-7

-------
in size)  must support great weights in a stacked blast furnace, the
coke must be  reasonably strong.   Maximum strength of the coke is
achieved by crushing the feed coal, then adjusting the bulk density
of the product by blending it with oil.  Water can also be added to
increase the bulk density of the coal.8  This "wet" charging of coal
typically contains  6  to  11  percent moisture  by weight.   "Dry"
charging is used in some  batteries.   In  this  type  of  system the
coal is  preheated to  remove  moisture and then fed  into the coke
oven by  conveyor or pipeline.6  However, industry representatives
indicate that pipeline charging was a theory that  did not work well
in practice and has generally been abandoned.9
     The coal in wet coal  batteries is heated in  the coke oven for
15 to 18 hours in the absence of oxygen.  Approximately 2/3 Mg of
coke can be  produced from one Mg of coal, excluding the coke fines,
which are referred to as "breeze."4
     After  the coke is  pushed,   it  is usually  cooled  by  a  wet
quenching process in which the quench car is  placed under a quench
tower and the  coke  doused with  water.10  Typically,  approximately
9-18 Mg  (10-20 tons) of hot (800°C/~1,500°F) coke are quenched by
22,700-45,400 liters (6,000-12,000 gallons) of water in a process
lasting  two  to three  minutes for blast furnace  coke.11   A. final
moisture content of 2.5-5.0 percent  results.9'10  After quenching,
the coke is removed from the car,  drained,  further  cooled,  and
sized.  This sizing  and  screening  is  performed  in order  to yield a
controlled size of coke for the blast furnace.11
      The properly  sized coke is  used in the production of iron.
The dust-like particles,  or  breeze, that  result from  the coke
handling process are used in  sintering  or elsewhere in the mill.
The gases formed by baking the coal are collected and sent to the
byproduct recovery  plants.   Water  sprays   cool  the  gases  to a
temperature of 80-lOO°C  (~200°F).  The  gases are then treated to
remove valuable  byproducts for use  or sale.    These  byproducts
include  light  oil,  tar, and ammonia.   The gas that remains after
the byproduct  recovery  processes  contains significant amounts of
methane and hydrogen and has a heating value  of approximately 20.5
MJ/Nm3 (Megajoules per normal  cubic meter).4  This gas is returned
                               2-8

-------
to the coke ovens for use as fuel  or may be used as fuel elsewhere
in the plant.6

2.2.5  Sintering

     Sintering fuses  raw materials such as  iron  ore fines, coke
breeze, and flux  (an  alkaline  material  such as limestone) into a
solid product of  sufficient  size  and  strength to be charged into
the blast  furnace.   The materials  to be  sintered are mixed with
water or are  left dry.   The mixture is  placed on a movable grate
known  as  a sinter  strand and  the coke breeze  is  ignited  by a
natural gas or fuel  oil burner.   The burning coke breeze in the
mixture   generates    enough   heat  to   sustain   combustion  at
1,300-1,480°C  (~2,400-2,700°F)  and achieve  the desired level of
fusion of iron particles.  The  large volume of combustion  gases is
drawn through the burning mixture  into windboxes below the strand,
then into  a common  duct that leads to a gas cleaner.   The fused
mass is cooled with  air or a  water spray,  crushed,  and  screened
prior to being charged into the blast furnace.  Undersized sinter
is recycled through the  sintering process.   Generally,  2.5 Mg of
raw materials,  including water and fuel, will  yield  1.0  Mg of
product sinter.12

2.2.6  Iron Production

2.2.6.1  Blast furnace
     The production of  iron for  steel-making is  achieved by the
reduction of iron ore  to iron in a refractory-lined blast  furnace.
Iron in the ore is typically found as hematite (Fe203) or magnetite
(Fe304)  .  In addition to the iron ore, coke,  sinter,  flux, and other
materials  are charged into the furnace  by  either  skip hoist or
continuous  conveyor.    The materials  added  to  the  furnace are
collectively  known  as the  "burden."   The typical  blast furnace
operates at temperatures greater   than ~1,650°C  (3,000°F) in the
lower part of  the  furnace.8  Heated air (-300-1,100°C, 600-2,000°F)
is injected into  the  furnace,  where it  heats and reacts with the

                               2-9

-------
coke to form carbon monoxide.   In the hot reduction environment,
the carbon monoxide reacts with the  iron oxide to produce molten
metallic iron and carbon dioxide.  The metallic iron flows to the
bottom of the furnace.  Impurities and other materials in the blast
furnace collect in a liquid slag layer on top of the molten iron,
which is now termed "hot metal."
     The hot metal and slag are removed from the blast furnace by
"tapping," which releases the material  through a taphole drilled
through a clay plug located at  the base of the furnace.   The hot
metal and slag  flow from the blast  furnace  into a trough.   The
slag, which floats  on  the hot metal, is  skimmed  off  the flowing
iron and  directed to  slag  runners,  which carry  the  slag  to  a
repository such as a slag pit.   Later, the slag is transported for
further processing.8   The hot metal  flows from the  trough into
runners that guide  the  hot metal to a torpedo  car.  The torpedo car
is essentially a refractory-lined tubular container mounted on a
railcar base.   At  the  conclusion  of  the  tapping process,  the
taphole is replugged with clay.
     Producing one Mg of hot metal typically requires the following
raw materials:  1.4 Mg of  iron  ore, 0.5 to  0.65 Mg of coke, 0.25 Mg
of flux and 1.8 Mg  of  air.   Byproducts include 0.2 to  0.4 Mg of
slag  and 2.5  to  3.5 Mg of  blast  furnace  gas  that  contains
approximately 0.05 Mg  of  dust and a  significant  amount  of carbon
monoxide.  Blast  furnace gas  has a low heating value but may be
used after cleaning for supplemental fuel.12

2.2.6,2  Direct reduction--an alternate method
     There are alternatives to the traditional blast furnace method
of producing iron.  The most widely known is  direct reduction, in
which natural gas is reformed into  hydrogen  and carbon monoxide,
then contacted with iron  ore at  800 to 900°C  (-1,500 to ~1,700°F),
which strips the oxygen from the iron.  The result  is a spongy form
of metallic iron known as directly reduced iron (DRI).  Currently
there is only one DRI plant in the United States because the cost
of natural gas has limited the implementation  of this process.  One
company, Hylsa S.A.  of  Monterrey, Mexico,  is attempting to develop

                              2-10

-------
a  pneumatic  transport   system  which  would  move  DRI   from  the
reduction facility  to the steel-making  furnace  while still hot,
resulting in  a significant  savings  in energy costs  and thereby
making  the  DRI  process  more  affordable.    Midrex,   Inc.,  of
Charlotte, North  Carolina  has developed a coal-based DRI process
known as  Fasmet®.   In this process,  ground iron  ore and coal are
mixed with water  and  bentonite, which  serves to  bind the mixture
together. The mixture  is  then  formed  into pellets,  which  are
predried  and heated  to  produce  the  reduction  gases.   DRI  is
generally used at  smaller capacity "minimills, " which produce steel
with electric arc furnaces.13

2.2.7  Hot Metal Transfer and Desulfurization

     From the blast furnace,  the torpedo car  carries the  hot metal
to the EOF  shop.   At  the EOF shop,  the  torpedo  car is  tilted to
transfer  the  hot metal  into a transfer ladle.    Typically,  the
transfer ladle is then moved to the desulfurization station, where
high-pressure nitrogen or argon is used to inject powdered reagents
through a lance into  the  hot metal.   The reagents react with the
sulfur in the metal and  draw the sulfur-compounds into a  slag that
floats on the  surface of the hot metal.   The slag layer is then
skimmed off  the metal.  Desulfurization reagents used have included
calcium carbide (CaC2)  , calcium carbonate (CaC03) , and salt-covered
magnesium.n
     Desulfurization is  used to improve the properties  of  the steel
in several ways.  Desulfurized steel generally is cleaner and has
fewer surface defects  than steel that does not undergo  the process.
The process  also tends to increase the malleability, strength, and
ductility of the steel.   Finally,  welds on desulfurized  steel are
less porous and are therefore stronger.14

2.2.8  Steelmaking

     After desulfurization,  the  molten iron  is  transported to a
furnace for conversion to steel.   In EOF Steelmaking,  steel scrap

                              2-11

-------
is fed into the furnace along with the molten iron.  Flux is added
to aid in the formation of a slag that removes excess silicon and
manganese  and  reduces  the  levels  of  sulfur  and  phosphorus.
Fluorspar is  also added to  increase  the fluidity of  the  molten
mixture.  Once the furnace is fully charged, oxygen is blown into
the molten metal to remove excess carbon and other impurities.
     In the last few decades,  the  open hearth furnace and EOF have
been the  predominant furnace types used to make steel  from hot
metal.   However, open hearth furnaces have recently been supplanted
by BOFs; by 1977,  open hearth furnaces  had virtually disappeared
from the  domestic steel industry, accounting for less  than five
percent of  production.5   There are  currently no operating open
hearth furnaces in the United States.   A brief  discussion  of the
open hearth furnace process  is  included here  for comparison with
the EOF process.

2.2.8.1  Open hearth furnace
     The open hearth process  employs a  furnace  with  a relatively
shallow refractory-lined basin.   Scrap  and  flux are  fed into the
furnace through doors  in its front,  while molten iron  is  poured
from a ladle through  the door.  The feed  varies but is typically an
equal mixture  of scrap  and fresh molten iron.  Heat for the process
is furnished by gas,  oil,  or tar burners  below and on the sides of
the furnace.   Oxygen is often injected below the surface  of the
melt to speed  the  process and to remove excess carbon, silicon, and
manganese.  The carbon is removed as gaseous  carbon  monoxide and
carbon dioxide, while  the manganese and silicon form oxides that
are removed with the slag.   When the steel has reached the desired
composition and temperature,  it is tapped  from  the bottom of the
furnace.  The  time required to produce a  "heat"  (batch) of steel in
this manner ranges from eight  to twelve hours, with  some older
models requiring up to 20 hours.  Some later models,  which employ
a water-cooled  oxygen  lance to add  oxygen to  the  furnace, have
produced heats in as little as four to five hours.12
                               2-12

-------
2.2.8.2  Basic oxygen process furnace
     Currently, the BOF is the most widely used process to convert
hot metal into steel.  In 1989, this process  produced 60 percent of
the domestic steel industry's output, the remainder being supplied
from steel  scrap  melted in  electric arc  furnaces.5   The furnace
consists of an open-mouthed,  pear-shaped vessel  with an alkaline
refractory lining.  The mouth of a standard vessel usually ranges
from 3.7  to 4.3  meters  (12 to  14  feet)  in diameter,  while the
vessel itself can be  -6.1  to 9.1 meters  (20 to  30 feet)  high.  A
BOF can produce up to  440 Mg  (~485 tons)  of steel  in a single heat.
The feed to a  BOF  is typically 70 percent or  more  hot metal and the
remainder scrap.   After the feed is charged, high-purity oxygen is
blown into  the  furnace in  one  of  two  ways.  The most prevalent
method involves the use of a water-cooled  lance  to inject oxygen
into the top of the  furnace.  In the newer Quelle process, or QBOP
furnace, oxygen and fluxing agents are injected into the bottom of
the furnace through  tubes known as tuyeres.   The Quelle process has
several advantages over the standard top blown technique.  A Quelle
furnace  fits  into many facilities  originally designed  for  open
hearth furnaces,  thus enabling  firms to avoid  excessive capital
costs in  replacing  open hearth  furnaces.   Quelle  furnaces  also
produce slightly higher yields of steel  and allow  a slightly higher
ratio of scrap to molten iron in the initial charge.3
     Oxygen is blown into the furnace in a part of the cycle called
the "oxygen blow" which lasts for approximately  8 to 20 minutes.
The carbon and silicon in the molten steel are oxidized and removed
during  this  period.   No  additional  fuel   is  necessary, as the
reaction of the oxygen with the  carbon  in  the  hot metal provides
all the heat needed to  sustain  oxidization  in  the furnace.   Flux
and fluorspar are added after the beginning of  the oxygen  blow.
Following the completion of the  blowing period, the composition of
the molten steel  is tested in the "turndown" period.   A second,
shorter, oxygen blow is conducted if necessary.   Next,  the molten
steel is poured into a teeming ladle through a taphole in the side
of the furnace.   It is at  this point in the process that alloying
agents are usually added.3   After the steel  has  been removed, the

                              2-13

-------
teeming ladle is removed, and the slag remaining in the vessel is
dumped into a slag car and removed.  A typical run in a EOF lasts
25 to 45 minutes.12

2.2.9  Ladle Metallurgy

     In some  facilities,  the steel  in  the teeming ladle  may be
taken to a ladle metallurgy station for further treatment.   Argon
stirring,   alloying,  and  other  metallurgical processes  may  be
performed on the metal at the ladle metallurgy station.

2.2.10  Casting

     From the BOF  or ladle metallurgy station,  the  molten steel
moves  to  the  casting  stage.   Traditionally,  molten  steel  was
poured, or teemed,  into large cast  iron ingot molds and allowed to
solidify in these molds.  The steel  was then removed, reheated, and
tested to ensure that the temperature was constant throughout the
entire ingot.  The  ingots were then  rolled into billets, blooms, or
slabs  depending on  their ultimate  destination.12   The  current
prevailing method  of casting  is known as continuous  casting.   In
1990,  67.1  percent of  the  steel  produced in the United  States
underwent  continuous  casting; Japan  and  Europe  used  continuous
casting for about 80 or 90 percent  of their steel production.5  In
this process,  the molten steel is poured  into a water-cooled mold.
The steel  emerges from  the bottom of the mold in a continuous slab,
bloom, or  billet that  is  then  cut into  appropriate  lengths  for
rolling.   The process is  more  efficient  than  traditional ingot
casting since  it  casts  steel into  shapes already suitable  for
rolling, thus  eliminating  the need for reworking.3  The trend in the
United States  is to roll billets and blooms from steel produced in
electric arc furnaces while producing slabs with steel from BOFs.
Pressure casting is a  third,  seldom-used method of forming steel
that uses pressure  from the outside air to force molten steel into
a slab-shaped ceramic mold.   The steel is  then solidified, removed
from the mold, and sent off for rolling.15

                              2-14

-------
2.2.11  Finishing

     Steel in bloom,  billet,  or slab form can have surface defects
that could lower the value of  the finished product if allowed to
remain.  To remove these defects, the steel is often subjected to
chipping,  grinding,  or scarfing.    Chipping  and grinding  are
relatively  simple  processes   used  with   minor  imperfections.
Scarfing is used  for more serious  defects.   In  scarfing,  high-
velocity streams of oxygen, along with acetylene or natural gas as
fuel,  are  applied to  the  steel  surface  until  it  reaches  a
temperature of  approximately  870°C  (1,600°F).   This  causes  the
steel at the surface to melt slightly.  Roughly three millimeters
of steel are then removed from each  side of the piece.6  Scarfing
is used primarily to remove defects caused by rolling ingots into
semifinished  forms,   i.e.,  slabs,  billets and  blooms,  and  is
therefore not typically used  when continuous casting  is employed.8

2.3  PM-10 SOURCE DESCRIPTIONS

2.3.1  Processes Described

     This section discusses emission factors for coke pushing, coke
quenching,  coke sizing/screening,  casthouse process fugitives,  hot
metal transfer,  and desulfurization.   Emissions from air pollution
control devices that  collect pollutants from  coke pushing,  coke
quenching,   coke   sizing/screening,   hot   metal   transfer,   and
desulfurization are considered  "process" emissions that are emitted
by a specific process and controlled by capture  devices assigned to
each process.   Uncaptured emissions  from the  above processes  are
considered  "process  fugitive"   emissions,  i.e.,  emissions   that
escape individual  capture devices or are emitted from uncontrolled
processes.   Emissions from other processes  are discussed briefly.
References  listed  in  Table  2-1  may   be  consulted  for  more
information on processes not included in this document.
     Tables 2-2 and 2-3 present uncontrolled  and  baseline PM-10
emission factors  for the  processes  examined  in  this  document.

                               2-15

-------















i
c^
Jj
EKI
55
S
H
CQ
CQ
H

ft
P4

O
_J
ROLLED PM-]
EH
I
z
S3

CM
1
CM
TABLE













I
4J
2
rl
o
JJ
0
fa

*4
o

a
a
1











• .
iasion Factoi
S
u













ource
w
C
o
•H
n
n
1






"O ^
CQ P * U ,* U
c c










^. "c ^
coc
0 4J 0
4-> -v 4J
*G "G *c A r4 a
O O O rH rH
4J 4J 4J W
>•« ^ •>. rH 0 CM
& Q XI CO • CM
rH rH rH • O
O *- 0
O tfl CO ~- — •
m H o "O
• • *0 d) T3
O I-l O 4) rl 0>
^ *- «- M 3 rl
3 O 3
•Q -O 73 O 04 O
d> d) d) 04 PI
O* (7* 0^ rH
rl M H rH (0 rH
<0 io (0 ia 4-> fl
X! £ A 4J d) 4J
o o o d> € a)
rH rH rH jJ
<0 <0 10 4-> O 4->
O O O O £ O
O O O A A
. f ^ , ij5 fc^
S? 2? Sf 2^ ^* Si
>T\ ^*
o^ o* o^ DI r^* CT*
tf) O ^* U") ^* O
(N) VO O rH O iH
o o o o o o


rt
u M
O1 O
C 4_*
•H -rl 0
C C M
d> o t>
« M w "c
en o M-I c o
%e a o  MH
X! g rH
d> d> d) 4J 3
y .ii ^ CO 4-J CO
O O O 10 O d)
o u u o a P















S
o
EH
U
En
§
H
CQ
CQ
H




0
H
i
04
H

§


*
?
3 CM

5 m
O ^«
c EH
c
(0
rl
d)
4J
(0 d)
S H
£>
O, <0
3 rH
0) -rl
CM g r~ (0
>1 4J
d) 4-> d) 0
0 rl 0 C
C -H C
d> -o d) n
rl M
d) A d) >
MH 4J MH (0
d) -rl d>"^
& s o; c
IB Q U _Q
M
O
ts
id
fe
C
o
•H
Q
a
•d
J
w

o
H
1,
g
•o
a
H
H
o
rl
•W
5








System

S
*»
g
o
CJ










'rocess
PI



•o
•O d)
73 4) 0<
d) Cn M
Ol M (0 rH rH H
rl (0 jfl (0 (0 (0
(0 .C O 4-> 4-) 4->
X! O d) d) d)
H 10
rH (0 O 4-1 4-) 4J
(0 O O O O O
00 X! X! X!
O -~
^ c — — —
— C 0 C C C
C O 4-i O O O
O 4-1 ^ 4-> 4-1 4J
4J ^ S3 ^ ^ ^
-^ X) rH £) XJ XI
XI rH H H rH
rH T~<
IT) O f-l CSJ CO
rH O • 00 O O
O O ^ O O O
Dt
O^ ty* £ O^ Cn en
:s ^ ^ s a ^
^> ^0. M "p. "o. "w
VO
If) CO O in rH rH
o o o 1-1 o o
o o o o o o






d)
4-1
rH
to "v5
rl d)
d) H )H O
XI MH d) -H
X! MH 4J M
? S d -0
rl XI S! <0
O MH
10 TJ O
C rl -H TJ
•H (0 d) M C
W 4-1 X) 10
3 U rH (0
4J d) « MH Tf
C 4-1 O
d> « 0 T3 O
> * -H C X!
rl 10
jj CU XI *O O^
d) 3 (0 d) T> C
S d> MH rH O -H
X rH O 4J
TJ « T3 O A 4J
(0 (0 4J >i T
C C O. d)
•d  d) c
d> d> a) "g
q-i MH MH S?
m m m ;>
W M W-H
• CM* B S 6 J
<* rH O O O "
M rl M
0) d) MH MH M-I '•H
o o j
c c -a -a -a <
d) d) d> d) 0)
rl M > > >
0) d) -H -H -H "
4_l 14-1 M M M ^
d) d) d) d> d) -*-1
oc a: Q Q p °
« J3 0 T3 01 Z
2-16

-------
Baseline  emissions  are  those  resulting  from  a  prescribed or
reference method  of air  pollution control,  not  necessarily the
predominant method.  The emission  factor data quality ratings are
given if  available.   The ratings  were  taken  from the U.S. EPA's
Compilation of Air Pollutant Emission Factors  (AP-42), which rates
the  quality  of emission  factors  from A to E,  with  A  being the
highest quality.   AP-42  indicates that high  ("A")  ratings were
given to  emission  factors  based on multiple observations at many
different plants,  while low  ("D"  or "E")   ratings  were given to
emission  factors  based  on  single observations  of questionable
quality or extrapolated  from other emission factors from similar
processes.   The  ratings given in  AP-42 are considered  a general
indicator of the accuracy and precision of a given  factor used to
estimate  emissions from  a  large number  of  sources.   If different
ratings are given for the total particulate emission  factor and the
size distribution (used to estimate a PM-10  emission factor from a
total particulate  emission factor), the lower  quality  rating is
shown.
     For  several  processes  in  Table 2-3,   emission factors vary
slightly for different  plant  sizes.  Emission factors represent the
composite of  fugitive  and process  emissions.   In general,  well-
operated fabric filters can remove more  than 99  percent  of the PM-
10 contained  in  the emission  stream.   However,  it is  generally
agreed that,  regardless of  the  amount  of  PM-10  in the emission
stream,  the airstream exiting the  fabric filter will contain some
minimum concentration of dust, on  the order of 0.003 gr/dscf  (for
typical  inlet  streams  with  concentrations  higher than  0.003
gr/dscf) .9'16
     To  calculate  emissions  from  a  specific  process,  it  is
appropriate  to  calculate  the  fugitive  emissions and  process
emissions separately.   The  fugitive  emissions  can  be calculated
using an uncontrolled emission factor and the capture efficiency of
the ventilation system.  The process emissions  (emissions that pass
through the control device)  may be calculated  one  of two ways:  by
either using the capture  and control efficiencies in conjunction
with  the  uncontrolled  emission  factor;  or  by  multiplying  the

                               2-17

-------
control device's minimum exit grain  loading  by the volume of air
moving  through  the control  device.    The final  emission figure
should be the sum of (a) the  fugitive emission calculation and (b)
the larger of the process emission estimates.  The emission factors
presented in Table  2-3 reflect the typical airflows associated with
different types of iron and steel mills.

