EPA-453/R-95-012
EPA-453/R-95-012
I
United States Office of Air Quality
Environmental Protection Planning and Standards
Agency __ Research Triangle Park NC 27711
Air
New Source Performance Standards,
Subpart Db - Technical Support for
Proposed Revisions to NOx Standard
ENVIRONMENTAL
PROTECTION
i AGENCY
DALLAS, TEXAS
-------
:r
EPA-453/R-95-012
New Source Performance Standards,
Subpart Db - Technical Support for
Proposed Revisions to NOx Standard
U.S. Environmental Protection Agency
Combustion Group/BSD
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1997
-------
DISCLAIMER
This document has been reviewed by the Emission Standards Division of the Office of Air Quality
Planning and Standards, EPA, and approved for publication. Mention of trade names or
commercial products is not intended to constitute endorsement or recommendation for use.
Copies of this document are available through the Library Services Office (MD-35), U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina 27711, or from the
National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161.
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1-1
2.0 CHARACTERIZATION OF INDUSTRIAL BOILERS ........ 2-1
2.1 Source Category Description 2-1
2.1.1 Source Category Definition 2-1
2.1.2 Current and Future Industry Description
2.2 Industrial Boiler Designs 2-8
2.2.1 Fundamentals of Boiler Design and
Operation 2-8
2.2.2 Furnace Configurations and Burner Types 2-10
2.2.3 Other Boiler Components 2-22
2.3 Fossil Fuel Characteristics 2-29
2.3.1 Coal 2-29
2.3.2 Oil 2-33
2.3.3 Natural Gas 2-34
2.4 Industrial Boiler NOx Formation Mechanisms . . 2-38
2.4.1 Nitrogen Oxide Formation . 2-38
2.4.2 Factors that Affect NOx Emissions . . . 2-43
2.5 Baseline Emissions 2-50
2.5.1 Baseline Emission Levels 2-50
2.5.2 Other Regulations on the Source Category 2-53
2.6 References 2-56
3.0 NITROGEN OXIDES EMISSION CONTROL TECHNIQUES .... 3-1
3.1 Combustion Controls for Coal-Fired Industrial
Boilers 3-3
3.1.1 Low NOx Burners for Pulverized
Coal-Fired Boilers 3-3
3.1.2 Low NOx Burners and Overfire Air for
Conventional Boilers 3-13
3.1.3 Air Staging for Fluidized Bed Combustion
Boilers 3-20
3.1.4 Air Staging for Spreader Stokers .... 3-20
iii
-------
TABLE OF CONTENTS (Continued)
Page
3.2 Performance of Combustion Controls on Coal-
Fired Boilers » . . . 3-21
3.3 Combustion Controls for Natural Gas- and
Oil-Fired Industrial Boilers 3-23
3.3.1 Flue Gas Recirculation 3-23
3.3.2 Low NOx Burners 3-25
3.3.3 Combinations of Combustion Controls . . 3-34
3.4 Performance of Combustion Controls for Natural
Gas- and Oil-Fired Boilers 3-34
3.5 Flue Gas Treatment Controls for Coal-, Natural
Gas- and Oil-Fired Boilers 3-37
3.5.1 Selective Noncatalytic Reduction .... 3-37
3.5.2 Selective Catalytic Reduction 3-48
3.6 Performance of Flue Gas Treatment Technologies
on Coal-, Natural Gas-, and Oil-Fired Boilers . 3-60
3.6.1 Selective Noncatalytic Reduction .... 3-60
3.6.2 Selective Catalytic Reduction 3-61
3.6.3 Recent Industrial and Utility Boiler
Permit Decisions Involving SCR .... 3-66
3.6.4 Analysis of Long-Term Continuous Emission
Monitoring Data From Coal-Fired Boilers
with SCR 3-70
3.7 References 3-80
4.0 MODEL BOILERS AND CONTROL OPTIONS 4-1
4.1 Selection of Model Boiler Parameters 4-1
4.1.1 Fuel Type 4-1
4.1.2 Furnace Type 4-3
4.1.3 Boiler Size 4-3
4.1.4 Capacity Factor 4-4
4.1.5 Baseline NOx Emission Rates 4-4
4.2 Model Boilers 4-4
IV
-------
TABLE OF CONTENTS (Continued)
Page
4.3 Control Options 4-6
i
4.3.1 Combustion Controls 4-6
4.3.2 Selective Noncatalytic Reduction .... 4-6
4.4 References 4-10
5.0 ENVIRONMENTAL AND ENERGY IMPACTS 5-1
5.1 Air Pollution Impacts 5-1
5.1.1 Primary Air Impacts 5-1
5.1.2 Secondary Air Impacts 5-3
5.2 Liquid Waste Impacts 5-7
5.3 Solid Waste Disposal Impact 5-7
5.4 Energy Impacts 5-8
5.5 References 5-14
6.0 MODEL BOILERS AND CONTROL OPTIONS COSTS 6-1
6.1 Costing Methodology 6-1
6.1.1 Total Capital Cost 6-3
6.1.2 Operating and Maintenance Costs .... 6-4
6.1.3 Calculation of Total Annual Cost, Cost
Effectiveness,, and Incremental Cost
Effectiveness 6-6
6.1.4 Other Cost Considerations 6-7
6.2 Costing Procedures 6-7
6.2.1 Combustion Controls 6-8
6.2.2 Selective Noncatalytic Reduction .... 6-9
6.2.3 Selective Catalytic Reduction 6-10
6.3 Model Boiler Cost Impacts 6-12
6.3.1 Combustion Controls 6-16
6.3.2 Selective Noncatalytic Reduction .... 6-18
6.3.3 Selective Catalytic Reduction 6-21
6.4 References 6-26
-------
TABLE OF CONTENTS (Continued)
Page
APPENDIX A ENVIRONMENTAL AND ENERGY IMPACTS FROM MODEL
INDUSTRIAL BOILERS
APPENDIX B COSTING PROCEDURES
APPENDIX C MODEL INDUSTRIAL BOILER EMISSIONS AND COST DATA
Vi
-------
LIST OF TABLES
Page
2-1 Boilers Subject to Subpart Db 2-2
2-2 Size and Fuel Distribution of Known Industrial
Boilers Subject to Subpart Db .' . . . 2-5
2-3 Size Distribution of Projected New Boiler
Population for 1996-2000 2-6
2-4 Classification of Coals by Rank 2-30
2-5 Sources and Typical Analyses of Various Ranks of
Coal 2-32
2-6 ASTM Standard Specifications for Fuel Oil 2-35
2-7 Typical Analyses and Properties of Fuel Oils .... 2-36
2-8 Characteristics of Selected Samples of Natural Gas
from United States Fields 2-37
2-9 Existing Subpart Db Nitrogen Oxide Emission Limits
for Fossil Fuel-Fired Industrial Boilers 2-51
2-10 Projected Steam Generating Capacity by Boiler Type
and Industry Segment 2-52
2-11 Baseline NOX Emissions by Boiler Type and Industry
Segment 2-54
3-1 NOX Emission Control Technologies for New Fossil
Fuel Industrial Boilers 3-2
3-2 Data Summary for Coal-Fired Industrial Boilers with
Combustion Controls 3-22
3-3 Data Summary for Oil- and Natural Gas-Fired Industrial
Boilers with Combustion Controls 3-36
3-4 Data Summary for Industrial Boilers with Selective
Noncatalytic Reduction 3-62
3-5 Data Summary for Boilers with Selective Catalytic
Reduction 3-63
3-6 List of New Industrial and Utility Boilers Using
Selective Catalytic Reduction 3-67
3-7 Summary Statistics for Carneys Point Unit No. 1 . . 3-73
vii
-------
LIST OF TABLES (Continued)
Page
3-8 Achievable NOX Emission Limits for Carneys Point
Unit No. 1 ' . . . 3-74
3-9 Summary Statistics for Carneys Point Unit No. 2 . . 3-77
3-8 Achievable NOX Emission Limits for Carneys Point
Unit No. 2 3-79
4-1 Effects of Model Boiler Parameters ... 4-2
4-2 Model Boilers 4-5
4-3 NOX Emission Control Options for Model Boilers . . . 4-7
4-4 Control Technology Performance Levels
(Annual Average) 4-8
5-1 Summary of Primary Air Impacts for Model Industrial
Boilers 5-2
5-2 Summary of Secondary Air Impacts for Model Industrial
Boilers 5-5
5-3 Summary of Energy Impacts for Model Industrial
Boilers 5-9
6-1 Capital and Operating Cost Components 6-2
6-2 Fixed and Variable O&M Unit Costs 6-5
6-3 Summary of Annualized Costs for Model Boilers . . . 6-13
6-4 Model Boiler Incremental Cost Effectiveness
Ranges 6-14
viii
-------
LIST OF FIGURES
Page
2-1 Single, Wall-Fired Boiler 2-11
2-2 Circular-type Burner for Pulverized Coal, Oil,^
or Gas . . 2-13
2-3 Firing Pattern in a Tangentially-Fired Boiler . . . 2-14
2-4 Burner Assembly of a Tangentially-Fired Boiler . . . 2-16
2-5 Spreader Type Stoker-Fired Boiler - Continuous
Ash Discharge Grate 2-18
2-6 Simplified Bubbling Bed AFBC Process Flow Diagram . 2-20
2-7 Watertube Design Configurations for Packaged
Boilers 2-23
2-8 Variation of Flame Temperature with Equivalence
Ratio 2-40
2-9 Conversion of Fuel-Bound Nitrogen-to-Nitrogen
Oxide in Pulverized-Coal Combustion 2-44
2-10 Fuel NOX to Fuel Nitrogen Content-Pulverized Coal,
Premixed 2-45
2-11 Comparative Physical Sizes of Boilers Firing
Different Fuels 2-49
3-la Controlled Flow/Split Flame™ Low NOX Burner .... 3-5
3-lb Internal Fuel Staged™ Low NOX Burner 3-5
3-2 Dual Register-Axial Control Flow (DRB XCL)™ Low
NOX Burner 3-7
3-3 RO-II Low NOX Coal Burner 3-8
3-4 Controlled Combustion Venturi™ Low NOX Burner . . . 3-9
3-5a Typical Fuel and Air Compartment Arrangement for
One Corner of a Tangentially-Fired Boiler 3-11
3-5b Plan View of Fuel and Air Streams in a Typical
Tangentially-Fired Boiler 3-11
3-6a Low NOX Concentric Firing System Fuel and Air
Compartment Arrangement 3-12
ix
-------
LIST OF FIGURES (Continued)
page
3-6b Plan View of Low NOX Concentric Firing System . . . 3-12
i
3-7 Low NOX Pollution Minimum burner 3-14
3-8a Typical Opposed Wall-Fired Boiler 3-16
3-8b Opposed Wall-Fired Boiler with Overfire Air .... 3-16
3-9a Conventional Overfire Air on an Opposed Wall-Fired
Boiler 3-17
3-9b Advanced Overfire Air on an Opposed Wall-Fired
Boiler 3-17
3-10 Low NOX Concentric Firing System 3-19
3-11 Flue Gas Recirculation System 3-24
TM
3-12 ROPM Burner for Natural Gas- and Oil-Fired
Boilers 3-26
3-13 Internal Staged Combustion™ Low NOX Burner .... 3-28
3-14 XCL™ Natural Gas- and Oil-Fired Low NOX Burner . . 3-29
3-15 Low NOX Swirl Tertiary Separation™ Low NOX Burner . 3-30
3-16 Rapid Mix Burner™ for Natural Gas- and Oil-Fired
Boilers 3-32
3-17 Pollution Minimum™ Burner for Natural Gas- and
Oil-Fired Boilers 3-33
3-18 Ammonia-Based Selective Noncatalytic Reduction . . . 3-38
3-19 Urea-Based Selective Noncatalytic Reduction .... 3-40
3-20 High-Energy Selective Noncatalytic Reduction
Process 3-42
3-21 General Effects of Temperature on NOX Removal . . . 3-44
3-22 General Effect of NH3:NOX Mole Ratio on NOX
Removal 3-46
J
-------
LIST OF FIGURES (Continued)
Page
3-23 Ammonia Salt Formation as a Function of Temperature
and NH3 and 803 Concentration .* . . . 3-47
3-24 Relative Effect of Temperature on NOX Reduction . . 3-49
3-25 Possible Configurations for Selective
Catalytic Reduction 3-50
3-26 Typical Configuration for a Catalyst Reactor .... 3-52
3-27 Example of Optimum Temperature Range for Different
Types of Catalysts 3-54
3-28 Configuration of Parallel Flow Catalyst 3-56
3-29 Effect of Temperature on Conversion of S02 to 803 . 3-57
3-30 Time Plot of Hourly NOX Emissions for July 1 to
September 30, 1995 (Carneys Point, Unit No. 1) . . . 3-72
3-31 Time Plot of Hourly NOX Emissions for July 1 to
September 30, 1995 (Carneys Point, Unit No. 2) ... 3-76
6-1 CC Incremental Cost Effectiveness vs. Boiler Size . 6-17
6-2 CC Incremental Cost Effectiveness vs. Capacity
Factor 6-19
6-3 CC + SNCR Incremental Cost Effectiveness vs. Boiler
Size (MMBtu/hr) • 6-20
6-4 CC + SNCR Incremental Cost Effectiveness vs.
Capacity Factor 6-22
6-5 CC + SCR Incremental Cost Effectiveness vs.
Boiler Size 6-23
6-6 CC + SCR Incremental Cost Effectiveness vs.
Capacity Factor 6-25
XI
-------
1.0 INTRODUCTION
This document supports regulatory development acttion
taken by the U. S. Environmental Protection Agency (EPA) under
section 407 of title IV of the Clean Air Act (hereafter
referred to as the Act) (42 U.S.C. 7411), as amended in 1990.
Section 407 of the Act presents the nitrogen oxides (NOX)
emission reduction program. It mandates that the EPA revise
existing new source performance standards (NSPS), developed
under section 111 of the Act, for NOX emissions from fossil-
fuel fired steam generating units, including both electric
utility and nonutility units. These revised standards
"...shall reflect improvements in methods for the reduction of
emissions of oxides of nitrogen." The group of fossil fuel-
fired steam generating units covered in this document are
those currently subject to 40 Code of Federal Regulations
(CFR) 60, subpart Db.
Title IV (Acid Deposition Control) of the Act was added
"...to reduce the adverse effects of acid deposition through
reductions in annual emissions of sulfur dioxide..., and, in
combination with other provisions,... of NOX emissions..."
Standards of performance for stationary sources,
developed under section 111 of the Act, are required to
reflect "... the degree of emission reduction achievable which
(taking into account the cost of achieving such an emission
reduction, and any nonair quality health and environmental
impacts and energy requirements) the Administrator determines
has been adequately demonstrated for that category of
sources." The standards developed under section lll(b) apply
only to new stationary sources that have been constructed or
modified after regulations are proposed by publication in the
Federal Register.
1-1
-------
The chapters of this technical support document (TSD)
present the technical information on which the regulatory
action under sections 407 and 111(b) is based. However, the
TSD is not the exclusive source of technical background
information. The docket, a public file maintained in
Washington, B.C., is another source of background information.
Types of information that may be included in the dock'et are
information provided by the affected industry, vendors, or
trade associatipns and information obtained through meetings,
phone contacts, or site visits.
This TSD presents a characterization of the affected
industry and information concerning the performance, cost, and
environmental impacts of the control techniques applicable to
fossil-fuel fired steam generating units (boilers). The model
boilers developed to evaluate the impacts of different control
options, and the incremental environmental, energy, and cost
impacts of the control options applied to the model boilers
are also presented.
Chapter 2.0, Characterization of Industrial Boilers,
describes the affected industry. The regulatory definition of
the industry and a description of the current and future
industrial boiler population are provided. Chapter 2.0 also
describes various types of industrial boilers and their NOX
emission characteristics, as well as design and operational
factors affecting NOX emissions. Finally, the projected
baseline emissions for the affected industry, in the absence
of additional controls resulting from revision to the current
standard, are estimated.
Chapter 3.0 presents NOX emissions control techniques
applicable to industrial boilers and discusses the emission
control performance of these techniques when applied to
different types of boilers firing different fuels. The
control techniques presented include combustion controls (CO
and flue gas treatment techniques. Available NOX emissions
test data from representative coal-, oil-, and gas-fired
industrial boilers equipped with CC and flue gas treatment
1-2
-------
controls are presented to reflect the performance of these
control techniques.
The model boilers developed to evaluate the impacts of
controlling NOX emissions from the affected industry are
presented in chapter 4.0. The boiler parameters considered in
developing model boilers and their impact on NOX emissions and
control technique performance and cost are discussed*
Chapter 4.0 also identifies the control options and the
associated performance levels.
Chapter 5.0 presents the incremental environmental and
energy impacts associated with the application of the control
options to the model boilers. Environmental impacts include
primary (NOX) and secondary air pollution impacts. Secondary
air pollution impacts include emissions of ammonia (NH3),
nitrous oxide (N20), and carbon monoxide (CO). Potential
liquid and solid waste impacts are also discussed. Energy
impacts include potential decreases in boiler efficiency,
which result in increases in fuel use, and potential increases
in electricity usage for some control options.
The estimated cost and cost effectiveness associated with
the application of the control options to the model boilers
are presented in chapter 6.0. The costing methodology and the
specific cost algorithms used are presented. Total capital
cost, operating and maintenance costs, average cost
effectiveness, and incremental cost effectiveness were
estimated for the application of each control option to the
model boilers. The effect of model boiler parameters and
different control techniques on cost and cost effectiveness is
discussed.
This document also contains three appendices. Appendix A
contains tables showing the primary air impacts (NOX emission
reductions), secondary air impacts (emissions of NH3, N2O, and
CO), and energy impacts resulting from the application of the
NOX control options to each model boiler. Appendix B
describes the cost procedures used for estimating the costs of
the NOX control options considered. Finally, appendix C
1-3
-------
presents the costs impacts associated with each control option
for the model boilers.
1-4
-------
2.0 CHARACTERIZATION OF INDUSTRIAL BOILERS
This chapter presents an overview and characterization of
industrial boilers. It is divided into five main sections:
source category description, industrial boiler designs, fossil
fuel characteristics, nitrogen oxides (NOX) formation
mechanisms, and baseline NOX emissions.
2.1 SOURCE CATEGORY DESCRIPTION
2.1.1 Source Category Definition
The revision of subpart Db of the New ISource Performance
Standards (NSPS) for NOX emissions applies to any fossil fuel-
fired steam generating unit that is capable of combusting more
than 100 million British Thermal Units per hour (MMBtu/hr)
heat input, for which construction or modification commenced
after the date of proposal and that is not subject to
subpart Da (electric utility steam generating units).
The revision of the standard also applies to natural gas-
or oil-fired duct burners used in steam generating units that
are components of combined cycle gas turbine systems if the
duct burners are not subject to subpart Da.
2.1.2 Current and Future Industry Description
Industrial boilers are used in a wide range of industry
sectors. The leading of these sectors are: petroleum, non-
manufacturing, boiler rental, and chemical.
2.1.2.1 Current Industry Description. Data available,
as of 1995, on sources subject to subpart Db are listed in
table 2-1. Information on industrial boilers subject to
subpart Db was obtained from three sources: a best available
control technology (BACT)/lowest achievable emission rate
(LAER) database search; copies of permits obtained from
various EPA offices; and telephone contacts with EPA regions.
State agencies, and boiler vendors. Based on the information
2-1
-------
TABLE 2-1. BOILERS SUBJECT TO SUBPART Db
Boiler Size
Company Name State (MMBtu/hr)
Coal-Fired
Goodrich Chemical Facility-Henry3
University of North Carolina3
Fort Druma
University of Northern Iowaa
Archbald Power Corp.a
T B Simmon-MSUa
Milwaukee County Power Plant3
Archer Daniels Midland3
Oil-Fired
Old Dominion Electric Cooperative
Sithe/Independence Power Partner
Okeelanda Corporation
South Carolina Electric & Gas Co.
Gas-Fired
Seminole Kraftb
Champion Paper
University of Cincinnati
Miami University
Minnesota Corn Processors Inc.
Gay lord Container Corp.*3
Willamette Industries Inc.
Corn Products
Bermuda Hundred Energy LTD
Partners"
Exxon Chemical Americas0
Saranac Energy Company
Clark County Ind. Council
Selkirk Cogen. Partners, L.P.C
American Crystal Sugar Co.
The Upjohn Company
Boilers
IL
NC
NY
IA
PA
MI
Wl
IA
Boilers
VA
NY 250
FL
SC
Boilers
FL
OH
OH
OH
MN
LA
LA
CA
VA
LA
NY
AR
NY
MN
MI
162.5
304
208
139.5
240
460
157
551.5
214
(x3)
205
190
174.7
(x3)
197
200
132
178.7
235
335
178
250
458
249
154.7
206
200.1
109.6
Control
Technology (d)
CC
CC
CC '
CC
CC
SNCR
SNCR
SNCR
Operation Hour
Restriction
CC
CC
CC
Good
Combustion
Practices
LNB
LNB
LNB
LNB
Multi-stage
LNB
LNB
LNB, Staged
Combustion
LNB
LNB
LNB+FGR
LNB
LNB
LNB+FGR
LNB+FGR
2-2
-------
TABLE 2-1. BOILERS SUBJECT TO SUBPART Db (Continued)
Company Name
James River Corp.
Boise Cascade Corporation
BAF Energy-Amer I Cogen. Project
SMUD/Campbell Soup
Helm Concentrates
Selkirk Cogen. Partners, L.P.C
Darling Delaware Co. Inc.
Northern Consolidated Power
Lake Cogen LTD0
State
MI
AL
CA
CA
CA
NY
CA
PA
FL
Boiler Size
(MMBtu/hr)
226.7
343.3
150 (x2)
100 (x4)
130
206 (x2)
110
100.4
150.5
Control
Technology (d)
LNB
LNB
LNB+FGR
LNB+FGR
LNB+FGR
SCR, LNB
SCR
Unknown
Unknown
a Circulating Fluidized Bed Combustion Boiler
b Packaged Boiler
c Duct Burner
d cc=Combustion Controls
SNCR=Selective Noncatalytic Reduction
LNB=Low NOX Burners
FGR=Flue Gas Recirculation
SCR=Selective Catalytic Reduction
2-3
-------
obtained, 8 coal-fired fluidized bed combustors (FBC), 6 oil-
fired, and 31 natural gas-fired industrial boilers subject to
the current subpart Db standard were identified. Table 2-2
presents these boilers and share of generating capacity by
size range and fuel type.
All of the known coal-fired boilers are FBC boilers. Of
these, five use only combustion controls (CC) and the*
remaining three are equipped with selective noncatalytic
reduction (SNCR) for NOX control. One of the oil-fired
boilers controls NOX emissions by restricting the operating
hours, the remaining three oil-fired boilers use CC. Of the
four oil-fired boilers, two are known to fire distillate oil.
