United States
         Environmental Protection
         Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-453/R-94-065
Final Report
September 1994
&EPA  ALTERNATIVE CONTROL

         TECHNIQUES DOCUMENT--

         NO,, EMISSIONS FROM
         IRON AND STEEL MILLS

-------

-------
                                             EPA-453/R-94-065
             ALTERNATIVE CONTROL
           TECHNIQUES DOCUMENT-
NOX EMISSIONS FROM 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
                     September 1994

-------

-------
             ALTERNATIVE CONTROL TECHNIQUES DOCUMENTS
     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
pollution control 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 last-- from the Library Services Office (MD-35),' U. S.
Environmental Protection Agency, Research Triangle Park, North
Carolina 27711 ([919]  541-5514)  or, for a nominal fee, from the
National Technical Information Service, 5285 Port Royal Road,
Springfield, VA 22161   ([800] 553-NTIS).

-------

-------
                        TABLE OF CONTENTS
Chapter
Page
          LIST OF FIGURES	  iv
          LIST OF TABLES	• . '  v

     1    INTRODUCTION	 1-1

     2    SUMMARY ............ 	 2-1
          2.1  SOURCES OF NOX EMISSIONS . .  .  .  .  .  ..  . .  2-1
          2.2  UNCONTROLLED NOX EMISSIONS	2-3
          2.3  NOX EMISSIONS REDUCTION TECHNIQUES  .  .  .  . .2-4
          2.4  COSTS AND COST EFFECTIVENESS OF NOX CONTROL
               TECHNIQUES 	....2-8
          2.5  IMPACTS OF NOX CONTROLS  .	2-8
               2.5.1  Air Impacts of Combustion
                      Modification Techniques   .  . .  .   . . 2-12
               2.5.2  Air Impacts of SCR and SNCR ..... 2-13
               2.5.3  Solid Waste Impacts .	2-13
               2.5.4  Energy Impacts	2-14

     3    INDUSTRY DESCRIPTION  . .	3-1
          3.1  BACKGROUND	 3-1
               3.1.1  A Historical Perspective  .  .  .  .   . . 3-2
               3.1.2  An Overview of the U.S.  Iron
                      and Steel Industry	   . .3-2
                      3.1.2.1  Integrated Producers .  .   . .3-5
                      3.1.2.2  Mini-Mills	3-5
                      3.1.2.3  Specialty Producers  .  .   . .  3-5
          3.2  INDUSTRY CHARACTERIZATION  .........  3 - 8
               3.2.1  An Overview of Iron and Steel
                      Manufacturing	  3-8
                      3.2.1.1  Raw Materials and Preparatory
                               Processes	3-8
                      3.2.1.2  Ironmaking .  .  .  .  .  .  .  . .  3-10
                      3.2.1.3  Steelmaking  .	3-11
                      3.2.1.4  Finishing  	  3-14
          3.3  PRODUCTION PROCESSES  WITH NOX EMISSIONS .  .   3-15
               3.3.1  Sintering  	  .....   .3-15
               3.3.2  Cokemaking  .....  	  3-17
               3.3.3  Blast  Furnaces and Blast Furnace
                      Stoves  .	•  .  .   3-18
               3.3.4  Basic  Oxygen Furnaces  .  ...  .  . '.  .  . .3-19
               3.3.5  Soaking-Pit and Reheat Furnaces    . .  3-20
               3.3.6  Annealing Furnaces  .  .  .	3-23
          3.4  REFERENCES		3-24

     4    UNCONTROLLED NOX EMISSIONS	.4-1
          4.1  MECHANISMS  OF NOX  FORMATION	4-1

-------
                TABLE OF CONTENTS (con.)

     4.1.1  Thermal NOX Formation	   4-1
     4.1.2  Fuel NOX Formation   ." .  .  .  .  .  .  .  .  .  .  4-4
     4.1.3  Prompt NOX Formation   ..........  4-6
     4.2  EMISSIONS OF NOX FROM IRON AND STEEL MILLS .  4-7
          4.2.1  Emission Factors  . . . . . . . . . .4-7
          4.2.2  Coke-oven Underfiring . . . . . . . .4-7
          4.2.3  Sintering .  .  . . . . . . . . . . . .4-9
          4.2.4  Blast Furnaces and Blast
                 Furnace Stoves	 . . . . .4-10
          4.2.5  Basic Oxygen Furnace  . . . . . . . .4-10
          4.2.6  Electric-arc Furnace  . . . . . . . .4-11
          4.2.7  Soaking Pits	4-12
          4.2.8  Reheat Furnaces	  4-i3
          4.2.9  Annealing Furnaces	. ...  4-14
          4.2.10 Galvanizing Furnaces  	 ...  4-15
          4.2.11 Summary   .	. .  4-15
     4.3  REFERENCES	4-18

5    NITROGEN OXIDES CONTROL TECHNIQUES  	  5-1
     5.1  BACKGROUND	 .  5 -1
     5.2  CONTROL TECHNIQUES APPLICABLE TO IRON AND STEEL
          FACILITIES	  5-1
          5.2.1  Control Techniques Applied  . . ; . .5-1
                 5.2.1.1  Low Excess Air . . . > . . .5-3
                 5.2.1.2  Low NOX Burners   .  .  .  .  ..  .  5-3
         /       5.2.1.3  Low NOX Burner Plus  Flue  Gas
                          Recirculation	 .5-8
                 5.2.1.4  Selective Catalytic
                          Reduction  . ,	5-8
          5.2.2  Other Control Techniques  	  5-11
                 5.2.2.1  Off-Stoichiometric or Staged
                          Combustion	 . .5-11
                 5.2.2.2  Fuel Switching 	  5-16
                 5.2.2.3  Selective Noncatalytic
                          Reduction  .	5-16
     5.3  NITROGEN OXIDES CONTROLS FOR SPECIFIC
          FACILITIES	 .  5-18
          5.3.1  Coke Ovens	5-18
          5.3.2  Sinter Plants	 .  5-19
          5.3.3  Blast Furnaces and Blast
                 Furnace Stoves  .	  5-21
          5..3.4  Basic Oxygen Furnace  ........  5-22
          5.3.5  Electric Arc Furnaces .  . . . . . . .  5-23
          5.3.6  Reheat Furnaces .  .  .	•'. .5-23
          5.3.7  Soaking Pits	 ...  5-2-5
          5.3.8  Annealing and Galvanizing	5-25
                           ii

-------
                     TABLE OF CONTENTS (con.)

                      5.3.8.1  Annealing Furnaces  .  .  .  .  . 5-25
                      5.3.8.2  Galvanizing Furnaces  .  .  .  . 5-27

          5.4  SUMMARY OF CONTROLLED NOX EMISSIONS   .  .  .  . 5-27
          5.5  REFERENCES	5-29

     6    COSTS OF NOX CONTROLS	 6-1
          6.1  STEEL REHEAT FURNACES   	 6-2
               6.1.1  Model Furnaces   . . .	6-2
               6.1.2  Costs and Cost Effectiveness of  LEA
                         Control for Reheat Furnaces  .  .    6-3
               6.1.3  Costs and Cost Effectiveness of
                      LNB Control for Reheat Furnaces  .  .  . 6-3
               6.1.4  Costs and Cost Effectiveness of
                      LNB Plus FOR Control	      6-5
          6.2  ANNEALING FURNACES ... .	6-8
               6.2.1  Model Furnaces	.6-8
               6.2.2  Costs and Cost Effectiveness of  LNB
                            Control	             6-11
               6.2.3  Costs and Cost Effectiveness of  LNB
                      plus FGR Control  . . .  ..  ..  .  .. 6-9
               6.2.4  Costs and Cost Effectiveness of
                      SNCR Control	6-9
               6.2.5  Costs and Cost Effectiveness of  SCR
                      Control of Annealing Furnaces  .... 6-14
               6.2.6  Costs and Cost Effectiveness of  LNB  plus
                      SNCR for Annealing Furnaces  ...    6-14
               6.2.7  Costs and Cost Effectiveness of  LNB
                      plus SCR Controls for Annealing
                      Furnaces  ...... 	 6-17
          6.3  GALVANIZING FURNACES 	  ..... 6-17
               6.3.1  Model Furnaces	6-17
               6.3.2  Costs and Cost Effectiveness of  LNB
                      plus FGR Controls for Galvanizing
                      Furnaces		.6-17
               6.3.3  Costs and Cost Effectiveness of  LNB
                      for Galvanizing Furnaces  	  6-19
          6.4  REFERENCES 	 ...........   6-22

     7    ENVIRONMENTAL AND ENERGY IMPACTS OF NOX CONTROLS  7-1
     7.1  INTRODUCTION    . .	  7 -1
     7.2  -AIR IMPACTS  .  . .	7-1
          7.2.1  NOX Emission Reductions    ........   7-1
          7.2.2  Air Impacts of Combustion Modifications .  7-3
          7.2.3  Air Impacts of SCR and SNCR    	7-4
     7.3  SOLID WASTE IMPACTS	7-6 .
     7.4  ENERGY IMPACTS  	 .........  7-6
     7.5  REFERENCES		 .  7-8
APPENDIX A  TABULATION OF UNCONTROLLED NOX EMISSIONS DATA.  .A-l
                               ill

-------

-------
                         LIST OF FIGURES
Number

  2-1

  3-1

  3-2
  5-1
  5-2
  5-3
  5-4
  5-5
  5-6
  5-7
Simplified steelmaking flow chart
Simplified steelmaking flow chart .  .  .
into major steel product forms  ....
Schematic representation of.progress of
refining in a top-blown basic-lined EOF
Staged-air low NOX burner    ......
Staged-fuel low NO^ burners	
Illustration of flue gas recirculation
Illustration of the SCR Principle   ".  .
Biased burner firing  .........
Burners out of service  (BOOS) 	
Overfire air  	   ......
Page

 2-2

 3-9
 3-9

3-21

 5-5
 5-6
 5-9
5-10
5-13
5-14
5-15
                                IV

-------

-------
                          LIST OP TABLES
Number
                                                             Page
  2-1     SUMMARY OF UNCONTROLLED NOX EMISSIONS	    2-5
  2-2     NOX CONTROLLED EMISSIONS AND                            .
          PERCENT REDUCTIONS  ...  	     2-7
  2-3     COSTS AND COST EFFECTIVENESS  - EXISTING REHEAT  FURNACES
          (300 MMBtu/hr)	  .    2-8
  2-4     COSTS AND COST EFFECTIVENESS  - EXISTING ANNEALING
          FURNACES  (200 MMBtu/hr) .............   2-9
  2-5     COSTS AND COST EFFECTIVENESS  - EXISTING GALVANIZING
          FURNACES  (100 MMBtu/hr} .	      2-11

  3-1     SUMMARY OF CURRENT STEEL  INDUSTRY FACTS   ....    3-4
  3-2     SUMMARY INFORMATION RELEVANT TO INTEGRATED
          STEEL PRODUCERS  ........	    3-6

  4-1     CALCULATED EQUILIBRIUM CONCENTRATIONS  (in ppm)
          OF NO AND N02 IN  AIR AND FLUE GAS	    4-3
  4-2     FACTORS CONTROLLING THE FORMATION OF THERMAL NOX    4-5
  4-3     NITROGEN OXIDE EMISSION FACTORS FROM AP-42
          AND NAPAP	    4-8
  4-4     SUMMARY OF UNCONTROLLED NOX EMISSIONS	4-12 -

  5-1     OVERVIEW OF NOX CONTROL TECHNIQUES   	    5-2
  5-2     NOX CONTROLLED EMISSIONS AND PERCENT REDUCTION  .  5-28

  6-1     COSTS AND COST EFFECTIVENESS OF LEA FOR
          REHEAT FURNACES.  .	'.....   6-4
  6-2     COSTS AND COST EFFECTIVENESS OF LNB FOR
          REHEAT FURNACES	6-7
  6-3     COSTS AND COST EFFECTIVENESS OF LNB/FGR FOR
          REHEAT FURNACES  	    6-7
  6-4     COSTS AND COST EFFECTIVENESS OF LNB FOR
          ANNEALING FURNACES	6-10
  6-5     COSTS AND COST EFFECTIVENESS OF LNB/FGR FOR
          ANNEALING FURNACES	'.    6-11
  6-6     COSTS AND COST EFFECTIVENESS OF SNCR FOR
          ANNEALING FURNACES	   6-13
  6-7     COSTS AND COST EFFECTIVENESS OF SCR FOR
          ANNEALING FURNACES   	   6-15
  6-8     COSTS AND COST EFFECTIVENESS OF LNB+SNCR
          FOR ANNEALING FURNACES  .	6-16
  6-9    -COSTS AND COST EFFECTIVENESS OF LNB+SCR FOR
          ANNEALING FURNACES .... 	   6-18
  6-10    COSTS AND COST EFFECTIVENESS OF LNB+FGR FOR
          GALVANIZING FURNACES	6-20
  6-11    COSTS AND COST EFFECTIVENESS OF LNB FOR
          GALVANIZING FURNACES 	 	   6-21
  7-1     NOX EMISSION REDUCTIONS--IRON AND  STEEL
          MILL PROCESSES   	    7-2

-------

-------
                             CHAPTER 1

                           INTRODUCTION
      Congress,  in  the  Clean Air Act Amendments  of  1990  (CAAA),
 amended  Title  I of the Clean Air Act  (CAA)  to address ozone
 nonattainment  areas.   A new Subpart 2 was added to Part D  of
 Section  103.   Section  183 (c) of the new  Subpart 2  provides that:

      [w]ithin  3 years  after the date of  the enactment of the
      [CAAA], the Administrator shall issue  technical documents
      which  identify alternative controls for all categories of
      stationary sources  of  ... oxides of nitrogen  which emit, or
      have the  potential  to  emit 25 tons  per year or more of such
      air pollutant.

 These documents are to be subsequently revised  and updated as
 determined by  the  Administrator.

      Iron and  steel mills have been identified  as  a stationary
 source category with emission sources that  emit more than  25 tons
 of nitrogen oxides  (NOJ  per year.   This  alternative control
 techniques  (ACT) document provides technical information on
various  NOX  controls for various  iron  and steel  mill processes
 that  State and  local agencies may use to develop and implement  •
 regulatory programs to control NOX  emissions from iron and  steel
mills.  Additional ACT documents have been  developed for eight
other stationary source  categories.

     The information in  this ACT document was generated from
previous EPA documents and  literature searches  and  contacts with
iron and steel manufacturers,  engineering firms, control
equipment vendors,  and Federal, State, and  local regulatory-
agencies .  Chapter 2 presents a summary  of  the  findings of this
study.  Chapter  3 provides a process description and industry
characterization of iron and steel mills.  A discussion of
uncontrolled NOX emission levels is presented in Chapter 4.
Alternative control techniques and achievable controlled emission
levels are discussed in Chapter 5.   Chapter 6 presents control
costs and cost effectiveness for each control technique.
Environmental and energy impacts associated with the use of NO,
control techniques  are discussed in Chapter 7.
                               1-1

-------

-------
                            CHAPTER 2

                             SUMMARY
2.1 SOURCES OF NOX EMISSIONS

     Integrated iron and steel mills produce steel by reducing
iron ore to iron in a blast furnace and, subsequently, removing
excess carbon and other impurities from the iron in a basic
oxygen furnace.  Other processes involve beneficiating iron ore
(e.g., pelletizing), recycling of iron-bearing materials  (e.g.,
sintering), coke-making, and steel finishing processes such as
shaping, annealing, and galvanizing.  All of these are high
temperature processes, usually involving the combustion of fossil
fuels, and all are potential sources of NOX emissions.

     Mini mills and specialty producers process steel through
some subset of the full range of processes found in integrated
iron and steel mills.  Typically, they enter the process by
melting scrap steel in an electric arc furnace, bypassing the
iron-making process and attendant support activities such as     -
sintering and coke making.

     The processing order is illustrated in the simplified flow
diagram in Figure 2-1.  In coke making, coal is destructively
distilled in coke ovens that are fired with, typically, a mixture
of coke oven gas (COG) and natural gas  (NG) and/or blst furnace
gas(BFG).  Coke oven underfiring is a high-temperature process
and NOX emissions from coke making are appreciable.

     In the sinter plant, iron ore fines, coke fines, other
iron-bearing materials, and (often) flux are well-mixed and
spread uniformly on a traveling grate and ignited, typically with
NG.  As the grate travels, a forced draft causes the coke fines
and other combustibles in the bed to burn.  The mixture is thus
heated to a fusion temperature, creating a sinter suitable for
use in the blast furnace.

     In the blast furnace, iron ore is reduced to molten iron
(also called pig iron or hot metal).  The blast furnace is a
closed system with no atmospheric emissions.  The effluent from
the furnace, blast furnace gas (BFG) that is rich in carbon
monoxide  (CO), is cleaned of particulates and used as a fuel in
the blast furnace stoves.  Each blast furnace has three or four
associated stoves that preheat the air blast supplied to the
blast furnace.  Because these stoves are heated primarily with
BFG, NOX emissions  from the stoves  have low concentrations.
                               2-1

-------
                                               CD

                                               O


                                               1
                                               M—

                                               O)



                                               CD


                                               "55


                                              "8

                                               •o
                                               o>
                                                Q.



                                               <75
                                               
-------
     Molten iron from the blast furnace, along with scrap steel,
is charged to the basic oxygen furnace  (EOF) where high-purity
oxygen is blown on the molten bath  (or upward through the bath in
the case of a Q-BOP  [basic oxygen process] type furnace).  The
oxidation of undesired elements in the bath  (carbon, silicon,
manganese, etc.) converts the hot metal into steel and brings the
bath to a suitable pouring temperature.  Combustible off-gases
from the process are typically collected in an open hood above
the furnace that admits air and are burned.  Some furnaces are
equipped with suppression-type hoods that prevent air from mixing
with the off-gases above the furnace, and the off-gas is
subsequently flared.  During the combustion of the off-gas,
thermal NOX is generated.

     Scrap steel is melted in electric arc furnaces (EAF's) where
radiant heat from an electric arc established between the
electrodes (usually powered from a three-phase utility-based
supply) and the scrap or molten steel bath is used to bring the
charge to pouring temperature.  Heating due to the electric
current passing through the scrap or molten bath is a minor part
of the total heat input.  The use of electricity for steel
melting transfers the generation of NOX from the  iron  and steel
mill to a utility generating plant.  However, oxygen and NG are -
sometimes used to preheat the charge.  Thus EAF's are NOX
emission sources.

     Molten steel from steel-making furnaces is typically
processed through a continuous caster where it is shaped into
slabs,  billets, or blooms.  Alternatively, it may be cast into
ingots and stored for subsequent processing.  Slabs, billets, and
blooms from the continuous caster are typically reheated to
suitable working temperatures in reheat furnaces prior to being
passed through mills for further shaping.  Ingots are typically
reheated in soaking pits prior to subsequent processing.  Reheat
furnaces and soaking pits are high-temperature, fossil fuel
(typically natural gas)  burning furnaces and are sources of NOX
emissions.

     Finishing processes such as annealing and galvanizing also
involve reheating steel products to suitable temperatures for
processing.  Consequently, these finishing processes are also
sources of NOX emissions.

2.2  UNCONTROLLED NOX EMISSIONS

     There are three fundamentally different mechanisms of NOX
formation.  These mechanisms yield  (1) thermal NOX,  (2)  fuel  NOX,
and (3) prompt NOX.   The thermal NOX mechanism arises from the
thermal dissociation and subsequent reaction of nitrogen  (N2)  and
oxygen (02) molecules in combustion air.   The fuel NOX mechanism
arises from the evolution and reaction of fuel-bound nitrogen
                               2-3

-------
compounds with oxygen.  The prompt NOX mechanism involves  the
intermediate formation of hydrogen cyanide  (HCN), followed by the
oxidation of HCN to NO.  Natural gas and most distillate oils
have no chemically bound fuel nitrogen and essentially all NO,
formed from the combustion of these fuels is thermal NOX.
Residual oils and coals all have fuel-bound nitrogen and,  when
these are combusted, NOX is  formed  by  all  three  mechanisms.

     Iron and steel mill processes tend to use gaseous fuels,
i.e., NG, COG, BFG, and oxygen, and the NOX generation tends to
be thermal NOX.   Exceptions  include sintering where  coke fines
are burned as a fuel and coke ovens where coal is destructively
distilled in' the absence of air.  Emissions from sintering and
fugitive emissions from coke ovens may be sources of fuel NOX
emissions.  Prompt NOX formation is not a  major  factor.  It forms
only in fuel-rich flames, which are inherently low NOX emitters.
Thermal NOX formation is the predominant mechanism of NOX
generation at iron and steel mills.

     Very little NOX emissions  data are available in the
literature relevant to iron and steel processes.  Table 2-1
summaries uncontrolled NOX emissions from the major  process
facilities found in iron and steel mills.  This summary is based -
on the available emissions data obtained during the preparation
of this document.  The data are presented in Appendix A of this
document.

     The uncontrolled NOX emissions tabulated in Table 2-1
indicate that coke-oven underfiring, sintering, reheat furnaces,
annealing furnaces, and galvanizing furnaces are facilities with
emission factors that range from 120 to 940 ppm at 3 percent  02.
Data available for both preheated  combustion air and cold
combustion air furnaces show that  NOX emission factors are much
higher when the  combustion air  is  preheated.  For some
facilities, e.g., reheat furnaces,  there is much scatter  in the
data, and the averages presented in Table 2-1 may not be
representative of individual furnaces.

2.3  NOX EMISSIONS REDUCTION TECHNIQUES         •

     Control techniques for NOX emissions can be placed into one
of two basic categories:  techniques designed to minimize NOX
generation and techniques to remove previously generated  NOX from
the waste'effluent stream.  Combustion modification  techniques
such as low-NOx burners (LNB's)  and flue gas recirculation fit
into the first category.  Add-on flue  gas  treatment  techniques
such as selective catalytic reduction  (SCR)  and selective
noncatalytic reduction  (SNCR) are  examples  of the second.

     Few facilities  found at iron  and  steel  mills have NOX
controls.  For many  facilities, a  suitable  control  technique  has
                                2-4

-------
CO
S3
O
CO
CO
H ^
W ^
••"•
Q §
S^
|g
o
CJ
D ^
o °<*°
S CO ©
g CO
|8f
w
!*
FIRING 1
PROCESS FACILITY







vo
vr>
o




o
o
00



1
. 1
)ke-oven underfiring
\^
o






r-

o




o
^j*
ro



'
Sintering








O5 ' '
o ; i i
• ii




o o o o
m VD co H
ex



1 1 1
1 ,
last furnace stoves
asic Oxygen Furnace
during 02 blow
during non blow
EAF
PQ m






«t^
H
O




i
I
i


4J
_J
to
 >
-H -H
4J 4-1
Regenera
Recupera
Reheat Furnace







*3*
H
O




O
ex
H


•H
i
T)
rH
O
U








in o in
H <* H
H O O




o o o
oo m ex
c^ m H

Q) 0)
> >
-H -H ^1
4J 4-) -H
Regenera
Recupera
Cold-A
Annealing Furnace








ex
H




O
<*<
CTl

<1)
>
-H
1J
Regenera
Galvanizing Furnace
w






0

o




o
m
CO

0)
-H
4-)
Recupera








^*
H
O




O
ex
H


-H
i
•d
rH
o
o


-2-5

-------
not been demonstrated.  These facilities include sinter plants,
coke ovens, blast furnace stoves, and steelmaking furnaces.  The
Japanese have investigated using SCR for sinter plants and coke
ovens.  These efforts appear to be experimental, and.SCR is not
used routinely on these facilities in Japan.  Iron and steel
process facilities known to have NOX  controls  applied  are  reheat
furnaces, annealing furnaces, and galvanizing furnaces.  Control
techniques known to have been applied to these facilities are as
follows:
     Reheat furnaces
     Annealing furnaces
     Galvanizing furnaces
Low excess air (LEA)
LNB'S
LNB plus flue gas
  recirculation  (FGR)

LNB
LNB plus FGR
SCR
LNB plus SCR
LNB
LNB plus FGR.
Other control techniques are discussed in this document.

     Table 2-2 summarizes controlled NOX emissions data and
estimates and percent reductions.  Percent reductions range from
13 to 90%.  For reheat furnces controlled emissions range from 25
ppm @ 3% 02 for cold air  furnace  controlled by LNB+ FGR to 560
ppm for a regenerative furnace controlled by LEA.  Emissions
reductions for the reheat furnaces with preheated combustion air
range from 0.03 Ib NOx/MMBtu for  an LEA-controlled recuperative-
fired furnace to 0.61 Ib NOx/MMBtu for regenerative-fired
furnace.  Due to low uncontrolled NOX emissions,  the emissions
reductions for cold air reheat furnace are much lower.

     For annealing furnaces, controlled emissions range  from 10
ppm for a cold-air fired furnace controlled by LNB+SCR to 390 ppm
for a regenerative-fired furnace controlled by LNB. Annealing
furnace emissions reductions range from 0.07 Ib NOx/MMBtu for a
cold-fired furnace with LNB to 0.85 Ib NOx/MMBtu for a
regenerative-fired furnace with LNB plus SCR controls.

     For galvanizing furnaces, controlled emissions range from 50
ppm for a'cold- air fired furnace controlled by LNB+FGR  to 470
ppm for a regenerative-fired furnace controlled by LNB.
Emissions reductions for galvanizing furnaces range from 0.07
Ib/MMBtu for a cold-air fired furnace with LNB to 0.69 Ib/MMBtu
to a regenerative-fired furnace controlled by LNB+FGR.
                               2-6

-------





CO
6
H
H
1
r«i
8
u
rt
H
i

CO
6;
O
H
CO
CO
1
3
0
o
u
i

CM
1
CM
W
€
i-5



Hy
fc O
^4 | 	 j

IB
s §
Pi

Cft ^""*
l^j 3 Ot
^^ **"r '"J
<*> fj p
S m °

w

l|i
CO ^ (14
tj pQ (3
A h^ O
M~ CQ
Pi

CQ->
co e S
03 -^.s
U m w
BBS

j
o
p^
o
o

w
^
"

w
r)
1


rou>c^ oooinoo o
t~H VC r*^ LO ^D ^0 GO 00 Q\ if}




™S?n r-c-r-oim^ r-
rngg ooooog o
0 CD 0 00000^ 0



_ CO VD
^^ i ^^ .. ^J \O vO VO CO ^^ ^3
HOO CNHHOOO CN
O Q . Q OOOOOO O






o*\ r^ co oo co co ^ c^ c*^
VO-l ^ i-J

O
O S
Mr i 	 i
^ H N
W ^ S
ffi ^ ^
W § ' >
3 
-------
2.4  COSTS AND COST EFFECTIVENESS OF NOX CONTROL TECHNIQUES

     Tables 2-3 thru 2-5 present costs and cost- effectiveness
estimates for NOX  controls  for reheat,  annealing and  galvanizing
furnaces, respectively.  These estimates are based on a limited
data base.  Controlled NOX  missions  (and,  consequently,  emissions
reductions) are often based on test data from a single furnace.
Costs are also based on limited data, often a single example
furnace, that do not account for site-specific factors.  All
costs are in April 1994$.

     Costs and cost effectiveness vary with furnace firing
capacity.  In discussing cost effectiveness in Tables 2-3 thru 2-
5, values used refer to the mid-size capacity furnace unless
stated otherwise and are for existing furnaces.  For1 combustion
modification controls such as LNB and LNB/FGR, costs and cost
effectiveness for new furnaces are approximately 1/3 - 1/4 the
corresponding number for new sources.  For add-on controls
only(SNCR and SCR), costs and cost effectiveness are, the same for
both new and existing sources.

     As shown in Table 2-3, the cost effectiveness of existing
mid-size reheat furnaces of all control techniques range from
$90/ton NOX removed (LNB on a regenerative-fired furnace)  to
$2,400/ton NOX removed (LEA on a cold-air fired furnace).   Only
LEA has cost effectiveness greater than $700/ton.  Capital costs
and annual costs for the various controls range from $190,000 to
340,000 and $51,000 to 83,000 respectively.

     For existing mid-size annealing furnaces  (Table 2-4) cost
effectiveness of all control techniques range from $200/ton NOX
removed  (for LNB at a regenerative fired annealing furnace) to
$7,900/ton NOX removed (for LNB+SCR  at a cold-air fired furnace).
Capital costs range from $540,000 to 4,500,000 and annual costs
from $77,000 to 870,000.

     For existing mid-size galvanizing furnaces  (Table 2-5), cost
effectiveness of all control techniques range from $110/ton for
LNB on a regenerative-fired furnace to $1,200/ton for LNB+FGR on
a cold-air fired furnace.  Capital costs for the two controls are
$250,000 and $ 380,000.  Annual costs are $58,000 and $70,000.

2.5  IMPACTS OF NOX CONTROLS

     All of the NOX control techniques listed in Section 2.3 have
the potential to impact other air emissions in addition to NOX,
and all may have energy impacts.  SCR units may have a  solid
waste impact in the disposal of spent SCR catalyst.  None of the
listed techniques have a wastewater impact.
                               2-8

-------
W5
o tj ^
w
c
Cj j^**v
(D ^
o
*c3 *^*
3 •*-» ^^
*2 C^ j?^
C r\ CP
ts.
•f !§„
u UE
1
I
2
1
u

"* — r fvf


°, 0 0
;-. en cs •



O 0 O
4) O
£ EH i_
 3 -a
&0 0 —
O O O
fti & U

a

0 0-0
O\ 10 O


o o o

vo vo VO

0 O O
en en en
fN «S CS

W rt
fl} 0} •**
CO, «?
Si "°

§

o o o
• »— i ^- VO


o • o o
en oo cvj

en en en
00 00 00

o o o
*T ^T Tt
en en en

•s -s
g g .B
1) i* 2
o o o
c^ rt o
o
1
2-9

-------
CO
e / — ^
|ll
(*-
c
.2
"8 "g
o
0*^
^
ctf co i~
P|
111
o t£
a
^c

8
•+-*
o
U

o o S


i S s

t-» t> r~
t-> r- r-
O 0 0
IO >O 1/1

0 »
2 2 .fc
| .§• ?
M o 2
o o o
(Vj py] rj

§

O O S


O 0 (-3
10 O S
-* «N ^

o o o
0 O 0
oo oo oo

.1 3
22-
o ai -a
C B. f
& s 1
0
S

o § S
OO fo 6^
10 ^^ en

S I-P

\O \o vo
CN es ts
o o o
^ ON ON
NO NO VO

5 ^
•a *±
TO Cv
c .0, *?
wi 5 2
U U O
(Vj fV| ^_)

u
C/3

o o o
o o o
i— CN r-

O : Q O

O ' O o
o : o o

.55 • .55
1 i -
c : a '?
p . a TJ
w o -S
O d3 O
C£i (Vj tj

0
00
2-10

-------
s

-H
J-)

0
O
CJ)
 I
CNJ
PQ
<
H
C ' — •>
111
u S tS-
§
e
= 'E
^ c
2
0*"^
Cti W3 ^
g II
— o*
-S to g
•a o 0
8-u -
u t^
1
H
e
;c
,
o
U
888
— m r~


o o o
00 00 r-.

^^ (^5 ^^
r~ r- r-
oo oo oo
o o o
o o o
10 >o vo
£ g
2 2 u
S SL '?
ass
«;> S o
p/j ^5 O
U
w
4-
^
P i 1
— «  >
s e
U O '3
CO.?

  • -------
    JE
     12
    CO
    CO ,
    « 1?
    a5 C'
    fc
    _o
    ^ c
    " O
    O
    Ctf C/3 ^^
    I5!
    III
    O £t
    H
    c
    ;c
    
    2
    •*->
    o
    u
    
    2 .8-8
    — . f* ON
    
    
    ^D O ^5
    r(- CS •<*
    CO —
    
    
    
    0 O O
    CM CM CN
    O ID
    £ .£
    4-< -4-J
    e e w
    g a •?.
    a s s
    D S O
    « Ot$ 0
    
    
    §
    
    o o o
    2 5 r4-
    
    
    0 0 0
    
    
    VO VO VO
    0 0 O
    oo oo oo
    co co co
    m o
    ^ •§ v.
    S & '?
    (D S "O
    g> 8' -o
    os « o
    o
    ^!
    §
                                 2-12
    

    -------
    2.5.1  Air Impacts of Combustion Modification Techniques
    
         All of the listed combustion modification NOX control
    techniques, LEA, LNB, and LNB plus FGR, may increase CO and
    unburned hydrocarbon (HC) emissions.  The NOX  reduction
    mechanisms inherent in these modification techniques are the
    reduction of peak flame temperatures, which is exponentially
    related to the formation of thermal NOX and the reduced
    availability of excess oxygen needed to form NOX.   Reducing the
    availability of oxygen to the combustion process increases the
    likelihood that some HC will not be burned and that some CO will
    not be oxidized to C02.   However,  if the control is properly
    designed and applied, NOX control  can be achieved without
    increasing CO or HC emissions.
    
         Data contained in the process heaters and glass ACT
    documents indicate that decreases in excess oxygen levels begin
    to impact CO emissions below 3 percent excess oxygen.  Data in
    the utility boilers ACT document show that CO emissions  increase
    for some boilers and decreace for others when implementing
    combustion modifications.  In a steel reheat furnace with LEA  for
    NOX control,  CO emissions of  39  ppm at 3 percent O2 and 26 ppm  at
    3 percent 02 are reported for the  uncontrolled and controlled
    tests, respectively.  These results indicate that CO emissions do
    not necessarily increase when LEA is implemented for NOX control.
    
         Other available information reports CO emissions  from an
    LNB-controlled reheat furnace to be 20  to  30 ppm at 3  percent  O2.
    Further, laboratory and field tests on  a galvanizing furnace
    yielded NOX emissions of 550  to  1,200 ppm at 3 percent 02 and,
    concurrently, negligible CO emissions.  Modifications  to the
    burners reduced NOX emissions to 350 to 430 ppm at 3 percent 02
    and, concurrently, increased CO emissions  to 30 to  60  ppm at 3
    percent 02.   The purpose of the  modifications  was to reduce NOX.
    
         No explicit data are available relevant to the impact of  NO,
    control techniques on HC emissions  from iron and  steel process
    facilities.  In general, controls that  reduce NOX by reducing the
    availability of excess oxygen in the high-temperature  regions  of
    a furnace would not  increase either CO  or  HC emissions unless
    oxygen availability was reduced excessively.   Limite d data  in
    the Industrial-Commercial-Institutional Boilers ACT document
    indicate that HC emissions do not change due to the
    implementation of  combustion modifications.
    
         Reheat, annealing, and galvanizing furnaces  predominantly
    use natural gas as a fuel.  Natural gas does not  contain sulfur
    and, consequently, SO2 is not emitted.  Gaseous fuels, including
    natural gas, can produce  soot and carbon black when burned if
    insufficient oxygen  is present.  However,  no  evidence  indicates
    that NO. controls increase particulate emissions in these iron
                                  2-13
    

    -------
    and steel facilities.
    2.5.2  Air Impacts of SCR and SNCR
    
         SCR units are add-on, flue gas treatment facilities that
    reduce NOX by injecting ammonia (NH3) upstream of a catalyst
    reactor.  Within the catalyst, NOX reacts with the NH3 and is
    reduced to N2 and water '(Hp) .  There is a potential for unreacted
    NH3 to  escape with the  flue gas from the SCR unit.   Any such
    emissions are referred to as ammonia slip.
    
         Two examples of SCR controls on annealing furnaces at iron
    and steel mills are reported.  One of these units is operational
    with more than 3 years' operating history, and one is still under
    construction.  In the case of the former, the NH3 slip was
    guaranteed to be less than 10 ppm initially and less than 12 ppm
    after 1 year.  The observed NH3 was initially less than 10  ppm.
    Subsequent observations are not reported.  The typical NH3/NOX
    molar ratio for this unit is 0.9.
    
         Other reports indicate that NOX removal rates of  70 to 90   -
    percent can be achieved with SCR using NH3/NOX molar ratios
    between 0.9 and 1.0, and that the NH3 slip will be between  5 and
    10 ppm.  These levels are considered to be well below health and
    odor thresholds.
    
    
         SNCR has not been applied to iron and steel mill process
    facilities.  In other SNCR applications, ammonia slip is
    controlled to acceptable levels by controlling the ammonia to NOX
    molar ratio.  These levels are similar to ammonia emissions from
    SCR applications, e.g., 10 ppm.
    
         Pilot-scale testing and chemical kinetic modeling of SNCR
    processes have shown that nitrous oxide  (N20)  emissions are a by-
    product of both ammonia (NH3). and  urea injection.   The N20
    formation resulting from these processes has been shown to be
    dependent on the reagent used, the amount of reagent injected,
    and the injection temperature.
    
         Full-scale tests on fossil-fuel-fired boilers have shown
    that direct emissions of N20 are less than 15 ppm and  do not
    generally correlate with NOX emissions.   N20 production is higher
    for urea injection than it is for NH3 injection.
                                         V
    
    2.5.3  Solid Waste Impacts
    
         The only NOX control  with solid waste impacts is  SCR due to
                                   2-14
    

    -------
    the disposal of spent catalyst materials.  Titanium dioxide and
    vanadium/titanium have been identified as the catalysts in the
    two SCR annealing furnace applications cited.  Other commonly
    used materials are vanadium pentoxide, tungsten trioxide,
    platinum, zeolites, and ceramics.  Of these, vanadium pentoxide
    is a toxic compound and a cause for concern.  However, worker?
    safety precautions adequately prevent any increased risk to
    workers handling the catalyst, and stack emissions of vanadium
    pentoxide are 1 million times less than industrial worker
    exposure.
    
         Most catalyst manufacturers arrange to recycle and
    reactivate the catalyst.  When that is not practical, the spent
    catalyst can"be disposed of in an approved landfill in accordance
    with the Land Disposal Restrictions in 40 CFR Part 268,
    Subpart D.
    
    2.5.4  Energy Impacts
    
         All of the combustion modification control techniques have
    the potential to impact energy requirements by affecting the
    thermal efficiency of the process.  No data are available to
    quantify the impact of these controls on iron and steel mill
    process facilities.
    
         SCR results in a pressure drop across the catalyst that
    requires additional electrical energy for the flue gas fan.  One
    estimates a cost of $537/yr assuming 8,000 hours operation per
    year and electricity at 8 cents/kWh.
                                   2-15
    

    -------
    

    -------
                                 CHAPTER 3
    
                           INDUSTRY DESCRIPTION
          Iron and steel  are essential  commodities  in a modern,
     industrialized society.   Steel  is  a widely used industrial
     material  affecting every facet  of  society.   Steel is  a major
     component of  every transportation  system,  present in  every
     motorized vehicle  and in the network of  highways,  bridges,  and
     traffic controls.  It is a widely  used construction material,
     present in every home,  skyscraper,  and dam.  Steel is essential
     for most  industrial  facilities, prevalent  in defense  hardware,
     and found in  most  durable goods.   The uses  of  steel are too
     prevalent and too  diverse to be listed completely or  to serve  as
     a basis for classification.
    
          Section  3.1 of  this chapter presents a brief background/
     historical discussion of the evolution of  iron and steel
     processing.   Section 3.2 is  a brief industry characterization
     including an  overview description  of the major production
     processes and an overview of the industry.   Section 3.3 presents
     a description of those processes with a  potential  for large NOX
     emissions.
    
     3 .1 BACKGROUND
    
          Iron (Fe)  is  one of the more  abundant  and widely distributed
     elements  in the  earth's  crust, constituting  not less  than 4
     percent of  the  total  crust.   It is  the fourth  most  abundant
     element in the  earth's crust, outranked  only by aluminum,
     silicon,  and  oxygen.   Pure iron is  a silvery white, relatively
     soft metal  that  readily  combines with other  elements,  e.g.,
     oxygen and  sulfur.   Consequently, native metallic  iron is rarely
     found in  nature.1'2  Iron oxides are the most prevalent natural
     form of iron.  That portion  of the  iron  oxides  that is  of
     commercial  significance,  i.e., economically  and spatially
     available  for industry use,  is referred  to as  iron  ore,  and  iron
     ore deposits are widely  distributed.  These  deposits  vary widely
     in mineralogy, chemical  composition,  and physical
     characteristics.  The United States  has abundant reserves of iron
     ore.   These reserves  are  grouped into five areas:   Lake Superior,
    Northeastern,  Missouri,  Southeastern, and Western.  Of  these, the
    Lake Superior reserves are the most  important.3
    
         Steel  is the generic name for a group of  ferrous metals
    composed principally  of  iron which, because  of  their  abundance,
    durability, versatility,  and low cost, are among the  most useful
    metallic materials known.  Most steels contain more than 98
    percent iron.   Steel also contains carbon,  up  to about  2 percent,
    and may contain other elements.   By  controlling  the carbon
                                   3-1
    

    -------
    content and alloying elements, and by proper selection of heat
    treatment in the finishing processes, steel can be produced with
    a wide range of mechanical and physical properties.
    