2.3.2  Coke Pushing

     Coke pushing produces varied amounts of emissions depending on
the execution of the coking process.   Most coke pushing emissions
consist of coke dust.  Coke  that  is  "green," or not fully coked,
emits far more particulates to the atmosphere  than does fully coked
coal.   The  remaining  volatile  components of  the  coal  burn  or
vaporize as the coke is pushed and exposed to the atmosphere.  The
emissions from the  pushing of green coke consist of coal dust, coke
dust, and condensed tars.  Optimization of the coking period can
significantly reduce these emissions.6
     Coke  pushing  emissions,   if  uncontrolled,   have  a  total
particulate  emission factor  of  0.58  kg/Mg (1.15  Ib/ton)  of coal
charged.   Of this  amount,  0.25  kg/Mg  (0.50 Ib/ton)  is  PM-10.12
Typical  control systems  for  coke  pushing  include  fixed  duct
systems,  sheds  or  moveable  hoods vented  to  venturi scrubbers or
baghouses, and  mobile  scrubber cars.   The  AP-42  PM-10 emission
factor for hoods venting to  a wet venturi scrubber is 0.08 kg/Mg
(0.16 Ib/ton).12    An  alternate  emission  factor  provided by  an
industry source is approximately  0.054  kg/Mg (0.11  Ib/ton).9  The
alternate  emission  factor  was  used  for calculations  in  this
document.

2.3.3  Coke Quenching

     During wet quenching,  water  is sprayed  onto the newly pushed
coke.  The thermal  shock that occurs when the  quench  water contacts
the  hot  coke fractures and  shatters the  coke, resulting in the
emission of  coke dust.  Testing emissions from coke quenching is

                              2-18

-------
difficult because of the  formation of a steam cloud when the quench
water contacts the hot coke.  One test showed particulate emissions
from quenching ranging from  0.29  to 1.22 kg of total particulate
per Mg of coal charged.  The amount of particulate  emissions varied
noticeably with the  cleanliness of  the  quench water.  When clean
quench water was used instead of recycled water,  the average amount
of particulate emitted dropped from 1.1  to 0.68  kg/Mg (2.2 to 1.36
Ib/ton).n   Another   reference  noted that emissions  for recycled
quench water were 1.5 to  3  times  greater  than  those  for clean
quench  water.6    The  percentage  of  total  emissions  that  are
classified  as  PM-10 emission  also varies,  depending  on whether
clean  or  recycled quench water is used  in  this operation.   When
clean water is used  for the quench operation, the mass percentage
of PM-10  emissions as a fraction of the total particulate emissions
is 30.1 percent.   For recycled quench water,  the mass  percentage of
PM-10 emissions is 22.8 percent.12  In general, coke manufacturers
use recycled quench  water, adding  clean water only as makeup for
the water lost as  steam or held in the coke.  Quenching is normally
conducted with "relatively clean water"  (less  than 1500  mg/L total
dissolved solids); however,  at  some facilities, water  containing
over 5000 mg/L total dissolved solids may  be used  (often called
"dirty water quenching")  ,16
     Aside from the  use of less dirty water,  the primary  control of
quench-tower emissions is  accomplished with  baffles as inertial
control  devices.   The baffles typically  cover  the  entire  area
inside of the quench  tower.   The  effectiveness of  baffles  in
controlling emissions varies  from 50 to 95 percent  depending on the
type.6
     The  AP-42 emission  factors  reflect   the  effects of  water
cleanliness and baffle use on the amount of emissions that result
from quenching.   Uncontrolled quenching  with dirty water  has  a
PM-10 emission factor of  0.60 kg/Mg (1.19 Ib/ton)  of  coal charged.
In one series of tests, the use of  baffles with recycled  water with
"typical" particulate loading reduced the PM-10 emission factor for
coke quenching to  0.21 kg/Mg  (0.42 Ib/ton) of coal  charged.  Use of
clean makeup water reduces  the emission  factor to  0.17 kg/Mg (0.34

                               2-19

-------
Ib/ton) of  coal  charged.   For comparison, the use  of both clean
makeup water and baffles reduces  PM-10 emissions by 95 percent, to
0.03  kg/Mg   (0.05  Ib/ton)  of  coal  charged;   however,  it  is  not
expected that this technique would be used in ordinary practice.12
     Other options, such as the Kress Indirect Dry Cooling  (KIDC)
system and  quenching  with inert  gas  have been tried in limited
efforts within the U.S.  or outside the U.S.16  These methods are
described in Chapter 3.

2.3.4  Coke Sizing and Screening

     Reference 17 gives an uncontrolled  PM-10 emission factor of
0.04 kg/Mg coal charged  (0.08 Ib/ton)  for coke handling operations.
This  emission  factor  is  assumed to be  an appropriate  emission
factor for coke sizing  and for coke  screening operations.   As in
the case for coke pushing emissions,  the emission factor given in
Table  2-3  for  coke  sizing  and   screening  is a  composite  of  a
fugitive emission calculation and a process emission calculation.
The fugitive emission portion of this  emission  factor  is calculated
using an assumed capture efficiency of 90 percent and  the reference
17 uncontrolled emission factor.   The process  emission portion of
this  emission  factor  is   calculated   using  typical  airflows
(10,000-40,000 cfm)  and a  minimum exit  grain loading  of 0.003
gr/dscf.9-17

2.3.5  Iron Production  (Casthouse)

     Emissions from the blast furnace casthouse  occur primarily
during tapping.   When  the  molten iron and  slag are  exposed to
ambient air, particulates, primarily iron oxides,  are formed that
may escape  to  the  ambient air through the sides  and roof  of the
casthouse.  Opening the  taphole,  particularly  if it is clogged, is
also a significant  source  of  emissions.12  Approximately 75 percent
of the casthouse emissions from  iron  production  consists of iron
oxides,  the  remainder  being  manganese,  silicon,  and  sulfur
compounds.18  Casthouse  emissions  are  sometimes uncontrolled.   In

                               2-20

-------
other cases, hoods above the taphole vented to a baghouse are used
to control  emissions.   Many blast  furnace operations reduce the
initial formation  of particulates  by preventing the molten metal
from contacting the outside air.  Some operations use metal covers
over  the  trough  and   runners  to  prevent  the hot  metal  from
contacting the atmosphere.  Flame suppression (removing oxygen from
the trough  and runner  areas  via  a  natural gas  flame)  can also be
used to reduce  iron oxidation.   Some new blast furnaces feature
runners  covered by hoods  vented   to baghouses  to  control  the
emissions  at  their source.   Uncontrolled casthouse  emissions
measured at the roof monitor have an AP-42 PM-10  emission factor of
0.15 kg/Mg  (0.31 lb/ton).12

2.3.6  Hot Metal Transfer

     The hot  metal transfer process occurs when molten  iron is
transferred from a torpedo  car to the charging ladle or a hot metal
mixer for the EOF.   Emissions from  this transfer operation include
small iron  oxide  particles as  well  as larger  flakes  of graphite
("kish").   Approximately 42 percent by weight of  hot metal transfer
emissions are  kish, with the remaining  58 percent  consisting of
particles of iron oxides.  Since  the kish generally has a diameter
of greater than 75  microns, it is  not considered  in this evaluation
of PM-10 emissions.  The iron oxides, however,  generally are less
than three microns in diameter.   A  baghouse is usually employed to
help control  these emissions.   The baghouse  is  fitted with  a
sparkbox  to protect the bags  from  accidental  ignition by  hot
particles emitted from the ladle.  Capture device options include
close-fitting  ladle hoods,  canopy  hoods  and  partial  building
evacuation.14
     Reference 17  gives a  PM-10  emission  factor  of  0.045 kg/Mg
(0.09 lb/ton)  hot metal for uncontrolled  emissions at the transfer
station.    As  in  the  case  for  coke  sizing  and  screening,  the
emission factor  given  in Table 2-3  for  hot metal  transfer  is  a
composite of a fugitive emission calculation and  a process emission
calculation. The fugitive emission portion of this emission factor

                              2-21

-------
is calculated using an assumed capture efficiency of 90 percent and
uncontrolled  emission  factor  from  reference  17.   The  process
emission  portion  of  this  emission  factor  is  calculated  using
typical airflows  (150,000-300,000  cfm)  and a  minimum  exit  grain
loading of 0.003 gr/dscf.9'17

2.3.7  Desulfurization

     Desulfurization also results in particulate emissions.   This
process is most often performed  in  the charging ladle and uses the
previously mentioned baghouse control system. The injection of the
desulfurization  reagents  causes turbulence  in the molten  iron,
which results in emissions.  These emissions typically consist of
iron  oxides,  sulfur oxides,  and  oxides  of the reagents.    The
removal of  the  slag  formed by  desulfurization also  results  in
particulate emissions.   A hood  such as the one  located over the
charging ladle is often used to control emissions  from the skimming
of slag.   Several  firms  use  various types  of hoods  to  control
emissions  from  desulfurization.    Few   quantitative  data  of
uncontrolled emissions  are available.14
     A study of emissions from the desulfurization of molten iron
was performed at the Kaiser Steel facility in Fontana,  California.
At this plant, CaC2 and CaC03 were injected into the molten iron by
means of a nitrogen lance.  Fumes from the process were collected
by side draft hoods and  then vented to a baghouse.   Tests were
conducted on the exhaust  prior to the collection  system and on the
dust  collected  in  the  baghouse  in  order  to  determine  the
composition of the exhausts.  The  testing  revealed a particulate
emissions rate of  0.637 kg/Mg (1.27 Ib/ton)  of metal treated.  Of
this material, 0.319 kg/Mg (0.64 Ib/ton), or 50 percent by weight,
was PM-10. The dust was mainly iron oxides, with concentrations of
arsenic, strontium, and chromium possibly high enough to preclude
landfilling without  treatment  for heavy metals.19  This  dust  is
often recycled through a sinter plant.9
     An uncontrolled emission factor for desulfurization is given
in AP-42.   For this study, the baseline control system is assumed

                              2-22

-------
to be a close-fitting hood vented to a fabric filter.  Assuming a
close-capture hood efficiency of 95 percent, a minimum exit grain
loading   as  described   earlier,   and   typical   airflows   for
desulfurization (50,000—100,000 cfm) generates the emission factor
given in Table 2-3.

2.3.8  Other PM-10 Sources

     Several  potential  PM-10  sources are  not  covered  in  this
document, including captured emissions such as those from the blast
furnace and steel  furnaces  as  well  as process  fugitive sources.
The  references  listed in Table  2-1 provide more  information on
those emissions sources.  Other activities that may produce PM-10
emissions are discussed briefly in the following sections.

2.3.8.1  Coal storage and handling
     Emissions resulting from the conversion of coal to coke can be
classified into two  types:  fugitive emissions  resulting from the
handling of materials, and emissions caused by the process itself.
Emissions can result from the transportation, unloading, handling,
piling,  crushing,  and  screening of  coal  as  well  as from  the
handling, screening,  and crushing of coke. Methods used to control
or  reduce  these  emissions  include  collection   (coke  handling,
crushing, screening)  and wet suppression  (storage piles).  For wet
suppression,  water  or  chemicals  are  sprayed  in  a mist  over
potential  emission sources  to prevent  particles  from becoming
airborne.  Generally, emissions from  coal  storage  piles might be
reduced as much as 90 percent with such a system.9'20

2.3.8.2  Charging and oven emissions
     Charging the  coal into the coking oven causes the formation of
smoke and combustion gases.   These products will leave the oven at
any place that is open to the atmosphere,  including the charge door
and the combustion stack from the battery.  An estimated 40 to 95
percent of this  particulate  matter has a diameter smaller than one
                               2-23

-------
micrometer.  On  the  average,  90 percent of the  particles have a
diameter of less than three micrometers.6
      The charging of wet coal from a larry car,  if uncontrolled,
emits about 0.24 kg  of  particulate per Mg  (0.48  Ib/ton)  of coal
charged.12  Most  modern  larry  cars also can include  the  use; of a
jumper pipe which provides suction to control  charging emissions.
Modern larry cars also employ  staged charging and steam aspiration
that keep  charging emissions  to a minimum.  Stage  charging is a
process which uses steam or liquor aspiration at the offtake piping
to create a vacuum  in the  oven being charged.  Material that would
otherwise be emitted  is evacuated to the collector main and cleaned
in the by-product plant.   Scrubber equipped larry cars,  used for
some time  but  no longer in wide use,  proved less  effective than
stage charging.9  Door leakage  can also  cause noticeable emissions.
Most doors leak during the initial portion of the  coking cycle when
oven pressures and  gap  sizes are  greatest; the doors  seal more
effectively later in the cycle.4

2.3.8.3  Sintering
     Sintering also  contributes  to  PM-10  emissions.   The windbox
exhaust is the  major  source of sintering emissions.  The emissions
from a typical windbox average  5.6  kg/Mg  (11.1  Ib/ton)  of sinter
produced.  The  PM-10  emissions factor for sinter windbox emissions
is 0.83  kg/Mg  (1.67  Ib/ton)  of finished sinter.   Most  sintering
processes have a mechanical collector such as  a cyclone to remove
the larger particles  in the exhaust.   The  exhaust  then  goes to a
secondary   collector  such   as  a   scrubber,   fabric   filter,
electrostatic  precipitator,  or  gravel bed filter.6   Most  U.S.
operations currently use scrubbers,  while some use baghouses.  ESP
use has been declining since changes were made in flux addition to
the sinter feed.

2.3.8.4  Flux handling
     Handling and transporting flux also releases particulates into
the atmosphere.   The composition  of  these emissions  naturally
depends on the  composition of the flux,  which is usually limestone,

                              2-24

-------
dolomite, or  another alkaline  material.    The flux  is normally
received in bulk  lots  by truck or rail and is moved by conveyor
belt to storage areas until it is needed.  Hoods cover areas having
a serious risk  of material loss.   These hoods normally vent to a
baghouse.3  No  emission  factors  are available for  flux handling.

2.3.8.5  Steelmaking
     EOF  operations produce  two  types  of  emissions:  "primary
emissions"  generated  during  the  oxygen  blow  and   "secondary
emissions" associated with charging and tapping. Primary emissions
generally  consist  of  iron  oxides,  slag  particles,  carbon,  and
carbon monoxide. These  fumes are mostly captured by a primary hood
located over the furnace mouth.   The  fumes must undergo  a cleaning
process to remove excess particulate matter.  The most common types
of cleaning processes employed are the open hood with electrostatic
precipitator (ESP),  open  hood with wet  scrubber,  and closed hood
with scrubber.3  If  an ESP is used, the carbon monoxide must first
be combusted by addition of excess  air at  the furnace mouth to
avoid the risk of explosion.
       For more information on  BOFs,  their emissions,  methods of
control, and  relevant   regulations,  agencies should  consult  the
sources listed in Table 2-1.
     Emissions from  the  secondary sources associated with a EOF are
usually not captured by the primary hood.  Most of these secondary
sources may also be  present in a shop employing an open hearth or
electric arc furnace.

2.3.8.6'  Charging of molten iron and of steel scrap
     The charging of the molten iron and the steel scrap into the
steel producing furnace may cause fairly significant emissions of
particulates.    Emissions are greater if  the scrap is  dirty or
contains excess moisture.  In some shops with an open hood, these
emissions are controlled to an extent by slowing the iron feed rate
and reducing the furnace  tilt in  order  to capture as much of the
emission stream as possible with the  hood.   Bottom blown or Quelle
furnaces produce greater quantities of emissions than standard top

                              2-25

-------
blown furnaces due to the continuous gas flow through the tuyeres
in a bottom blown furnace.3

2.3.8.7  Tapping, slag pouring, and turndown
     Iron oxides are emitted when  the hot steel is tapped from the
furnace  into  ladles.   When alloying materials  such as  nickel,
silicon, or manganese are added, emissions increase noticeably.  In
some BOFs,  the ladle is  enclosed and the fumes are captured by the
hood system on the furnace:  in some shops,  the hood system is also
used to attempt to control emissions from the tilting or "turndown"
of the furnace. Turndown may be performed to check the composition
of the  molten steel or to pour slag  from the top of  the  molten
steel.  A few furnaces have sliding doors in front  of the furnace
that can be  closed  to  improve the collection efficiency  of the
primary hood.3 Flame suppression can be used to remove oxygen from
the tapping area, thereby reducing oxide formation  and emission.

2.3.8.8  Scarfing
     Very few data are available on the quantity and composition of
emissions that result from scarfing.  The scarfing  process may be
carried out either by hand with torches  or by jets  of oxygen from
a machine similar in design  to a  rolling mill.   Machine scarfing
emissions  are  normally controlled by   either  an  electrostatic
precipitator  or a  scrubber.   Emissions  from  hand  scarfing are
almost  always uncontrolled.    It  is estimated that  uncontrolled
machine scarfing produces 0.05 kg/Mg (0.1  Ib/ton) total particulate
emissions.12

2.3.8.9  Slag handling
     The molten slag from steel making is dumped either in the shop
or at a dump  site elsewhere  in  the mill area, allowed to solidify,
and then removed for disposal.  The pouring and cooling  of slag
emits particles to the atmosphere.  In addition,  the use of heavy
equipment to  remove the slag produces large quantities of dust.
Usually, neither of  these emission sources is controlled.
     Slag  recycling and  materials  recovery also  may  generate

                               2-26

-------
fugitive emissions.  Magnets remove the metallic  components of the
slag for reuse.   Portions of the slag that are especially basic may
be reused as flux.   The remaining,  unusable portion of the slag is
landfilled.   This  operation  generates  emissions  from materials
transportation and storage.   Measures must  be  taken  to prevent
components of the slag from leaching into nearby  groundwater.3  No
firm emission factors are available for slag handling and disposal.

2.3.8.10  Deskulling
     Both  transfer ladles and  torpedo  cars  build up  crusts  of
solidified  metal around their  edges  after  repeated use.   This
buildup  is  known  as  "skull"  and  may eventually prevent  proper
operation of the ladles and cars.  Oxygen lances are used to burn
off skulls  on transfer ladles;  iron oxides  are emitted from this
process.  Torpedo cars are deskulled with jackhammers rather than
with oxygen lances; this process is also likely to emit dust to the
atmosphere.3  There  are no established  emission  factors  for the
deskulling processes.

2.3.8.11  Waste disposal
     Two sources of dust result from the disposal of solid wastes
from the  steelmaking process.   First, the dust  collected  by the
various control devices on or around the EOF must be removed.  In
some cases the dust may be recycled to the blast furnace for use in
iron production.   However,  the dust often has high concentrations
of tin and zinc oxides, which would adversely  affect the quality of
metal if fed into  the  furnace.   In these cases,  the  dust must be
dumped  onto  an  open  pile  or   into a landfill.    While  the
transportation and subsequent storage of this dust allows some of
it to escape into the atmosphere, a more significant problem is the
possibility of metals leaching into groundwater from the dust.3  No
reliable emissions data have been  compiled for disposal  of waste
and slag.
                               2-27

-------
2.4  MODEL PLANTS AND EMISSIONS

2.4.1  Introduction

     This  section  describes  operating   parameters   and  PM-10
emissions from  the  processes listed  in Section 2.2.1  for model
plants representing typical  integrated  iron  and steel mills that
use  BOFs.   The  model plant  operating  parameters  and  emissions
estimates are based  on data collected from published literature or
on engineering estimates.
     The small  plant  includes two 136  Mg  (150 ton)  BOFs  in its
steel making shop; this model is representative of some older mills
and of mills that produce some specialty steels.  The large plant
includes three 272 Mg  (300 ton) furnaces, typical of  a modern high-
volume steel production  facility.  The  medium plant includes two
272  Mg  (300 ton) furnaces,  midway between  the small  and large
facilities in  size.   The small  and  large model plants'  furnace
complement  and  production   data  were  taken  or  developed  from
Reference 3.   Metal  production  figures in  Reference 3  are not
necessarily proportional  to  the number of furnaces  and  furnace
capacity.   According  to the reference,  the  production  numbers
reflect the variations  in the operating schedules  possible with
multiple  furnaces  (e.g., a  two furnace  facility  may  alternate
operation of each furnace, or a three-furnace shop may operate one
or two  furnaces while  the  third  is  idle).    Table 2-4 contains
operating parameters  assumed for  these model plants.   Emission
factors  also  were  taken from  the  literature and/or  based  on
engineering estimates as described earlier in this chapter.12

2.4.2  Model Plant Potential PM-10 Emissions

     Table  2-5  lists PM-10  emission  estimates for uncontrolled
model plants.   These calculated emissions are often  referred to as
"potential emissions."   The  potential  PM-10  emissions  for each
process  and model  plant are  calculated using  the  model plant
parameters given in Table 2-4 and  the emissions factors given in

                              2-28

-------
















W
EH
1
2
SI

g
H
EH
S
w
01
o

t^
pR
ix

^^
h
U
Q
2
*

•
I
CM

M
3
9
^























i
&

0
8*
3







I
B
•H
*o
J













Plant
.
H
§
W


























-% co
o
o
co
CM
t-
CM








~ CM
O
0
CO
CN
r-
CM







~ CN
CO
c
o
4J

O
in
rH


§
vo
CO
rH

















CO
0)
>> 0
4-1 re
•H C
o n
re 3
CX 4-1
re
U 

4J re re
34: 4J
3 re 4J
o o o
MOB
X!
^
VO

o
o
o
rH








[->
VO

O
O
O
rH







r-
vo
•
o
o
o
I
rH





























d)
O
U
H
re
o
o
m m
00 VO
• CO
o
0
I-
o








in in
oo vo
• (O
o
o
o







in in
co vo
• CO
o
o
t-
•
0














•o
0) M
u re
3 0)
n >,
0

(X 0)
Q.
H
fl) Oi

re
H M
re (u
4J D<
fl) O
n
o re
X Q










^
CM
C
O
•H

O
C/l
41
U
O
cM
CX

O
91
O
C
3
•o
C
cfl
n
o
CO
^
3


H-l
o
3
C
01
4J
O

cfl
C
O
•H
^
0
a
o
a
!— (
•H
irt
til
— I
01 01 01 co
IM VM M-i 4J
41 41 4) 4)
Di pi Oi E
« £ o -o
     CO

     §
     H
     CO
     CO
     H
     S
     U
     §
     EH
     a
     o
     U
     m

     CM





i
H
fit
0
r?
3











^J
g
H
Oi
§
•rl
•o
£









4J
H
CU

^j
,
^J
9
w
W





























^7
10
SL
^*
s
o
~

^^
M
K.
^^
s*
^




id
ft
^^
M
§

"•'

i?

ft

S



_
M
ft
a
g
o
4J
^^




*^
M
*j

8[L
"s,^
!











a
a
0
o
o
u











o
m
o
w



0
in
en







o
n
^^





o
rH
m







o
CO
CM









o
vo
CM
















01
c
•rl
m
3
a

0)
^
o
U


o
0
m
CM



o
0
CO

CM





o
o
^
i-H
^•*



0
0
CM
rH






o
00
vo









o
rH
VO














01
C
•rl
r*
O
0)
cr

0)
«X
o
U


o
p»
rH




O
m
rH







,_)
S






CM
CO








in
3









rH
^*










C
•H
C
0)
sere

LjJ

O>
C
-H
N
-rH
CO

1)
J^
O
0


o
m
r-




o
CO
vo







00
o
*••*





0
t-
co







V
o
CM









o
o\
rH









U
O
4->

i
VH
O
o
M
0)
n
3
O
f^
4->
CO
re
O


o
CM
eg




o
o
CM







o
CM
i-H
*••*





CO
0
i-H







CTl
m









^*
in














M
0)
00
C
re
M
4J
,_,
re

p

j_)
o
X


o
CO
in




o
oo
^*







o
cr>
CM
*"*





0
vo
CM







O
rH









O
CO
iH













C
0
-H
1 1
re
N
-H
M
3
i-H
3
CO
0)
Q
D)
2§
-r4
4J
Cj8
rH
4> U
J3
4J 5
m O •
M M tn
UH 4J
0) -H
4) "O u*
3 C "O
O^-r(
4J C
g O O
s: m •*
fjv vl) C
3 3 Oi
O rH -H

5 o
•a "i 4J
4) 3
TJ cro
C *U 4^
3 _
O *J ^5
M O 41
c c-o
3 C
•o « p
c e M
(Q
tl) 01
M -Q 1J
o ^J
y 4J 4)
O *~*
IH -H m
C *J
o w
•H C-l
M (0 ^
£ 41 C
41 3 -H
4) > 41
1.4
o n -^
M O ti
r* 4-j
3 4)
M •
tj. 4) T
C -C 1
^| H CN
in
3 -0
• C
41 1
4J CN CN
fl 1
rH "O CN
3 C
(J fl 0)
—I. 0)
O 1 -O
CM fD
4) E->
U °>
fl 0) E
-H O
4) .Q M
ra W
4J C 0)
10 3
-H (0 tji
r; 4) VH
4J ^*
3 4J
C Cn o
•H VM fl
v
M T) 0)
M ^
4) T3 tj>
Sec
£= 3 —^
3 O n
55 M 3
2-29

-------
Table 2-2.  Potential PM-10 emissions estimates are given for all
processes where reasonable operating parameter and emission factor
data were available or could be estimated.