The type of NOX controls for gas-fired boilers range from the
use of good combustion practices to using selective catalytic
reduction (SCR); however, none of them use SNCR.
2.1.2.2 Future Industry Description. New industrial
boiler capacity (boilers with capacities & 100 MMBtu/hr heat
input) for the 5-year period following promulgation of the
revised NSPS was estimated based on information obtained from
the American Boiler Manufacturers Association (ABMA) . 1>2>3 The
year 2000 is used to represent the fifth year of the revised
standard. The number of boilers, capacity breakdown, and
boiler type are presented in table 2-3. Only 5 percent of the
projected boilers fire coal; 18 percent fire residual oil;
8 percent fire distillate oil; and 69 percent fire natural
gas. A comparison of the size distribution indicates that
about 61 percent of the new boiler population is projected to
be between 100 and 200 MMBtu/hr in heat input and the majority
of these are natural gas-fired, packaged boilers. Boilers
ranging from 200 to 300 MMBtu/hr make up 18 percent of the
projected population; 300 to 500 MMBtu/hr boilers make up
12 percent; and boilers with heat inputs greater than
500 MMBtu/hr make up less than 10 percent of the projected
boiler population.
2-4
-------
J5|»
Gj
Q
E5 O
t*j p
fcEn
S ft
O «
H ID
gco
m o
H E-I
Di
E-* EH
CO U
H W
p b
§
W CO
Cn CO
P W
i
M ^
c n"u
a) 5 rt
0) O
cu
4J r^
-H ^.
u n
nj -i-J
rt S
us
^j
tw 0) C
0 r-l 0
-H -H
AJ O 4J
C XI rt
Q) rH
U rH ^
r-l OS ft
(1) 4J O
Oi O ft
•^
14-1
o m
S-l 0)
fl -H
E o
S
en >-i
rt\
rt 3
m
N g
CO ~^
rH
OJ
fi
inooH CNCNO Hvoin
H H n CN
cr>(NCN ocrio inc^oo
inmH cnvD usinin
^*r^O iHiH 00^<4*
•> «. CN CN
H H
rH _ ^(OJro
r*1 ^^ CN ^_j cj ^ p. CN
nrOCN rHinO — - • H
0 O O 0 O O
o in o in o in
CN m CN m CN m
VI VI VI VI VI M
*O t3 *O *O *O 'O
G CH C EH C C
nj rt (d (d rt (tf
ooo ooo ooo
ooin ooin ooin
HCNrn HCNrn HCNm
Al Al Al AIAIAI MAIM
rH
rt! rH CD
O -H rrj
CJ O O
2-5
-------
o
o
o
vo
o\
H
«
o
0
H
5
I— ^
tr
o
cu
tf
u
H
o
CQ
iJ
H
a;
EH
w
D
Q
25
H
*j
|
Q
U
B
U
g
cu
fo
o
2
0
H
EH
£3
CQ
H
a
EH
en
H
o
H
N
H
W
n
1
CM
W
CQ
EH
O >
_. -P
it ^^ ,j
cJS o
41 n «
s°s
01 U
4-> £
O 3
Id 4J
04 m
31
0
*•« 5
C 4J «
a
^
^
*J
g
^5 ^D Ct fO 'I1 C? '^ ^3 ^5 C3 ^3 ^3 ^3 ^^ *^*
ooorHo o^rooo ooovoo
r~ oo «H in n
n CM CM ao CM
CO ^ rH CM ^
ooom*r o in o o o o o o •<* on CN ro in en
IIIIA IIIIA IIIIA
OOOO OOOO OOOO
OOOO OOOO OOOO
0
o w £
04 W »
,
id
s
in
H
1-1
in
en
m
t-i
CM
CM
•J
•^
H
rH
i^
s
OrHtHcnO VO r-t in O O
oooo^ooo ooooo^oo
in oo vo i-t !»• 1-1
n o <* t-- in r~
H t-t «n in ^
o in ro in o en CM CM o o
in tH
ooooo ooooo
ooooo ooooo
otmmcncft cMmincncri
IIIIA IIIIA
oooo oooo
oooo oooo
ficMrnin 1-1 CM n in
•a
Q)
4J -O
U 0)
0) CF>
U«f
IQ
?j^
u
•0 «
^ a.
o
h
i-H
O
id
•§
•H
n
0)
o
rn
en
VO
t-i
00
VO
VO
4:
o
H
•"
C
r-
J
T
a
0
PS
2-6
-------
o
o
o
CM
VO
s
£
S5
O
H
j|j
2
c^
cu
2
1
H
0
CQ
J
i4j
H
B
CO
H W
&
Ely
55 g
§1
S
H
1
,.
O
O
B
CQ
M
e
H
Q
H
tS3
H
CO
•
cn
1
S
n
g
H
£.3 o
8 R o
0) U
o.
bx |3
4J.C
•rl ^s.
U 9
40 4J
(X03
«|
*° 0
JJ *•! **
o> o;~j
O 4J Q
14 A
m
N+ft
«Cj
w —
$
0
r-l
o
(0
H
i
O O CM O O
o «n r» oo o r» o «N o o
cri t vo «o o «*
vo ^ r- vo CM H
o •-! o ft o in o 1-1 o o
oni-ioo t- r-i m o o
rt
ooooo ooooo
ooooo ooooo
CM
•H
Q
00
!«,
•-4
o>
00
o
Total
r-l
O
O
4i
f^
ft
4*
n
-i
a
o r«- 01 r-i CM in m r-i o o
iH CM
O ft t* O i-l «* O 00 O O
en ft o *t oo o oo
VD CM Oi O 00 i-l CM
CM in
H CM
o oo vo o fn in vo 1-1 o o
*
Or* CM r-l 01 CM i-4« O O
ro CM 1-1 r» CM
1-4
ooooo ooooo
ooooo ooooo
CM n in en at CM n m en a\
i i i i A i i i i A
OOOO OOOO
OOOO OOOO
r-i CN n in r-tcMroin
•a
43 TJ
1 s
i •JJ
2 «
3 *
fa
r-l
u n
9 «
•P C9
n
z
0
VO
•H
CM
CM
VO
0\
VO
n
vo
CM
^J
Q
a
O
H
^
N
* «O 00 CM O O
ro iH rt i-l CM O
i
«n CM n ft oo
vo ro o vo in i-i
CM oo o> n in o>
* VO CO CM O CM
n 1-1 H H CM o
•H 00 CM « in O
VO iH i-l O
00 00 Is* 00 O i-l
(N VO t r-l CM 00
CM l*>
ooooo
ooooo
CM n in a\ o
1 1 1 1 A
OOOO
OOOO
ft CM n in
0
N
co a
r-l
C M
£
n ft
40 ft
a a
B Ml
rH >
•d 6
H
C
o
43
rH M
8J^ tt
o n
4j>
• •
M h N
•^ « -H
V4-OTJ
0 «•*
> 0 9
» tj •
iH M r«4
2 a fa
(0 *
1 1 O
CM U (M
2-7
-------
When the capacity share is considered, 15 percent of the
projected boiler capacity is made up of coal-fired boilers;
17 percent by residual oil-fired boilers; 8 percent by
distillate oil-fired boilers; and 60 percent by natural gas-
fired boilers. Of the total capacity of projected boilers,
boilers with heat inputs ranging from 100 to 200 MMBtu/hr make
up 34 percent. Boilers with heat inputs from 200 to1
300 MMBtu/hr account for 16 percent; from 300 to 500 MMBtu/hr
account for 18 percent; from 500 to 900 MMBtu/hr account for
12 percent; and greater than 900 MMBtu/hr account for
20 percent.
2.2 INDUSTRIAL BOILER DESIGNS
The basic purpose of an industrial boiler is to convert
the chemical energy in a fuel into thermal energy as steam for
process heat. To achieve this objective, two fundamental
processes are necessary: combustion of the fuel by mixing
with oxygen, and the transfer of the thermal energy from the
resulting combustion gases to working fluids such as hot water
and steam. The physics and chemistry of combustion, and how
they relate to NOX formation, are discussed in section 2.4.
2.2.1 Fundamentals of Boiler Design and Operation
An industrial boiler consists of several major
subassemblies, including the fuel preparation system, air
supply system, combustion system, furnace, and convective heat
transfer system. The fuel preparation system, air supply, and
combustion system are primarily involved in converting fuel
into thermal energy in the form of hot combustion gases. The
last two subassemblies are involved in the transfer of the
thermal energy in the combustion gases to steam. In
cogeneration facilities, the steam is superheated and expanded
in a steam turbine to produce electricity. The steam leaving
the turbine is used as process heat.
The NOX formation potential of a boiler is determined by
the design and operation of the fuel preparation equipment,
air supply, combustion, and furnace subassemblies. The
potential for reducing NOX after it forms is primarily
2-8
-------
determined "by the design of the furnace and convective heat
transfer system and, in some cases, by the operation of the
air supply system.
Three key thermal processes occur in the furnace and
convective sections of a boiler. First, thermal energy is
released during controlled mixing and combustion of fuel and
oxygen. Oxygen is typically supplied in two, and sorrtetimes
three, separate air streams. Primary air is mixed with the
fuel before introducing the fuel into the burners. In a coal-
fired boiler, primary air is also used to dry and transport
the coal from the fuel preparation system (e.g., the
pulverizers) to the burners. Secondary air is supplied
through a windbox surrounding the burners, and is mixed with
the fuel after the fuel is injected into the burner zone.
Finally, some boilers are equipped with tertiary air
(sometimes called "overfire air"), which is used to complete
combustion in boilers having staged combustion burners. A
detailed discussion of the importance of each of these air
supplies as it relates to NOX formation and control is
presented in section 2.4.
Large industrial boilers include multiple, closely-spaced
tubes filled with high-pressure water. Water flows into these
"water tubes" at the bottom of the furnace and rises to the
steam drum located at the top of the boiler. In the second
key thermal process, a portion of the thermal energy formed by
combustion is absorbed as radiant energy by the furnace walls.
During the transit of water through the water tubes, the water
absorbs this radiant energy from the furnace. Although the
temperature of the water within these tubes can exceed 370 °C
(710 °F) at the furnace exit, the pressure within the tubes is
sufficient to maintain the water as a liquid rather than
gaseous steam.
At the exit to the furnace, typical gas temperatures are
1,000 to 1,300 °C (1,600 to 2,400 °F), depending on fuel type
and boiler design. At this point, the gases enter the
convective pass of the boiler, and the balance of the energy
2-9
-------
retained by the high-temperature gases is absorbed as
convective energy by the convective heat transfer system
(superheater, reheater, economizer, and air preheater). In
the convective pass of boilers with air preheaters, the
combustion gases are typically cooled to 135 to 180 °C (275 to
350 °F).
The fraction of the total energy that is emitted as
radiant energy depends on the type of fuel fired and the
temperature within the flame zone of the burner. Because of
its ash content, coal emits a greater amount of radiant energy
than a flame produced from burning gas, which is relatively
transparent. As a result, coal-fired boilers are designed to
recover a significant amount of the total thermal energy
formed by combustion through radiant heat transfer to the
furnace walls, while gas-fired boilers are designed to recover
most of the total thermal energy through convection.
The design and operating conditions within the convective
pass of the boiler are important in assessing NOX control
options because two options--SNCR and SCR--are designed to
operate at temperatures found in and following the convective
pass.
2.2.2 Furnace Configurations and Burner Types
There are a number of different furnace configurations
used in industrial boilers. For purposes of presentation,
these configurations have been divided into five groups:
wall-fired, tangentially-fired, stoker, FBC, and packaged
boilers.
2.2.2.1 Wall-Fired. Wall-fired boilers are
characterized by multiple individual burners located on a
single wall or on opposing walls of the furnace. Depending on
the design and location of the burners, wall-fired boilers can
be subcategorized as single-wall or opposed-wall.
2.2.2.1.1 Single-wall. The single-wall design consists
of several rows of circular-type burners mounted on either the
front or rear wall of the furnace. Figure 2-1 shows the
burner arrangement of a typical single, wall-fired boiler.4
2-10
-------
Burner B
Burner A
AirA-
AirB-
AirC-
AirD-
FuelA
FuelB
FueIC
FuelD
Burner D
Burner C
Figure 2-1. Single, wall-fired boiler,
2-11
-------
In circular burners, the fuel and primary air are
introduced into the burner through a central nozzle that
imparts the turbulence needed to produce short, compact
flames. Adjustable inlet vanes located between the windbox
and burner impart a rotation to the preheated secondary air
'from the windbox. The degree of air swirl, in conjunction
with the flow-shaping contour of the burner throat, '
establishes a recirculation pattern extending into the
furnace. After the fuel is ignited, this recirculation of hot
combustion gases back towards the burner nozzle provides
thermal energy needed for stable combustion.
Circular burners are used for firing coal, oil, or
natural gas, with some designs featuring multi-fuel
capability. A circular burner for firing pulverized coal,
oil, or natural gas is shown in figure 2-2.5 To burn fuel oil
efficiently, circular burners are equipped with atomizers.
Atomization provides high oil surface area for contact with
combustion air. The oil can be atomized by the fuel pressure
or by a compressed gas, usually steam or air.
In natural gas-fired burners, the fuel can be supplied
through a perforated ring, a centrally located nozzle, or
radial spuds that consist of a gas pipe with multiple holes at
the end.
2.2.2.1.2 Opposed-wall. Opposed, wall-fired boilers are
similar in design to single, wall-fired boilers, differing
only in that two furnace walls are equipped with burners and
the furnace is deeper. The opposed-wall design consists of
several rows of circular-type burners mounted on both the
front and rear walls of the furnace.
2.2.2.2 Tangent ially- Fired. The use of tangentially-
fired industrial boilers is uncommon; however, because there
are some exceptions, they are included here. The
tangentially-fired boiler is based on the concept of a single
flame zone within the furnace. As shown in figure 2-3, the
fuel-air mixture in a tangentially-fired boiler projects from
2-12
-------
Gos-fired lighter
Oil
Cool
Figure 2-2. Circular-type burner for pulverized
coal, oil, or gas.
2-13
-------
•o
0)
M
i
c
(0
•M
(0
c
JJ
(0
E1
•H
M
^4
fe
I
CM
0
I
•H
fa
2-14
-------
the four corners of the furnace along a line tangential to an
imaginary cylinder located along the furnace centerline.6 As
shown in figure 2-4, the burners in this furnace design are in
a stacked assembly that includes the windbox, primary fuel
supply nozzles, and secondary air supply nozzles.6
As fuel and air are fed to the burners of a tangentially-
fired boiler and the fuel is combusted, a rotating "fireball"
is formed. The turbulence and air-fuel mixing that take place
during the initial stages of combustion in a tangentially-
fired burner are low compared to other types of boilers.
However, as the flames impinge upon each other in the center
of the furnace during the intermediate stages of combustion,
there is sufficient turbulence for effective mixing and carbon
burnout.7 Primarily because of their tangential firing
pattern, uncontrolled tangentially-fired boilers generally
emit relatively lower NOX than other uncontrolled boiler
designs.
The entire windbox, including both the fuel and air
nozzles, tilts uniformly. This allows the fireball to be
moved up and down within the furnace in order to control the
furnace exit-gas temperature and provide steam temperature
control during variations in load. In addition, the tilts on
coal-fired units automatically compensate for the decreases in
furnace-wall heat absorption due to ash deposits. As the
surfaces of the furnace accumulate ash, the heat absorbed from
the combustion products decreases. The burners are then
tilted upwards to increase the temperature of the flue gas
entering the convective pass of the boiler. Furnace wall
fouling will cause the heat to rise in the furnace normally
resulting in downward tilts, while fouling in the convective
sections can cause the reverse. Also, when convective tube
fouling becomes severe, soot blowers are used to remove the
coating on the tubes. The sudden increase in heat absorption
by the clean tubes necessitates tilting the burners down to
2-15
-------
2-16
-------
their original position. As the fouling of the tubes resumes,
the tilting cycle repeats itself.
Tangentially-fired boilers commonly burn coal. However,
oil or gas are also burned by inserting additional fuel
injectors in the secondary air components adjacent to the
pulverized-coal nozzles as shown in figure 2-4.
2.2.2.3 Stoker-Fired. There are several types'of
stoker-fired boilers used by industries. The most common
stoker type is the spreader stoker. Spreader stokers burn
finely crushed coal particles in suspension, and larger fuel
particles in a fuel bed on a grate as shown in figure 2-5.8
Relatively low combustion gas velocities through the boiler
are necessary to prevent fly ash erosion, which results from
high flue-gas ash loadings.
Spreader stokers use continuous-ash-discharge traveling
grates, intermittent-cleaning dump grates, or reciprocating
continuous-cleaning grates. They are capable of burning all
types of bituminous and lignitic coals. Heat input capacities
usually range from 5 to 550 MMBtu/hr due to material handling
limitations.
2.2.2.4 Fluidized Bed Combustion. Fluidized bed
combustion is an integrated technology for reducing sulfur
dioxide (SC>2) and NOX emissions during the combustion of coal.
Fluidized bed combustion boilers inherently emit low levels of
NOX due to the relatively low combustion temperatures and are
an option for new boilers or repowering existing boilers. In
a typical FBC boiler, crushed coal in combination with inert
material (sand, silica, alumina, or ash) and/or a sorbent
(limestone) are maintained in a highly turbulent suspended
state by the upward flow of primary air from the windbox
located directly below the combustion floor. This fluidized
state promotes uniform and efficient combustion at furnace
temperatures between 815 and 870 °C (1,500 and 1,600 °F)
compared to 1,370 and 1,590 °C (2,500 and 2,900 °F) for
conventional coal-fired boilers. Fluidized bed combustion
2-17
-------
0)
-p
R)
-H
2-18
-------
boilers are capable of burning a wide range of fuels and sizes
range between 8 and 1,070 MMBtu/hr.
Fluidized bed combustion technologies based on operation
at atmospheric and pressurized conditions have been developed.
The atmospheric FBC (AFBC) system is similar to a conventional
industrial boiler in that the furnace operates at or near
atmospheric pressure.9 There are two major categories of AFBC
boilers: bubbling bed, shown in figure 2-6, and circulating
bed. Pressurized FBC (PFBC) operates at pressures greater
than atmospheric pressure and recovers energy through both
heat transfer to a working fluid and the use of the
pressurized gas to power a gas turbine. No PFBC boilers are
currently in operation for industrial applications and it is
unlikely that such systems will be used in the near future due
to the developmental status of this technology.10
In the bubbling bed AFBC design, coal and limestone are
continuously fed into the boiler from over or under the bed.
The bed materials are suspended by the combustion air blowing
upwards through the fluidizing air nozzles. Some bed material
is entrained in the upflowing flue gas and escapes the
combustion chamber. Approximately 80 to 90 percent of this
fly ash is collected in the cyclone and is then either
discarded or reinjected into the bed. Reinjection of ash
increases combustion efficiency and limestone utilization. In
general, combustion efficiency increases with longer residence
times and greater ash recycle rates. Fly ash not collected in
the cyclone is removed from the flue gas by an electrostatic
precipitator (ESP) or fabric filter.
The circulating fluidized bed design is a more recent
development in AFBC technology. The two major differences
between circulating and bubbling AFBC's are the size of the
limestone particles fed to the system and the velocity of the
fluidizing air stream. Limestone feed to a bubbling bed is
generally less than 0.1 inches in size, whereas circulating
beds use much finer limestone particles, generally less than
0.01 inches. The bubbling bed also incorporates relatively
2-19
-------
Convection
4
Pass
Coal Llmsstons
re-
o-
Frssboard
Splash Zons
Bsd
Transport Air
Phildlzlng Air
Flu* Gat
^Cyclon*
FUcycl*
~ Distributor
Plats
Plenum
Foresd Draft Air
[Comprsssor]
Wasts
Wasts
Figure 2-6. Simplified bubbling bed AFBC process
flow diagram.
2-20
-------
low air velocities through the unit, ranging from 4 to 12 feet
per second (ft/sec).9 This creates a relatively stable
fluidized bed of solid particles with a well-defined upper
surface. Circulating beds employ velocities as high as
30 ft/sec.11 As a result, the bed materials are entrained in
the transport air/combustion gas stream. These solids are
then separated from the combustion gases by a cyclone* or other
separating device and circulated back into the combustion
region, along with fresh coal and limestone. A portion of the
collected solids are continuously removed from the system to
maintain material balances. Circulating beds are
characterized by very high recirculated solids flow rates, up
to three orders of magnitude higher than the combined
coal/limestone feed rate.9
Circulating AFBC's predominate new FBC installations, in
part due to their improved performance, enhanced fuel
flexibility, and ability to operate at higher heat input
rates.12 Bubbling beds recover heat from the combustion gases
by use of in-bed and convective water tubes, whereas
circulating beds recover essentially all of their heat using
convective tubes located downstream of the cyclone. Some
specific advantages of circulating bed over bubbling bed
designs include:
• Higher combustion efficiency, exceeding 90 percent;
• Greater limestone utilization, due to high recycle
of unreacted sorbent and small limestone feed size
(greater than 85 percent SC>2 removal efficiency is
projected with a Ca/S ratio of about 1.5, with the
potential for greater than 95 percent SC-2 removal
efficiency);
• Potentially fewer corrosion and erosion problems,
compared to bubbling bed designs with in-bed heat
transfer surfaces;
• Less dependence on limestone type, since reactivity
is improved with the fine particle sizes; and
2-21
-------
• Reduced solid waste generation rates, because of
lower limestone requirements.
2.2.2.5 Packaged Boilers. Industrial boilers that are
shop-assembled and shipped to the site as a complete unit are
called packaged boilers. These are usually oil- or gas-fired
and are of the watertube design. The major type of watertube
design used is the horizontal bent tube, classified by the
number of drums, headers, and tube configuration, with the
latter being the most distinguishing factor. Figure 2-7 shows
the three most common tube configurations used in packaged
units. The "A" type has two small lower drums, or headers,
and a large upper drum for steam and water separation. Most
steam production occurs in the center furnace wall tubes
entering the drum. The "D" type, the most flexible design and
the most widespread, has two drums and a large-volume
combustion chamber that is easy to outfit with a superheater
or economizer. The "0" configuration's symmetry exposes the
least amount of tube surface to radiant heat.13
2.2.3 Other Boiler Components
This section discuses additional boiler components
including pulverizers (fuel preparation system), air supply
system, and superheaters/reheaters, economizers, and air
heaters (heat transfer system).
2.2.3.1 Pulverizers. The only fuel preparation system
discussed here is the pulverizer which is used in industrial
boilers that burn pulverized coal. Pulverized coal is favored
over other forms of coal because pulverized coal mixes more
intimately with the combustion air and burns more rapidly.
Pulverized coal also burns efficiently at lower excess air
levels and is more easily lit and controlled.14
To achieve the particle size reduction required for
proper combustion in pulverized coal-fired boilers, machines
known as pulverizers (also referred to as "mills") are used to
grind the fuel. Coal pulverizers are classified according to
their operating speed. Low-speed pulverizers consist of a
rotating drum containing tumbling steel balls. This
2-22
-------
A-type
D-type
o-type
Figure 2-7. Watertube design configurations for packaged
boilers.