    3.1.1 A Historical Perspective
    
          Facilities for reducing iron ore (iron oxide) to metallic
    iron have evolved to the modern blast furnace that produces iron
    from ore and other iron-bearing materials  (e.g., sinter, pellets,
    steelmaking slag, and scrap), coke, and flux (limestone and
    dolomite).  In the blast furnace, a blast of heated air and, in
    most instances, a gaseous or liquid fuel, are injected near the
    bottom of the furnace.  The heated air burns the fuel and the
    coke to produce the heat required by the process and the reducing
    gas that removes oxygen from the ore.  The reduced iron melts and
    pools in the bottom of the furnace where it is periodically
    drained through tapping holes as molten pig iron or hot metal.4
    
         Modern steel making dates from the introduction of the
    pneumatic or Bessemer process in 1856.  The Bessemer process
    involved forcing air through molten pig iron to oxidize the major
    impurities (silicon, manganese, and carbon).  The Bessemer
    process was quickly followed by development of the open-hearth
    furnace that evolved into the principal means of producing steel
    throughout the world.5  Open-hearth furnaces  were  used in the
    United States to produce steel into this decade, but all have now
    been replaced by modern basic oxygen furnaces(BOF).
    
         In the modern basic oxygen furnace,  high-purity oxygen is
    blown onto the surface of the hot metal  (top-blown or basic
    oxygen process) or up through the hot metal  (bottom-blown or Q-
    BOP process), oxidizing or removing excess carbon, silicon,
    manganese, and other impurities from the hot metal to produce
    steel of a desired composition.6
    
    3.1.2 An Overview of the U.S. Iron and Steel Industry
    
         Steel production is an international industry serving a
    highly competitive, international market.  The United States is
    one of about 85 steel-producing nations and accounts ;for about 12
    percent of the 1991 world production.  In 1990, the United States
    produced about 99 million tons of raw steel.  Of this total, 37
    percent was produced in EAF's and the balance predominantly by
    BOF's.  In 1991, the United States produced about  87 million tons
    of raw steel.  In 1991, 76.percent of the steel produced was-
    processed through continuous casting machines rather than through
    ingot casting.  Continuous casting is projected to account for
    about 84 percent in 1995.  It should be noted that 1991 was not a
    good year for steel production.  World steel production was down
    about 5 percent from 1990, and U.S. production was down 12
    percent.  Shipments to both the automobile and construction
                                   3-2
    

    -------
    industries were down significantly in 1991.  The automobile
    industry is the largest single consumer of steel in the United
    States, and the construction industry is.also a large consumer.7
    
         Emphasizing the international character of the steel
    industry, the import share of the United States steel market,
    during the first half of 1991, was 8.24 million tons or 18.9
    percent.  Exports for the same period were 3.2 million tons.
    
         Annual steel production capacity in the United States_in
    1991 was about 120 million tons.  High and low capacities in
    recent times were 160 million tons in 1977 and 112 million tons
    in 1988.  In 1990, the steel industry in the United States was
    composed of approximately 300 companies, of which 83 produced raw
    steel at 127 locations.8
    
         Table 3-1 summarizes data pertinent to the iron and steel
    industry to characterize the current status of the industry.
    Basic oxygen converters  (BOP's and Q-BOP's) remain the major
    source of steel production, but electric arc furnaces now account
    for about 37 percent.  Open hearth furnaces are no longer in use
    in the United States.
    
          A 1991 directory of iron and steel plants in North America
    defines three categories of steel producers as follows:9
    
         •    Integrated Steel Producers are defined as those
              companies having blast furnace or direct reduction
              facilities and whose principal commercial activity is
              the production and sale of carbon steel.
    
         •    Specialty Producers are defined  as those companies _
              whose principal commercial activity is the production
              and sale of stainless steels, alloy steels, tool
              steels, bars, wires, pipe, etc.
    
         •    Mini-Plants or mini-mills are defined as those
              companies whose production is based on electric
              furnace-continuous  caster-rod/bar mill operations,
              generally rolling carbon steel products--rebar, rounds,
              flats and small shapes.
    
    The differences between these categories are not always  clear.
    Some companies in all three categories use EAF's to melt  steel
    and have  continuous casters.  The integrated producers  are  the
                                    3-3
    

    -------
    TABLE 3-1.  SUMMARY  OF  CURRENT STEEL INDUSTRY PACTS7'1
    
    U.S. percentage of
    world production (%)
    Total U.S. production
    (106 tons)
    Total U.S. shipments
    (106 tons)
    Total production by (%) :
    Basic oxygen converter
    EAFs
    Open hearths
    U.S. raw steel capacity
    (106 tons)
    Number of companies
    Raw steel producers
    Imports (105 tons)
    Percent by continuous
    casting (%)
    1991
    11.6
    87
    79
    37
    120
    
    83
    
    75.8
    1992
    
    92.9
    82.2
    
    113
    
    
    
    
    1993
    
    96.1
    88.5
    
    110
    
    
    
    
                             3-4
    

    -------
    most distinct group because they alone produce iron from iron ore
    in blast furnaces or direct reduction iron  (DRI) facilities.
    Integrated producers also use coke in blast furnaces and either
    produce coke on site or purchase it from coke producers.  As of
    May ,1991, there was only one DRI facility in the United States.10 ..
    Mini-mills use only electric arc furnaces and use scrap steel as
    feed stock.
    
    
         3.1.2.1 Integrated Producers.  The integrated producers,
    although much fewer in number than specialty producers and mini-
    mi 11s, produce the greater volume of steel.  In 1990, integrated
    producers accounted for about 70 percent of raw steel production
    in the United States.7'11  The referenced 1991 directory lists 20
    integrated steel producers in the United States.  However, it is
    not clear that all of those listed conform to the preceding
    definition of an integrated producer, i.e., use blast furnaces or
    DRI facilities and produce carbon steel.  Those known to be
    integrated producers are listed in Table 3-2 along with other
    available information.9
    
         3.1.2.2 Mini-Mills.  The mini-mills are a growing segment of
    the steel-producing companies that, in 1991, accounted for more  -
    than 20 percent of production in the United States.  It has been
    suggested that the term mini-mill is a misnomer.  The steel-
    melting facility used by mini-mills  (i.e., EAF's) can produce up
    to 130 tons of steel per hour, not far from the 200- to-300-
    tons/hr typical of a EOF.12   Once viewed as suppliers of
    unsophisticated, low-quality products requiring minimal
    technology, mini-mills are now recognized as playing a growing
    technological role and for having made permanent inroads into the
    traditional domain of integrated producers.13  In 1990, Nucor
    Corporation and North Star Steel Company, two mini-mill
    producers, ranked seventh and eighth, respectively, among United
    States steel producers.  Nucor, for example, has an annual
    production capacity of 3 million tons, and North Star an annual
    capacity of about 2.5 million tons.  By contrast, USX Corporation
    (formerly U.S. Steel Corporation), the nation's leading steel
    producer, shipped about 12 million tons in 1990.9>12  The 1991
    directory of iron and steel plants lists 42 mini-mill companies.
    
         3.1.2.3 Specialty Producers.  It is more difficult to
    characterize the specialty producers than integrated or mini-mill
    producers.  There are more companies, a total of about 120; they
    produce a wide variety of specialty products; and it is difficult
    to ascertain the starting point in their production process from
    the literature. 'Company sizes vary widely  (e.g., the number of
    employees number from as few as 8 to as many as 5,000) and annual
    capacities range from 62,000 to 1,500,000 tons of product.  As a
    group, they operate EAF's and a host of other furnaces including
    annealing and reheat furnaces.  By contrast, some weld or do
                                   3-5 ,
    

    -------
    
    
    
    
    
    ffv
    to
    PJ
    H
    0
    P
    Q
    o
    Pi
    Pi
    j
    W
    to
    _
    p
    uQ
    M
    3
    5
    W
    c
    H
    O
    IH
    
    E?
    5jj
    1
    §
    H
    NPORMAT
    H
    1
    to
    
    
    
    •
    CS*
    1
    f)
    
    W
    t"3
    1
    •|
    4J .P
    a A
    So1
    ft i-i
    
    
    
     1
    ft vo
    tt rH
    00
    CJ eo
    i-l
    BETHLEHEM STEEL
    Bethlehem, PA
    (215) 694-2424
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    -O
    0)
    •» N
    0)-H
    «a
    * rH ?
    S C 01
    -H -H
    4J a
    Ol -H
    -rl • i-l
    ft CD 4J
    rH 01
    blooms ,
    chemica
    sheet ,
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    _
    b
    
    
    
    
    
    o
    o
    r-
    tN
    
    
    
    
    o
    o
    0
    o
    o
    k. '
    H
    
    
    
    ^
    4J
    4)
    (1) •
    J3 o
    to *o
    Q) Q
    o •
    0)
    c a
    O-H
    w a
    •H
    • 0) m
    Q) JJ rH
    .y *-H
    O rH O
    cj au
    
    
    
    
    
    
    
    
    
    
    
    
    m
    GENEVA STEEL
    Provo, UT 8460
    (801) 227-9090
    
    
    _
    4
    
    
    
    
    
    eri
    H
    c»
    IN
    
    
    •d
    o
    o JH o
    o ao
    o o
    • i-i •
    o  tN
    • u •
    rH 05 H
    
    
    ••0
    0) 0)
    4-* N
    m-H
    • rH
    m (6
    rH Ol
    u a
    -rl -H
    0) -U
    x; to
    u
    rH <1>
    (I) 0)
    O J3
    cj tn
    
    
    
    
    
    
    • in
    cj n
    2i CM
    H H
    I
    • i-l
    i-3 0
    H a\
    GULF STATES STE
    Gads den, AL 35
    (205) 543-6100
    
    
    ^
    M
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •*
    * s
    u
    (D fl5
    S-°
    rH •
    am
    i-H
    Sheets,
    chemica
    billet
    
    CJ
    25
    H
    
    k.
    co
    Pa
    H
    Oi
    
    CO ro
    D o
    P vo
    INLAND STEEL IN
    Chicago, IL 60
    (219) 399-1200
    M
    
    SC
    5
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •
    Id m
    rH a
    "a
    4J •
    0) Ol
    rH C
    rH -rl
    0 <0
    a o
    H U
    
    
    
    
    
    
    
    
    
    in
    • H
    £3
    M ^
    LTV STEEL CO.,
    Cleveland, OH
    (216) 622-5000
    - ~-
    
    ..
    S
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    a
    •H
    4J
    m
    4J"
    0)
    0) -a
    J3 0>
    m H
    •H
    i«J
    4J O !>
    «} rH rH
    i-l rH «5-
    ft «J 01
    
    
    
    
    
    
    
    
    
    
    
    m
    co
    H
    McLOUTH STEEL
    Trenton, MI 48
    (313) 285-1200
    
    
    
    
    ,_
    tu
    p
    JIj
    4J
    
    
    
    <—•
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    3-6
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ^^
    •0
    Q)
    §
    •H
    
    §
    O
    ^•*>
    (N
    
    TABLE 3
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    "§
    •H
    ij ii
    SB)
    0
    H 0
    O) rH
     co
    U 4J
    Coal , chemi
    sheet , ingo
    galvanized,
    Ingot, slab
    
    
    
    
    
    • in
    o< «* H
    e! in CN
    O U3 i-H
    U •«* CO
    • ^f
    J o O
    
    MHO HO
    (ft (*) W *• rH
    § (C CS CO H H
    K & O •#
    o rt -— *• w _n f^
    H J3 en O tn i
    EH CQ H E3 rt O
    2 5EJ ^^ Pi Q CO
    
    
    
    $
    
    
    
    
    0
    r~
    o
    o
    o
    o
    o
    
    fc
    ^
    
    
    
    
    
    
    0)
    t 6^
    rH
    V H
    8 1
    
    Strapping,
    tubing
    gavlvanized
    
    
    
    
    
    
    
    • rH
    Da CJ
    te H
    ^5 ^D
    u H in
    t-
    W < rt
    U CU i
    EH H
    rH CTl
    !5rH
    
    BJ 1-1 CS
    
    CO fc —
    «.
    Ol H
    
    S'S
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    »
    
    o
    Cn
    C -
    "H iJ O
    C"H rH
    O ft rt
    
    Coke, pig i
    pipe, rods,
    galvanized,
    
    \o
    r*
    T
    ON
    rH
    CS
    in
    rH
    
    grH
    
    «rt
    • Dim
    i!5
    U tn —
    4J (N
    X -U H
    D ft -"
    
    
    
    §
    
    
    
    
    
    o
    0
    0
    o
    o
    in
    
    H
    
    
    
    
    
    
    •O
    i-H r-H
    rH  S
    CO t-
    a c"?
    O 0
    EH 4-J -^^>
    Pi M *J*
    H -H O'
    » S ~-
    
    
    
    §
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •0
    V
    4J N
    O-H
    Cn C co
    C rt 4J
    -H > V
    rH 3
    * flj 'o
    Coke , sheet
    tinpiate, g
    roofing pro
    8
    
    w.
    B
    CO
    
    K ro
    O o
    0J O
    »D VD
    CQ f^
    co o
    EH o
    H ^ M
    1 .T}l
    2J C CN
    H -H
    r-3 rH •—*.
    PQ O ^
    ta o) o
    J5 5 -"
                                                               m
                                                               Q)
                                                               OJ
                                                               CO
    
                                                               >1
                                                               ^)
    
                                                               m
                                                               C
                                                               o
                                                               •rH
                                                               OJ
                                                               rt
                                                                Cu
                                                                (1)
                                                                4J
    
                                                                2
                                                                tn
                                                                0)
                                                                4J
                                                                C
                                                                H
    3-7
    

    -------
    other single-function operations that do not include the onsite
    combustion of fuels.  Products may range from a single item such
    as cold-drawn wire to highly specialized steels.9   In terms  of
    their contribution to domestic capacity and production, they are
    grouped with mini-mills.
    
    3.2 INDUSTRY CHARACTERIZATION
    
    3.2.1 An Overview of Iron and Steel Manufacturing
    
         Figure 3-1 is a simplified flow diagram illustrating the
    principal steps involved in the production of steel.  All of
    these processes, and more, are found in modern integrated steel
    mills.  An overview of these processes is presented here.
    Individual processes that produce NOX are discussed in more
    detail in Section 3.3 of this chapter.
    
         3.2.1.1 Raw Materials and Preparatory Processes.  With
    reference to Figure 3-1, the basic raw material input for the
    production of iron and steel is iron ore,  which is reduced to
    metallic iron in the blast furnace.  The principal iron ore
    beneficiating processes are pelletizing and sintering.
    Pelletizing is usually accomplished in specially designed
    furnaces located at or near iron ore mines rather than at iron
    and steel mills.  Its purpose is to pelletize fine, low-grade
    ores prior to shipping to an iron and steel mill.   Pelletizing is
    rarely practiced in an iron and steel mill.
    
         In contrast to pelletizing, sintering is often practiced at
    integrated iron and steel mills.  Its primary purpose is to
    agglomerate dusts and fines from other process, e.g., ore fines,
    coke fines, and flue dust, into particles with suitable mass,
    size, porosity, and strength to charge into the blast furnace.
    Fines charged to a sintering furnace differ in size from those
    charged to a pelletizing furnace, and they originate at the iron
    and steel mill rather than at the mine.  Alternatively, fine
    particles charged into the blast furnace will be blown out by the
    rapid countercurrent flow of the furnace gases.
    
         Another basic material used in the conversion of iron ore
    into metallic iron is coke.  Coke is the primary residue that
    remains when a blend of pulverized coking coals is heated
    gradually to high temperatures, about 900 to 1,100  °C (1,650 to
    2,010 °F) .in the absence of air for approximately 18 hours.
    About 90 percent of the coke produced in the United Stated is
    charged to blast furnaces for the production of pig iron.14
    
         The conversion of coal to coke is performed in long, narrow,
    slot ovens, i.e., by-product coke ovens, which are usually
    designed and operated to permit the separation and recovery of
    the volatile materials evolved from the coal during the coking
                                   3-8
    

    -------
    
         pl§
          *- re
      IS u
      £g
      0) h;
           i
          CD
          C -
          3-J2
          go.
    
          to
      1
      (A
    
      o
            »
           -8
    
    CO
    is
    0) !=
    «5
    en
          o
    
          t o
    
          43 (0
    
          CD
      CO
    
    
      2
      re
    
    
    
      p
    t^
                 oc
                 o
         Q)
         U
              II
    «!«
    c-^: o
    (75 £
         o
         g
        i
                        tsl
                           t
                           re
    
                           u
    
    
                           I
                           «P
    
                           O)
                           c.
    
                           (0
    
    
                           "33
                           v
                             o.
    
    
                             (75
                             CO
    
    
    
                             I
                             O)
        J2 o
         '
         |
          -
                CO
    
    
                15
                    8
                    o
                       §
                       TO
                       .5
                       •s
                       Q.
             3-9
    

    -------
    process.  In addition to coke, the primary product, a number of
    by-products may also be recovered including  breeze, crude tar,
    crude light oil, ammonia, and coke oven gas.  These may be used  -,
    on site or marketed.  Coke is likely to be used on site but may
    be marketed, coke oven gas is likely to be used on site, and
    breeze is likely to be recycled to the sintering process.15
    
         A third raw material input to the iron and steel making
    process is flux.  In the iron smelting process, i.e., when iron
    is separated from the ore by fusion in a blast furnace, a flux is
    used to enhance the process by combining with ash in the coke and
    gangue in the ore to make a fluid slag that can be readily
    separated from the molten iron.  Selection of the proper flux for
    a given process is a well-established process requiring knowledge
    of the composition and properties of the materials involved.
    Limestone and/or dolomite are the fluxes used in blast furnaces.
    The proportions of each depend principally on the constituents of
    the slag and the amount of sulfur that the slag must remove.16
    
         In the basic oxygen steelmaking process, lime normally is
    added as calcined or burnt lime or burnt dolomite.  The calcium  -
    oxide in either the burnt lime or-burnt dolomite fluxes the
    silica formed upon the oxidation of silicon in the hot metal.17
    
         Often, limestone is also charged into the sinter plant to
    produce a precalcined or self-fluxing sinter.  The use of self-
    fluxing sinter has both economic and performance advantages.18
    
         3.2.1.2 Ironmaking.  Iron ore, coke, selected fluxes  (e.g.,
    prepared limestone), and sinter are charged to the blast furnace
    where iron is reduced from its ore.  The product of the blast
    furnace contains more than 90 percent iron and is referred to as
    pig iron or, if in a molten form, hot metal.  Most of the blast
    furnace product is transported directly to onsite steelmaking
    furnaces as hot metal.
    
         The blast furnace charge of iron-bearing materials
    (including iron ore, sinter, mill scale, steelmaking slag, and
    scrap), coke, and flux  (limestone and/or dolomite) is placed in
    the furnace, and a blast of heated air and, in most instances, a
    gaseous, liquid, or powdered fuel is introduced through openings
    near the bottom of the furnace just above the furnace's hearth
    crucible.  The blast of, heated air burns the injected fuel and
    most of the coke to produce the heat required by the> process and
    the reducing gas that removes oxygen from the ore.  The reduced
    iron melts and pools in the bottom of the hearth.  The flux
    combines with the- impurities in the ore to produce a slag, which
    also melts and accumulates on top of the liquid iron.
    Periodically, the iron and the slag are drained from the furnace
    through tapping holes.4
                                   3-10
    

    -------
         The blast air delivered to the process through the furnace
    tuyeres is preheated by passing it through regenerative blast-
    furnace stoves that are heated primarily by combustion of the
    blast furnace off-gas.  Blast furnace gas is often enriched with
    fuel of a higher calorific value such as natural gas to achieve
    higher temperatures.  Thus, some off-gas energy is returned to
    the blast furnace in the form of sensible heat.  This procedure
    enhances the efficiency of the process by lowering fuel
    requirements.
    
         A modern blast furnace will typically have three or four
    blast furnace stoves as auxiliaries.  These are alternately fired
    with blast furnace gas to raise the temperature of the stoves'
    brick lining and then, by reversing the gas flow, preheat the
    supply of blast air to the blast furnace to temperatures of 760
    to 1,150°C (1,400 to 2,100 °F).19
    
         To produce a metric ton of pig iron requires about 1.7
    metric tons of ore or other iron-bearing material, 450 to 650 kg
    of coke and other fuel, about 250 kg of limestone or dolomite,
    and 1.6 to 2.0 metric tons of air.4
    
          Although the blast furnace remains the dominant source of
    iron for steelmaking, there are processes that produce iron by
    the reduction of iron ore below the melting point of iron.  These
    are classified as direct reduction processes, and the products
    are referred to as direct-reduced iron(DRI).  DRI produces
    several percent of the total iron produced worldwide. The major
    part of DRI production is used as a substitute for scrap in
    electric-arc steelmaking furnaces.20 There  are many DRI
    processes.  The objective of these processes is to improve iron
    ore until it is sufficiently iron-rich to be charged to an
    electric arc furnace.  A typical process uses heat to drive
    oxygen from the ore, leaving an iron-rich residue called sponge
    iron.6
    
         3.2.1.3 Steelmaking.  Steel is made from molten iron  (hot
    metal), delivered directly from a blast furnace, and scrap in
    steelmaking furnaces. Two basic types of steelmaking furnaces
    dominate this process: (1) ; the EOF and  (2)  the EAF.  The open-
    hearth steelmaking furnace, developed in the 19th century and a
    major source of steel well into the latter half of the 20th
    century, is no longer used in the United States.7  There are
    three basic types of the EOF.  All use oxygen of high purity
    (>99.5 percent) to oxidize excess carbon, silicon, and other
    impurities in the hot metal, thereby producing steel.  One type
    of EOF blows the oxygen on the top of the hot metal pool.  In the
    United States, this process is called the basic oxygen process or
    BOP.  The second type blows the oxygen through tuyeres in the
    bottom of the furnace and is referred to as the bottom-blown or .
    Q-BOP process. The BOP process is the most widely used form in
                                   3-11
    

    -------
    the United States.
    other two.
    The third type of EOF is a combination of the
         To make steel in a EOF, molten pig iron (hot metal) and
    usually scrap steel are charged to the furnace.  The scrap
    consists of the by-products .of steel fabrication and wornout,
    broken, or discarded articles containing iron and steel.  The
    furnace is mounted in a trunnion ring to facilitate tilting
    during processing.  At the beginning of a heat, i.e., the
    processing of a batch of materials to make steel, the furnace is
    tilted to receive a charge of materials through the open top.
    Scrap steel, if any, is charged first.  Scrap can form up to 30
    percent of the total charge  (up to 45 percent if preheated).1
    The furnace is then tilted in the opposite direction to receive a
    charge of hot metal (molten iron) from a transfer ladle.  The
    furnace is then returned to the vertical position and a water-
    cooled retractable lance is inserted through the open top of the
    furnace and positioned above the bath level.  A water-cooled hood
    is positioned over the open top.  A jet of gaseous oxygen is
    blown at high velocity onto the surface of the hot metal bath.
    No external heat source is required during this process.         -
    
         Slag-forming fluxes, e.g., burnt lime, dolomitic lime, and
    fluorspar, are added in controlled amounts through a chute built
    into the side of the hood.  The flux is added shortly before or
    after the oxygen jet is started.                   '''.-..
    
         The oxygen striking the surface of the molten bath
    immediately forms iron oxide, part of which disperses rapidly
    through the bath.  The iron oxide reacts with carbon in the
    molten bath to form carbon monoxide, which gives rise to a
    violent circulation that accelerates the refining process.
    Impurities such as carbon, manganese, silicon,  sulfur, and
    phosphorus are oxidized and transferred to the slag.  These
    oxidizing reactions take place very rapidly.  A 300-metric-ton
    heat (or charge), for example, can be processed in about 30
    minutes.6
    
         During the oxygen blowing, gases emitted by the process are
    collected by the water-cooled hood and conducted to a cleaning
    system where solids are removed from the effluent gas before it
    is discharged to the atmosphere.  The solids may be returned to
    the sintering process or discarded, depending on whether or not
    they contain contaminants.  When the oxygen blowing is completed,
    the lance is withdrawn and the molten steel is checked for
    temperature and composition.  If the steel is too hot, it is
    cooled by the addition of scrap or limestone.  If the steel is
    too cold,  the oxygen lance is reinserted and additional oxygen is
    blown.   When the temperature and composition are satisfactory,
    the steel is tapped into a waiting ladle where alloying materials
    are added.  Subsequently, the remaining slag is dumped from the
                                   3-12
    

    -------
    furnace, and the furnace is returned to a charging position to
    begin the next heat.1'6'22
    
         The bottom-blown or Q-BOP process is also used in the United"'
    States.  The Q-BOP furnace has tuyeres or double pipes in the
    bottom of the furnace.  Oxygen is blown into the furnace through
    the center pipe and natural gas or some other hydrocarbon is
    blown into the furnace through the annular space between the two
    pipes.  Considerable heat is generated when the oxygen oxidizes
    the carbon, silicon, and iron in the molten bath.  The
    hydrocarbon stream provides essential cooling by thermal
    decomposition.  Without this cooling, the refractories
    surrounding the oxygen jet would be destroyed.1'6
    
         Although the BOP and Q-BOP processes have many similarities,
    there are also significant differences.  The Q-BOP process
    operates much closer to equilibrium conditions between the metal
    and the slag and is therefore much lower in oxidation potential
    than the BOP.  The manifestations of this characteristic of the
    Q-BOP are lower iron losses as ferric oxide (FeO) to the slag,
    higher manganese recoveries, faster blow times, improved
    phosphorus and sulfur control, and lower dissolved oxygen and
    nitrogen contents.  Because of these improvements, yield of the
    Q-BOP is 1.5 to 2.0 percent higher than the BOP,  i.e., there is
    less iron loss as FeO to the slag.  The lower FeO generation,
    however, results in lower scrap melting in the Q-BOP as compared
    to the BOP.  Process control of the Q-BOP is much easier than the
    BOP because of the highly consistent metallurgical behavior of
    the Q-BOP reactions and the absence of the variability caused by
    oxygen lance practices as encountered in the BOP.21
    
         The second major classification of steelmaking furnaces is
    the EAF.  In 1991, EAF production accounted for about 37 percent
    of the total steel produced in the United States and Canada.7
    Almost all of the balance was produced in basic oxygen furnaces.
    
         Electric arc furnaces produce liquid steel primarily by
    melting steel scrap.  However, metallic iron,  including pig iron
    and direct-reduced iron,  are sometimes added to the charge.
    Generally, steel is produced in an EAF from a metallic charge of
    about the same composition as the steel to be made.1   Reactions
    taking place in the EAF are similar to those in the. basic oxygen
    furnace.  The charge is melted, impurities are oxidized, fluxes
    are added to aid in the formation of slag, and alloying elements
    are added to achieve the desired composition.
    
         There are several different variations of the EAF.  However,
    the direct-arc,  three-phase electric furnace is the most common.23
    In operation, the three-phase electric current flows from one
    electrode, through an arc between the electrode and charge,
    through the charge,  then through an arc between the charge and a
                                   3-13
    

    -------
    second electrode.  The charge is heated primarily by radiation
    from the arc although heat developed by current through the
    electrical resistance of the charge makes a slight contribution.
    Three electrodes are used in the EAF, and the current flow to and"
    from the electrodes is consistent with the ,features of the
    utility-based, three-phase power supply.
    
         Electric arc furnaces offer several advantages including low
    construction cost, flexibility in the use of raw materials (steel
    can be made directly from scrap without having a source of molten
    iron), the ability to produce steels over a wide range of
    compositions, improved process control, and the ability to
    operate below full capacity.  The improved process control makes
    the EAF well suited for stainless steel and other high-alloy
    steels.  Economic constraints tend to favor the use of EAF's for
    low to medium tonnage steelmaking facilities.1'24
    
         Molten steel, whether from a EOF or an EAF, is tapped from
    the furnace to a ladle and transported to a finishing process.
    Specialty steels, i.e., "clean" steels which meet extraordinarily
    stringent requirements of certain critical applications, may be  .
    subjected to additional processing in the ladle.  These processes
    are generally referred to as ladle metallurgy and as secondary
    steelmaking.  There are numerous ladle metallurgy processes
    including ladle temperature control, composition control,
    deoxidation, degassing, cleanliness control, and others.25
    
         3.2.1.4 Finishing.  Although there are many variations, most
    steel follows one of two major routes to final processing.  Both
    routes produce solid, semifinished products: ingots or cast
    slabs.    The more traditional route is the ingot route.  Molten
    steel is poured from the ladle into an ingot mold where it cools
    and begins to solidify from the outside toward the center.  When
    the ingot is solid enough, the mold is stripped away and the
    ingots are transported to a soaking pit or reheat furnace where
    they are soaked in heat until they reach a suitable, uniform
    temperature throughout.  The reheated ingots are then transported
    to a roughing mill where they are shaped into semifinished steel
    products, usually blooms, billets, or slabs.  Blooms are large
    and mostly square in cross section, and they are frequently used
    in the manufacture of building beams and columns.  Billets, which
    are made from blooms, are also mostly square, but are smaller in
    cross section and longer than blooms.  Billets are processed
    further to produce bars, pipes, and wire.  Slabs are the wide,
    semifinished product from which sheets, strip, and other flat
    rolled products are made.26  The roughing mills may be the first
    in a series of continuous mills that continue to shape the steel
    into more useful, more finished products.
    
         More recently, molten steel is increasingly routed to a
    continuous casting process that bypasses the soaking pit or
                                   3-14
    

    -------
    reheat furnace process associated with the ingot route.  In this
    process, the molten steel is lifted in the ladle to the top of
    the continuous caster where it is allowed to flow into a         .
    reservoir called the tundish and from the tundish into the molds
    of the continuous casting machine.  These molds are cooled with
    water so that a thin skin forms on the outside of the liquid
    metal.  Emerging from the mold, the continuous column of steel is
    further cooled by a water spray, causing the skin to thicken.
    The steel is further shaped by various designs of casters as it
    continues to flow.  As in the ingot process, the steel may be
    shaped into blooms, billets, or slabs and, subsequently, into
    more useful products.26  The continuous casting process has a
    significant advantage in that the soaking-reheating step is
    eliminated.  Thus, it offers a fuel savings and the potential to
    reduce undesirable emissions including NOX.
    
         Downstream of the series of mills (roughing stands and
    finishing stands) that shape the steel blooms, billets, slabs,
    and other semifinished products are a multitude of processing
    steps that may be used to produce finished steel products.  The
    actual processes followed may include a combination of many, few,-
    or perhaps none of this multitude of options.  The processes that
    may be encountered include pickling, annealing, galvanizing,
    coating, painting, and additional mechanical finishing steps.
    Some of these steps require additional heating or perhaps
    reheating and, thus, pose a potential for further combustion of
    fuels with the attendant potential for generating additional NO,
    emissions.  These include annealing and all reheating furnaces.
    
    3.3 PRODUCTION PROCESSES WITH NOX EMISSIONS
    
         In this section, iron and steel processes with a potential
    to produce NOX emissions  are discussed in detail.   Sintering,
    coke ovens, soaking pits, and reheat furnaces are identified as
    some of the more important iron and steel combustion processes
    and, thus, some of the largest sources of NOX emissions.
    
    3.3.1 Sintering
    
         The sintering and pelletizing processes are the two most
    important processes used to improve iron ore prior to making
    iron.  Of-these, pelletizing is usually accomplished near iron-
    ore mines where ore is ground to fine particles, separated from
    the gangue minerals either magnetically or by flotation, and then
    made into pellets.  This process is usually carried out at or
    near the mines to reduce the cost of shipping the beneficiated
    ore to iron and steel mills.  At iron and steel mills, a parallel
    process, sintering, takes fine iron-bearing materials recovered
    from ore handling and other iron and steel operations and  fuses
    these fine particles into materials suitable for charging to
    blast furnaces or direct reduction facilities to make iron.26
                                   3-15
    

    -------
         In sintering, the mixture of iron ore fines and other iron-
    bearing materials (iron-bearing dusts and slags) and fluxes (lime
    or dolomite) are thoroughly mixed with about 5 percent of a
    finely divided fuel such as coke breeze or anthracite and
    deposited on a traveling grate.  The traveling grate is shaped
    like an endless loop of conveyor belt forming a shallow trough
    with small holes in the bottom.  The bed of materials on the
    grate is ignited by passing under an ignition burner that is
    fired with natural gas and air.  Subsequently, as the grate moves
    slowly toward the discharge end, air is pulled down through the
    bed.  As the coke fines in the bed burn, the generated heat
    sinters, or fuses, the fine particles.  The temperature of the
    bed reaches about 1,300 to 1,480 °C  (2,370  to
    2,700 °F).   At  the discharge  end of  the  sintering machine,  the
    combustion will have progressed through the thickness of the bed.
    The sinter is then crushed to eliminate lumps, is then cooled,
    and finally screened.1'18
    
         Good sintering requires thorough mixing of the feed
    materials and careful deposition on the traveling grate so as to
    achieve a uniform bed.  Proper ignition is also important, and   _
    some processes replace part of the solid fuel with gaseous fuel
    to improve ignition and improve the sinter.  Plants using this
    process have approximately 25 percent of the length of the sinter
    bed covered with a gas-fired ignition-type hood.  Temperature in
    the hood ranges from about 1,150 °C  (2,100  °F)  where ignition
    begins to about 800 °C (1,500 °F) at the exit of the hood.
    Depending upon characteristics of the ore materials and sintering
    conditions, average production rates of 22 to 43 metric
    tons/m2/day (2.3  to  4.4  tons/ft2/day) of grate area are expected,
    and rates in excess of 49 metric tons/m2/day
    (5 net tons/ft2/day)  have  been attained.27
    
         The major source of energy used in the production of sinter
    is the carbon content of coke breeze and flue dust.  The amount
    of ignition fuel required is about 140 J/g  (0.12 MMBtu/ton) of
    sinter produced.  The total fuel requirement,  including coke
    breeze, is about 1.74 kJ/g (1.5 MMBtu/ton)  of sinter produced.27
    
         Beyond the sintering grate, the sinter is cooled so that it
    can be transported by conveyor belts.  The exhaust air from these
    coolers is normally at too low a temperature to permit economical
    recovery of heat.  Recent developments in sinter cooling have
    been directed toward on-strand cooling, which could improve heat
    recuperation, sinter quality, and dust collection.  The use of
    sinter in the blast furnace charge, and especially fluxed sinter,
    also improves blast furnace performance.  One improvement is  that
    less coke is required in the blast furnace.  Available data
    indicate that for each net ton of limestone removed from the
    blast-furnace burden and charged to the sinter plant to make
    fluxed sinter, approximately 182 kg  (400 Ib) of metallurgical
                                   3-16
    

    -------
    coke is saved.18  This also has a potential for reducing NOX
    emissions through fuel conservation.
    
    3.3.2 Cokemaking
    
         Coke is an essential component in the production of pig iron
    (hot metal) in a blast furnace.  Consequently, coke ovens are
    found in most integrated steel plants.  Coke ovens are
    constructed in batteries that contain as few as 10 to more than
    100 ovens.  Coking chambers in a battery alternate with heating
    chambers so that there is a heating chamber on each side of a
    coking chamber.  Underneath the coking and heating chambers are
    regenerative chambers, a brick checkerwork that preheats
    combustion air to conserve fuel, increase thermal efficiency, and
    give a higher flame temperature.  The ovens are constructed of
    refractory brick.
    
         In operation, selected blends of coal are charged to the
    ovens through openings in the tops.  Subsequently, the coal
    undergoes destructive distillation during a heating period of
    about 16 to 18 hours.  At the end of this portion of the coking  -
    cycle, doors at both ends of the oven are opened and the
    incandescent coke is pushed from the oven into a railroad-type
    car called a quench car.  The coke is then transported to the end
    of the battery to a quench tower where it is deluged with water
    to end the burning process.  The operation of each oven in the
    battery is cyclic, but the batteries usually contain a large
    number of ovens so that the yield of by-products is essentially
    continuous. The individual ovens are charged and discharged at
    approximately equal time intervals.  Practically all of the coke
    produced in the United States is made by this process.13
    
         The product of a coke battery is either furnace coke
    (requiring 15 to 18 hours of distillation) or foundry coke
    (requiring 25 to 30 hours of distillation).M The by-products of
    the process are driven off during distillation and are recovered
    in the by-product coke-making process through condensation. These
    by-products include tar, light oils, and heavy hydrocarbons.  The
    noncondensable gaseous product remaining is known as coke oven
    gas, which, on a dry basis, has a heating value of about 22
    MJ/Nm3 (570  Btu/ft3) .  Approximately 35 percent of the coke oven
    gas produced is used in heating the ovens.24'27
    
         Excess coke-produced in a cokemaking operation that is sold
    to others is referred to as merchant coke, and merchant coke can
    be either furnace coke or foundry coke.
    
         In 1983, about one-sixth of the total bituminous coal
    produced in the United States was charged to coke ovens.  On the
    average, 1.4 kg of coal is required for each kilogram of coke
    produced.  Each kilogram of coal carbonized requires 480 to 550
                                   3-17
    

    -------
     kJ (450 to 520 Btu)  of energy.
     1,480 °C (2,700 °F) ."
    Flue temperatures are as high as
          Current annual coke production in the United States is about'
     29  million tons.   Of this amount,  about 72 percent or 21 million
     tons comes from steel plant coke ovens (furnace coke)  with the
     balance from independent coke producers (merchant coke).  All of
     this coke is produced in 12 States,  and approximately 70 percent
     is  produced in five States: Pennsylvania,  West Virginia, Ohio,
     Indiana,  and Illinois.   The only nonrecovery coke-oven facility
     producing metallurgical coke in the United States is located in
     Virginia.28
    
          The  underfire combustion of fuel  in coke ovens produces NOX,
     among other emissions.   Because of the many different emissions
     from coke ovens,  coke ovens will be subj ected to more stringent
     environmental regulations in the 1990's and beyond.   Apart from
     these environmental concerns,  most coke ovens in the United
     States have reached obsolescence.   The International Iron and
     Steel Institute projects that approximately 40 percent of the
     coke ovens in North America must be replaced within the next 10
     years unless technology to extend their life is developed.   The
     costs of  replacing these facilities, already considerable,  are
     significantly increased by the costs of meeting increasingly
     stringent environmental regulations.   These issues are of much
     concern to the steel industry because  coke is an essential
     material  for the  integrated steel  producers.8'28
    
     3.3.3  Blast  Furnaces and Blast Furnace Stoves
    
          Iron ore is  converted to molten iron  in a blast furnace.   A
     charge consisting of iron ore,  sinter,  limestone,  and  coke  is
     charged to the top of the vertical-shaft blast furnace.   The coke
     provides  thermal  energy for the process and a reductant  gas  for
     the  iron-ore conversion,  and it serves as  the pathway  passage.for
     gases  that pass through the furnace burden.   The limestone
     becomes calcined,  melts,  reacts with,  and  partially  removes
     sulfur from  the molten  iron.
    
         Heated  air is  injected through tuyeres  near the bottom  of
     the  furnace  and moves upward through the burden,  consuming the
     coke carbon  and thereby providing  energy for the process.  Blast
     furnace gas  leaves  the  furnace  through offtakes  at the top of the
     furnace,  is  cleaned of  particulates, and subsequently  used as a
     fuel.  The BFG contains  about  1  percent hydrogen and about 27
    percent carbon monoxide,  and it  has a  heating value  that  ranges
     from about 2,540  kJ/Nm3  (65 Btu/ft3) to 3,600  kJ/Nm3  (92
    Btu/ft3) ,27'32  Between 2.2 and 3.5 kg of BFG is generated for each
    kilogram of pig iron produced  in the blast  furnace.27
    
         Molten  iron  and  slag are  tapped periodically  from the bottom
                                   3-18
    

    -------
    of the furnace."  In operation, the blast furnace is kept under
    pressure by an arrangement of pressure locks that function to
    seal the top of the furnace.  These locks are in the form of two ..
    inverted cones, called bells, which operate sequentially to
    maintain a seal in the furnace.  Thus, the blast furnace is a
    closed unit with no atmospheric emissions.27
    
         A blast furnace typically has about three blast furnace
    stoves associated with it.  These stoves primarily burn BFG in a
    process designed to preheat the air used to combust the fuel in
    the blast furnace they serve.  In an installation with three
    associated blast furnace stoves, for example, two are generally
    being preheated while the third is supplying preheated air to the
    furnace.  While being heated, hot gases from the blast furnace .
    are cleaned, cooled, and some portion is then routed to the
    stoves that are being heated where it is burned, often with other
    gases, to heat a thermal storage, regenerative checkerwork
    (refractory material) in the stove.  The heat thus stored is then
    used to preheat combustion air to the blast furnace.  The BFG may
    also be burned elsewhere in a steel plant.  The blast-furnace
    stoves require very large quantities of fuel for heating.
    However, because the stoves are heated primarily with BFG, the
    flame temperature is reduced, which reduces NOX generation.27
    
         The BFG burned in the blast furnace is often enriched by the
    addition of fuels with a. higher calorific value such as coke-oven
    gas or natural gas 'to obtain higher hot-blast temperatures, often
    in excess of 1,100 °C (2,010 °F) .  This enhances the efficiency
    and productivity of the process.19
    
    3.3.4 Basic Oxygen Furnaces       :
    
         Most of the steel produced in the United States is made
    using an oxygen steelmaking process.  There are three oxygen
    steelmaking processes: a top blown process called the BOP
    process, a bottom blown process called the Q-BOP process, and a
    third process, which is a combination of the other two.  In all
    three processes,  scrap and molten iron  (hot metal) are charged to
    the furnace and high-purity  (>99.5 percent) oxygen is used to
    oxidize excess carbon and silicon in the hot metal to produce
    steel.  The major differences between, the three processes are in
    the design of the furnace and the equipment for introducing
    oxygen and fluxes.21  These differences do  not  have any
    significant impact on the potential for NOX generation  and are
    not discussed at length here.  The BOP process is the most
    commonly used process in the United States.
    