2.4.3  Model Plant Baseline Emissions

     Table  2-6 is  similar to  Table  2-5,  but  Table 2-6  lists
emission  estimates  for model plants  that  use  pollution control
methods assumed to be in use at typical iron and steel facilities.
These  calculated  emissions are  often referred to  as  "baseline
emissions"  because they  represent  the  emissions expected  from
plants that use currently accepted pollution  control techniques.
The  baseline   emissions for  each process  and  model plant  were
calculated using the  model plant parameters in  Table 2-4 and the
emission factors given in Table 2-3.
     The  emissions estimates  given  in  Tables  2-5  and  2-6  are
considered  representative of  typical plants;  however,  process
parameters and pollution  control practices will be different for
each  integrated  mill.  Extreme  caution should be  exercised in
interpreting the emissions estimates  in Tables 2-5 and 2-6 because
they are  calculated  from  very  general  operating parameters given
for the model plants and general emission factors.  Any assessment
of emissions from existing plants should be made on an individual
site-specific  basis for each  integrated mill  examined.   Sample
calculations   of  two  baseline  emission  figures  are  given  in
Appendix B.
                               2-30

-------




to
H
CO
CO
w
^J
^\
Vj
Oi

Q
B
o
M
a
H
CO

§
fc
CO
S
2
04

»4
M
1
5
Pi
Du
»**
CO
§
u
H
CO
CO
H
rl
W

o
H
1
S
CM
H
2
H
^q
W
^0
•s
m
W


*
VO
1
CM
a
a



"M
«
3
>i
Plant
(tons/

i
^
«
S §
«J 4J
H —*
04

H
H
Jit
id
ft
**S.
fcvt
S












n
n
8
01






O
ro
(N






o
rH
IN




O
CM






O
rH
rH






rH
VO








VO
in















Ifl
Cn
c
•H
ro
Q.
0)
M
O
U




in
o
rH






1>
rH
rH




P
m






(N
vo






CO
(N








rH
ro














a
O>
C
-rl
a
quenc

U
U
<0
^
Cn
sizin
0)
x
o
u




o
VO






m
in





CO
00






o
00






VO
rH








tn
rH











M
0)
•o 41 -o c 3 O M C 3 T3 C m tn ^ O 4J O m HH c O -H tn tn -H E 41 •o 41 -o C O M C 3 Oi C •H m 4-1 in •H £ c -H tn M 41 i 3 2 o -H 3 m 4J 3 cr o 41 U 4J T3 O 41 C T) >i 3 "S O 6 M 4) 41 rH )H J3 nj m 4J 4) rH tn S) f-i tq JZ 4J 4J 0) C-H -H .C 4-1 in 41 C 3 -H rH m (o > 4) M - 3 4) CP M -H 0 fe, ^M 41 M • 41 ^< JS 1 EHCN "O T f!3 1 oj ro 1 •O csi c m in 41 O rH 1 J3 CM (0 E-> to 41 E rH O J3 *-i m u-i E-< to C 1) -H ^] 3 w ty 4) -^ >q 3 4J _cr u *fcH TJ X T) 41 41 T) CP c c 3 -H O 0) M 3 o* c ~H T) £ 3 O M O 4-J 41 T3 10 (U 3 rH IB >-t 04 E-> C m £ 4-1 in 41 CT> M nj ^H 41 ^4 ro CT> C •H £ O c 41 3 cr 41 .V o o V4 o tn 41 3 rH >-l 0^ s Q 2-31
-------
2.5  REFERENCES FOR CHAPTER 2
1.   U.S. Department of Commerce.   1990  U.S.  Industrial Outlook.
     Chapter 16: Ferrous Metals.  Washington,  B.C.

2.   Gale Research,  Inc.  Ward's Business Directory of U.S. Private
     and Public Companies-1991.  Volume 4.  Detroit, MI.  1991.

3.   U.S. Environmental Protection Agency.  Revised Standards for
     Basic  Oxygen  Process Furnaces-  Background  Information  for
     Proposed Standards.  EPA-450/3-82-005a. Office  of Air Quality
     Planning and Standards,  Research Triangle Park, NC.  December
     1982.

4.   U.S. Environmental Protection  Agency.   Coke Oven Emissions
     from  Wet-Coal  Charged   By-Product  Coke   Oven  Batteries-
     Background    Information    for    Proposed     Standards.
     EPA-450/3-85-028a.    Office of  Air  Quality  Planning  and
     Standards,  Research Triangle,  Park,  NC.   April 1987.

5.   Standard  and  Poor's  Industry  Surveys.     Steel  and  Heavy
     Machinery.   Volume 158,  No. 30.   August 9,  1990.

6.   U.S. Environmental Protection Agency.  Control Techniques for
     Particulate Emissions  from Stationary Sources -  Volume  2.
     EPA-450/3-81-005b.    Office of  Air  Quality  Planning  and
     Standards,  Research Triangle Park,  NC.   September 1982.

7.   Letter from David C. Ailor, American Coke  and  Coal Chemical
     Institute,   to  James   H.  Maysilles,  U.S.  Environmental
     Protection Agency.  Comments on draft ACT document for PM-10
     emissions from iron and steel  mills.  Letter dated September
     25,  1992.

8.   Lankford,  W.T.,   et  al.,   eds.    The Making,  Shaping,  and
     Treating of Steel.  Tenth Edition.   Association  of Iron and
     Steel Engineers.   Pittsburgh,  PA.   1985.

9.   Letter from Thomas W. Easterly,  Bethlehem Steel Corporation to
     James  H.  Maysilles,  U.S.  Environmental  Protection Agency.
     Comments on draft  ACT document for PM-10  emissions from iron
     and steel mills.   Letter dated August 19,  1992.

10.  U.S. Environmental  Protection Agency.   Industrial Process
     Profiles for  Environmental Use.    Chapter  24: The  Iron  and
     Steel Industries.   EPA-600/2-77-023x.  Office of Research and
     Development, Research Triangle Park, NC.   February 1977.

11.  U.S. Environmental  Protection Agency.    Coke Quench  Tower
     Emission  Testing  Program.    EPA-600/2-79-082.    Industrial
     Emissions  Research Laboratory,  Research Triangle  Park,  NC.
     April 1979.


                               2-32

-------
12.  U.S. Environmental  Protection Agency.   Compilation  of Air
     Pollutant  Emission  Factors.    AP-42,   Fourth Edition  with
     Supplements.  Office of Air  Quality Planning and Standards,
     Research Triangle Park, NC.  September 1985.

13.  Parkinson,   G.      "Steelmaking   Renaissance."     Chemical
     Engineering, Volume 98, No. 5.  New York, NY.  May 1991.

14.  U.S. Environmental Protection Agency. Fugitive Emissions from
     Integrated Iron and Steel Plants.  EPA-600/2-78-050.   Office
     of Research and Development,  Washington, D.C.  March 1978.

15.  Varga,  Jr. and H.W. Lownie, Jr.  A Systems Analysis Study of
     the Integrated  Iron and Steel  Industry.   HEW  Contract No.
     22-68-65.  Battelle Memorial  Institute, Columbus, Ohio.   May
     1969.

16.  Buonicore,  A.J.,   and  W.T.   Davis,  eds.     Air  Pollution
     Engineering Manual.  Van  Nostrand Reinhold.  New York,  NY.
     1992.

17.  U.S. Environmental Protection Agency. A IRS Facility Subsystem
     Source Classification  Codes and  Emission Factor Listing for
     Criteria Air  Pollutants.   EPA-450/4-90-003.  Office  of Air
     Quality Planning  and Standards,  Research Triangle Park,  NC.
     March 1990.

18.  Jeffery,  J. and J. Vay.  Source Category Report for the Iron
     and Steel  Industry.   EPA-600/7-86-036.  U.S. Environmental
     Protection Agency, Research Triangle Park, NC.  October 1986.

19.  U.S.   Environmental   Protection   Agency.      Hot   Metal
     Desulfurization,  EOF  (Basic  Oxygen Furnace) Charging  and
     Oxygen   Blowing:      Level   1  Environmental   Assessment.
     EPA-600/2-82-036.   Research Triangle Park,  NC.  March 1981.

20.  U.S. Environmental Protection Agency.   Control Techniques for
     Particulate Emissions  from Stationary Sources   -  Volume  1.
     EPA-450/3-81-005a.   Office   of  Air   Quality Planning  and
     Standards, Research Triangle  Park,  NC.   September 1982.
                              2-33

-------
                            CHAPTER 3
                   EMISSIONS  CONTROL  TECHNIQUES
3.1  INTRODUCTION
     This chapter contains descriptions and data  for various PM-10
alternative  control  techniques  (ACTs)  for the  specific emission
sources described in Chapter 2 of this document.  Also, pollution
prevention  (source  reduction)  techniques,  which can  prevent or
reduce emissions from mills and associated fugitive area sources,
are provided.  Finally,  typical emission reductions are discussed.
     The emission control techniques discussed are emission capture
and collection  devices  that are  considered to be ACTs for PM-10
sources.   This  chapter focuses  on retrofit  control  techniques,
where  a  retrofit is  considered  to be  the  replacement  of,  or
addition  to,  pre-existing  equipment,  and provides  additional
information  on  ACTs  for  newly  constructed  facilities.    The
discussion of each control technique addresses design parameters,
operating parameters, and variables affecting operation.

3.2  SOURCE REDUCTION IN IRON AND STEEL MILLS

     Source  reduction  techniques  are  practices  that  reduce the
amount  of  any   hazardous  substance,   pollutant, or  contaminant
entering  any  waste  stream prior  to  recycling, treatment,  or
disposal.  Process improvements  may reduce the number of process
steps,  and  their  implementation  can result  in  reduced  energy
consumption and/or materials use.  These reductions in energy and
resources  use  may translate directly  or  indirectly  into lower
pollutant emissions.
     Because of  their design,  non-recovery coke  ovens inherently
release less particulate matter  to the atmosphere.  Non-recovery
coke ovens are  operated under  negative pressure, which minimizes
the emission of particulate from the oven.  Only  one non-recovery
coke oven is currently used in the United States.
                               3-1

-------
     Coke pushing operations provide  some  opportunity for source
reduction.   Coke  pushing is the  operation in which  hot  coke is
removed or "pushed"  from  the coking oven.  The  PM-10 emissions from
coke pushing  are  greater if the  coal is not  fully  coked before
pushing.  Optimizing the coking period can significantly reduce the
pushing emissions.  An increase in  the temperature  of the coking
oven,  an  increase  in  the  coking  period,   and  a  more  even
distribution of heat  in  the coking oven can  all  increase coking
efficiency and decrease pushing emissions.  The first two steps are
related, with the temperature of the oven walls and the amount of
time required for proper  coking being  interdependent.  There is an
upper temperature limit above which the  refractory  lining of the
oven  will  be  damaged  by  increased   heating.     Uneven  heat
distribution in the oven  can cause areas within the coal charge to
receive less heat  than necessary for full  coking.   If a flue or gas
nozzle is closed or defective in one section of the oven, then the
coal  in  that  section  may  not  coke completely,  resulting  in
increased emissions when it is pushed.   Moisture pockets within
charges of coal can also  cause  "greenness" or undercoking, as some
of the heat  provided by the oven must be used to heat and evaporate
the excess moisture.1
     Controlling the level of suspended solids in the coke quench
water is another  example of  source  reduction.   In general,  water
not evaporated or caught  by the coke is collected  in the quench
tower sump and recycled  for  later quenches.   Makeup  water can be
supplied by clean  water or by wastewater from other plant processes
such as  the coal  chemical  plant.   One  test  showed  particulate
emissions from quenching  that ranged from 0.29 to  1.22 kg of total
particulate matter per Mg of coal  originally charged.   When clean
makeup water was used  instead of coal  chemical plant makeup water,
the average amount of particulate  matter  emitted dropped from 1.1
to 0.68 kg/Mg.2 Another source confirmed this observation, noting
that when clean  quench  makeup water was used  instead  of  coal
chemical  plant wastewater,  average emissions were 1.5  to  three
times smaller.1
                               3-2

-------
     Some  firms  are  attempting  to reduce  their need  for coke.
Furnaces may  be  modified to allow powdered  coal injection (PCI)
directly into the combustion area.  One reference indicates  that up
to 50 percent of the coke used  in  steelmaking could be replaced by
coal.3
     Another  emission reduction  alternative  may be  to entirely
eliminate coke production from the steelmaking process.   The German
firm Deutsche Voest-Alpine Industrieanlagenbau has developed a two-
step iron production  process that uses non-coking coals.  First,
the iron ore  is  reduced in a shaft furnace.   The reduced ore is
then melted in a "melter-gasifier, " which both melts the ore and
produces the reducing gas for the  reduction furnace.  This process
reportedly avoids the emissions normally generated by coke-making
in addition to  lowering operating costs up to  25 percent.   This
process is in commercial use in South Africa but has not yet been
adopted by U.S.  steel producers.   As  of  1991,  one  unit had been
designed (in South Africa) with an operating capacity of  800,000 to
1,000,000  tons  per  year.   Several American firms  are  currently
examining the viability of the process.3
     Emissions from the  casthouse  area may  also be reduced by
employing pollution prevention  practices.  Some  new blast furnaces
feature  covered  runners  or other devices  that  prevent contact
between  the  molten  pig  iron  and the  ambient  air.    Thus,  the
formation of  particulate pollutants may  be  inhibited.4  Another
approach is to blanket the troughs and runners with a natural gas
flame  that  consumes  the  local   oxygen,  thus  suppressing  the
formation of iron oxides.
     Changes  in  some  work  practices  can  prevent  or  reduce
particulate emissions from open dust sources.  These work practices
focus on the operation of equipment used to transport, store,  and
transfer materials.    In the  case  of  unpaved  and  paved  travel
surfaces, emissions  can be reduced by decreasing vehicle speed and
weight, using dust  suppressants,  and/or employing road sweeping.
For materials  handling operations, emissions  can be reduced by
decreasing drop  height and  increasing  bucket  capacity.   Finally,
                               3-3

-------
emissions from wind erosion can be reduced by decreasing the size
of the active area of a storage pile or exposed ground surface.5

3.3  CONTROL EQUIPMENT

     Iron and steel mill process PM-10 emissions can be controlled
by systems consisting of two basic components,  a capture device and
a collection device.   (The collection device is also  known as a
control or removal device.)  Figure 3-1 shows a typical emissions
control scheme.  The particulate matter  is  captured in the local
air stream  by  a hood  or  other capture  equipment  and sent,  to a
control device,  which separates the particulate matter from the air
stream and  sends  the  cleaned air into the  stack.   The recovered
particulate matter  is  then removed and  recycled,  landfilled,  or
reused.
     This section discusses equipment used to capture both primary
and secondary emissions.   By definition,  primary emissions in the
EOF shop are those generated by the furnace and related equipment
while  the  furnace  is operating.  Primary  capture equipment  is
designed to capture  these primary emissions and consists of various
types of hoods  and enclosures.  Secondary emissions are defined as
emissions from  process steps that  are  not collected  by primary
emissions  control  systems.    Secondary capture   involves  using
building evacuation, booths and/or  hooding, and  enclosures.   The
control devices discussed here are those most frequently employed
in  the  industry  and  include  fabric  filters   (baghouses),  wet
scrubbers, and electrostatic precipitators  (ESPs).

3.3.1  Primary Control Equipment

     The operations generating primary emissions that are examined
in detail in this document include  coke  pushing,  coke quenching,
coke sizing and  screening, hot metal  transfer, and desulfurization.
Primary control equipment  is designed to capture  emissions from
these areas and direct them to collection systems.   Also, a brief
                               3-4

-------
O



I
                                                 0.2 ID
                                                o o
                                                   o
                                                             CO
                                                             c
                                                            _o
                                                            °eo
                                                             t/i


                                                            u

                                                            o
                                                                                          ro

                                                                                          (1)
                                            3-5

-------
discussion is provided  on primary furnace  emissions  and various
capture and collection equipment associated with them.

3.3.1.1  Primary control of BOF emissions
     Fumes from  the BOF generally  consist  of iron  oxides,  slag
particles, and carbon monoxide.  These fumes can be captured by a
hood located over the furnace mouth and routed to a control device
for PM-10 collection.   The most common types of  control systems
employed are the open hood with ESP,  open hood with scrubber, and
closed hood with scrubber.6
     The particulate emissions  from open hearth furnaces are mainly
(up to 90 percent) iron oxides, primarily in the  form of Fe20;J.  Due
to the predominance  of small particles  in these emissions,  high-
efficiency controls  such as wet venturi scrubbers and ESPs must be
used  in  conjunction with  appropriate  hooding.   Fabric  filter
baghouses may also  be used if the  exhaust gases are  cooled to
reduce the risk of accidentally igniting a bag.1

3.3.1.2  Primary control of other process emissions
     Control  systems  for  non-furnace  primary  emissions  vary
substantially in design between different sources.  Some systems,
such as  those for coke pushing and  coke  quenching,  are unique to
the specific operation.   Systems for casthouses are also somewhat
site-specific.    The following sections describe  control methods
that are available for each type of source.

     3.3.1.2.1  Coke Pushing-.   Three  methods are commonly used for
controlling emissions from coke pushing.   The first method is the
use of a mobile  scrubber  car, which consists of an enclosed quench
car coupled  with a   wet  venturi  scrubber car.  The  enclosed car
captures any exhaust gases  and particulates that  result from the
pushing  operation,   then vents  them into  a specially  equipped
exhaust  gas  cleaning car  for treatment.l   An  advantage  of the
enclosed quench  car  is the relatively low gas flow rate required
for operation,  which results in lower utility  costs.  However, the
                               3-6

-------
added weight of the control system may require that  the tracks and
support  system for the  car be strengthened  before use.8  Also,
industry  representatives indicate  that  the  enclosed  quench car
requires frequent maintenance,  leading many in the coke industry to
abandon its use.9
     The second available method uses a full-length shed  over the
exits  from  the oven.    The  shed traps the  pushing  emissions and
prevents their escape into the atmosphere.   The emissions  are then
drawn  into a control device  such as a scrubber or baghouse.1  One
advantage of the shed system is that its continuous operation also
captures emissions  from coke oven  door leaks.  It does,  however,
require large volumes of airflow to be effective.8
     The third method is the traveling hood.   This system  consists
of a mobile hood that is attached to the quench car  or to  the coke
guide.  The mobile hood with quench car combination can result in
very good capture efficiency, nearly at the  same level as  the shed
with baghouse, and  with lower  airflows.10   The emissions  from the
hood are vented to a fixed duct exhaust main,  which  is routed to a
collection device such as a scrubber or a baghouse.  A disadvantage
of  the  traveling  hood is  frequent  downtime  (5  percent)  for
maintenance and repairs.9

     3.3.1.2.2   Coke  Quenching.   After pushing,  the  coke is hot
enough to burn in the  presence of  air.  To prevent  the coke from
burning, most coke production  facilities cool the coke with water
in a quench tower.  In the  quench tower, water is sprayed  onto the
hot coke.   Emissions  from quenching  are  typically larger-sized
particulates created by the breakup of the hot coke when the quench
water contacts the coke. The particulate is carried up the quench
tower on the steam plume.
     Quenching is normally  conducted  with  relatively clean water
(less  than  1500  mg/L  total  dissolved solids) ;  however,  at some
facilities,  water containing over 5000 mg/L total dissolved solids
may be used (often called "dirty water quenching").16
                               3-7

-------
     Other  options  have  been  used  in  the  past  to  prevent
particulate emissions from coke  cooling.   The Kress Indirect Dry
Cooling  system is  unique in  that  it  eliminates  the need for
particulate control.  In  the Kress  system,  coke  is pushed into a
large box that mates directly to the door jamb.  Once the coke is
in the box, the box is sealed and  taken  to a cooling rack.   When
the coke in the box has been cooled (by  thermal  transfer through
the  box) ,  it  can  be  handled  as  dry,  cool  coke.   There are
essentially no pushing or quenching emissions from this system.16
The system has not been tried widely in the United States.
     Quenching of coke with inert gas under controlled conditions
can also eliminate quenching emissions and recover heat energy from
the hot  coke.   While  this  process has  been used  in  Europe and
Japan, it has not been tried in the United States.