2-23
-------
pulverizer type can be used with all types of coal, but is
particularly useful for very abrasive coals having a high
silica content.
Most medium-speed pulverizers are ring-roll and ball-race
mill designs, and are used for all grades of bituminous coal.
Their low power requirements and quick response to changing
boiler loads make them well-suited for industrial boiler
applications and they comprise the largest number of
pulverizers in use overall. High-speed pulverizers include
impact or hammer mills and attrition mills and are also used
for all grades of bituminous coal.
The capacity of a pulverizer is affected by the
grindability of the coal and the required fineness. The
required fineness of pulverization varies with the type of
coal and with the size and type of furnace, and usually ranges
from 60 to 75 weight-percent passing through a 200 mesh
(74 micrometers [/xm] ) screen. To ensure minimum carbon loss
from the furnace, high-rank coals are frequently pulverized to
a finer size than coals of lower rank. When firing certain
low-volatile coals in small pulverized coal furnaces, the
fineness may be as high as 80 weight-percent through a
200 mesh screen in order to reduce carbon loss to acceptable
levels.15
Coal enters the pulverizer with air that has been heated
to 150 to 400 °C (300 to 750 °F), depending on the amount of
moisture in the coal. The pulverizer provides the mixing
necessary for drying, and the pulverized coal and air mixture
then leaves the pulverizer at a temperature ranging from
55 to 80 °C (130 to 180 °F).16
The two basic methods used for moving pulverized coal to
the burners are the storage or bin-and-feeder system, and the
direct-fired system. In the storage system, the pulverized
coal and air (or flue gas) are separated in cyclones and the
coal is then stored in bins and fed to the burners as needed.
In direct-fired systems, the coal and air pass directly from
2-24
-------
the pulverrzers txr the burners and the desired firing rate is
regulated by the rate of pulverizing.
2.2.3.2 Air Supply System. Key air supply system
components are fans and windboxes. The purpose of these
components are to supply the required volumes of air to the
pulverizers and burners, and to transport the combustion gases
from the furnace, through the convective sections, and on to
the air pollution control equipment and stack.
The location of the fan determines the static pressure of
the boiler, which can be characterized as forced-draft,
balanced-draft, or induced draft. A forced-draft boiler
operates at static pressures greater than atmospheric, a
balanced-draft boiler operates with static pressures at or
slightly below atmospheric, and an induced-xlraft boiler
operates at less than atmospheric pressure. Four types of
fans are used: forced-draft, primary-air, induced-draft, and
gas -recirculation.
Forced-draft fans are located at the inlet to the
secondary air supply duct. These fans supply the secondary or
tertiary air used for combustion. The air is typically routed
through the air preheater and then to the windbox. Forced-
draft fans are used on both forced-draft and balanced-draft
boilers.
Primary-air fans are located before or"after the fuel
preparation systems, and provide primary air to the burners.
Induced-draft fans are generally located just before the
stack. These fans pull the combustion gases through the
furnace, convective sections, and air pollution control
equipment. Induced draft fans are used on balanced-draft
boilers to maintain a slightly negative pressure in the
furnace. Induced draft fans are used on induced-draft boilers
to maintain negative static pressure. In this arrangement,
the induced-draft fans are also designed with sufficient
static head to pull secondary air through the air preheater
and windbox.
2-25
-------
Gas recirculation fans are used to transport partially
cooled combustion gases from the economizer outlet back to the
furnace. Gas recirculation can be used for several purposes,
including control of steam temperatures, heat absorption
rates, and slagging. It is also sometimes used to control
flame temperatures, and thereby reduce NOX formation on gas-
and oil-fired boilers. k
The second part of the air supply system is the windbox.
The windbox is essentially an air plenum used for distributing
secondary air to each of the burners. The flow of air to
individual burners is controlled by adjustable air dampers.
By opening or closing these dampers, the relative flow of air
to individual burners can be changed. To increase or decrease
the total air flow to the furnace, the differential pressure
between the windbox and furnace is changed by adjusting the
fans. In boilers having tertiary (overfire) air injection,
tertiary air can be supplied from the windbox supplying
secondary air or by a separate windbox. Separate windboxes
allow greater control of the tertiary air supply rate.
2.2.3.3 Superheaters/Reheaters. Industrial boilers in
cogeneration facilities produce electricity in addition to
process heat/steam. To produce electricity, a steam turbine
converts thermal energy in superheated steam to mechanical
energy (rotation of the turbine and electrical generator
shaft). The amount of electricity that can be produced by the
turbine-generator system is directly related to the amount of
superheat in the steam. If saturated steam is utilized in a
steam turbine, the work done results in a loss of energy by
the steam and subsequent condensation of a portion of the
steam. This moisture, in the form of condensed water
droplets, can cause excessive wear of the turbine blades. If,
however, the steam is heated above the saturation temperature
level (superheated), more useful energy is available prior to
the point of excessive steam condensation in the turbine
exhaust.17
2-26
-------
To provide the additional heat needed to superheat the
steam recovered from the boiler steam drum, a superheater is
installed in the upper section of the boiler. In this area of
the boiler, flue gas temperatures generally exceed 1,100 °C
(2,000 °F) . The superheater transfers this thermal energy to
the steam, superheating it. The steam is then supplied to the
turbine. In some turbine designs, steam recovered from the
turbine after part of its available energy has been used is
routed to a reheater located in the convective pass just after
the superheater. The reheater transfers additional thermal
energy from the flue gas to the stream, which is supplied to a
second turbine.
Superheaters and reheaters are broadly classified as
convective or radiant, depending on the predominate mechanism
of heat transfer to the absorbing surfaces. Radiant
superheaters usually are arranged for direct exposure to the
furnace gases and in some designs form a part of the furnace
enclosure. In other designs, the surface is arranged in the
form of tubular loops or platens of wide lateral spacing that
extend into the furnace. These surfaces are exposed to
high-temperature furnace gases traveling at relatively low
speeds, and the transfer of heat is principally by radiation.
Convective-type superheaters are more common than the
radiant type. They are installed beyond the furnace exit in
the convection pass of the boiler, where the gas temperatures
are lower than those in the furnace. Tubes in convective
superheaters are usually arranged in closely-spaced tube banks
that extend partially or completely across the width of the
gas stream, with the gases flowing through the relatively
narrow spaces between the tubes. The principal mechanism of
heat transfer is by convection.18
The spacing of the tubes in the superheater and reheater
is governed primarily by the type of fuel fired. In the
high-gas-temperature zones of coal-fired boilers, the
adherence and accumulation of ash deposits can reduce the gas
flow area and, in some cases, may completely bridge the space
2-27
-------
between the tubes. Thus, in coal-fired boilers, the spaces
between tubes in the tube banks are increased to avoid excess
pressure drops and to ease ash removal.18 However, because the
combustion of oil and natural gas produces relatively clean
flue gases that are free of ash, the tubes of the superheaters
and reheaters can be more closely spaced in oil- and natural
gas-fired boilers and the superheaters and reheaters »
themselves are more compact.
2.2.3.4 Economizers. Economizers improve boiler
efficiency by recovering heat from the moderate-temperature
combustion gases after the gases leave the superheater and
reheater.
Economizers are vertical or horizontal tube banks that
heat the water feeding the furnace walls of the boiler.
Economizers receive water from the boiler feed pumps at a
temperature appreciably lower than that of saturated steam.
Economizers are used instead of additional steam-generating
surface because the flue gas at the economizer is at a
temperature below that of saturated steam. Although there is
not enough heat remaining in the flue gases for steam
generation at the economizer, the gas can be cooled to lower
temperatures for greater heat recovery and economy.
2.2.3.5 Air Preheaters. On some industrial boilers, air
preheaters are installed following the economizer to further
improve boiler efficiency by transferring residual heat in the
flue gas to the incoming combustion air. Heated combustion
air accelerates flame ignition in the furnace and accelerates
coal drying in coal-fired units.
In large pulverized coal boilers, air heaters reduce the
temperature of the flue gas from the 320 to 430 °C level (600
to 800 °F) at the economizer exit to the 135 to 180 °C level
(275 to 350 °F) at the air heater exit. This energy heats the
combustion air from about 25 °C (80 °F) to between 260 and
400 °C (500 and 750 °F) .19
2-28
-------
2.3 FOSSIL FUEL CHARACTERISTICS
This section contains information on the three fossil
fuels fired in industrial boilers: coal, oil, and natural
gas.
2.3.1 Coal
Coals are classified by rank, i.e., according to their
progressive alteration in the natural metamorphosis from
lignite to anthracite. Volatile matter, fixed carbon,
inherent moisture and oxygen are all indicative of rank. The
American Society for Testing and Materials (ASTM) classifies
coals by rank, according to fixed carbon and volatile matter
content, and heating (calorific) value. Calorific value is
calculated on a moist, mineral-matter-free basis and shown in
table 2-4.20 The ASTM classification for high rank (older)
coals uses volatile matter and fixed carbon contents. The
coal rank increases as the amount of fixed carbon increases
and the amounts of volatile matter and moisture decrease.
Moisture and volatile matter are driven from the coal during
its metamorphism by pressure and heat, thus raising the
fraction of fixed carbon. Low rank (younger) coals are
classified by calorific value and caking (agglomerating)
properties.
The components of a coal are customarily reported in two
different types of analyses, known as "proximate" and
"ultimate." Proximate analysis separates coal into four
fractions: (1) moisture; (2) volatile matter, consisting of
gases and vapors driven off when coal is heated; (3) fixed
carbon, the coke-like residue that burns at higher
temperatures after the volatile matter has been driven off;
and (4) mineral impurities, or coal ash, left when the coal is
completely combusted.
In addition to proximate analysis, which gives
information on the behavior of coal when it is heated,
"ultimate analysis" identifies the primary elements in coal.
These elements include carbon, hydrogen, nitrogen, oxygen, and
sulfur.
2-29
-------
o
fS
w
\^
1
s_,
OQ
C/D
w>
*"* 10
.<*>
?!
ID ^
SE .1
OJ ^
r" 4)
*3 c
JU'z
o .
t_
Q
S_X
W
(A
(C
«5"S CO
- 0)
c/>
01 C
X •»-
u.
>
0
D)
£
£ Of
tO 4-»
t- U
E *«
O (D
0)
^
c
ID
.C
tfl
VI
41
C
CO
O I—
ID 41
D *-<
cr ID
LLJ 0)
u
c
.c
V-
(A
(A
a
c
O 1—
«fc
§-"«
tu a
c
a
H-
tfi
tfi
0
—^
c
to
t- j:
o >-
LU (
U
§-
I
^
i
i
(0
t
c
O)
c
(D
L-
O)
O)
(0
Q
c
1 1 1
1 1 1
CM CO ~»
i l\l CO
i co r\j
to ev 4)
o ••-
U ID U
•» L. 41 «
*^ .e *i u
'o C 'ij 4
CO ID ID (
t- : I. CD
JZ co j: —
^ AJ . J |
C 4> C 5
< Z < w
•« *- CM to
O) D)
C C
eg co
c^ E §
E a> 01
O O) O)
U ID CO
i i i O O O
i i i 000
o o in
>* to «-
""
I I O O O O
i i O O O 0
o o in in
-*" to" «- o"
rg «— i it
INJ K) i i i
«— rvi ro i '
o co o > i
CO N- U 4^U — 'VI —• VI — '(A
S ^ D "o D OC CDC OC
I |c gC li I'i |§
'2 O '^ C "'" ^ J ^ J 4> J
'
£ ^ fvl t^ «* i^
?
CO
u
~ai
CTl
CO
5
c
o o o
o o o
in in in
«-" o" ex
"
0 0 O
o o o
in in to
o o co
i i i
| ] I
i i i
i i i
; : ;
VI VI VI
r D 3
en o o o
>S c c c
1 § § 1
5 ••- co -f- co ••- co
4V XI O XI O XI O
.„ xi o xt u xi u
ri 3 3 3
2 (/) < GO CO (A U
M
i
•*
•" «- C\J to
Ul
c
4-*
ID
L.
B
O
O)
Ol
CO
c
o
c
i
o o
0 0
to to
CO ^
O I
O I
to
1 1
1 1
; ;
I 1
I '
; ;
u < •
IP- 41 4)
11 1
—1 Ij °Jj
•
•* «-^ 0
41
xi
CO
u
&
(D
0
II
[
2-30
-------
Coal analyses may be given on several bases, according to
the application. For coal classification, the moist, mineral-
matter-free basis is generally used. For combustion
calculations, coal is analyzed as-received, including moisture
and mineral matter. Table 2-5 presents sources and analyses
of various ranks of as-received coals.21'22 The nitrogen
contents of these coals are generally less than 2 percent and
does not vary systematically with coal rank.
Various physical properties of coal such as the type and
distribution of mineral matter in the coal and the coal's
"slagging" tendencies are of importance when burning coal.
Mineral matter influences options for washing the coal to
remove ash and sulfur before combustion, the performance of
air pollution control equipment, and the disposal
characteristics of ash collected from the boiler and air
pollution control equipment. Slagging properties influence
the selection of boiler operating conditions, such as furnace
operating temperature and excess air levels, and the rate and
efficiency of coal conversion to usable thermal energy.
2.3.1.1 Anthracite Coal. Anthracite is a hard,
slow-burning coal characterized by a high percentage of fixed
carbon, and a low percentage of volatile matter. Anthracite
coals typically contain 0.8 to 1.0 weight-percent nitrogen.23
Because of its low volatile matter, anthracite is difficult to
ignite and is not commonly burned in boilers. In the United
States, commercial anthracite production occurs almost
exclusively in Pennsylvania.
2.3.1.2 Bituminous Coal. By far the largest group,
bituminous coals are characterized as having a lower
fixed-carbon content, and higher volatile matter content than
anthracite. Typical nitrogen levels are 0.9 to 1.8 weight -
percent.23 Bituminous coals are the primary coal type found in
the United States, occurring throughout much of the
Appalachian, Midwest, and Rocky Mountain regions. Key
distinguishing characteristics of bituminous coal are its
2-31
-------
r»
|
o
U
[n
O
CO
«
S
en
D
o
H
a
>
b
O
CO
W
>H
J
<
2
<
H
O4
EH
<
CO
w
0
g
o
in
W
m
E-"
*
c
5
D
*
t
E
"x
I
A
- 1
TT CO
J
o
i
o
1
1
1
to
1
I o
x 1
"- 3
V k
s I
> E
0
|
O
9
-2
S
>•
O
w
!
1
Clawlficati
o
CO
tD
6
S
oo
0
o
to
n
ID
CO
CD
N
«
-
^
Middle
Newport
£
Meta-anthraclte
o
(M
N
o
n
6
0
CD
O
^
O
PM
(D
a
a
n
(^
S
Q
O
S
e>
6
S
Oft
00
6
(D
CO
O
CO
t—
ID
«
5
O
J
i
2
Anthracite
0
00
o
09
q
c*
03
0>
o"
.-
09
01
to
u
5
O
o
S
o
o
00
00
^
«•
ID
eo
-
to
o
CD
<
?
s
i
Lower Hart
I
*
i
S
8
00
op
o
Cfi
0
n
a
*"
t
w
CO
f*»
^
g
1
1
00
CD
CO
6
tD
O
s
h"
fv
CO
<
n
d
Z
1
Wyoming
1
5
|
i
1
i
ID
O
CO
CD
00
6
rv
a>
n
09
2
1
o
o
o>
£
o
IB
O
•"
CO
n
CO
^f
5
-
<
1
E
Upper KItta
Clearfleld
d
1
2
Medium-volatile
§
10
S
^
10
—
ID
r-
ID
CD
1
§
ID
CD
Ifi
ID
CO
6
tfi
0
ifi
ID
CD
0
Plttaburgh
4
2
|
|
|
r
§
o
CO
rs
10
CO
o'
10
d
z
a
1
^
5
|
CO
s
00
CD
O
00
ID
^
pi
i
0
ID
O
$
o
o
CO
s
00
«.
0
CD
OJ
CO
»
^
0
»
•-
iO
s
!
-
1
1
U
£
o
X
3
ID
a
s
CO
o
ID
*
ID
CO
CD
1
S
s
o
^
•-
*
r*
*
09
S
0>
•"
<
n
1
t
i
i
<
s
(O
o
a
ID
ID
CO
rv
CM
CO
*~
CO
ID
t^
5
!
je
O
0
8
2
a
LO
6
5
CO
»
f«j
S
2
"•
<
Monarch
c
CO
1
0
Subbituminou*
§
09
r^
O
CO
CO
6
09
">
rv
O
ID
LO
d
«
to
CO
5
o
§
ID
O
V
O
CO
to
o
6
0)
ID
ID
S
CO
rv
CO
n
•8
McLean
O
Z
V
"c
O
n
N
CD
O
O
ID
E
! *
I i
2-32
-------
relative volatile matter and sulfur content, and its slagging
and agglomerating characteristics. As a general rule, low-
volatile-matter and low-sulfur bituminous coals are found in
the Southern Appalachian and the Rocky Mountain regions.
Although the amount of volatile matter and sulfur in coal are
independent of each other, coals in the northern and central
Appalachian region and the Midwest frequently have medium to
high contents of both.
2.3.1.3 Subbituminous Coal. Subbituminous coals have
still higher moisture and volatile matter contents. Found
primarily in the Rocky Mountain region, U. S. Subbituminous
coals generally have low sulfur content and little tendency to
agglomerate. The nitrogen content typically ranges from 0.6
to 1.4 weight-percent.23 Because of the low sulfur content in
many Subbituminous coals, their use grew rapidly in the 1970's
and 1980's when lower SC>2 emissions were mandated. Their
higher moisture content and resulting lower heating value
influence the economics of shipping and their use as an
alternate fuel in boilers originally designed to burn
bituminous coals.
2.3.1.4 Lignite. Lignites are the least metamorphesized
coals and have a moisture content of up to 45 percent,
resulting in lower heating values than higher ranking coals.
The nitrogen content of lignites generally range from 0.5
to 0.8 weight -percent ,23 Commercial lignite production occurs
primarily in Texas and North Dakota. Because of its high
moisture content and low heating value, lignite is generally
used in facilities located near the producing mine.
2.3.2 Oil
Fuel oils produced from crude oil are used as fuels in
industrial boilers. The term "fuel oil" covers a broad range
of petroleum products, from a light petroleum fraction similar
to kerosene or gas oil, to a heavy residue left after
distilling off fixed gases, gasoline, gas oil, and other
lighter hydrocarbon streams.
2-33
-------
Specifications established by the ASTM for grades of fuel
oil are shown in table 2-6.24 Fuel oils are graded according
to specific gravity and viscosity, the lightest being No. 1
and the heaviest No. 6. Typical properties of the standard
grades of fuel oils are given in table 2-7.25'26-27
Compared to coal, fuel oils are relatively easy to burn.
Preheating is not required for the lighter oils, andkmost
heavier oils are also relatively simple to handle. Ash
content is minimal compared to coal, and the amount of
particulate matter (PM) in the flue gas is correspondingly
small.
Although more expensive than residual oils, distillate
oils are often burned in the industrial boilers. Emissions of
PM and SC>2 are lower when burning distillate oils compared to
residual oils or coal.
The U. S. supply of fuel oils comes from both domestic
and foreign production. The composition of individual fuel
oils will vary depending on the source of the crude oil and
the extent of refining operations. Because of these factors
and the economics of oil transportation, fuel oil supplies
vary in composition across the United States, but are
relatively uniform with the exception of sulfur content. In
general, ash content varies from nil to 0.5 percent, and the
nitrogen content is typically below 0.4 weight-percent for
grades 1 and 2, 0.2 to 2.5 weight-percent for grades 4 and 5,
and 0.3 to 3.0 weight-percent for grade 6.25
2.3.3 Natural Gas
Natural gas is a desirable fuel for steam generation
because it is practically free of noncombustible gases and
residual ash. When burned, it mixes very efficiently with
air, providing complete combustion at low excess air levels
and eliminating the need for particulate control systems.
The analyses of selected samples of as-collected natural
gas from U. S. fields are shown in table 2-8.28 Prior to
distribution, however, most of the inerts (carbon dioxide
2-34
-------
CO
>3
H
O
£
§
o
H
H
b
H
O
W
a
w
I
• £
il
o
r
li
Kinematic vlacotlty, contlatokoa
Saybolt vlacotlty, tec
Distillation tamperaturas, f
CCI
(3.091 J.ZZ I IV
8
<£
¥•
U.
¥•
§
I8
D
c
S
si
* >i
0 *
1*8
ill
if*
&
Ije
O ""
i
c:
S
i
c
S
1
c
2
^'
1
i
I
1
1
1
1
I
I
i
n
i
to
'
«
:
!
'
O 00
to co
10 *s
'
II
••
to
d
i
o
lac) 001
1
-i
g
S
q
(N
'
'
OS
n
i
o S
^ o
CP 0
o w
f 00
!
1
to
o
o
o
85
I8CI OOI
CM
O
z
:
'
(b
£
:
'
10
to
'
o
6
i
6
85
130(66)
1
i
§
8
o
1
o
6
8
"
130(66}
to 7
if
•
i
i
(0
1
o
1
o
10
o
to
n
i
1
o
d
'
o
q
13O/65)
to *?
t
1
s
'
O
O
f)
to
s
I
S
i
i
•
|
1 60 (66)
eg
i
I
i
I
«
2-35
-------
rH
VO f*
V
•
Q *""*
3
rH
•H
01
3
^ rH
.. "rH
o
rH
0)
^rH
""o
g
CD
3
§
0)
-a
id
^
o
rH
irl
3
•6
__j
a
Light
residual
£rH
|d
rH T3
-H
C eu
0) rl
dl
4J
rH
rH
-H
4->
to
-H
P
4J "a?
id C
rH 0)
1 ff*
rH UJ
-n O
4J M
CO 0)
a
•i*t
u
id
rH
ffl
f" Black ]
^
u
rH
03
)H
0)
4J
X!
Oi
•H
Color
CN
i-t
t~
rH
rH
CN
CN
CO
0
IT.
0
0
vo
^
£•
•H
^
id
^
Oi
H
Si
H
VO
00
O
CN
in
en
o
en
CN
en
0
in
VD
00
o
H
in
fN
00
o
b.
0
o
vo
o
vo
„
>•,
4J
•H
^
It)
0
u
-H
(l_|
-rl
U
0)
ft
CO
04
H
CN
.