         In the BOP process, oxygen is blown downward through a
    water-cooled lance onto a bath of scrap and hot metal.  Heat
    produced by the oxidation of carbon, silicon, manganese, and
    phosphorus is sufficient to bring the metal to pouring
                                   3-19
    

    -------
     temperature, and auxiliary  fuel  is not required.1
    
         Oxygen striking the  surface of the molten bath immediately  -.
     forms  iron oxide.  Carbon monoxide generated by  the reaction of
     iron oxide with carbon  is evolved, giving rise to a violent
     circulation that accelerates refining.  The oxidizing reactions
     take place so rapidly that  a 300-metric-ton heat, for example,
     can be processed in about 30 minutes.  The intimate mixing of
     oxygen with the hot metal permits this rapid refining.1   Figure
     3-2 is a schematic representation of the refining in a top-blown
     BOP.   It illustrates the  decreasing content of several
     undesirable elements in the molten metal bath during the period
     of an  oxygen blow.21
    
         Scrap use in a BOP furnace  is limited to about 30 percent of
     the charge.  If the scrap is preheated, the scrap content can be
     increased to about 45 percent.1  In addition to scrap preheating,
     other  fuels may be burned to dry out refractories and to keep the
     BOF furnace from cooling  between heats.  The use of fuel for the
     latter two purposes amounts to about 230 kJ/kg  (200,000 Btu/ton)
     of steel produced.                                               _
    
         About 470 kJ of carbon monoxide are produced in the BOF per
     kilogram of steel produced  (400,000 Btu/ton).  Typical practice
     is to  burn combustible  gases in  water-cooled hoods mounted above
     the BOF vessel.  In most  cases,  the BOF vessels are equipped with
     open hoods that admit air for combustion of carbon monoxide on a
     relatively uncontrolled basis.   Some new plants use suppressed
     combustion hoods, which do  not inspire air and burn off-gases.
     New BOF capacity is expected to  continue this trend, which may
     cause  a decrease in total NOX emissions.   During the combustion
     of the waste gas, the potential  for NOX production  exists.
    
     3.3.5  Soaking-Pit and Reheat Furnaces
    
         Soaking-pit and reheat furnaces are large furnaces used to
     raise  the temperature of  steel in the course of processing to a
     temperature suitable for  hot working or shaping.  They are
     designed to accommodate the steel being processed at a suitable
     rate,   heat it uniformly,  and hold it at a desired temperature for
     a specified length of time.  There are numerous design.
    variations.  Two major variations are batch-type and continuous -
     type furnaces.  Within  the  latter category are variations on how
     the charge is moved through the  furnace.   These variations
     include roller hearth furnaces in which the material moves as the
     series of rollers that  constitute the hearth rotate, walking-beam
     furnaces in which material  is moved in a controlled step-wise
    manner, pusher-type furnaces in  which a continuous line of
    material is pushed over skids, and rotary-hearth furnaces with
     circular hearths that rotate in  a horizontal plane.  Other
    variations involve the  introduction and removal of the charge and
                                   3-20
    

    -------
    8 40
    
    2
    O
    
    
    d 30
    ut
       20
     o
     o
     ffi
     e
       10
        0
               !\
          SV1
                                            ooe
    006 a
        oc
        o
        z
        A.
        v>
    004 O
    
        &
    
    
        ec
    
    002 z
        a.
         0  2  4  6 ' 8  10   12  14  16  18  20  22
    
                    TIME (MINUTES)
    Figure 3-2.  Schematic representation of progress
    
                 of  refining  in  a  top-blown basic-lined
    
                 BOF,21
                          3-21
    

    -------
    the circulation of heat.  The circulation of heat in fuel-fired
    soaking pit and reheat furnaces is accomplished by natural
    convection and stack draft.29
    
         The sizes of soaking-pit and reheat furnaces are usually
    described by their hearth area.  Soaking-pit furnaces range from
    about 100 ft2 to  over 300  ft2, and reheat furnaces range up to
    4,000 ft2 of  hearth area.29  The capacity of  these  furnaces  is
    determined primarily by the area of the surface of the material
    to be heated; the shape, thickness, and composition of the
    material; and the temperature of the material and the furnace.
    The thermal efficiency  (defined in this case as the amount of
    heat required to raise the temperature of the charge from its
    initial to its desired temperature as a percentage of the gross
    heat input to the furnace) also varies widely due to differences
    in the temperatures of the heated stock and of the charged
    material, provisions for heat recovery, furnace insulation,
    operating schedules, and heating requirements. Large production-
    line furnaces, such as continuous furnaces with recuperators and
    good insulation,  generally give over 30 to 40 percent thermal
    efficiency over an average month's operation.29
    
         Soaking-pit furnaces provide uniform heating of ingots to
    the desired temperature with a minimum of surface overheating.
    The normal range for heating ingots is between 1,180 °C and
    1,340 °C  (2,150°F and 2,450  °F) ,29  Soaking pits also function as a
    reservoir to correct irregularities in the flow of ingots between
    the steelmelting facilities and the primary rolling mills.
    
         The use of continuous casters is affecting the operations of
    both soaking pits and reheat furnaces.  There is less need for
    soaking pits as more steel is routed to reheat furnaces and, in
    general,  the steel is entering reheat furnaces at higher
    temperatures.  The net effect is that, in the aggregate, less
    fuel is being consumed in soaking pits and reheat furnaces.  In
    1991, about 76 percent of the steel produced was processed by
    continuous casting rather than forming ingots that subsequently
    needed to be processed in a soaking-pit furnace.7   Although the
    trend to continuous casting will continue, some soaking-pit
    capacity will likely be retained to cope with upset or irregular
    conditions.
    
         Reheat furnaces function to adjust or to maintain steel at a
    suitable temperature as the steel is processed.  They are similar
    in many respects to soaking-pit furnaces.  Reheat furnaces are
    being integrated into continuous casting facilities for slabs,
    blooms, and billets.  When these are charged hot into a reheat
    furnace,  a moderate fuel savings results.  Fuel inputs to large
    soaking-pit and reheat furnaces range from 1.2 to 5.4 MJ/kg (1.0
    to 4.7 MMBtu/ton) heated.27
                                   3-22
    

    -------
    3.3.6 Annealing Furnaces
    
         Flat-rolled steel products such as coils of sheet steel are
    sometimes annealed to enhance some physical properties of the
    product.  Annealing involves subjecting the product to a
    supplemental heat treatment under controlled conditions.  Only a
    portion of the steel produced is annealed.  It is usually cold-
    reduced coils that are annealed because hot-rolled coils are
    "self-annealed.n  The great bulk of annealing is done at
    temperatures of about 675 °C (1,250°F).  Some products are
    annealed at much higher temperatures[760 °C to  1,210  °C,
    (1,400 °F  to  2,200  °F) ] .30  Annealing is done in a protective,
    deoxidized atmosphere that is predominately nitrogen.
    
         Annealing can be a batch or a continuous process although
    the trend is strongly toward continuous annealing.  Batch
    annealing is done in a box-type furnace that consists of a
    stationary base, several stools on which coils of steel are
    stacked, individual cylindrical covers for each coil stack  (to
    provide for the protective atmosphere), and the furnace, which is
    lowered by crane over the base with its load, stools, and
    cylindrical covers.  Subsequently, the charge is heated slowly
    but uniformly to a specified temperature, soaked for a period of
    time, and then allowed to cool.  After a period of cooling, the
    furnace is removed to begin a cycle on another base.  However,
    the inner covers are left in place to preserve the protective
    atmosphere.  After the charge has cooled to about 150 °C
    (300 °F),  the charge can be exposed to air without oxidizing.  In
    this cycle, the cooling period takes at least as long as heating
    and soaking combined.
    
         Continuous annealing is done in large furnaces in which the
    steel coil is threaded vertically around rollers located at the
    top and bottom of the furnace.  Thus,  the residence time of the
    steel in the furnace is dramatically increased as it passes
    continuously through the furnace.  A typical, continuous
    annealing furnace will have several zones including a gas-fired
    heating zone, an electrically heated holding zone, an
    electrically heated slow cooling zone, and a fast cooling zone.
    Steel coil will thread through these zones at a rate of about
    1,500 ft/min (7.6 m/s).   Threading back and forth, the steel will
    make 10 passes through the heating zone in about 20 seconds.
    Subsequently, it will be cooled to about 540 °C (1,000 °F)  in the
    slow cooling zone and then to 115 °C  (240 °F) in the fast cooling
    zone.  The entire process takes about 2 minutes and is carried
    out in an atmosphere of nitrogen (95 percent) and hydrogen
    (5 percent) .31
                                   3-23
    

    -------
    

    -------
    3.4  References
    
    1.   Kirk Othmer Encyclopedia of Chemical Technology.  New York,
         John Wiley.  3rd ed.  1981.  Vol. 13.  p. 552.
    
    2.   Lankford, W. T. ,  N. L. Samways, R. F. Craven, and H. E.
         McGannon.  The Making, Shaping and Treating of Steel, Tenth
         Edition.  Pittsburgh, Herbich & Held.  1985.  p. 256.
    3.
    4.
    .5.
    6.
    7.
    Ref .
    Ref .
    Ref.
    Ref.
    Labee
    2
    2
    2
    2
    /
    / P
    / P
    / P
    / P
    C.
    . 257.
    . 541.
    . 24.
    . 599.
    J. , a
    8.
    
    
    
    
    9.
    
    
    10.
    
    
    11.
    
    
    
    
    12.
    
    
    13.
    
    
    14.
    
    
    
    
    
    15.
    
    16.
    
    17.
    Labee, C. J., and N. L. Samways. Developments in the Iron
    And Steel Industry, U.S. and Canada-1991. Iron and Steel
    Engineer.  February 1992.
    
    Labee, C. J., and N. L. Samways. Developments In The Iron
    And Steel Industry, U.S. and Canada-1990. Iron and Steel
    Engineer.  February 1991.
    
    Directory, Iron and Steel Plants.  Association of Iron and
    Steel Engineers.  Pittsburgh, PA.  1991.
    American Iron and Steel Institute.
    Report 1990.  Washington, DC.
    Annual Statistical
    Telecon, Parker C., Research Triangle Institute, with
    Steiner, B., American Iron and Steel Institute.  February
    12, 1992.  Iron and steel industry
    
    Parkinson, G.  Steelmaking Renaissance.  Chemical Engineer.
    May 1991.
    
    Usher, Thomas J. Steel Industry in the Nineties. Iron and
    Steel Engineer. February 1991.
    
    U.S..Environmental Protection Agency.  Benzene Emissions
    from Coke By-Product Recovery Plants - Background
    Information for Proposed Standards.  Research Triangle Park,
    NC.  Publication No. EPA-450/3-83-0162.  May 1984.
    
    Ref. 2, p. 156.
    
    Ref. 2, p. 325.
    
    Ref. 2, p. 326.
                                   3-24
    

    -------
    18,
    
    19.
    
    20.
    
    21.
    
    22,
    
    23.
    
    24.
    25
    
    26.
    
    
    27
    
    
    
    28
    
    
    
    
    29
    
    30
    
    31
    
    32
    Ref. 2,
    
    Ref. 2,
    
    Ref. 2,
    
    Ref. 2,
    
    Ref. 2,
    
    Ref. 2,
                 P-
    
                 P-
    
                 P-
    
                 P-
    
                 P<
    
                 P-
    307.
    
    557.
    
    507.
    
    612.
    
    611.
    
    627.
    Stallings, R. L. (Research Triangle Institute).  Organic
    Emissions from Ferrous Metallurgical Industries:
    Compilation of Emission Factors and control Technologies.
    Prepared for U.S. Environmental Protection Agency.  Research
    Triangle Park, NC.   Publication No. EPA-600/2-84-003.
    September 1983.  108 pp.
    Ref. 2, p. 671.
    
    American Iron and Steel Institute.
    Washington, DC. February 1989.
                                             Steelmaking Flowlines.
         U.S. Environmental Protection Agency.  Control Techniques
         for Nitrogen Oxides from Stationary Sources -Revised Second
         Edition.  EPA-450/3-83-002 .  January 1983.  pp 5-75, 76.
    
         Prabhu, D. U. , and P. F. Cilione.  1990 Clean Air Act
         Amendments:  Technical/Economic Impacts.  U.S. Coke and
         Steelmaking Operations.  Iron and Steel Engineer. January
         1992.
    
         Ref 2., pp. 841-843.
    
         Ref. 2, p. 1111.
    
         Ref. 2, p. 1116
    
         Letter and attachments from Finke, H. P., Bloom Engineering
         Company, Inc. to Neuffer, W. J., EPA/OAQPS/ISB.  May 14,
         1993.
                                   3-25
    

    -------
                                CHAPTER 4
                        UNCONTROLLED NOX EMISSIONS
         This chapter presents available information  on uncontrolled
    NOX emissions from process facilities at iron and steel mills.
    Section 4.1 discusses the mechanisms of NOX formation.  Section
    4.2 discusses emission factors and emissions  from specific
    sources, and Table 4-4 presents a summary of  emissions  from these
    sources.  Appendix A is a tabulation of the available emissions
    data.
    4.1  MECHANISMS OF NOX FORMATION
    
         Nitrogen oxides refer to the combination of nitric  oxide  and
    nitrogen dioxide.  However, flue gas resulting from the
    combustion of fossil fuel consists primarily of NO  with  NO
    representing 90 to 95 percent of the total NOX due to kinetic
    limitations in the oxidation of NO to NOj.1 There are three
    fundamentally different mechanisms of NOX formation. These
    mechanisms are (1) thermal NOX,  (2)  fuel NOX, and  (3)  prompt  NOX.
    The thermal NOX mechanism arises from the thermal dissociation
    and subsequent reaction of nitrogen and  oxygen molecules in
    combustion air. The fuel NOX mechanism is the evolution and
    reaction of fuel-bound nitrogen compounds with oxygen. The prompt
    NOX mechanism involves  the intermediate formation of hydrogen
    cyanide, followed by the oxidation of HCN to NO. Natural gas and
    most distillate oils have no chemically  bound fuel  nitrogen  and
    essentially all NOX formed from the  combustion of these fuels is
    thermal NOX.  Residual  oils and coals  all have fuel-bound nitrogen
    and,  when these are combusted, NOX is formed by all three
    mechanisms. The formation of prompt NOX is only significant in
    very fuel-rich flames. These three mechanisms are discussed  in
    more detail in the following subsections.
    
    4.1.1  Thermal NOX Formation
    
         At the temperatures encountered in  combustion  air,  both N2
    and O2  molecules  are dissociated into their respective atomic
    states, N-and O.  The subsequent reaction of these atoms  to create
    thermal NOX is  thought  to proceed through mechanisms first
    formulated by Zeldovich:1'2
    
         N2 +  0 <*  NO + N
    
         N + O2 5^  NO + 0
                                      2  x 1014 exp (-76500/RT)   (4-1)
    
                                  kf=  6.3 x 109 exp  (-6300/RT)   (4-2)
    In each reaction, kf is  the  forward rate constant for that
    reaction. The high activation energy of  reaction (4-1),  76.5
    kcal/mol, shows that this reaction  is the most temperature
    sensitive.  These relationships assume  that the  combustion  has
    

    -------
    reached equilibrium.  Because  it has a  large  activation  energy,
    reaction  (4-1) is generally believed to be  rate  limiting.   Oxygen
    atom concentrations are assumed to have reached  equilibrium
    according to
                                                                 (4-3)
    where M denotes any third substance  (usually N2) ,
         Experiments at atmospheric pressure have  indicated  that,
    under certain conditions, the amount of NO  formed  in heated N2/
    O2/  and argon (Ar)  mixtures can be expressed as3
    [NO] =
                             [N2] [02]1/2t
    (4-4)
    where
          [NO]      = mole fraction
    
         kk = constants
    T
    
    t
                     temperature, and
    
                     time.
    Although equation  (4-4) cannot adequately describe NO  formation
    in a turbulent flame, it does point out several  features  of
    thermal NOX formation.   Like reactions (4-1)  and (4-2) , it
    reflects a strong, exponential relationship with NO  emissions and
    temperature. It also suggests that NO formation  is directly
    proportional to N2 concentration,  residence time,, and the square
    root of 02 concentration.   .Moreover,  equation (4-4)  also suggests
    several control measures for reducing thermal NOX formation:
    
         • Reduce local nitrogen concentrations at peak  temperatures,
    
         • Reduce local oxygen concentrations at peak temperatures,
    
         • Reduce the residence times at peak temperatures, and
    
         • Reduce peak temperatures.
    
    These control measures are discussed further in  Chapter 5 .
    
         For the purposes of the computations presented  in Table 4-1,
    the N2 and 02 concentrations in flue gas are defined  to be 76
    percent and 3.3 percent, respectively.  The strong,  exponential
    dependency on temperature  is evident in these tabulations, and a
    dependency on 02  concentration can be inferred.   Note,  for
    example, that the equilibrium NO, is  much higher in air where the
    02  concentration  is higher  than it is in flue gas.
                                   4-2
    

    -------
    TABLE  4-1. CALCULATED EQUILIBRIUM CONCENTRATIONS OF NO AND N02 IN
                           AIR AND FLUE GAS4(ppm)
         Temperature
      K
                      Air
               NO
                                  Flue gas
                    NO,
               NO
                   NO,
       300
    
       800
    
     1,400
    
     1,873
       80
    
      980
    
    2,060
    
    2,912
        3.4{10)-10
    
        2.3
    
      800.
    
    6,100.
     2.1(10)"
    
     0.7
    
     5.6
    
    12.
        I.KIO)'10
    
        0.8
    
      250.
    
    2,000.
    3.3(10)-3
    
    0.1
    
    0.9
    
    1.8
                                     4-3
    

    -------
         In practice, flue gas NOX concentrations  tend to be much
    higher than suggested by the computations in Table 4-1.  After
    NOX is  formed  in the  high temperatures  of a  flame,  the rate  of
    its decomposition (the reverse of reactions (4-1) and  (4-2)) is
    kinetically limited at the lower temperatures and lower N and 0.
    atomic concentrations characteristic of the postcombustion zone
    of the flame.   Thus,  although NO, is  thermodynamically unstable
    even at high temperatures, its decomposition is kinetically
    limited. The result is that the NOX concentration in flue gas is
    higher than  predicted by equilibrium and depends, to a large
    extent, on the mixing of fuel and combustion air in the flame.
    
         The factors discussed above affect thermal NOX formation on
    a macroscopic scale.  However, local microscopic conditions
    ultimately determine the amount of thermal NO, formed. These
    conditions are intimately related to a host of variables such as
    local combustion intensity, heat removal rates, and internal
    mixing effects.
    
         Studies on the formation of thermal NO, in gaseous flames,
    for example, have confirmed that internal mixing can have large
    effects on the total amount of NOX formed.3  Burner swirl,
    combustion air velocity, fuel injection angle and velocity,  quarl
    angle,  and confinement ratio all affect the mixing between fuel,
    combustion air, and recirculated products. Mixing, in turn,
    alters the local temperatures and specie concentrations, which
    control the rate of NOX formation.
    
         Generalizing these effects is difficult because  the
    interactions are complex.  Increasing swirl, for example, may
    both increase entrainment of cooled combustion products  (lowering
    peak temperatures) and increase fuel/air mixing  (raising local
    combustion intensity). Thus, the net effect of increasing awirl
    can be to either raise or lower NOX formation.   In summary,  a
    hierarchy of effects, depicted in Table 4-2, produces local
    combustion conditions that promote thermal NOX formation.3
    
    4.1.2  Fuel N0r Formation
    
         Nitrogen oxides can also be produced by oxidation  of
    nitrogen compounds contained in fossil fuels.  The mechanisms are
    more complex than the Zeldovich model for thermal NOX formation.
    Several studies indicate that the fuel-bound nitrogen compounds
    react to form NOX in  two separate mechanisms,  one a solid-phase
    char reaction  (with solid fuels) and the other a homogeneous gas-
    phase reaction resulting from evolution and cracking  of volatile
    compounds  (solid and liquid fuels).1
                                   4-4
    

    -------
    
    
    
    
    
    m K
    O
    s
    
    >Jj
    H 0
    03 r^
    E M-4
    -H
    M 73
    Pi pj
    nJ
    0 JJ 0
    V-i nj SH
    4J 0 4J
    rH nj g nj
    0 >H -H >H
    > 0 4J 0
    d) pj p[
    • >~i € d) £s
    dj j-i Q}
    a 4J 3 4J
    0 CQ
    CD ^ O 4«i
    >i rt ft rt
    X 0 X0
    O ft H ft
    
    
    
    
    
    
    
    >i a o a
    4J O 0 -H O
    •rl -H g 4J -H
    m 0 jj 3 m jj
    C JJ CO rH Vi SH
    0 rt 3 H i (H JJ
    fi g O -H CO 0
    •H rH O JJ ft) -H fl
    ns u e -v. 73 o
    « > -H rH N
    O O H C SH co
    3 O 0 M 4J
    4J C! -H JJ -H 0
    (C U) JJ JJ (t rH
    >H -H -H (C ft CQ rH
    0 co CQ ft73 O0 JH
    ft 0O C r( rH -H
    g 73 ft £3 (d 73 U S
    0 >i gO -H CO
    JJJJ X O -H i-H M-l JJ
    -H O O 4J0 O5-I 5-1
    4J U X3 U3 rt 0
    0O 0 rH 01W 0ft Cl
    rH rH >-| 0 -n N >H
    C0-H 3 C! «H -H M 3
    H>[X( Cu HO COO «
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    CO
    CO
    
    
    rH G
    nj O
    C -H
    M 4-J
    0 -H
    4J 73
    X73
    cq nj
    4-5
    

    -------
         The char nitrogen reaction is not well understood although
    data show that the char nitrogen conversion to NOX is weakly
    dependent on the flame temperature and strongly dependent on
    stoichiometric ratio, fuel/air mixing, and on the char
    characteristics.1"3  The precise relationships,  however, are not
    known. Conversion rates to NOX of  15  to 35 percent have been
    documented.1
    
         The gas-phase reaction is postulated to include  a number of
    intermediate species  (e.g., HCN, HOCN, NHj)  which are produced at
    rapid reaction rates.  The decay rate of the intermediate species
    into N2 (fuel-rich)  and NOX  (fuel-lean) is slower by at least an
    order of magnitude.  These reaction rates are strongly dependent
    upon the stoichiometric ratio and the gas phase fuel  nitrogen
    concentration and weakly dependent upon the flame temperature and
    the nature of the organic nitrogen compound.  It is the weak
    influence of temperature on gas-phase NOX conversion that reduces
    the effectiveness of NOX controls  that rely on temperature effects
    in the combustion of nitrogen-bearing fuels.1
    
         The relative contribution of fuel NOX and thermal NOX to total
    NOX emissions  from sources  firing  nitrogen-containing fuels has net
    been definitively established.  Estimates indicate that fuel NOX is
    significant and may even predominate.3  In one laboratory study,
    residual oil and pulverized coal were burned in an argon/oxygen
    mixture to eliminate thermal NOX effects.   Results show that fuel
    NOX can account  for over 50 percent  of total NOX production  from
    residual oil firing and approximately
    coal firing.5
               80  percent  of  total
            NOX from
         The nitrogen content of most residual oils varies  from  0.1  to
    0.5 percent.  Nitrogen content of most U.S. coals lies  in  the  0.5
    to 2 percent range.  Fortunately, only a fraction of the fuel
    nitrogen is converted to NOX for both oil and coal firing.3
    Furthermore, the percent of fuel nitrogen converted to  NOX ,
    decreases as nitrogen content increases.  Thus, although fuel  NOX
    emissions undoubtedly increase with  increasing fuel nitrogen
    content, the emissions increase is not proportional.  In fact,
    data indicate only a small increase  in NOX emissions as  fuel
    nitrogen increases.3
    
    4.1.3  Prompt NOX Formation
    
         Prompt NOX is produced by the formation first of intermediate
    HCN via the reaction of nitrogen radicals and hydrocarbons,
              NO + HC + H,
    HCN + H20
    (4-5)
    followed by the oxidation of the HCN to NO.  The formation  of
    prompt NOX has a weak temperature dependence and a short lifetime
    of several microseconds.  It is only significant in very  fuel-rich
    flames, which are inherently low-NOx emitters.1
                                    4-6
    

    -------
    4.2 NOX  EMISSIONS  FROM IRON AND  STEEL MILLS
    
         Integrated iron and steel mills import three basic raw
    materials and process these to produce steel in a variety of
    compositions and forms.  The conversion processes, described in
    Chapter 3 of this document, tend to be high-temperature processes
    in which large quantities of fuels are consumed.  As a
    consequence, most of these processes are sources of NOX emissions.
    
         Mini-mills and specialty producers do not have the full range
    of processing facilities characteristically found in integrated
    plants.  They do not, for example, produce coke or sinter for use
    in blast furnaces or operate blast furnaces to produce pig iron or
    hot metal.  Instead, they may produce steel by melting scrap in an
    electric-arc furnace, and they may use some subset of the
    processes found in integrated mills to produce a variety of steel
    products. They, too, may use a number of high-temperature, fuel-
    consuming processes and produce NO, emissions.   In this section,
    the major sources of NOX emissions inherent  in steelmaking plants
    are identified along with the fuels and process temperatures
    characteristically used in these processes.  Further, estimates of
    uncontrolled NOX emissions are also given.
    
         The following subsections discuss uncontrolled NOX emission
    factors and NOX emissions from specific iron and steel process
    facilities.  These data appear in the literature or were provided
    by Section 114 responses for a variety of units, many of which
    cannot be easily  interpreted for other conditions.  These data are
    summarized in tabular form in Section 4.2.11 and are tabulated in
    Appendix A.
    
    4.2.1  Emission Factors
    
         NOX emissions data in the literature relevant to iron and
    steel mill processes are limited.  This lack of a good  emissions
    database is reflected in the major compilations of emission
    factors.6-8  Table 4-3 is a compilation of  emissions  factors from
    AP-426-8 and from the National  Acid Precipitation Assessment Program
     (NAPAP) .7>s  Note that the units used for these factors are unusual
    and not easily  interpreted or used.  Also, note that  their quality
    ratings are poor  (D  for AP-42 and E  for NAPAP), i.e.,  they are
    based on a single observation of questionable quality  or  an
    extrapolation from another factor for a similar process.   If  these
    emission factors  are  to be used,  it  has been  recommended  that
    AP-42 factors be  used whenever possible and that  the  NAPAP
    emission factors  be  used when AP-42  factors are unavailable.6
    These emission  factors  are not used  in  this document.
    
    4.2.2   Coke-oven  Underfiring
    
         As described in Chapter  3,  coke is made  by  the  destructive
    distillation of  coal in the absence  of  air.   During  distillation,
                                     4-7
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    1
    (Sj
    a
    
    Z
    S
    N
    I
    S
    u
    s
    0!
    b
    2
    <
    b
    2
    0
    H
    CO
    CO
    H
    W
    a
    H
    g
    53
    w
    8
    g
    H
    2
    
    1
    ^*
    H
    r3
    I2C
    pj
    "^
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    m
    u
    •rl
    c
    
    
    
    
    
    
    
    J?
    4J
    •H *-i
    C PJ
    cT^ls
    rH £5
    
    
    
    
    
    
    
    ;.
    -rl C-3
    !z *^» 0j
    35
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    m
    tion/proces
    S
    o
    
    o
    
    
    
    T^ T3 'O *O T5 *O
    O Q) 0) 0) 0) Q)
    0> 01 0) 01 01 01
    *O (fl (tJ flj (0 fC (0
    (U ,cj pC »C 1^1 .C J2
    c u o o u u o
    3 pH iH rH i-l i-l iH
    »Q rt (fl rt nj nj nJ
    o o o o o o
    •Jj U U O O U O
    10 CO OQ CQ CQ 02
    v C C C! C C C
    o O O O O O O
    rH EH EH EH EH EH H
    
    
    
    
    
    
    
    
    
    O O PO ^l H
    • • O VO O O
    CO O ... .
    04 00 O O O 0
    
    
    
    
    
    
    
    
    ro H '
    O O
    0 0
    
    
    
    
    
    
    
    
    
    
    
    
    
    Cn
    14-1
    e
    m
    m j-i
    ^ (fl
    O Cn Q)
    U C OJ rH
    ffl cn-ri M
    (fl 4J Cn C Vi (fl a)
    Cn uc cn-H -H Q) T3
    3 -H a J3 >t-i I-H -H
    to n3 Cn-H u v< to
    CO O l-l J3 C (I) W i
    d)OO VinStaaiTJOft
    U to O &J5 3 3 C 0 0
    O CQ tj ' t^ Oi ^y D O ^
    r) >,
    D< ffl
    •o
    0)
    U
    3
    •5
    o
    a
    
    ^4
    0)
    JJ
    c
    •H
    m
    m
    1
    
    
    
    
    
    
    
    
    
    
    f>
    o
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ffg
    TJ>8
    a) c
    JJ-H
    C S
    -H
    CO
    
    
    
    
    
    
    •a TI -a TJ -a T) -a
    o)  a) a) d) a)
    u u u u u u u
    3333333
    *O *O *O *o *o *o *t3
    o o o o o o o
    Di Qi Qi Oi Oi d ft
    Illiiil
    H EH EH EH EH EH EH
    
    
    
    
    
    
    
    
    
    00 C4 in
    rH O)O O O H O\
    o Ss o o o o H
    
    
    
    
    
    
    
    
    
    
    
    
    I
    rH
    «1
    u c
    2 g_
    o T
    » o
    10 4J
    " AJ tt) 0)
    0) U U
    U (fl (fl -
    m u en c u
    c a a C c
    fi I -H 3 -H
    3 t) 4J «4H N
    C 44 O tD
    O O b (fl D) •>
    •H 0 m J3 O 0 C C
    4J rl 4-> (I) -rl -H
    U (fl -rl C O 4J 1J
    3 ft d) 3 (fl —
    •o u a •• o d)
    OTHOiOtn3lHCn
    rl W C C C 4J C
    ft jj -H -H -iH i -H
    d)rHOO<80d)O
    0) b3 CO CO E^ CJ tC CJ
    4J
    CO
    
    
    
    
    
    
    
    
    
    
    
    
    MH
    tt t Q
    o
    SO
    -H
    -H JJ
    4J Rl
    r m 5} co
    m ra S °>
    m W n d)
    •§ 8 ° 8
    d) O,8,&&?
    ^SsSt
    rH LJ TO Tj 01
    2 fl) S« Q) "jl
    O1 ^ jj
    o o S
    r* i
    K,^ H S^ *^
    «w jj «3 .j
    "H -rl "
    1 •«&««
    •C Qfr£3 Qi O
    S «S «j-S
    s3s«H-
    0 g 0 g £
    o o
    L^ U1 L| & S
    
    
    « .O 0
    4-8
    

    -------
    a noncondensable gaseous product known as coke-oven gas  (COG),
    with a, heating value of about 22 MJ/Nm3  (570  Btu/f t3) , is evolved.
    The major combustible component of COG is methane; thus, it is
    very similar to natural gas.  Some 30 to 40 percent of_the COG is
    used to heat the coke oven, and the remainder is used in other
    heating processes in the plant.  Coking is accomplished at
    temperatures that range from about 900 to 1,100 °C (1,650 to 2,010
    °F) ,  and flue-gas  temperatures  can reach about  1,480°C (2,700 °F) .
    A regenerative, reversible checker-work is used to preheat the
    combustion air used to burn the COG, yielding fuel conservation
    and a higher flame temperature.  Nitrogen oxides emissions from
    coke underfiring result from the high temperatures and the large
    quantity of fuel consumed.9  Uncontrolled NOX emissions reported
    for coke ovens are tabulated in Appendix A.
    
         The wide range of NOXemissions from heating of coke ovens is
    due to wide variations in the geometry of the combustion chambers
    and the combustion variables of fuel and air mixtures,
    temperature, humidity, and other factors.  While the principle of
    all coke ovens is the same, each coke oven and coke-oven operation
    is unique.
    
         Much of the coke oven data included in Appendix A is
    incomplete such that it cannot be compared on a common basis with
    other data.  The data that can be compared on a common basis  (ppm
    @ 3% 02,  Ib/MMBtu,  and Ib/ton of product)  are summarized below.
    n(No. Samples)
    Min. Value
    Max. Value
    Avg. Value
    Std. Dev.
    ppm @ 3% 0;
    
        8
      254
     1452
      802
      448
    Ib/MMBtu
    
      11
       0.10
       2.06
       0.66
       0.72
    Ib/ton
    
     12
      0.15
      2.15
      0.88
      0.64
    4.2.3  Sintering
    
         Sintering is a continuous operation in which  feed materials,
    including coke breeze or fines and flue dust, are  deposited in a
    uniform bed on a traveling grate and  ignited by passing under
    burners fired with natural gas.  Beyond the ignition  region,  air
    is pulled,down through the bed.  The  coke  fines in the bed
    continue to burn, generating the heat required for the sintering
    to proceed.  Ignition temperatures range from about 1,150 °C
     (2,100 °F) ,  where' ignition begins,  to about 800°C   (1,500°F)  at the
    ignition hood exit. As the coke fines in the bed burn, bed
    temperatures reach 1,300 to 1,480 °C  (2,370 to 2,700 °F) .
    Available uncontrolled NOX emissions data for sinter plants are
    tabulated in Appendix A.  These data  vary  widely,  and much  of
    these data are incomplete, i.e., test conditions are  seldom
    specified.  A summary of these data is tabulated below.  The data
    in terms of Ib/ton show a wide range  from  0.007 to 0.70.
                                     4-9
    

    -------
                      ppm @ 3%
                   Ib/MMBtu
                  Ib/ton
    n(No. Samples)        2
    Min. Value          284
    Max. Value          395
    Avg. Value          339
    Std. Dev.
                      4
                      0.22
                      0.59
                      0.47
                      0.17
                  16
                   0.007
                   0.70
                   0.34
                   0.28
    4.2.4  Blast Furnaces and Blast Furnace Stoves
    
         The blast furnace is fueled by the combustion of coke that,
    along with iron ore and flux, is a component of its charge, and
    the off-gas is recovered and burned for its fuel value. It is a
    closed unit that does not have atmospheric emissions.  However,
    the associated blast furnace stoves do generate NOX emissions.
    These stoves, typically three or four for each blast furnace, burn
    the blast furnace gas  (BFG) to preheat the combustion air supplied
    to the blast furnace.  The BFG has a low heating value ranging
    from 3.0 to 3.5 MJ/Nm3  (80  to 95  Btu/ft3) .9 More recently,  values
    ranging from a low of 2.5 MJ/Nm3  (65  Btu/ft3)  have been reported
    with changes in blast furnace burdens cited as the cause of the
    deterioration.16  Blast furnace gas burns with a low flame
    temperature.  Its combustible component is carbon monoxide, which
    burns clean.6  However,  the BFG is  often enriched with other fuels
    (e.g., natural gas or COG)  to obtain blast-air temperatures as
    high as 1,100 °C (2,010 °F) .3
    
         The available uncontrolled NOX emissions  data for blast
    furnace stoves are tabulated in Appendix A.  These data vary
    widely in terms oflb/MMBtu and Ib/ton, and much of these data are
    incomplete, i.e., test conditions are seldom specified.  A summary
    of these data is shown below.
    n(No. Samples)
    Min. Value
    Max. Value
    Avg. Value
    Std. Dev.
    ppm @ 3% 0;
    
        1
    
    
       28,
    Ib/MMBtu
    
      11
       0.00.2
       0.057
       0.021
       0.019
    Ib/ton
    
     10
      0.003
      0.072
      0.037
      0.022
    4.2.5  Basic Oxygen Furnace
    
         The basic oxygen process of steelmaking is an exothermic
    process that does not require the burning of fossil fuels.  The
    process does produce off-gases, principally carbon monoxide, which
    are usually burned prior to discharge into the atmosphere.  Some
    newer BOF facilities use suppressed combustion hoods, which
    suppress the influx of air, and the off-gases are subsequently
    flared.  A potential for NOX generation does  exist during the
    combustion of waste gases.  Nitrogen oxides production from this
    process has been estimated to range from 30 to 80 ppm, or 180 to
    500 ng NOx/kg  (0.36  to  1  Ib N0x/ton) of steel produced.10
    

    -------
         Appendix A presents NO,emissions data for BOFs, including
    Q-BOPs, and these data are summarized below.  Data for which the
    status of the furnace is unknown  (i.e., whether or not the furnace
    is in an 02blow period)  are not included in the summary.  A wide ._
    range in terms of ppm@ 3%02 is  reported (18-180)  but in terms of
    Ib/ton the range is much narrower  (0.042-0.22) .
    n(No. Samples)
    Min. Value
    Max. Value
    Avg. Value
    Std. Dev.
                           During 0*, Blow Period
    ppm (aver)
      12
      12.3
      84.0
      24.0
      19.6
    ppm @ 3% 0?
        7
       18
      180
       58
       56
    Ib/ton
      7
      0.042
      0.222
      0.119
      0.059
    n(No. Samples)
    Min. Value
    Max. Value
    Avg. Value
                         Durincr Non  0, Blow Period
    ppm  (aver)
       2
      14.3
      14.5
      14.4
    ppm @ 3% 0?
         2
       200.
       366.
       283.
                                                   Ib/ton
    4.2.6  Electric-Arc Furnace
    
         The electric arc furnace largely  transfers  the  generation of
    NOX emissions from the steelmelting facility to a utility plant
    where it is  easier to control.  The only use  of  fossil  fuels  in
    the electric arc facility  is for  scrap preheating, which may  or
    may not be practiced.  However, some EAFs  also fire  oxy-fuel
    burners in addition to electric arcs during meltdown.17'  The
    available NOX emissions data,  presented in Appendix A and
    summarized below, suggest  that concurrent  oxy-fuel firing during
    meltdown does  increase NOX emissions above the emissions from
    electric-arc melting alone as NOX increased from 12 to 98 ppm.  The
    range of NOX emissions is narrow from 0.5  - 0.6 Ib/ton and  83  - 100
    Ib/heat.
                     EAF With Concurrent  Oxy-fuel  Firing
     n(No.  Samples)
     Min. Value
     Max. Value
     Avg. Value
     Std. Dev.
     ppm (avg.)
         6
        80.
       110.
        98.
        10.
     Ib/ton
      6
      0.50
      0.60
      0.54
      0.05
     Ib/heat
       6
      83
      100
      89
       8.2
    

    -------
                   EAF Without Concurrent Oxy-fuel Firing
     n(No.  Samples)
     Min. Value
     Max. Value
     Avg. Value
    ppm (avg.)
        2
        7
       17  '
       12
     4.2.7  Soaking  Pits
    
         Soaking pits  reheat  or hold slabs,  ingots, and other forms of
     steel  to temperatures suitable  for further shaping and processing.
     They predominantly use natural  gas for fuel although some COG and
     fuel^oil are used.  In recent years, the use of soaking pits has
     declined as the use of continuous casters increased.
    