     3.3.1.2.3  Coke Handling.   Coke handling consists of several
operations, including  sizing, screening, and conveying.  Emissions
from coke sizing and screening are typically controlled by a hood
or enclosure vented to a baghouse.
     Emissions from conveyors may be controlled in the same manner.
Coke conveyors often need  no controls as the coke  is  a strong  solid
and, unless it is  extremely dry, does not release particulate when
being  conveyed.   Conveyors  carrying dusty  materials  are   often
covered, and any  dust  that  may be  generated  falls  back onto the
material being conveyed.  Conveyor  transfer  points are sometimes
controlled by an enclosure, which, for very dusty  materials, may be
vented to a baghouse.10
     The crushing operation may  be  controlled by wet suppression
(see Reference 12)  with a  particulate matter control efficiency of
around  95  percent.   However, most  steel  producers  avoid adding
additional moisture to the coke.   Foggers  that wet the emissions
only  and not  the  coke  itself  may be  preferable  for  control.
Alternatively, these emissions  can be controlled  like coke sizing,
screening, and crushing emissions by enclosing the area and venting
                               3-8

-------
to  a baghouse.   This  type of  system  would  control  the PM-10
emissions without adding excess moisture.13

     3.3.1.2.4  Casthouse.  Casthouse emissions may be controlled
by any of  three methods.   The first method involves local hoods
vented to  a baghouse, which  can capture emissions  from various
sources within the casthouse.  These local hoods can have a capture
efficiency up to  95  percent  or greater depending on their design
and location.  A second option for casthouse emissions control is
total building  evacuation of the  casthouse area  exhausted  to a
baghouse.
     These  control   systems  can   have  an  overall  collection
efficiency of 95  percent  or  greater for PM-10.  One advantage of
total building evacuation  is the absence of bulky structures (i.e.,
hoods and  ductwork)   near the casthouse operations,  leading to
greater ease of casthouse operations and maintenance.6  Figure 3-2
shows a casthouse with an emissions control system.14
     A third method  for casthouse emissions  reduction involves the
use of suppression techniques.  Suppression techniques are designed
to  limit  or control  the  formation of  particulate (metal  oxide)
emissions  by  either  reducing contact  between  the hot  metal  and
ambient air/oxygen or disallowing formed particulate from escaping
the  casthouse.    The major techniques  used today  include flame
suppression, steam suppression,  and covers.   Any combination of
these three types can be integrated into one  suppression system for
a given facility.
     Flame suppression is accomplished  by  directing  natural  gas
fired from a  nozzle  (or  nozzles)  toward the  surface  of  the  hot
metal.  The gas  is ignited by the high temperature of the metal and
burns away the oxygen near the surface of the metal so that metal
oxide formation is hampered.  Flame suppression has been shown to
be effective on  particulate  emissions  from  troughs,  runners,  and
ladles.
     Steam suppression can control  PM-10 emissions in one  of  two
demonstrated methods.    As  one  example,  taphole emissions  are
                               3-9

-------
                                                  c
                                                  o
                                                  u

                                                  o
                                                  a
                                                  d


                                                 I
                                                  tt
                                                 X)
                                                  4)
                                                  a

                                                  3


                                                 ,8
                                                 4J
                                                  a
                                                  (0
                                                 U
                                                  I
                                                 CO

                                                  4)



                                                  E
                                                 •H
3-10

-------
controlled by steam agglomeration of particulate into larger sizes
which then settle inside the casthouse.  As a second example, iron
runner emissions can be controlled by a steam blanket that covers
the hot metal and prevents contact with oxygen.  Steam suppression
has been shown to  be  somewhat  effective  on particulate emissions
from the taphole, trough, runners, and ladles.
     Cover suppression  involves  the  use of  tight  fitting covers
that prevent  the hot metal  from contacting ambient air.   These
covers have proven effective for  controlling particulate emissions
from the trough and runners.  However, PM-10 emissions can escape
when the trough covers are removed for such activities as taphole
drilling.

3.3.2  Secondary Control Equipment

     This section  discusses control  of emissions  from hot metal
transfer and  desulfurization and from  secondary  control  of  EOF
emissions during charging, tapping,  and slagging.

3.3.2.1  Secondary control of EOF emissions
     The primary emission control system for a EOF is designed to
capture emissions  released during the oxygen blow,  in which the
furnace is  in the vertical  position  with  the mouth at  the top.
This  primary hood  may  not effectively  capture   emissions  from
furnace  charging,  steel  tapping, and  slag pouring  operations,
during which the furnace is rotated or tilted and its mouth is no
longer completely under the hood.7
     Two secondary control options currently used are close-in or
local capture hoods and roof-mounted hoods, with the latter option
being more frequently  employed.  Secondary hoods may be designed to
capture charging, tapping, and  slag pouring emissions.  Along with
secondary  hood  design,   which   is   usually  facility-specific,
operating  practices,   particularly  for  furnace  charging,  are
critical to effective  secondary emissions control.  The furnace is
tilted towards the transfer ladle no more than is  necessary to
                               3-11

-------
begin hot metal  pouring.   As the transfer ladle  is  emptied,  the
ladle and furnace are progressively rotated towards each other only
as much as necessary to  transfer metal while keeping the ladle and
furnace under the secondary hood as much as possible.7
     The  secondary  hood is typically evacuated through  a fabric
filter.  This secondary  control system is sometimes part of a more
comprehensive system in which other  secondary  hoods  that capture
emissions from non-furnace operations, including hot metal transfer
and desulfurization, may be evacuated through the same filter.

3.3.2.2  Control of other process emissions
     This  section   contains  control  options  for  the  hot  metal
transfer and desulfurization processes.   Typically,  the  emission
control systems  for  these processes are site-specific and thus are
inherently  different.    Figure 3-3  shows a   typical  hot  metal
transfer  station.6   There are  several  options  for  collecting
emissions from the hot metal transfer operation. The first capture
option is the installation of a close-fitting ladle hood around the
ladle at the transfer station.   The second alternative is the use
of a canopy hood, which  is  generally placed several feet above the
transfer  station.4   Due  to this positioning, the  canopy  hood may
require a larger volume  of  air to capture emissions as effectively
as close  fitting hoods.   However,  emissions  from the hot  metal
transfer are carried upward by the buoyant pressure of the heated
air.  Also,  if  a baghouse is used, the gas from  a close-fitting
hood may be too hot  and will require dilution air to cool the gases
for the baghouse.   Hoods for non-furnace  emissions may be either
fixed or movable, depending on the application.   Fixed hoods are
less burdensome  from an operational  standpoint since  they can be
located farther away from the process being controlled.  However,
they  require larger airflows.   Movable hoods must  be  in  close
proximity to the  process they control and are therefore more likely
to be an operational hindrance, but they require less airflow than
fixed hoods.  In  some facilities,  emissions from hot metal transfer
and desulfurization  are  captured by a side  draft hood.15  Regardless
                               3-12

-------
                              Exhaust to
                               baghouss
Figure 3-3.  A hot metal transfer and  skimming station.
                          3-13

-------
of the  hood  type employed  (canopy, movable,  close-fitting,  side
draft, etc.), the emissions are normally  vented to a baghoiise or
scrubber.    Since  the  hot metal  transfer  and  desulfurization
operations are often conducted at the same station within a mill,
emissions from both operations may be captured and collected by the
same control system.   Flame  suppression  (see discussion in Section
3.3.1.2.4) is another control option at  the hot  metal  transfer
station.7

3.3.3  Control System Performance

     The  effectiveness  of  the  control  systems discussed  here
depends mainly on the capture system's  performance.   In order to
effectively  collect  emissions  they must  first  be routed, via a
capture system (composed of  a capture device and ductwork), to the
collection device.   The capture efficiencies of hood designs used
for primary and secondary emissions vary widely.  A single system
efficiency value cannot  be assigned  to  these designs  because
capture depends upon installation and site-specific parameters such
as the hood placement and design face velocity.   This variability
makes  site-specific  capture  efficiency  determination  virtually
mandatory to ensure a high level of emissions capture.

3.3.4  Control Devices

     The primary and secondary devices that capture PM-10 emissions
from iron and steel mill  processes  comprise  the first part of an
effective control system.   The  second half of  the  system is the
control  (collection)  device.   The  control equipment  cleans  (or
removes) PM-10 from the air streams before exhausting to the stack.
     The  fabric  filter  is  the  most  common  control  device for
primary and secondary particulate emissions.  However, some mills
employ high efficiency wet scrubbers or ESPs for EOF shop primary
emissions  and   other   processes   depending   on  the  emissions'
characteristics.   For  example,   EOF  primary  emissions  are hot
                               3-14

-------
 (1,650-3,000°F) and  contain carbon monoxide  (CO),  providing the
opportunity for energy recovery from CO combustion for use within
or outside the plant.  These emissions are normally controlled by
an  ESP  or  wet  scrubbing  device.    Emissions  from  casthouse
operations may contain CO, sulfur dioxide (S02) , or high humidity,
which would require the use of  a wet scrubbing device.16  Table 3-1
lists  the  particulate  removal efficiencies  of  various  control
devices.4 The reasons for the predominance of fabric filtration use
in mills for PM-10 emissions control are discussed below.

3.3.4.1  Fabric filters
     The fabric filter  is a versatile type  of equipment  used to
collect  solid  particulates  from  an air   stream.    For  PM-10
emissions,  a collection efficiency of  99  percent or greater can be
expected.  As the  filter medium becomes caked with  particulate, the
collection efficiency actually increases, as does the proportion of
smaller  particles  collected.17   This section  discusses  the basic
principles of  operation  for a  fabric filter  and how collection
efficiency can be maximized.
     A  fabric  filter  system  (baghouse)  consists  of  several
filtering elements (bags) and a bag cleaning system all contained
in a  main shell  structure  that  is  equipped with  dust  hoppers
 (Figure 3-4).18  Particulate-laden gases are passed  through the bags
so that  particles are retained on the fabric in a filtering dust
layer that enhances  the  fabric filter's  performance.   The major
fabrics used in iron and  steel mill bags are polyester,  woven and
felted fiberglass, and Teflon  fluorocarbon materials, though the
specific   choice   of  material  depends  on  the  application.16
Typically,  a baghouse  is divided  into  several  compartments  or
sections, each containing several bags.  In larger installations,
an extra section is often provided to allow one compartment to be
out of service for cleaning at any given time without affecting the
overall efficiency of the system.18
     The basic mechanisms  used for cleaning particulate-laden gases
in a  fabric  filter  are  inertial  impaction,  diffusion,  direct
                               3-15

-------
         TABLE 3-1.  TYPICAL PARTICULATE COLLECTION EFFICIENCIES OF  CONTROL
                     DEVICES USED IN IKON AND STEEL MILLS4'3

                                                   Efficiency (percent)
                                                     by Particle Size
                                                       (micrometers)
Type of Device
Wet scrubber
- high efficiency
- medium efficiency
- low efficiency
Gravity collector
- high efficiency
- medium efficiency
- low efficiency
Centrifugal collector
- high efficiency
- medium efficiency
- low efficiency
Electrostatic precipitator
- high efficiency
- medium efficiency
- low efficiency
Fabric filter
- high temperature
- medium temperature
- low temperature
0 - 2.5

90
25
20

3.6
2.9
1.5

80
50
10

95
80
70

99
99
99
2.5 - 6

95
85
80

5
4
3.2

95
75
35

99
90
80

99.5
99.5
99.5
6 - 10

99
95
90

6
4.8
3.7

95
85
50

99.5
97
90

99.5
99.5
99.5
    Data represent an average of actual efficiencies.  Efficiencies are representative of well-
    designed and well-operated control equipment.   Site-specific  factors  (e.g., type of
    particulate  being collected,  varying  pressure drops  across scrubbers,  maintenance of
    equipment, etc.) will affect overall control efficiencies.  Efficiencies shown are intended
    to provide guidance for estimating control equipment performance when source-specific data
    are not available.
interception,   and   sieving.     The  first   three  mechanisms  are

prevalent  during  the first  few minutes  of  cleaning with  new  or

recently cleaned bags.    However,  the  sieving  action  of  the dust

layer  accumulating  on the fabric  surface soon predominates..   This
                                      3-16

-------
SHAKER
MECHANISM
OUTLET
PIPE
INLET
PIPE
3AFFLE
DLATE
1USTY AIR
.IDE
FILTER,
BAGS
                                                              CELL
                                                             ".PLATE
                                                             HOPPER
                Figure 3-4.   Fabric  filter.18
                                3-17

-------
sieving  mechanism   leads   to  high  efficiency   collection  of
particulates unless defects such as pinhole  leaks  in the bags or
cracks in the filter cake appear.19
     In fabric filtration,  both the collection efficiency and the
pressure drop across the bag surface increase as the dust layer on
the bag builds up.   Since  the system cannot  continue to operate
efficiently with the pressure  drop increasing  without limit, the
bags are cleaned periodically by a reverse airflow, pulse-jet, or
shaker mechanism.   The  dust  is collected  below the  filters in
hoppers and is either recycled or sent to a landfill.
     The fabric filter  capacity can be varied  widely with little
effect  on  efficiency.    This  inherent   flexibility permits  an
increase in capacity within reasonable limits by increasing system
fan horsepower.   An  oversized unit  is   more  desirable  than an
undersized unit because the dust loading  and gas volume can surge
during many mill operations.  A sudden increase  in volume flow rate
may decrease  the  dust  collection efficiencies on  other  types of
collectors but, within limits, does not affect the performance of
a fabric filter.   Sporadic  overloading can be readily accommodated
by a baghouse but this does not mean that  a fabric filter should be
operated  at wide variations   from  the  equipment  manufacturer's
recommendations .20
     Fabric  filters have  the disadvantage  of  being unable to
control  high-temperature  gas  streams.    Often  dilution air is
required  to  cool  certain gas  streams to  a  temperature  for
sufficient  baghouse  operations.   Also,  baghouses may not be the
most  suitable control  device for all  gas  streams  (i.e.,  high
moisture  content,  condensing  organics)  even though  they  have a
theoretical control efficiency advantage.
      As the data in Table  3-1 indicate,  the collection efficiency
of a fabric filter typically exceeds that of any other applicable
control device.   Fabric filters also have  many other advantages
that  make  them  suitable  for  control  of particulate emissions.
Fabric filters consume less energy (lower pressure drop, less fan
horsepower  required) than  either scrubbers or ESPs for equivalent
                               3-18

-------
outlet particulate concentrations.  They are efficient collectors
of very  fine emissions and  are  tolerant of  fluctuations  in the
inlet particle size distribution  (which affects ESPs).   Finally,
fabric filters collect particulate emissions as a dry dust, which
is easier  to handle  or recycle  than  the wastewater and sludge
collected from scrubbers.   However,  if desired,  the  dust from a
fabric filter can be wetted in a pug mill or pelletized before it
is recycled or landfilled,  decreasing handling problems associated
with the fine dust.
     Two types of  fabric  filter systems used in the industry are
the positive-pressure  type and the  negative-pressure  type.  They
are distinguished by how  the  exhaust air stream is moved through
the baghouse. Positive-pressure fabric filter systems are those in
which the effluent gases are  forced through the  fabric filter by a
fan placed  between the emissions  capture system and the fabric
filter.  Bag inspections and maintenance are easier to perform than
on negative-pressure   fabric  filters.    The  compartments  can  be
entered while the positive-pressure fabric filter  is in operation,
assuming that the temperature inside the compartments is low enough
for worker  safety.   Uncleaned air entering  the fabric  filter is
filtered  through the  cloth  and  then   vented  to the  atmosphere
through louvers,  a stack,  stub stacks,  or  a ridge vent (monitor) .20
     The alternative to the positive-pressure system is a negative-
pressure or suction-type fabric filter.  In this  system, the fan is
placed on the clean air side  of the fabric filter and effluent gas
is drawn through  the  fabric.   These  negative-pressure  filters
usually require less fan maintenance and less operating horsepower
than the pressure  type; however,  there are disadvantages.19  They
need to withstand the suction created by the fans,  and good sealing
is necessary to prevent the introduction of dilution air.  Despite
these disadvantages,  negative-pressure  systems are dominant in the
iron and steel  industry  because  of the  large-size particulates
sometimes present in mill  emissions, which can quickly destroy fan
blades if  not  filtered before entering the  fan.11   It  may  be
necessary to precoat  the  bags with another  material  in  order to
                               3-19

-------
prevent them  from being fouled or  plugged by tar  from the coke
pushing operation.1

3.3.4.2  Wet scrubbers
     Another  type of particulate  collection device  is  the  wet
scrubber.    Wet scrubbers  are broadly  employed  in industry  for
particulate control.  Generally, mechanically generated dust  can be
handled with medium or low  energy scrubbers, but fine particulates
can only be collected efficiently with higher energy scrubbers, as
shown by the data in Table 3-1.
     No wet scrubbing device,  however,  is  comparable  to a fabric
filter  for overall  particulate control  efficiency.    The  most
frequently employed wet  scrubbing design for PM-10 control is the
high energy venturi  scrubber.   In the venturi  scrubber, shown in
Figure 3-5, the water is injected into the inlet  gases,  which are
                <*"
passed through a venturi throat.  In spite of the relatively short
residence   time,   the  extreme  turbulence  in  the venturi  throat
promotes very  intimate  contact between the particulate  and  the
water.   The  wetted  particles  and  droplets are  collected in  a
cyclone spray separator.11
     Properly designed plant-size high-energy wet venturi scrubbers
are capable of  collecting  up to 99 percent  of total  particulate
emissions   from  various  processes  in integrated  mills,  including
those from the steelmaking  furnace  (EOF).  The efficiency of these
scrubbers   is  greater  than  90 percent  for particulate  removal
regardless of the process or type of particle being collected.21
     One  disadvantage   of  wet  scrubbers  is that  most  require
settling equipment  of appreciable  size, in which  the solids  can
separate from the water, and a recirculating system to reuse the
water  because raw overflow  from the  scrubber  system  cannot be
discharged into streams  or sewers.   Scrubber wastes include mill
PM-10 emissions considered hazardous to the environment that must
be treated.  Currently,  wet scrubbing technologies are  addressing
the wet waste  problem by recovering some reusable material  from the
waste stream, thereby reducing material and energy costs.
                               3-20

-------
tf
o
s
P4
pa
CQ
w
^
H
                                                                                          0)
                                                                                          O
                                                                                          to
                                                                                          a
                                                                                          0)
                                                                                          -u
                                                                                          a)
                                                                                           a)
                                                                                           ^
                                                                                           3
                                                                                           tn
                                                           Pi
                                                           H
                                                           a
                                          3-21

-------
3.3.4.3  Electrostatic precipitators
     The ESP  is a third  type of particulate  collection device.
Figure 3-6  shows a  single-stage  ESP,  which  is  the  type  most
applicable  to  the   iron  and  steel  mill  industry  due  to  the
industry's typically high flow rates.18   The ESP uses electrical
forces to deflect particulates out  of  the  effluent stream and onto
collector  plates.    In  the  ESP,  the  emission stream  is  passed
through a region in which  the gases are ionized.  Particles pick up
the charge from the ionized gas and are captured on the polarized
collector plates.16
     The plate-wire precipitator is  the most  common single-stage
ESP in current use.   It is well-suited for handling large airflow
volumes.   The ESP housing encases  several  electrically grounded
plates set at specific intervals with rows of discharge electrode
wires between the plates.  These electrodes are  long wires weighted
and hanging between the plates or  wires supported there by rigid
frames.16   High voltages  at  the  discharge electrodes  ionize gas
molecules  as  the  air  stream  flows  through  the  ESP.    The
electrically grounded plates function as collection electrodes and,
by electrostatic attraction,  capture the charged particles.22
     After time, particulate  builds into a layer  on the collection
plates  and must  be  removed without  reentrainment  in the  gas
stream.16'23   In a  "dry"  ESP,  the  collector plates  are cleaned by
rapping; in a "wet"  ESP, the plates are cleaned by a water spray.
If water is used,  it  must  be  cleaned before it  can be recirculated
or discharged.20
     Wet scrubbers and fabric filters  are  the dominant particulate
control devices  in  the  industry, partly  because ESPs often cost
much  more  to  install  and  maintain  than  high energy  venturi
scrubbers or baghouses with similar capacity.19  Recent enhancements
to precipitators  have  increased their efficiency while reducing
capital and operating costs.   Pulse energization and intermittent
energization  apply  a cyclic  high voltage  to  the  base voltage,
allowing the base voltage  to be reduced. These  high voltage bursts
                               3-22

-------
Ml  RAPPERS
HT CABLE  FROM
RECTI FIER
              .COLLECTING
              'PLATES
                                                           HOPPERS
                                               •»IRE-TENSIONING
                                                WEIGHTS
                                            HOPPER BAFFLES
           Figure 3-6.   Electrostatic precipitator.
                                                      is
                              3-23

-------
result  in  better  collection  efficiency while  the base  voltage
reductions provide energy cost  savings.  Also, wider plate spacing
and  the reduction  in  the  number  or  discharge  electrodes  has
resulted  in  capital   cost  savings  of  10-20  percent.    These
improvements may eventually  make  ESPs a more viable option for use
in the iron and steel  industry.11

3.4  NEW CONSTRUCTION CONTROLS

     There is no substantial technological difference between ACT
capture and control systems  for new construction and ACT retrofit
technology for most integrated  iron and steel mills.  However, the
retrofit installation  of a PM-10 control system in an existing mill
is typically  more expensive than new construction  at a  new or
expanding facility.  New construction costs are lower since the new
control system can be  designed and installed without adjusting or
changing existing equipment.  Typical industry estimates indicate
a  25  to  50  percent  increased  cost   for retrofit  versus  new
construction.16'23   As  a  result,  the  cost  per ton of  controlled
particulate can be lower for new construction than it  is for the
retrofit of a complete control  system.  Note, however, that the 25
to 50 percent increased cost is only an estimate and may be higher
(or lower)  on a case-by-case basis.  Additional information on ACT
system costs is presented in Chapter 5 of this document.