00
in
ro
en
r-
CN
"
vo
o
CN
^
O
00
vo
bl
o
o
vo
^
C
O
i-H
rH
id
01
V,
I-*
o
,
o
vo
fi
o
o
in
o
in
H
00
vo
CN
VO
H
0
o
0
rH
CO
0)
0
m
•rl
jj
£
0)
u
^
>,
4->
-H
m
Visco
in
VD
0
0
H
0
^
<1)
N
jj
0)
PQ
0)
N
| Below
(K
O
4J
C
-H
0
ft
M
§
O<
O
O
rH
35
(minimum)
^
P
\f\ r4
U 1 Tl
H C
•H
U
>p^
M
5
ft
CO
0
4J
tJ
-H
0)
J3
Atmosp
CM
o
^
D»
C
&
E
3
&
^
o
MH
&
S
o
o
o
m
H
~
3
in .§
-H
u
>r]
r<
0)
ft
(Q
0
3
U
-rt
CU
Atmosp
b.
o
v
Oi
-H
N
-H
Q
4J
id
^4
o
*4-l
I
o
•
CN
H
O
in
in
0)
U
a)
EH
0)
u
id
<*>
^
0)
3
•0
•H
CO
0)
i-l
C
Carbo
o
•
M
t
m
o
in
CN
i
CN
O
in
*
J^
«.
C
0)
Nitro
in
•
o
H
r-
rH
rH
en
i-H
H
^
a
o
o
o
^
o
in
H
o
o
o
CO
•«*<
H
o
0
o
vo
I-l
o
o
0
rH
rH
O
0
o
rH
C
O
rH
i-H
id
01
4J
0
H
E-<
r>
i
CN
w
I
2-36
-------
CO
O
,_T
3
H
P
rt«
[g
OS
CO
CO Q
W i-3
O, H
S Pt,
CO CO
W
Q E-"
w <;
U CO
W
J P
W W
CO EH
H
O B
CO 2
U 0
H Pi
EH PL,
CO
H
W
«jj
p2
J?
*TJ
U
,
00
CN
W
EH
rd
GQ
rd
O
«M
O
CD
U
j_i
3
CO
'O
C
rd
•
o
55
0)
rH
&
g
rd
CO
fc->
ing
j
^ 5
"
jTj
m g
V— J
6
\~s
CN .
CO
^
iH >?*
IM
m
u
-H
4J
CO
-H
^1
CD
4_)
o
rd
td
u
rHOrHOrHOOrHrH
•H rH -H C^ -H 00 *& -H -H
S3 "S3 • S3 • *S3S3
^ vo o oo
00
,-IOrHOr-lrHOrHrH
-H O -H O -H -H O -H -H
S3 *S "S5S3 'S3S
o in in
cr\
CNromrHincNOinoo
rHmo oomoo
CTi
rHOrHOrHOOiHiH
•H O -H 00 -H C^ in -H -H
S3 -S3 -S3 . ' S3 S3
*J< • ^ O O
00 H
iHOrHOrHrHOrHrH
-H ^ -H 00 -H -H 00 -H -H
S3 "S3 • S3 S3 • S3 S3
n in o
00 H
CD
CD -d
T5 CD -H
-H "O 1W
rH X -H rH
0 0X3
> G O co
O-H
>i d CD g -d a a
.Q 0) 0) C (U CD
tjicjcDCDGCcntHCTi
dip ordrHCOoocDO
>-ij5 ^ifdX3J3 r-i tnl-i
-u SJSPLlWUrjSSOK
a
CD
3
4J
CO -H
0) 4->
CO CO 'i1 VO
Jx, fi 'tfffi ffi CN CO
rH O CNK CN CNO O CN CN CN
rdU KUCJUUUS3OK
s
rH in *3* O iH
•H CO 00 CTl ^
S3 ....
O *f CN H
CN VO H
rH OO VO VO rH
-H VO CN O -H
S • • -S3
CN cn co
CN VO
^f C^ C^ ^-P OO
m CN H t*> in
o m en in H
CN VO
rH O CN VO CN
-H m r- c*- CN
53 . . . .
CO ^* O iH
CN >
iH en in CN rH
•H in oj CN -H
S3 • • • S3
en in H
CN C-
c c
CD CD
M tnfl end
4J 3 O O O 0)
? u-i V-^ Q V-i l^n
rH *d >-l -U >i
>^ 3 >i ni -H X
^3 co 55 U S3 O
dU°
1^
CD
1 1
(d
-H CN CN CN
iJ CO W U S3 0
rH
ID
o
co
.
0
o
o
VD
.
O
r>
VD
I P»
U 1
o
vo
vo
«
o
vo
en
vo
«
o
•£•
-H
rd
O
rH
CD
—
jx,
4-)
-H
rd
SH
CD
O
-H
l*-l
-H
U
0)
a
CO
«* o
r> vo
cr\ H
>.
o
CN
CN «*
*O CN
O 00
•. ^
H rH
CN
«3* r^
vo >
CTi O
CN
CN
VO «3<
H 0
H CTi
>. -
H CN
CN
cr\ o
CN r>
H H
- -
H CO
OJ
CD
ffi
,
a
-H
O
en
US
fJH
O
0)
3 O rH
rH VO CD
rd 3
4-) 4J M-l
rd u-i O
-------
and nitrogen), sulfur compounds, and liquid petroleum
gas (LPG) fractions are removed during purification processes.
As a result, natural gas supplies burned by industrial boilers
are generally in excess of 90 percent methane, with nitrogen
contents typically less than 0.4 percent.29-30'31
2.4 INDUSTRIAL BOILER NOX FORMATION MECHANISMS
Nitrogen oxide emissions from combustion devices are
comprised of nitric oxide (NO), nitrogen dioxide (N02), and
nitrous oxide (N20) .' For most combustion systems, NO is the
predominant NOX species. This section discusses how
differences in boiler design, fuel characteristics, and
operating characteristics can affect NOX emissions.
2.4.1 Nitrogen Oxide Formation
The formation of NOX from a specific combustion device is
determined by the interaction of chemical and physical
processes occurring within the furnace. This section
discusses the three principal chemical processes for NOX
formation. These are: (1) "thermal" NOX, which is the
oxidation of atmospheric nitrogen; (2) "prompt" NOX, which is
formed by chemical reactions between hydrocarbon fragments and
atmospheric nitrogen; and (3) "fuel" NOX, which is formed from
chemical reactions involving nitrogen atoms chemically bound
within the fuel.
2.4.1.1 Thermal Nitrogen Oxide Formation. Thermal NOX
results from the oxidation of atmospheric nitrogen in the
high-temperature, post-flame region of a combustion system.
During combustion, oxygen radicals are formed and attack
atmospheric nitrogen molecules to start the reactions that
comprise the thermal NOX formation mechanism:
0 + N2 ^ NO + N . (2-1)
N + 02 ** NO + 0 (2-2)
N + OH ^ NO + H (2-3)
"N20 is not considered a component of NOX for regulatory
purposes under the Clean Air Act.
2-38
-------
The first reaction (equation 2-1) is generally assumed to
determine the rate of thermal NOX formation because of its
high activation energy of 76.5 kcal/mole. Because of this
reaction's high activation energy, NOX formation is slower
than other combustion reactions causing large amounts of NO to
form only after the energy release reactions have equilibrated
(i.e., after combustion is "complete"). Thus, NO formation
can be approximated in the post-combustion flame region by:
[NO] = ke-K/T [N2] [0211/2 t (2-4)
where:
[ ] are mole fractions,
k and K are reaction constants,
T is temperature, and t is time.
The major factors that influence thermal NOX formation
are temperature, oxygen and nitrogen concentrations, and
residence time. Of these four factors, temperature is the
most important. Thermal NOX formation is an exponential
function of temperature (equation 2-4). One of the
fundamental parameters affecting temperature is the local
equivalence ratio.b Flame temperature peaks at equivalence
ratios near one as shown in figure 2-8.32 If the system is
fuel-rich, then there is not sufficient oxygen to burn all the
fuel, the energy release is not maximized, and peak
temperatures decrease. If the system is fuel-lean, there are
additional combustion gases to absorb heat from the combustion
reactions, thus decreasing peak temperatures. A premixed
flame0 may exist at a wide range of equivalence ratios, and
thus premixed flames have a wide range of peak temperatures.
""Equivalence ratio is defined as the fuel/oxidizer ratio
divided by the stoichiometric fuel/oxidizer ratio. The
equivalence ratio is given the symbol .
CA premixed flame exists when the reactants are mixed prior to
chemical reaction.
2-39
-------
s
I
V
a
.0)
03
1.0
Fuel-lean
Fuel-rich
&= (F/A)/(F/A)
Stoich
Figure 2-8. Variation of flame temperature with
equivalence ratio.
2-40
-------
However, a non-premixed flamed will generally react near an
equivalence ratio of one, causing high peak temperatures.
The temperature is also related to the heat release per
unit of burner zone volume. Units with large heat release
rates per unit volume generally experience higher
temperatures, creating high NOX levels.
2.4.1.2 Prompt Nitrogen Oxide Formation. Prompt NOX is
the formation of NOX in the combustion system through the
reactions of hydrocarbon fragments and atmospheric nitrogen.
As opposed to the slower thermal NOX formation, prompt NOX
formation is rapid and occurs on a time scale comparable to
the energy release reactions (i.e., within the flame). Thus,
it is not possible to suppress prompt NOX formation in the
manner as thermal NOX formation. However, the contribution of
prompt NOX to the total NOX emissions of a system is rarely
large.33
Although there is some uncertainty in the detailed
mechanisms for prompt NOX formation, it is generally believed
that the principal product of the initial reactions is
hydrogen cyanide (HCN) or CN radicals, and that the presence
of hydrocarbon species is essential for the reactions to take
place.34 The following reactions are the most likely
initiating steps for prompt NOX:35
CH + N2 ** HCN + N (2-5)
CH2 + N2 ** HCN + NH (2-6)
The HCN radical is then further reduced to form NO and other
nitrogen oxides.
Measured levels of prompt NOX for a number of hydrocarbon
compounds in a premixed flame show that the maximum prompt NOX
is reached on the fuel-rich side of stoichiometry.36 On the
fuel-lean side of stoichiometry, few hydrocarbon fragments are
free to react with atmospheric nitrogen to form HCN, the
precursor to prompt NOX. With increasingly fuel-rich
dA non-premixed flame exists where the reactants must diffuse
into each other during chemical reaction.
2-41
-------
conditions, an increasing amount of HCN is formed, creating
more NOX. However, above an equivalence ratio of
approximately 1.4, there are not enough 0 radicals present to
react with HCN and form NO, so NO levels decrease.
2.4.1.3 Fuel Nitrogen Oxide Formation. The oxidation of
fuel-bound nitrogen is the principal source of NOX emissions
in combustion of coal and some oils. All indications' are that
the oxidation of fuel-bound nitrogen compounds to NO is rapid
and occurs on a time scale comparable to the energy release
reactions during combustion. Thus, as with prompt NOX, the
reaction system cannot be quenched as it can be for thermal
NOX.
Although some details of the kinetic mechanism for
conversion of fuel nitrogen to NOX are unresolved at the
present time, the sequence of kinetic processes is believed to
be a rapid thermal decomposition of the parent fuel-nitrogen
species, such as pyridine, picoline, nicotine, and quinoline,
to low molecular weight compounds, such as HCN, and subsequent
decay of these intermediates to NO or nitrogen (N2). In
stoichiometric or fuel-lean situations, the intermediates will
generally react to form NO over N2, whereas in fuel-rich
systems, there is evidence that the formation of N2 is
competitive with the formation of NO. This may, in part, be
the cause of high NOX emissions in fuel-lean and
stoichiometric mixtures and lower NOX emissions in fuel-rich
systems.
Several studies have been conducted to determine factors
that affect fuel NOX emissions. One study on coal combustion
found that under pyrolysis conditions, 65 percent of the fuel
nitrogen remained in the coal after heating to 750 °C
(1,380 °F) but only 10 percent remained at 1,320 °C
(2,400 °F) ,37 This suggests that the formation of NOX may
depend upon the availability of oxygen to react with the
nitrogen during coal devolitization and the initial stages of
combustion. Consequently, if the initial stage of combustion
occurs under fuel-rich conditions, NOX formation will be
2-42
-------
suppressed because the oxygen concentration is lower relative
to fuel-lean conditions. This promotes the formation of N2
over NO. If the initial stage of combustion occurs under
fuel-lean conditions, the formation of NO will be promoted,
resulting in greater NOX emissions than under fuel-rich
conditions.
During another study, fuel NOX was measured in a* large
tangentially-fired coal utility boiler. Figure 2-9 shows that
fuel NOX formation correlated well with the fuel oxygen/
nitrogen ratio), which suggests that fuel oxygen (or some
other fuel property that correlates well with fuel oxygen)
influences the percentage of fuel nitrogen converted to fuel
NOX.38 This corresponds to previous observations that greater
levels of NOX are found in fuel-lean combustion environments.
There is no readily apparent correlation between the
quantity of fuel nitrogen in coal and fuel NOX as shown in
figure 2-10.39 Note, however, that most of the tested coals
contained approximately 1.0 percent nitrogen or higher.
2.4.2 Factors that Affect NOX Emissions
The formation of thermal, prompt, and fuel NOX in
combustion systems is controlled by the interplay of
equivalence ratio with combustion gas temperature, residence
time, and turbulence (sometimes referred to as the "three
T's"). Of primary importance are the localized conditions
within and immediately following the flame zone where most
combustion reactions occur. The equivalence ratio and the
three T's are determined by factors associated with burner and
boiler design, fuel characteristics, and boiler operating
conditions. This section discusses how boiler design, fuel
characteristics, and boiler operating conditions can influence
baseline (or uncontrolled) NOX emission rates.
2.4.2.1. Boiler Design Characteristics. There are a
number of different furnace configurations used in industrial
boilers. These include wall, tangential, spreader stoker,
PBC, and packaged designs. Background information on each of
these boiler designs was presented in earlier sections. Each
2-43
-------
S2
Q>
"55
£
16
14
12
10
8
6
4
2
0
6 10 14 18 22 26
Ratio of Coal Oxygen to Coal Nitrogen
30
32
Figure 2-9. Conversion of fuel-bound nitrogen-to-nitrogen
oxide in pulverized-coal combustion.
38
2-44
-------
CM
o
1400
1200
1000
800
600
400
- Lignite
Subbit.
CCoal
High-Vol.
BitC
High-Vol.
BitB
High-Vol.
Bit A
Med.-Vol.
Bit
Low-Vol.
Bituminous
I I I I I I I 1_
1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0
% Nitrogen in Fuel (DAF)
Figure 2-10. Fuel NOX to fuel nitrogen content-pulverized
coal, preroixed. 9
2-45
-------
configuration has design characteristics that partially
determine the uncontrolled NOX emissions of the boiler.
2.4.2.1.1 Wall-fired boilers. There are two types of
dry-bottom wall-fired boilers that are likely to be subject to
the revision to the current standard. They are single- and
opposed-wall.
Single, wall-fired boilers consist of several rows of
circular burners mounted on either the front or rear wall of
the furnace. Opposed, wall-fired boilers also use circular
burners, but have burners on two opposing furnace walls and
have a greater furnace depth.
Circular burners introduce a fuel-rich mixture of fuel
and primary air into the furnace through a central nozzle.
Secondary air is supplied to the burner through separate
adjustable inlet air vanes. In most circular burners, these
air vanes are positioned tangentially to the burner centerline
and impart rotation and turbulence to the secondary air. The
degree of air swirl, in conjunction with the flow-shaping
contour of the burner throat, establishes a recirculation
pattern extending several burner throat diameters into the
furnace. The high levels of turbulence between the fuel and
secondary air streams creates a nearly stoichiometric
combustion mixture. Under these conditions, combustion gas
temperatures can be high and contribute to thermal NOX
formation. In addition, the high level of turbulence can
cause the amount of time available for fuel reactions under
reducing conditions to be relatively short, thus increasing
the potential for formation of fuel NOX.
2.4.2.1.2 Tangentially-fired boilers. The burners in
tangentially-fired furnaces are incorporated into stacked
assemblies that include several levels of primary fuel nozzles
interspersed with secondary air supply nozzles and warmup
guns. The burners inject stratified layers of fuel and
secondary air into a relatively low turbulence environment
outside the center fireball. The stratification of fuel and
air creates fuel-rich regions in an overall fuel-lean
2-46
-------
/ environment. Before the layers are mixed, ignition is
initiated in the fuel-rich region. Near the highly turbulent
center fireball, cooler secondary air is quickly mixed with
the burning fuel-rich region, insuring complete combustion.
The off-stoichiometric combustion reduces local peak
temperatures and thermal NOX formation. In addition, the
delayed mixing of fuel and air provides the fuel-nitrogen
compounds a greater residence time in the fuel-rich
environment, thus reducing fuel NOX formation.
2.4.2.1.3 Fluidized bed combustion boilers. Fluidized
bed combustion technology is designed to reduce emissions of
SC>2 and NOX. Boilers using FBC designs have lower
uncontrolled NOX emission rates than conventional boilers.
The primary reason for the low NOX emission rates from FBC
boilers is the absence of thermal NOX emissions due to the low
combustion temperatures. An FBC boiler may typically operate
at 800 to 900 °C (1,480 to 1,650 °F) while a conventional
( boiler may operate at 1,370 to 1,540 °C (2,500 to 2,800 °F) .
Another influence on the NOX emissions from an FBC
boilers is the quantity of calcium oxide, used for SC>2
emissions control, present in the bed material. Higher
quantities of calcium oxide result in higher base emissions of
NOX. Therefore, as 862 removal requirements increase, base
NOX production may increase.12
2.4.2.1.4 Stoker-firing. Stokers are generally low
capacity boilers which burn crushed coal particles in
suspension, while larger particles are burned in a fuel bed on
a grate. They typically are operated with high levels of
excess air to insure complete combustion and to maintain
relatively low grate temperatures. The low uncontrolled NOX
emissions are believed to be a function of the lower furnace
temperatures [~1,090 °C ("2,000 °F), compared to temperatures
of 1,370 to 1,570 °C (2,500 to 2,800 °F) found in other boiler
types].
f
2-47
-------
2.4.2.1.5 Packaged boilers. For smaller packaged
natural gas- or oil-fired boilers, NOX emissions generally
depend more on fuel and heat release rate. In general,
packaged boilers with higher heat release rates tend to have
higher levels of NOx.40
2.4.2.2 Fuel Characteristics. In the combustion of
"clean" fuels (fuels not containing nitrogen compounds, such
as natural gase) , the thermal mechanism is typically the
principal source of nitrogen oxide emissions. However, as the
nitrogen content of the fuel increases (table 2-7),
significant contributions from the fuel nitrogen occur.41
Thus, the nitrogen content of the fuel is a partial indicator
of NOX emission potential.
The type of fuel dictates certain design characteristics
of a given boiler. Because of residence time requirements and
the impact of ash on tube erosion and fouling, boilers
designed for coal have larger furnace volumes than boilers
designed for oil or gas as shown in figure 2-II.42
2.4.2.3 Boiler Operating Conditions. During the normal
operation of an industrial boiler, factors that affect NOX
continuously change as the boiler goes through its daily
operating cycle. During a daily operating cycle, the
following factors may change and affect NOX formation:
• Steam requirements,
• Excess oxygen, and
• Burner secondary air register settings.
The effects of changing steam requirements on NOX
emissions are varied and complex. With an increase in load,
furnace temperatures increase and excess oxygen levels
typically decrease. In wall-fired boilers, an increase in
load generates higher turbulence which in turn increases the
local temperatures in the furnace. This increases the
°The nitrogen present in natural gas exists almost
exclusively as elemental nitrogen and not as organic nitrogen
compounds.
2-48
-------
£H
Si
/
to
ra
CD
1
QoQn°o
to
r-|
0)
4J
C
0)
b
O
«M
«w
•H
•O
•H
U-l
to
h
0)
n
;g
8
QC
\[
ra
I
s
3
•P
CO
0)
10
w
n
0)
N
O
-H
10
>i
a,
-P
«J
a
•• *H
0) r-l
£ •
O CM
2-49
-------
potential for thermal NOX formation. With other burner and
boiler designs, however, the decrease in C>2 at increased load
levels can decrease both fuel and thermal NOX formation.
Because of these many varied effects, it is difficult to
predict the overall effect of changing operating load on NOX
emissions. The net effect will depend on the changes made to
other boiler operating parameters during load changes".
The effect of excess oxygen or burner secondary air
register settings on NOX emissions can vary. Altering the
excess oxygen levels may change flame stoichiometry.
Increasing secondary air flow may increase entrainment of
cooler secondary air into the combustion regime, lowering
local temperatures, and increase fuel and air mixing, altering
equivalence ratio. The net result of both actions may be
either to raise or lower NOX emissions, depending on other
unit-specific parameters.
2.5 BASELINE EMISSIONS
This section presents baseline emission rates (Ib/MMBtu)
and projected total baseline emissions (tpy). This section
also discusses existing federal and state regulations for this
source category.
2.5.1 Baseline Emission Levels
Table 2-9 presents baseline emission rates (Ib/MMBtu heat
input) for each fuel and boiler type as required by subpart
Db. The new steam generating capacity projected for the
five-year period (1996-2000) was presented earlier in
table 2-3. A breakdown of these data by boiler type and
industry segment are presented in table 2-10. For each
industry segment, capacity utilization factors were assigned
based on current use patterns. To estimate baseline emissions
in the year 2000 for the source category, the emission rates
from table 2-943 were combined with the steam generating
capacity and capacity utilization factors presented in
table 2-10. An emission rate of 0.02 Ib/MMBtu less than the
current emission limit was used to represent average annual
2-50
-------
TABLE 2-9. EXISTING SUBPART Db NITROGEN OXIDE EMISSION
LIMITS FOR FOSSIL FUEL- FIRED INDUSTRIAL BOILERS
NOX emission
limits3
Fuel/boiler type _ (Ib/MMBtu)
Coal
%
Pulverized Coal -fired (PC) 0.70
Spreader Stoker (SS) 0.60
Fluidized Bed Combustion (FBC) 0.60b
Residual Oil
Low Heat Release Rate Units - -
Field-erected (FE) 0.30
High Heat Release Rate Units - -
Packaged (Pkg) 0.40
Distillate Oil /Natural Gas
Low Heat Release Rate Units - -
Field-erected (FE) 0.10
High Heat Release Rate Units - -
Packaged (Pkg) _ 0.20 _
aBased on 30 -day rolling averages.
emission rate used in calculation of baseline
emissions for FBC was 0.4 Ib/MMBtu based on information in
Reference 43.
2-51
-------
s^
CQ
^H
EH
H
U EH
< fc
rtj S
u o
w
O CO
55
H >H
S EH"
2 CO
W D
H
W EH
EH
L) Cd
w w
f~2 J "^
O H
Pi O
CU PQ
O
1-1
CN
0)
rt
^
|f
"m
z£
u
Ul J=
"" ^
(3*00
Z £
Xl"'
^^
™*
2
°£
° £
c_
UJ .C
o m
o £
O)^
°"i
"E
^
t_
UJ -C
•« 4->
o m
£
L
jr
u >
S|
^
.c
en 3
c/> *-*
CD
i
^
.c
U 3
O. V
to
X
X.
ZL.
— O
S. R
B U.
U
r
%l *•
ti
•T^ 2
to"6
N.