         The available NOX emissions data for soaking pits  are included
     in Appendix A.  These data correspond to a range of conditions
     that are not always specified.  A summary of these data are
     tabulated below:
    
              Summary  of all Appendix A Soaking Pit Data
         n(No. Samples)
         Min. Value
         Max. Value
         Avg. Value
         Std. Dev.
         Ib/MMBtu
           5
           0.064
           0.148
           0.108
           0.039
    Ib/ton
    9
    0.091
    0.361
    0.175
    0.078
    ppm @ 3% 0,
    3
    49
    689
    307
    337
         A sub-set of the Appendix A data for soaking pits includes
    more^test-condition information than is available for the
    remainder of the data.  It is clear that these data correspond to
    a complete cycle of the soaking pits.  Hot and cold charges are
    specified and the temperature of the combustion air is given.  A
    summary tabulation corresponding to this sub-set is presented
    below, and a distinction is made between soaking pits with
    preheated combustion air and those with cold or ambient combustion
    air.  NOX for the preheat  air is approximately  doubled  that  for  the
    cold air soaking pit.                                        '
    
         Summary of a Sub-set of Appendix A Soaking Pit Data
                            Ib/MMBtu
                      Cold Air
         n(No. Samples  2
         Min. Value     0.064
         Max. Value     0.069
         Avg. Value     0.066
         Std. Dev.
               Preheat Air
                3
                0.121
                0.148
                0.135
                0.016
      Ib/ton
    Cold Air
      2
      0.091
      0.104
      0.098
    Preheat Air
       3
       0.164
       0.194
       0.164
       0.016
    

    -------
    4.2.8 Reheat Furnaces
    
         Most reheat furnaces are either recuperative- or
    regenerative-fired, i.e., they preheat the combustion air in order
    to increase fuel efficiency.  Some use cold combustion air.  The '•
    temperature of the combustion air has a large impact on
    uncontrolled NOX emissions.   Increasing  the  combustion  air
    temperature from 38 °C (100 °F) to 540° C (1,000 °F),  for example,
    will increase uncontrolled NOX emissions by  a  factor of about  6.
         Combustion air preheated in regenerators has a much higher
    temperature than air preheated in recuperators.  While the higher
    combustion air temperature increases furnace fuel efficiency,  it
    also increases NOX generation and NOX emissions.  Consequently,
    regenerative-firing is not usually practiced without combustion
    modifications for NOX control,  but  there may be exceptions.
    
         Appendix A contains a tabulation of available uncontrolled
    emissions data for reheat furnaces.  These data correspond to a
    variety of condition, some known and some unknown.  Summaries of
    the Appendix A data follow.  The first summary represents all of
    the Appendix A data except three values that are more than an
    order of magnitude lower than any other value and two values from
    an ejector stack that may not be representative.  Included are
    data for both regenerative- and recuperative-fired furnaces and
    others for which the firing configurations are not specified.
    This tabulation shows a wide range  (0.023 - 0.91 Ib/MMBtu).
              Summarv of Appendix A Reheat Furnace Data
         n(No. Samples)
         Min. Value
         Max. Value
         Avg. Value
         Std. Dev.
    Ib/MMBtu
      28
       0.023
       0.909
       0.226
       0.198
    ppm @ 3% 0,
      14
      65.
     740.
     292.
     166.
    Ib/ton
    11
     0.054
     0.327
     0.198
     0.084
         The following is a summary of the Appendix A data for
    recuperative-fired reheat furnaces.
    
    Summary of •• Recuperative -fired Reheat Furnace Data From Appendix A
                        Ib/MMBtu
         n(No. Samples) 16
         Min. Value      0.080
         Max. Value      0.40
         Avg. Value      0.200
         Std. Dev.       0.101
                ppm @ 3%
                  8
                 65.
                326.
                220.
                 81.5
                   Ib/ton
                   8
                   0.157
                   0.327
                   0.237
                   0.058
         The following is a summary of the Appendix A data for
    regenerative-fired .reheat furnaces.  Data from the recuperative-
    fired furnace is much lower than regenerative reheat  furnaces  (0.2
    vs. 0.91 Ib/MMBtu and 220 vs. 740 ppm at 3% 02.  ,
                                    4-13
    

    -------
    Summarv of Regenerative-fired Reheat Furnace Data From Appendix A
         n(No. Samples)
         Min. Value
         Max. Value
         Ave. Value
    Ib/MMBtu
      2
      0.675 ,
      0.909
      0.792
    ppm @ 3%
           2
         550.
         740.
         645.
         None of the reheat furnace emissions data in Appendix A are
    known to correspond to cold combustion air firing.  For the
    purposes of this Chapter, the uncontrolled emissions data for
    cold-air-fired reheat furnaces are estimated using controlled
    emissions data from an LNB-controlled reheat furnace (80 ppm. at 3%
    02)  assuming a  control  efficiency of  27  percent.18  Thus,
    uncontrolled emissions from cold-air-fired reheat furnaces are
    estimated to be 110 ppm at 3% 02 (0.135  Ib/MMBtu) .
    
    4.2.9  Annealing Furnaces
    
         Annealing processes vary greatly depending on specific
    objectives and starting materials.  In general, annealing furnaces
    subject steel products to a planned time-temperature profile in a
    reducing atmosphere with heating supplied from gas-fired, radiantf-
    tube burners.  There are two basic approaches.  Batch annealing
    typically includes slow heating of the steel to a desired
    temperature, soaking at temperature, and controlled slow cooling,
    a process that may take several days.  In continuous annealing,
    steel sheet is passed continuously through an annealing furnace
    where it is subjected to a desired time-temperature profile.
    Temperatures typically peak at about 675 °C  (1,250 °F).  However,
    some processes may require temperatures as high as 760 to 1,200 °C
    (1,400 to 2,200 °F).
    
         Uncontrolled NOX emissions  data  from operating annealing
    furnaces are not available.  The emissions data included in
    Appendix A are based on estimates and on laboratory measurements
    of emissions from furnaces without NOX controls.   These  data were
    provided for new furnaces or, in one case, for a furnace still
    under construction.  In each case, the furnace is regenerative
    fired.  For the purposes of this Chapter, the average of the two
    values is used as the uncontrolled emissions from a regenerative-
    fired annealing furnace, i.e., 775 ppm at 3% 02 (0.952  Ib/MMBtu).
    The actual value would be affected by the temperature of the
    preheated air.  The 1,000 ppm value is possible for regenerative-
    fired furnaces without Nox controls  (because  the  air preheat is
    very high with regenerator technology), but this configuration is
    not used in practice (there may be exceptions) .16
    
         In lieu of uncontrolled emissions data for recuperative- and
    cold-air-fired annealing furnaces, estimated values are used for
    the purposes of this Chapter.  Values of 204 and 450 ppm at 3% 02
    (0.25 and 0.55 Ib/MMBtu) have been suggested as reasonable for
    recuperative-fired furnaces.16'19   The  average  of, these is  327 ppm  at
    3% 02 (0.402 Ib/MMBtu).   A value of  120  ppm at 3% 02  (0.147
    Ib/MMBtu) has been suggested as reasonable for cold-air-fired
    furnaces.19
    

    -------
    4.2.10  Galvanizing Furnaces
    
         Appendix A lists uncontrolled NOX emissions from two
    regenerative-fired galvanizing furnaces.14   (These furnaces are
    described as galvanizing/galvannealing and aluminizing furnaces.)-.
    One estimate is based on measurements at laboratory facilities
    with NOX control  features disabled,  and the second is based on
    field measurements.  These values are 1,000 and 880 ppm at 3% 02,
    respectively, and the average is 940 ppm at 3% 02 (1.15 lb/MMBtu) .
    
         No uncontrolled NOX emissions  data are available for
    recuperative- or cold-air-fired galvanizing furnaces.  In lieu of
    any data, values for annealing furnaces are used as estimates for
    the purposes of this Chapter, i.e., 327 ppm at 3% 02 (0.402
    lb/MMBtu) and 120 ppm at 3% 02 (0.147  lb/MMBtu)  for recuperative-
    fired and cold-air-fired galvanizing furnaces, respectively.16'19   In
    practice, the actual value would depend on many variables
    including furnace and combustion air temperatures.
    
    4.2.11  Summary
    
         Table 4-4 is a summary tabulation of uncontrolled NOX
    emissions from the major NOX-emitting  sources  at iron and steel
    mills.  These values are based on the available emissions data
    presented in Appendix A and, in instances where no data are
    available, on estimates.  There is not an abundance of data, and
    test conditions are often not specified.  For example, NOX
    emissions will vary with furnace operating temperature and
    combustion air temperature, and these conditions are usually not
    specified.  In each case, the text of this Chapter and Table 4-4
    includes information indicative of the quantity and quality of the
    data.  The uncontrolled NOX emissions  data  tabulated in Table 4-4
    are used to calculate NOX emissions  reductions and cost
    effectiveness of NOX controls  in  Chapters 5 and 6,  respectively.
                                    4-15
    

    -------
    I
    H
    CO
    CO
    H
    $
    Q
    
    I
    §
    O
    CJ
    
    
    
    
    
    m
    C
    o
    •H
    m
    m
    "g
    W
    M
    o
    $5
    
    CD
    en
    i
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    m
    c
    o
    •H
    to
    m
    •H
    K
    O
    ^
    0)
    ?
    Q)
    S
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    g
    4J
    ^N^
    ja
    
    
    
    
    3
    
    rQ
    5
    S
    \
    rH
    
    O
    
    tV*
    g (Y)
    Q<
    a©
    
    a
    g
    4J
    ^s^
    A
    rH
    
    
    
    
    
    C
    *J
    g
    yj
    ^o
    rH 3
    
    
    
    
    
    C
    o"
    ,<,,
    g fO
    Pi
    ft ©
    
    
    
    ^i
    4J
    •H
    rH
    •H
    U
    n)
    ^
    03
    m
    0)
    u
    O
    H
    o<
    
    
    in
    H
    •
    CN
    in
    H
    
    o
    
    
    vo
    0
    
    CN
    
    O
    H
    O
    
    
    
    I CN
    T* in
    in TJI
    CN H
    «N
    H
    CN
    CO
    CO
    
    0
    
    
    
    
    H
    H
    
    CO
    in
    vo
    0
    
    
    
    
    
    CO
    
    
    o
    o
    CO
    
    
    
    
    
    
    
    H1
    C-H
    0) )H
    ^ -rl
    O M-l
    i rl
    0) CD
    O C
    U 3
    i-l
    ^
    CO
    •
    o
    1
    r-
    0
    o
    
    o
    
    
    
    
    
    1
    ro eri
    H CO
    CN in
    O 0
    in
    Cn
    ro
    i
    CO
    ^*
    CN
    VO
    H
    H
    *3*
    ro
    
    0
    
    
    
    
    ^
    
    
    in
    vo
    <"
    o
    
    
    
    
    
    CN
    
    
    O
    ^*
    ro
    
    
    
    
    
    
    
    Cn
    C
    •H
    CD
    4J
    C
    •H
    CO
    CN
    r-
    o
    •
    o
    i
    ro
    o
    o
    
    o
    
    
    
    
    
    i
    CN t--
    o in
    o o
    O 0
    
    
    i
    
    
    
    
    o
    H
    t-
    m
    0
    
    0
    
    
    
    
    H
    H
    
    H
    CN
    O
    0
    
    
    
    
    
    H
    
    
    CO
    CN
    
    
    
    
    
    
    CD
    U
    
    m o
    i-H U
    CQ CO
    CN
    CN
    CN
    •
    O
    I
    ^4 1.
    O
    
    O
    
    
    
    
    
    
    1 1
    
    
    
    vo
    o vo
    co ro
    H <
    t O
    CO 0
    rH CN
    
    f- I
    crt
    H i
    H
    
    o
    
    
    1 1
    
    
    
    
    
    
    1 1
    
    
    
    
    
    
    t-CN
    
    
    co ro
    in co
    CN
    Cn
    C
    -H S
    •O * O
    3 f) l~l
    O ^
    c c c
    <1)-H „ 0
    0)"-^ O C
    X1 to Oi Gi
    O 0) C C
    U -H -H
    U 16 -~ )-l M
    -H (3 0< 3 3
    CO H O "O "O
    «! 3 CQ
    03 4-1 O
    
    
    O
    vo
    «
    0
    1
    O i
    in
    
    o
    
    
    
    
    
    
    
    i i
    
    
    
    
    i
    f^
    H B
    I Q
    t- a
    
    VO i
    
    ^ 1
    in
    
    0
    
    
    i i
    
    
    
    
    i i
    
    
    
    
    
    
    i
    
    CN
    
    
    i
    CN
    
    __,
    ~ 0)
    Cn 3
    
    -H i
    • 4J-H X
    O C"H O
    'h 0)
    (d >-l rH JJ
    i 1-1 Q) 3
    U 3 3 O Cn
    -H d) U iw J3 C
    JH U C i 4J -H
    4J (8 O ><-H ^1
    U C U X S -H
    CD }-i — • o ^*m
    rH 3
    H M-l
    
    crt o
    rH rH
    • •
    O 0
    1 1
    VD CM
    rH O
    
    p o
    
    
    
    
    
    I i
    rH CO *f CA
    CN Tjl VO VO
    H H O O
    O O P O
    
    
    
    i i
    
    
    
    
    ro CN
    
    vo en
    rH O
    
    o o
    
    
    
    
    
    fi CN
    
    in vo
    ro vo
    rH O
    o o
    
    
    1 1
    
    
    
    
    1 . 1
    
    
    
    ^
    •rl
    n)
    
    •H
    *i a
    03 0 •
    S" 1
    (d u
    Cn 0)
    C X! -O
    •HO) rH
    X iH O
    it! ft U
    0
    co
    
    CN
    ro
    •
    p
    
    • in i
    H
    
    0
    
    o
    *4*
    •
    0
    
    in en o
    f^ 0 CO I
    VO CD P
    P P 0
    p
    *J( VO
    t^ CN
    i ro i
    p i
    in in
    in vo
    
    l CO i
    
    i ro i
    CN
    
    0
    rH
    
    (U
    4J
    VO O
    CN H S5
    
    CN o in
    en o ro
    t-- N rH
    o o o
    
    rH
    0)
    ^
    O
    CN CO 2
    
    
    in P o
    I1 CN H
    VO CN H
    
    
    
    •0
    CO i I 0)
    CD 0) CD tH
    U > > -H
    id -H -H IH
    54J 4J I
    (d (d rl
    3 ^i v< -H
    b cu cu (d
    c a •
    4J CD 3 T3
    (d Cn-O u T) r-i
    Q) CD CP CD (D O
    ,C Pi rl Pi rl U
    CD -H -H
    K  > -H
    3 -rl -H MH
    M-l 4J 4-> i
    (d cd M
    (71 ^| ^ -H
    Co) o id
    -rl C pt i.
    rH 0) 3 T!
    (d Cn-O u -O IH
    CD Q) CD 0) 0) O
    C Pi rl PJ rl U
    S 'H "H
    
                                                    4-16
    

    -------
    1
    H
    CO
    CO
    H
      X
    
    g,
    
    
    Q
    W
    §
    O
    U
    o
    
    >H
    CO
    I
    
    
    
    m
    g
    -H
    m
    a
    -H
    w
    M
    o
    55
    
    0)
    Ol
    c
    »!
    
    
    
    
    
    
    
    
    
    
    
    
    a
    C
    0
    -H
    (Q
    m
    w
    
    if
    O
    tn
    ($
    0)
    £
    
    
    
    
    
    
    
    
    
    
    c
    o
    4J
    A
    i-l
    
    
    
    ^
    4J
    CD
    
    g
    •x^
    .Q
    r-H
    es
    O
    <*>
    C f*l
    a
    a®
    
    c
    8
    4J
    ~~x,
    rH
    
    
    
    
    
    C
    
    4J
    1
    -x.
    -1 3
    
    
    
    
    
    C
    o
    <*>
    & M
    a©
    
    
    
    >•
    -H
    r-H
    -H
    U
    ft)
    1^4
    *-x^
    to
    m
     > -H
    -H -H >H
    OlW 4J i
    C IS IB >H
    
    NO) 0)  (1) 0) Q) 0) 0
    rH o; n a M u
    It! -H -H
    C5 "4-1 <4-l
                                                                    4J
                                                                    C
                                                                    s
                                                                       m
                                                                          4J 1C
    m  o
    0)  >H  -
    3 t-i  M  ft!
    I-H    «H  g m
    itf -O    -HO
    !>  m -O  w
       4J  CL>  m p
    >H  It) 4J  0) tt)
    0)  g  3     -H
    £ -H  a e r-i
    e -u  E  o
    3  to  o  ^ C
    E! W U  fc H
                                                                 C  i-i  CN m ^
    
                                                                    0)  0) 0) 0)
                                                                    4J  4J 4J 4J
                                                                    O  O O O
                                                                    2  a a a
                                                                                 4-17
    

    -------
    

    -------
    4.3  REFERENCES
    1.
    5.
    
    6
    7.
    8.
    9.
    
    10,
    
    11.
    
    12
    Campbell, Lisa M.,  D. K. Stone, and G. S. Shareef (Radian
    Corporation).  Sourcebook: NOX Control Technology Data.
    Prepared for U.S. Environmental Protection Agency.  Research
    Triangle Park, NC.   Publication No. EPA-600/2-91-029, July
    1991.
    
    Zeldovich, J.  The Oxidation of Nitrogen in Combustion and
    Explosions.  Acta.  Physiochem. 21(4).  1946.
    
    U.S. Environmental Protection Agency.  Control Techniques for
    Nitrogen Oxides Emissions from Stationary Sources-Revised
    Second Edition.  EPA-450/3-83-002.  January 1983.
    
    U.S. Environmental Protection Agency.  Control Techniques for
    Nitrogen Oxide Emissions from Stationary Sources.  Pub. AP-
    47.  National Air Pollution Control Administration,
    Washington, DC.  1970.
    
    Ref. 3, p. 4-4.          .                                   -
    
    Barnard, W.R.  (E.H. Pecham and Association, Inc.).  Emission
    Factors for Iron and Steel Sources-Criteria and Toxic
    Pollutants. Prepared for U.S. Environmental Protection
    Agency.  Research Triangle Park, NC.  Publication No. EPA-
    600/2-90-024.  June 1990.  7 p.
    
    Stockton, M.B., and J.H.E. Stalling  (Radian Corporation).
    Criteria Pollutant Emission Factors for the 1985  NAPAP
    Emissions Inventory.  Prepared for U.S. Environmental
    Protection Agency.   Research Triangle Park, NC.   Publication
    No. EPA-600/7-87-015.  May 1987.  212 p.
    
    U.S. Environmental Protection Agency.  Compilation of Air
    Pollutant Emission Factors, Vol. 1:   Stationary Point and
    Area Sources.  Research Triangle Park, NC.  Publication No.
    AP-42  (GPA 055-000-00251-7), Fourth Edition  (including
    supplements A, B, C, and D).  September 1991.
    
    Ref. 3, pp. 5-72, 5-73.
    
    Ref.-3, p. 5-75.
    
    Ref. 3, p. 5-77.
    Hovis, J. E. Low NOX Burners for Metallurgical Furnaces.
    and Steel Engineer.  pp. 43-46.  June, 1986.
    Iron
                                    4-18
    

    -------
    13.  Letter and attachments from Sulc, D. A., Nucor Steel,
         Crawfordsville Division, to Jordan, B. C., EPA/OAQPS.
         October 6, 1992.  Response to Section 114 letter on iron and
         steel mills.                                                 ;
    
    14.  Letter and attachments from Gilbert, F.C., North American
         Manufacturing Company, to Jordan, B.C., EPA/OAQPS.  November
         11, 1992.  Response to Section 114 letter on low-NOx burners.
    
    15.  Letter and attachments from Harmon, M. L., LTV Steel Company,
         to Jordan, B. C., EPA/OAQPS.  July 2, .1992.  Response to
         Section 114 letter on iron and steel mills.
    
    16.  Letter and attachments from Finke, H. P., Bloom Engineering
         Company, Incorporated to Neuffer, W. J., EPA/OAQPS/ISB.
         May 14, 1993.
    
    17.  Letter and attachments from Felton, S. S., Armco Steel
         Company, L. P., to Jordan, B. C., EPA/OAQPS.  June 26, 1992.
         Response to Section 114 letter on iron and steel mills.
    
    18.  Letter and attachments from Gilbert, F. C., North American  - .
         Manufacturing Company, to Parker, C. D., Research Triangle  -
         Institute.  May 14, 1993.  Response to a request for
         information.
    
    19.  Letter from Dickinson, G., Engineered Combustion Systems,
         Inc., to Parker, C. D., Research Triangle Institute.  May 19,
         1993.  Response to a request for information.
                                   4-19
    

    -------
                                CHAPTER 5
    
                    NITROGEN OXIDES  CONTROL TECHNIQUES
    5.1  BACKGROUND
    
         Because the types of process facilities involved in the
    making of iron and steel are numerous and different, they are
    discussed separately in this chapter.  Control techniques
    applicable to each process facility are identified in the
    appropriate subsection.
    
         Control techniques for NOX emissions  can be  placed into one
    of two basic categories:  techniques designed to minimize NOX
    generation and techniques designed to remove previously generated
    NOX from the waste  effluent  stream.   Combustion modification
    techniques, including low NOX burners(LNB's)  and  flue gas
    recirculation(FGR), fit into the first category.   Add-on flue gas
    treatment techniques such as selective catalytic reduction(SCR)
    and selective noncatalytic reduction(SNCR) are examples of the
    second.  Table 5-1 summarizes these techniques and indicates that
    combustion modifications tend to be less effective than flue gas -
    treatment as a NOX  control.   Of these techniques,  low excess
    air(LEA), LNB's, FGR, SCR, and combinations of these controls are
    known to be used to control NOX emissions  from  iron and steel
    process facilities.  It should be noted that LNB's and often FGR
    incorporate staged combustion in their burner design.
    
    5.2  CONTROL TECHNIQUES APPLICABLE TO IRON AND STEEL FACILITIES
    
         Historically,  few facilities found at iron and steel mills
    have NOX controls.  Requests for information  from the industry
    and control equipment vendors on NOX  controls yielded information
    on only a few facilities with NOX controls.   Those facilities are
    either new or are still under construction, and little
    performance data are available.  The control techniques known to
    be used  (LEA,  LNB's,  FGR, and SCR) are discussed in Section
    5.2.1.  These techniques are applied to reheat furnaces,
    annealing furnaces, and  galvanizing furnaces.   The applications
    are discussed in Section 5.3.  Other control techniques that may
    be applicable to iron and steel processes are discussed in
    Section 5.2.2.
    
    5.2.1  Control Techniques Applied
    
         Control techniques known to have been used on iron, and steel
    process facilities are LEA  (reheat furnaces), LNB  (reheat
    furnaces and galvanizing furnaces),  LNB plus FGR  (reheat
    furnaces, annealing furnaces, and galvanizing furnaces), SCR
    (annealing furnaces), and LNB plus SCR (annealing furnaces).
                                   5-1
    

    -------
              TABLE 5-1  OVERVIEW OF NOX CONTROL TECHNIQUES1
    Technique
    Emission
    reduction
    Comments
    Combustion
    modification
    
      Low excess air
    
      Staged
      combustion
    
      Flue gas
      recirculation
    
      Low-N0x
      burners
    10 to 50+
    percent
     One or more of
     these has  the
     potential  for
     application to  any
     combustion unit.
     Performance and
     costs are  highly
     site specific.,
    Flue gas treatment
    
      Selective noncatalytic      30  to  80
      reduction                   percent
                      Ammonia and urea
                      inj ection are  used
                      on fossil-fuel-
                      and refuse-fired
                      units  in the U.S.,
                      Japan,  and Europe.
      Selective catalytic
      reduction
    50 to 90+
    percent
     Used in U.S.  on
     gas  turbines,
     interal combustion
     engines,  process
     heaters and some
     boilers.   Used  on
     gas-,  oil-,  coal-,
     and  refuse-fired
     units  in Japan  and
     Europe.
                                   5-2
    

    -------
    These control techniques are discussed in the following
    subsections.
         5.2.1.1   Low Excess Air.  Low excess air(LEA) is a
    combustion modification technique in which NOX formation is
    inhibited by reducing the excess air to less than normal ratios.2..
    It reduces the local flame concentration of oxygen, thus reducing
    both thermal and fuel NO,, formation.   It  is  easily implemented
    and is used extensively in both new and retrofit applications,
    either singly or in combination with other control measures.  It
    can be used with all fuels and all firing methods.2'3
    
         The potential of LEA as a NOX control technique  is limited
    by the onset of smoke or CO emissions.  Tests on utility boilers
    have indicated that LEA firing can reduce NOX emissions between
    16 and 21 percent as compared to baseline levels.2  Other sources
    have suggested reductions of up to 15 percent.4   In the case of.
    utility boiler applications, LEA also increases thermal
    efficiency.  It decreases the volume of combustion air to be
    heated, allowing more heat of combustion to be transferred, thus
    lowering fuel requirements for a given output.3   To maintain
    proper control of the furnace pressure, positive pressure must be
    maintained in the furnace to prevent the influx of tramp air.s
    
         A number of other factors affect the excess air levels that
    can be implemented.  These include the type of fuel fired,
    uniformity of the air/fuel ratio, air and fuel control lags
    during load swings, and other combustion control features such as
    staging of fuel or air.2
    
         For utility boilers, LEA firing is considered a routine
    operating procedure and is incorporated in all new units.
    Because it is efficient and easy to implement, expectations are
    that LEA will be used increasingly in other applications as well.
    Although it is a feasible technique for furnaces, specifically
    commercial furnaces, the trend in NOX  control for these sources
    has been in improved burner design.2
    
         In a reheat furnace application at a steel mini-mill, LEA
    as a NOX  control  technique yielded a NOX emission reduction of
    about 13 percent.6
    5.3.6.
    This application is described in Section
         5.2.1.2  Low NOr Burners.   Low NOX burners(LNB) have been
    used since the early 1970's  for thermal NOX control.   These
    specially designed burners employ a variety of principles
    including LEA, off-stoichiometric or staged combustion  (OSC), and
    FGR.  The objective in the application of LNB's is to minimize
    NOX formation while  maintaining acceptable combustion of carbon
    and hydrogen in the fuel.2
                                   5-3
    

    -------
          Literature  references  to  LNB  applications  discuss  industrial
    and utility boiler  and process heater  applications primarily.
    However,  they  are applicable to other  combustion processes
    including reheat furnaces and  continuous  annealing and
    galvanizing line furnaces.7'8 Annealing and galvanizing line
    furnaces  are more like boilers and process  heaters in that  their
    operating temperatures are  in  the  same range.   Reheat furnaces
    and soaking pits, in general,  operate  at  much.higher
    temperatures.
    
          The  differences between a low NO, burner and a burner
    featuring LEA  or FGR, for example, are not  always clear.  In
    general,  LNB's implement LEA,-  OSC, FGR, or  a combination of these
    techniques.  In  a stricter  sense,  LNB's have been defined as
    burners that control NO^ formation by carrying out the combustion
    in stages (OSC)  and, further,  by controlling the staging at and
    within the burner rather than  in the firebox.3  Consistent with
    this  definition, there are  two distinct types of designs for
    LNB's:  staged air  burners  and staged  fuel  burners.  Staged air
    burners are designed to reduce flame turbulence, delay  fuel/air
    mixing, and establish fuel-rich zones  for initial combustion.
    The reduced availability of oxygen in  the initial combustion zone
    inhibits  fuel  NO, conversion.   Radiation of heat from the primary-
    combustion zone  results in  reduced temperature  as the final
    unburned  fuel  gases mix with excess air to  complete the
    combustion process.  The longer, less  intense flames resulting
    from  the  staged  stoichiometry  lower peak  flame  temperatures and
    reduce thermal NOX formation.3
    
          Figure 5-1  is  an illustration of  a staged-air LNB.  All of
    the fuel  and less-than-stoichiometric  air (primary air) are
    initially supplied  to the primary  combustion zone.  Staged  or
    secondary air  is supplied beyond the primary zone where
    combustion of  the fuel is completed, again  under off-
    stoichiometric conditions.3   Staged air burners  generally
    lengthen  the flame  configuration so their applicability is
    limited to installations large  enough  to  avoid  flame impingement
    on internal surfaces.  The  installation of  replacement burners
    may require substantial changes  in burner hardware, including air
    registers, air baffles and  vanes,  fuel  injectors, and throat
    design.   Existing burners can  incorporate staged air burner
    features  by modifying fuel  injection patterns,  installing air
    flow baffles or vanes,  or reshaping the burner  throat.  Staged
    air burners are effective with  all fuel types.3
    
         Figure 5-2 is  an illustration of a staged-fuel LNB.3 Staged
    fuel burners mix a portion  of the  fuel  and  all  of the air in the
    primary combustion  zone.  The high level  of excess air greatly
    lowers the peak flame temperature achieved  in the primary
    combustion zone,  thereby reducing  formation of  thermal NO,.  The
    secondary fuel is injected  at high pressure into the combustion
                                   5-4
    

    -------
    Staged air is mixed
    with the combustion
    products from the
    primary zone.  This
    lowers the peak flame
    temperature which
    limits the formation
    of NO. 	
       Staged  air
     Secondary air
    Sub-stoichiometric
    conditions in primary zone
    increase the amount of
    reducing agents (H2 & CO).
                                                Primary Air
                Figure 5-1..  Staged-air Low NOX burner.
                                      5-5
    

    -------
    Secondary Combustion
                                                High Air-
                                                Ratio  in
         Secondary Fuel
    Primary Fuel
    Connection
    to-Fuel
    Primary Zone
                                                          Combustion Air
                                                Secondary Fuel Connection
               Figure  5-2.   Stagsd-fusl  Low  NOX burners.
                                   5-6
    

    -------
    zone through a series of nozzles that are positioned around the
    perimeter of the burner.  Because of its high velocity, the fuel
    gas entrains furnace gases and promotes rapid mixing with first-
    stage combustion products.  The entrained furnace gases simulate
    flue gas recirculation.  Heat is transferred from the first-stage.
    combustion products prior to the second-stage combustion.  As a
    result, second-stage combustion is achieved with lower partial
    pressures of oxygen and temperatures than would normally be
    encountered.3
    
         Unlike the staged air burner, staged fuel burners are
    designed only for gas firing.  The staged fuel burner is able to
    operate with lower excess air levels than the staged air burner
    due to the increased mixing capability resulting from the high-
    pressure second-stage fuel injection.  An additional advantage of
    the staged fuel burner is a compact flame.  Cooling of the
    combustion products from the first-stage zone in the staged air
    burner is accomplished primarily by radiation to the process.
    However, in a staged fuel burner the entrained furnace products
    give additional cooling to the flame.  This particular
    characteristic permits more intense combustion with reduced NOX
    levels.3
    
         In addition to the common viewpoint of LNB as a burner that
    stages combustion within the burner rather than in the firebox
    other designs also include FGR within the burner.  The FGR
    feature may be a more effective control than OSC.  For example, a
    radiant tube burner that uses a vitiated air stream (i.e., a
    burner featuring FGR) is being developed.  The combustion takes
    place within the burner, and reduced NOX generation has been
    demonstrated.9
    
         Full-scale tests on boilers in Japan have shown NOX
    emissions reductions from 40 to 60 percent with gas- fired LNB's.
    Other subscale test results with LNB's include 55 percent NOX
    reduction with specially designed nozzles and burner blocks, and
    55 percent NOX  reduction with designs that create fuel-rich and
    fuel-lean combustion zones (staged combustion).2
    
         Other estimates on NOX reductions for LNB's  include  20 to 50
    percent NOX reductions  in oil-  and gas-fired package boilers
    using shaped fuel injection ports and controlled air-fuel mixing,
    and 55 percent (typical) NOX  reductions in process furnaces using
    self-recirculating (an FGR feature) and staged combustion type
    burners.  In summary, LNB's appear to offer potential NOX
    reductions around 50 percent in boiler and process heater
    applications.2
    
         There are three known applications of LNB's used by the iron
    and steel industry in the United States.  LNB's are being used at
    a reheat furnace, an annealing furnace, and a galvanizing
                                   5-7
    

    -------
    furnace.  The use of LNB's may change the heat transfer
    characteristics and the oxidizing/reducing atmosphere of a
    furnace.  This is not a problem for new furnaces as the design
    and construction of new furnaces can accommodate these changes.
    For existing furnaces, in particular annealing furnaces, which
    operate at a  very specific flame and furnace geometries to
    achieve a specific "set point" at which steel processing is most
    efficient, major problems may occur for a specific furnace
    without a large amount of equipment reconstruction.
    
    
         5.2.1.3  Low NOT Burner Plus  Flue  Gas  Recirculation.   As
    discussed earlier, LNB's are often designed to incorporate OSC
    and FGR within the burners.  However, additional, external FGR is
    added to enhance NOX control in some furnace installations.  Flue
    gas recirculation is implemented by recycling a portion of the
    flue gas to the primary combustion zone.  This principle is
    illustrated for a boiler in Figure 5-3.  It reduces NOX formation
    by two mechanisms.  The recycled flue gas contains combustion
    products that act as inerts during combustion and lower the peak
    flame temperature, reducing thermal NOX formation.   To a lesser
    extent, FGR also reduces thermal NOX formation by lowering the
    oxygen concentration in the primary flame zone.  These factors
    lower flame temperature, altering heat distribution and reducing
    fuel efficiency.
    
         Flue gas recirculation has been applied principally to
    boilers and to a few process heaters.  Emissions tests on utility
    boilers with flue gas recirculation control have achieved NOX
    reductions of 13 to  60 percent.2
    
         Controlled NOX emissions data are available for two reheat
    furnaces, an annealing furnace, and a  galvanizing furnace with
    LNB plus flue gas recirculation control.9'10  These data are
    discussed in Section 5.3.
    
         5.2.1.4  Selective Catalytic Reduction.  SCR is  the most
    advanced of the flue-gas treatment methods  for reducing NOX
    emissions and the one on which the great majority of  flue  gas
    treatment units is based.2  The principle of SCR is illustrated
    in Figure 5-4.  SCR  units use  NH3 to selectively reduce NOX.   The
    ammonia, usually diluted with  air or steam,  is injected through a
    grid system into the flue/exhaust gas  stream upstream of  a
    catalyst bed.  On the catalyst surface, the ammonia  reacts with
    NOX to form molecular nitrogen and water.   Depending on system  •
    design, NOX removal rates of 80 to 90 percent and higher are
    achievable.3
                                    5-8
    

    -------
    Secondary
    air wind-
       box
                                                              Suck
                                                             breechlnf
                  Figure 5-3.   Illustration  of flue gas recirculation
                                                                         11
                                    5-9
    

    -------
         The major reactions that occur in SCR are  the  following:
    4NO
    
    2N02
    4NH3
    
    4NH3
                               4N2 + 6H20 , ' and
                                3N
                                      (5-1)
                                 (5-2)
    Of these, reaction  (5-1) predominates since  flue gas NOX consists
    primarily of NO.  The optimum temperature range for these
    reactions is generally 260 to 540  °C  (500 to 1,000 °F) .  At
    higher temperatures, the NH3 oxidizes to NOX  or ammonium nitrate
    and ammonium nitrite.  Lower temperatures do not provide
    sufficient energy to initiate the  reaction.1  The catalysts used
    in SCR units are predominately oxides of titanium and  vanadium.2
    However, platinum,  zeolites, and ceramics are also used.1  In-
    gas-fired applications, catalyst pellets in  a fixed bed are
    commonly used.  For oil- or coal-fired applications where  the
    flue gas contains particulate matter, catalyst designs usually
    include honeycomb,  pipe, or parallel shapes,  which allow the flue
    gas to pass along the catalyst"surface.2
                        NH,
                 Gts
          +
       •—NO,
    
       •— "P.
       NH.
    O—NO,
    X-NH,
    O—NO,
    X-NH,
                                       0-HjO
    Figure 5-4  Illustration of the SCR principle.12
         The effectiveness of the SCR in removing NOX is also
    dependent on the NH3:NOX mole ratio.  As the NH3:NOX mole  ratio
    increases, the NOX reduction increases and ammonia slip  (i.e.,
    unreacted ammonia slipping through the catalyst bed)  increases.
    In practice, removal efficiencies of up to 90 percent are
    commonly achieved with ammonia slip values below  5 ppm.  On-line
    NOX analyzers  and feed-back control  of NH3 injection are  required
    to maintain control of the SCR process.
    
         Site-specific factors other than temperatures. and NH3:NOX
    ratios that affect NOX emission rates achievable with SCR include
    the ratio of flue gas flow rate to catalyst volume  (the  inverse
    of residence time), catalyst poisoning  (by metals, acid  gases, or
    particulate entrainment), and pressure drop across the catalyst
    which reduces fuel efficiency.  Most of these potential  problems
    have been addressed successfully in commercial operations.3
                                   5-10
    

    -------
         There has been limited experience with SCR at iron and steel
    mill.  This is primarily due to the lack of stringent NOX
    regulations.  In the U.S., SCR has been applied to oil and gas
    fired combustion units such as gas turbines, internal combustion .
    engines, and process heaters and several coal-fired boilers.  At
    iron and steel mills, there are 2 SCR units in the U.S.  An SCR
    unit is being used to control NOX emissions from a gas-fired,
    radiant tube, continuous annealing furnace at a steel mini-mill
    in the United States.  This furnace also has LNB's.  Controlled
    emissions from this unit are about 33 ppm at 3 percent O2.14 A
    second SCR unit, currently under construction, will be used to
    control NOX emissions from an annealing furnace  at an integrated
    steel plant in the United States.  This furnace does not have
    LNB's.    The unit has a guaranteed NO, reduction of  90 percent.15
    
         Techniques for NOX control adopted by the Japanese iron and
    steel industry are, for the most part, based on combustion
    modifications.  However, the Japanese have investigated  "flue gas
    denitrification" using the SCR process for sintering plants, coke
    ovens, and reheating furnaces.7  They concluded  that  the SCR
    process was impractical for use on sintering plants due  to the
    many technical problems that remain to be solved.16  These
    problems include the limited service life of the  catalysts,
    energy requirements to raise the exhaust gas temperature, and the
    large installation space required.  However, in some
    environmentally critical areas, flue gas denitrification units
     (presumably SCR) are used on sinter plants.  SCR  applications to
    coke ovens and reheat furnaces appear to be experimental.  They
    are not routinely used on these facilities in Japan.8
    
    5.2.2  Other Control Techniques
    
         LEA, LNB's, LNB plus FGR, SCR, and SCR plus  LNB are NOX
    control techniques that have been used to reduce  NOX  emissions
    from iron and steel process facilities in the United States.
    Other techniques that may be applicable, some of  which have been
    used in Japan, are discussed in this section.
    
         5.2.2.1  Off-Stoichiometric or Staged Combustion(OSC).  OSC
    reduces NOX generation by carrying out initial combustion in a
    primary, fuel-rich combustion  zone and subsequently completing
    combustion at lower  temperatures in a second, fuel-lean  zone.
    This technique is essentially  the control  technique implemented
    through the use of LNB's; however, it can be  implemented using
    conventional burners by staging the combustion  in the  furnace or
    firebox rather than within the burner itself as  is done  with
    LNB'S.
         In practice, OSC is  implemented through biased burner  firing
     (BBF), burners out of service  (BOOS), or overfire air  (OFA).  BBF
    and BOOS are very similar and  are generally applicable  to
                                   5-11
    

    -------
    furnaces with multiple burners arranged in rows.  In BBE, fuel
    distribution among the burners is redistributed so that the lower
    rows of burners are fired more fuel-rich than the upper rows, as
    illustrated in Figure 5-5.  The additional air needed to complete
    combustion is provided through the upper rows of burners, which  ..
    are fired fuel-lean.  In BOOS firing, selected burners or rows of
    burners are made inactive  (taken out of service) by turning off
    their fuel supply and using them to admit air only to the
    furnace.  The total fuel demand is supplied by the remaining or
    active burners.  Thus, the active burners are fired fuel-rich,
    and the additional air required to complete combustion is
    supplied through the BOOS or air-only burners.  This principle is
    illustrated in Figure 5-6.  Both BBF and BOOS implement staged
    combustion by creating two combustion zones within the firebox:
    one fired fuel-rich or with low excess air and the second fired
    with excess air at a reduced temperature.  These methods are
    applicable to all fuels, and both fuel- and thermal-NOx formation
    are reduced.2
    
         These techniques, BBF and BOOS, are attractive options for
    existing facilities with suitably configured burners since few
    equipment modifications are required.  In some cases, process
    facilities may need to be derated if the active burners (BOOS) or-
    fuel-rich burners (BBF) have limited firing capacities.
    
         Implementing BBF or BOOS requires careful, accurate
    monitoring of the combustion process and the flue gas in order to
    control the process.  Alternatively, furnace efficiency and
    safety may be compromised.2
    
         Emissions tests of BOOS firing on utility boilers have
    indicated average NOX  reductions  of  31  to  37 percent  for coal,
    oil, and natural gas firing compared to baseline values.2   It
    seems reasonable that BBF would yield similar results due to the
    similarity of the two techniques.
    