3.5  CONTROL METHODS FOR OPEN FUGITIVE DUST SOURCES

     Fugitive  dust  sources refer to  non-process generated air
pollutants that enter  the atmosphere without first passing through
a stack or duct designed to  direct or control their flow.  Several
non-furnace mill  activities generate  these  fugitive  emissions,
including vehicular travel on paved and unpaved roads, wind erosion
from storage piles, and materials transfer to or from vehicles or
storage piles.  Reference 12 describes control methods designed to
reduce PM-10 emissions from these sources.
                               3-24

-------
3.6  REFERENCES FOR CHAPTER 3


1.   U.S. Environmental Protection Agency.  Control Techniques for
     Particulate Emissions  from  Stationary  Sources  -  Volume 2.
     EPA-450/3-81-005b.    Office  of  Air  Quality  Planning  and
     Standards, Research Triangle Park,  NC.  September 1982.

2.   Laube, A.H., and  B.A. Drummond.   Coke Quench Tower Emission
     Testing Program,   EPA-600/2-79-082.  Industrial Environmental
     Research  Laboratory,  U.S.  Environmental  Protection Agency,
     Research Triangle Park,  NC.  April 1979.

3.   Parkinson,  Gerald.    "Steelmaking  Renaissance."   Chemical
     Engineering.  Vol. 98,  No.  5, pp 30-35.   May 1991.

4.   U.S.  Environmental  Protection Agency.   Compilation  of Air
     Pollutant  Emission  Factors  - AP-42.    Fourth Edition with
     Supplements.  Office of Air Quality Planning and Standards,
     Research Triangle Park,  NC.  September 1985.

5.   Cowherd Jr., C. and  Kinsey,  J.S.  Identification, Assessment,
     and    Control    of   Fugitive    Particulate    Emissions.
     EPA-600/8-86-023.    Office  of  Air  Quality  Planning  and
     Standards,  U.S.   Environmental  Protection  Agency,  Research
     Triangle Park, NC.  August 1986.

6.   U.S. Environmental Protection Agency.   Revised Standards for
     Basic Oxygen  Process  Furnaces - Background Information for
     Proposed Standards.  EPA-450/3-82-005a.  Office of Air Quality
     Planning and Standards, Research  Triangle Park, NC.  December
     1982.

7.   Telephone discussion with J.  Maysilles (Office of Air Quality
     Planning and Standards, U.S.  Environmental Protection Agency)
     and S. Snow (Alliance Technologies Corporation).  October 17,
     1991.  Discussed  Maysilles'  observations of iron and steel
     mill controls from plant trips.

8.   Goldman,  L.J.  and D.W. Coy.   Technical Support for Control of
     Co^e Pushing.  Contract No. 68-01-4141, Task 33.   Research
     Triangle Institute,  Research Triangle Park,  NC.   Undated.

9.   Letter from David C. Ailor,  American  Coke  and  Coal Chemical
     Institute,  to   James   H.    Maysilles,   U.S.   Environmental
     Protection Agency.  Comments on draft ACT document for PM-10
     emissions from iron and steel mills.  Letter dated September
     25,  1992.

10.  Letter from Thomas W. Easterly, Bethlehem Steel Corporation to
     James H.  Maysilles, U.S.  Environmental Protection Agency.
     Comments on draft ACT document for PM-10 emissions from iron
     and steel mills.   Letter dated August 19,  1992.
                               3-25

-------
11.  U.S. Environmental  Protection  Agency.   Coke  Net Quenching-
     Background  Information  for  Proposed Standards--Draft  EIS.
     Office  of  Air  Quality  Planning  and  Standards,  Research
     Triangle Park,  NC.  March 25, 1982.

12.  Cowherd, C., G.E. Muleski, and J.S.  Kinsey,  Control of Open
     Fugitive  Dust  Sources.    EPA-450/3-88-008.    Office  of  Air
     Quality Planning and Standards, U.S. Environmental Protection
     Agency, Research Triangle Park,  NC.   September 1988.

13.  Ohio Environmental  Protection Agency.   Reasonably Available
     Control Measures for Fugitive Dust Sources.   Columbus, Ohio.
     September 1980.

14.  Nicola,  Arthur  G.     "Blast  Furnace  Casthouse  Emissions
     Control."  Jron and Steel Engineer,  56  (8).   Association of
     Iron and Steel  Engineers.  Pittsburgh,  PA.  August 1979.

15.  U.S.   Environmental   Protection   Agency.       Hot   Metal
     Desulfurization,  EOF  (Basic  Oxygen Furnace)  Charging  and
     Oxygen   Blowing:      Level  1   Environmental   Assessment.
     EPA-600/2-82-036.      Industrial   Environmental   Research
     Laboratory,  Research Triangle Park,  NC.   March 1981.

16.  Buonicore,  A.J.,  and  W.T.   Davis,   eds.    Air  Pollution
     Engineering Manual.   Van Nostrand Reinhold,  New York.  1992.

17.  Foundry  Ventilation and  Environmental Control.    American
     Foundrymen's Society Inc.,  Des  Plaines,  IL.  1972.

18.  U.S.   Environmental   Protection   Agency.      Air  Pollution
     Engineering Manual  - AP-40.  Second  Edition.   Office of  Air
     Quality Planning and Standards,  Research  Triangle Park,  NC.
     May 1973.

19.  Fennelly,  P.P.  and  P.O.  Spawn.    Air  Pollution  Control
     Techniques for Electric  Arc  Furnaces in  the  Iron and Steel
     Foundry Industry.   EPA-450/2-78-024.  Office  of Air Quality
     Planning and Standards, U.S. Environmental Protection Agency,
     Research Triangle Park,  NC.  June 1978.

20.  Foundry Ventilation Manual.  American  Foundrymen's Society,
     Des Plaines, IL.  1985.

21.  Strauss,  W.     Industrial  Gas  Cleaning.    Second  Edition.
     Pergamon Press, New York.  1975.

22.  U.S. Environmental Protection Agency.  Control Techniques for
     Particulate  Emissions  from Stationary  Sources  -  Volume 1.
     EPA-450/3-81-005a.    Office  of  Air  Quality  Planning  and
     Standards, Research Triangle Park,  NC.   September 1982.
                               3-26

-------
23.  U. S. Environmental Protection Agency.  Control Technologies
     for  Hazardous  Air Pollutants.   EPA-625/6-86-014.    Air  and
     Energy  Engineering Research  Laboratory,  Research  Triangle
     Park, NC.  1986.
                               3-27

-------
                            CHAPTER 4
                      ENVIRONMENTAL IMPACTS

4.1  INTRODUCTION

     The  primary  purposes  of  this  chapter  are  to  identify
alternative  control  techniques   (ACTs)  in  addition  to  those
discussed in Chapter 3 for the iron and steel processes discussed
in Chapter  2  and to  estimate  PM-10 emissions  from model plants
employing  each option.    The  chapter  presents  the  alternative
control techniques  and their respective  PM-10  emission factors,
emissions,  and  emission  reduction   efficiencies   from   the
uncontrolled case.   Since  implementation of these  air emission
control techniques  may have  secondary environmental  impacts on
water  pollution,  solid waste  disposal,  and  energy consumption,
these topics are discussed briefly.

4.2  PM-10 EMISSIONS IMPACT

     Many capture systems  discussed  in  Chapter  3  can attain high
capture efficiencies.  Hoods, enclosures, and building evacuation
systems can attain  capture efficiencies  in  excess of 90 percent.
Of the collection devices  discussed  in  Chapter  3, fabric filters
and wet venturi scrubbers are well-suited for use in iron and steel
mills.  Total  PM-10 control efficiencies of 90 percent or greater
can be  attained by  control systems using  these   devices  if  the
capture and collection devices are  properly designed, constructed,
and operated.   Note  that  the  capture  efficiencies  and control
efficiencies  used  in  the  calculations  here  are  hypothetical
efficiencies thought  to be  within the  realm of  actual  control
system performances. The efficiencies assumed in this document are
not represented as  necessarily  typical  or recommended for actual
facilities.
     In this chapter, two alternative control techniques (ACTs) are
presented for coke pushing,  one  ACT  is  presented  for coke sizing
and screening,  and two ACTs are presented for controlling emissions

                               4-1

-------
from the casthouse.  The ACTs for the casthouse and for coke sizing
and screening result in improved PM-10 control from those emission
points in comparison to the baseline cases presented in Chapter 2.
The ACTs for coke pushing are  comparable  to the baseline control
option for PM-10 control efficiency and presented for comparison.
The ACTs presented here reflect the  variety  of techniques that are
in current  use within the  industry,  but are not  comprehensive.
Other techniques may be used successfully within the industry.
     No ACTs are presented for  coke  quenching, hot metal transfer,
or desulfurization.   The  control methods presented in the baseline
case  were  considered  the  principally  feasible control  methods
available for these processes.   Therefore,  any ACTs presented for
these processes would either be significantly less efficient than
the assumed baseline, or prohibitively expensive.
     It is emphasized that  the PM-10 calculations in this chapter
are made for the model plants given in Chapter 2  using emission
factor data taken from various sources;  they should not be directly
applied  to  operating  sources  that  possess  unique  operating
practices and conditions.  The  emission information presented here
is intended to provide a basis for  understanding emission factor
calculation methods and relative magnitudes  of emission sources in
example plants.  Assessment of  emissions from existing plants must
be made  on  the  basis of  operating information for the  actual
plants, rather than on the operating information assumed for the
model plants in this document.

4.2.1  Alternative Control Techniques for Coke Pushing, Coke Sizing
       and Screening, and Casthouse Emissions

     Emission  factors  for  the  baseline and alternative control
techniques  are summarized  in  Table 4-1.   The  baseline control
technique assumed for coke  pushing is a moveable hood (also known
as a  "traveling hood" or  "fixed duct"  system)  vented  to  a wet
venturi scrubber.  ACTs considered for coke  pushing are the use of
a  shed vented  to a fabric  filter  ["Coke  pushing  (Al)"]  in the
tables]  and a  mobile  scrubber car  ["Coke  pushing  (A2) "] ..   The

                               4-2

-------























1
^
gj
S5
0
H
CO
CO
H




O
H


Oi
Q
W
lj
CONTROL

•
H

^
3
a
H
















M
O
4«)
u
«
fa
c
o
•rl
n
Q
•d
a
w

o
H
I
£
04


H
H
0
V4
£
8
















4^
n
>i
CO
H
S
1
O
O











n
n
M
0
w
rt
s








•s
en
M
10

O
•H
10
O
O

^^
C
o
4J

x7
H

O

rH
O
Cn
S
$
IT)
O
•
O









M
0)
S
3
t-l
enturi sc
>
4->
a>
3
•o
C
10
T3
O
O
X




•J
^^

CF>
C
•rl
£1
n
3
a
0)
X
o
o


•a -a
0) 4)
Cn 0**
M M
10 10

0 0
rH i-l
10 10
0 0
0 0

-*-» *— V
c c
o o
4-> 4J

A XI
•H rH

\o o

o o
Cn Cn
•5 ^
51 x
CO lO
O 0
• •
0 0













0)
4J M
rH 10
=3 o
•H 0)
lf} _f)
(0 3
*W ^
0
T3 to
C
m a>
•o -H
0) A
£ 0





o o
iH CM
5 5


0* Cn
c c
n «
3 3
a a

S
0^
3

fl
U
10
0)
rH
o




m
en
c

r1,
uenc
cr
0)
^
o
u
•a -o
0) 0)
en en
U M
m m
A A
o o

H rH
10 10
O 0
O O

<*** *~+
C C
o o
4-> 4-1
•»»* -^
X) XI


O T
iH O
0 0

O O
2 S
Cn Cn
X X
o o
o o
• •
0 0







M
0)
4-)
0
-H
M
Xl
<0
0)
d c
•H M
M 3
rQ W
(0 O
M-t rH
O
•a c
§ *
T3 (0
O 4-1
O 0


•a -a
m <
Cn (ji
0) (1)
0) 0)
M M
O O
to to
\ \
Cn tn
c c
-H -H
N N
•H -H
n n
a> a)
x x
o o
o u




•H
10

§

4-1
O
A

4M.
C
o
4J
^s^
£1
rH

rH

O
Cn
^
X
IT)
iH
•
O















*o
0)

iH
O
4->
0
O
C




M
O
1
lj_J
O
o
M
0)
to
3
O
_£«
4-> 0*
to —
(0 D3
O —




•-)
T5
4-i
i

4J
O
J£

**.
c
o
4J
^^
jQ
rH

vo

o
tp
^
0^
(T)
o
•
0



&-I
(U
4J
rH


•H
M
(0
MH

•a
c
(0
o
4J
10
3
O
10
^
y>
Cn
-H
•a
-H
CD



M
o
4J
I
4-1
0
0
M
0)
to
3
0
.Cn
4J ^
(0 rH


4J
o
A

*-*.
C
O
4-)
^^.
XI
rH

cy»

o
en
*•
X
1*
o
•
o









M
4J
rH
:r*
d fabric :
c
10

to
'O
o
o
A
(0
o
o






M-l
o

M •^•^
 a>

4-> 4J
o o
JC A

*— - *•-••
C C
0 0
4J 4->
•*^ ^.
£1 S3
rH H

OJ fO

o o
en Cn
•^ ^
X X
iH rH
O O
• •
O O





M
0)
4->
rH
M 0
4J M
rH XI
»M (0
M-l
O
•H -O
M C
10
O
T3 O
C A
O^
•o c

O 4J
A 4J
a a>
o n
c o
o o



-a
BJ
0) —
m
to C
c o
(0 -rl
M 4J
4J rO
N
H -H
4J 3
(J) M-l
6 H
3
4-> W
O 0>
ffi Q






















w 07
O 0
o rl

M M
IU (LJ

T1 Tl
aj a)
•IJ J_)
c c
0} i])
n n
0) 0)
M M
p, pt

n n
a) a)
>i>,
to co

a) a)
3 3
cr er
-H -H
|||
4J 0) 0)
>1
o o

3 4j 4j
cr c c
-H O O
coo
S S -H -H
>.rH (0 (0
to O C C

O C 4-> 4J
,4 0 H H
jj O fl 10
0 J 0 0
o > * *
•H 4J 4->
"'j ,.
n\ fl) JJ Tl
rrt JJ W C '
nt rH t fsi
X)  C
M n M
0) W 0) "5
>W -H <*-! J)
0) g 0) >
ii
...
" C TJ C J2
iH O U) O ^
fl)  -H o -H • ;
n-i M o u 0)
J5* £ rt
o-o ?«P 2
4-3

-------
baseline emission  factors are discussed  in Section 2.3.2.   The
emission factors given in Table 4-1  for coke pushing controlled by
a  shed vented  to a  fabric  filter are  composites of  fugitive
emission calculations and  process  emission  calculations.   The
fugitive emission portion of these emission factors is calculated
using  an assumed shed capture efficiency of  97 percent  and the
uncontrolled PM-10 emission factor for AP-42.1  The process emission
portion of  these  emission  factors  is  calculated  using  typical
airflows  (350,000-1,000,000  acfm)   and  an assumed  minimum grain
loading of  0.003 gr/dscf.2'3   The  emission factor  for  the mobile
scrubber car is based  on  performance  ratings provided by  an
industry representative.2
     The baseline  control technique assumed  for coke  sizing and
screening is a hood vented to a fabric filter.  The ACT presented
for coke sizing  and screening is a total enclosure vented  to a
fabric  filter.    The  ACT emission  factors for  coke  sizing and
screening are based  on  an  uncontrolled  emission  factor  of  0.04
kg/Mg coal charged (0.08  Ib/ton)  taken from Reference 4, an assumed
total enclosure  capture  efficiency  of 98  percent,  and an assumed
fabric filter minimum exit grain loading of 0.003 gr/dscf.2'3'4
     In the  baseline case, emissions from the casthouse are assumed
to be uncontrolled and emitted through the casthouse roof monitor.
The  AP-42  PM-10  emission  factor  for  uncontrolled  casthouse
emissions is 0.15 kg/Mg hot metal produced  (0.31 Ib/ton).'  Two ACTs
are presented for  controlling emissions from  the  casthouse.   The
first casthouse ACT ["Casthouse (Al)"]  is total building evacuation
to  a   fabric  filter.    The assumed building evacuation  capture
efficiency  is  95  percent,  and  the assumed  minimum  exit grain
loading is  0.003 gr/dscf.2'3  Historically, building evacuation has
been rarely  used, due  to  cost and operating drawbacks, but this ACT
option is presented here for comparison of emission reductions and
(in Chapter 5)  control system costs.
     The second  casthouse ACT ["Casthouse  (A2)"]  is a combination
of  local hoods   over  the trough and runners  vented to  a fabric
filter.  The assumed capture device efficiency is 80 percent, and
                               4-4

-------
the  assumed  fabric  filter  minimum exit  grain loading  is 0.003
gr/dscf ,2'3
     The  above   assumptions  in  combination  with  the  AP-42
uncontrolled  PM-10  emission  factor  are  used to  generate  the
casthouse ACT emission factors in Table 4-1.
     Table 4-2 summarizes the control system efficiencies for the
processes and  control systems  in Table 4-1;  these efficiencies
represent the PM-10 emission reduction from the uncontrolled case.
Emissions  from  each process  in  Table  4-1 are   summarized  in
Table 4-3.

4.3. WATER POLLUTION IMPACT

     Of the  control techniques presented  here, coke pushing and
coke  quenching  operations   use  water  for  control of  airborne
particulate.    Disposing of  the  used water  can  create potential
water quality impacts that should be considered when evaluating the
environmental impact of implementing  these  options.   These impacts
are briefly discussed in this section.
     Used water  from coke pushing and coke quenching particulate
control contains solid  emissions from  these processes.   These
emissions may include coke  (carbon),  crude tar,  crude light oil,
ammonia,  phthalates, and other volatile organic gases.5'6
     Wastewater  composition  differences  between  the baseline and
alternative control technique cases are expected to be a function
of  quantity  more than  type of  substance.   All  control options
presented use water for control of coke quenching operations.  The
baseline and second alternative  control  options  for coke pushing
use water, but the  first alternative control option uses a fabric
filter for control  of coke  pushing emissions.   Therefore, model
plants under the first  alternative technique would generate less
wastewater than  those using  the baseline  or  second alternative
technique controls.
     It common practice  for coke oven facilities and iron  and steel
mills to  recycle water used in pollution  control  systems.   For
water discharged from the facility, the Clean Water Act, along with

                               4-5

-------
        TABLE 4-2. PM-10 CONTROL SYSTEM EFFICIENCIES FOR
                     BASELINE  AND  ALTERNATIVE  CONTROL
                     TECHNIQUE (ACT)  SYSTEMS
              Process
  Control
Efficiency%
         Control system
   Coke pushing  (B)a


   Coke pushing  (Al)


   Coke pushing  (A2)

   Coke quenching  (B)

   Coke sizing/screening
   (B)

   Coke sizing/screening
   (A)

   Casthouse (B)

   Casthouse (Al)


   Casthouse (A2)


   Hot metal transfer  (B)


   Desulfurization (B)
     78b


     88°

     79b

     90b

     86b


     96C


       0

     85C


     72C


     72C


     88C
Hood vented to venturi scrubber
Shed (97%)  vented to fabric
filter

Mobile scrubber car

Clean makeup water and baffles

Hood (90%)d vented to fabric
filter

Total enclosure (98%) vented to
fabric filter

Uncontrolled

Building evacuation  (95%)
vented to fabric filter

Local hood  (80%) vented to
fabric filter

Hood (90%)d vented to fabric
filter

Hood (90%)d vented to fabric
filter
  •»(B)" - baseline control system.
   "(A)" - alternative control technique system.
   "(Al)" - 1st of two alternative control technique systems presented for single process.
   "(A2)" -  2nd of two alternative control technique systems presented for single process.
  bCalculated from AP-42 controlled and uncontrolled emission factors.  No efficiencies
  given for individual capture and collection devices.
  °Control efficiency calculated from assumed capture efficiency and a fabric filter minimum
  exit grain loading of 0.003 gr/dscf  (References 2 and 3).
  ""specific capture device not given in AP—42.  Capture efficiency assumed
  Note:  percentages calculated from unrounded figures used in Table 4-3.
any  local or  State  regulatory standards,  require  that  facilities

limit their release  of such compounds  as oil and  grease and other

organics.   In addition, there  are generally limits  on the amount of

total suspended  solids  (TSS) in wastewater released from the mills.

Facilities must  also adjust wastewater  pH where necessary to comply

with applicable  regulations.   Wastewater from  iron  and steel mills

is  not  expected to  exhibit  biochemical oxygen  demand  (BOD)  and

therefore should not need  secondary treatment  for that  condition.
                                     4-6

-------


CO
W
CO
CO
H
O
^J
ftj
ft
Q
CJ
3
H
CO

^
P5
O
fa
CO
2
|K
3
04
M
U
Q
y
A
M
5
g

CO
55
O
H
CO
CO

*g
W
o
rH
%•
04
Q
*j
W
hH
o
2

p
CJ



CO
1


a


r-»






"M
i,
1 1

A.
Ml
0
8»

5 «
x,.





^^
n
4
*^
•u 2
g §
H *>
Ot *"*
|

"0 LJ
3 s

^^

g



"S"

li


c o
« -u
Pi
H
•!
co 10
&
t?i
A
**









w
a
Of
u
8
Ml









o o
CO 
C
•H
C
(D
0)

0
n
Cn
C
-H
N
•H
n
0)
•X ^-*
O ffl
u —


r-







^








•a1
—







*?