8
CM
O>
CM
^J"
g
in
CM
CM
CM
g
O
«-"
CM
*-__
to
o
in
CM
o
t-
•o
CM
to
in
ru
o
u
3
4-*
U
ra
H-
g
CO
X
g
z
•4- CM
«- CM
-* to
^
O O
in
^*
to o
«-
0 0
to o
CM
0 0
0 0
o o
o o
in o
to oo
0 0
01
O)
CO
t_
o> in
oi ^
I
"E o,
U. 1—
S
to
o-
o
CM
^T
o
o
o
o
o
o
o
o
•o
o
B
•g
CO
u
-1
to 5
CM" to
«- in
>O 10
to" «-"
in to
>i eo
*- CM
O O
«- o
«- o
o ~»
in
CM
o -o
CM
O >»
CM
0 ~t
00
to
-*
O O
00 00
o o
CO
u
S. 1
CO ^
a. u
CM
CM
in
o
oo
c
u
4-*
O — •
a a
M- *J
i g
ID Of
£
C.
• 01
O —
•— o
£ CQ
CM
in
o
o
o
o
o
N.
o
CM
o
o
o
in
CM
o
CO
u
1
(A
ID to
V L.
•^ 0)
z u
CM t^
in CM
-* CM
Kl
to oo
o"
CM
0 g
o
in
o oo
§
CM
o in
0
«-;'
h* in
to «-
CM O
in
o o
o
o .»
CM
o «-
"^
in
CM
o
v>
01
O)
01
o
u
1
U)
8 «
£ +*
u o
—
"8
D)
CO
u
H
a.
II
at
a.
.-
"8
^
L.
OJ
1
H-
11
UJ
Q
1
t/i
1-
"iT a
v c
i- ii
— 13
^ z
o —
0 0
N CO
l_ — ••
— VI
!!
a. o
to
2-52
-------
emissions. Annual emissions were estimated with the following
equation:
Tons per year NOX = size * CF * ER * 4.38
where: size = boiler heat input capacity (MMBtu/hr)
CF = capacity factor (decimal fraction)
ER = annual average NOX emission rate
(Ib/MMBtu)
Based on the above equation, assumptions, and use of a range
of emission rates for each boiler/fuel type, the baseline NOX
emissions in the fifth year is estimated to be 59,600 tpy.
Table 2-11 shows the baseline NOX emissions by boiler type and
industry sector. As shown in the table, the chemical industry
segment has the largest baseline NOX emissions (20,100 tpy)
with over half coming from PC boilers (10,400 tpy). Also, PC
boilers have higher baseline NOX emissions followed closely by
distillate oil-fired packaged boilers.
2.5.2 Other Regulations on the Source Category
Multiple regulations affect this source category. In
addition to subpart Db, the primary Federal program affecting
this source category is the New Source Review (NSR) program.
The NSR program regulates NOX emissions from new industrial
boilers through Title I of the Clean Air Act and the
Prevention of Significant Deterioration (PSD) program.
Title I, implemented at the State or Federal level, links NOX
emissions to ozone formation and requires LAER controls for
new sources located in ozone non-attainment areas. The PSD
program, also implemented at the State or Federal level,
regulates NOX emissions from new sources that are located in
ozone attainment areas. Under the PSD program, the source
must implement BACT.
Under these two programs, and considering local air
quality conditions, limits more stringent than a national
standard (NSPS) may be set on a case-by-case basis. For
2-53
-------
^"*
P4
EH
u_I
Q
g
M
g
<
W
dl
EH
pans
. ~] ^_^
H pi
O fd,
CQ Cd
CQ Pi
W
CO CU
O CO
ii h^
CO O
CO H
i i ^__^
2
W H
XW
0 S
S3 O
W
W CO
S3
H
W
CO
CQ
,_j
H
U
a ^
CQ
3
EH
(0
o
t—
en
o.
cJ
UJ
u.
s
o
z
O)
.*
a
.
o
o
u_
o
O
D)
^£
O.
d"
UJ
u.
0*
0£.
CJ
CO
"~
(S>
V)
0-
•~ o
O 4-»
(D U
a. to
(0 s-
u
L.
•° £
fc£
^ «
O -F
"£
o
o
^o*
in
o
>t
PJ
0
o
o
o
o
f=
^x
*—
o
o
Kl
PJ"
in
PJ
o
|
U
CO
1
1
Q
z
o
>o
PJ
o
PJ
PJ
o
in
PJ
«-
o
ro
o
o
o
o
o
o
in
o
01
D)
10
L,
CO
•g
0
u.
o o o o
O in N- h-
pj r^ \o o
^± «—
PJ
o o o o
0 CO O CO
PJ «— ^t PJ
«— PJ
0 5 eo §
in PJ co
PJ
o o o o
O CO
"~~
o o o o
o o o o
«- 0
CO CO
o o o o
in
PJ
(
o o o o
CO
^
ro
o o o o
-4-
pj"
O O O O
in
o
00 ~
•— o
^ ta
g
PJ
o
o
oo
o
o
Q
§
0
o
o
in
PJ
o
in
L.
0)
4J
s
0
CO
u
CO
'a
c/l
0
S §
o ro
*-T ^^
o o
Sr^
ro
ro"
g |
in
o o
PJ
o
PJ
0 S
^
o o
o
•o
P ^
r*- >*
ro
CM
0 0
in
in
in
0 0
xf
f\J
0 g
eo
in
PJ
o
1A
01
O!
01
O
U
•g
CO
(a
"o -•
o a
*- 4_|
U O
•b"
o>
u
01
01
2
01
•4-
II
UJ
•g
l_
_L
o
1
V)
01
^*
II
o
a:
c
o
i
y
•o
J3
^ t3
IM c.
"D *4-
'5 i)
— ' CD
H- O)
II — •
CD
U 1-
**" ffl
L" ^
U II
"0 C3
M Z
OJ 01
"D t-
CO •—
01 H-
Ol !~ 1
U CL —
CO VI —
~0. II °
01
CA CO
£ ..-
*" "S '£
01 i- CO
O O
"S ° °
"? "8 "
o -2 o>
1- L. CO
01 It
Vi > O
§ =1
•t- Q.
U II
-------
example, a recently permitted coal-fired boiler in Iowa was
required to achieve a NOX emission limit of 0.12 Ib/MMBtu.44
Other coal-fired boilers are required to meet limits ranging
from 0.25 to 0.6 Ib/MMBtu.
Because of the active regulation of this source category,
the actual NOX baseline is expected to be lower than the
estimate of 59,600 tpy. As a result, the potential emission
reductions achieved by revision of the current standard will
be lower.
2-55
-------
2.6 REFERENCES
1. Memorandum from Gideon, L. and M. Gundappa, Radian Corp.
to James Eddinger, U.S. Environmental Protection Agency,
Industrial Boiler 5-year Growth Estimates. November 23,
1994.
2. A statistical study report for Radian Corporation
prepared by American Boiler Manufacturers Association
(ABMA). October 1994.
3. The Boiler Market Outlook. A Special Report, The WEFA
Group. Prepared for the American Boiler Manufacturers
Association (ABMA). January 1994.
4. Singer, J. G. (ed). Combustion, Fossil Power Systems,
Third Edition. Combustion Engineering, Inc. 1981.
p. 13-3.
5. Baumeister, T., E. A. Avallone, and T. Baumeister, III
(eds.). Mark's Standard Handbook for Mechanical
Engineers, Eighth Edition. McGraw-Hill Book Company.
1978. p. 9-12.
6. Ref., 4, p. 13-4.
7. Ref. 4, p. 13-5.
8. Ref. 4, pp. 13-19.
9. White, D. M., and M. Maibodi. "Assessment of Control
Technologies for Reducing Emissions of S02 and NOX from
Existing Coal-Fired Boilers." Prepared for the U. S.
Environmental Protection Agency. Air and Energy
Engineering Research Laboratory. Final EPA Report
No. 600/7-90-018. April 1990. pp. 3-13 through 3-15.
10. Alternative Control Techniques Document--NOX Emissions
for Industrial/Commercial Institutional (ICI) Boilers.
EPA-453/R-94-022. U.S. Environmental Protection Agency.
Office of Air Quality Planning and Standards. Research
Triangle Park, NC. March 1994. p. 3-16.
11. Makansi, J. and R. Schwieger. "Fluidized-bed Boilers."
Power. May 1987. pp. S-l through S-16.
12. State-of-the-Art Analysis of NOX/N20 Control for
Fluidized Bed Combustion Power Plants. Acurex
Corporation. Mountain View, CA. Acurex Final Report
90-102/ESD. July 1990.
13. Ref. 10, pp. 3-26 - 3-27.
2-56
-------
14. Steam. Its Generation and Use. Babcock & Wilcox. New
York, NY. 1975. p. 9-1.
15. Ref. 4, p. 12-7.
16. Ref. 5, p. 9-11.
17. Ref. 14, p. 12-8.
18. Ref. 5, p. 9-20. k
19. Ref. 4, p. 5-11.
20. Ref. 4, p. 2-3.
21. Ref. 5, p. 7-4.
22. Energy Information Administration. Coal Data - A
Reference. DOE/E14-0064. Washington, B.C. September,
1978. p. 6.
23. Bartok, B., Sarofim, A. F. (eds.) Fossil Fuel Combustion,
A Source Book. John Wiley & Sons, Inc. 1991. p. 239.
24. Ref. 14, p. 5-17.
25. Ref. 23, p. 230.
26. Ref. 4, p. 2-31.
27. Control Techniques for Nitrogen Oxides Emissions from
Stationary Sources - 2nd Edition. Prepared for the
U. S. Environmental Protection Agency. Publication No.
EPA-450/1-78-001. January 1978. p. 3-8.
28. Ref. 14, p. 5-20.
29. Radian Corporation. Eagle Point Cogeneration Facility
West Deptford Township, New Jersey Compliance Test Report
- Unit B, July 1992. p. 2-20.
30. Radian Corporation. Atlantic Electric Sherman Avenue
Generating Station Combustion Turbine Unit 1, Vineland,
New Jersey Compliance Test Report. Appendix A, pp. 2
through 5. July 1991.
31. Telecon. Rosa, J., Tenneco Gas, with Quincey, K. Radian
Corporation. September 29, 1992. Typical nitrogen
content of pipeline quality natural gas-ACT document on
NOX emissions.
32. Glassman, I., Combustion, 2nd ed., Academic Press,
Orlando, Florida (1987). p. 20.
2-57
-------
33. Ref. 23, p. 231.
34. Ref. 32, pp. 330 through 337.
35. Ref. 32, p. 331.
36. Ref. 32, pp. 333 through 334.
37. Ref. 4, p. 4-34.
*
38. Ref. 4, p. 4-35.
39. Ref. 4, p. 4-34.
40. Ref. 10, p. 4-3.
41. Ref. 23, pp. 230 through 231.
42. Letter and attachments from Smith J. R., Houston Lighting
& Power, to Neuffer, W. , U. S. Environmental Protection
Agency. December 15, 1992. Discussion of NOX RACT.
43. New Source Performance Standard, Subpart Da - Technical
Support for Proposed Revisions to NOX Standard (TSD).
EPA-453/R-94-012. U.S. Environmental Protection Agency,
Research Triangle Park, NC. October 1995. p. 4-6.
44. Permit issued by the State of Iowa Department of Natural
Resources to the Archer Daniels Midland Company.
August 31, 1993. DNR Permit # 93-A-324.
2-58
-------
I
3.0 NITROGEN OXIDES EMISSION CONTROL TECHNIQUES
Chapter 3.0 describes the control technologies available
for reducing nitrogen oxide (NOX) emissions from new fossil-
fuel-fired industrial boilers. In addition, the factors
affecting the performance of these control technologies and
the demonstrated performance levels are discussed.
All of the control methods can be grouped into one of two
fundamentally different techniques—combustion controls (CC)
and flue gas treatment controls. Combustion controls reduce
NOX emissions by suppressing NOX formation during the
combustion process while flue gas treatment controls reduce
NOX after formation.
Combustion controls are the most widely used method for
controlling NOX formation in industrial boilers. Most new
coal-fired industrial boilers are expected to use some form of
CC to reduce NOX formation. Although flue gas treatment
methods can often achieve greater NOX control than CC, they
have not been applied to many industrial boilers in the United
States. The types of NOX controls currently demonstrated or
applicable for new fossil-fuel-fired industrial boilers are
presented in table 3-1.
This chapter describes each of these NOX control
technologies. Section 3.1 describes the CC techniques for new
coal-fired industrial boilers. Section 3.2 presents the
demonstrated performance of these control techniques.
Sections 3.3 and 3.4 present process descriptions and
performance levels respectively of CC applied to oil- and
gas-fired boilers. Sections 3.5 and 3.6 discuss the design
and performance of flue gas treatment techniques for all three
fuel types (coal, oil, and gas).
3-1
-------
TABLE 3-1. NOX EMISSION CONTROL TECHNOLOGIES
FOR NEW FOSSIL FUEL INDUSTRIAL BOILERS
control options
Fuel
applicabilit
Combustion control techniques
Low NOX burners for
conventional boilers
Low NOX burners + overfire
air for conventional boilers
Air staging for fluidized
bed combustion boilers
Air staging for spreader
stoker boilers
Flue gas recirculation
Flue gas treatment techniques
Selective noncatalytic
reduction
Selective catalytic
reduction
Coal, natural
gas, oil
Coal, natural
gas, oil
Coal
Coal
Natural gas, oil
Coal, natural
gas, oil
Coal, natural
gas, oil
3-2
-------
3.1 COMBUSTION CONTROLS FOR COAL-FIRED INDUSTRIAL BOILERS
Combustion control techniques include low NOX burners
(LNB) and LNB with overfire air (OFA) for conventional
boilers, and air staging for fluidized bed combustion (FBC)
and spreader stoker (SS) boilers. Each of these control
technologies is discussed in the following sections.
3.1.1 Low NOv Burners for Pulverized Coal-Fired Boilers
** £ 1
3.1.1.1 Process Description. Low NOX burners have been
developed by boiler and burner manufacturers for new
applications, and are applicable to both wall- and
tangentially-fired boilers.* Low NOX burners limit NOX
formation by controlling the stoichiometric and temperature
profiles of the combustion process. This control is achieved
with design features that regulate the aerodynamic
distribution and mixing of the fuel and air, thereby yielding
one or more of the following conditions:
1. Reduced oxygen (©2) in the primary flame zone, which
limits fuel NOX formation;
2. Reduced flame temperature, which limits thermal NOX
formation; and
3. Reduced residence time at peak temperature, which
limits thermal NOX formation.
Low NOX burner designs can be divided into two general
categories: "delayed combustion" and "internal staged".
Delayed combustion LNB are designed to decrease flame
turbulence (thus delaying fuel/air mixing) in the primary
combustion zone, thereby establishing a fuel-rich condition in
the initial stages of combustion. This design departs from
traditional burner designs, which promote rapid combustion in
turbulent, high-intensity flames. The longer, less intense
flames produced with delayed combustion LNB inhibit thermal
NOX generation because of lower flame temperatures.
"The firing system in tangentially-fired boilers consists
, of coal and air nozzles and will, be referred to as low NOX
( burners in this document.
3-3
-------
Furthermore, the decreased availability of oxygen in the
primary combustion zone inhibits fuel NOX formation.
Internally staged LNB are designed to create stratified
fuel-rich and fuel-lean conditions in or near the burner. In
the fuel-rich regions, combustion occurs under reducing
conditions, promoting the conversion of fuel nitrogen to
nitrogen (N£) and inhibiting fuel NOX formation. In the fuel-
lean regions, combustion is completed at lower temperatures,
thus inhibiting thermal NOX formation.
3.1.1.1.1 Wall-fired coal boilers. A number of
different LNB designs have been developed by burner
manufacturers for use with wall-fired boilers firing coal.
Several of these burner designs are discussed below.
The Controlled Flow/Split Flame (CF/SF) burner shown in
figure 3-la is an internally-staged design that stages the
secondary air and primary air and fuel flow within the burner
throat.1 The burner name is derived from its operating
characteristics: 1) controlled flow is achieved by the dual
register design, which provides for the control of the inner
and outer air swirl, and allows independent control of the
quantity of secondary air to each burner, and 2) the split-
flame is accomplished in the coal injection nozzle, which
segregates the coal into four concentrated streams. The
result is that volatiles in the coal are released and burned
under more reducing conditions than would otherwise occur
without the split flame nozzle. Combustion under these
conditions converts the nitrogen species contained in the fuel
volatiles to N2, thus reducing NOX formation.
The Internal Fuel Staged™ (IFS) burner, shown in
figure 3-lb, is similar to the CF/SF burner and is also for
coal-fired boilers.1 The two designs are nearly identical,
except that the split-flame nozzle has been replaced by the
IFS nozzle, which generates a coaxial flame. The two designs
are best suited for large industrial boilers with burners of
at least 40 MMBtu/hr.
3-4
-------
id
•P-
to •
u
H 0)
<0 X
Eg
c o
H iH
ft
r4
I
01
14
4J »-l
^H 0>
•H C
x
•
0)
oE
^ «»
O -t
U «M
H
I
O
0)
3-5
-------
The Dual Register Burner - Axial Control Flow™ (DRB-XCL)
wall-fired LNB operates on the principle of delayed
combustion. The burner diverts air from the central core of
the flame and reduces local stoichiometry during coal
devolatization to minimize initial NOX formation. The DRB-XCL
is designed for use without compartmented windboxes, and the
flame shape can be tuned to fit the furnace by use pf
impellers. As shown in figure 3-2, the burner is equipped
with fixed spin vanes in the outer air zone that move
secondary air to the periphery of the burner.2 Also,
adjustable spin vanes are located in the outer- and inner-air
zones of the burner. The inner spin vane adjusts the shape of
the flame, which is typically long. The outer spin vane
imparts swirl to the flame pattern. The flame stabilizing
ring at the exit of the coal nozzle enhances turbulence and
promotes rapid devolatization of the fuel. An air-flow
measuring device located in the air sleeve of each burner
provides a relative indication of air flow through each burner
and is used to detect burner-to-burner flow imbalances within
the windbox.
The RO-II burner consists of a single air inlet, dual-
zone air register, tangential inlet coal nozzle, and a flame-
stabilizing nozzle tip. Figure 3-3 shows the key components
of the burner. Combustion air is admitted to both zones of
the air register and the tangential inlet produces a swirling
action. The swirling air produces a "forced vortex" air flow
pattern around the coal jet. This pattern creates local
staging of combustion by controlling the coal/air mixing, thus
reducing NOX formation.
The Controlled Combustion Venturi™ (CCV) burner for wall-
fired boilers is shown in figure 3-4.* NOX control is
achieved through the venturi coal nozzle and low swirl coal
spreader located in the center of the burner. The venturi
nozzle concentrates the fuel and air in the center of the coal
nozzle, creating a very fuel-rich mixture. As this mixture
3-6
-------
Pilot Fixed
Manifold Vanes
Conical
Diffuser
Adjustable
Vanes
Spin Vanes
Figure 3-2. Dual register-axial control flow low NOX burner.
3-7
-------
g
m
o
u
o
m
2
•H
3-8
-------
0)
g
tf
o
•H
W
3
.u
C
s
c
o
-H
OJ
10
3
O
O
2
4J
d
o
u
ro
0
•H
Cn
3-9
-------
passes over the coal spreader, the blades divide the coal
stream into four distinct streams, which then enter the
furnace in a helical pattern. Secondary air is introduced to
the furnace through the air register and burner barrel. The
coal is devolatized at the burner exit in a fuel-rich primary
combustion zone, resulting in lower fuel NOX conversion. Peak
flame temperature is also lowered, thus suppressing,, the
thermal NOX formation.*
3.1.1.1.2 Tanaentiallv-fired coal boilers. As discussed
in chapter 2 and shown in figure 3-5a, the traditional burner
arrangement for tangentially-fired coal boilers consists of
corner-mounted vertical burner assemblies from which fuel and
air are injected into the furnace.5 The fuel and air nozzles
are directed tangentially to an imaginary circle in the center
of the furnace, generating a rotating fireball in the center
of the boiler as shown in figure 3-5b . Each corner has its
own windbox that supplies primary and secondary air through
the air compartments located above and below each fuel
compartment.
The low NOX concentric firing technique for
tangentially-fired boilers is shown in figure 3-6a5. This
technique changes the secondary air flow through the windbox;
however, the primary air is not affected. A portion of the
secondary air is directed away from the fireball and toward
the furnace wall as shown in figure 3-6b . In addition,
"flame attachment" nozzle tips that accelerate the
devolitization of the coal are used. This configuration
suppresses NOX emissions by providing an O2~rich environment
along the furnace walls. This can also reduce the slagging
and tube corrosion problems often associated with combustion
slagging. Several systems are available that use the
concentric firing technique in combination with OFA. These
systems are classified as a family of technologies called the
Low NOX Concentric Firing System™ (LNCFS) and are discussed in
section 3.1.2.
3-10
-------
14
0»
ri
-&
T3 flj
U-O
iH -H Q)
•W AJ «M
«w
2
(0
0)
I
•rH
&4
«5 W .
e c a)
O l-l r-l
U O-H
M
•H L
nJ C ^j
O ffi
C W 'f4
R) O
H « tn
«B
in
ro
g.
•H
3-11
-------
<
"55
u.
+
^
<
^
co
E
"w
9^ *
CO
0
O
"<
£•
n
T3
O
O
O>
^
_co
'x
5
V.
^\
§
4J
CO
X>
O co
s c
O-H
K-I G
0)
C O
to d
rH O
Oi U
X)
vc
ro
u -H
171 m
c
•H JJ
U C
•H 0)
«w e
Q)
en
-H -O U
ecu
QJ m nj
u
c
O 0)
-" 2
«w e
X 4J
o e u
S Q) (0
*££
o >,o
^ to u
*J
C
w«
^£
o
(0
«
-
•-
JO
U- <— .
4^
I4i
' \
\
CO
o
O
w ®
<2 j^
r<
1o ±: g
5 <|
CO
o
O
Q.
f3
1O1
8
o
.j=f
S £
-------
The Pollution Minimum (PM) burner has also been developed
for tangentially-fired boilers. As shown in figure 3-7, the
PM burner system for coal-fired boilers uses a coal separator
that aerodynamically divides the primary air and coal into two
streams, one fuel-rich and the other fuel-lean.6 Thus, NOX
emissions are reduced through controlling the local
stoichiometry in the near-burner zone. The PM burner may also
be used with OFA systems.
3.1.1.2 Factors Affecting Performance. The
effectiveness of LNB depends on a number of parameters. Low
NOX burners are generally larger than conventional burners and
require more precise control of fuel/air distribution. The
performance of delayed combustion LNB depends partially on
increasing the size of the combustion zone to accommodate the
longer flames. Because of this, boilers equipped with these
LNB typically have larger furnace volumes than those equipped
with conventional burners. Flame impingement on furnace walls
or superheater tubes can be minimized or eliminated by
adjusting burner tilt, coal/primary air velocity, secondary
air velocity, biased burner firing, and ensuring that
superheater tubes are not located in the path of the flames.