         A third type of staged combustion, illustrated in Figure 5-
    7, is OFA.  With OFA,  the burners are fired more fuel-rich than
    normal, and the additional air needed to complete combustion is
    admitted through overfire air ports or an idle top row of
    burners.  This technique is applicable with all fuels.  In
    emissions tests on utility boilers,. OFA has achieved average NO^
    reductions of 24 to 59 percent with oil, coal, and natural gas
    firing compared with baseline levels.2
    
         These staged combustion techniques, especially OFA, are more
    easily implemented on large, utility-scale boilers than on .
    smaller units and furnaces.2 As  furnace size  decreases, furnace
    volume decreases more quickly than furnace wall area.  Thus,
    residence times for fuel combustion may become a problem in
    smaller units.  Moreover, space for additional ductwork, furnace
                                   5-12
    

    -------
    Fud-rich
      Zoo*
                    Figure 5-5.   Biased  Burner Firing!!
                                  5-13
    

    -------
    Figure 5-6.  Burners out of service.
                                        11
                5-14
    

    -------
                                    Ovcrfire Air
                                      Ports
    Figure 5-7.  Overfire air.
                 5-15
    

    -------
    penetrations, and fans may be a problem if these additions are
    needed.  These factors are potential impediments to implementing
    these OSC techniques in iron and steel process facility furnaces,
    especially on a retrofit basis.
    
         There is no evidence that FGR has been used to control NOX
    emissions from iron and steel process facilities.  However, a
    radiant tube LNB featuring a vitiated air stream concept  (i.e.,
    FGR) is being developed.7  As  discussed in Section 5.2.1.3,  FGR
    has been used with LNB's to reduce NOX emissions  from a reheat
    furnace.
    
         5.2.2.2  Fuel Switching.  Fuel switching and  fuel
    denitrification are possible NOX control techniques.   In Japan,
    these techniques are currently used or have been used to  reduce
    NOX emissions from reheat furnaces,  sintering furnaces,  and coke
    ovens  (denitrification of fuel only).8
    
         Natural gas is an attractive fuel from a NOX control
    perspective because of the absence of fuel-bound nitrogen (fuel
    NOX)  and the flexibility it provides for implementing other
    combustion modification control techniques.  Because natural  gas -
    is already the principal fuel used in reheat furnaces in  the
    United States, fuel switching is not a reasonable  NOX
    control option for reheat furnaces  (or soaking pits, for  the  same
    reason) in the United States.  Moreover,  for large fuel-burning
    processes, fuel choice decisions tend to  be based  on fuel costs
    and  on other regulatory constraints.
    
         Coke ovens reduce coal to coke in a  reasonably closed
    reducing environment.  There are fugitive emissions of  NOX from
    leaks and during charging and pushing operations.  Coke oven
    underfiring  is done predominantly with  coke oven gas, blast
    furnace gas, natural gas  ,or some combination of these  three.   It
    is not clear that fuel switching is a reasonable option for NOX
    reduction in the case of coke ovens in  the United  States, and it
    is not clear from the literature to what  extent the Japanese  use
    fuel switching as a coke oven NOX control strategy.  Two
    strategies are cited.  These are  (l)  switching to  lighter fuel
    oil  (not an  option if COG, BFG, and NG  are baseline fuels), and
     (2)  taking  credit for the  nitrogen  removal that is incidental to
    the  desulfurization of COGV(an oxides of  sulfur control
    measure).&
    
         5.2.2.3  Selective  Noncatalytic  Reduction.   SNCR is
    generally an applicable  NOX.control technique for  applications
    that are suitable  for SCR.  SNCR is very  similar  to SCR:  i.e.,
    NH3 or urea is injected into the flue gas stream where the
    temperature  is suitable  for the  reaction  with NOX  to proceed.
    However, SCR and SNCR differ  in that  their ideal  reaction
                                   5-16
    

    -------
    temperatures differ.  Depending on catalyst type, the optimum
    temperature for SCR reactions is 260 to 450 °C  [500 to 840 °F],
    and the optimum temperature for SNCR is 870 to  1,250 °C  (1,600 to
    2 300 °F) .  Generally, 40 to 70 percent NOX reduction is achieved
    with NH3:NOX molar ratios of 1:1 to 2:1.  This control effiency   ,
    generally results from the difficulty of achieving uniform mixing
    of NH3  with the flue gas  and from the variations of flue gas
    temperature and composition usually present.
    
         There are two commercially available  SNCR  processes  that  use
    either ammonia or urea.  The SNCR process  that  uses ammonia  is
    the Thermal DeNOx (TDN)  system developed by EXXON.  In this
    process, gaseous ammonia is injected into  the air-rich  flue  gas
    stream where it reacts with NOX in accordance with two competing
    reactions:3
    2NO
    4NH
    4NH3  +  502
                 20
                     4NO
                            3N2 + 6H20,  and
    (5-3)
    
    (5-4)
    Thus, the TDN process has  the potential  to  either reduce NOX in
    accordance with  reaction  (5-3)  or increase  NOX in accordance with
    reaction  (5-4).  Temperature is the primary variable for
    controlling  the  rates of these  reactions.   In the temperature
    range of  870 to  1,200 °C  (1,600 to 2,200 °F),  reaction (5-3)
    dominates, resulting in a  reduction of NOX.  Above 1,200  °C
     (2,200  °F),  reaction  (5-4)  dominates,  resulting in an increase of
    NO.  Below 870 °C (1,600 °F),  neither reaction is of sufficient
    activity  to  either increase or  decrease  NOX.  The reduction of NOX
    is  maximized in  the relatively  narrow range of 870 to
    1 040  °C   (1,600 to 1,900  °F) with an optimum temperature of
    about  950  °C (1,750 °F) .   The favorable, 170 °C (300 °F) reaction
    window for reaction  (5-3)  can be lowered to about 700 to 820 °C
     (1,300  to 1,500  °F) by  introducing hydrogen, a readily oxidizable
    gas.3-17
    
         Without the use of a  catalyst to increase the reaction
    rates,  adequate  time at optimum temperatures must be available
    for the NOX  reduction reaction  to occur.  Design  considerations
    should allow ample residence  time and good mixing in the required
    temperature  range. Long residence times (>1 second) at optimum
    temperatures tend  to promote  relatively high performance even
    with less than optimum initial  mixing or temperature/velocity
    gradients:   However, when the NH, injection zone  is  characterized
    by  low temperature arid/or steep temperature declines, a loss of
    process efficiency results.
    
          The  initial ratio of ammonia injected to NOX concentrations
     is  another parameter to control in the process.  Maximum NOX
     reductions  (50 to 70  percent)  require 1.5 to 2.0 NH3:NOX injection
     ratios.  At these ratios,  significant concentrations of NH3 can
                                    5-17
    

    -------
    exit the convective zone, creating corrosive ammonium sulfates
    and/or a visible NH3 stack plume.   Unreacted ammonia emissions
    from a TDN system are usually higher than from SCR systems.3
    
         The second of the two commercially available SNCR processes
    is the Nalco/Fuel Tech NOXOUT process.  In this  process,  a urea-
    type compound is injected at a favorable location to reduce NOX
    according to the following reaction:             •
         NH2 +  NO -» N2 + HjO
    (5-5)
    The exact chemical mechanism is not fully understood, but it
    involves the decomposition of urea  [CCNH^O].  The likely
    decomposition products include the NHj groups.   The reaction
    takes place at temperatures of 870 to 1,200- °C  (1,600 to
    2,300°F) .3
    
         Originally developed by the Electric Power Research
    Institute (EPRI) in the early 1980's, the process is currently
    licensed by Nalco/Fuel Tech where patented process modifications
    and control techniques, combustion unit computer modelling and
    proprietary additives that allow NOX reduction capability over a
    temperature range of 760 to 1,200 °C  (1,400 - 2,300 °F) have been,
    developed.   As with the other post-combustion NOX control
    systems, temperature is the primary control variable for the
    selective reactions.  NOX reductions up to 80  percent are
    achievable with this technology.  The performance of the urea
    injection process is limited by the time/temperature/flow
    characteristics of the flue gas.  Residence time in the
    temperature window and the urea-to-NOx ratio  impact the
    performance in a manner similar to that for Thermal DeNOx.   The
    NO, reduction capability is  limited by mixing  because the
    reaction times are relatively quick.3
    
         The SNCR technique is discussed here because of its
    successful application to facilities that are similar to iron and
    steel facilities, e.g., hot-side, coal-fired utility boilers,
    heavy oil fired industrial boilers, and wood-waste fired co-gen
    plants.  In general, SNCR is applicable in applications where a
    temperature suitable for a favorable reaction is acheivable.
    SNCR has not been used for any iron and steel mill processes.
    
    5.3  NITROGEN OXIDES CONTROLS FOR SPECIFIC FACILITIES
    
         In this section, the NOX emission sources at iron and steel
    mills are discussed with a view toward identifying and evaluating
    applicable control techniques.
    
    5.3.1  Coke Ovens
    
         Coke ovens produce metallurgical coke from coal by the
                                   5-18
    

    -------
    distillation of volatile matter.  A by-product of the process is
    coke oven gas, a fuel commonly used (along with blast furnace
    gas) to fire the ovens.  Although NOX  emissions  tend to  be
    minimized by slow mixing in the combustion chamber, they are
    nonetheless substantial because of the large quantity of fuel
    consumed.  Coke ovens are among the major NOX emission sources at
    iron and steel mills.2  Coke  ovens  with NOX controls in the United
    States have not been found.
    
         The Japan Iron and Steel Federation has reported the
    installation of flue gas denitrification (SCR) units on coke
    ovens.  They also cite low-air-ratio combustion and
    denitrification of fuel as NOX control techniques applicable to
    coke ovens.  Denitrification of fuel may refer to the practice of
    desulfurizing coke gas, which also has the effect of removing
    nitrogen and, thus, reduces NOX emissions.   Flue gas
    recirculation is also cited.  The Japanese acknowledge that there
    are many limitations and difficulties associated with applying
    these controls and some may be applicable only to new ovens.  No
    details are provided about the effectiveness or levels of
    performance of these controls.
    
    5.3.2  Sinter Plants
    
         Sintering is also a major source of NOX emissions from iron
    and steel mills.  In general, only integrated steel plants
    operate sinter plants where the sinter is charged to a blast
    furnace.
    
         Sinter is produced from iron ore fines, limestone  fines,
    coke fines or breeze, flue dust, and possibly other additives._
    These materials are mixed, uniformly distributed onto a traveling
    grate, ignited, and heated to a fusion temperature that causes
    the materials to agglomerate into sinter.  Subsequently, the
    sinter is cooled, crushed, and screened for  size prior  to
    charging to a blast furnace.
    
         The major source of energy used  in the  production  of sinter
    is  the carbon content of the coke breeze and flue dust.  An
    auxiliary fuel such as natural gas or coke oven  gas  is  used to
    ignite the bed.  This ignition fuel typically represents about  8
    to  10 percent of the  thermal energy required by  the process.
    Following ignition, combustion is  continued  by  forcing  air
    through  the bed as it travels with the grate.  The  combustion of
    the coke breeze and flue dust  (about  5 percent  of  the bed
    material) creates sufficient heat  and temperature  [about 1,300  to
    1,480  °C  (2,370 to 2,700 °F)]  to sinter the  fine ore  particles
    into porous  clinkers  that, following  cooling and screening,  are a
    suitable burden for a blast  furnace.17 No information was
    obtained on NOX emission control techniques for sinter plants in
    the United States.
    
          Japanese sources cite sinter  plants as  a major source  of NOX
    
                                   5-19
    

    -------
    emissions at steel plants and report on research on SCR
    applications, an electron beam radiation process, and technology
    for reducing nitrogen in the coke breeze as they apply to sinter
    plants.  The results of the research on reducing N2 in the coke
    breeze are not readily available; however, it is noted that many .
    problems remained to be solved.16
    
         More recently, the Japanese have reported that some steel
    mills have controlled NOX generation at  sinter plants  by
    pretreating the sinter materials, i.e., lowering the oxidation
    level of the sintering atmosphere through adjustment of the coke
    blending ratio, basicity, etc.7  Most of the  NOX in sinter plant
    flue gas appears to come from coke in the sinter bed, which
    usually contains about 1 percent nitrogen.  Japanese research in
    this area progressed from batch tests to a continuous test plant
    with a capacity for treating 20 tons of coke per day.  The
    process was to preheat pulverized coke in a two-stage, fluidized-
    bed heater and then use an electric furnace to heat the coke to
    1,700 °C (3,090 °F) using an electric current through the coke
    itself.  The retention time is about 2 hours.  The coke is
    subsequently cooled.19
    
         About 70 percent of the N is removed from the coke by
    heating to 1,700 °C (3,090 °F).  It is claimed that by removing
    70 percent of the N from coke, the NOX in the flue  gas was also
    reduced by about 70 percent.  This indicates that, in the sinter
    plant, most of the NOX is generated from N in the coke.19
    
         It was estimated that the coke denitrification process
    requires about 30 kWh of electricity per ton of sintered iron
    ore.  Concern was also expressed about the flue gas from the
    treatment process, i.e., its composition and treatment.  The
    Japanese were also uncertain as to whether the process could be
    commercialized.
    
         The Japanese also report that some sintering plants, located
    in areas where national emission standards are not met
    sufficiently to ensure environmental quality, are equipped with
    flue gas denitrification units.  These are SCR units, and there
    are still technical problems with this application, e.g., the
    catalyst life is limited and large amounts of additional heating
    energy are required.8                                   ;
    
         In one sinter plant application, the commercial SCR plant
    was constructed as a permitting compromise when a large, new
    blast furnace was constructed.  The capacity of the SCR unit was
    750,000 Nm3/hr.  The sinter  flue  gas  contained considerable
    amounts of particulates and S03 and,  thus,  was difficult to
    treat.  The flue gas treatment consisted of a flue gas
    desulfurization (FGD)  process followed by a wet electrostatic
    precipitator (ESP).  The cleaned gas was then heated to 400 °C
    (750 °F)  for SCR treatment.   Sinter flue gas typically has a
    temperature of about 150 °C (300 °F), 150 to 300 ppm of NOX
    (mostly NO), 150 to 400 ppm of SOX,  and  200 to 300  mg/Nm3 of
                                   5-20
    

    -------
    particulates containing ferric oxide ;and potassium chloride
    vapor.  It is not an easy gas to treat with SCR because the
    effluent gas has to be heated and because particulates and
    potassium chloride vapor in the effluent tend to contaminate the
    catalyst.19
    
         The following information was obtained about this major SCR
    unit:19
         Completed
         Treated facility
         Capacity
         Load factor
         Pretreatment of gas
         SCR inlet
         Reaction  temperature
         NOX:NH3  ratio
         Catalyst  life
         Nox removal
    1976
    Iron ore sintering plant
    762,000 Nm3/hr
    70 to 100 percent
    Electrostatic precipitator
    Flue gas desulfurization
    Wet ESP
    Heating
    200 to 300 ppm NOX
    5 to 20 ppm SOX
    3 to 10 mg/Nm3  particulates
    11.2 percent 02
    380-390 °C  (716-734 °F)
    1:1
    1 year
    95 percent
     In  1979,  the Japanese  assessment  of  this  control  application was
     that  the  investment  and operating costs were  so high that  the
     process could  not  be widely used.19
    
     5.3.3 Blast Furnaces  and Blast Furnace Stoves
    
          The  blast furnace is a tall,  shaft-type  furnace where
     iron-bearing materials (iron ore,  sinter,  slag, scrap,  etc.)  are
     reduced to iron (pig iron or hot  metal).   The burden or charge to
     the blast furnace  also includes flux (limestone or dolomite)  and
     coke. A  preheated air blast is charged to the  blast furnace
     through tuyeres, nozzle-like openings near the  bottom of the
     furnace.   The  coke in the burden  produces the heat required for
     smelting  the charge, and it provides carbon and carbon monoxide
     required  for reducing the iron ore.   Moreover,  because the coke
     retains its strength at temperatures above the  melting
     temperature of pig iron and slag, it provides the structural
     support that keeps the unmelted burden materials  from falling
     into  the  hot,  molten metal that collects  in the hearth.  A
     description of the blast furnace  and its  operation can be found
     in Chapter 3 of this document, and a very detailed discussion can
     be found  in Reference 18.
    
          The  blast furnace itself is  a closed unit  with no
     atmospheric emissions.  The hot air blast reacts  with the coke in
     the furnace to produce more carbon monoxide than is needed to
     reduce the iron.ore.  The excess,CO leaves the  top of the blast
                                    5-21
    

    -------
    furnace with other gaseous products and particulates.  This blast
    furnace effluent, called BFG, is captured, cleaned, cooled, and
    subsequently used as a fuel in other processes.  BFG contains
    about 1 percent hydrogen and 27 percent CO and has a heating
    value of about 2,540 - 3,600 kJ/Nm3 (65  -  92  Btu/ft3) .5
    
         A blast furnace typically has three or four blast furnace
    stoves as auxiliaries.  These stoves are used to heat the blast
    air required by the blast furnace.  They function somewhat like a'
    regenerator in that one or more of the stoves are on blast (i.e.,
    they are being used to heat the blast air being used by the blast
    furnace} while the others are being heated.  They differ from a
    regenerator in that the source of heat is the combustion of a
    fuel rather than the extraction of heat from an effluent stream.
    
         Between 2.2 and 3.5 kg of BFG are generated for each
    kilogram of pig iron produced in a blast furnace.  About 18 to 24
    percent of this gas is used to heat the checker brick in the
    associated stoves.2  The  remainder is  used elsewhere in the
    plant.  Because of its low heating value, the BFG used in the
    stoves is often enriched by the addition of fuels with a much
    higher calorific value such as natural gas or coke-oven gas to
    obtain higher blast air temperature, often in excess of 1,100 °C -
    (2,010
         Blast furnace stoves are a NOX emissions concern because
    they consume large quantities of fuel.  However, the primary fuel
    is BFG, which is largely CO, has a low heating value, and
    contains inerts, factors that reduce flame temperature.  Thus,
    the NOX concentration in blast furnace stove  flue gas tends to be
    low and the potential for NOX
    small.2
                                  reduction is considered to be
         Apparently there are no blast furnace stoves with NOX
    emissions controls in the United States.  The Japanese also
    indicate that NOX emissions from blast furnace stoves are not a
    concern for the reasons noted above.  They also suggest that LEA
    combustion can be implemented, but they do not explicitly claim
    to have implemented LEA.8
    
    5.3.4  Basic Oxygen Furnace
    
         The EOF is the principal steelmaking process used in the
    United States.  In operation, the EOF is tilted to receive a
    charge of scrap metal and, subsequently, molten pig  iron from a
    blast furnace.  After charging, it is returned to a  vertical
    position and blown with high-purity oxygen, which converts the
    charge to steel.  It is finally tapped and the molten steel is
    poured into a transfer ladle.  Following tapping, the EOF is
    again tilted to discharge the remaining slag.  A more .detailed
    description of the EOF facility and the oxygen steel -making
                                   5-22
    

    -------
    process can be found in Chapter 3 of this document and in
    Reference 18.
    
         In the EOF, oxygen is blown onto the bath at supersonic
    velocity (to penetrate slag and metal emulsions) where it
    oxidizes impurities in the bath to create steel.  The oxidation
    of carbon produces about 467 kJ of CO per kilogram of steel
    (400,000 Btu/ton) .2  Typically,  the off-gas,  which is mainly CO,
    is burned in an open, water-cooled hood mounted above the EOF.
    These hoods admit air for combustion of the CO on an uncontrolled
    basis.  Some newer BOF's use suppressed combustion hoods that do
    not admit air or burn the off-gas.  In these instances, the
    effluent gas is flared.
    
         The combustion of the waste gas results in the generation of
    NOX.   There is  no information available  to indicate that existing
    BOF's have any kind of NOX emission controls or that any control
    techniques are suitable.2
    
         A variation on the EOF  is the Q-BOP  furnace.   It differs
    from the BOF principally in  that oxygen is blown  through the
    molten metal bath  through tuyeres  in the  bottom of  the vessel
    rather than onto the top surface of the bath.  From a NOX
    emissions and controls perspective, the two processes appear to
    be essentially the same, i.e., there is no  indication that  Q-BOP
    furnaces have NOX emissions controls or that any control
    techniques are suitable.  In each  case, NOX is generated during
    the oxygen blow when CO  in the off-gas is combusted either  at the
    mouth of an open,  combustion-type  hood or at the  flare  of a
    suppressed hood system.20
    
    5.3.5  Electric Arc  Furnaces
    
         The use of electricity  to melt steel scrap in the  EAF
    transfers NOX generation from the steel mill to a utility power
    plant.  There  is no  information  that NOX emissions controls  have
    been  installed  on  EAF's  or that  suitable  controls are available.
    
    
    5.3.6  Reheat  Furnaces
    
          Reheat  furnaces bring steel to a uniform temperature  of
    about  1,180  to  1,340 °C  (2,150  to 2,450  °F),  a temperature
    suitable  for hot working.18   They are major  fuel - consuming
    facilities  and major sources of  NOX emissions.
    
          LEA,  LNB's, and LNB plus  FGR are being used to control NOX
    emissions  from reheat  furnaces  in the United States.5-10'11'21    Major
    modifications  to furnace structure and refractories to install
    alternate burners  may be required in  some existing reheat
    furnaces.    In an  LEA application, the  control was retrofitted to
                                   5-23
    

    -------
    a 98-ton/hr  (112-MMBtu/hr) capacity reheat furnace.  LEA was
    selected as the most cost-effective option because the burner
    system did not need replacement.  Alternatively, another control
    option would likely have been chosen.  A test report shows the
    LEA control efficiency to be 14 percent.5  Assuming uncontrolled ,
    emissions of 220 ppm at 3 percent 02 (0.200 Ib NOx/MMBtu)  (from
    Table 4-4) for a recuperative furnace and a control efficiency  of
    13 percent, controlled NOX emissions are estimated to be 190  ppm
    at 3 percent 02 (0.174  Ib  NOx/MMBtu) .  The corresponding NOX
    reduction is 0.026 Ib NOx/MMBtu.   For regenerative furnaces,
    controlled emissions are 560 ppm at 3% 02 and 0.689 Ib/MMBtu.
    For cold-air fired reheat furnaces, LEA lowers emissions to  96
    ppm at 3%02 and 0.117  Ib/MMBtu.
    
         LNB or FOR may require major modifications to furnace
    structures and refractories to install alternate burners.  These
    combustion modifications do not necessarily -reduce efficiency and
    capacity.  The addition of LNB to a recuperative-fired furnace  is
    not likely to affect either capacity or fuel requirements.   The
    addition of FGR to a recuperative-fired furnace with or without
    LNBs will likely involve derating the furnace.  The addition of
    LNB plus FGR to a regenerative fired furnace will not likely
    reduce capacity.22
    
         Average controlled NOX emissions from two reheat
    furnaces(regenerative-firing) with LNB plus FGR controls are 150
    ppm at 3 percent O2 (0.18  Ib NOx/MMBtu) .to>"  Thus,  the control
    efficiency is 77 percent and the NOX removed is 0.61 Ib NOx/MMBtu.
    This efficiency is also used for recuperative and cold air
    firing.
    
         Two reheat furnaces with LNB NOX control,  one using
    regenerative preheated combustion air and one using cold
    combustion air, have controlled NOX emissions of 220 ppm at 3
    percent O2 (0.27 Ib NOx/MMBtu) and 80 ppm at 3 percent 02 (0.10 Ib
    NOx/MMBtu) ,  respectively.11'24  There are no uncontrolled NOX
    emissions data available for either of these furnaces.
    Uncontrolled NOX emissions from reheat furnaces using preheated
    regenerative combustion air are estimated to be 645 ppm at 3
    percent 02 (0.792 Ib NOx/MMBtu)  as  shown  in  Table 4-4.   Thus,  for
    the reheat furnace using regenerative preheated combustion air,
    the control efficiency is estimated to be 66 percent, and the NOX
    removed is 0.52 Ib NOx/MMBtu.
    
         From Table 4-4, the uncontrolled NOX emissions for cold-air
    fired reheat furnaces are 110 ppm @ 3% 02 (0.135 Ib/MMBtu).   The
    control efficiency is therefore 27% or 0.035 Ib/MMBtu.  For
    recuperative-fired reheat furnaces, the controlled efficiency for
    LNB is estimated to be 45% or controlled emissions are 120 ppm@
    3% 02 (0.11 Ib  NOx/MMBtu) .  The NOX removal  is 0.09 Ib/MMBtu.
    
         In Japan,  fuel switching, denitrification of fuel, LEA, and
    LNB's have been used to reduce NOX emissions from reheat
    
                                   5-24
    

    -------
    furnaces.8'19  No details about the performances of these controls
    are available.
    
    5.3.7  Soaking Pits
    
         Soaking pits have many similarities to reheat furnaces.
    Both are used to raise or maintain the temperature of steel in
    preparation for hot working or shaping.  They burn the same fuels
    (natural gas predominantly, but also COB, BFG, and oil) and
    operate at approximately the same temperatures.  Soaking pits
    tend to be smaller than reheat furnaces and, in the aggregate,
    use much less fuel.  In 1979, for example, soaking pits used only
    about one-fifth the fuel used in reheat furnaces.   Subsequently,
    the use of soaking pits has continued to decline as the practice
    of continuous casting has increased.
    
         No examples of soaking pits with NOX emissions controls  were
    found.  However, the similarity of reheat furnaces and soaking
    pits indicate that controls suitable for reheat furnaces would
    also be suitable for soaking pits.
    
    5.3.8  Annealing and Galvanizing
    
         Annealing and galvanizing are-two finishing processes often
    practiced in  iron and steel mills.  They are separate,
    independent processes; however, both can be accomplished in  a
    single continuous-type facility that includes, for example,  a
    continuous annealing furnace followed immediately by a
    galvanizing facility in which the continuously moving steel  sheet
    is immersed in a molten zinc bath.  As compared with most iron
    and steel process, which take place at temperatures of 1,090  to
    1,650 °C  (2,000 to 3,000 °F) , annealing and galvanizing are
    accomplished  at moderate temperatures usually below 540 °C
     (l,000°F); however, some annealing processes may require higher
    temperatures.  Because of'these much lower temperatures, NOX
    emissions  from these processes should be  lower-  These two
    finishing processes are discussed independently in the following
    subsections.
    
          5.3.8.1   Annealing Furnaces.  Annealing may be accomplished
    in a  large, batch-type  furnace processing tons of  coiled steel  in
    a single, multiday cycle or  in a large continuous-type furnace
    processing a  continuous sheet of steel passing through the
    furnace.  "Annealing is  a highly  specialized technology that  can
    alter the  properties of steel in useful  ways.  In  general,
    annealing  relieves cooling stresses  induced by hot-or-cold-
    working  and  softens the steel to  improve its  machinability or
     formability.   This is  accomplished by  subjecting  the  steel to a
     controlled thermal profile or cycle  with moderate  peak
     temperatures.
    
          SCR plus LNB, LNB plus  FGR,  and SCR are  being used  in the
     United  States to  control NO, emissions from annealing furnaces.
                                   5-25
    

    -------
    Information is available on  four annealing  furnaces,  one  that  is
    operational and three that are under const rue t ion.u>14>15  All are
    radiant tube, continuous annealing furnaces firing natural gas.
    Due to the many exhaust collection points at annealing  furnaces,
    there may be problems in installing SCR at  existing  furnaces.
    
         The control efficiency  of the SCR unit alone on the  LNB-
    plus-SCR controlled furnace  is consistently 90 percent  or more,
    and controlled emissions at  or near capacity firing  rates average
    33 ppm at 3 percent 02 (0.04 Ib NOx/MMBtu) .M There are  no
    uncontrolled emissions data  available for this furnace, only NOX
    concentration data referenced to the inlet  of the SCR.
    Uncontrolled NOX emissions  from two annealing furnaces are
    reported to be 1,000 ppm at  3 percent O2 (1.23  Ib NOx/MMBtu)  (from
    Table 4-4).  Thus, the control efficiency of the LNB plus SCR
    control is estimated to be 97 percent.  Individual control
    efficiencies of 53 percent for the LNB's and 93 percent for the
    SCR, for example, would yield this result.   The NOX removed by
    this control combination is  1.19 Ib NOx/MMBtu.
    
         Laboratory data are available for one  of the two LNB-plus-
    FGR-controlled annealing furnaces under construction.   Controlled-
    NOX emissions  from this  furnace are reported to average 180 ppm
    at 3 percent 02(0.22 Ib NOx/MMBtu)."  Uncontrolled emissions,
    from Chapter 4,  are 1,000. ppm at 3 percent  02 (1.23 Ib NOx/MMBtu) .
    Thus the control efficiency  is estimated to be 82 percent, and
    the NOX removed  is  estimated to be 1.01  Ib NOx/MMBtu.
    
           Controlled NOX  emissions for the  SCR-controlled annealing
    furnace under construction are estimated using 1,000  ppm  at 3
    percent O2  as  uncontrolled  emissions  (from Chapter 4) and
    assuming a control efficiency of 85 percent.  This yields
    estimates of controlled emissions and NOX removed of  150 ppm at 3
    percent 02  (0.18  Ib NOx/MMBtu) and 1.1 Ib NOx/MMBtu,  respectively.
    
         Although no data or examples are available, controlled NOX
    emissions and NOX removed for annealing  furnaces  with only LNB's
    for control are estimated by assuming a control efficiency of
    50 percent.  The resulting values are controlled NOX emissions of
    500 ppm at 3 percent 02  (0.61 Ib NOx/MMBtu)  and 0.61  Ib  NOx/MMBtu
    removed.
    
         No evidence of the use  of SNCR for NOX control on annealing
    furnaces (or other iron and  steel facilities) has been  found.
    However,  because of SNCR's similarity to SCR, NOX emissions from
    annealing furnaces with SNCR controls are estimated.  As
    discussed in Section 5.2.2.3, favorable temperature windows for
    the desired reactions differ between SCR and SNCR, and  the point
    of injection for the ammonia or urea reactant will differ.  These
    are design issues that will  have to be considered if  SNCR is
    adapted to annealing furnaces.  Assuming a  control efficiency  of
    60 percent for the SNCR control unit yields  controlled  emissions
    and NOX removed  estimates of  400  ppm  at  3  percent 02  (0.49 Ib •
    
                                   5-26
    

    -------
    NOx/MMBtu)  and 0.74  Ib NOx/MMBtu, respectively.2
    
         Similarly, controlled emissions and NOX removed estimates
    are made for LNB-plus-SNCR-controlled annealing  furnaces by
    assuming a control efficiency of 80 percent  (50  percent for  LNB   -.
    control and 60.percent for SNCR control).  The resulting
    estimates are controlled NOX emissions of 200 ppm at 3 percent 02
    (0.25 Ib NOx/MMBtu)  and 0.98 Ib NOx/MMBtu removed.
    
         5.3.8.2  Galvanizing Furnaces.  Steel products  are often
    coated with a protective layer of zinc, aluminum,  terne  (a lead-
    tin alloy) or a zinc-aluminum alloy to provide protection  against
    corrosion.  This process is  called galvanizing.  At  iron and
    steel mills, steel sheets,  strips, and other products with a
    light cross section are typically galvanized.  A typical molten-
    zinc bath temperature, for  example, is 450  °C (840 °F).l
    
         Controlled emissions data are available for two,  preheated
    combustion air galvanizing  furnaces with LNB plus  FOR NOX
    controls and one cold combustion air  galvanizing furnace with LNB
    NOX controls.11'21  Controlled  NOX emissions from the preheated
    combustion air furnaces average 220 ppm at  3 percent 02 (0.27  Ib  -
    NO /MMBtu).  Given uncontrolled emissions (from  Chapter 4) of  940-
    ppm at 3 percent 02 (1.15 Ib NOx/MMBtu) ,  the LNB  plus FGR control
    efficiency is  77 percent and the NOX removed is  0.88  Ib NOx/MMBtu.
    
         Controlled NOX emissions data for the cold  combustion air,
    relatively low-temperature,  LNB-controlled  galvanizing furnace
    are given  as  21 ppm at 3 percent O2 (0.026 Ib NOx/MMBtu) .21   In
    Chapter  4, uncontrolled NOX  emissions for this furnace were
    estimated  from the  controlled emissions  assuming a 34 percent
           _   _ _ i  a        _«*  ^f**. 	*	_J  '._ ._. ._ ^ J «•« ^ >^ *9  4— ^* V1^/-^ f*\  ^ "1 ^
     control  efficiency.
     NO./MMBtu.
    The NO, removed is  estimated to be 0.014 Ib
     5.4  SUMMARY OF CONTROLLED NOX EMISSIONS
    
          Table 5-2 summarizes controlled NOX emissions  data,and
     estimates for controlled facilities taken from preceding sections
     of this chapter.  The values tabulated in Table 5-2 are used in
     Chapter 6 to estimate cost effectiveness of the control
     techniques.   No controls are reported for coke ovens, sinter
     plants, blast furnaces or blast furnace stoves, basic oxygen
     furnaces, electric arc furnaces, or soaking pits in the United
     States.  the Japan Iron and Steel Federation reports experimental
     work with coke oven controls  (LEA, fuel denitrification, and flue
     gas denitrification), but no definitive data are provided.8  They
     also report experimental work with sinter plant controls
     (pretreating sinter materials and SCR).19  Again very little
     information is provided.  This work appears to be  experimental/
     very selectively applied, and not appropriate for  application in
     the United States.
                                    5-27
    

    -------
    CO
    
    
    O
    H
    I
    p<
    I
    H
    CO
    CO
    H
    I
    8
    N
     i
    in
    H
    
    I
               o
               H
            W
    
            P4
               Q
               W
            CO
                  OJ
            8
                        m   vo
                        H   VO
                                          o   o   o
                                          in   us   ko
                                                         in   o  o
                                                         oo   oo  a\
                                  H
    
    
                                  S
                                          c>-
                                          o
                                               r-
                                               o
                                                    O
    
                                                    o
                                                         o
    
                                                         o
                                     n
                                     o
                        <;
                        W
                        1
                        w
    oo
    
    o
     •
    o
                                  o
                                   •
                                  o
                                               VO   
    
                                                                          O
                                                         5-28
    

    -------
    5.5  REFERENCES
    
    1.   Mclnnes, R.,  and M. B. Van Wormer.  Cleaning Up NOX
         Emissions.  Chemical Engineering.  September 1990.
    2.
    3.
    4.
    
    5.
    
    
    
    
    6.
    
    
    
    
    7.
    
    
    8.
    
    
    
    
    9.
    
    
    10
        U.S. Environmental Protection Agency.  Control Techniques
        for Nitrogen Oxides Emissions from Stationary Sources-
        Revised Second Edition.  EPA-450/3-83-002.  January 1983.
    
        Campbell, Lisa M., D.K. Stone, and G.S. Shareef  (Radian
        Corporation).  Sourcebook:  NOX Control Technology Data.
        Prepared for U.S. Environmental Protection Agency.  Research
        Triangle Park, NC.  Publication No. EPA-600/2-91-029.
        July 1991.
    
        Plant Engineering.  March  1991.
    
        Letter and attachments  from Finke, H.P.,  Bloom Engineering
        Company, Inc., to Neuffer,  W.J.,  EPA/ESD. May 14,  1993.
        Comments on draft ACT document..
    
        Letter and attachments  from Lee,  D.,  Cascade Steel Rolling, .
        Mills, Inc., to  Jordan, B.C., EPA/ESD.  July 6,  1992.
        Response to Section 114 letter on iron and  steel mills.
    
        Finke, H.P.  Low-N0x Recuperative Burner Technology.  AISI
        Conference Paper. May  1988.
    
        The Japan Iron and Steel  Federation.   NOX Regulations and
        Control Technology in  the Japanese Iron and Steel Industry.
        AISI Conference  Paper,  May.  1988.
         Finke,  H.P.   Low NOX Recuperative Burner Technology.  AISI
         Conference Paper.,  May 1988.
    
         Letter and attachments from Postlethwait,  W.,  Nucor Steel,
         Darlington Division,  to Jordan,  B.C.,  EPA/OAQPS.
         December 2,  1992.  Response to Section 114 letter on iron
         and steel mills.
    
    11.  Letter and attachments from Gilbert, F.C., North American
         Manufacturing Company, to Jordan, B.C.,  EPA/OAQPS.
         November 11, 1992.   Response to. Section 114 letter on low-
         NOX burners.
    
    12.  Joseph, G.T., and D.S. Beachler  (Northrop Services, Inc.).
         APTI Course 415, Control of Gaseous Emissions, Student
         Manual.  Prepared for U.S. Environmental Protection Agency.
         Research Triangle Park, NC.  Publication No. EPA-450/2-81-
         085.  December 1981.
    
    13.  SCR Committee, NOX  Control Division,  Institute of Clean Air
         Companies, Inc.  White Paper - Selective Catalytic Reduction
                                    5-29
    

    -------
         (SCR)  Controls to Abate NOX Emissions.   September  1992.
    
    14.  Letter and attachments from McMahon, K.R.,  USS-POSCO
         Industries, to Neuffer, W.,  EPA/ISB.  August 17, 1992.
         Response to Section 114 letter on iron and steel mills.
    
    15.  Letter and attachments from Anderson, D.M., Bethlehem Steel
         Corporation, to Jordan, B.C., EPA/ESD.  July 13, 1992.
         Response to Section 114 letter on iron and steel mills.
    
    16.  lijima, K.  Environmental Control Measures in the Japanese
         Steel Industry.  Industry and Environment,  April/May/June.
         1980.
    
    17.  Ref.  1, p. 3-42.
    
    18.  Lankford, W.T., N.L. Samways, R.F. Craven,  and H.E.
         McGannon.  The Making, Shaping, and Treating of Steel, Tenth
         Edition.  1985.
    
    19.  U.S.  Environmental Protection Agency.  NOX  Abatement for
         Stationary Sources In Japan.  Research Triangle Park, NC. . _
         Publication No. EPA-600/7-79-205.  August 1979.             .
    
    20.  Barnard, W.R.  (E.H. Pechan and Associates,  Inc.).  Emission
         Factors for Iron and Steel Sources-Criteria and Toxic
         Pollutants.  Prepared for U.S. Environmental Protection
         Agency.  Washington, DC.  Publication No. EPA-600/2-90-024,
         June 1990.
    
    21.  Letter and attachments from Harmon, M.L., LTV Steel Company,
         to Jordan, B.C., EPA/ESD.  July 2, 1992.  Response to
         Section 114 letter on iron and steel mills.
    
    22.  Letter and attachment from Steiner, B.A., American Iron and
         Steel Institute, to Neuffer, W.J., EPA/ESD. May 25, 1993.
         Comments on draft ACT document.
    
    23.  Kobayashi, H.  Oxygen Enriched Combustion System  Performance
         Study, Volume  II - Market Assessment, Phase I - Final
         Report.  DOE/ID/12597-3, DE89 004061.  September  1988.
    
    24.  Letter and attachments from Sulc, D.A., Nucor Steel,
         Crawfordsville Division, to Jordan, B.C., EPA/OAQPS.
         October 6, 1992.  Response to Section 114 letter  on iron and
         steel mills.
                                   5-30
    

    -------
                                CHAPTER 6
    
                          COSTS OF NOX CONTROLS
    
    
         in Chapter 5, reheat furnaces, annealing furnaces, and
    galvanizing furnaces were identified as iron and steel mill
    processes to which NOX emissions  control  techniques  have  been
    applied.  In this chapter, costs and cost-effectiveness^estimates
    are developed for controls applied to these process facilities.
    The applicable control techniques for these furnaces are as
    follows:
         Facility
    
         Reheat furnace
         Annealing furnace
         Galvanizing furnace
    Control technique
    
    LEA
    LNB
    LNB plus FGR
    
    LNB
    SCR
    SNCR
    LNB plus FGR
    LNB plus SCR
    LNB plus SNCR
    
    LNB
    LNB plus FGR
         Controls applicable to reheat  furnaces are also considered
    to be applicable to soaking pits because of the similarity  of
    soaking pits and reheat furnaces.   They have similar purposes
     (i.e., they keep or reheat steel to temperatures  suitable for
    further processing), burn the  same  fuels  (usually natural gas),
    and  operate at about the same  temperature  (e.g.,  1,320  °C  [2,400
    °F]) .
    
         Costs and cost-effectiveness estimates are developed for
    controls  for models of each of the  three process  facilities
    described above.   For each facility,  the models correspond,
    approximately, to  small, middle, and  large sizes  selected from
    information obtained from iron and  steel industry personnel and
    control apparatus  vendors.
    
         All  costs are in April 1994 dollars.  Costs are based on the
    Chemical  Engineering Plant Cost Index.1
                                    6-1
    

    -------
         No differentiation is made in this chapter between new and
    retrofit installations of SCR and SNCR NOX control  techniques.
    These control technologies are add-on units, and cost differences
    would be site-specific.  In the case of LNB and LNB plus FGR
    controls, new burners and associated peripherals (ducts,
    plumbing, and controls) may not be required in existing burners.
    For new units the costs associated with combustion modifications
    are based on the incremental increase in costs for these burners
    over the cost of an ordinary burner.  In the case of LEA, the
    total capital investment (TCI) costs are for monitors and
    combustion controls that would be required for either new or
    retrofit installations.
    