CM
•^






CM







Cn
c
-H
c
0)
0)
M
O
n
C
-rl
N
-H
n
0)
^ ^-.
0 <
u —


o
m
rH






o
rH







~
[2.







CM
t-







O
»r
^•*





CO






M
O
4-)
-H
C
i

4-1
O
O
0)
CO
3
O
x:

« r7
m <
0 —


0 0
H VO
CM v






o m
rH







O CO
rH CO
r~H ^™*







o o
O CO
rH






VO VO
in rH
*-^ *™^





I-H in
in rH






M —
o o
^J *— *
-H
c w
o a)
n
U-4 C
o id
0 M
M 4-1
 ~
(0 CM 4->
id < o
O — as


vo
VO






o








vo
CO







CO
CO







co
i-H
*~"





vo
rH









^^
pa
**"*
c
o
•H
n)
N
-H
M
3
M-l
3
n
0)
Q

n
0)
MX)
n o> c
CD 3 3
U rH O
3 id M
Cn >
1=5 rH W
4j a a
3 3
CX CJ* 9
"o> * 2
3 -P id
0 0 4J
U C
** *!c
« «"0 «^
W 05 (p
w n -O  1
•U 4J xj 4) CM
C C o 3
0) a>  4J en C n
>i c id 3
n H i-H .3 tj<
o o n rH -.3
tt) H H 3 1
ere "c TJ "* o
•H O O > 3 id en
^J JJ -rl -H fj JJ C
W 4J -U ,H -H
>iH id id ,Q c n

n n u
rH 4J d) CO J) 00 95
4J O Id Id "* 3 -H
C fl) Cn*J
O fl) O O ^-4 ;r^ (rt
0 > 3 9 3 H
-H 4J t) -J3 X) 3
0) J-> jj tt) O
•H C O O m C Id
H M -H 3 0

Id H rH CM ^ O
X5 id c C M
1 1 -H id >w
ii x:
• • 0) 4J TJ
LLr7cM M U C
03 i4^ rtj rfl n tt) 'H
^^ ^-* ^^ *— * g j^ id
SEES § 4J 4J
2 id X3
« jiT M O






































4
en
•H
•o
3
X
2
3


^.
Ou

tt) S
M5
3 ll
Cn

4ji?
C _j
id «
0 g
HH «
Cn 4)
•H S
CO 3
§£

4-) S4
rri
o s
-Po
4-7

-------
4.4  SOLID WASTE IMPACT

     For many  of  the processes examined here,  fabric filters are
used  to control  particulate  emissions.    Nearly  all  materials
(especially coke recovered  from the  coke  pushing operation using
the first control  technique)  recovered from  fabric  filters can be
reused  in  the  mill or  sold  for  reprocessing.7   If wastewater
treatment ponds or tanks are used,  sludge from the ponds must also
be recycled or discarded.   Where  materials  are discarded, these
operations generate an  additional solid waste impact  from the mill.
Table 4-4 lists some materials commonly  found in the particulate
generated by iron and steel mill operations.

           TABLE 4-4. COMPOUNDS COMMONLY EMITTED AT IRON
                      AND STEEL MILLS*'b
Coking Emissions
Coke (C)
Crude tar
Crude light oil
NH3
CO
CH4
Phthalates



Blast
Fe
Fe203
Si02
A1203
CaO
ZnO
P4
S
Mn
C
Furnace Emissions
MgO
Cr203
MnO
V205
Na2O
Ti02
Pb
BaO
K2O

"Reference 1.
^Reference 6.
     Sludge  volumes  can  sometimes  be  reduced  if the  sludge  is
dewatered  by gravity  settling,  filtering,  or  other techniques.
Solid waste from mills is usually classified as nonhazardous and is
managed  at nonhazardous  waste  landfills.   However,  solid  waste
generated  from  capture of desulfurization  emissions may  contain
significant concentrations of toxic metals,  necessitating  disposal
at a hazardous waste landfill.   Testing  at Kaiser  Steel's  Fontana,
                               4-8

-------
CA  mill  revealed  significantly  high  concentrations   of   lead,
mercury, barium, antimony, strontium, arsenic, copper, manganese,
and chromium in the desulfurizing process air stream.8  However,  an
industry  commenter  indicates  that  the  Kaiser  results  may  be
atypically high for the industry.9  Regulatory  limits  for  solid
waste and hazardous  waste  disposal are typically provided within
legislation  such as the  Resource  Conservation  and Recovery Act
(RCRA), the  Superfund  Amendments  and Reauthorization Act  (SARA) ,
and various State and  local rules.

4.5  ENERGY IMPACT

     Nearly all  particulate  control  techniques require  energy  to
induce  airflow,  pump water,  or perform other tasks.    Table 4-5
summarizes the energy  impacts  associated  with  implementing the
alternative control techniques at the model plants.  Energy impacts
are calculated for fans (used for  fabric filters and wet scrubbers)
and water pumps  (used for wet scrubbers), the two principal energy
users within the air pollution control systems. Energy impacts for
fabric  filters  and wet scrubbers were  estimated  using techniques
presented in  EPA guidance.10'11  Reference  10  contains  an energy
consumption  equation  for  fans  that is based on airflow, system
pressure drop,  hours  of  operation,  specific gravity  of  air, and
fan/motor efficiency:
       E  =      0.746xQxPDxsx HRS
                           6,356ri
     where
       E  =    energy  consumption in kwh/yr
       Q  =    airflow  rate in acfm
      PD  =    system  pressure  drop in  inches  H20
       s  =    specific gravity  of gas relative to  air
     HRS  =    hours of operation per year
       Tj  =    combined fan and motor efficiency
For  these  calculations,  PD  was   assumed  to  be  15  inches  H20,
specific gravity was 1.0, hours  of operation were  8,760,  and  r\ was
equal to 0.65.

                                4-9

-------
     Coke  pushing  (A2)   power  requirements  were  derived  from
Reference  11,  where energy consumption  figures  of  400-600  kW to
power  the  scrubber car  transport are  reported.   An  industry
reference provided  other  data  indicating that the combined power
requirement for the scrubber car transport and scrubber operation
was on  the order of 3,500 kw.2  The  energy  requirements for the
mobile scrubber car in Table  4-5  were  calculated using this energy
consumption factor,  the  coke  throughput specified  for  the model
plants, and an assumed operating  time of 8,760 hours per year.  All
other  process  energy  requirements   were  calculated  using  the
equation  from Reference  10.     It  should be  noted that  energy
requirements for emissions control  from both primary and secondary
sources are typically less than one percent  of  the total energy
used for all mill operations.12
     Energy use  for an air pollution  control  system is directly
proportional  to the  air  flow  rate through  the  system  and the
pressure drop necessary to induce the airflow.  The flow rates used
to estimate energy consumption  are  summarized in Table 4-6.  These
airflows were  derived  from published   reports  (reference numbers
given  in  the  table)  that provided specific airflows  or  ranges of
airflows based on either  a model  plant size or throughput.  These
airflows were  adapted  to the throughputs and  plant  sizes in this
document to determine the ACT model plant and process flow rates.
     Pressure drops for fabric  filters  may range  from 0.75 to  4 kPa
(3 to 16 in. H20),  while pressure drops for wet-venturi scrubbers
may  range from  1.5  to  20 kPa  (6 to  80 in.  H20) .    As stated
previously, the  energy requirements for  fabric  filter  systems  in
Table 4-5 were estimated  assuming a static pressure drop of  4 kPa
(15 in. H20).
                               4-10

-------
   TABLE  4-5.  ENERGY REQUIRED  BY  IMPLEMENTATION OF PROCESS
                 CONTROL  OPTIONS
Process
Coke pushing (Al)a
(shed with fabric filter)
Coke pushing (A2)b
(mobile scrubber car)
Energy Consumed (kWh/yr)
Small Medium Large
Plant Plant Plant
8.3 x 106 17 x 106 24 x 106
10.5 x 10s 15.8 x 10s 26.3 x 106
+87.6 x 109 +131 x 109 +219 x 109
Btuc Btu Btu
 Coke  quenching
  (baffles with clean  makeup
 water)

 Coke  sizing/screening (A)
  (enclosure with fabric
 filter)

 Casthouse (Al)a
  (building evacuation with
 fabric  filter)

 Casthouse (A2)a
  (local  hoods with  fabric
 filter)

 Hot metal transfer
  (canopy hood with  fabric
 filter)

 Desulfurization
  (close-fitting ladle  hood
 with  fabric filter)
  n/avd
n/av
n/av
0.2 x  106     0.5  x  106      1.0 x  106
7.1 x  106      14 x 106       21 x 106
4.8 x  106     9.5  x 106       14 x 106
3.6 x  106     7.1  x 106       11 x 106
1.2 x  10s     2.4 x  106      3.6 x 10*
•"(A)" - alternative control technique system.
 "(Al)" - 1st of two alternative control technique systems presented for single process.
 "(A2)" - 2nd of two alternative control technique systems presented for single process.
"See Reference 11
°The mobile scrubber car uses electricity to move the car, and uses petroleum fuel (typically
No.  2. fuel oil)  to heat water for the scrubber system.
''n/av - not available
                                   4-11

-------
  TABLE  4-6.   AIRFLOWS  REQUIRED  FOR  PROCESS CONTROL  OPTIONS
                                                  Airflow  (acfm)
Process
Coke pushing (Al)a
(shed with fabric filter)
Coke pushing (A2)b
(mobile scrubber car)
Coke quenching
Small
Plant
350,000
n/avc
n/appd
Medium
Plant
700,000
n/av
n/app
Large
Plant
1,000,000
n/av
n/app
  (baffles with clean make-up
 water)

 Coke  sizing/screening (A)e
  (enclosure with fabric filter)

 Casthouse (Al)e
  (building evacuation with fabric
 filter)
 10,000


300,000
20,000
600,000
40,000
900,000
Casthouse (A2)a
(local hoods with fabric filter)
Hot metal transfer6
(canopy hood with fabric filter)
Desulfurization6
(close-fitting ladle hood with
fabric filter)
200,000
150,000
50,000
400,000
300,000
100,000
600,000
450,000
150,000
•Reference 2.
''Reference 11.
cTypical flowrates are on the order of 10,000 acfm per ton of coke  (Reference 3).  Reference
10 gives energy  consumption  in  IcWh;  this figure is used for the  energy consumption
calculations  in Table 4-7.
dnot applicable.  This control  technique does not require  an induced airflow.
•Reference 3.
                                   4-12

-------
4.6  REFERENCES FOR CHAPTER 4


1.   U.S. Environmental  Protection Agency.   Compilation  of Air
     Pollutant  Emission Factors.    AP-42.   Fourth  Edition  with
     Supplements.  Office of Air  Quality  Planning and Standards,
     Research Triangle Park,  NC.  September 1985.

2.   Letter from Thomas W. Easterly, Bethlehem Steel Corporation to
     James  H.  Maysilles, U.S.  Environmental Protection  Agency.
     Comments on draft ACT document for PM-10 emissions from iron
     and steel mills.  Letter dated August 19, 1992.

3.   Buonicore,  A.J.,   and  W.T.   Davis,   eds.     Air  Pollution
     Engineering Manual.  Van Nostrand Reinhold,  New York.  1992.

4.   U.S. Environmental  Protection Agency.  AIRS Facility Subsystem
     Source Classification Codes and  Emission Factor Listing for
     Criteria Air  Pollutants.   EPA-450/4-90-003.  Office of Air
     Quality Planning and Standards,  Research Triangle Park, NC.
     March 1990.

5.   U.S. Environmental  Protection Agency.  Control Techniques for
     Particulate    Emissions    from    Stationary   Sources.
     EPA-450/3-81-005b.     Office   of  Air  Quality  Planning  and
     Standards, Research Triangle Park,  NC.   September 1982.

6.   Laube,  A.M.,  and B.A.  Drummond,  Coke Quench Tower Emission
     Testing  Program.    EPA-600/2-79-082.    U.S.  Environmental
     Protection Agency,  Industrial Emissions Research Laboratory,
     Research Triangle Park,  NC.  April 1979.

7.   Letter from R. Wade Kohlmann, Indianapolis  Coke, to James H.
     Maysilles, U.S.  Environmental Protection Agency.  Comments on
     draft ACT  document  for  PM-10 emissions  from iron and steel
     mills.   Letter dated August 44, 1992.

8.   U.S.   Environmental   Protection   Agency.       Hot   Metal
     Desulfurization, EOF  (Basic  Oxygen Furnace)   Charging and
     Oxygen   Blowing:      Level   1  Environmental   Assessment.
     EPA-600/2-82-036.   Industrial Emissions Research Laboratory,
     Research Triangle Park,  NC.  March 1981.

9.   Letter from Robert  E. Sistek,  LTV  Steel Company to James H.
     Maysilles, U.S.  Environmental Protection Agency.  Comments on
     draft ACT  document  for  PM-10 emissions  from iron and steel
     mills.   Letter dated August 25, 1992.

10.  U.S. Environmental  Protection Agency.   OAQPS  Control  Cost
     Manual.   Fourth Edition.   EPA 450/3-90-006.  Office of Air
     Quality Planning and Standards,  Research Triangle Park, NC.
     January 1990.
                               4-13

-------
11.   U.S. Environmental  Protection Agency.    Envirotech/Chemico
     Pushing Emissions Control System Analysis.   EPA-340/1-83-019.
     Office of Air Quality Planning and Standards, Washington,  DC.
     April 1983.

12.   U.S. Environmental Protection Agency.   .Revised Standards  for
     Basic  Oxygen Process  Furnaces—Background Information  for
     Proposed Standards.  EPA-450/3-82-005a.  Office of Air Quality
     Planning and Standards, Research Triangle Park, NC.  December
     1982.
                               4-14

-------
                            CHAPTER 5
                      CONTROL COST ANALYSIS
5.1  INTRODUCTION

     This chapter  presents  cost analyses of  the  ACT options for
PM-10  control  from  the  model  iron  and  steel  mill  processes
presented in  Chapter  4.   Equipment  costs,  capital  costs, annual
costs (including operation and maintenance) ,  and cost effectiveness
are  calculated  for each process  control option  and model plant
size.  Also  included in this chapter are the  comments  of industrial
representatives who reviewed the  draft version of this document.
All cost estimates  given here  apply only to the emission sources
described in  Chapter  4,  and not to  other PM-10 sources that may
exist in the mills.  Section 5.2 summarizes the design parameters
assumed  for  each  process   control  system  and  the  associated
purchased equipment costs.   Section 5.3 describes and summarizes
the capital costs for each process control option, including cost
algorithms used  in the  calculations.   Section 5.4  describes and
summarizes  the  annual  costs  for each  process  control  option,
including cost  algorithms  used in  the calculations.  Section 5.5
summarizes the  cost effectiveness  estimates for each process ACT
option and model  plant  size.   Finally Section 5.6  contains cost
comparisons provided by industry representatives  (asked to review
a draft version of  this document)  for most  of the control option
costs  presented in  this  chapter.    These  cost  comparisons  are
provided to show the variability of cost estimates within  the iron
and steel industry.   To assist readers interested  in completing
their  own  cost  estimates for  existing  facilities,  this  chapter
makes  detailed   reference  to  the  texts,  tables,   figures,  and
equations  used  to  develop  the  cost  estimates  contained  in
Sections 5.2,  5.3, and 5.4.
                               5-1

-------
5.2  DESIGN PARAMETERS AND PURCHASED EQUIPMENT COSTS

     The process ACT system options presented in Chapter 4 achieve
PM-10 emissions control by the addition of capture and collection
equipment.  These additions  increase both the  capital and annual
costs for operating each controlled process.  To develop purchased
equipment  costs  for   each   process   control   option,   detailed
assumptions were made about the types and sizes of equipment needed
and the operation of the equipment.   These assumptions were based
on cost algorithms presented  in the EPA  cost manuals titled Capital
and Operating Costs of Selected Air Pollution Control Systems and
OAQPS Control Cost  Manual.1"2  Note that  the  costs  and estimating
methodologies given here are directed towards  a "study" estimate of
30 percent accuracy. Study estimates are used to roughly evaluate
the  economic  feasibility  of a  project  using  relatively  small
amounts of data and without expending great effort.  Other general
classifications  of  estimates include  the  "order-of-magnitude"
estimate, less  accurate than a study  estimate;  "scope,"  "budget
authorization" or "preliminary" estimates, nominally of 20 percent
accuracy; "project control" or "definitive" estimates, nominally of
10 percent  accuracy; and  "firm,"  "contractor's," or "detailed"
estimates, nominally of 5 percent  accuracy.2  Study estimates have
been extensively used  in regulatory development, while  the more
detailed techniques have been more typically used in actual plant
practice.  This section presents the design parameters and costing
methodologies as well  as the associated  equipment  costs  derived
from them.

5.2.1  General Equipment Assumptions

     Two  types  of  equipment parameter assumptions were  used for
this study:  general and process-specific. General assumptions are
described  in  this   section  and  apply  to  all  process  control
equipment  discussed herein.   Process-specific  assumptions  are
described in subsequent sections and are dependent on the process
and model process size being controlled.

                               5-2

-------
     Table 5-1 lists  the  general  equipment assumptions that were
made for each type of control equipment including ductwork,  fan and
motor system, and baghouse parameters.  Each of these assumptions
is required as an input  into the  cost methodologies presented in
the cost manuals.1-2   The  ductwork  and fan/motor system costs were
developed from Reference  1;  baghouse, capital,  and annual costs
were derived from Reference 2.    Equipment assumptions  for  the
mobile scrubber car were not necessary because the costs were given
directly in Reference 3.

5.2.2  Ductwork Cost Methodology

     The cost algorithm for determining ductwork costs depends on
two key input variables:   airflow and duct length.  Each of these
parameters is process-specific and was chosen based on the given
model  plant   sizes.    This  section  describes  the  ductwork  cost
algorithm contained in Reference 1.  Therefore, the page numbers,
tables,  and   figures  cited  in  this  section  refer  to  the  page
numbers, tables,  and  figures  contained in Reference 1.   They are
italicized to avoid  confusion with tables  and figures appearing in
this ACT document.
     First,  a duct  velocity of 20.3 m/s   (4,000  fpm)  was chosen
based on the dust type to be captured (page 4-16).  Duct velocity
and airflow were then used to calculate the required diameter of
the ductwork  from  a given equation  (page 4-17).   Duct  diameter
determines the costs for  straight  steel duct, elbows, and dampers.
Using general ductwork assumptions from Table 5-1,  namely carbon
steel  construction  and  0.476 cm  (3/16  in.) duct  thickness,  the
straight steel duct,  elbow,  and damper costs were read from cost
curves given  in Figures 4.7,  4.10, and 4.14,  respectively.  Note
that the cost for straight steel  duct is  given as dollars/linear
foot and therefore must be multiplied by duct  length to calculate
the total straight steel duct cost.  Also, elbow and damper costs
are per  item  and must be multiplied by the  number  of elbows  and
dampers to determine their total costs.  The  sum of straight steel
                               5-3

-------
 TABLE 5-1.
EQUIPMENT PARAMETER  ASSUMPTIONS FOR  ALL FABRIC
FILTERS
 Equipment
                                     Assumption
 Ductwork
       Material of  construction
       Duct  thickness  (cm)
       Elbows
            Radius/diameter
            Angle  (degrees)
            Material of construction
       Pressure drop across ductwork (kPa)
       Dampers

 Fan/Motor
       System pressure drop  (kPa)
       Material of  construction
       Fan/motor efficiency  (%)
       Motor cover
       Starter
       Fan type
       Location of  fan
       Altitude of  plant  (m)
       Inlet/outlet dampers

 Baghouse
       Gas inlet temperature  (°C)
       Baghouse life/individual bag life  (yrs)
       Operator time  (hours/shift)
       Maintenance  time  (hours/shift)
       Pressure drop across baghouse (kPa)
       Bag material
       PM-10 removal efficiency  (%)
       Baghouse maintenance  (man-minutes/bag)
	Pulse-jet	
                                     Carbon  steel
                                     0.476

                                     1.5
                                     90
                                     Carbon  steel
                                     1.25
                                     Automatic control
                                      3.75
                                      Carbon  steel
                                      65
                                      Drip proof
                                      Magnetic
                                      Backward curved
                                      After the fabric filter
                                      0
                                      1 each
                                      66
                                      10/2
                                      8
                                      1-2
                                      2.5
                                      Polyester
                                      99

                                      10
                                   5-4

-------
duct,  elbow,  and  damper  costs  determines  the  total  purchased
equipment cost of ductwork.