The fuel-rich operating conditions of LNB generate
localized reducing conditions in the lower furnace region and
can increase the slagging tendency of coal. To reduce this
potential, some combustion air can be diverted from the burner
and passed over the furnace wall surfaces, providing a
boundary air layer that maintains an oxidizing atmosphere
close to the tube walls. Some LNB operate with a higher
pressure drop or may require slightly higher excess air levels
in the furnace at full load to ensure good carbon burnout,
thus increasing fan requirements.
3.1.2 Low NOX Burners and Overfire Air for Conventional
Boilers'
3.1.2.1 Process Description. Low NOX burners and OFA
are complementary combustion modifications for NOX control
that incorporate both the localized staging process inherent
3-13
-------
0)
•H
c
O
•H
I
O
0)
I
•H
3-14
-------
in LNB designs and the bulk-furnace air staging of OFA. Low
NOX burners are described in section 3.1.1.
Overfire air is a combustion control technique whereby a
percentage of the total combustion air is diverted from the
burners and injected through ports above the top burner level.
The total amount of combustion air fed to the furnace remains
unchanged. In the typical boiler shown in figure 3»8a, all
the air and fuel are introduced into the furnace through the
burners, which form the main combustion zone.5 For a boiler
equipped with an OFA system, such as in figure 3-8b,
approximately 5 to 20 percent of the combustion air is
injected above the main combustion zone. Since OFA
introduces combustion air at two different levels in the
furnace, this combustion hardware is also called air staging.
Use of OFA reduces the amount of air in the burner zone
to below that required for complete combustion and delays
burn-out of the fuel-rich combustion gases. This decreases
the overall rate of combustion and results in a less intense,
cooler flame, which suppresses the formation of thermal NOX.
3.1.2.1.1 Wall-fired boilers. In wall-fired boilers,
LNB can be coupled with either conventional OFA or advanced
OFA (AOFA). Conventional OFA systems such as in figure 3-9a,
use a single windbox to supply air to the burners and OFA
ports.5 Because air flow to the OFA ports is taken from the
same windbox, the ability to control air flow distribution may
be limited.
Advanced OFA systems have separate windboxes and ducting,
and the OFA ports can be placed to achieve optimum air mixing
with the fuel-rich combustion products. The AOFA systems, as
shown in figure 3-9b, usually inject more air at greater
velocities than conventional OFA systems, giving improved
penetration of air across the furnace width and greater NOX
reduction.5
3.1.2.1.2 Tangentially-fired boilers. In tangentially-
fired boilers, OFA is incorporated into the LNB design,
3-15
-------
o<
^ O
in u
\ / 1
/
,
\
1
•H
(0
0)
•C -H
O
I O
llg
IIfl
o o
00
.o
= "1
A)
Q) ^
W 0)
O rH
£X*H
a o
o u
ll *
\
/
°
in
CO
I
n
0)
fc
&
•H
ft,
. "O 03
• rou.
8
ffi
: -" v
T
3-16
'D
C
.fc'D (D
3
0)
•O rH
Q) -H
W O
O .Q
(X
D.TJ
O Q)
b
rH-H
10
-------
0)
c o
O.Q
(0
0)
k
•rt
-O
-------
forming a LNB and OFA system. There are three possible
arrangements as shown in figure 3-10.7 For LNCFS Level 1™, )
OFA is integrated directly into the existing windbox by
exchanging the top coal nozzle with the air nozzle directly
below it. This OFA arrangement is referred to as close-
coupled OFA (CCOFA). In LNCFS Level 2™, OFA is supplied by a
windbox which is separated from the main windbox and located
at a higher elevation in the furnace. This OFA arrangement is
referred to as separated OFA (SOFA). The quantity and
velocity of the air injected into the furnace through a SOFA
windbox can be higher than those levels possible with a CCOFA
windbox, thus providing better mixing. LNCFS Level 3™ injects
OFA through both windboxes for maximum control and flexibility
of the staging process.
3.1.2.2 Factors Affecting Performance. The design and
operational factors affecting the NOX emission control
performance of LNB are discussed in section 3.1.1.2. These
factors also apply when LNB and OFA are combined. For OFA "x
systems, the number, size, and location of the OFA ports as )
well as the OFA jet velocity must be adequate to ensure
complete combustion. Improper design could lead to an
increase of incomplete combustion products (unburned carbon,
CO, and organic compounds), tube corrosion, and lower and
upper furnace ash deposits (slagging and fouling).
To have effective NOX reduction, there must be adequate
separation between the top burner row and the OFA ports.
However, efficient boiler operation requires maximizing the
residence time available for carbon burnout between the OFA
ports and the furnace exit, which means locating the OFA ports
as close to the burners as practical. These conflicting
requirements must be considered when designing OFA systems.
Increasing the amount of OFA can reduce NOX emissions;
however, this means that less air (oxygen) is available in the
primary combustion zone. The resulting reducing atmosphere in
the lower furnace can lead to increased slagging and tube >.
3-18
-------
CCOFA
CCOFA
coal
coal
offset air
coal
offset air
coal
oil
coal
air
LNCFS
Level 1
SOFA
coal
offset air
coal
offset air
coal
offset air
coal
oil
coal
air
LNCFS
Level 2
SOFA
CCOFA
CCOFA
coal
coal
offset air
coal
offset air
coal
oil
coal
air
LNCFS
Level 3
Figure 3-10. Low NOX concentric firing systems.
3-19
-------
corrosion and can change furnace heat release rates, the exit
temperature of the flue gas, and steam generating efficiency.
3.1.3 Air Staging for Fluidized Bed Combustion Boilers
3.1.3.1 Process Description. Air staging is widely used
to control NOX emissions from FBC boilers. In an FBC boiler
that is not equipped with air staging, all the combustion air
is introduced through primary air orifices located below the
bed. With air staging, a portion of the total combustion air
is introduced into the combustor through secondary air ports
located along the freeboard section of the combustor. Usually
about 40 percent of the total combustion air is introduced as
secondary air. The total amount of air to the combustor
remains unchanged.
The lower oxygen concentration in the dense portion of
the bed suppresses the conversion of fuel nitrogen to NO.
This also promotes the formation of CO and H2 which
catalytically reduces any NO formed to N2 in the presence of
CaO and char catalysts present in the bed.
3.1.3.2 Factors Affecting Performance. The
effectiveness of air staging depends on the degree of staging
applied, the overall oxygen level, and the composition of the
c
bed material. NOX formation decreases as the degree of
staging increases; however, this can lead to increases in CO
emissions and unburned carbon losses. Increasing the
residence time between the primary and secondary air injection
levels also inhibits NOX formation. However, there are
practical limits on how high the secondary air can be
introduced in the freeboard area without affecting combustion
efficiency, corrosion, and steam temperature control.
3.1.4 Air Staging for Spreader Stokers
3.1.4.1 Process Description. Air staging is a common
method of NOX control for spreader stokers. Stoker units
naturally operate with a form of combustion staging due to the
design of the boiler. Volatiles from the fuel bed are driven
off and burned above the bed level. The coal solids are then
3-20
-------
burned on a bed with lower combustion temperatures. With air
staging, more air is introduced through overfire air (OFA)
ports above the fuel bed while reducing the undergrate
airflow. The total amount of combustion air is unchanged.9
3.1.4.2 Factors Affecting Performance. The
effectiveness of air staging depends on the location of the
OFA ports, the degree of staging applied, and the overall
oxygen level. NOX levels decrease as the height of the OFA
ports over the fuel bed decreases. There is a limit to the
degree of air staging that can be applied to stoker boilers.
In many cases, the only method of cooling the grate is the
flow of combustion air underneath it. If this flow is
decreased too much during air staging, the grate could
overheat. Reduced undergrate air flow can also cause the
formation of clinkers. For coals with low ash fusion
temperatures, significant clinker formation can be caused by
the excessively high bed temperatures resulting from
combustion with insufficient amounts of excess air. There is
also the possibility of creating local reducing zones with low
oxygen which may form harmful corrosion products.8
3.2 PERFORMANCE OF COMBUSTION CONTROLS ON COAL-FIRED BOILERS
Nitrogen oxide emissions data for coal-fired boilers were
obtained from various publications and boiler vendors. Only
those boilers larger than 100 MMBtu/hr heat input capacity,
built or retrofitted after June 1984 (promulgation date of the
current NSPS subpart Db), and had CC for NOX control were
included. This data reflects the performance of current CC
technologies on representative boilers of the subpart Db
population. Table 3-2 presents plant names, boiler status
(new or retrofit), rated boiler capacity (MMBtu/hr), boiler
type, coal type, NOX control, test load, test period, and
measured NOX emission levels (Ib/MMBtu). The data are sorted
by boiler type.
For the seven pulverized coal (PC)-fired boilers shown in
table 3-2, the NOX emissions range from 0.25 to 0.59 Ib/MMBtu.
3-21
-------
OQ
S
o »-»-«-»- o o
J» -»-*•*
Kl M Kl K>
ic
«^
81
5 s a
K K S o ? o ft g 5 8 8 8 S 2 * fe 5535 5
o o d )Q °, jjj o ooooooooo oooo o
G> O O
S S S Si
5 S g g
Illl
S,000
jD^
w"S
§1
UUUUUUUUU
cDcommoameacnm
t; b S £ & J5 S 15 £
v>
»
* o3
2 as
o
O fl
I &S ^ § 2 v g
3o.9au.3u.~
K
1
I
I
+*
•
1
i!
*< c
— «
o «
u
I
i H
Is
o »
s^ I
« -
• JJ u TJ
3-22
-------
Boiler capacity ranges from 140 to 1,176 MMBtu/hr. six of the
seven boilers are retrofit,'and one is new. NOX emissions
range from 0.30 to 0.59 Ib/MMBtu for boilers equipped with
LNB. For boilers equipped with LNB+OFA, NOX emissions are
lower and range between 0.25-0.49 Ib/MMBtu.
Nine FBC boilers are presented in table 3-2. Boiler
capacities range from 130 to 500 MMBtu/hr. Eight of these are
circulating fluidized bed combustion (CFBC) boilers, and one
is a bubbling fluidized bed combustion (BFBC) boiler. NOX
emissions range from 0.14 to 0.50 Ib/MMBtu for the CFBC
boilers. For the BFBC boiler, NOX emissions were measured at
0.17 Ib/MMBtu.
Test data for five spreader stokers are presented in
table 3-2. Boiler capacity ranges from 134 to 325 MMBtu/hr.
Nitrogen oxide emissions for all of the boilers are
approximately 0.50 Ib/MMBtu based on a 30-day rolling average.
3.3 COMBUSTION CONTROLS FOR NATURAL GAS- AND OIL-FIRED
INDUSTRIAL BOILERS
The combustion control techniques applicable to new
natural gas- and oil-fired boilers include LNB and flue gas
recirculation (FGR). Additionally, these controls can be
combined with OFA. These control technologies are discussed
in the following sections.
3.3.1 Flue Gas Recirculation
3.3.1.1 Process Description. Flue gas recirculation is
a flame-quenching strategy in which the recirculated flue gas
acts as a thermal diluent to reduce combustion temperatures.
It also reduces excess air requirements, thereby reducing the
concentration of oxygen in the combustion zone. As shown in
figure 3-11, FGR involves extracting a portion of the flue gas
from the economizer or air heater outlet and readmitting it to
the furnace through the furnace hopper, the burner windbox, or
both. To reduce NOX, the flue gas is injected into the
windbox.15 The degree of FGR varies between 10-20 percent of
the combustion air. Windbox FGR is primarily effective at
3-23
-------
9
*J
(0
>i
Ul
o
•H
4J
(0
rH
fl
O
o
0)
w
(0
Q)
3-24
-------
reducing thermal NOX and is not used for NOX control on coal-
fired boilers in which fuel NOX is a major contributor.
3.3.1.2 Factors Affecting Performance. Since FGR is
used to suppress temperature within the flame, the
effectiveness of this technique depends on the burner heat
release rate and the type of fuel being burned. When burning
heavier fuel oils, less NOX reduction is expected than when
burning natural gas because of the higher nitrogen content of
the fuel.
3.3.2 Low NO^ Burners
3.3.2.1 Process Description. The fundamental NOX
reduction mechanisms in natural gas- and oil-fired LNB are
essentially the same as in coal fired LNB discussed in
section 3.1.1.1. However, many vendors of LNB for natural
gas- and oil-fired boilers incorporate FGR as an integral part
of the LNB.
3.3.2.1.1 Wall-fired boilers. As with pulverized
coal-fired LNB, there are a number of different oil- and
natural gas-fired LNB available from manufacturers. Several
of these are discussed below.
The wall-fired ROPM™ burner for natural gas- or oil-
firing is shown in figure 3-12.16 Combustion in a ROPM™ burner
is internally staged, and takes place in two different zones;
one under fuel-rich conditions and the other under fuel-lean
conditions. Gaseous fuel burns under pre-mixed conditions in
both the fuel-lean and fuel-rich zones. However, to maintain a
stable flame with liquid fuels, burning occurs under diffused-
flame conditions in the fuel-rich mixture.
The natural gas-fired ROPM™ burner generates a fuel-rich
flame zone surrounded by a fuel-lean zone. The burner
register is divided into two sections. Natural gas and
combustion air supplied via an internal cylindrical
compartment produces the fuel-rich flame. The fuel and air
supplied via the surrounding annular passage produces the
fuel-lean zone.16
3-25
-------
•II
„*!
IN
Jil
ils
(0
M
a)
H
•H
•0
0)
C
I
t>J
H
I
O
I
3-26
-------
The oil-fired ROPM™ atomizer sprays fuel at two different
spray angles, creating two concentric hollow cones. The inner
cone creates a fuel-rich flame zone; the outer cone forms the
fuel-lean flame zone. The inner fuel-rich flame zone has
diffusion flame characteristics that help maintain overall
flame stability. The ROPM™ technology also uses FGR to
achieve NOX reductions. «.
The Internal Staged Combustion™ (ISC) wall-fired LNB
incorporates low excess air in the primary combustion zone,
which limits the oxygen available to combine with fuel
nitrogen. In the second combustion stage, additional air is
added downstream to form a cooler, oxygen-rich zone where
combustion is completed and thermal NOX formation is limited.
The ISC design, shown in figure 3-13 can fire natural gas or
• i 17
oil.
TM
In addition to the XCL burner for coal-fired boilers,
the XCL™ as shown in figure 3-14 is also available for wall-
fired boilers burning natural gas and oil.2 This design
enables the use of an open windbox (compartmenta1 windbox is
unnecessary). Air flow is controlled by a sliding air damper
and swirled by vanes in the dual air zones.
The Swirl Tertiary Separation (STS) burner for natural
1 fi
gas- and oil-fired boilers is shown in figure 3-15. In this
design, the internal staging of primary and secondary air can
be adjusted depending on required NOX control and overall
combustion performance. The ability to independently control
swirl of the primary and secondary air streams provides
flexibility in controlling flame length and shape, and ensures
flame stability under low-NOx firing conditions. A separate
recirculated flue gas stream forms a distinct layer between
the primary and secondary air. This layer of inert flue gas
delays the combustion process and reduces peak flame
temperatures and oxygen concentrations in the primary
combustion zone, thus controlling both thermal and fuel NOX.18
3-27
-------
COOLER OXYGEN RICH ZONE
REDUCES THERMAL NOx
PRIMARY
COMBUS
AIR (STAGING AIR I GAS SPUDS
Figure 3-13. Internal staged combustion™ low NOX burner.17
3-28
-------
I
I
X
o
I
•o
0)
-H
O
•g
U
«0
4J
(0
c
X
I
<*»
0)
3-29
-------
0)
X
i
o
•H
(0
&
0)
to
&
to
0)
EH
to
X
o
2;
I
in
r-l
n
0)
b
3
CT
•H
3-30
-------
The wall-fired RMB™ (Rapid Mix Burner) for natural gas-
or oil-firing is shown in figure 3-16.™ The RMB™ burner mixes
gaseous fuel and combustion air near the burner exit. Flame
temperatures are lowered by mixing flue gas recirculation with
combustion air upstream of the burner. This flue gas
recirculation/combustion air mixture creates a strong internal
recirculation zone after passing through the axial swirl
vanes. The flame here is compact and stable. Combustion
« Q
gases from the furnace are drawn into the recirculation zone.
The gas-fired RMB™ injects the gas through hollow bases
drilled through the axial swirl vanes, which are attached to
the gas reservoir. This produces rapid mixing at the burner's
exit.18
The oil-fired RMB™ burner has a standard atomizer gun
assembly down the center of the burner. The fuel is rapidly
mixed at the end of the burner by the advanced atomizer tip
« A
and the internal flue gas recirculation.
3.3.2.1.2 Tanaentially-fired boilers. The
tangentially-fired Pollution Minimum™ (PM) burner is shown in
figure 3-17.20 The burners are available for natural gas or
oil firing. Both designs are internally staged and
incorporate FGR within the burners.
The gas-fired PM burner compartment consists of two fuel-
lean nozzles separated by one fuel-rich nozzle. Termed "GMW
(gas mixing), this LNB system incorporates FGR by mixing a
portion of the flue gas with combustion air upstream of the
burner. When necessary, FGR nozzles are installed between two
adjacent PM burner compartments.20
The oil-fired PM burner consists of one fuel nozzle
surrounded by two separated gas recirculation (SGR) nozzles.
Within each fuel compartment a single oil gun with a unique
atomizer sprays fuel at two different spray angles. The outer
fuel spray passes through the SGR streams produce the fuel-
lean zones. The inner concentric spray produces the fuel-rich
zones. The SGR creates a boundary between the rich and lean
3-31
-------
W
u
-------
00
I
o>
o
•s
m
u
W
0
•8
* Fu«IGa«(C)
V'
\\
\ \
•V •
'^—
(M) ««O I»«U -» c^
q
lu
•sf
-\ !
if
Li'
4
I
. 1
L
I,
U
o
u
r\ I'lCVJ ~
MTU
1 'I'1 !" J
I ! I 1
m
i
o»
i
0)
c
u
2
I
O
-H
•M
O
Oi
r*
tH
I
0)
3-33
-------
flame zones, thereby maintaining the NOX reducing
characteristics of both flames.20
3.3.2.2 Factors Affecting Performance. The factors
affecting the performance of oil- and gas-fired LNB are
essentially the same as those for coal-fired LNB discussed in
section 3.1.1.2 of this document. However, the overall
success of NOX reduction with LNB may also be influenced by
fuel grade and boiler design. For example, the most
successful NOX reductions are on natural gas and light fuel
oil firing in boilers initially designed to burn these fuels.
Also, boilers designed with larger furnace volumes per unit
output are more conducive to NOX reduction with LNB than
boilers designed with a smaller furnace.
3.3.3 Combinations of Combustion Controls
3.3.3.1 Process Description. Large NOX reductions can
be obtained by combining combustion controls such as FGR, LNB,
and OFA. The types of combinations depend upon the furnace
design and fuel type. The process descriptions for FGR and
LNB are in sections 3.3.1.1 and 3.1.1.1, respectively. The
process description for OFA applied to coal-fired boilers in
section 3.1.2.1 is also applicable for natural gas- and
oil-fired boilers.
3.3.3.2 Factors Affecting Performance. The same basic
factors affecting the performance of individual combustion
controls apply to these controls when they are used in
combination. These are described in sections 3.3.1.2, 3.1.1.2
and 3.1.2.2 for FGR, LNB, and OFA, respectively.
3.4 PERFORMANCE OF COMBUSTION CONTROLS FOR NATURAL GAS- AND
OIL-FIRED BOILERS
As with coal-fired boilers, NOX emissions data for oil
and natural gas-fired boilers were obtained from various
publications. Only data for boilers greater than 100 MMBtu/hr
in capacity, built or retrofitted after June 1984 (the
promulgation date of the current NSPS subpart Db), and had CC
for NOX control were included. This data reflects the
performance of current combustion control technologies on
3-34
-------
representative boilers of the subpart Db population.
Table 3-3 presents plant names, boiler status (new or
retrofit), rated boiler capacity (MMBtu/hr), boiler type, fuel
type, NOX control, tested load, test period, and measured NOX
emission levels (Ib/MMBtu). The data are first sorted by fuel
type (oil or natural gas) and then by boiler type (field-
erected or packaged) and NOX control technique.
Data on only two oil-fired boilers (one residual and one
distillate oil-fired) were available and are presented in
table 3-3; both boilers are packaged units. The residual oil-*
fired boiler emits 0.22 Ib/MMBtu of NOX. At 63 percent load,
the distillate oil-fired boiler emits 0.05 Ib/MMBtu.
There are 16 natural gas-fired boilers presented in
table 3-3. Of those, 3 are field-erected, and 13 are packaged
units. Boiler capacities range from 220 to 380 MMBtu/hr for
the field-erected units and from 100 to 205 MMBtu/hr for the
packaged units. For field-erected boilers with LNB, the NOX
emissions range from 0.10 to 0.17 Ib/MMBtu. Nitrogen oxide
emissions from one field-erected boiler with LNB, FGR, and OFA
were measured at 0.14 Ib/MMBtu. For the two packaged gas-
fired boilers equipped with LNB, NOX emissions range from
0.03 to 0.11 Ib/MMBtu. For the two packaged gas-fired boilers
equipped with FGR, NOX emissions were measured at
0.06 Ib/MMBtu. The NOX emissions from eight packaged boilers
equipped with LNB and FGR range from 0.01 to 0.09 Ib/MMBtu.
The last packaged boiler shown in table 3-3 uses LNB and OFA
to control NOX emissions to 0.08 Ib/MMBtu.
Data showing the effect of load on NOX emissions are
presented for two packaged natural gas-fired boilers in
table 3-3. The United Airlines boiler with LNB was tested at
loads ranging from 24 to 100 percent of rated capacity and the
NOX emissions measured 0.03 Ib/MMBtu. The data indicates that
NOX emissions are unaffected by changes in load. Similar
results were obtained on the Morningstar boiler with LNB and
FGR. The NOX emissions remained constant at 0.01 Ib/MMBtu
when load varied between 15-100 percent of rated capacity.
3-35
-------
P
H
O
0
H
g
to
H
Q
M
CJ
fa
L GAS-
ITROLS
»*j g
go
81 H
^ m
H °
O ju
° s
1
W
<2
tj|
Q
•
m
i
1
TABLE
g
•*.
ae
— 3
31
~* *
sTS
M —
41 t-
|
"gS
*•*
1
• u
4->
§
U
0*
It
M. «V
-V
m **
xC
** £
P
D-OJ
31
i!
*i
41 *^
Of
a
~ o
o o
ae ae
\ 3
S g
? *
5 5
o o
oc o
£c u
0. 0.