         In lieu of definitive cost data, the following generally
    accepted relationship for scaling costs between facilities of
    different size or capacity is used as needed in this chapter:2
         Costs for size 1
         Costs for size 2
                     [ Capacity of size 1] °'6
                     [ Capacity of size 2 ]
    (6-1)
         Capital costs are annualized as needed in this chapter using
    the following capital recovery factor:3                          -
                   CRF •
                                                            (6-2)
    where
         CRF
         i
         n
    capital recovery factor,
    annual interest rate, and
    system economic life in years,
    An interest rate of 7 percent is used in this chapter.  The
    system economic life used, n, was taken from equipment life
    predictions provided by Section 114 respondents.  Cost
    effectiveness is determined by dividing the total annual cost of
    the NOX controls by the tons  of  NOX removed per year by the
    control.
    
    6.1  STEEL REHEAT FURNACES
    
    6.1.1  Model Furnaces                              .
    
         Steel reheat furnace capacities of 75, 150, and 250 tons of
    steel processed per hour are reasonably representative of small,
    mid-size, and large furnaces, respectively.  Therefore, these   ..
    capacities are selected as model facilities.  Further, a linear
    regression of available data from 10 examples that define both
    production capacities  (in tons per hour) and fuel firing
    capacities  (in million Btu's/hr) indicates that the corresponding
    fuel-firing capacities for these three sizes are 140, 300, and
                                    6-2
    

    -------
    520 MMBtu/hr,  respectively.4"9  The regression equation is
    
              MMBtu/hr = 2.17 (ton/hr)  - 21.1              (6-3)
         The 10 reheat furnaces used to generate Equation (6-3) use
    preheated combustion air, i.e., they are either recuperative or
    regenerative furnaces.  However to cover the entire population,
    reheat furnaces are separated into regenerative, recuperative and
    cold-air furnaces.  The uncontrolled emission factors are shown
    in Table 4-2 and are listed below.  These emission rates_are
    assumed to be relatively constant over a wide range of firing
    rates and furnace capacities.
    Firing Type
    Regenerative
    Recuperative
    Cold-air
    Uncontrolled Emissions
         ppm @ 3% O?          Ib/MMBtu
         645                 0.792
         220                 0.200
         110                 0.135
    6.1.2  Costs and Cost Effectiveness of LEA Control for Reheat
           Furnaces               ,                                .
    
         Costs and cost-effectiveness estimates for LEA controls for
    reheat furnaces are based on a single example for which costs and
    NOX emissions estimates are available.4  NO, emissions  for this
    furnace are estimated to be reduced by 13% by low excess air.
    The TCI costs and annual operating costs for this 112-MMBtu/hr
    furnace are $101,000 and $12,500/yr, respectively.4  In Table 6-
    1, these costs are scaled to the model reheat furnaces using
    Equation  (6-1).  Capital costs range from $120,000 to 260,000.
    The TCI costs are annualized using Equation  (6-2), assuming an
    economic life of 10 years and 7 percent interest.  For each
    furnace, the annual costs are the sum of the annualized TCI and
    operating costs.  Annual costs range from $32,000 to 70,000.  The
    NOX reductions are computed as the difference between
    uncontrolled and controlled emissions assuming 8,000 operating
    hours annually at capacity rate.  The cost effectiveness ranges
    from $330/ton for the  520-MMBtu/hr regenerative reheat furnace  to
    $3,200/ton for the 140-MMBtu/hr cold-air reheat furnace.   For the
    same size furnace, cost effectiveness for recuperative and cold-
    air fired furnaces are approximately four and six times that of
    regenerative furnaces.
     6.1.3   Costs  and  Cost  Effectiveness  of  LNB  Control  for
     Furnaces
                                              Reheat
          The  costs  and cost  effectiveness  of  LNB  controls  for
     regenerative  and recuperative-fired reheat  furnaces  presented
     here  are  based  on NCL emission reduction of 66%. LNB costs were
                                    6-3
    

    -------
    Cost
    Effectiveness
    ($/ton)
    §
    11
    °1 J.
    O
    "^ w g"
    60 ^
    Ijf
    5 U %
    Firing Type
    ctf t) _p
    M!
    s a 2
    C*i C*} fl
    3 2 S
    «N CN CS
    g g
    1 1 *.
    1 §• 3
    I 1 a
    i
    M!
    § ' s a
    !/•> \r> 10
    a s s
    III
    g g
    11 *.
    S 8. '?
    a g 2
    O 4J O
    8
    CO
    o 8 8
    co co 2K
    .CO * »
    ro Q O
    cs * ^
    00 0
    c^ c^ c**
    cs cs cs
    CO CO CO
    o o o
    \o \o vo
    CS CN 
    -------
    determined by using the costs for LNB/FGR for reheat furnaces
    shown in Section 6.1.4   and multiplying these costs by 0.67.
    The ratio in TCI's for LNB versus LNB+FGR capital costs is 0.67
    for 265 MMBtu/hr natural gas fired boiler in the
    Industrial/Commercial/Institutional Boiler ACT Document. 10   This,
    data is used as data was limited for the iron and steel industry.
    The only data for reheat furnace with LNB were for a 440 MMBtu/hr
    furnace where the PEC costs were $300,000 and the total burner
    system PEC plus direct and indirect installation costs for a 233
    TPH(555 MMBtu/hr) furnace under construction are $ 3,110,000.7
    As shown in Table 6-2, TCI for LNB controls for existing model
    furnaces range from $ 150,000 to 310,000 and for new furnaces
    from $40,000 to 90,000.  These costs are annualized using
    Equation (6-2), assuming a 5-year economic life and 7 percent
    interest (CRF « 0.244).  As shown on Table 6-2, annual costs
    range from $ 38,000 to 74,000 for existing furnaces and from
    $10,000 to $21,000 for new furnaces.  A control efficiency of 66%
    was used.  Cost effectiveness for new reheat furnaces range from
    $20/ton for the 520-MMBtu/hr regenerative-fired reheat furnace to
    $ 200/ton for the 140-MMBtu/hr cold-air-fired reheat furnace.
    Cost effectiveness for existing furnaces range from $70/ton for
    the 520-MMBtu/hr regenerative-fired furnace to $760/ton for the  -
    140-MMBtu/hr cold-air-fired reheat furnace.
    
    6.1.4  Costs and Cost Effectiveness of LNB Plus FGR Control
    
         Purchased equipment costs (PEC) were provided for three
    examples of LNB plus FGR controls for reheat furnaces.6'7
    Estimates of direct and indirect installation costs were also
    provided for one of these examples, and the sum of these costs is
    106 percent of the PEC.6  Therefore,  to convert PEC into total
    capital investment(TCI) a factor of 2.8 is used which is
    consistent with the factors used in other NOX ACT documents.
    The example furnace capacity is 78 tons/hr  (assuming a hot
    charge) and 154 MMBtu/hr.  Thus, it is reasonably consistent with
    Equation (6-3).  Its burner system PEC is $735,000 and thus, the
    TCI for this example furnace is $ 2,060,000 for retrofit
    installations (78 TPH) .  For a new 78 tons/hr reheat furnace the
    capital cost is 0.275($2.0 million) or $ 550,000.  The portion of
    the TCI attributable to the LNB plus FGR NOX control system over
    the TCI of a conventional burner system is 15% of the total.11
    Therefore, the TCI is $ 230,000.
    
         In Table 6-3, the TCI for the NOX control system for this
    example furnace is scaled to the model furnaces using the
    relationship defined in Equation  (6-1).  It is recognized that
    costs for a given  facility, new or retrofit, are site-specific
    and may differ greatly from the values used in this chapter.
    Capital costs vary from $ 62,000 to $130,000 for new sources and
    from $230,000 to 460,000 for existing sources.
                                    6-5
    

    -------
    Cost
    Effectiveness
    ($/ton)
    J-e
    1 *i
    iil
    «3 u S
    Firing Type
    Furnace
    Capacity
    I (MMBtu/hr)
    |V§1
    -.3 O O Q
    (fl {T) ^M \O
    X
    SW1 O
    ^H *
    | . . .
    ! « « «
    | , , .
    •a g g g
    1 1
    88,'?
    & 1 1
    Oi OJ O
    •-
    Vi O O
    
    -------
    Cost
    Effectiveness
    ($/ton)
    *Ji
    1 n|
    x^-
    3 -a §~
    Firing Type
    u £>=5
    M -| S
    ^ o o o
    2^r ^* ^O
    ^ ^M CS
    •g O O O
    S 8 8
    1 a .» a-
    f
    ,tS "^
    | S S '8
    a*
    •5 o o o
    IS en en en
    •g ts «s cs
    II
    'I ! •*
    ft, s -o
    g1 8 "o
    C^ P^ O
    s
    en en °
    a I I.
    en oo «N
    en en en
    es es «N
    CO CO CO
    00 OO 00
    . . .
    CO CO CO
    1 1
    1 1' I
    43 45 O
    06 06 U
    '
    O Q O
    « 5 "2
    8 1 1
    ^ en «s
    O O O
    en en en
    O O O
    CO CO CO
    1 I.I
    | |
    § 8 -a
    If 1
    06 06 U
    e^
    IT)
    6-7
    

    -------
         The annual costs in Table 6-3 are the annualized costs of
    the TCI.  In lieu of definitive cost data, it is assumed that
    other cost components, e.g., annual operating costs, are
    approximately the same as comparable costs for conventional
    burners and should not be charged to NOX control.   In annualizing,
    these costs, a 5-year life is assumed for the LNB system
    (estimates from 3 to 7 years are provided), and a 7 percent
    interest rate is used.  These costs are annualized using the
    capital recovery factor  (CRF) defined in Equation  (6-2).3  The
    CRF is 0.244.  Annual costs range from  $15,000 to 32,000 for new
    reheat furnaces and $ 56,000 to 110,000 for existing furnaces.
    
         The NOX reductions  in Table  6-3  are computed from the
    controlled and uncontrolled emissions assuming 8,000 hours of
    operation per year.  A control efficiency of 77% is used to
    determine emission reductions.  For new reheat furnaces, the cost
    effectiveness ranges from $ 30/ton for the 520-MMBtu/hr
    regenerative-fired reheat furnace to $ 930/ton for the 140-
    MMBtu/hr cold-air fired reheat furnace.  The cost effectiveness
    for retrofit reheat furnaces are 3-4 times the cost effectiveness
    of new reheat furnaces.
    6.2  ANNEALING FURNACES
    
    6.2.1  Model Furnaces
    
         Information was obtained on four annealing furnaces, one
    that is operational with an operating history and three that are
    under construction.  All are strip-anneal rather than batch-type
    furnaces.  The operating furnace has LNB plus SCR for NOX
    emissions control.12  Of the furnaces under construction, two will
    have LNB plus FGR and the third will have SCR for NOX control.5>7>13
    One of the LNB-plus-FGR-controlled furnaces is nearing completion
    and some emissions data are available.  The furnace capacities
    are given or estimated from data as 127 ton/hr and 95 MMBtu/hr,
    126 ton/hr and 180 MMBtu/hr, and 140 ton/hr and 95 MMBtu/hr.  No
    explanation for the sharp difference in product capacity to
    firing capacity ratios is provided.  For the purposes of this
    section, furnaces with firing capacities of 100, 200, and 300
    MMBtu/hr are selected as model annealing furnaces.
    
         The uncontrolled emission factors used in this section to
    calculate-emission reductions are listed below:
    Firing Type
    Regenerat ive
    Recuperative
    Cold-air
    PPM@ 3% P.,
    645
    220
    110
    LB/MMBTU
    0.792
    0.200
    0.135
                                   6-8
    

    -------
    6.2.2  Costs and Cost Effectiveness of LNB Control of Annealing
           Furnaces
    
         The control efficiency.used for LNB control is 50%. _  Costs
    and cost effectiveness are shown in Table 6-4.  Cost estimates
    for LNB controls are calculated by multiplying the cost of
    LNB/FGR as given in Section 6.2.4 by 0.67 as was done for reheat
    furnaces.  Only cost data for LNB/FGR were available for
    annealing furnaces.  The results of these estimates are presented
    in Table 6-4.   As was done earlier for reheat furnaces,  the
    capital costs for new furnaces was determined by multiplying the
    costs for existing furnaces by 0.275. Capital costs for retrofit
    applications vary from $ 360,000 to 690,000 and for new sources
    from $100,000 to 190,000.  To determine annual costs a life of 10
    years and an interest rate of 7% were used.  Annual costs for
    existing furnaces vary from $50,000 to 98,000 and for new
    annealing furnaces from $14,000 to 27,000.      For existing
    annealing furnaces, cost effectiveness ranges from $170/ton for a
    300 MMBtu/hr regenerative furnace to $l,680/ton for a 100
    MMBtu/hr regenerative furnace.  For new annealing furnaces, the
    cost effectiveness range from $50/ton for the 300-MMBtu/hr
    regenerative annealing furnace to $470/ton for the 100-MMBtu/hr  _
    annealing furnace fired with cold-air.
    
    6.2.3 Costs and Cost Effectiveness of LNB plus FGR
    
         The control efficency for this control was estimated to be
    60%.  This reflects the efficiency achieved by this control for
    boilers.  Costs for an LNB+FGR controlled, 180-MMBtu/hr annealing
    furnace are estimated as $2,400,000 for the burner system PEC.5'
    Using the 106 %  (for installation costs costs) and 15 %  (LNB_plus
    FGR  cost premium over a conventional burner system). TCI estimate
    of $760,000 for LNB+FGR control for 180 MMBtu/hr  is obtained.
    Equation 6-1 is used to calculate costs for the model furnace
    sizes.
    
         In Table 6-5, the costs and cost effectiveness of the LNB
    plus FGR control for model annealing  furnaces are given.  For  new
    annealing furnaces, capital costs range from  $150,000 to
    $280,000; annual costs from $21,000 to 40,000 and cost
    effectiveness from $60/ton to 530/ton.  Only .the  cold-air
    furnaces had cost  effectiveness greater than  $220/ton.  For
    existing annealing furnaces, capital  costs range  from $ 530,000
    to 1.03 million; annual  costs from  $75,000 to 146,000 and cost
    effectiveness from $210/ton to  1,900/ton.
    
    6.2.4   Costs and Cost Effectiveness  of SNCR for Annealing
            Furnaces
    
         SNCR has not  been applied  to any iron and  steel mill
    facilities.  However, the  similarity of SNCR  to SCR suggests  that
                                    6-9
    

    -------
    3
    1
    3
    §
    u
    i
    u
    f2
    I
    u
    w
    w
    s
    1.11
    ur
    C2
    o 'w
    §1!
    PS ~
    E
    J ^ -
    <£
    
    •1 » i
    u ° S
    
    ^
    eo
    •c
    £
    Furnace
    Capacity
    (MMBtu/hr)
    S o o o
    jjr f^ 8O f-
    
    fill
    i
    
    5? o o
    5) oo cr>
    1 2 2.2
    
    .S3 o - t~- t-
    O 0 O
     >o
    
    
    >o in n <— > O
    *— ' CO
    
    g 1 i
    
    00 ^ Os
    R B S
    
    OO 00 OO
    ON ON ON
    ON ON ON
    
    
    ON ON ON
    4) O
    •5 '-IS
    1 1 I
    O S O
    •
    

    -------
    *-> § c
    ur
    i'lt
    ^ £
    1 "I
    •^^
    H!
    Firing Type
    o £» ^
    >_ 3C YMI £f}
    ^* CS V)
    S"
    S 8 g 8
    iS
    8 ••
    |/8 s a •
    ! s s P
    ^H « »-l
    1 8 § -8 -
    u
    43 1>
    1 1 ^ •
    ^11
    « a; u
    g
    ...
    8 S i
    O O O
    o o o
    en en _ en
    § a a.
    ° (M CS
    2 2 2
    00 00 00
    •111
    60 W *5
    O 45 Q
    eS «! O
    •-
    • a,
    i ij
    ti a
    
    § § g
    a a. i
    o o o
    Sen en
    O O
    ti U
    ! i
    CCu
    1 fr 1
    II -3
    g
    en
    -11
    

    -------
    SNCR is potentially applicable to facilities  that are  controlled
    with SCR provided the flue gas stream to be controlled has a
    suitable temperature for the SNCR reaction to proceed.  As
    discussed in Chapter 5, a temperature suitable  for  the SNCR
    control reaction to proceed is much higher than the temperature
    required for the SCR control reaction.   SNCR is considered
    feasible for annealing furnaces and the potential control costs
    and cost effectiveness of SNCR were developed.
    
         Cost estimates have been developed for SNCR applications  to
    process heaters.  In the process heater ACT,  TCI for an SNCR unit
    for an 80-MMBtu/hr heater is $392,000.14  In lieu of better data,
    it is assumed that this TCI is the same as for  an SNCR unit for
    an annealing furnace of the same firing capacity.   In  Table 6-6,
    TCI for the model annealing furnaces are scaled from the process
    heater example using Equation  (6-1).  Capital costs range from
    $460,000 to 880,000.
    
         Operating costs are modeled in Reference 14 as follows:
    
         1. NH3  Costs  -  (Q)*(lb NOx/MMBtu)*  [0.370 (lb per  mole
                        NH3/lb per mole NOX) ] * (mole  NH3/mole
                        N0x)*($/lb NH3)*(8,760 hr/yr) * (CF) ,  (6-5)
    
         2. Electricity Costs =  (0.3 kWh/ton NH3) * (tons NHj/yr)  *
                        ($/kWh), and                             (6-6)
         3. Fuel Penalty Costs
    (0.015)*(Q)*(8,760  hr/yr)*($ fuel
     costs/MMBtu)*(CF),             (6-7)
    where Q is the furnace-firing capacity in MMBtu/hr and CF  is  the
    capacity factor, i.e., operating hours per year/8,760.   The tons
    of NH3/yr value in Equation (6-6)  is computed as (Q)*(controlled
    emissions-uncontrolled emissions) * [0.370 (lb per mole NH3/lb per
    mole NOX)]*  (moles NH3/moles NOX) * (CF) * (8760/2000) .   The
    0.3 kWh/ton NH3 value is  taken from Reference 14,   The fuel
    penalty cost estimate (1.5 percent  or the 0.015 factor in
    Equation 6-7) represents a loss in  thermal efficiency  in a
    process heater and is used here to  account for possible thermal
    losses from using SNCR on an annealing furnace.4
    
         For the purposes of this section, the parameter values used
    in the operating cost models are as follows:
              Natural gas costs
              NH3 cost
              Moles NH3/moles NOX
              $/kWh
              CF
      $  3/103 ft3
      $  164/ton,
      1.5,
      0.07,  and
      8000/8760.
                                   6-12
    

    -------
    S
    u
    i
    <
    ttl
    ctf
    1
    I
    w
    &
    s
     ca
     ea
     •4
    •JU
    O JjHj ^p;
    o C-
    X *"^ «•?'*
    0 3 H
    -3i iil/
    PS
    ilsl
    60 ,>>
    1 i § 1
    Pi
    Su
    60
    "S
    Furnace
    Capacity
    (MMBtu/hr)
    V§ 1
    
    
    o o o
    a. 2 *
    
    o o o
    ir> 
    O O O
    CO 00 CO
    o o o
    
    1 1
    ^ 2 «-
    1 1 1
    pg 05 
    -------
         The annual operating costs are estimated using the preceding
    models and parameters.  The total annual costs are the sum of the
    annualized TCI plus the annual operating costs.  Costs are the
    same for both existing and new furnaces although there may be a  ,
    wide variation in costs from one furnace to another.  As shown on
    Table 6-6, the total annual costs range from $150,00 to 370,000.
    
         The NOX reductions  in Table  6-6  are estimated  with a  NO^
    reduction efficiency of 60%.  Eight thousand hours of operation
    per year at furnace capacities is assumed.  The cost
    effectiveness ranges from $540/ton for the regenerative-fired
    300-MMBtu/hr annealing furnace to $3,800/ton for the 100-
    MMBtu/hr cold-air furnace.,
    
    6.2.5 Costs and Cost Effectiveness of SCR Control
    
         The control efficiency assumed for SCR is 85%. The PEC for
    an SCR unit in  a construction permit for an 95 MMBtu/hr
    annealing furnace is $1.25 million.15   Converting this to April
    1994 $(multiplying by 1.018), converting PEC to TCI by
    multiplying by 2.06 and using Equation 6-1, the capital costs as -
    shown in Table 6-7 range from $2.7 million to $5.1 million.  The-
    annual operating and maintenance costs were determined by using
    the costs in the Process Heaters ACT Document for the 69 MMBtu/hr,
    distillate oil-fired heater and using Equation 6-I.14  The  total
    annual costs are the sum of the annual o&m costs and the
    annualization of the total capital costs.  An interest rate of 7%
    and a life of 10 years were used to calculate the annualization
    of capital costs.  As shown on Table 6-7, annual costs range from
    $550,000 to $1,040,000.  Also on this Table, cost effectiveness
    varies from $l,100/ton to $ll,000/ton.  Only the regenerative
    fired annealing furnaces have cost effectiveness .less than
    $2,000/ton.
    6.2.6  Costs and Cost Effectiveness of LNB plus SNCR for
           Annealing Furnaces
    
         The control efficiency for LNB plus SNCR was estimated to be
    80%.  Capital and annual cost estimates for these controls are
    taken by adding the respective numbers in Tables 6-4 (for LNB)
    and 6-6  (for SNCR).  The resulting estimates of costs, NOX
    reduction, and cost effectiveness are tabulated in Table 6-8.
    For existing annealing furnaces, capital costs range from
    $820,000 to $1.6 million; annual costs from $200,000 to 470,000
    and cost effectiveness from $520/ton to 5,000/ton.  For hew
    annealing furnaces, capital costs range from $560,000 to $1.1
    million; annual costs from $160,000 to 400,000 and cost
                                   6-14
    

    -------
    Cost
    Effectiveness
    ($/ton)
    §1!
    tr
    pi
    Firing Type
    Furnace
    Capacity
    (MMBtu/hr)
    ** ** v^*
    t"M '^f ^^
    S 2 • o '
    CN CO V}
    CO "^
    v> 
    Vi  _^
    1 I
    S 8, '?
    0 3 •«
    00 O --^
    o C O
    ft (A U
    8
    US
    0 O O
    Cn Ch Cn
    q q q
    1 1
    88.'?
    8) S S
    5y S o
    CA & U
    '
    !- ! !
    82°
    IS!
    ui vi >/S
    Q> 4)
    _> £
    I \ 1
    Bi f* O
    8
    6-15
    

    -------
    s
    CO
    
    3
    a
    Cost
    Effectiveness
    ($/ton)
    *!§
    §11
    pi **~*
    *T3 •>?*
    PI
    Firing Type
    I ^1
    1,11
    ••= g s 8
    •2 5 g 8
    x « C3 Vi
    H
    8^D f^
    rj ^
    S O O O
    2 o
    o 6 6
    •I g S S
    •g d e> e
    « g
    's '2
    1 f 1 :
    & 1 a
    8
    g 8 I
    •* a S
    "« — cr,
    IS •
    OO 00 OO
    s • s §
    CO CO CO
    00 OO 00
    odd
    «, 2" 2
    4) O
    III
    1
    111
    •8' 1 1
    I.I!
    1 1 1
    o o o
    r~ r- r-
    ^ ^ ^
    vo vq vo
    1 1
    & S «-
    S 2, '?
    1 1 3
    a; as u
    8
    en
                                        6-16
    

    -------
    effectiveness from $440/ton to $4,000/ton.  Only the cold-air
    annealing furnace has cost effectiveness greater than $2,000/ton.
    
    6.2.7  Costs and Cost Effectiveness of LNB plus SCR for Annealing.
    Furnaces
    
         The control efficiency for LNB plus SCR is 90%.  Capital and
    annual costs for the LNB-plus-SCR are estimated by adding the
    respective numbers in Tables 6-4 (for LNB) and 6-7  (for
    SCR) .  NOX  reductions used in these computations  are based on an
    assumed 8,000 hours of operation per year at furnace-firing
    capacities. As shown in Table 6-9,  for existing annealing
    furnaces, capital costs range from $ 3.1 million to 5.8 million;
    annual costs from $600,000 to $1.14 million and cost
    effectiveness from $l,lOO/ton to $12,000/ton. For new annealing
    furnaces, capital costs range from $2.8 million to 5.3 million;
    annual costs from $560,000 to 1.07 million and cost effectiveness
    from $1,000 to 11,000/ton.  Only the regenerative furnace had
    cost effectiveness less than $2,000/ton.
    
    6.3  GALVANIZING FURNACES
    
    6.3.1  Model Furnaces
    
         Information was obtained on three galvanizing  furnaces.  One
    is a 79-ton/hr furnace, and the firing capacity of  this furnace
    is estimated from a  test report to be 150-MMBtu/hr.5  Firing
    capacities of the other two are given as  38 MMBtu/hr and
    179 MMBtu/hr.7  Using these furnaces as representative of
    galvanizing furnaces, firing capacities of 50, 150, and 200
    MMBtu/hr are selected as model furnaces.  The 150-MMBtu/hr
    furnace has LNB's for NOX control,  is a cold combustion air
    furnace, and its operating temperature is several hundred
    Fahrenheit degrees below many other galvanizing furnaces.16   The
    other two furnaces have LNB plus FGR for  NOX control and are
    preheated combustion air  furnaces.
    
         The uncontrolled emission factors used to calculate  emission
    reductions are shown below:
    Firing Type              ppm @ 3%O?          Ib/MMBtu
    Regenerative              940                 1.15
    Recuperative              330                 0.40
    Cold-Air  -                120                 0.14
    
    6.3.2  Costs and Cost Effectiveness of LNB Plus  FGR Controls for
           Galvanizing Furnaces
    
         Burner system purchased equipment costs for the  38-MMBtu/hr
    and 179-MMBtu/hr furnaces are given as $511,000 and $1,430,000,
    respectively.7  Assuming the 10.6 percent  (for installation costs)
                                   6-17
    

    -------
    %
    I
    C6
    
    ?
    M
    
    2
    
    §
    en
    Ed
    I
    I
    i
    I
    H
    Cost
    Effectiveness
    ($/ton) |
    
    o "5*
    K *** .?"*
    ill
    erf s~'
    lA
    3 In c5
    
    I
    oo
    E
    Furnace
    Capacity
    (MMBtu/hr)
    I!!!
    1 § 8 1
    fl - * a
    CO ~"
    1 R 1 1
    1
    >5J 00 00 00
    F— 1 ...
    a1
    • •< *• 4 t ^u
    W
    
    O O
    1 I -a-
    & 1 I
    erf erf o
    8
    51$
    I '•§' 1
    — CO P>
    g g 2
    \O CN «-i
    222
    OO OO 00
    OO OO OO
    •*•*•«*
    in v> \r>
    
    
    o «
    erf erf" C5
    •
    >^ pf vo
    § 8 '§
    -« P< r-
    R o 9
    I- 5 S
    § § §
    f«M V"4 V««
    CO CO CO
    00 CO 00
    
    
    fl> 
    -------
    and 15 percent (cost premium for LNB plus FGR as compared with
    conventional burner systems) factors, the TCI costs for NOX
    control for the two burner systems are $158,000 and $442,000,
    respectively.  These costs were converted to April 1994 $ by
    multiplying by 1.018.  A cost was obtained per MMBTU ($4,240 for -.
    38 MMBtu/hr and $2,510/MMBtu) for the 179 MMBtu/hr.  For the 50
    MMBtu/hr model furnace, the capital cost for existing galvanizing
    furnaces was determined by multiplying 50 by $4,240.  For the 150
    and 200 MMBtu/hr furnaces, $2,510/MMBtu was used.  The capital_
    costs for new galvanizing furnaces were determined by multiplying
    the capital costs for existing furnaces by 0.275.  As shown on
    Table 6-10, capital costs for existing furnaces range from
    $210,000 to 500,000 and for new furnaces from $60,000 to 140,000.
    
         Operating costs are assumed to be approximately the same as
    operating costs for a conventional burner system and, thus,  not
    chargeable to NOX control.   Capital  costs were  annualized using
    an useful life of 9 years and an interest rate of 7% giving a CRF
    of 0.153.  Annual costs for existing furnaces range from $32,000
    to 77,000 and for new furnaces $9,000 to 21,000.
    
         The control efficiency estimated for LNB + FGR is 60%.
    The emission reductions shown on Table 6-10 are also based on an-
    assumed 8,000 operating hours per year at furnace-firing
    capacity.  Cost effectiveness for existing furnaces range from
    $140/ton to 2,000/ton.  For new furnaces, cost effectivenes range
    from $40/ton to 560/ton.  '
    
    6.3.3  Costs and Cost Effectiveness of LNB for Galvanizing
    Furnaces
    
         In lieu of any cost data for the LNB-controlled galvanizing
    furnace, the cost estimates used in Table 6-10 for the LNB-plus-
    FGR-controlled model furnaces are multiplied by 0-.67 to obtain
    the costs for LNB only controls.  This is the same factor used
    earlier to calculate costs  for these combustion controls for
    reheat and annealing furnaces.
         LNB is assumed to reduce NOX by 50%.  This estimate is  used
    to compute the NOX reductions included in Table 6-11.   These
    reductions are based on an  assumed 8,000  operating hours per year
    at furnace- firing capacity.  The capital costs range from
    $140,000 to 340,000 for existing furnaces and 40,000 to 90,000
    for new furnaces.   Annual  costs range from $21,000 to 52,000 for
    existing furnaces and from  6,000 to  14,000 for new furnaces.
    Cost effectiveness range from  $llO/ton  to $l,500/ton for
    existing furnaces and from  $30/ton to 430/ton.
                                   6-19
    

    -------
    1
    o
    g
    •<
    o
    
    I
    ft
    -J
    fe
    o
    H
    u
    Cd
    
    fe
    
    8
    en
    
    I
    U
    Ed
    J
    CO
    ts If
    ur
    di
    OS
    - ^
    |0|
    
    1 « §
    <3°S
    Firing Type
    111
    ^ o g o
    S1
    1 '8 1 I
    Ed
    i-H .
    | . . ,
    *
    | 8
    ^^ ^5 ^3 vg^
    
    .5 m 0 O
    \3 O ^_ ^-
    | H " N-
    .1 I •
    !}!
    1 1 a
    o
    g O O
    ? 2 8
    1 3 °
    rf —<
    a a a'
    oo oo oo
    8 ,8 8
    
    OO 00 00
    CO CO CO
    > 5
    '' 1 '-a
    1 1 1
    8
    o ° 2
    * "-1 O
    — tn
    III
    8 S .P
    cv c< ts
    
    § i 5
    
    •:> •
    1 i
    1 ! 1
    1— C
                                               6-20
    

    -------
    o
    <
    i
    I
    J
    (El
    O
    
    8
    H
    O
    w
    
    I
    on
    
    S
    
    8
    v£
    
    w
              1
              § 'e'
    
              .£ 2
            S  tO
            S  o
            U
            W)
            III
            S &§
                                    22
                                    -t
    M
    
    
    I   «
    |   2
    
    a
                                8
    
    
                            «s
    
                            
    
                                           K
                                                    S
                                                    O
                                                    0
                                           8
                                                          oo   o
                                                          en   ON
                                                          §
                                                               CS
                                                               ON
                                                          CO   CO
    
                                                          bo
                                                                   o
                                                                   ss
                                                 O
                                                 o
                                      6-21
    

    -------
    

    -------
    6.4  REFERENCES
    
    1.   Economic Indicators:  Chemical Engineering Plant Cost Index.
         Chemical Engineering. July 1994. p.174.
    
    2.   Cooper, C. D.f and F. C. Alley.  Air Pollution Control:  A
         Design Approach.  Boston.  1986.  73 pp.
    
    3.   Vatavuk, W. M. Cost Estimating Methodology.  In:  OAQPS
         Control Cost Manual, Fourth Edition, Vatavuk, W. M.  (ed.).
         Research Triangle Park, NC, U.S. Environmental Protection
         Agency.  Publication No. EPA 450/3-90-006.  January, 1990.
         p 2-13.                       .    .     .
    
    4    Letter and attachments from Lee, D., Cascade Steel Rolling
         Mills, Inc., to Jordan, B. C., EPA/OAQPS.  July 6^1992.
         Response to Section 114 letter on iron and steel mills.
    
    5.   Letter and attachments from Harmon, M. L., LTV Steel
         Company, to Jordan, B. C., EPA/OAQPS.  July 2, 1992.
         Response to Section 114 letter on iron and steel mills.
    
    6.   Letter and attachments from Postlethwait, W., Nucor  Steel,
         Darlington Division, to Jordan, B. C., EPA/OAQPS.
         December 2, 1992.  Response to Section 114 letter on iron
         and steel mills.
    
    7.   Letter and attachments from Gilbert, F. C., North American
         Manufacturing Company, to  Jordan, B. C.,  EPA/OAQPS.
         November 11,  1992.  Response  to Section 114  letter on  low-
         NOX burners.
    
    8.   Letter and attachments from Finke, H.  P., Bloom Engineering
         Company, Inc.,  to Jordan,  B.  C., EPA/OAQPS.  October 14,
         1992.  Response to  Section 114  letter  on  low-NOX burners.
    
    9.   Letter and attachments  from Felton,  S. S., Armco Steel
         Company, L. P., to  Jordan, B.  C.,  EPA/OAQPS.  June 26,  1992.
         Response to Section 114  letter on  iron and steel mills.
    
    10.  Alternative Control  Techniques Document-- NO,. Emissions from
         Utility Boilers.  U.S.  Environmental Protection Agency.
         Research Triangle Park,  NC.   Publication  No. EPA-453/R-94-
         023.  March  1994.
    
    11.  Telecon.  Gilbert,  F.,  North  American  Manufacturing  Company,
         with  Parker,  C., Research Triangle Institute.   Section 114
         letter response.
                                   6-22
    

    -------
    12.  Letter and attachments from McMahon, K. R., USS POSCO, to
         Jordan, B. C., EPA/OAQPS.  August 17, 1992.  Response to
         Section 114 letter on iron and steel mills.
    
    13.  Letter and attachments from Sapia, J. P., Bethlehem Steel
         Corporation, to Method, T. J., Indiana Department of
         Environmental Management.  January 18, 1991.  Construction
         permit application.
    
    14.  Alternative Control Techniques Document--NOX Emissions from
         Process Heaters.  U.S. Environmental Protection Agency.
         Research Triangle Park, NC.  Publication No. EPA-453/R-93-
         015.  February 1993.  pp 6-6 to 6-7.
    
    15.  Letter and attachments from Sapia, J. P., Bethlehem Steel
         Corporation, to Method, T. J., Indiana Department of
         Environmental Management.  January 18, 1991.  Construction
         permit application.
    
    16.  Telecon.  Sistek, B., LTV Steel Company, with Parker, C.,
         Research Triangle Institute.  February 24, 1993.  Response
         to Section 114 letter.
                                   6-23
    

    -------
                                CHAPTER 7
    
             ENVIRONMENTAL AND ENERGY IMPACTS OF NOX CONTROLS
    
    7.1  INTRODUCTION
    
         This chapter discusses the environmental and energy impacts
    of the NOX emissions  control  techniques  identified in Chapter  5.
    The control techniques that have been applied or are potentially
    applicable to iron and steel mill process facilities are the
    following:
         Facility
    
         Reheat furnace
         Annealing furnace
         Galvanizing furnace
    Control technique
    
    LNB
    LEA
    LNB plus FGR
    
    LNB
    SCR
    SNCR
    LNB plus FGR
    LNB plus SCR
    LNB plus SNCR
    
    LNB
    LNB plus FGR
    Of these control techniques, LNB's, FGR, and LEA are combustion
    modification techniques, and SCR and SNCR are postcombustion
    control techniques.  Potentially, all techniques can impact other
    air emissions in addition to NOX,  and all may have energy
    impacts.  None have water pollution impacts.  However, SCR may
    have a solid waste impact.
    
    7.2  AIR IMPACTS
    
    7.2.1  NOX Emission Reductions
    
         NOX emission reductions for the emission control techniques
    are summarized in Table 7-1.  Control efficiencies range  from 13
    percent for LEA on a reheat furnace to  90 percent for LNB plus
    SCR on an annealing furnace.  Emission  reductions range from  40
    tons/yr for LNB control on a  150-MMBtu/hr,  cold-air galvanizing
    furnace to 730 tons/yr for LNB plus FGR controls on a 300-
    MMBtu/hr regenerative reheat  furnace.   The  cold-air galvanizing
    furnace reduction is low because  the uncontrolled NOX emissions
    are low for this relatively low-temperature furnace.
                                    7-1
    

    -------
    S3 oj
    o G
    CQ tjc<
    ivj ^ t"i ^^ •
    rn O CJ CM <—
    CO !S D E-" Q
    1 r8
    i4 O H
    |— I I I U1
    £ X E"* ^ j3
    8 ° °* H
    
    H W ^CD
    w rt g
    § w
    Sf ° S
    O H — S
    H §g^ji
    i Q *— * S
    CO § §
    § rt
    H
    B g^S
    g CO O S
    g H rf g
    0 SS|
    CQ
    CQ
    H
    8 #
    o
    8
    H
    C-
    W
    Q H
    P &
    B
    w
    o
    
    I
    
    o 3 § g ° g 3 § S
    i— 1 rH *"• ""I t"
    
    
    
    _OO OOOOOO
    *»o oo VD o o c^ in oo
    ^HH rHOJCNMCNCS
    
    
    
    
    OOO OOOOOO
    dCNn ooinin'^fooo
    tH VO f*» CO ^J* ^^ \& ^O **&
    
    
    
    
    
    OOO OOOOOO
    OOO OOOOOO
    mmro oaojojcNcsc^
    
    
    (V (V P, ct<
    CD CD fe CJ
    pTi p^| ^. ^~* 1^/3
    "^g ^ 1' ™ " B 1
    § § a J
    
    
    CD
    § H
    w d
    W H -
    H g
    
    
    o o
    «* in
    
    
    
    
    o o
    CM •*
    H H
    
    
    
    
    0 O
    «* CN
    rn ^*
    
    
    
    
    
    o o
    in in
    H rH
    
    
    «
    CD
    fn
    14 03
    J
    
    
    1
    N
    y
    3
    >
    CD
    7-2
    

    -------
    7.2.2  Air Impacts of Combustion Modifications
    
         Combustion modification techniques applied to process
    facilities at iron and steel mills are LNB's, LNB's with external
    FGR, and LEA.  LNB's incorporate one or more of the following
    combustion modification techniques:  LEA, off-stoichiometric or
    staged combustion (either staged air or staged fuel), and FGR.
    Any of these combustion modification techniques have the
    potential to increase carbon monoxide and unburned hydrocarbon
    emissions.  The NOX  reduction mechanisms  inherent  in these
    modification techniques are the reduction of peak flame
    temperatures that are exponentially related to the formation of
    thermal NOX and the  reduced availability  of  excess oxygen needed
    to form NOX.   Reducing the availability of oxygen  to the
    combustion process increases the likelihood that some HC will not
    be burned and that some CO will not be oxidized to C02.   However,
    it is not necessarily true that CO or HC emissions will increase
    if combustion modification techniques are used to limit NOX
    emissions.  If the control is properly designed and applied, NOX
    control can be achieved without increasing CO or HC emissions.
    Experimental evidence from tests with process heaters,  for
    example, indicate that decreases in excess oxygen levels begin ta
    impact CO emissions below about 3 percent excess oxygen.1  Also,_ -
    data from glass melting furnaces indicate that CO emissions begin
    to increase at excess oxygen levels of  3.2 percent.2  Limited
    data for ICI boilers  indicate that HC emissions do not  change due
    to the implementation of combustion modifications.3  Data in the
    utility boiler ACT show that CO emissions increase for  some
    boilers and decrease  for some boilers when implementing
    combustion modifications.4
    
         For the steel reheat furnace with LEA for NOX control,  CO
    emissions of 39 ppm at 3 percent O2 and 26 ppm at  3 percent O2 are
    reported for the uncontrolled and controlled tests,
    respectively.5  These results indicate that  CO emissions do not
    necessarily increase  when LEA is implemented for NOX control.   It
    should be noted that, in this example, LEA is actually
    implemented using a PID control loop that functions to  satisfy  a
    CO ppm setpoint input.  As the furnace operates independently to
    achieve individual  zone temperature setpoints, the CO control
    loop continuously adjusts zone oxygen  setpoints to minimize
    excess oxygen via the air-to-fuel ratio  in each zone without
    violating the CO controller  setpoint.
    