5.2.3  Fan/Motor System Cost Methodology

     The cost algorithm for determining the fan/motor system cost
depends primarily on two key input variables:  airflow and system
pressure drop.   Airflow  is  a process-specific  parameter chosen
based on model plant size, while pressure  drop was assumed to be
the same for all process control systems  discussed in this document
(see Table 5-1).   This  section describes the fan/motor system cost
algorithm contained in Reference 1.   Therefore the page numbers,
tables,  and figures  cited  in this  section  refer  to  the  page
numbers, tables,  and figures  contained  in  Reference  1.   They are
italicized to avoid confusion with  tables and figures appearing in
this ACT document.
     First, a fan sizing  factor must be determined from Table 4.10
for the assumed airflow temperature and facility altitude.  Next,
the pressure drop is divided by the fan sizing factor to estimate
the  pressure  drop  at standard  conditions.    By combining  the
pressure drop at standard  conditions  with  airflow, the fan price
can be read from  Figure 4.33.   Also, inlet and  outlet damper costs
can be found in Figure 4.35 with the same inputs.
     Similarly,  the motor  and starter costs are  found  in Figure
4.34  as a  function of  airflow and  pressure  drop at  standard
conditions. The motor and starter costs, however,  require an extra
iteration.   When brake horsepower (bhp) is  found in Figure 4.34, it
must be multiplied by the fan sizing factor to determine the actual
bhp requirement.   Motor and starter costs are then found based on
the  actual bhp.    The  total  purchased equipment  cost  for  the
fan/motor system  is the sum of fan,  inlet and outlet damper, motor,
and starter costs.
                               5-5

-------
5.2.4  Baghouse Cost Methodology

     The cost algorithm for determining baghouse costs depends on
two key  input  variables:   airflow and  air-to-cloth (A/C)  ratio.
Each of these parameters is process-specific and was chosen based
on a given process model size.  This section describes the baghouse
cost algorithm contained  in  Reference 2.   Therefore,  the  page
numbers, tables, and  figures  cited in this section refer  to the
page numbers, tables,  and figures contained in Reference 2.  They
are italicized to avoid confusion with tables and figures appearing
in this ACT document.
     First,  an air-to-cloth ratio must be either assumed from Table
5.1 or calculated from equation 5.11.  Table 5.1 lists A/C ratios
for different dust  types based on the baghouse and filter material
used to  control a process.    For  iron compounds,   the A/C ratio
ranges from 2.0 to 11.0  feet/minute.   The actual  airflow is then
divided by the A/C ratio to determine the net cloth area required
of the filtering media.   Table 5.2  is a guide to estimating, based
on net cloth area, the gross cloth area required for  such things as
downtime and cleaning.
     Next,  the type  of  baghouse,  either shaker,   pulse-jet,  or
reverse air, must  be  determined.   The gross cloth area  is  then
applied to the appropriate cost curve given in Figures 5.2 through
5.7.  Any additional  costs,  such as those for a  stainless steel
frame  or  insulation add-on,  are  also  determined  from  the  cost
curves.  The costs for the baghouse were developed using the field-
construction curves in  the manual rather than  the  prefabricated
construction to reflect  the  likelihood of extra  costs associated
with custom fitting or retrofitting existing equipment.
     Finally, the  costs  for bags and bag mounting hardware must be
calculated.    Initially, bag  diameter and  type  of material  are
chosen from Table 5.7, which determines the cost per  square foot of
bag area.  Also, the type and cost of mounting hardware are found
in Table 5.7.  The  total  purchased  equipment cost of a baghouse is
the sum of baghouse, additions,  bags, and  mounting hardware costs.
                               5-6

-------
Polyester bags were  assumed for the examples  in  this study, but
Nomex bags are another typical bag type used in the industry.

5.2.5  Process-specific Equipment Assumptions and Equipment Costs

     This section contains the detailed process-specific parameters
that were used to determine  purchased equipment costs.  Tables 5-2
through  5-7  contain the  specific  equipment parameters  for each
process  control  option.   Also contained  in  the tables  are the
purchased equipment  costs that were derived  from  the  equipment
parameters.
     Table 5-2 contains  data for the first ACT  for coke pushing
["coke pushing (Al) "  in the table] ,  which utilizes a  shed vented to
a baghouse.   The shed serves as a partial  enclosure  of  the coke
pushing operation that reduces PM-10 emissions and lowers airflow
requirements.  No process-specific data are given in this section
for the  second ACT for coke  pushing  due to the unavailability of
such data from the published information used for mobile scrubber
car costs.3
     Table  5-3  lists  the  parameters  and  costs  for  the  coke
sizing/screening option.  This option utilizes a total enclosure of
the  sizing/screening area  vented  to  a  baghouse.    The  total
enclosure ensures high  capture efficiencies  while  keeping the
volumetric flow rate relatively low.
     Table  5-4 contains  data for  the  first  ACT  for  casthouse
emissions ["casthouse (Al) "  in the table], which is the evacuation
of the casthouse  to a baghouse.  Building evacuation captures PM-10
emissions from all areas of the casthouse, including the taphole,
troughs, and runners.  This type of control system requires large
airflows to control  large volumes of  space and  therefore large and
costly fans,  motors,  ducts, and baghouses.  Data for  the second ACT
for casthouse  emissions   ["casthouse  (A2)"  in the  table],  which
incorporate  local  hooding  to  a  baghouse,  are  contained  in
Table 5-5.  Local hoods  reduce  the  required airflow volumes from
building evacuation; however, they require  more  space around the
blast furnace and can be difficult to retrofit.

                               5-7

-------
TABLE 5-2.  CONTROL EQUIPMENT PARAMETERS
            FOR COKE PUSHING (Al) - SHED
AND PURCHASE COSTS
WITH BAGHOUSE

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for
temperature)
Baghouse
Gross cloth area required
(m2)
Air-to-cloth ratio (m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs
($1,000)*
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000)

Small


4
60.96
3.22
20.3
4
2

10
119
66
0.869

7,876
1.37
594

184
338
218
1,071
1,811
Plant Size
Medium


8
60.96
3.22
20.3
8
4

20
119
66
0.869

15,752
1.37
1,188

368
676
436
2,142
3,622

Large


15
60.96
3.22
20.3
15
7

37
119
66
0.869

28,945
1.37
2,183

676
1,242
801
3,936
6,655
Costs are  in April  1991 dollars.
                              5-8

-------
TABLE 5-3.
CONTROL EQUIPMENT PARAMETERS AND PURCHASE COSTS FOR
COKE SIZING/SCREENING - ENCLOSURE WITH BAGHOUSE

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for
temperature)
Baghouse
Gross cloth area required
(m2)
Air-to-cloth ratio
(m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs
($1,000)»
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000)

Small


1
22.86
0.54
15.2
1
1

3
13.4
66
0.869

413
1.37
17

8
9
17
399
433
Plant Size
Medium


2
22.86
0.77
15.2
2
1

3
26.8
66
0.869

619
1.37
34

15
21
24
418
478

Large


3
22.86
1.09
15.2
3
2

3
52.2
66
0.869

1,239
1.37
68

27
47
35
473
582
 Costs are in April 1991 dollars.
                              5-9

-------
  TABLE 5-4. CONTROL EQUIPMENT PARAMETERS AND PURCHASE COSTS FOR
             CASTHOUSE  (Al) - EVACUATION TO BAGHOOSE
Plant Size

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for temperature)
Baghouse
Gross cloth area required
(m2)
Air-to-cloth ratio (m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs ($1,000)*
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000) 1,
Small


3
22.86
2.98
20.3
3
2

10
89
66
0.869

6,813
1.37
509

160
128
183
975
446
Medium


4
22.86
4.21
20.3
4
2

10
179
66
0.869

13,006
1.37
1,019

296
260
382
1,533
2,471
Large


5
22.86
5.16
20.3
5
3

10
268
66
0.869

19,324
1.37
1,528

418
414
593
2,101
3,526
a Costs are in April  1991  dollars.
                               5-10

-------
TABLE 5-5.
CONTROL EQUIPMENT PARAMETERS AND PURCHASE COSTS
FOR CASTHOUSE (A2) - LOCAL HOODING AND BAGHOUSE
Plant Size

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for
temperature)
Baghouse
Gross cloth area required
(m2)
Air-to-cloth ratio (m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs ($1,000)*
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000)
Small


3
22.86
2.43
20.3
3
2

8
82
66
0.869

4,645
1.37
340

112
101
136
780
1,129
Medium


4
22.86
3.44
20.3
4
2

8
127
66
0.869

8,919
1.37
679

210
202
195
1,165
1,772
Large


5
22.86
4.21
20.3
5
3

8
179
66
0.869

13,006
1.37
1,019

296
325
306
1,533
2,460
a Costs are in April 1991 dollars.
                           5-11

-------
TABLE 5-6.  CONTROL EQUIPMENT PARAMETERS AND PURCHASE COSTS
            FOR HOT METAL TRANSFER - CANOPY HOOD WITH BAGHOUSE

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for temperature)
Baghouse
Gross cloth area required (m2)
Air-to-cloth ratio (m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs ($1,000)*
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000)
Plant Size
Small Medium


2 3
22.86 22.86
2.11 2.98
20.3 20.3
2 3
1 2

6 6
89 179
66 66
0.869 0.869

3,623 6,813
1.37 1.37
255 509

85 160
56 128
110 229
688 975
939 1,492

Large


3
22.86
2.98
20.3
3
2

6
179
66
0.869

6,813
1.37
509

160
128
229
975
1,492
a Costs are in April 1991 dollars.
                            5-12

-------
TABLE 5-7.  CONTROL EQUIPMENT PARAMETERS AND PURCHASE COSTS
            FOR DESULFURIZATZON - LADLE HOOD AND BAGHOUSE

Equipment Parameters
Ductwork
Number of takeoffs
Length per takeoff (m)
Diameter (m)
Duct velocity (m/sec)
Number of elbows
Number of autodampers
Fan/Motor
Number of motors
Motor size (kW)
Gas temperature (°C)
Fan sizing factor
(correction for temperature)
Baghouse
Gross cloth area required
(m2)
Air-to-cloth ratio (m/min)
Total flow rate (actual
thousand m3/h)
Equipment Purchase Costs ($1,000)*
Capture devices
Ductwork
Fan/motor system
Baghouse
Total Purchase Costs ($1,000)

Small


2
22.86
1.22
20.3
2
1

3
67
66
0.869

1,548
1.37
85

32
30
42
501
605
Plant Size
Medium


3
22.86
1.72
20.3
3
2

3
127
66
0.869

2,581
1.37
170

62
67
71
594
794

Large


3
22.86
1.72
20.3
3
2

3
127
66
0.869

2,581
1.37
170

62
67
71
594
794
a Costs are in April 1991 dollars.
                            5-13

-------
     Table 5-6  lists the parameters  for the hot  metal  transfer
control  option  of a  movable  canopy hood.   This  movable  canopy
reduces the required  airflow by allowing the  hood  to travel with
the  transfer  ladle,   thus  reducing the  emissions   coverage  area
versus a fixed canopy hood.
     Table 5-7 contains data for the desulfurization control option
of a close-fitting ladle hood  evacuated to a baghouse.  A ladle
hood is  ideal for capturing  emissions  from  the  desulfurization
process since  desulfurization takes place while the  hot metal is in
the ladle.
     All Tables 5-2 through 5-7  contain process-specif ic parameters
which were estimated  via other documents,  industry contacts,  and
engineering judgement.  Table 5-8 lists conversion  factors used to
convert the design parameters given in Tables 5-1 through 5-7 from
metric to English (inch-pound)  units.

                  TABLE  5-8.  CONVERSION  FACTORS
Original Unit
meters
meters
centimeters
square meters
meters per
second
cubic meters
per hour
liters
Megagrams
degrees Celsius
kilopascals
kilowatts
Multiplied by
3.281
39.370
0.394
10.764
196.9
0.589
0.264
1.102
9/5 + 32
4.015
1.341
Yields
feet
inches
inches
square ft .
feet per minute
cubic feet per
minute
gallons
short ton
degrees
Fahrenheit
inches of water
brake horsepower
     These  control  device  parameters  and  plant  sizes  may  be
different from any given  facility  in  the United States; however,
they were developed for cost comparison purposes only.  The costs

                               5-14

-------
given in Tables 5-2 through 5-7 are in April 1991 dollars and may
be updated with cost indices obtained from Chemical Engineering.4

5.3  BASIS FOR CAPITAL COSTS

     Capital costs represent the costs associated with purchasing
and  installing  new or  retrofitted equipment  (see  Section 5.2).
These costs are usually divided into three  categories:   (l)  the
base costs of purchasing the control equipment and some auxiliary
equipment;  (2) the costs of installing the equipment; and  (3)  the
indirect labor costs of  retrofitting  and testing the equipment.
Capital costs were estimated by obtaining base equipment costs from
EPA  cost  manuals  or   from   background  sources,   then  adding
installation and indirect costs derived from the base costs using
algorithms provided in EPA cost manuals.1'2  It was recognized that
retrofitting new  control equipment in an existing  plant is more
expensive than installing new equipment in a new plant.  Therefore,
the total retrofitted capital  costs are assumed to be equal to the
total new  capital  cost  of new  installation  plus a "retrofitting
effort" of 25 percent of the total new capital cost.5
      Table  5-9  lists  the  costs  and  cost  factors  derived from
Reference 2 for fabric filter capital and annual operating costs.
Capital cost factors are described here while annual cost factors
are discussed in  Section 5.4.   Capital  costs  were  determined by
summing  the  capital cost  factors and multiplying  by the total
purchased  equipment  cost.   The  total  purchased equipment cost
includes  the  costs  of equipment  (control  device,  fan/motor,
ductwork) and freight and sales tax.  The following equation shows
the calculation:

                   TCI =  ( EC + FST ) x Sum of CCF

      where TCI = total capital investment
             EC = equipment costs  (from Tables 5-2 through 5-7)
            FST = freight and sales tax
     Sum of CCF = sum of capital cost factors
                               5-15

-------
 TABLE  5-9   GENERAL COSTS AND COST FACTORS FOR FABRIC FILTERS8
 Parameter	Factor or Cost

 Capital Costs
       Factor of Equipment  Costs
             Freight and sales  tax                                0.08
       Direct Installation  Cost Factors
             Foundation and supports                               0.04
             Handling and erection                                0.35
             Electrical                                           0.08
             Piping                                               0.01
             Insulation for ductwork                               0.07
             Painting                                             0.02
       Indirect Installation Cost  Factors
             Engineering and supervision                          0.10
             Construction and field expenses                        0.20
             Contractor's fee                                     0.10
             Start-up and performance tests                        0.02
             Contingencies                                         0.03
 Annual Costs
       Direct Costs
             Electricity ($/kWh)                                   0.075
             Compressed air ($/m3/hour)                             0.094
             Operator labor rate  ($/hour)                         13.41
             Supervisor rate (fraction of operator                 0.15
             labor)
             Maintenance labor  rate ($/hour)                      14.75
             Material (fraction of maintenance                     1.00
             labor)
       Indirect Costs
             Annual interest rate  (%)                             10.0
             Overhead (fraction of operator,                       0.60
             supervisor,  maintenance and material
             costs)
             Property tax (factor  of TCI)                          0.01
             Insurance (factor  of  TCI)                             0.01
	Administration (factor of TCI)	0 .02
a See Reference 3  for mobile scrubber car cost factors.
                                  5-16

-------
The estimated  total and  retrofit  capital costs  for  the process
control options are given in Tables 5-10 through 5-15.

5.4  BASIS FOR ANNUAL COST ESTIMATES

     In  contrast to  capital  costs,  which  represent  the  costs
associated with  purchasing  and installing new equipment, annual
costs represent yearly disbursements for operating and maintaining
the control systems and the annualized costs of capital recovery.
Annual costs are usually  divided  into direct and indirect annual
costs.   Direct annual costs include  utilities,  operating labor,
maintenance labor,  maintenance material,  dust disposal and sludge
disposal.  Indirect annual  costs  include  overhead,  property tax,
insurance, general administration and annualized capital recovery
charges.  Annual costs were  derived by using  cost algorithms given
in Reference 2 for fabric filters.
     Table 5-9 lists the costs and cost factors used to determine
annual operating costs for fabric  filters.  As shown, annual costs
are the sum of direct and indirect costs.   The following equation
shows the calculation for total annual costs:
                 TAC = UT + OML  + RP + OH + CR + GAC

     where TAC = total annual cost
            UT = utilities cost
           OML = operating and maintenance labor and materials
            RP = replacement parts
            OH = overhead (60 percent of OML)
            CR = capital recovery
           GAC = general administrative costs (property tax, etc.)

Capital recovery costs  were based on a 10 percent  interest rate
with a 10-year equipment  life.  The annual  costs  for the process
control  options,  assuming  capital equipment  expenditures on  a
retrofit basis,  are given  in  Tables  5-17 through  5-22.   Annual
costs published  in  Reference 3 for the mobile  scrubber car were
given in a different format than those of the fabric filter given
in Reference  2,  and  appear separately in  Table  5-16.   Capital
recovery costs  given  in Table 5-16 are  estimates  of  the annual

                               5-17

-------
  TABLE 5-10.  TOTAL CAPITAL COSTS  FOR COKE PUSHING  (Al)  ($1,000)
Plant Size

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
Small

1,811
145
1,956

78
685
156
20
137
39
1,115

196
391
196
39
59
881
3,952
4,940
Medium

3,622
290
3,912

156
1,369
313
39
274
78
2,229

391
782
391
78
117
1,759
7,900
9,875
Large

6,655
532
7,187

288
2,516
575
72
503
144
4,098

719
1,438
719
144
216
3,236
14,521
18,151
a Costs are in April 1991 dollars.
b For retrofit capital costs multiply new capital costs by 1.25.
                                  5-18

-------
           TABLE 5-11.  TOTAL CAPITAL COSTS  FOR COKE
                        SIZING/SCREENING  ($1,000)*
Plant Size

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
Small

433
35
468

19
164
37
5
33
9
267

47
94
47
9
14
211
946
1,182
Medium

478
38
516

21
181
41
5
36
10
294

52
103
52
10
15
232
1,042
1,303
Large

582
47
629

25
220
50
6
44
13
358

63
126
63
13
19
284
1,271
1,589
a  Costs are in April 1991 dollars.
b  For retrofit capital  costs multiply new capital costs by 1.25.
                                 5-19

-------
  TABLE  5-12. TOTAL CAPITAL COSTS FOR CASTHOUSE  (Al)  ($1,000)*
Plant Size

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
Small

1,446
116
1,562

62
547
125
16
109
31
890

156
312
156
31
47
702
3,154
3,943
Medium

2,471
198
2,669

107
934
214
27
187
53
1,522

267
534
267
53
80
1,201
5,392
6,740
Large

3,526
282
3,808

152
1,333
305
38
267
76
2,171

381
762
381
76
114
1,714
7,693
9, 616
a Costs are in April 1991 dollars.
b For retrofit capital costs multiply new
capital costs by 1.25.
                                  5-20

-------
  TABLE  5-13. TOTAL CAPITAL COSTS FOR CASTHOUSE  (A2)  ($1,000)*
Plant Size

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
Small

1,129
90
1,219

49
427
98
12
85
24
695

122
244
122
24
37
549
2,463
3,079
Medium

1,772
142
1,914

77
670
153
19
134
38
1,091

191
383
191
38
57
860
3,865
4,831
Large

2,460
197
2,657

106
930
213
27
186
53
1,515

266
531
266
53
80
1,196
5,368
6,710
a Costs are in April 1991 dollars.
b For retrofit capital costs multiply new capital costs by 1.25.
                                  5-21

-------
TABLE 5-14. TOTAL CAPITAL COSTS FOR HOT METAL TRANSFER
            ($l,000)a

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
a Costs are in April 1991 dollars.
b For retrofit capital costs multiply new

Small

939
75
1,014

41
355
81
10
71
20
578

101
203
101
20
30
455
2,047
2,559
capital costs
Plant Size
Medium

1,492
119
1,611

64
564
129
16
113
32
918

161
322
161
32
48
724
3,253
4,066
by 1.25.

Large

1,492
119
1,611

64
564
129
16
113
32
918

161
322
161
32
48
724
3,253
4,066

                       5-22

-------
  TABLE 5-15. TOTAL CAPITAL COSTS FOR DESULFURIZATION ($1,000)*
Plant Size

Purchased Equipment Costs
Removal & auxiliary equipment
Freight & sales taxes
Subtotal
Direct Installation Costs
Foundation & supports
Handling & erection
Electrical
Piping
Insulation for ductwork
Painting
Subtotal
Indirect Costs
Engineering & supervision
Construction & field expenses
Contractors fee
Start-up and performance test
Contingencies
Subtotal
Total New Capital Costs
Total Retrofit Capital Costsb
Small

605
48
653

26
229
52
7
46
13
373

65
131
65
13
20
294
1,320
1,650
Medium

794
64
858

34
300
69
9
60
17
489

86
172
86
17
26
387
1,734 1,
2,168 2,
Large

794
64
858

34
300
69
9
60
17
489

86
172
86
17
26
387
734
168
a Costs are in April 1991  dollars.
b For retrofit capital costs multiply new capital costs by 1.25.
                                 5-23

-------
        TABLE 5-16.  TOTAL ANNUAL COSTS FOR COKE PUSHING
                      (A2)  ($l,000)a
Plant Size

Direct Costs
Power cost
Heat
Maintenance
Subtotal
Indirect Costs
Capital recovery
Subtotal
Total Annual Costs
Small

788
787
332
1,907

500
500
2,407
Medium

1,185
1,177
498
2,860

1,000
1,000
3,860
Large

1,973
1,968
830
4,771

1,250
1,250
6,018
a Costs are in April 1991 dollars.
maintenance required to successfully operate a mobile scrubber car
already in use.   These costs do not represent the capital recovery
of an investment in  a  new mobile  scrubber  car  system.   The costs
given  in Reference  3  were  indexed  from  1974  dollars to  1991
dollars.   It is  recognized that  this  method  may not  yield an
accurate  representation  of  current annual costs;  however,  other
references were  not readily available.   Agencies  interested in
conducting a more detailed estimate of  mobile  scrubber car costs
should consult with an air pollution control equipment vendor.

5.5  COST EFFECTIVENESS

     Cost effectiveness is defined as the total annual costs per Mg
of PM-10 emissions reduced.  The cost effectiveness is calculated
by dividing  the incremental cost  of  implementing  a  new process
control   option   (the  annualized   cost   of   control   system
modifications) by the  additional mass  of PM-10  removed by the new

                               5-24

-------
    TABLE 5-17. TOTAL ANNUAL COSTS FOR COKE PUSHING  (Al)
                 ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

682
135
129
129
29
1,104

236
49
49
99
796
1,229
2,333
Medium

1,364
270
258
258
58
2,208

472
98
98
198
1,592
2,458
4,666
Large

2,507
496
474
474
107
4,058

867
180
180
364
2,926
4,517
8,575
a Costs are in April 1991 dollars.
                                5-25

-------
 TABLE 5-18.  TOTAL ANNUAL COSTS FOR COKE SIZING/SCREENING
              ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

20
34
16
16
1
87

40
12
12
24
192
280
367
Medium

39
34
16
16
2
107

40
13
13
26
212
304
411
Large

78
34
16
16
5
149

40
16
16
32
257
361
510
a Costs are in April 1991 dollars.
                                5-26

-------
  TABLE 5-19. TOTAL ANNUAL COSTS FOR  CASTHOUSE (Al)  ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

584
135
129
129
25
1,002

236
39
39
79
635
1,028
2,030
Medium

1,169
135
129
129
47
1,609

236
67
67
135
1,083
1,588
3,197
Large

1,753
135
129
129
71
2,217

236
96
96
192
1,545
2,165
4,382
a Costs are in April 1991 dollars.
                                5-27

-------
       TABLE 5-20. TOTAL ANNUAL COSTS FOR CASTHODSE  (A2)
                    ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

400
135
129
129
17
810

236
31
31
62
496
856
1,666
Medium

779
135
129
129
32
1,204

236
48
48
97
111
1,206
2,410
Large

1,169
135
129
129
47
1,609

236
67
67
134
1,078
1,582
3,191
a Costs are in April 1991 dollars.
                                5-28

-------
        TABLE 5-21. TOTAL ANNUAL COSTS FOR HOT METAL TRANSFER
                     ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

292
135
129
129
13
698

236
26
26
51
413
752
1,450
Medium

584
135
129
129
25
1,002

236
41
41
81
655
1,054
2,056
Large

584
135
129
129
25
1,002

236
41
41
81
655
1,054
2,056
* Costs are in April 1991 dollars.
                                5-29

-------
TABLE 5-22. TOTAL ANNUAL COSTS FOR DESULPURIZATION ($1,000)*
Plant Size

Direct Costs
Utilities
Operating labor
Maintenance
Replacement parts
Bags
Subtotal
Indirect Costs
Overhead
Property tax
Insurance
Administration
Capital recovery
Subtotal
Total Annual Costs
Small

97
135
129
129
6
496

236
17
17
33
267
570
1,066
Medium

195
135
129
129
9
597

236
22
22
43
350
673
1,270
Large

195
135
129
129
9
597

236
22
22
43
350
673
1,270
a Costs are in April 1991 dollars.
                             5-30

-------
process control option  (amount of emissions reduced from baseline
emissions).   The cost effectiveness for each process control option
and model plant site is listed in Table 5-23.