O 0
in o
oc ee
«
fc 6
'o w
i|I
.J: 8- c.
u. o o
i * §
0X0
»- »- •- •- ooooooooo o o o o o o o o o oooooooooo o
O O 0 O OOOOOOOOO O O 0 0 O 0 O 0 0 OOOOOOOOOO 0
t
SliafSSzISi^SSSSiSiaiiafSSilSiiSilSliilSliSe^
z z z z |4IO» MtlU > (O C
"- S * "• M "S 1 -8 « * c
iT "E -o c c obv&ou x u
i B - C Ci — u. U U P> • O « 4J
»> a — e o) — S 5 . u. . . c "-
a — — o i- B> x — • u M « "" -o * 5 — _,CI"^8 '
"a o t> ti i- T) ^»ToTo5<°:5ii.eB»>t
1- «J W — > a 41 .^ > l_ — 41 «3 — O) •- O
D 0 i Q. > *J C4lU«)4^3DlCC ^
4->ONEi.». e— cV'-coicoi- — '
0 oT 3 ^ i c *Ji-5Eo5omi-o i
"S
£
L.
^
H
9 u-
O> O
AB «H
H O
g'S
^r c
*o S
. 41 W
!-*••«
« 4»
"o *5 5
^ £
« N
^g je*
a L*
IAJ •» C •
fc-!*
• » O Xl **
Jjfl
E.i!
• J3 O TB
3-36
-------
Carbon monoxide emissions remained constant at below 1 ppm at
all loads.
3.5 FLUE GAS TREATMENT CONTROLS FOR COAL-, NATURAL GAS- AND
OIL-FIRED BOILERS
Two commercially available flue gas treatment
technologies for reducing NOX emissions from fossil fuel
industrial boilers are selective noncatalytic reduction (SNCR)
and selective catalytic reduction (SCR). SNCR involves
injecting ammonia (NH3) or urea into the flue gas of an
industrial boiler to yield nitrogen (N2) and water. The NHs
or urea must be injected into specific temperature zones in
the upper furnace or convective pass of the boiler for this
method to be effective.23 The other commercially available
flue gas treatment method, SCR, involves injecting NH3 into
the flue gas upstream of a catalyst. The catalyst promotes
reactions that converts NOX to N2 and water at lower
temperatures than required for SNCR.
Selective catalytic reduction technologies have been
applied to commercial-scale natural gas-fired boilers,
primarily in California. SNCR has been applied to a number of
industrial coal-fired boilers in the United States.
3.5.1 Selective Noncatalytic Reduction
3.5.1.1 Process Description. The SNCR process involves
injecting a nitrogen-bearing chemical (usually NH3 or urea)
into boiler flue gas at temperatures of 870 to 1,040 °C (1,600
to 1,900 °F). The NH3 or urea reacts with NOX in the flue gas
to produce N2 and water.
As shown in figure 3-18, for the NH3-based SNCR process,
NH3 is injected into the flue gas where the temperature is 950
± 30 °C (1,750 ± 50 °F).2* The primary reaction equation is:
2NO + 2NH3 + 1/2 02 -» 2N2 + 3H2O (3-1)
Competing reactions that use some of the NHs are:
4NH3 + 502 - 4NO + 6H2O (3-2)
4NH3 + 3O2 -*• 2N2 + 6H2O (3-3)
3-37
-------
o
;g
S E
§ {
11-
<0
U)
CO
CM
N
o
•H
4J
U
0)
M
O
•e
0)
I.
CO
"5.
o.
D
CO
£
CO
I
OT
I
3-38
-------
To maximize reaction 3-1, NH3 must be injected into the
optimum temperature zone, and must be effectively mixed with
the flue gas. When the temperature exceeds the optimum range,
reaction 3-2 becomes significant, NH3 is oxidized to NOX, and
the net NOX reduction decreases.25 Reaction 3-3 represents the
oxidation of N2, which results in a net loss of NH3 for
reduction of NOX. k
If the temperature of the combustion products falls below
the SNCR operating range, the NH3 does not react and is
emitted to the atmosphere. Ammonia emissions, referred to as
slip, should be minimized because NH3 is a pollutant and can
also react with sulfur oxides in the flue gas to form ammonium
salts, which can deposit on downstream equipment such as air
heaters. A small amount of hydrogen or another hydrocarbon
reducing agent (not enough to appreciably raise the
temperature) can be injected with the NH3 to lower the
effective temperature range to approximately 705 °C
(1,300 °F).
As shown in figure 3-19 for the urea-based SNCR process,
an aqueous solution of urea [CO(NH2)2] is injected into the
flue gas at one or more locations in the convective pass.23
The urea reacts with NOX in the flue gas to form N2, water,
and carbon dioxide (003)• Aqueous urea has a maximum NOX
reduction activity at approximately 900 to, 1,150 °C (1,650
to 2,100 °F) ,26 Proprietary chemical .enhancers can be used to
broaden the temperature range in which the reaction occurs.
Using enhancers and adjusting the concentrations can expand
the effectiveness of urea to 820 to 1,150 °C (1,500 to
2,100 OF) ,23
Although the exact reaction mechanisms are complex, the
overall reaction mechanism is:
CO(NH2)2 + 2NO + 1/2O2 -» 2N2 + C02 + 2H2O (3-4)
As indicated by the above reaction, one mole of urea
reacts with two moles of NO. However, because of competing
3-39
-------
NOxOUT Process
Boiler
Figure 3-19. Urea-based selective noncatalytic reduction.
23
3-40
-------
reactions, more than stoichiometric quantities of urea must be
injected to achieve the desired level of NOX removal.23 These
competing reactions can result in emissions of nitrous oxide
(N2O), carbon monoxide (CO), and unreacted NH3.
Another version of the SNCR process uses high energy to
inject either aqueous NH3 or urea solution as shown in
figure 3-20." The solution is injected into the flue gas
using steam or air as a diluent at one or more specific
temperature zones in the convective pass. Additionally,
methanol can be added in the process to further reduce NH3
slip. This system is based on the same concept as other SNCR
systems except that the pressurized reagent mixture is
injected into the cross-flowing flue gas with high-velocity
nozzles. High-energy injection is especially applicable to
units with narrow reagent injection windows because this
?ft
system improves flue gas mixing.
Hardware requirements for SNCR processes include reagent
storage tanks, air compressors, reagent injection grids, and
an NH3 vaporizer (for NH3-based SNCR). Injection equipment,
such as a grid system or injection nozzles, are needed at one
or more locations in the convective pass. A carrier gas, such
as steam or compressed air, is used to provide sufficient
velocity through the injection nozzles to ensure thorough
mixing of the reagent and flue gas. For units that vary loads
frequently, multi-level reagent injection is used. A control
system consisting of a NOX monitor and a controller/processor
(to receive NOX and boiler data and to control the amount of
reagent injected) is also required.
3.5.1.2 Factors Affecting Performance. Six factors
influence the performance of urea- or NH3~based SNCR systems:
inlet NOX level, temperature, mixing, residence time, reagent-
to-NOx ratio, and fuel sulfur content.
The NOX reduction reactions are directly affected by
inlet NOX concentrations. Lower inlet NOX concentrations
3-41
-------
-------
reduce the reaction kinetics and hence the achievable NOX
emissions reductions.
As shown in figure 3-21, the gas temperature can greatly
27
affect NOX removal and NH3 slip. At temperatures below the
desired operating range of 900 to 1,150 °c (1,650
to 2,100 °F), the desired NOX reduction reactions begin to
diminish and unreacted NI*3 emissions (slip) increase. Above
the desired temperature range, NH3 is oxidized to NOX/
resulting in low NOX reduction efficiency and low reactant
utilization.27
The temperature in the upper furnace and convective pass,
where temperatures are optimum for SNCR, depends on boiler
load, fuel, method of firing (e.g., off-stoichiometric
firing), and extent of heat transfer surface fouling or
slagging. The flue gas temperature exiting the furnace and
entering the convective pass is typically 1,200 °C ± 110 °C
(2,200 °F ± 200 °F) at full load and 1,040 °C ± 70 °C
(1,900 °F ± 125 °F) at half load. At similar loads,
temperatures can increase by as much as 30 to 60 °C (55 to
110 °F) depending on the extent of ash deposition on heat
transfer surfaces. Due to these variations in the
temperatures, it is often necessary to inject the reagent at
different locations or levels in the upper furnace or
convective pass for effective NOX reduction.
The third factor affecting SNCR performance is mixing of
the reagent with the flue gas. The zone surrounding each
reagent injection nozzle will probably be well mixed by the
turbulence of the flue gas. Mixing in regions distant from an
injection nozzle depends on adequate reagent velocity and
momentum for penetration. Because of reduced flue gas
turbulence, stratification of the reagent and flue gas will
probably be a greater problem at low boiler loads.27
The fourth factor which affects SNCR performance is the
residence time of the injected reagent within the required
temperature window. If the residence times are too short.
3-43
-------
S
400
350-
300
'
250
200
150-1
100
50 1
1000
\ NH3 Slip
NO
Initial
NH3:NOx = 2:
1200 1400 1600
Temperature F
1800
Figure 3-21. General effects of temperature on NOX removal.27
3-44
-------
there will be insufficient time for completion of the desired
reactions between NOX and NH3. Due to residence time and
temperature constraints, small packaged watertube boilers and
boilers with varying steam loads are difficult applications.26
The fifth factor in SNCR performance is the ratio of
reagent to NOX. Figure 3-22 shows that at an ammonia-to-NOx
ratio of 1.0, NOX reductions of less than 40 percent are
29
achieved. Based on this figure, by increasing the NH3:NOX
ratio to 2:1, NOX reductions of approximately 60 percent can
be obtained. Increasing the ratio beyond 3:1 has little
effect on NOX reduction. Since NH3tNOx ratios higher than the
theoretical ratio are required to achieve desired NOX emission
reductions, a trade-off exists between NOX control and the
presence of excess NH3 in the flue gas. Higher NH3 feed rates
increase costs.
The sixth factor in SNCR performance is the sulfur
content of the fuel. Excess NH3 can react with sulfur
compounds in the flue gas to form ammonium sulfate salt
compounds that deposit on downstream equipment. As shown in
figure 3-23, depending on the concentrations of NH3 and 803 in
the flue gas, ammonium bisulfate (NH4HS04) or ammonium sulfate
28
t(NH4)2S04] will form at temperatures below 260 °C (500 °F).
For example, ammonium bisulfate and ammonium sulfate can plug
and corrode air heaters which typically operate at
temperatures of less than 260 °C (500 °F). Because natural
gas and oil do not contain as much sulfur as coal, the fuel
sulfur content may not be as much a factor for natural gas-
and oil-fired boilers as it is for coal-fired boilers.
Most SNCR experience has been on boilers less than
2,000 MMBtu/hr in size. In larger boilers, the physical
distance over which reagent must be dispersed increases and
the surface area/volume ratio of the convective pass
decreases. Both of these factors are likely to make it more
difficult to deliver the reagent in the proper temperature
window, achieve good mixing between the reagent and flue gas,
3-45
-------
Performance of
Actual Commercial
Installation
i i i
1 2 3
Initial Mole Ratio of NH3 to NOx
Figure 3-22.
General effect of NH3:NOX mole
ratio on NOX removal.29
3-46
-------
500
100
500
SO .Concentration, ppm
3
Figure 3-23. Ammonia salt formation as a function of
temperature and NH3 and 803 concentration.
3-47
-------
and provide sufficient residence time for the mixture in the
temperature window. For larger boilers, more complex reagent
injection, mixing, and control systems may be necessary.
Potential requirements for such a system could include high
momentum injection lances, more engineering design, and
physical/mathematical modeling of the process as part of the
overall system design.
3.5.2 Selective Catalytic Reduction
3.5.2.1 Process Description. Selective catalytic
reduction involves injecting NH3 into boiler flue gases in the
presence of a catalyst to reduce NOX to N2 and water. The
catalyst lowers the activation energy required to drive the
NOX reduction to completion and decreases the temperature at
which the reaction occurs. The overall SCR reactions are:30
4NH3 + 4NO 4- O2 -» 4N2 + 6H2O (3-5)
8NH3 + 6NO2 -» 7N2 + 12H20 (3-6)
There are also undesirable reactions that can occur in an SCR
system, including the oxidation of NH3 and S02 and the
formation of sulfate salts. Potential oxidation reactions
are:31
4NH3 + 5O2 -> 4NO + 6H2O (3-7)
4NH3 + 3O2 -» 2N2 + 6H20 (3-8)
2NH3 + 202 -» N20 + 3H20 (3-9)
2SO2 + 62 -*• 2S03 (3-10)
The reaction rates of both desired and undesired reactions
increase with increasing temperature. The optimal temperature
range for achieving NOX reduction is shown in figure 3-24.n
Figure 3-25 shows several SCR configurations that have
been applied to coal-fired power plants in Europe or Japan.32
The most common configurations are diagrams la and Ib, also
referred to as "high dust" and "low dust" configurations,
respectively. Diagrams Ic and Id represent applications of
3-48
-------
NOx
CONVERSION
COMPOSITE
OF SCR NOx
AND NH3 OXIDATION
REACTIONS
MAXIMUM
CONVERSION
OPERATING
OPERATING
— WINDOW
TEMPERATURE
Figure 3-24.
Relative effect
on NOX reduction
temperature
3-49
-------
L
NH,
I
Ash
Waste Waste
Water Solids
(ib)
x/x/x/
i
Ash
L
NH,
Waste Waste
Water Solids
(1C)
L
NH3
I
Ash Dry Scrubber Waste
(id)
L
NH,
I
Ash
X/X/X/
I
Dry Scrubber Waste
X/X/X/
U----CJ
Waste Waste I
Water Rnliric 1—*•
Ash Water Solids
NH8
\/
SCR ESP or WetSO8 Spray
Heater FF Scrubber Dryer
Figure 3-25. Possible configurations for selective
catalytic reduction.
3-50
-------
spray drying with SCR. Diagrams la-id are called "hot-side*
SCR because the reactor is located before the air heater.
Diagram le is called "cold-side" SCR because the reactor is
located downstream of the air heater, particulate control, and
flue gas desulfurization equipment.
The SCR system hardware includes the catalyst material;
the NH3 system—including a vaporizer, storage tank> blower or
compressor, and various valves, indicators, and controls, the
NH3 injection grid; the SCR reactor housing (containing layers
of catalyst); transition ductwork; and a continuous emission
monitoring system. Anhydrous or dilute aqueous NH3 can be
used; however, aqueous NH3 is safer to store and handle. The
control system can be either feed-forward control (the inlet
NOX concentration and a preset NH3/NOX ratio are used) or
feed-back control (the outlet NOX concentration are used to
tune the NH3 feed rate), or a combination of the two. The
individual catalyst honeycombs or plates are combined into
modules, and the modules are applied in layers. Figure 3-26
shows a typical configuration for a catalyst reactor.
The catalyst must reduce NOX emissions without producing
other pollutants or adversely affecting equipment downstream
of the reactor. To accomplish this, the catalyst must have
high NOX removal activity per catalyst unit size; tolerance to
variations in temperature due to boiler load swings; minimal
tendency to oxidize NH3 to NO, and S02 to 803; durability to
prevent poisoning and deactivation; and resist erosion by fly
ash.
SCR catalysts are typically composed of an active
material and a catalyst support material. The active compound
promotes the NH3/NOX reaction and may be composed of a
precious metal (e.g., Pt, Pd), a base metal oxide, or a
zeolite. The entire catalyst cannot be made of these.
materials because they are expensive and structurally weak.
The catalyst support (usually a metal oxide) provides a large
3-51
-------
flow directing layer
catalyst element
catalyst module
catalyst layers
Figure 3-26. Typical configuration for a catalyst reactor.
33
3-52
-------
surface area for the active material, thus enhancing the
contact of the flue gas with the active material. Figure 3-27
shows examples of relative optimum temperature ranges for
precious metal, base metal, and zeolite catalysts.31
Some manufacturers offer homogeneous extruded monolithic
catalysts that consist of either base metal oxide or zeolite
formulations. The specific formulations contain ingredients
that have mechanical strength and are stable. This type of
catalyst is comparable in price to composite catalyst and has
been installed in Europe and Japan.34
The precious metal catalysts are typically platinum (Pt)
or palladium (Pd) based. They are primarily used in clean
fuel applications and at lower temperatures than the base
metal oxides or zeolite catalysts. The NOX reduction
efficiency of precious metal catalysts is reduced above 400 °C
(700 °F) because the NH3 oxidation reaction is favored.31
The most common commercially available base metal oxide
catalysts are vanadium/titanium based, with vanadium pentoxide
(V2°s) used as the active material and titanium dioxide (TiC>2)
or a titanium oxide-silicon dioxide (SiC>2) as the support
material.35 Tungsten oxide (WO^) can also be added to V/Ti to
increase resistance to catalyst deactivation and decrease
oxidation of 802 to 803. Metal oxides are among the best
catalysts for SCR because of their high activity at low
temperatures (<400 °C) and because of their high resistance to
poisoning by sulfur oxides.36
The zeolite catalysts are crystalline aluminosilicate
compounds. These catalysts are characterized by
interconnected systems of pores two to ten times the size of
NO, NH3, S02, and ©2 molecules. They absorb only the
compounds with molecular sizes comparable to their pore size.
The zeolite catalyst is reported to be stable over a wider
temperature window than other types of catalyst.
3-53
-------
-§
-8
-8
-S
8
k
CM
I
o
M
§>
3-54
-------
The SCR catalyst is usually offered in extruded honeycomb
or plate configurations as shown in figure 3-28.33 Honeycomb
catalysts are manufactured by extruding the catalyst-
containing material through a die of specific channel and wall
thickness. The pitch, or number of open channels, for coal-
fired applications is larger than the pitch for oil or natural
gas applications due to the increased amount of particulate
matter with coal-firing. Plate catalysts are manufactured by
pressing a catalyst paste onto a perforated plate or by
dipping the plate into a slurry of catalyst resulting in a
thin layer of catalyst material being applied to a metal
screen or plate.
3.5.2.2 Factors Affecting Performance. The performance
of an SCR system is influenced by six factors: flue gas
temperature, fuel sulfur content, NH3/NOX ratio, NOX
concentration at the SCR inlet, gas flow rate, and catalyst
condition.
Temperature greatly affects the performance of SCR
systems, and, as discussed earlier, each type of SCR catalyst
has an optimum operating temperature range. Below this range,
NOX reduction does not occur or occurs too slowly, which
results in NH3 slip. Above the optimum temperature, the NH3
is oxidized to NOX, which decreases the NOX reduction
efficiency.
The second factor affecting the performance of SCR is the
sulfur content of the fuel. Approximately 1 to 4 percent of
the sulfur in the fuel is converted to 803. The 803 can then
react with ammonia to form ammonium sulfate salts, which
deposit and foul, downstream equipment. Options for minimizing
formation of ammonium sulfate salts are to minimize NH3 slip,
select a catalyst with a low S(>2 to 803 conversion rate, or
burn a low sulfur coal. As shown in figure 3-29, the
conversion of SC>2 to 803 is temperature dependent, with higher
conversion rates at the higher temperatures.37 The temperature
sensitive nature of the conversion of SC>2 to 803 is especially
3-55
-------
yyyv
vvvv
honeycomb
plate
Figure 3-28. Configuration of parallel flow catalyst.
33
3-56
-------
0)
0>
To
0°
C/3
2-
1-
0-
482
572 662
Temperature, (°F)
752
Figure 3-29. Effect of temperature on conversion of
S02 to SO3.
3-57
-------
important for boilers operating at temperatures greater than
370 °C (700 °F) at the economizer outlet.37 Potential reaction
equations for ammonium sulfate salts are:36
NH3(gas) + SO3(gas) + H20(gas) -» NH4HSO4(liquid) (3-11)
NH4HS04(liquid) + NH3(gas) -> (NH4)2SO4(solid) (3-12)
With the use of medium- to high-sulfur coals, the •>
concentration of S03 will likely be higher than experienced in
most SCR applications to date. This increase in SO3
concentration has the potential to affect ammonium sulfate
salt formation. However, there is insufficient SCR
application experience with medium- to high-sulfur coals to
know the nature of the effects. Applications of SCR with
medium- to high-sulfur coals may need to incorporate ways to
minimize the impacts of ammonium sulfate salt formation and
deposition.
The third factor affecting SCR performance is the ratio
of NH3 to NOX. For NOX reduction efficiencies up to
approximately 80 percent, the NH3-NOX reaction follows
approximately 1:1 stoichiometry. To achieve greater NOX
removal, it is necessary to inject excess NH3, which results
in higher levels of NH3 slip.
The fourth factor affecting SCR performance is the
concentration of NOX at the SCR inlet. The NOX reduction is
relatively unchanged with SCR for inlet NOX concentrations
above 150 ppm.39 However, at inlet concentrations below
150 ppm, the reduction efficiencies decrease with decreasing
NOX concentrations.
The fifth factor affecting SCR performance is the gas
flow rate. Gas flow through the reactor is expressed in terms
of space velocity and area velocity. Space velocity (hr-1) is
defined as the inverse of residence time. It is determined by
the ratio of the amount of gas treated per hour to the
catalyst bulk volume.40 As space velocity increases, the
contact time between the gas and the catalyst decreases. As
3-58
-------
the contact time decreases, so does NOX reduction. Area
velocity (ft/hr) is related to the catalyst pitch and defined
as the ratio of the volume of gas treated per hour to the
apparent surface area of the catalyst. At lower area
velocities, the NOX has more time to react with NH3 on the
active sites on the catalyst; at higher area velocities, there
is less time to react.
The sixth factor affecting SCR performance is the
condition of the catalyst material. As the catalyst degrades
over time or is damaged, NOX removal decreases. Catalyst can
be deactivated by attrition, cracking, breaking, or from
fouling by solid particle deposition in the catalyst pores and
on the surface. Similarly, catalyst can be deactivated or
"poisoned11 when certain compounds (such as arsenic, lead, and
alkali oxides) react with the active sites on the catalyst.
Poisoning typically occurs over the long term, whereas fouling
can be sudden. When the maximum temperature for the catalyst
material is exceeded, catalysts can be thermally stressed or
sintered, and subsequently deactivated. As the catalyst
degrades by these processes, the NH3/NOX ratio must be
increased to maintain the desired level of NOX reduction.
This can result in increased levels of NH3 slip. However, the
greatest impact of degradation is on catalyst life. Because
the catalyst is a major component in the cost of SCR, reducing
the life of the catalyst has a serious impact on the cost.
The first layer of catalyst is typically a "dummy" layer
primarily used to straighten the gas flow and reduce erosion
of subsequent catalyst layers. The dummy layer is made of
inert material that is less expensive than the active catalyst
material. A metal grid can also be used as a straightening
layer. To maintain NOX removal efficiency, several options
exist for replacing active catalyst material as degradation
occurs. First, all the catalyst may be replaced at one time.
Second, extra catalyst may be added to the reactor, provided
3-59
-------
extra space has been designed into the reactor housing for
this purpose. Or, third, part of the catalyst may be
periodically replaced, which would extend the useful life of
the remaining catalyst.