         Other available  sources report CO emissions  from
    LNB-controlled  reheat furnaces to be in  the  range  of 20 to  30 ppm
    at  3 percent 02.6  Further,  laboratory  and field tests on a
    galvanizing  furnace yielded  NOX emissions of 550 to 1,200 ppm at
    3 percent 02 and,  concurrently,  negligible CO emissions.
    Modifications  to the  burners reduced NOX emissions to 350 to 430
    ppm at 3 percent O2 and,  concurrently,  increased CO emissions to
                                    7-3
    

    -------
    30 to 60 ppm at 3 percent 02.7  The purpose of the modifications
    was to reduce NOX,  and further modifications  were planned.
    
         No data are available relevant to the impact of NO, control
    techniques on HC emissions from these iron and steel facilities. •
    In general, controls that reduce NOX by reducing the availability
    of excess oxygen in the high-temperature regions of a furnace
    would tend to increase both CO and HC emissions if the  control
    were not properly designed and applied.  In each case,  given the
    availability of oxygen, available CO and HC molecules could be
    further oxidized according to the following reactions:
    2CH4
                     30
               2 CO + O2
    2 CO + 4^0
    
    -*    2CO,
    (7-1)
    
    (7-2)
    The reaction rates for HC to HjO + CO and CO to C02 differ; thus,
    they are competing reactions.  Nevertheless, reducing the
    availability of oxygen to reduce NOX can also inhibit these
    reactions, resulting in an increase  in both  CO and HC emissions
    if the oxygen availability is reduced excessively.
    
         Reheat, annealing, and galvanizing  furnaces predominantly
    use natural gas as a fuel.  Natural  gas  does not contain sulfur,
    and thus, SO2 is not  a  pollutant of concern.8 Gaseous fuels,
    including natural gas,  can produce soot  and  carbon black when
    burned if insufficient oxygen is present.9  However,  no
    information on the effect of NOX controls on particulate
    emissions in iron and steel facilities is available.
    
    7.2.3  Air Impacts of SCR and SNCR
    
         As described in Chapter 5, SCR  units are add-on, flue gas
    treatment facilities that reduce NO, by injecting ammonia
    upstream of a catalyst reactor.  Within  the  catalyst, NOX reacts
    with the NH3 and is reduced to N2 and H20.  There is a potential
    for unreacted NH3 to  escape with the flue gas from the SCR unit.
    Any such emissions are referred to as ammonia slip.
    
         Section 114 respondents cite two instances of SCR  controls
    on annealing furnaces at iron and steel  mills.  One  of  these
    units is operational with more than  3 years' operating  history,
    and one is still under construction.10'"   In the  case  of  the
    former, the NH3  slip  was  guaranteed to be less than 10 ppm
    initially and less than 12 ppm after 1 year.  The observed NH3
    was initially less than 10 ppm.  Subsequent  observations are not
    reported.  The typical NH3/NOX molar  ratio for this unit is O.9.11
    
         For the unit under construction, a  NOX reduction of 90
    percent is guaranteed.10  The anticipated NH3/NOX molar ratio is
    1.0.  There is no NH3 slip guarantee.  The construction permit
                                   7-4
    

    -------
    requires 80 percent NOX reduction and a minimum NH3/NOX molar
    ratio of 0.8.  There are no requirements relevant to NH3 slip.
    
         Other reports indicate that NOX removal rates of 70 to 90
    percent can be achieved with SCR using NH3/NOX molar ratios
    between 0.9 and 1.0, and that the NH3 slip will be between 5 and
    10 ppm.12  These levels are considered to be well below health and
    odor thresholds.13  The Occupational Safety  and Health
    Administration  (OSHA) standard for NH3 in the workplace is 25 to
    30 ppm, and in the atmosphere it tends  to form nonhazardous_
    compounds and disappear within 2 weeks.14  In other applications
    in the United States, permitted NH3 slip levels are often on the
    order of 5 to 10 percent.  In Europe,  5 ppm is typically
    specified and achieved, and levels as  low as 0.5  to  2.0 ppm have
    been achieved; and in Japan, levels as  low  as  4 ppm  are
    specified.13
    
         No examples of  SNCR applications  to  iron  and steel mill
    process facilities have been found.   In other  SNCR applications,
    ammonia slip is controlled to acceptable  levels by controlling
    the ammonia or urea  to NOX molar ratio.  These levels are similar
    to ammonia emissions from  SCR applications, e.g.,  10 ppm.       - -
    
         Pilot-scale testing and chemical  kinetic  modeling of SNCR
    processes have  shown that  nitrous  oxide emissions are a by-
    product of both ammonia and urea injection. The  N20 formation
    resulting  from these processes has been shown  to  be  dependent  on
    the reagent used,  the amount of  reagent injected,  and the
    injection  temperature.15
    
         Full-scale tests on  fossil-fuel-fired boilers have shown
    that direct  emissions of  N2O are less than  15 ppm and do not
    generally  correlate with  NOX emissions.  Results from tests on  a
    small  pilot-scale  combustor using NH3  injection and encompassing
    a range of test conditions show a maximum N2O production of about
    45 ppm corresponding to an initial NO level of 700 ppm, an NH3
     injection rate corresponding  to a NH3/NO molar ratio of 2,  and  a
     temperature  of 880 °C (1,620  °F).   The corresponding NOX
     reduction was  75  percent.  The peak NOX reduction  corresponding
     to an  initial  NO  level of 700  ppm and an N/NO ratio of 2 was
     about  95  percent  at 980 °C (1,800 °F).  The corresponding N2O
    production was about 21 ppm or 3 percent of the NOX  reduction.
    
         N2O production  is higher for  urea injection  than for NH3
     injection.  Given the same initial NO level (700 ppm) and_
     reactant-to-NOx ratio  (corresponding to a NH3/NO molar  ratio of
     2),  the peak N20 production was  about  160 ppm  at  980 °C (1,800
     °F)  and the corresponding NOX reduction was at a  peak value of
     about  82  percent.   Thus the N2O  production  was about 28 percent
     of the NOT reduction.15    .      ;
                                    7-5
    

    -------
         Other pilot-scale tests have shown that, with urea
    injection, the level of CO present affects the NOX removal
    efficiency and the NOX removal  temperature dependency.   The
    presence of CO also affects N20 by-product emissions.   In the
    temperature range of 820 to 870 °C (1,500 to 1,600 °F), for
    example, increasing CO from 0 to 1,000 ppm caused an increase in
    N20 emissions  from 10  to  35  ppm.   The initial NOX level was 125
    ppm and the NH3/NO ratio  was 2.16
    
    7.3 SOLID WASTE IMPACTS
    
         Combustion modifications do not have solid waste
    environmental impacts.  There is a potential for SCR to have
    solid waste impacts in the disposal of spent catalyst materials.
    
         Titanium dioxide and "vanadia/titania" have been  identified
    as the catalysts in the two SCR annealing furnace applications
    cited.10'11  Other commonly used materials are  vanadium pentoxide,
    tungsten trioxide, platinum, zeolites, and ceramics.12'"  Of
    these, vanadium pentoxide is a toxic compound and a cause  for
    concern.  However, worker safety precautions adequately prevent
    any increased risk to workers handling the catalyst, and stack
    emissions of vanadium pentoxide are 1 million times less than
    industrial worker exposure.13
    
         Most catalyst manufacturers arrange to recycle and
    reactivate the catalyst.  When that is not practical,  the  spent
    catalyst can be disposed in an approved landfill in accordance
    with the Land Disposal Restrictions in 40 CFR Part 268,
    Subpart D.  The volume of catalyst materials has been  low  and
    their useful lives are 4-5 years for coal-fired boilers, 7-10
    years for oil-fired boilers, and more than 10 years for gas-fired.
    boilers.13
    
    7.4 ENERGY IMPACTS
    
         All of the combustion modification control techniques have
    the potential to impact energy requirements by affecting the
    thermal efficiency of the process.  No data are available  to
    quantify the impact of these controls on iron and steel mill
    process facilities.  A lack of discussion of these issues  in theft
    literature suggests that these impacts, if any, are not a  major
    concern in installations using combustion modifications.
    
         In the case of the SCR, the pressure drop across  the
    catalyst requires additional electrical energy for the flue gas
    fan.  Typical flue gas velocities over the catalysts are about 20
    ft/s, and the pressure drop is about 3 inches of water.  The
    volume flow of one typical unit is 1,550 ft3/m.   Given these
    parameters and assuming an efficiency of 0.65 for the motor and
    fan, the additional power demand is 0.84 kW, which corresponds to
                                   7-6
    

    -------
    $540/yr assuming 8,000 hours of operation per year and
    electricity at 8*/kWh.17
                                    7-7
    

    -------
    

    -------
    7.5  REFERENCES
    1.
         U.S.  Environmental Protection Agency.   Research and
         Development,  Emissions from Refinery Process Heaters
         Equipped with Low-N0x Burners.  Industrial Environmental
         Research Laboratory.   Research Triangle Park, NC.
         EPA/7-81-169.  October 1981.
    
         Barklage-Hilgefort, H.,  and W. Sieger.   Primary Measures for
         the NOX Reduction  on Glass Melting Furnaces.  Glastech.
         62(5):151. 1989.
    
         Alternative Control Techniques Document --NO,. Emissions  from
         Industrial/Commercial/InstitutidnaKICI)  Boilers.   U.S.
         Environmental Protection Agency.   Research Triangle Park,
         NC.  Publication No.  EPA-453/R-94-022.   March 1994.
    
    4.   Alternative Control Techniques Document--NOX Emissions  from
         Utility Boilers.  U.S. Environmental Protection Agency.
         Research Triangle Park,   NC.  Publication No. EPA-453/R-94-
         023.  March 1994.           ,                       -
    2.
    3.
    5.
    6.
    7.
     8.
     9.
     10.
     11.
         Letter and attachments from Lee, D., Cascade Steel Rolling  -
         Mills, Inc., to Jordan, B.C., EPA/OAQPS.  July 6, 1992.
         Response to Section 114 letter on iron and steel mills.
    
         Letter and attachments from Dickinson, G., Engineered
         Combustion Systems, Inc., to Jordan, B.C., EPA/OAQPS.
         October 12, 1992.  Response to Section 114 letter on low NOX
         burners.
    
         Letter and attachments from Gilbert, F.C., North American
         Manufacturing Company, to Jordan, B.C., EPA/OAQPS.
         November 11, 1992.  Response to Section 114 letter on low
         NOX burners.
    
         Perry, R.H., C.E. Chilton, and S.D. Kirkpatrick.  Perry's
         Chemical Engineers' Handbook, Fourth Edition.  New York,
         McGraw-Hill Book Company.  1963.  p. 9-8.
    
         Devitt, T.W. Fossil Fuel Combustion.  In:  Handbook of Air
         Pollution Technology,  Calvert, S.,  and H. England  (eds.).
         New York, John Wiley and Sons.  January 1984.  pp. 375-417.
    
         Letter and attachments from Anderson, D.M., Bethlehem Steel
         Corporation, to Jordan, B.C., EPA/OAQPS.  July 13, 1992.
         Response to Section 114 letter on iron and steel mills.
    
         Letter and attachments from McMahon, K.R., USS POSCO, to
         Jordan, B.C., EPA/OAQPS.  August 17, 1992.  Response to
         Section 114 letter on  iron and steel mills.
                                    7-8
    

    -------
    12.  Garg, Ashutosh.  Trimming NOX From Furnaces.   Chemical
         Engineering,  p. 122-128.  November 1992.
    
    13.  Institute of Clean Air Companies, Inc.  Selective Catalytic
         Reduction (SCR) Controls to Abate NOX  Emissions,  A White
         Paper.  Washington, DC.  September 1992.  25 pp.
    
    14.  Makansi, Jason.  Ammonia:  It's Coming to a Plaint Near You.
         Power.  p. 16-22.  May 1992.
    
    15.  Muzio, L., and T. Montgomery.  N20 Formation  in Selective
         Non-Catalytic NOX Reduction Processes.  .Paper presented at
         1991 Joint Symposium on Stationary Combustion NOX Control.
         Washington,  DC.  November 1991.
    
    16.  Teixeira, D.  Widening the Urea Temperature Window.   Paper
         presented at 1991 Joint Symposium on Stationary Combustion
         NOX  Control.  Washington, DC.   November 1991.
    
    17.  Vatavuk, W.M. Cost Estimating Methodology.  Ins  OAQPS
         Control Cost Manual, Fourth Edition.  U.S. Environmental
         Protection Agency.  Research Triangle Park, NC. Publication
         No.  EPA 450/3-90-006.  January 1990. p. 2-13.
                                   7-9
    

    -------
                    APPENDIX A.
    
    TABULATION OF UNCONTROLLED NOX EMISSIONS DATA
             A-l  Coke Ovens (underfiring)
             A-2  Sinter Plants
             A-3  Blast Furnace Stoves
             A-4  Basic Oxygen Furnaces
             A-5 Electric-Arc Furnaces
             A-6  Soaking Pits
             A-7  Reheat Furnaces
             A-8  Annealing Furnaces
             A-9  Galvanizing Furnaces
    

    -------
    

    -------
    
    
    
    
    
    
    
    
    
    
    s?
    •H
    
    
    t!
    0)
    
    !§
    
    CQ
    C
    I
    1
    a
    
    
    
    
    
    
    
    
    
    
    1
    o
    4J
    C
    Comme
    *
    IW
    Cd
    
    4J (1)
    m jj
    0) (8
    E-T3
    Sir
    S*
    3 m
    T3 a
    0 0
    
    O, «"
    01
    4J
    Emissions uni
    m
    c
    o
    •H
    m
    m
    X-H
    o e
    55 4)
    dl
    -5 3
    JJ
    4J CQ
    c8 2
    Oi 2
    a —
    "
    t-H
    s
    EL,
    
    
    rH
    0)
    4J
    0
    55
    
    
    n
    
    CO
    00
    
    n
    
    
    
    
    
    ^_^
    M
    0
    55 rH
    (8
    m o
    (8 3 O
    H CQ C
    T3 rt 2 C
    CTl O
    m in 01
    fO fl O rH
    
    n
    . X
    s e^
    CU pi .0
    pt P[^-
    CO CO r-
    •* CS 0
    r-
    
    
    
    
    
    
    O
    o
    
    
    
    
    s
    CTl
    *
    
    CO
    
    O
    O
    rH
    
    
    
    
    
    
    M
    O
    <*>
    n
    e E~-
    in o
    n cs
    o cs
    t-H
    
    
    
    
    
    O
    U
    
    
    cs
    0)
    4J
    O
    55
    CD
    w .
    cs
    o
    \
    rH
    
    .-*
    
    
    
    
    
    
    O
    <*= O
    CS
    C- M Z
    CQ
    ® ® 2
    6S|
    t—
    H CN M
    
    rH H
    
    
    ^
    ro
    vo
    O
    EL,
    CQ
    C3
    O
    U
    
    
    cs
    0)
    4J
    o
    s
    0
    CTl
    M
    
    rH
    CTl
    m
    
    
    
    
    
    
    
    
    O
    04
    <*> 0
    • <*> 3
    vo ro -U
    CQ
    ® ® 2
    2
    p) f^[ ^£
    VO
    0
    CTl O CS
    
    rH rH
    
    
    
    in
    CO
    CO
    O
    EL,
    CQ
    U
    O
    U
    

    -------
    •o
     to
    I
    4J
     O
     u
    -S,
    M
    PH
    P-i
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    fl
    •H
    •H
    4-4
    
    0)
    rj
    g
    
    
    CO
    rl
    0)
    
    ^
    CD
    0)
    
    *p?
    u
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    tn
    
    a
    §
    a
    E
    o
    u
    ~
    ""•
    
    0)
    Pi
    
    4J 4)
    1 to 4J
    4) m
    H TJ
    c
    0 —
    •H M
    1 1 Jf*
    0 ~-s
    3 tn
    •a c
    0 O
    ~~"
    
    tn
    4j
    •H
    C
    3
    
    tn
    c
    0
    -rl
    tn
    tn
    •rj
    £
    H
    ta
    c
    0
    •rl
    ta
    ta
    X-rl
    o E
    •g 4)
    n'l
    •rl 3
    U
    U £0
    m 2
    4) 2
    K —
    
    
    rH
    4)
    3
    
    
    
    
    
    
    
    
    
    O
    en
    ^
    N
    0
    Cn
    ^,
    H
    H
    
    
    
    
    
    
    M
    O
    
    c>P W
    ro o
    
    o #> 3
    H ro 4J
    a
    © © 2
    2
    e e •>-.
    Qj flj .Q
    
    m
    in
    rH •* O
    co p-
    eg *f
    
    
    
    
    
    
    O
    O
    ro
    0
    «
    O
    O
    U
    
    
    
    
    
    
    
    
    &
    o\
    ^
    M
    
    CO
    CO
    •*s^
    r-
    
    
    
    
    
    
    
    N
    0
    M
    tX> 0
    
    • #> 3
    r- n 4J
    CD
    © © 2 ^l
    2 fi
    
    ft ftrH rH
    (N
    ro
    •*
    ro co o CA
    eg 01 ^*
    rO ^ rH
    
    
    
    
    
    
    in
    ^J4
    ro
    O
    CO
    o"
    o
    0
    
    
    
    
    
    
    
    
    z
    en
    *
     3
    H ro 4J
    .a
    © © 2 M
    2 . i
    g £ ^^^^x
    Pt Pi.n .0
    O4 OirH r-H
    CO
    in
    cs j
    rH CN
    
    
    
    
    
    
    ID
    0
    rH
    
    
    O
    O
    U
    
    
    
    
    
    
    
    
    
    ^
    ^•^
    cn
    
    o
    o-v
    ^>^
    CN
    in
    0
    
    
    
    
    
    
    
    
    . —
    
    01 (tf 3
    > 
    
    Q4 iir-l rH rH
    oV
    ro vo
    CO O (N ^< ^
    p- ui o eg o
    co cn ro
    
    
    
    
    
    U)
    •
    in
    ro
    H
    
    
    O
    O
    O
    
    
    
    
    
    
    
    
    
    ^
    ~~-
    cn
    
    0
    cn
    *-^
    cs
    ro
    p-
    r>
    
    
    
    
    
    
    
    
    »—
    
    Oi rij 3
    > 4) 4J
    m a,m c
    *— •* ^— ' ^S* Li A
    A M u
    
    2 S *"*•* ***NI ^>*
    Q. O, O ^Q f>
    Qj CUrH rH rH
    
    ro rH
    p* o eg cn ^
    P- in 0 rH 0
    co o*\ ro
    
    
    
    
    
    VD
    •
    Ul
    ro
    H
    
    
    O
    O
    u
    
    
    
    
    
    
    
    
    
    2;
    ^.^
    cn
    
    o
    a\
    ^*^
    og
    in
    u>
    
    
    
    
    
    
    
    
    —
    
    01 ffl 3
    > 4) JJ
    r8 CUCQ C
    ^^ ^-^ 5?" Ll A
    •£< rl w
    
    S £ "^ ^x*^^-
    P, p, Q Q Q
    (i O
    a a
    — k 0
    x: 4J
    £ ***» ^^
    O4X) XI
    
    CO
    Oft?
    o ^ o
    Cn rH
    
    
    
    
    
    
    
    
    
    
    
    
    o
    o
    u
                                                            A-2
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    t
    •H
    -H
    4-1
    r-l
    CD
    "S
    
    
    CQ
    fl
    (U
    0)
    0
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    CQ
    
    g
    
    g
    o
    U
    I
    «j
    41
    
    JJ 4)
    CO JJ
    4) Hi
    EH T3
    C
    0 ~
    •H M
    JJ XI
    U ^
    3 CQ
    T3 C
    0 O
    rl JJ
    
    2
    •rH
    C
    3
    CQ
    s
    0
    •H
    §
    H
    a
    	
    CQ
    C'
    C
    a
    Cf
    
    O £
    S 4)
    S
    c -^
    •H 3
    JJ
    JJ CQ
    (S S
    4) S
    EC —
    ffl
    Be.
    
    
    
    t
    
    
    
    
    
    
    
    
    0
    •«— '
    i CD
    ro
    
    C?
    **-"
    0V
    0
    G\
    *^»
    CO
    CO
    rH
    ro
    
    
    
    
    
    
    
    
    __
    IS C
    XI JJ
    ftXl X!
    ftr-l rH
    
    C\
    O U> O
    O ^* r-
    0 CO
    C*J
    
    
    
    t
    
    
    
    
    
    
    O
    b
    CQ
                                                       CU
    
                                                       0
    
                                                       c
                                                       0
    
                                                      XI
                                                       CQ
                                                       tn
    
                                                      •H
                                                       rl
    
                                                      73
                                                       m
                                                       ca
                                                       0)
                                                       CQ
                                                       O
                                                       O
                                                       O
                                                       T3
                                                       IS
    
                                                       XI
    
                                                       •H
    
    
                                                       CQ  4)
                                                       0)  >
                                                       rH  O
                                                       ft     CQ
                                                       g T3  QJ
                                                       IS  C  g
                                                       CQ  0 -H
                                                        I   CJ JJ
                                                       tN  4)
                                                           CQ JJ
    
                                                       • •  IS  4)
                                                       4)      M
                                                       r-i  a  co
                                                       ft 0 M-l
    
                                                       IS  4) -H
                                                       CQ  WO
                                                           M
                                                    4) Xl  IS JJ
                                                   r-l IS XI  IS
                                                   Xi r)  O
                                                    is tn     >•«
                                                           tn M
                                                               0)
     o
    •H
                                                           c
                                                           •H  JJ
    
                                                    .       3   IS
                                                   tJ  M m  g
                                                    O      i   O
                                                    CJ  IW
                                                        0
                                                    JJ       >
                                                    c   • c
                                                    is  tn CD
                                                   ^  5 S  5
                                                    o.  «; o  Q
               4-1
    
              ' is
                                                      •  4)
                                                        JJ
                                                        0
                4)
                JJ
                o
    A-3
    

    -------
    s
    0
    M
    tn
    CO
    CQ H
    55 H
    5H
    H H
    CO t3
    CO H
    H CJ
    
    § r*
    
    XCO
    O CO
    & to
    §1
    ^ &
    >4
    Q J
    Qj |_*|
    pj H
    5^5 £
    O
    P
    6g
    CO
    •
    C4 P
    1 jgj
    ^i ^
    
    a
    X O
    M p{
    S H
    U
    CM
    PH
    <
    
    
    
    
    
    
    
    «— — »
    CQ
    ^^
    o
    rt
    4J
    CQ
    
    X
    0
    f
    -H
    gj
    *— •
    
    C/}
    4J
    C
    5
    O4
    S-i
    0)
    4J
    rt
    -H
    CO
    
    
    
    
    
    
    
    
    o
    JJ
    d
    E
    E
    0
    U
    
    *
    *— *
    ,
    U-l
    a)
    erf
    4j Q>
    01 JJ
    1) nt
    EH TJ
    O -^
    •H ^
    JJ 45
    3 a
    •0 d
    o o
    M JJ
    ft —
    
    
    
    a
    jj
    -r)
    d
    3
    
    01
    o
    •H
    m
    CO
    •H
    E
    H
    a
    d
    o
    •r)
    m
    CQ
    X-H
    O E
    
    -H 3
    JJ
    JJ CQ
    CS E
    0) S
    K —
    
    
    
    
    i-H
    (U
    3
    Cu
    
    
    
    
    
    
    
    
    a
    •^^
    CTI
    
    ^
    rH
    r-
    00
    •v^
    in
    
    
    
    i1
    
    r-
    in
    H
    
    
    
    
    ^^
    01
    
    (S
    
    "Si
    >3
    to JJ
    — CQ d
    son
    > E JJ 43
    S3 3 3
    
    
    m oo
    rH 0
    o N cf> in
    in o o oo
    *j* ^*
    
    
    p-
    •
    ro
    ^
    cs
    
    O
    
    
    aT
    >^<
    o
    0
    
    
    rH
    0)
    u
    0
    
    
    hH
    ^-*
    CTl
    
    *
    rH
    
    
    
    
    
    
    
    0
    •
    O
    00
    H
    
    
    
    
    .M.
    01
    
    ^ rt 'oi
    
    nf O) nj
    0) > 3 —
    ft 18 JJ
    — — ea d
    E 0 M
    > > 3S AJ.C
    Qj OiiH i-H rH
    
    
    in CTI
    CTI ro
    O O *}" «a* d
    CN in o o UD
    \O lf> 00
    
    
    ^4
    •
    ^
    t-
    ^
    
    O
    
    
    0)
    l>
    0
    u
    
    
    
    
    
    
    
    0
    -»-•
    CTl
    
    *
    tN
    
    CTl O
    ' CO CTl ^-^
    "*^ n ***» &•
    CTl 0 CTl -"
    
    
    
    
    
    
    
    
    1
    (B ^
    — O1
    M (8
    O *^ ^^
    01
     (M >
    m o n!
    » -^^
    in *>
    rH m 3
    jj
    © ® CQ
    E E ^^--
    pt pi n Q
    O. Oli-H rH
    
    
    CTl
    CO
    o in cs
    CTl rf O n
    CO CO CO
    CJ M
    
    in
    •
    01
    (T)
    "*
    o
    o
    CD
    
    oT
    ^*
    o
    o
    
    
    
    
    
    
    
    
    £5
    —
    CTl
    
    ».
     3
    n JJ
    ® CQ .
    ® S in
    IE-^
    Cu CU o _Q
    Cb CXrH rH
    
    
    ^t*
    vo
    o in M
    in in o m
    CTI CTI n
    m cs
    
    in
    •
    m
    rH
    «*
    o
    o
    u
    
    0)
    r^
    O
    u
    
    
    
    
    
    
    
    
    
    
    ^^^
    <:
    ^-^
    CTl
    0
    CTl
    ^^
    rH
    
    
    
    
    
    0
    in
    H
    
    
    
    
    
    
    
    ^x
    •^^> pirf
    01  0)
    ^"^ *— • J^
    M 0
    Pi Q{) ^Q VQ
    Ol GirH rH
    
    
    p»
    CTl
    in in in o
    in u> ^< o
    rH
    
    i
    
    
    
    
    
    
    
    
    i
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ^-^
     0)
    -53*
    
    >H O
    E E^.^.
    Pi Pi _f^ c^
    CU O-K
    ^
    01 rt
    > 0)
    <8 O4
    .— — • d
    >>|S
    a ax) xi
    ft fti-H rH
    
    
    P-
    H
    CN P- ^< rH
    P- P- CTl O
    rH
    
    t
    
    
    
    
    
    O
    O
    u
    
    a)
    ^*
    o
    • u
    A-4
    

    -------
    3
     8
              S
    •*"      -H
     CO
    4J
    
    
     s
    rH
    CM
    
     rH
     
                     m  4J
     o -.
    
    ~4J 43
     U^
     3  m
    •a  c
     o  o
    Emissions
                      S   H
                                             CN ID o
    
                                             rH CN O
                                                                     M 0
                                                                  > A U
                                                                     ~-
    
                                      > X!
    
                                                                                          OUr-H rH
                                      m
                                      m
                                      H
                                                                                     R1
    
                                                                                    i'   c
                                                                                        r4 0
                                                                                     >  43 W
    
                                                                                     ma xi
                                                                                     CUrH 1-H
                          O 0 rH
                          M in O
                                 o
                                0
    0
    u
                            O
                            U
          en
       in ^*
    o cs in
    
    <4< vo o
    ^ co
                                                          0
                                                          U
          rH
          O
    M vo in
    
    o co o
    ^< P-
                                                                                                  co
                                                                                                  en
                                                                                            o  
    -------
    
    
    
    
    
    
    
    
    
    
    
    
    •r-~»»
    ca
    o
    to
    
    CO
    X
    O
    =9
    a
    •H
    &
    
    W
    4J
    9
    rH
    04
    
    t 1
    o
    •H
    CO
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1 *""
    1
    L
    MJ
    a)
    
    -•
    l| CQ 4J
    11 fl) nJ
    EH TJ
    1 C
    
    1 vr^ ^
    II O^S-
    3 a
    •a c
    o o
    11 M xJ
    * ""*
    
    i
    
    i
    i
    j CO
    4J
    "c
    3
    tn
    c
    o
    •rl
    a
    m
    'E
    
    m
    c
    0
    •H
    a
    X-H
    0 6
    S 0)
    17
    'J3
    •rl 3
    4J
    ij ca
    nJ •£•
    m E
    
    rH
    0)
    3
    
    
    
    £J
    
    01
    N
    O}
    •^
    in
    
    
    
    
    
    n
    in
    H
    
    
    
    
    
    
    
    
    
    ?
    fd
    -- c
    rl O
    
    O.J3 JQ
    
    
    
    i-H
    (Tl
    O 00 U)
    in in o
    in o
    H
    
    
    i
    
    
    
    
    i
    
    
    
    
    
    P?
    
    a\
    
    at
    ^^
    in
    
    
    
    
    
    n
    in
    rH
    
    
    
    
    
    
    
    
    
    ?
    rt
    ^^ C
    >H O
    
    e\\
    pi Q f>
    O* i^ i~H
    
    
    
    CO
    o^
    o t*» vu
    r** vo o
    w o
    H
    
    
    1
    
    
    
    
    1
    
    
    ca
    u
    •H
     (U
    13
     o
     o
    Pk
                                                                    rt
                                                                    u
                                                                    0.
                                                                    a.
                                                                    rt
    0
    o
    
    4J
    a
    a
           A-6
    

    -------
    en
    OJ
    JJ
    CO
    
     g
    
     §
     en
     tC
    iH
    PQ
            ta
            JJ
    
            0)
            g
    
            0
            u
            H-l
            0)
    JJ  0)
    tn JJ
    0)  (8
    H
     3  tn
    
     0  0
     H  JJ
    p.  •—
     ta
    jj
    •H
     d
     3
    
     to
    
     o
    -H
     ta
     01
    •H
    
    U
         to
         c
         o
        -H
         tn
         m
      X-rl
     O  g
     a  «>
    •rl  3
        Jj
    JJ • CQ
     m  2
    
    ?E  *—*•
                      CO
    
                      in
                       o
                       u>
                       tH
                                 ?
                                 (8
                      tn 3
                      >  -W
               i-
               — -"— S
                         2
               g  g
                             >
                        Cil Qj
                                             CO
    
                                             CO
    •  *>  3
      d JJ
         ca
                                 r-  01
                                 Ul  t-
                                 O  O  VD
                       o  o in o  o
                         4)  3
     (8  OiJJ
    — ^- CO
           S M
     >  >  2 .C
     g
                                                      &, O.rH rH
                         CO     -*
                         iH-cn-fM
                   r> o o  co  o
                                                              in co o co  o
                                                      0
                                                      r-
                   C3  C3
                   fc  O
                   CQ  U
                                                                        ijl (S
                                                                        >  4)  3
                                                                        ro  Q.JJ
                                                                       ~^~- CQ
                                                                              2
                                                                        >  >  2
                                                                        g
                                                                         O. &rH r-l iH
                              in
                              (N
                        r> o o o
                                      co in o  o o
                                      H CN     CN
                                      CM
                                      r-
                        U  O
                        fa  O
                        m  u
    tn rt
    > a)
    rt
    
          2  i-i C
    >
    g
    Ot ftXl Xl -Q
    Q, ft.r-1 i-t
                                                                                                 IM
                                                                                              O 0  •
                                           o n o  vo o
                                           CJ M     H
                                           O
                                           co
                                           r-
                                                                                            ua
                                                                                            ft.  O
                                                                                            CQ  U
    ppmv (avg
    Ib/MMBtu
    
    Ib/hr
    Ibt
                          VD,
                          rH
                          0
                                                                                                                           0  CM 0
                        u o
                        fo O
                        ca u
                                                          A-7
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    in
    SP
    >
    
    S^
    4-<
    CO
    
    CD
    0
    (U
    e
    rj-
    
    ra
    rH
    CQ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    I
    0
    u
    
    1 v
    erf
    
    4J 4)
    (0 JJ
    1 0) a
    1 Hr°
    §~
    •rl JH
    0^
    
    3 0
    -0 n
    0 O
    1 M JJ
    o. ~-
    
    
    4->
    •H
    
    3
    §
    °
    •H
    tfl
    a
    I •r|
    1
    
    0
    •H
    
    X-H
    O E
    g a)
    •H>
    u
    4J CO
    (8 2
    0) 2
    a —
    «
    3
    
    
    
    
    W
    
    en
    CO
    CO
    "^
    o
    rH
    
    r-
    O
    M
    
    
    
    
    
    
    
    
    
    
    ?3
    18 4J
    — CQ C
    2 in 0
    ttrH rH rH
    VD CO
    H *f ro
    o en o
    c\i o vo o
    rH O O O
    
    
    
    VO
    •
    VO
    ro
    <*
    o
    fa
    CQ
    
    
    
    CB
    
    m
    00
    00
    ^.
    o
    rH
    
    [^
    O
    C*3
    
    
    
    
    
    
    
    
    
    
    (8 4-1
    -~CQ C
    2 >H 0
    > 2 43 4J
    E ^"v.-^
    G 3
    (8 4J
    *^ CQ 13
    2 SH 0
    > 2 43 4J
    &iXl XI A
    QjrH rH rH
    
    rH O
    O.rH ^<
    in o vo o
    co o in o
    
    
    
    in
    .
    0
    o
    in
    0*0
    fa O
    CQ U
    
    
    
    3
    
    en
    
    n
    
    *x^
    ro
    
    o
    
    H
    
    
    
    
    
    
    
    
    
    
    1?
    > 3
    (8 4J
    2 M 0
    > 2 43 4J
    O)i-H rH rH
    
    «* O'COV
    H en ^*
    CO O [*• O
    O O VO O
    rH
    
    
    in
    
    o
    o
    in
    O O
    ta O
    m u
    
    
    
    3
    
    en
    
    ro
    (Tv
    
    ro
    
    o
    
    rH
    
    
    
    
    
    
    
    
    
    
    "Si
    > 3
    (8 4J
    — CQ 13
    2 rJ 0
    > 2 43 4J
    OuXI X! XI
    CUrH rH rH
    
    ^ ^
    rH in ^^
    O O (N O
    rH O t~ O
    rH
    
    
    in
    
    0
    O
    in
    
    fa O
    CQ U
                                 (U
                                rH
                                O4
                                a
                                18
    
                                vw
                                •H
                                 a)
                                •o
                                 0
                                 o
    
                                4J
                                 (3
    A-8
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    
    
    
    CO
    0)
    o
    
    fl
    n
    fc
    
    _<
    0)
    en
    1
    o
    •H
    to
    rt
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    CQ
    ,j
    C
    
    01 4J
    41 n!
    H Tl
    
    C
    0 '—
    
    'jJ X!
    O •-•»
    3 n
    •o c
    0 0
    M 4J
    o, ^
    m
    
    •H
    a
    3
    Emissions
    01
    a
    o
    •H
    CQ
    01
    X-H
    o e
    S 0)
    X
    •H -
    4-
    jj m
    rt 2
    4) 2
    K •—
    c-i
    4)
    
    h
    
    
    
    
    
    
    
    
    
    
    
    
    3
    C*
    ^
    rH
    
    c-
    CO
    *-^
    ui
    
    
    
    0
    rH
    VO
    
    
    
    
    
    
    
    
    "w «
    > 41
    (S ft C
    •^^ M 0
    E e\\
    ft ftXt XI
    
    04
    O4
    1* O4
    ^ O LO O
    CO VO CO
    H rH
    
    
    
    
    
    
    
    
    O
    
    
    
    
    
    
    
    
    
    
    
    
    O
    0\
    «
    og
    O 0
    u\ o\
    *^^^~^
    0 H
    rH H
    
    
    
    
    
    
    
    
    
    
    
    O
    rH
    04
    H
    ® 0
    4J
    
    2
    o
    Vi> O
    
    
    
    
    
    
    
    
    
    0
    ^
    o
    C
    4)
    
    01
    •H
    01
    C
    0
    41 U
    E C
    0 -H
    W
    
    •3
    0^
    ^
    04
    
    00
    00
    ^-^
    ^
    
    
    
    
    
    
    
    
    
    
    
    CJ
    o
    m
    CO
    ® 0
    jJ
    E-^
    
    OS
    ro o
    rg o
    
    
    
    
    
    
    
    
    
    o
    O
    4)
    i-i
    O
    01 !>i
    C 0
    0
    4) -H vo
    Ol ffl
    (8 01 >H
    k -H 4)
    4) E >
    > 41 0
    <
    
    
    
    
    og
    H
    
    "\
    O
    rH
    
    
    
    
    
    
    
    
    
    
    
    O
    *
    rH
    C
    ® ^ 0
    A J->
    fti-H i-l
    
    fO
    C- rH
    O] «* 0
    rH O4
    
    
    
    
    
    
    
    
    
    0
    01
    4)
    r-i
    V
    01 >i
    C 0
    o
    41 -H 00
    01 m
    m m M
    M -H 0)
    4) S >
    > 4) 0
    <
    
    
    
    
    04
    H
    a\
    *^
    rH
    H
    
    
    
    
    
    
    
    
    
    
    
    
    
    ft
    ft
    
    
    •*
    rH
    og
    
    
    
    
    
    
    
    
    
    o
    O
    4)
    rH
    U
    m S
    C CJ
    o
    4) -H in
    O! 01
    nj 01 JH
    r4 -H 4)
    41 £ >
    > 4) 0
    rt
    
    
    
    
    og
    
    04
    CPl
    ^•^
    VO
    
    
    
    
    
    
    
    
    
    
    
    o
    co
    co
    rH
    ft
    ft
    
    
    04
    rH
    04
    
    
    
    
    
    
    
    
    
    O
    01
    01
    i-H
    O
    to >-.
    C 0
    o
    41 -H VO
    Ol ffl
    SH -H 4)
    4) E >
    > 4) 0
    "*•
    
    
    
    
    05
    
    04
    C\
    ^^v
    vo
    
    
    
    
    
    
    
    
    
    
    M
    0
    5
    vo
    rH
    ft
    ft
    
    
    <*1
    M
    rH
    
    
    
    
    
    
    
    
    (N
    O
    
    
    
    
    
    
    
    
    
    
    
    
    
    H
    - "*-'
     4)
    a ft 5
    >-•«— 0 in
    4J X!
    E E-^^
    rtj pj j^ j2
    ft fti-H r-i
    
    0
    in
    C\ O CO r-
    r~ m 04 o
    rH VO VO
    
    
    
    
    
    
    
    
    
    O
    
    
    
    
    
    
    
    
    
    
    
    
    
    H
    **-'
    
    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    to
    o
    8
    CO
    0)
    0
    CD
    |
    £
    a
    CD
    01
    >i
    x
    5
    asic
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    10)
    4J
    C
    CD
    E
    E
    O
    CJ
    •k
    *•*
    M-l
    Crf
    iJ <1)
    tQ jj
    0) m
    S-i t>
    
    
    
    3 "a"
    -a c
    1 t
    1 ^
    5
    
    C
    3
    to
    c
    0
    •H
    (Q
    1
    H
    
    m
    C
    0
    •H
    Q
    X-H
    o e
    2 0)
    'r»
    •rl 3
    4J
    (tf 3£
    QJ 2*
    Npl ^^ ,,
    •3
    3
    b
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    H
    «
    
    en
    **%.
    CO
    
    
    
    co
    rH
    
    
    
    
    
    
    __^
    1^4
    tn m
    > 41
    ma c
    A 4J
    Pl p| tQ ^Q
    ft ftrH rH
    
    
    
    VO
    CN
    ro in rH
    VO CN CN O
    H in in
    
    
    
    
    
    
    
    fN
    0
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    a
    en
    ro
    a\
    ^.
    ro
    
    
    
    vo
    CN
    ro
    
    
    
    
    
    
    
    ^_^
    tn
    m c
    — H 0
    ftrH r-H
    
    
    
    1*
    ro
    o o o
    CO rH O
    rH
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    CK
    «
    ro
    c\
    \.
    ro
    
    
    
    M
    ro
    
    
    
    
    
    
    
    ^^
    Cn
    (8 C
    x: 4J
    ftXI Xt
    ft rH rH
    
    
    
    in
    CN
    0 CN 0
    VO CO O
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    S
    C*
    ro
    Ol
    *^
    ro
    
    
    
    CN
    
    
    
    
    
    
    
    ^.^
    Cn
    S r.
    (o j5
    ^ r< O
    .C JJ
    ftXI XI
    ftrH rH
    
    
    
    VD
    ro CN
    o in o
    r~ co o
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ^
    0
    §
    ffl
    u
    rs
    t,
    M
    G
    0)
    01
    S"1
    i?
    