5.6  CONTROL OPTION COSTS FROM INDUSTRY REPRESENTATIVES

     The  cost  comparisons  in  this  section  are  provided  to
illustrate the variability in control option cost estimates for the
iron and steel industry.  These comparative costs were provided to
EPA by  industry representatives  who reviewed a draft  version of
this document.  The discussion here  focuses on reviewers' comments
on  the  methodology  used  to  derive  control  option  costs  in
Sections 5.2,  5.3,  and  5.4 of  this document and  their own field
experience with costing these control options.
     The costing methodologies  (derived from EPA guidance) used in
this  chapter were  sound  according  to one  reviewer's  comments;
however,  certain   parameters   used '  in   the   methodology  were
questioned.   For example,  the assumed control system pressure drop
of 15 inches  H^O was stated to  be low as  pressure  drops  on some
systems often exceed 20 inches  H2O.  This same reviewer questioned
the airflow  rates  that were  assumed  for  each process  size and
control option. The airflow rates have since been revised to be in
line  with the  reviewers'   comments;  however,  the  control  system
pressure drop  assumption  remains  at 15 inches  water column.   The
reviewer  also  stated   that the  assumed  air-to-cloth ratio  of
4.5 feet/minute in Tables 5-2  through 5-7 was somewhat  low for
modern baghouse designs; however,  this  value was left unchanged in
the costing methodology.6   (The  effect of changing the air-to-cloth
ratio is relatively small.)  Changes in these parameters noted by
reviewers could either increase  or decrease the  capital and annual
costs derived  in this document.
     Other  comments  on  the  parameters  used  in  the  costing
methodology focused on various capital and annual cost factors that
are given in Table 5-9.  General comments were that  Freight and
Sales Tax (8%) , Foundation and Supports  (4%) , and  the  assumed
annual interest rate  (10%)  were too low for most  iron and steel

                              5-31

-------
  TABLE 5-23.  COST EFFECTIVENESS FOR EMISSIONS  REDUCTIONS FROM
               UNCONTROLLED CASE
PM-10
Reduction
Annual Coats from
($1,000)' Uncontrolled Cost Eff.
(Mg/yr) ($l,000/Mg)
Coke Pushing (Al)
Small
Medium
Large
Coke Pushing (A2)
Small
Medium
Large
Coke Sizing/Screening
Small
Medium
Large
Casthouse (Al)
Small
Medium
Large
Casthouse (A2)
Small
Medium
Large
Hot Metal Transfer"
Small
Medium
Large
Desulfurizationb
Small
Medium
Large

2,333
4,666
8,575

2,407
3,860
6,018

367
411
510

2,030
3,197
4,382

1,666
2,410
3,191

1,450
2,056
2,056

1,066
1,270
1,270

230
450
830

200
410
750

39
79
150

150
300
550

130
270
490

39
78
140

120
230
420

10
10
10

12
9.5
8.0

9.3
5.2
3.5

14
11
8.0

12
9.0
6.5

37
26
14

9.3
5.5
3.0
'  Costs are in April 1991 dollars.
b  Control options for desulfurization and hot metal transfer are considered to
  be baseline.
                                 5-32

-------
applications.  One reviewer stated that specifically for the Coke
Pushing  (Al) control  option  (shed with baghouse),  Foundation and
Supports are very expensive due to the massive structure associated
with a  shed.   Other specific comments  were  that  Engineering and
Supervision costs are more likely near 20 percent rather than the
10 percent  factor  used in the  calculations.   Also  the reviewer
stated that contingency costs may be at least 10 percent as opposed
to the  3 percent factor used  in  the methodology.   Finally,  the
reviewer    stated    that    from    their   experience,    annual
Maintenance/Replacement costs typically run at 10 percent of each
facility's  total annual replacement  costs.   However,  the costing
methodology assumed that replacement costs equal 100 percent of the
maintenance  costs  plus  fabric filter  bag purchases.    For  one
example given in Table 5-17,  coke  pushing  (Al), a small facility's
maintenance/replacement  costs  would  be  $1.3 million  in  the
reviewer's example; however,  these costs sum to $287,000 using the
OAQPS costing methodology.6
     Several industry  reviewers provided  comments  relating their
actual costing experience to the  control  costs estimated in this
chapter.6-7'8  These comments were consolidated and are presented in
Table 5-24.  Nearly all the  cost  comments provided by reviewer's
related their own retrofit capital cost experience instead of total
annual costs.  These  reviewer's costs were up to  4 times greater
than the retrofit capital costs estimated in this chapter.  These
differences  may  result  from  several  factors,   including  the
differences in  capital cost estimating factors  discussed earlier in
this section and the  "study" estimate error  inherent  in the cost
methodology discussed  in Section  5.2.   Also,  as  noted earlier in
this document,  several site-specific parameters  may  impact  the
control  costs  derived in  this  chapter.   These  costs  may include
building  and site  preparation  costs,  stack  costs,   or  downtime
losses.   It is therefore re-emphasized that  the  costs presented
here are provided  for  comparative  purposes  only;  evaluation of
actual facility  costs should be  developed specifically  for each
individual site.
                               5-33

-------
CO


H





CO
CO
D
CM
in

I


**H
41
DC


n
IH 4)
<0 3
"-CH
-H "0
ot>

10 -H
OEH
-,
au
c °
i5 *"^
^«

~t
HJ
1
G
•u n
*3 ^
IH 0
•UO
41
«
4J
O


s

r"3
VJ
•U
C
(H

Q,





O
•*H
I

fi
4J
0
VJ


4)
0
3
0
O
1
VO
4)
j}
3


S

-H
"2



Q

1





•H
•H
01
*








^
1
to



-
*^J
&
q
•-»
43
3
Q.
01
•x
3


_
0



*





^a






i
i

c
o
^J
"H
g
S
o
o

O
o\
,H
u
w
§<




|
'




q
3
•M
en
»-l








_
1
10



-
iq;
tf.
c
•H
n
s
a
41
-V
3


4)
to
m



vo




<*»•
CM
S





i
i




c
o
-H
•H
O

CM

q-u
10 C
43 10
•U—1
0.
IH
tn3
IH <0
10 E
*U
CM«<



-
^
&.
c
•H
3
a.
4)

3



to
IQ
vo
1
VO
41
3
EH


Ol
01
H
*
c
0
*H
^fif-H
• •H
CM E

















„,
10
to



CM

&
c
v>
3
Q.
41

O
O



EH
(J



VO





U
*

q
O
^1
(*)>-H
••H
"e




Oi
C
o
•H
•H
E
O
^
C
43 c;
•u H 0,
41
••H'-l
10 "0
E E
S°




CM
•»;
o
c
•H
43
3
a
01
•X
o
o



o
to



^





n









10
to





•H
^»
01
ID
3
O
43
4J
V)
3



EH
o
41
-Q



CJ\

r, *
*




1 1
1 1





C C
0 0
€ E
"H 0
0








§
"S
*



4)
<0

•u
"«
u
0)

4J
§



tj
"O
1

0. «»
X)
s


o\
o\
., *
S




i i
I i





q q
O 0
E E
CM o
CM "»








1 i
•o -o
* *




C q
O 0
u u
10 t
N N
•H "H
3 3
"H "H
3 3
v> n
4) 01
Q Q



EH (J
u to
                                                                         CN
                                                                         O(
                                                                         41
                                                                         X)
                                                                   O-u
                                                                   «g
                                                                   *> §
                                                                   •H E
                                                                   43 O
                                                                    R«J
                                                                            43
                                                                            u

                                                                            c
                                                                            •H

                                                                            4-1
                                                                            q
                                                                            10
                                                                             41
                                                                            TI
                                                                         «^
                                                                         3   "
                                                                         S   S
                                                                     IH OITJ
                                                                    4) D-u q
                                                                   ->l 4 3-M
                                                                   -Q 4J-U
                                                                    "0-H-^-u
                                                                   •u ^ C
                                                                   T3 Inq
                                                                    Oi  >-. »
                                                                    U q  <0
                                                                    q-H »> 4
                                                                    41  -H
                                                                    IntJ 10 «1
                                                                    41 41 O-u
                                                                   •in »)HH w
                                                                    41 

                                                                             S


                                                                             O
pushing
                                                                            .
                                                                    0 OCJ   fi
                                                                    M C  IH O
                                                                       -
                                                                                •u n
                                                                                q <0
                                                                                i3 41
                                                                                «i.q
                                                                                "•5
                                                                                •043
                                                                                u-u
«a 10
S5
S-.
•H *
q y
                                                                                o c
                                                                                 "» IH
                                                                                •U m
                                                                                41 u
                                                                                £•*
                                                                                4t u
                                                                                43 iq

                                                                                •oS-
                                                                                PiS
                                                                                      O
                                                                                      41
IS,
o q
O--I
                                                                                      O ^j
                                                                                        IH
                                                                                        01
                                                                                      O IH
                                                                                      -2B
                                                                                 :s   «5
                                                                                      o 2 o
      w >^ -»i
      ^«s-
      S2S
          Cn
      VDO?

      »ol
_i   - SOT3
C q-u *tO 3
S Cr>q njOl^
g-asti  .s
a-gSgq^
                                                                    L, 4)  01 « 10
                                                                    •0 S 41 3 41 Q.

                                                                    »1 » O 0

               I
               O
              .0
                                                                                              §3

                                                                                              5§
                                                                                              O *

                                                                                              tJ 4)
                                                                                              0»)
                                                                                              •U-H
             41-H

             oS
             S5
             cS
             8*
             O q,
               43
        C 01
        O O
         flj
        S5
        15

       *3
       ^^5
        4-fl
        OI L,

        :?
                                                                                             ,

                                                                                        3a  Is
                                                                                      43 Q 41 C 43 to
                                                                                     flEnO Q,H EH O) «
                                                    5-34

-------
5.7  REFERENCES FOR CHAPTER 5


1.   Neveril, R.B.   Capital and Operating  Coats  of Selected Air
     Pollution Control Systems.  EPA 450/5-80-002.  Office of Air
     Quality Planning and Standards, U.S. Environmental Protection
     Agency, Research Triangle Park, NC.  December 1978.

2.   U.S. Environmental  Protection Agency.   OAQPS  Control  Cost
     Manual.  Fourth  Edition.   EPA 450/3-90-006.   Office of Air
     Quality Planning and Standards, U.S. Environmental Protection
     Agency, Research Triangle Park, NC.  January 1990.

3.   McClelland,  R.O.  Coke Oven  Smokeless Pushing System Design
     Manual.   EPA 650/2-74-076.   U.S.  Environmental  Protection
     Agency, September 1974.

4.   "Chemical Engineering  Plant Cost  Index and Marshall & Swift
     Equipment Cost  Index,"  in Chemical  Engineering,  Vol.  98,
     No. 7.   July 1991.

5.   U.S. Environmental  Protection  Agency.   Control Technologies
     for Hazardous Air Pollutants.   EPA 625/6-86-014.   Air and
     Energy  Engineering  Research Laboratory, U.S.  Environmental
     Protection Agency, Research Triangle Park,  NC.  June 1986.

6.   Letter from Thomas W. Easterly, Bethlehem Steel Corporation,
     to James H.  Maysilles, U.S. Environmental Protection Agency.
     Comments on draft ACT document for PM-10 emissions from iron
     and steel mills.  Letter dated August 19, 1992.

7.   Letter from R.  Wade Kohlmann,  Indianapolis Coke, to James H.
     Maysilles,  U.S.  Environmental  Protection Agency.  Comments on
     draft ACT document  for PM-10 emissions  from iron and steel
     mills.   Letter dated August 4, 1992.

8.   Letter from David C. Ailor, American Coke and Coal Chemicals
     Institute,   to   James  H.  Maysilles,   U.S.  Environmental
     Protection Agency.  Comments on draft ACT document for PM-10
     emissions  from   iron  and  steel  mills.     Letter  dated
     September 25,  1992.
                               5-35

-------
          APPENDIX A
INTEGRATED IRON AND STEEL MILLS
 AND BLAST FURNACE  COKE  OVENS
              A-l

-------
      TABLE A-l.
INTEGRATED IRON AND
FURNACE COKE OVENS
STEEL MILLS AND BLAST
             Plant
                           Location
 Acme Steel
 Armco Inc.
 Armco Inc.
 Bethlehem Steel
 Bethlehem Steel
 Bethlehem Steel
 Bethlehem Steel
 Geneva Steel
 Gulf States Steel
 Inland Steel
 LTV Steel
 LTV Steel
 LTV Steel
 LTV Steel
 McClouth Steel
 National Steel
 National Steel
 Rouge Steel
 Sharon Steel
 USS Division of USX
 USS Division of USX
 USS Division of USX
 USS/Kobe Steel
 Warren Consolidated
 Industries
 Weirton Steel
 Wheeling-Pittsburgh  Steel
 Wheeling-Pittsburgh  Steel
                       Chicago,  IL
                       Middletown,  OH
                       Ashland,  KY
                       Bethlehem,  PA
                       Burns Harbor,  IN
                       Lackawanna,  NY
                       Sparrows  Point,  MD
                       Provo,  UT
                       Gadsden,  AL
                       East Chicago,  IL
                       Cleveland,  OH
                       Pittsburgh,  PA
                       Indiana Harbor,  IN
                       East Chicago,  IN
                       Trenton,  MI
                       Ecorse, MI
                       Granite City,  IL
                       Dearborn, MI
                       Parrel1,  PA
                       Fairfield,  AL
                       Clairton, PA
                       Gary, IN
                       Lorain, OH

                       Warren, OH
                       Weirton,  WV
                       Mingo Junction,  OH
                       East  Steubenville, WV
References:
U.S. Environmental Protection Agency.  Cost Analysis for the Coke Oven NESHAP.
Office of Air Quality Planning and Standards, Research Triangle Park, NC, April
1992.
"Blast  Furnace Roundup."   33 Metal Producing, Volume 29,  No.  5.  Penton
Publishing, Inc.,  Cleveland, OH.  May 1991.   Inc., Cleveland,  OH.  May 1991.
                                A-2

-------
            APPENDIX B
SAMPLE EMISSION FACTOR CALCULATIONS
                B-l

-------
Example Calculation for Derived Emission Factors

If controlled emission factors were not available from published texts,
an emission factor was derived using the following methodology:

1) Calculate fugitive emission factor (emissions not captured at the source) using:
           (a)  controlled emission factor from published source
           (b)  assumed capture system efficiency

2) Calculate process emission factor (emissions released at control device outlet) using:
           (a)  assumed minimum exit grain loading
           (b)  assumed inlet temperature
           (c)  assumed hours of operation
           (d)  assumed flowrate (for each size model plant)
           (e)  assumed throughput (for each size model plant)
           (f) if flowrates and throughputs are not linearly proportionate, use the average
              process emission factor for step 3

3) Add fugitive  emission factor to the average process emission factor to obtain total emission factor.


The following example illustrates the use of this procedure to generate
a controlled emission factor for the hot metal transfer process.

1) Calculate fugitive emission factor:
           (a)  controlled emission factor	0.09 Ib/ton

           (b)  assumed capture efficiency	90 percent

           Fugitive e.f.  =   0.09 Ib PM-10 emitted        x  (100%-90%)
                               ton hot metal produced

           Fugitive e.f.  =           0.009 Ib PM-10 emitted
                                        ton hot metal produced
                                                          B-2

-------
2) Calculate process emission factor:
           (a) assumed minimum exit grain loading (assume all is PM-10)..   0.003 gr/dscf

           (b) assumed inlet temperature	100 C

           (c) assumed hours of operation	8,760

           (d) assumed flowrate for each size model plant (acfm)	small              150,000
                                                                        medium            300,000
                                                                        large              300,000

           (e) assumed throughput for each size model plant (tons/yr)	  small            1,316,276
                                                                        medium          2,632,552
                                                                        large            4,856,604

           (f) calculate average process emission factor

           Process e.f. =  150,000 acfm x 273 C x    0.003 gr  x 60 min  x  8,760 hr  x      year       x     115
           (small)                       373 C         dscf       hr          year   1,316,276 tons     7,000 gr

             Process e.f. =        0.0188  Ib PM-10 emitted
             (small)                     ton hot metal produced

           Process e.f. =  300,000 acfm x 273 C x    0.003 gr x 60 min  x   8,760 hr  x       year       x     1 Ib
           (medium)                     373 C         dscf       hr          year   2,632,552 tons     7,000 gr

             Process e.f. =        0.0188 Ib PM-10 emitted
             (medium)                   ton hot metal produced

           Process e.f. =  300,000 acfm x 273 C x    0.003 gr x 60 min  x   8,760 hr  x       year       x     1 Ib
           (large)                       373 C         dscf       hr          year   4,856,604 tons     7,000 gr

             Process e.f. =        0.0102  Ib PM-10 emitted
             (large)                      ton hot metal produced

           Process e.f. =          (0.0188 + 0.0188 + 0.0102) / 3
           (average)

             Process e.f. =        0.0159  Ib PM-10 emitted
             (average)                   ton hot metal produced

3) Calculate total  emission factor (fugitive emission factor plus average process emission factor):


           Total e.f. =    Fugitive e.f. +  Process e.f. (avg)

           Total e.f. =      0.009   +    0.0159

           Total e.f. =             0.249  Ib PM-10 emitted
                                        ton hot metal produced
                           or
                                  0.125  kg PM-10 emitted
                                        Mg hot metal produced

Note that derived emission factor values in Table 2-3 and Table 4-1 are reported with only one significant figure.
                                                     B-3

-------
           REPORT DOCUMENTATION PAGE
                                       rOr.T ~~c :,. -rC

                                       OMB No 3704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existmq aaia sources
gathering and maintaining the data needed, and completing and reviewing the collection of information Send comments regarding this burden estimate c" anv otner aspect of tn:
collection of information, including suggestions for reducing this burden, to Washington Headquarters Services. Directorate tor information Operations and Reports 1215 Jefterso
Davis Highway. Suite 1204, Arlington. VA 22202-4302. and to the Off ice of Management and Budget. Paperwork Reduction Project (0704-0188). Washington DC 20SQ3
 1. AGENCY USE ONLY (Leave blank)
2. REPORT DATE
   February
                                      1994
3. REPORT TYPE AND DATES COVERED
   Final
4. TITLE AND SUBTITLE
 Alternative Control Techniques Document  --
 PM-10 Emissions  From Selected Processes  at
 Coke Ovens  and Integrated Iron and  Steel  Mills
6. AUTHOR(S)
 Philindo J.  Marsosudiro and W.  Scott Snow
                                 5. FUNDING NUMBERS

                                     68-DO-0121
                                     WA 2/117
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
 TRC  Environmental  Corporation
 100  Europa  Drive,  Suite 150
 Chapel Hill,  NC   27514
                                 8. PERFORMING ORGANIZATION
                                   REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
 U.S.  Environmental Protection Agency
 Emission Standards Division  (MD-13)
 Office of Air Quality  Planning and  Standards
 Research Triangle Park,  NC  27711
                                 10. SPONSORING / MONITORING
                                   AGENCY REPORT NUMBER

                                  EPA-453/R-94-005
11. SUPPLEMENTARY NOTES
 EPA Work Assignment Manager:
           James  H. Maysilles    919-541-3265
12a. DISTRIBUTION/AVAILABILITY STATEMENT

 Release unlimited
                                 12b. DISTRIBUTION CODE
13. ABSTRACT (Maximum 200 words)
 The  purpose of this document  is to  provide guidance  to state and loc
 air  quality management  agencies for determining reasonably available
 control technologies  (RACT) and best available  control technologies
 (BACT) that apply to  PM-10  sources  in the iron  and steel  industry.
 Emission  sources  addressed  are coke pushing, coke quenching, coke
 sizing and screening, casthouse operations, hot metal transfer,  and
 desulfurization.   These sources were selected for analysis because
 they are  not presently  regulated under New Source Performance
 Standards (NSPS).   Emission control system descriptions,  environment
 and  energy impact assessments, and  control cost analyses  are
 presented.  The principal emission  collection devices used are  the
 fabric filter and wet venturi scrubber.
14. SUBJECT TERMS
Iron production Iron and steel industry Cost of
Steel production Fabric filter PM-10
Coke oroduction Wet venturi scrubber control
17. SECURITY CLASSIFICATION
OF REPORT
UNCLASSIFIED
18. SECURITY CLASSIFICATION
OF THIS PAGE
UNCLASSIFIED
19. SECURITY CLASSIFICATION
OF ABSTRACT
UNCLASSIFIED
15. NUMBER OF PAGES
128
16. PRICE CODE
20. LIMITATION OF ABSTF
UL
NSN 7540-01-280-5500
                                     Standard Form 298 (Rev
                                     Prescnoed by ANSI Std  Z39-18
                                     298-102
                                                                              2-!

-------