    •H
    at
    (8
    a
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    3
    0
    i-H
    CQ
    
    PI
    0
    Tl
    Cn o
    C -H
    •H .S-I
    r< O
    3 fti
    Q
    
    CQ
    m
    v
    
    en
    **s^
    VO
    
    
    
    in
    CN
    CN
    
    
    
    
    
    
    CJ
    o
    
    dp
    ro
    £
    ft
    
    
    
    
    
    ro
    m
    CN
    
    
    
    
    
    
    
    CJ
    0
    3
    0
    
    CQ
    
    PJ
    0
    73
    en o
    C -H
    •H rl
    M 0)
    3 a-
    Q
    
    CQ
    en
    *
    cs
    c\
    -^^
    VO
    
    
    
    in
    CN
    CN
    
    
    
    
    
    
    PI
    O
    
    dP
    ro
    ®
    E
    ft
    ft
    
    
    
    
    
    in
    o
    CN
    
    
    
    
    
    
    
    PJ
    o
    
    PJ
    O 73
    1 0
    C -H
    0 !H
    2 41
    a>
    en
    C 3
    -H 0
    
    3 CQ
    Q
    
    cn
    en
    •»*.'
    CN
    a\
    "^^
    vo
    
    
    
    in
    CN
    CN
    
    
    
    
    
    
    O
    
    *>
    ro
    6 E
    p, p,
    ft ft
    
    
    
    
    
    ro
    ^ vo
    rH VO
    ro
    
    
    
    
    
    
    
    
    
    P4
    o 73
    I. 0
    C -H
    O ^
    2 0)
    0<
    Cn
    C 3
    -rH 0
    JH r-H
    3 CQ
    a
    
    CQ
    en
    •*"
    CN
    
    ^^
    VO
    
    
    
    in
    cs
    cs
    
    
    
    
    
    
    CN
    O
    
    dP
    ro
    E E
    ft ft
    ft ft
    
    
    
    
    
    in
    ^ o
    rH 0
    CN
    
    
    
    
    
    
    
    
    A-10
    

    -------
    o
    C
          ~-   II   J
           o a
                     e  E--.
                     a, a xi
                     O< CUrH
                     H O
     Cn  (S
     >  01 jj
     us  Cb m d
    — — 0) OS
          J3 JJ
     e  s --^ •>
     a  a-ja A xi
     ft  O41-H i-H rH
                                                 CO
                                              o in
                                              VO t-t
                                      CN O   0) 4J
     (8  CU <8 C
    —'-' ID 0
          A w
     g  e---^-
     a  a, x> xi xi
     O,  OirH rH
                     o o in o o
                     rH CO CO
                     H m
                                                          U
                      en
                      ID
     JJ
     •H
     O
     •rl
     SH
    
     O
     01
     rH  N
     H o
     en m
     > 0) JJ
     m a. m c
    -'>-' 0) O S
          si jj
     e s-^--
     a nx> xt X)
     Ol CUH r-H
                                               O
                                            f- ID
                     H 0 O\ O
                     O rH O\
                                                                          6
                                                                        •H
                                                                        o
                                       O
                                       0)
                                       rH
                                       ca
                                                                                        in
                                                                                        u>
                                                                                        rH
                                                                                         en m
                                                                                         > 01 JJ
                                                                                         « tt rt
                                                                                        — ^ 01
         A JJ
    e  e^~~~~~
    a CU.Q
    fi4
                                               CT> 0
                                               0 M
                                            O IT) t-
                     VD O TJ<
                     CT» CN CO
                                         fc.
                                         U
                                         s
                                       S in
                                       A!   •
                                       VD CS
                                       •^ m
                                       VD
                      JJ
                      -H
                      O
                      •H
                      M
                      JJ
                      O
                      01
                      rH  M
                      W o
     en  (B
     >  01 JJ
     m  ft rt  C
    ^--— 01  O S
          J3 JJ
     S  e^--'~^
     a  O4X< xi xi
     Q,  CUrH rH
                              O rH
                              O CN
                           in in
    
                     O O CN O O
                     CO ^ 00
                        CO
                                                                            CC4
                                                                            o
    
                                                                         S  o
                                                                         .M   •
                                                                         CN  0
                                                                         CO  •*
                     JJ
                     •H
                      O
                     ••H
    
                     JJ
                      O
                      0)
                     rH   M
                     ra  o
                                                      A-ll
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    rrt
    Furnaces
    o
    u
    •H
    
    JJ
    U
    (U
    rH
    *^
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    I ^
    || n) -H
    •I *y E
    II *"" ^"*
    1 0)
    •u 6
    II ® *H
    
    II °
    II 4J
    II §
    II 6
    O
    0
    "«
    a:
    
    I 4J (B
    If CQ 4J
    II 4) (8
    
    c IT
    IOfl
    I -H (1)
    3 0
    •on
    [I O 0
    
    II
    11
    c
    o
    co m
    U 4J
    •H -H
    II E C
    W 3
    m
    o
    •H
    0)
    m
    X-rl
    O E
    S 0)
    4J '
    •H
    JJ
    C
    a
    3
    a
    
    rH
    a)
    3
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    M
    
    rH
    CFl
    v^
    CO
    tons/hr
    0
    o
    H
    
    d
    O
    *>
    rH
    
    @ rl
    j^
    g "s^s
    Qj^Q
    QjrH
    
    
    
    
    • •
    ^ C*J
    H CO
    
    
    
    1
    
    
    
    
    1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    CM
    
    rH
    Ot
    •*^
    CO
    tons/hr
    0
    o
    rH
    
    N
    O
    rH
    
    © rl
    ff*
    E "**-..
    Q.J2
    lirH
    
    
    
    
    • *
    cs r--
    rH U>
    
    
    
    1
    
    
    
    
    1
    
    
    
    
    
    
    
    
    
    O
    a)
    c
    rl
    .Q
    niJ
    0 O
    
    CM
    H
    O\
    "^
    O
    H
    tons/hr
    0
    0
    rH
    
    
    
    
    
    
    
    E
    O4
    a
    
    
    
    
    •
    r^
    
    
    
    
    i
    
    
    
    
    i
    
    
    
    
    
    
    
    
    
    0
    (U
    
    M
    ^
    Mt-l
    o o
    
    M
    rH
    Cft
    ^^
    o
    rH
    tons/hr
    O
    0
    H
    
    
    
    
    
    
    
    E
    
    Qj
    
    
    
    
    •
    [•«•
    1-1
    
    
    
    1
    
    
    
    
    
    
    
          .-§
        JJ  4J
        (S  rH
        o  a)
        43  E
         0 -H
        4J  W
    
        in T3
        vo
        H T>
            (1)
        >u  m
         0  3
    
         C  
        0  »
            ^
     C)  -O -H
    rH  0) "4-1
    .a  m
     ns  (S  N
     O  J3 rH
    •H     rH
    •-H  C  as
     CU 0 -rl
     O.4J TS
     (S  ^  nj
        J3  rl
        C rH
      «-rl  
    -------
    H
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    03
    I
    3
    r^H
    «
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    mments
    o
    u
    -~
    —
    IU
    Q)
    | 1 (J)
    en 4->
    EH T)
    
    
    ffl
    •JJ
    £
    c
    0
    -H
    m
    CO
    KJ
    to
    ^
    Q
    (Q
    X-H
    O S
    - i*^j d)
    g
    • EH
    43 »—
    g i-t r*!
    0 -H 0
    0  0 0 43
    e ^ -H a o
    o oio a
    JJ > Tjl  *
    ^^ — rH IS
    Dl Dl O 0)
    > > >, ID rH
    ^-. ""^ ' ljl^-fc ~— " Ijl " — " O t>i
    v> ^ >> ^ ^, o
    fll 4J ^-p*  oo
    OJ in O in rH CO rH
    OOlOinOO OrHO
    CO O^
    
    
    
    in in
    O rH
    in in
    
    co in
    
    
    „ _ '
    c c
    
    ±) \£
    , m m in
    O IS C rH
    — 43 0 IS •
    4-> -H O
    c— o a
    O Dl o pj
    JJ > TJ« S-U
    •~v. IS M JJ -rl
    O CU
    r-H —
    
    M
    01
    .-rH
    
    CO
    CO
    > —
    IS — D
    Dl >
    
    !_ ! E1 §^f
    ^ > ^ ^ >• o
    m^(C(B3nJ3 30 —
    OJ4J — ' . M
    OJ fC to
    rH 13 Tf 43 rl
    O 0) --s. O 0) -H -^ .
    >t m m Dl iS ^J1
    0 IS G rH iH 4J
    ^— 43 O IS IS • IS •
    4J O 43 43 ttl 0
    a — -H o g 43 a
    S £5 &2 8 |.H
    
    
    
    H
    w
    H
    r-
    CO
    CO
    ?
    f8 *" * Ol
    Ol *>
    
    01 pi o o)
    S > >i 01 rH
    ffl ^, IS — 0 rH 0
    	 — Ol-~— Dl — O >i
    2 > A: > x o
    n(3iS
    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    tn
    i!
    cu
    tn
    •H
    3
    0
    CO
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    1
    1
    Comments
    I
    
    •s
    erf
    I JJ 01
    0) JJ
    I 0) IS
    HT3
    
    I 1 (H
    1) IS 18
    r-H 13 Tl 43 >H
    O ID ^ O -rl
    >, ra m is •*
    O IS C rH JJ
    — 43 O IS • IB •
    jj O _O m o
    CJ e r* b^
    — •* ri C A A
    O OlO ft O 0)
    1 1 ^ ^t ^, U ^ 4J
    •^ IS CN JJ Oi-rl
    43 O Ot
    rH Si —
    
    P
    °I
    H
    C-
    CO
    CO
    "5
    IS —171
    Ol >
    V > IS
    — . — rH IS
    01 01 o u
    > > >1 U rH
    IS — IS^ 	 O rH O
    — — pi— — pi — o >,
    »3(SiS3iS 3O —
    U JJ — 4) JJ — JJ~-
    ft42 43 ft 42 43 43 43 43
    OirH rH OirH rH rH i-H rH
    
    
    in m *t" PI
    CO Tjl Ul Ul VD rH
    o o o in o PI en
    r-ocNinoo ocn •
    CO VO • O
    o
    
    
    o o
    c- r-
    
    
    r- vo
    rH
    
    ~O\ "oi
    a c
    •rl -H
    JJ y
    IS IS
    a o
    £ *}.
    
    T)l O
    0) rH
    rH
    V
    §1
    u —
    0)
    C Ol
    0 >, rl
    0) IS IS
    rH 'O 'O 43
    o ai~^~ o . —
    >, tn en a) PI
    O IS C rH Ol
    — 43 0 as JH •
    JJ O is O
    o "oi o "a o
    \ is oa jj jj -H
    42 O 04
    
    JH
    en
    ^
    r-
    CO
    ^~,
    CO
    "5 •
    is — "oi
    0) >
    4) > itf
    	 . — . rH IS
    Ol 01 O 4)
    > > >, 41 rH
    IS — IS — O rH O
    2~>2^^ ">,&•
    IS3JSIS3JS 30 —
    ft 43 43 O442 43 43 43 43
    ftrH i-H fti-H rH rH i-H rH
    
    
    rH O . rH ^l
    cn co en rH 01 TC vo
    H Ul O CO rH VO rH
    CN O ^* in O O O rH O
    rH O
    d rH
    
    
    O O
    cn o
    i" in
    
    t~ vo
    r^
    
    "S "9i
    C C
    •H -H
    
    IS IS
    V 0
    £ JB.
    
    *p cn
    
    
    
    O
    Z
    
    C
    0
    
    Ib/ton based
    29 tons/hr
    production
    
    s
    cn
    •*"
    a\
    CO
    -^,
    in
    
    
    
    
    o
    <(P M
    T*4 O
    cn <*>
    rH PI
    ® ® rl O
    ft ft 42 43
    ft ftrH i-H
    
    
    in
    t~ in H
    t^ O> T}4 O
    in co
    vo
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    2
    
    §•
    0
    
    a
    o
    
    Ib/ton based
    29 tons/hr
    production
    
    5?
    en
    •*.'
    en
    CO
    in
    
    
    
    
    ct
    o
    if o1
    rH
    rH PI
    ® ® rl 0
    Eg^<
    ft ft 43 43
    ft ftirH rH
    
    
    in
    PI rH (N rH
    Tjl Cft Tj< O
    in T|I
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    O .
    
    t
    U
    
    a
    0
    
    Ib/ton based
    29 tons/hr
    production
    
    en
    en
    •*
    en
    CO
    tn
    
    
    
    
    o
    o
    #> OJ
    in o
    Til JP
    rH PI
    CU Q*»Q »Q
    ft OirH rH
    
    
    " . VO
    CO rH rH
    in rH in O
    VO CO
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    8
    u
    
    
    
    
    
    
    5
    en
    N
    CO
    CO
    ^
    
    
    
    
    Oi
    O
    vo
    rH
    a
    ® o
    jj
    ft 42
    ftrH
    
    
    H
    VO
    PI
    in o
    pi
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    8
    CJ
    g
    in
                                                      ID
                                                      rH
                                                      43
                                                      2
                                                      3
                                                      43
                                                      JJ
                                                      -H
                                                      •H
                                                      ••s
                                                      JJ
                                                      m
                                                      m
                                                      ra
                                                    «s
                                                    is m
                                                    O-ri
                                                    •H
                                                    rH HI
                                                    ft C
                                                    Ol-H
                                                    IS rH
                                                     .a.
                                                    ^8
    
                                                    IS 4)
                                                    rH 43
                                                    d4 EH
                                                    *
                                                    — 0)
                                                      4J
    A-14
    

    -------
    0)
    CJ
    rd
    
    fi
    4-)
    td
    0)
    
    
    Comments
    £
    
    Crf
    0)
    4J
    fS
    Tl
    4J
    ta
    m
    E-i
    oT^-
    c£"
    O -H ^~
    -HUM
    4-> CIS Xt
    0 ft^
    3  ^ > >
    0) flj 0 flj 'H
    fa (8 1! fa (S
    a) o M o M T)
    U P* U) Ch CA  3 -H
    >H CN O 4J rH O UH
    3D  3
    ro JJ
    I ® S
    it
    
    
    o
    \o o
    CN
    ro
    
    •t
    
    
    
    
    SC
    fa
    CJ
    00
    o
    Si -:^-
    fa
    Qj o fa •
    Flue gas tern]
    = 900-1200
    = 500-600°:
    beyond Rec
    Recuperative
    fired
    
    in
    
    
    
    
    
    o
    ^^
    vr,
    
    
    
    i
    
    
    
    
    
    
    
    
    
    
    o
    *> 3
    ro 4J
    CQ
    B
    
    
    CN ;
    ro o
    rH
    
    1
    
    
    
    
    
    u
    •z.
    
    0.
    80" Hot Stri
    Mill
    Recuperative
    fired
    
    CN
    
    
    
    
    0
    
    o
    rH
    
    
    
    1
    
    
    
    
    
    
    
    
    M
    O
    M
    *> 0
    CQ
    E e^
    PI pi n
    Qj CUrH
    
    
    CO
    ro
    ID in o
    CN rH
    ro ro
    
    CN
    
    CN
    
    <*
    C
    z
    o
    o
    u
    
    
    110" Plate
    Mill
    Recuperative
    Fired
    
    CN
    
    
    
    
    O
    ^
    H
    H
    
    
    
    1
    
    
    
    
    
    
    
    CM
    O
    
    <*> CM
    ro o
    CN * 3
    rH (0 4J
    CQ
    6 g^.^
    Qu ftXt X
    
    en
    CN rH
    in m o ro
    o rH ro
    rH CN
    
    m
    
    in
    rH
    rH
    
    Z
    
    
    160" Plate
    Mill,
    continuous
    Recuperative
    fired
    
    CN
    
    
    
    
    0
    ov
    H
    rH
    
    
    
    1
    
    
    
    
    
    
    
    N
    O
    
    P- 0
    o #> 3
    rH rO 4-)
    CQ
    ® ® S rl
    S Xt
    g e^-^-
    O, QjrH r-(
    
    r-
    00
    CN rH
    ro ro o o\
    ro ro P-
    rH CN
    
    oo
    
    ID
    in
    CN
    S
    •*
    a
    o
    u
    
    
    160 " Plate
    Mill
    Recuperative
    fired
    
    CN
    
    
    
    
    
    CO
    00
    *^^
    00
    
    
    =
    
    
    
    
    
    
    
    
    
    N
    0
    dp O
    03
    • *> 3
    oo ro -U
    CQ
    ® ® S
    g E--.
    CL, Qjr— 1
    
    p-
    CN
    CN
    in in o
    CN 00
    rH rH
    
    
    
    tj4
    P-
    rH
    rH
    O
    U
    O
    o
                                                 A-15
    

    -------
    •O
     01
    
     §
    •H
    JJ
    
     §
     O
    r^
    
    
    rt!
    
    
    M
    H
    S4
    
    S
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    w
    0
    o
    tc
    
    £
    jj
    rt
    
    rH
    
    (D
    Pi
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    II
    1
    I
    1
    
    
    II to
    I -u
    1 c
    1 °
    1 e
    e
    0
    II u
    
    DJ
    -0
    w
    1 m
    1 ffl
    "
    Ls
    1 C -U
    0 -H —
    •H 0 H
    i> (tf X!
    3 m ta
    Tl 0 C
    a. o —
    a
    t!
    1 ""H
    C
    3
    2
    g
    »
    a>
    •y
    w
    c
    o
    •H
    a
    m
    o*-g
    8 4)
    1
    5>
    4J
    4J CO
    s s
    
    K —
    
    
    "0
    
    a
    
    
    
    0)
    ^
    •r) TJ
    4J <
    (S CO
    E h -a m ns
    W Q) 4) (U rH
    £G &i M ,O i
    3 -H 0)
    = O H-l 0 !q
    co 0) nj
    vo cri K
    cs
    
    •
    CO
    00
    *v^
    °
    
    1
    
    
    
    
    
    
    o
    
    *> 0
    cn
    • *> 3
    H n 4J
    CQ
    @ © 2
    2
    pt p, ^Q
    
    O
    
    0
    cn in o
    VO \fl
    
    
    
    ro
    
    cs
    VD
    H
    
    
    
    
    C3
    3
    la
    4J C
     r(
    x; -H
    0) 0) 4J CN
    cti o m
    w
    4J d 
    CQ Oi <
    r-
    
    
    H
    cn
    "^^
    r-
    
    1
    
    
    
    
    
    
    
    
    0
    c*P 13
    n jj
    CQ
    ® 2
    e^
    OirH
    
    cn
    o
    cn
    o o
    r-
    
    
    
    
    
    
    
    
    
    
    
    O
    *
    
    p,
    •H
    
    JJ
    CQ m
    
    4J M-l [fl
    O rH 0 JJ
    K rH , ffl
    •H • 
    oo ft.
    cn
    U)
    
    cn
    00
    ~^^
    <*•
    
    i
    
    
    
    
    
    
    
    
    o
    *> 3
    n jj
    CQ
    ttrH rH
    
    
    r-t
    ro
    O O (N
    vo r*i
    (N rH
    
    
    
    
    O
    cn
    n
    
    
    
    
    O
    
    
    
    
    
    
    
    
    i
    
    
    
    
    
    
    _^^
    a\
    
    CO
    f^.
    •v^
    *
    
    r-
    
    
    
    
    
    
    
    
    
    3
    4J
    CQ C
    2^0
    c^* xi xi xi
    ftrH rH i-H
    I-
    CS •<*"
    fO rH
    0 rH rH
    mono
    
    
    in
    00
    
    ^
    cn
    
    rH
    0)
    3
    &4
    rH
    U) -iH
    «: 0
    
    
    
    
    
    
    
    t
    
    
    
    
    
    
    ^ 	 	
    cn
    
    03
    
    >~
    H
    
    H
    in
    
    
    
    
    
    
    
    
    
    3
    CQ C
    S rJ 0
    OiXI XI XI
    ftrH r-H rH
    cn n
    O3 ^*
    CN in
    OHO
    mono
    cs
    
    
    
    in
    
    in
    
    H
    rH
    0)
    3
    Cu
    i-H
     2 Si 4J
    O
    -------
    •d
    0)
    3    g
    
    
    
    
    m
    «
    I
    8
    
    v
    
    
    
    
    
    1
    
    
    •^
    >-
    \ & <*
    1 «> II
    
    *
    •0 II
    «
    11 fl)
    11 ^
    S
    | O -rl —
    1 -H U rl
    1 4->  0) 3
    18 ft 4->
    ft ftja .a
    ft ftrH rH
    
    O
    ro
    O 00
    00 CM O rH
    CO VO
    
    .
    1 *»« II ";
    1 (8 S 1 ro
    *~ I*
    
    1 1
    
    «
    1 3 | 0
    III b |S
    
    
    
    
    1
    
    
    •j
    " • " .
    
    
    
    
    t-
    00
    
    in
    
    '
    
    
    
    
    
    
    
    
    3
    "5i <8
    >  4) 3
    flj ^ft-U
    ft ftrH rH
    
    
    in en
    o o
    co c^ o ro
    Tl< VO
    
    
    ro
    *
    in
    
    
    
    •z
    •' —
    o
    18
    4J
    W
    M
    O
    4J
    O
    1)
    •n
    U
    «
    en
    -
    
    
    
    
    o
    en
    
    CN
    
    r
    
    
    
    
    
    
    
    o
    <*>  N
    in o
    • *> 3
    r- ro JJ
    CQ
    S S
    Q. pt Q _Q
    ft ftrH 1-H
    
    
    in ro r-
    CM vo o in
    en vo CN
    rH M rH
    
    
    fS
    
    ro
    <
    H
    )
    1
    )
    3 U
    j S
    
    
    
    
    i
    
    M
    en
    vo
    
    
    
    
    en
    CO
    
    
    CJ
    0
    rH
    
    
    
    
    s
    
    O
    in o
    ™ *> 3
    co ro 4J
    CQ
    © © «£ JH
    e e^-^-
    CU *0i _Q _Q
    ft ft rH rH
    
    
    rH
    o ro in
    o in o in
    CO in rH
    rH CS H
    
    vo
    M
    r-
    ro
    
    
    
    O
    •z,
    
    
    
    
    1
    
    a
    en
    ^4
    
    
    
    
    o\
    CO
    
    
    17
    o
    CTl
    rH
    
    
    
    
    
    d
    O
    in o
    ™ *> 3
    r- ro JJ
    CQ
    © @ S r4
    e e *-«.• —
    0j f^j jQ Q
    ft ftrH rH
    
    
    in ro in
    vo co o in
    cr\ in rN
    rH CM rH
    
    en
    *t
    0
    <*
    
    
    
    O
    Z
                                            A-17
    

    -------
    •o
     a)
    §
    
    O
    r-
    
    
    <
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    W
    0)
    u
    fH
    ^
    •"<
    S
    4J
    nj
    0)
    •*•*
    QJ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    1
    
    
    S
    
    I
    u
    
    H
    «
    s
    18
    •0
    1
    ti
    S
    EH
    B
    0-H~
    I -H 0 rl
    II JJ (8 x!
    0 Q,---.
    II 3 (8 m
    i T> o a
    loo
    I rl H JJ
    1 CK 0 —
    m
    
    «
    1 3
    II
    Emissions
    \
    "£
    II Q
    _-
    a §
    "rT
    Ml
    
    JJ CQ
    rt i
    « s
    I
    1
    
    
    4) O
    > fa •
    •H T> .M
    JJ X! 4) O
    18 • JJ rH 4)
    M "O *H fQ »G
    4) 4) 3 18 O >,
    CM tQ r~
    4) -rl ffl -H JJ C
    O1M-I CQ T3 O O
    4) £3 Qi
    cc! ij w
    
    CO
    
    
    
    
    
    1
    
    
    
    CJ
    oo
    
    
    
    
    
    
    
    M
    O
    *>
    e
    Q<
    
    
    
    o
    in
    in
    i
    
    
    
    
    
    
    'Z
    
    4)
    
    •rl
    JJ
    *
    4) 41
    3 -H
    O MH
    4)
    -
    CTl
    CS
    
    
    
    o
    CTl
    ^»
    rH
    
    .
    
    
    
    
    
    
    CS
    0
    cjp (^
    P- 0
    cs <*> 3
    H n JJ
    CQ
    © © 2 >H
    66-^
    Qi QiXI A
    Q, Q,rH rH
    
    P-
    cs in
    ^t* p" o n
    0 CS rH
    
    CO
    •
    CO
    ^
    
    o
    o
    u
    cs
    
    4)
    JJ
    0
    
    rH
    fa
    4) o
    JJ O
    0 oo
    H
    CTl
    rH
    
    
    
    
    r-
    00
    in
    ^_^
    VD
    U>
    rH
    
    
    
    
    ^-^
    Oi
    ^s^8
    
    "Si <8
    > 4) 3
    18 Qi JJ
    — ~CQ C
    2 M 0
    > > 2 X! JJ
    s e^^^
    Qi QiXI X) X)
    Qi QirH rH rH
    O p-
    ^ CO
    rH •* M
    in in o CTI o
    cs vo m
    rH rH
    
    vo
    .
    rH
    CO
    CS
    
    O
    cs
    
    4)
    jj
    0
    S3
    rH
    fa
    4) o
    JJ O
    0 CTl
    £•£ ^t^
    W
    CTl
    1H-
    
    
    
    
    P-
    oo
    in
    "••^~,
    in
    p~
    rH
    
    
    
    
    
    
    ^^
    ^-^ .V
    Oi 18
    > 4) 3
    18 Qi JJ
    — — CO C
    !it!£
    PI Pi-H .n .0
    Qi Qir-H i-H rH
    V£ ^
    VO Tf
    rH W> CS
    in in o cs o
    VO CS -41
    H CS
    
    in
    
    u>
    in
     4) 3
    18 Qi JJ .
    •— — CQ C
    ' 2 rl 0
    > > 2 J3 JJ
    Qi QiXI A A
    Qi QirH rH i-H
    O U3
    rH \&
    rH O CS
    O CS O rH 0
    rH CO CS
    
    ^ ~ '
    .
    O
    CTl
    rH
    
    O
    U
    ,^
    
    d)
    JJ
    o
    '5.
    .
    fa
    4) o
    JJ 0
    0 in
    S •*
    '— ' •'
    CTl
    rH
    
    
    
    
    P-
    00
    in
    cu
    VO
    P-
    in
    rH
    
    
    
    
    
    
    ^^
    **• .y.
    Oi 18
    > 4) 3
    18 QiJJ
    -"~ CQ C
    > > 2 XI JJ
    Q, Q,X> Xl Xt
    Qi QirH rH i-H
    Vfl P-
    CS CTl
    rH 0 rH
    in oo o rH o
    cs in c>
    i-f rH
    
    CTl
    
    in
    if
    
    
    O
    U
                                                                A-18
    

    -------
    I
    •ri
    4J
    
    §
    U
    
    
    
    -
    
    
    
    
    
    
    
    
    
    
    
    tfl
     I \
    ^•^ CJ
    cu ~—
    >,
    C 4J
    0 -H ^~
    •rH U rH
    4J (B X!
    fj f^^^
    3 (S to
    Tl 0 C
    0 0
    tj tjj _JJ
    cu o — •
    a
    4-1
    C
    ^
    i suoTssTUig
    0)
    0
    •H
    to
    a
    X-H
    O S
    
    
    Aj
    
    01
    O
    PI
    
    
    
    
    
    
    
    
    ^ 	 	 l£
    > 4) 3
    CS 0U4J
    > > S 42 4J
    E E -^\\
    PI pt.n >Q ^
    Qj pji-H i — 1 r-
    
    vo r*
    C^ 0
    ui o o eg o
    II r^ in -^
    a 4) II r-l fN
    •u 1 in
    4J ffl I
    SI 5
    a — 11 CM
    '£ 1 8
    CN
    
    fl>
    JJ
    o
    rH
    4) o
    4J O
    0 (N
    a •<*<
    in
    m
    -i
    t-
    00
    in
    
    
    oT
    .^
    r-
    EN
    
    
    
    
    
    
    
    
    — A:
    01 m
    > 4) 3
    rt ft j->
    — ~CQ • C
    S H 0
    •> > 2 A 4J
    E E ^.->-v-.
    ft ftXI XI XI
    ft ftr-l r-H r-t
    
    rH 0
    rH 01 C
    o in o in o
    •cf 0. PI
    i-H rH
    00
    in . • .
    0
    o
    o
    cs
    
    4)
    JJ
    0
    0) o
    4J 0
    0 VO
    a <*
    H
    Ol
    H
    C-
    CO
    in
    
    
    sr
    >«-^
    00
    rH
    H
    
    
    
    
    
    
    
    
    ~*i
    01 a
    > 4) 3
    (S ft4J
    ~-^ CQ C
    ft ftXl XI XI
    ft ftr-H r-l r-l
    
    O VD
    O C^
    rH CN H
    O O O O O
    o in PI
    
    
    0
    PI
    u
    —
    
    4)
    4-J
    0
    a
    4) o
    4J 0
    0 N
    H
    Ol
    H
    r-
    00
    m
    
    
    eu
    •— '
    H
    0
    H
    
    
    
    
    
    
    
    
    "01 m
    > 4) a
    rt ft 4J
    — ^ ffl C
    S rH 0
    > > S X! 4J
    E E \\\
    ft ftXl XI XI
    ft ftr-l i-l rH
    
    in c^
    01 in
    O 00 i-l
    in in o 01 o
    Ol 0 CN
    1-1
    r-
    PI
    rH
    n
    o
    o
    0
    0)
    4)
    > 4)
    •rH 4J
    4J 0
    m a
    rH 13 ^
    4) 0>
    ft rH fa
    3 -rH 0
    O 14-1 C-
    4) 1*
    ce! •*
    
    S
    Ol
    Ol
    PI
    
    
    
    
    
    
    
    
    
    
    
    
    
    ppmv (avg)
    Ib/MMBtu
    Ib/hr
    Ib/ton
    PI 00
    03 in
    o ^ o
    01 o m o
    i-H 0 i-H O
    
    in
    0
    rH
    O
    O
    U
    CS
    4)
    r> 4)
    •rH 4J
    4J O
    (« a
    4) 4)
    ft fH fa
    3 -H o
    O 4-4 PI
    4) VO
    
    fti
    Ol
    PI
    Ol
    n
    
    
    
    cu
    w
    in
    
    
    
    
    
    
    
    
    ppmv (avg)
    Ib/MMBtu
    Ib/hr
    Ib/ton
    00
    PI (N
    0 0 rH i-l
    Ol O VO O
    O O O O
    
    in
    o
    u>
    rH
    O
    U
                                               A-19
    

    -------
    I
    4J
     8
     u
    0)
    o
               4J
               rt
               (U
               a)
                        0)
    
                        E
                        0
                        O
                        a)
                       •OS
                        0)
                        jj
                        nj
                       •a
                        (0
                        (I)
                       EH
                       --> O
                       a. •—
               -H ^
    •rl  O
    JJ  (8 JS
     O  Qj"^
     3mm
    TJ  o  a
            •H
            c
                                
    M  C
    3  (8
    JJ
    n>   >
    M  Cn
    (D  r;
    O.-H
    E  Cn
    (II  M
                                                             (U
                                                             X!'
                                                  •o
                                                  C  * 4J
    
                                                      0)  11
                                                  Cn  c x!
                                                  com
                                                  •H  N  M
    
                                                  "  Cn
                                                      C  M
                                                    -H -H
                                                         (8
                                                  C  (8
                                                  -HOC
                                                      CD  0
                                                  •O    -H
                                                  Q)  C JJ
                                                  ^ -H  m
                                                  •H     3
                                                  MH T3 XI
                                                  i    -H  O
                                                  •H M-l
                                                  JJ  I   O
                                                                           0)
    
                                                                       Cn 0
                                                                       C  N
                                      0)  0
                                      x:  m
    
                                      C  fit
                                      •H  0
                                         JJ
                                                  ^ -H
                                                   -H
                                      -H M-l
                                      JJ  I
                                      It  ^
                                      >H -H
                                      o)  m
    
                                      3 T3
                                      O rH
                                      0)  O
                                      ft! O
                                                                   w  n
    
                                                                    (U  0)
                                                                   jj  JJ
                                                                    o  o
                                                                                A-20
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    u
    0)
    u
    (0
    a
    r-l
    £
    HN
    tn
    C
    -H
    rH
    (TS
    OJ
    £
    C
    rf
    •^H
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    II
    II
    
    
    
    
    j
    •*
    Comtne:
    U-)
    4)
    04
    4-J OL
    U CTJ I
    H-n
    c
    0 —
    •H M
    4J X!
    o •»•.
    3 m
    •0 C
    0 0
    
    0)
    m >
    C -H
    T3 0 -U
    d) -H (8
    4J CQ ^ ^
    (8 [Q 4) 4)
    g-H C H
    -HE 0) -H
    4J 0) 0T4-I
    ffl 4)
    pq O4
    
    
    CO
    
    
    
    
    1
    
    H 4J 1
    01 ~1
    .
    4J
    •H
    c
    3
    c
    0
    -H
    CO
    tfl
    1
    to
    c
    0
    •H
    CD
    m
    X-H
    0 E
    a 4>
    cs
    -S^
    4J
    4-> m
    is s
    4) S
    EC —
    i-H
    0)
    
    
    
    o
    *> 3
    m 4->
    a
    ® s
    
    e \
    Oj »fl
    &! 1-H
    
    in
    
    u>
    
    0 0
    in
    in
    
    
    
    
    
    vo
    H
    
    
    0
    t!
    4)
    ca m
    4) m w
    J> ^Q (8
    -H -O
    <8 C J3
    M -a o as
    4) 0) -H r-4
    C H CQ
    4) -rl CO C
    01 ^W -H 0
    0) E
    04 H ,
    
    
    03
    co
    
    
    
    i i
    
    
    
    
    
    CJ
    o
    *> 3
    00 4-J
    a
    © g
    s
    e -^
    pt jQ
    Cb rH
    
    
    ro
    CS
    * *
    O rH
    0
    O
    H
    
    
    
    
    VD
    CO
    rH
    
    U
    T3
    4)
    co m 11
    4) nj 4J|I
    ^> j3 (C II
    -H 73
    4-i m ||
    is c ja
    M T3 0 - rt H
    4) 4) -H rH II
    014-1 -H 0 11
    0)6
    Oi H ||
    
    II
    GO I!
    a
    
    
    
    t i
    
    
    
    
    
    
              •a
              o
              o
    
              4J
              a
              CU
    A-21
    

    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    w
    (U
    u
    Furna
    1
    M
    •H
    C
    III
    _5
    rH
    m
    0
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    
    1
    1
    I
    1
    1
    1 °
    1 w
    g
    
    e
    u
    cS
    U 01
    a u
    01 rt
    EH Tl
    
    1 -rl "i7
    1 4J J-*
    1 Produc
    | (tons/
    *
    c
    3
    m
    c
    0
    •rl
    CD
    m
    •rl
    E
    W
    §
    0
    •rl
    ID
    0)
    X-rl
    0 £
    S3 0)
    17
    JC
    c--.
    •H 3
    
    «m
    (B S
    0) S
    a —
    r-l
    (1)
    3
    
    Dl
    C (8
    1 -H 4J X
    11) ~^rH IB O
    > Dl 18 T3 S
    •rl CO)
    4J -H C T3 JJ rH
    IB N C 0) 3 0
    rl TJ -rl 18 m O fc
    fl) O C > (8 X! 4J
    C ^ (8 rH ,Q 4J C
    IU -rl > (8 1 -rl O
    OlM-l rH Ol^J 3 0
    (U flf flj
    Oi O J
    00
    
    H
    c\
    ^
    
    
    1
    
    
    
    M
    o
    oft 3
    n 4J
    CQ
    ® s
    s ^
    pj Q
    O. rH
    
    M
    C\J
    • •
    0 rH
    O
    0
    "*
    
    
    
    
    
    CTl
    t-
    ^
    
    C5
    s
    
    
    
    1
    01
    > Cn
    •H c;
    AJ -H
    IB N
    rl T3 -rl
    o) o> a
    a rl -rl
    
    DlM-l 3
    0) rH
    
    00
    
    CTl
    00
    
    
    
    1 1
    
    
    
    o)
    O
    wP 3
    m 4J
    CQ
    © S
    H, 1
    
    ft M
    
    t^
    O
    . .
    in H
    p-
    co
    
    
    
    
    
    
    00
    m
    
    
    C5
    Z,
    
           A
            (8
           •H
           rH
            a
            (8
            4)
           •O
            O
            O
    A-22
    

    -------
                   References for Appendix A
    l.
    2.
    3.
    4.
    5.
    6.
    7.
    8.
    9.
    Letter and attachments from Felton, S. S., Armco Steel
    Company, L. P., to Jordan, B. C., EPA/OAQPS.  June 26, 1992.
    Response to Section 114 letter on iron and steel mills.
    
    Letter and attachments from Anderson, D. M. , Bethlehem Steel
    Corporation, to Jordan, B. C., EPA/OAQPS.  July 13, 1992.
    Response to Section 114 letter on iron and steel mills.
    
    Letter and attachments from Keeney, L., Rouge Steel Company,
    to Jordan, B. C., EPA/OAQPS.  June 30, 1992.  Response to
    Section 114 letter on iron and steel mills.
    
    Letter and attachments from Starley, J. R. , Geneva Steel, to
    Jordan, B. C., EPA/OAQPS.  July 6, 1992.  Response to
    Section 114 letter on iron and steel mills.
    
    Letter and attachments from Lee, D., Cascade Steel Rolling
    Mills, Inc., to Jordan, B. C., EPA/OAQPS.  July 6, 1992.
    Response to Section 114 letter on iron and steel mills.
    
    Letter and attachments from Harmon, M. L., LTV Steel
    Company, to Jordan, B. C., EPA/OAQPS.  July 2, 1992.
    Response to Section 114 letter on iron and steel mills.
    
    Letter and attachments from Postlethwait, W. , Nucor Steel,
    Darlington Division, to Jordan, B. C., EPA/OAQPS.  December
    2, 1992.  Response to Section 114 letter on iron and steel
    mills.
    
    Letter and attachments from Gilbert, F. C. , North American
    Manufacturing Company, to Jordan, B. C. , EPA/OAQPS.
    November 11, 1992.  Response to Section 114 letter on low-
        burners.
    Letter and attachments from Steiner, B.A., American Iron and
    Steel Institute, to Neuffer, W. J., EPA/OAQPS /BSD. May 25,
    1993.  Comments on draft ACT document.
                              A-23
    

    -------
    

    -------
    TECHNICAL REPORT DATA
    (Please read Instructions on reverse before completing)
    1. REPORT NO. 2.
    EPA-453/R-94-065
    4. TITLE AND SUBTITLE .
    Alternative Control Techniques Document~NOx Emissions from
    Iron and Steel Mills
    7. AUTHOR(S)
    Bill Neuffer
    9. PERFORMING ORGANIZATION NAME AND ADDRESS
    U.S. Environmental Protection Agency
    Office of Air Quality Planning and Standards
    Emission Standards Division (MD-13)
    Research Triangle Park, NC 27711
    12. SPONSORING AGENCY NAME AND ADDRESS
    Director
    Office of Air Quality Planning and Standards
    Office of Air and Radiation
    U.S. Environmental Protection Agency
    Research Triangle Park, NC 27711
    3. RECIPIENT'S ACCESSION NO.
    5. REPORT DATE
    September 1994
    6. PERFORMING ORGANIZATION CODE
    8. PERFORMING ORGANIZATION REPORT NO.
    10. PROGRAM ELEMENT NO.
    1 1 . CONTRACT/GRANT NO.
    13. TYPE OF REPORT AND PERIOD COVERED
    14. SPONSORING AGENCY CODE
    EPA/200/04
    15. SUPPLEMENTARY NOTES
    16. ABSTRACT
    This document describes available control techniques for controlling NOX emissions from various
    processes at iron and steel mills. These processes include reheat furnaces, annealing furnaces,
    galvanizing furnaces, soaking pits, sintering, blast furnaces, basic oxygen furnaces, and electric-arc
    furnaces. Discussions of NOX formation and uncontrolled emission levels are included. NO, control
    t«rhnini,P.s include low excess air. low NO, burners, flue gas recirculation, selective noncatalytic
       mClUUlJllCO lllV.I.U\lW 1AJW WAWWOO ttAl.j-.lWTT  A.^V^jj «_r***.**»*..w»5 *-..-•— 9—-	  i
       reduction and selective catalytic reduction.  Achievable NOX emission levels, costs and cost
       effectiveness and environmental impacts for these controls  are presented.
    17.
                                          KEY WORDS AND DOCUMENT ANALYSIS
                       DESCRIPTORS
                                                       b. IDENTIFIERS/OPEN ENDED TERMS
                                                                                             c. COSATI Field/Group
      Iron and steel mills
      NOX control techniques
      Low NOX burners
      Selective noncatalytic reduction
      Selective catalytic reduction
    Air Pollution control
    18. DISTRIBUTION STATEMENT
    
    
       Release Unlimited
    19. SECURITY CLASS (Report)
        Unclassified
    21. NO. OF PAGES
    133
                                                       20. SECURITY CLASS (Page)
                                                           Unclassified
                                           22. PRICE
                                                EPA Form 2220-1 (Rw. 4-77) PREVIOUS EDITION IS OBSOLETE
    

    -------
    

    -------