The factors that affect the performance of SCR systems
may not be as severe on natural gas- and oil-fired
applications as they are on coal-fired boilers. Of wthe
factors listed above, one factor which will not have as much
of an effect is the fuel sulfur content because these fuels do
not contain as much sulfur as coal. Therefore, there will not
be as much 803 in the flue gas to react with excess ammonia
and deposit in downstream equipment.
Another parameter which will not have as much impact in
natural gas- or oil-fired boilers is the condition of the
catalyst material. The SCR catalyst material can still be
damaged by sintering or poisoned by certain compounds.
However, since natural gas- and oil-fired boilers do not have
as much fly ash as coal-fired boilers, the pores in the
catalyst will not plug as easily and the surface of the
catalyst is not scoured or eroded due to the fly ash
particles.
3.6 PERFORMANCE OF FLUE GAS TREATMENT TECHNOLOGIES ON
COAL-, NATURAL GAS-, AND OIL-FIRED BOILERS
This section presents the demonstrated performance of the
gas treatment technologies (SNCR and SCR) applied to
industrial boilers.
3.6.1 Selective Noncatalvtic Reduction
Nitrogen oxide emissions data were obtained from various
publications for boilers larger than 100 MMBtu/hr in capacity,
built or retrofitted after June 1984 (the promulgation date of
the current NSPS subpart Db), and use SNCR for NOX control.
This data reflects the performance of SNCR for representative
boilers of the subpart Db population. Data was not available
for oil- or natural gas-fired boilers with SNCR; therefore,
3-60
-------
only data for coal-fired boilers are presented in this
section. Table 3-4 presents plant names, boiler status (new
or retrofit), rated boiler capacity (MMBtu/hr), boiler type,
coal type, reagent type, tested load, test duration,
uncontrolled and controlled NOX levels (Ib/MMBtu), and NH3
slip (ppm @ 3 percent 02)• The data is shown by boiler type.
Eleven coal-fired boilers equipped with SNCR are
presented in table 3-4. Boiler capacities range from 120 to
750 MMBtu/hr. Three pulverized coal-fired boilers use urea as
the SNCR reagent, two of which are tangential fired and one is
wall-fired. Nitrogen oxide emissions from the two tangential
PC-fired boilers were reduced by 30 percent resulting in
controlled NOX emissions of 0.19 Ib/MMBtu. Ammonia slip
levels were maintained below 5 ppm during these tests. On the
wall PC-fired boiler, NOX levels were reduced by 60 percent
resulting in a controlled NOX level of 0.11 Ib/MMBtu.
Four FBC boilers are presented in table 3-4. Three of
these use urea, and one uses NH3 as the SNCR reagent and NOX
reductions of 78 to 88 percent were achieved. Controlled NOX
levels ranged from 0.03 to 0.16 Ib/MMBtu. Finally, for the
four spreader stoker boilers using NH3 as the SNCR reagent,
NOX reductions of 56 to 63 percent were achieved, resulting in
controlled NOX emissions of 0.15 to 0.18 Ib/MMBtu. Ammonia
slip levels were less than 20 ppm.
3.6.2 Selective Catalytic Reduction
Nitrogen oxide emissions data were obtained from various
publications for boilers larger than 100 MMBtu/hr in capacity,
built or retrofitted after June 1984 (the promulgation date of
the current NSPS subpart Db), and use SCR for NOX control.
This data reflects the performance of SCR for representative
boilers of the subpart Db population. The data are presented
in table 3-5. Because data for only two industrial boilers
with SCR was available, table 3-5 also includes data for SCR
applications on utility boilers. Additional information on
pilot-scale SCR applications is contained in Section 3.6.2.2
3-61
-------
S5
O
y
Q
H
OS
0
H
El
X
S
|
O
H
B
w
W
CO
H
CO
w
l_]
H
O
oa
•4
i_i
g
CO
D
Q
_/
O
fc
s
CO
^i
Q
1
n
H
J
1
«
fj
I
•2 M ^^ ^^
i— E.*
I"" Eg
^
__
|ix|
& I
5
2
•S3 g
If "
x| "8
Sc t ~
5 ^
§
ii T
^fc
«j -a
888
H- -* w
4^
M
s **
ac
m
|
7
(.
II
09 v
ti-
ll*
S £
^ t o '
z ° *•**
* ac
j
4-*
1
(
(
££ ^ 1^5 ^ ^ ^ ^^^^
in in oe Sf£<- £ Si Si 8 ° «3 °
* * V V V V
oo o te te ee £p 3 S m 14 - »- o ^ o o »- «-
N N N >r CW f^ ^ ^ ^ *^
oo o ee ee ee o oo oooo
ee K te ee ee ee ee ee ee te te te te
te ee te in in g as eeae oeaeaeae
• • a • c
SS 8 ? 1 ? ? 1111
33 3 3 I 33 I I I I
MM M M M
ft 5 g O O
*^ M- nf ••_ *^ OC*** BCOCBCAC
l§ fi i 1 g S ""
••_ *«~ ••» ••» ••—
CD CD CD CD CD
gg - oi « Li £ MMMM
99 fl U- U. U. U. MMMC/
»- »- 3 u u o u
MllAin Kl ND CO ON CO CO CO a
i g) f.
T! * *
t Ji. te ee 3 — _ — •——_— —
~- £f.x S 4 o * • * *
i c TI i
i I&2 3 c £ «• 2 ? *
1-S i E
Ml 9 H
1
1
1
H
%
g
1
1
o
1
•8
N
2
"5
g
*J
§
u
M
u
to
ft*.
u
?""
<*-
s
M
i
• »
L.
I.
5
1
V
*
/
3-62
-------
o
H
EH
U
Q
W
U
M
EH
EH
rt^
U
W
H
EH
W
W
W
EH
H
cn~
Pd
W
)— i
O
CO
MH
O
P^
g
3
&
w
ri
1
in
CO
H
1
01
g
01
t_
01
BC.
g
O
Z 3 ^
L!
o*9
Z 4J
CO
01 I.
CO
1 ~I
|~»3
TJ
Cfl — I.
0) I- jC
1- CJ -X
Q.
4-«
c 4,
0£ 01 X
CJ O) S
C/> CO >•
0)
3 t-i *-*
"*~ CO *- ° ^
CJ _• *"' —
(/> M-
TJ
CO
0
0>
01
5-*^ *^
4J .c
(/) Q
"is'i
u_
4-»
0
a
CD
O
4-* CO
3 CO
>v> c -
(" C ^^ C/l C/l
cu 3 « & O z
_:,
3 4J o — • CO — •
*D (A CU U — ' —
C 0) — ' CD 0) 4J
— 3 uj no 3
intnininmininininininininc^ >o*o>o ^O^^o r^
at at ct at at at at at at at at at at \r\ " S"^. " » ro 00^3 ^-VL^
cJoo'oocaocsoooo'oo dc)O odd
» «M CM co o >o r>- N- -o ro CM ~»
ro ro ro ro ro ro ro ro ro ro ro ro -* o ae at at at at at ° • u
'ox 3
CO 4^ O
CO U "-
O CO 01
Q. 4->
I. CD CO
oooooooooooooo ooo ooo o " ° .T S
OOOOOOOOOOOOOO OOO OOO O ^ c. 0)
*~'~*^*~*~*~'~*~*~*~"~*~'~*~ *~*~*~ «-«-*- T- OOJ—.C.
£— c —
•*• i^ ID
O A (. 4J
0) CD
o o oo ooluoiu
c c c c m ><- o o .e ^
E E !_,'"'§ uin XR1"81
4-'CO'^ 4^CD^^ 4-•_« CDCU
— O CM — O CM — O «- — O Z CO — O — in O E
on o ^ m u ^ coo^mcjv^ u o -^ o rM in *-<
CD O.
XX —
J-T— T>_
-j 3 01 CJ Ol «
1 M 4 S •% I i 3 «
•« ^ i S ^ £ .1.1 gl
fM CM CM>* >* in OO^
j£ ^ CMFO fO tM t. £ oT
& &.2"crt
CO CO--
«- *^ CM^> 4-> ceM^^cc o y o o ro4J
•O *O CCD *Q. •(X4^*C1. -y m
WO «OO 2O COt/) COl/>t7>Ol/>4-.4JZOi
X Z- X Z> -I.OZZL.OZZC Z Z — — S
014^ c/i o>4^ c/) c cy>oi"- tot— i»-f IU_H>>«!
C C 4J O. CC4-1Q. CD<-O.O4J4->OIU4^4->OI4-'U«-'OIOO«IU
C- — — Crt L. •—•— O) . (/) c CD — 1- 1- CO ••- L. C CO •— C_ f_ S* c — .
COOCZ COOCZ OOZ OI4->CQ.«l4^CCL30lCa4JuocOTl
ua.3^ ua.r>^ -Jz^zooD^zoo^^zcozj^oja)— -a
L. L. CO + *J
L.
a> at at at o
X X — — cj CJ Q.
en t» ra oflcoacco cnc/ioi
C C C CO CO — 4^ 4J u
toe roc OCUOLOUOLOOJOC 3333O
L.CO C-fO C_CO.«->^4^CJ-*-»-4^CJfH-O "D*DZ
• 0)Q. *OIQ. • 0! Q. — • > o — • > o 4^"-cA aeac:ii
cncg WCS «rtCEJ3i-oic«3 — — cjcj^ccn
oucj 3uu =iucjo.c/iujua.iu<9c/>cjui _ ee
ID J3 u ^ z
3-63
-------
of the subpart Da Technical Support Document.42 Table 3-5 \
presents the company name, the unit name and standard which '
the boiler is subject to, boiler capacity, fuel type, test
load (percent), the SCR flue gas flow rate (106 Ib/hr), the
SCR reagent type, test period (hr), ammonia slip (ppm), outlet
NOX (Ib/MMBtu), and NOX reduction (percent).
NOX emissions data were obtained for a 380 MMBtu/hr
natural gas-fired packaged boiler operated by the Westinghouse
Electric Corporation. This boiler is equipped with FGR,
oxygen trim, and SCR for NOX control. During three 30-minute
tests performed on this boiler, NOX emissions were measured at
0.01 Ib/MMBtu. For the 110 MMBtu/hr natural gas-fired
packaged boiler operated by the Darling Delaware Co., the NOX
emissions were also 0.01 Ib/MMBtu. The NH3 slip for this
boiler was 3 ppm.
Performance data on commercial size installations of SCR
systems were collected on three coal-, one coal-/gas-, and one
^
oil-/gas-fired utility boilers. Data on two of the coal-fired I
boilers were collected during the final 100-hour acceptance
tests conducted in May 1994 for the Carneys Point Generating
Plant. This is the first U.S. coal-fired plant equipped with
an SCR. The plant is nominally rated at 285 MW and includes
two coal-fired boilers. The boilers burn a 2 percent sulfur,
eastern bituminous coal. Besides the SCR system, each boiler
is equipped with LNB and an AOFA system for control of NOX
emissions. The high-dust SCR system, which is located between
the economizer and air heater consists of a homogenous
honeycomb type catalyst. The major catalyst constituents are
V2°5' Ti°2» an<^ WC>3 . Aqueous ammonia containing approximately
27 percent NH3 is used as the SCR reagent. The system is
designed for 63 percent NOX removal (0.1 Ib/MMBtu outlet NOX
level) with an NH3 slip of 5 ppm (at 7 percent ©2). Permitted
NOX levels are not to exceed 0.17 Ib/MMBtu (3-hour rolling
average) ,48
3-64
-------
As shown in table 3-5, for both boilers, the short-tena
NOX emissions (4-hour averages) at full load were below
0.15 Ib/MMBtu with NH3 slip levels less than 0.3 ppm at
7 percent O2. No adverse effects on the catalyst or on
downstream equipment were reported during inspections
following the acceptance tests.*
The technical feasibility of SCR was also demonstrated on
a 344 MW, wet-bottom, coal-fired utility boiler at Mercer
Station. The boiler burns low-sulfur, eastern bituminous coal
as the primary fuel and natural gas as the secondary fuel.
The performance of an in-duct SCR system and of its
combination with an air-heater SCR has been evaluated on this
boiler. The in-duct SCR reactor, located in an expanded
section of a horizontal duct between the boiler economizer and
air heater, processed approximately 25 percent of the flue gas
from Unit 2. The reactor contains V/Ti plate type catalyst
elements oriented vertically to minimize fly ash deposition.
For the air-heater SCR, the existing hot-end air heater
baskets were replaced with catalytic air heater baskets. The
catalyst composition was identical to the in-duct SCR
catalyst. The entire system was designed to achieve an outlet
NOX of 0.2 Ib/MMBtu (approximately 88 percent reduction for
coal)/6
Results shown in table 3-5 indicate that the in-duct SCR
achieved NOX emissions of 0.20 Ib/MMBtu for coal-firing and
0.11 Ib/MMBtu for gas-firing with NH3 slip levels of 10 ppm
(NH3 slip measured at the inlet to the air heater for both
fuels). Higher NOX emissions of 0.30 Ib/MMBtu for coal and
0.27 Ib/MMBtu for natural gas were measured with NHs slip
levels of 5 ppm (at the air heater inlet). The addition of
the air-heater SCR decreased NOX emissions to 0.06 Ib/MMBtu
for coal and 0.04 Ib/MMBtu for natural gas with less than
5 ppm NH3 slip at the air-heater outlet. Information on
catalyst volume or space velocity was not reported.
3-65
-------
The effect of catalyst exposure time on catalyst activity
was examined on Huntington Beach, Unit No. 2. This is an oil-
or gas-fired boiler rated at 215 MW. The SCR unit processed
approximately one-half of the boiler flue gas (107.5 MW) and
was designed for 90 percent NOX reduction at 10 ppm NH3 slip
for oil-fired flue gas. Total SCR operating time was
approximately 25,000 hours during the test period between 1982
and 1986. About one quarter of this operation was on flue gas
from low sulfur oil (0.25 percent by weight) and the remainder
of the operation was on natural gas.
The results indicate that during the initial 2,000-
7,000 hours of operation, 90 percent NOX reduction was
achieved with NH3 levels of 14 ppm. After 17,000 hours, NH3
slip increased to about 40 ppm at 90 percent NOX removal. The
higher NH3 slip levels were believed due to catalyst
deactivation.
3.6.3 Recent Industrial and Utility Boiler Permit Decisions
Involving SCR
Under the New Source Review (NSR) program, State
permitting agencies have set NOX emission limits for new
boilers that are more stringent than those contained in the
current NSPS, subparts Db and Da. Only those decisions that
involve the installation of SCR systems are discussed in this
section.
Table 3-6 lists new industrial and utility boilers that
are either already operating with, or are expected to be
built, with SCR for NOX emissions control. The list includes
two gas-fired industrial boilers subject to the NSPS,
subpart Db, and six coal-fired utility boilers, two each in
New Jersey, Florida, and Virginia under subpart Da. To date,
there are no known new oil- or natural gas-fired boilers that
have been permitted with SCR under subpart Da. However,
several existing gas-fired utility boilers operating in
California are being retrofitted with SCR systems to meet
local NOX regulations.
3-66
-------
w
W
J^
>4
£4 55
M O
J H
• J P.
Q W
O
hH H
H 3
its
H U
>
W £•!
2 U
fa 3
o w
CO
£4
(A
H
J
•
\O
1
U
M
H
«
tl
i
£• «
•II
^-1
ac
i !i
! i»
E
*• .0 ^
J»l
™* 8 ^
>Cw ^
g^w
S3
"5
fc,
C 1. C
i>«- U kl
>- 3 t O)
M Q."-
£"" ^ "s £
W-S
41
" ^
"E «
k> n
"c ""^
a o
4>
a
VI
Company Name
(Plant Name)
g
L.
I
Industrial Boilers (NSPS
3
u
*
1
to
o
o
in
o
o
i
S
o
^J
S
^
^
S ^
i
w
«*
2
s
w
p-
Uestinghouse Electric Co
5
«
l]c
P w
O
o
i
s
o
1
2
5
^S
?i
i
^
^
3
Darling-Delaware Company
Incorporated
•
0
f1"
i
s
£
3
S
1,
i
K 0
0 O
^*
i
u *•
*• ^
*~
fM
•0
^
•8
** ^s
•s*
Chambers Cogeneratlon Li
Partnership (Carneys Pol
(M
in
in
^
i
Kl
t§
O O
^^
i
u "*
4\J ^
•^
^
S
Logan Energy Service Com
(Logan Generating Plant)
R
*^
||
N.
O
i
o °
*• PJ
(M
Ik
g
•^
IA
Orlando Utilities Conrnis
(Stanton)
in
o
in
|i
t:
o
$
u °
*• «M
Mg
«~
k)
(L
U.S. Generating Company
(Indiantown Cogeneration
Sn
JQ
II
si
in o
o o
^*
i
o1^
*• ^
~i
«••
^%
•8
kt
•
Birchwood Power Partners
(SEI, Birchwood, Incorpo
in
in
ft
||
ft°
00
I
u> *~
"• ^-
s-'l
<0
1
**
^
^
Cogentrlx, Incorporated
(Cogentrlx of Dindwiddie
Incorporated)
t
^
•«—
•g
•
o
"a
\
S>
3
1
ee
u
VI
•
•>
•; t
* 6
« "5
At *2
! •
& r i
" £ 2 .
§""* 8 S
5 -d ~ «
C * • « k>
1 1 1 «• ! s
f 1 ! « * i
J> — 0 • k.
I C 1 8. • 2
« • *»
3-67
-------
As shown in table 3-6, the two gas-fired industrial
boilers began operation in 1989. The 380 MMBtu/hr (heat
input) gas-fired boiler operated by Westinghouse Electric
Corporation operates at high turndown ratio (10:1) each day.
During low-load operation, the temperature of the flue gas at
the inlet to the SCR system is lowered which adversely affects
the achievable NOX reduction efficiency.43 To account for this
decreased performance at low loads, the NOX emissions from the
boiler are regulated on a sliding scale based on load. At 75-
100 percent of full load, NOX emissions are limited to
12 ppmvd at 3 percent C«2 (0.015 Ib/MMBtu) based on a 3-hr
average. This limit is increased progressively as load is
decreased until at 10-30 percent of full load, NOX emissions
are limited to 24 ppmvd (0.030 Ib/MMBtu). On a mass basis,
NOX emissions are limited to 140 lb/day.*3
The 110 MMBtu/hr (heat input) boiler owned by Darling-
Delaware Company is limited to 9 ppmvd at 3 percent 02
(0.01 Ib/MMBtu) averaged over a minimum of 15 minutes. During
startup, NOX emissions are limited to 40 ppmvd at 3 percent ©2
(0.05 Ib/MMBtu) .*9
The two PC-fired utility plants in New Jersey went into
commercial service in 1994. These units are located in the
northeast ozone transport region (NOTR), which is classified
as a nonattainment area for ozone. Permit levels for NOX
emissions from these boilers were set initially at
0.17 Ib/MMBtu (3-hour average) with NH3 slip levels of
10 ppmvd at 7 percent 02 (3-hour average). However, the
permit conditions require that the SCR systems be designed for
outlet NOX emission levels of 0.10 Ib/MMBtu (3-hour average).
During operation, the SCR systems must be optimized to achieve
the design NOX emission limit by catalyst addition and/or
replacement as necessary, to the extent that catalyst
addition/replacement does not exceed 50 percent of initial
catalyst over a 5-year operating period. Final NOX emission
limits will be set between 0.10 and 0.17 Ib/MMBtu based on the
3-68
-------
NOX levels that are demonstrated to be continuously achievable
over the first 5-year operating period. ' '
The two PC-fired plants in Florida are located in ozone
attainment areas. The 464 MW, Stanton Unit No. 2 is permitted
for a NOX emission rate of 0.17 Ib/MMBtu (30-day rolling
average) with NH3 slip levels below 30 ppmv (wet,
uncorrected)." This unit is designed to achieve 0.1 Ib/MMBtu
and is guaranteed by the vendor to achieve the permitted level
of 0.17 Ib/MMBtu.56 The 330 MW, Indiantown Cogeneration unit
is permitted at 0.17 Ib/MMBtu NOX (24-hour block average) with
NH3 slip levels below 50 ppmv.
Finally, Virginia has issued permits for one PC-fired
boiler and eight SS boilers with permit conditions that are
similar to those issued for the New Jersey boilers. The SCR
system on the 220 MW PC-fired boiler in Virginia must be
designed to achieve a NOX emission level of 0.10 Ib/MMBtu (30-
day rolling average). During operation, if this emission
limit cannot be maintained, then a 4th layer of catalyst must
be added to the original 3-layer catalyst bed in the SCR
system and/or existing catalyst must be replaced to the extent
that catalyst replacement does not exceed 50 percent of the
design catalyst volume within each 3-year operating period.
If this proves ineffective then the maximum NOX emission limit
of 0.15 Ib/MMBtu must not be exceeded. NH3 slip levels are
limited below 25 ppmv in the flue gas. Fuel sulfur levels are
limited to below 1.0 percent by weight (annual average) and
1.2 percent per shipment.55
The SCR systems for the eight SS boilers in Virginia must
be designed to achieve a NOX emission level of 0.10 Ib/MMBtu
(30-day rolling average). If catalyst addition/replacement
(to the extent that this does not exceed 50 percent of design
catalyst volume in a 3-year operating period) is ineffective
in maintaining the 0.10 Ib/MMBtu level than the maximum NOX
emission level of 0.25 Ib/MMBtu must not be exceeded. The NH3
3-69
-------
slip level is limited to 25 ppmv and fuel sulfur is limited to
1.1 percent by weight per shipment.
3.6.4 Analysis of Long-Term Continuous Emission Monitoring
Data from Coal-fired Boilers with SCR
Continuous emission monitoring data was analyzed from
Carneys Point Units 1 and 2 for the period July 1 to
September 30, 1995." The objective of the data analysis was
to assess the long-term NOX emission levels that can be
continuously achieved by these coal-fired boilers. Each data
set contained hourly-averaged values of NOX emissions and
oxygen levels measured over a 3-month operating period.
Boiler load information for the same time period was not
available; therefore, correlation of NOX emissions as a
function of load could not be determined. For each boiler,
data was analyzed for which there were at least 18 hours of
data per day.
The Carneys Point Generating Plant consists of two
142.5 MW, wall-fired, pulverized coal boilers. The plant is
operated by U.S. Generating Company and supplies up to 184 MW
to Atlantic City Electric Company and up to 1 million Ib/hr of
steam and up to 40 MW of electricity to E.I. DuPont
deMenours & Company's Chambers Works. Each boiler is equipped
with eight Foster Wheeler ISF low NOX burners in four rows of
two burners each on the front wall. Additional NOX reduction
is achieved with an AGFA system consisting of four circular
air ports on the front wall and four circular ports in the
rear wall. There is also a spray-dryer type of flue gas
desulfurization system designed to reduce SC>2 by 90 percent or
more and a reverse-air type fabric filter baghouse removes the
particulate matter.*5
The SCR unit is located between the economizer and air
heater on each unit. There are two catalyst layers with space
provisions for a third layer. The first two layers are
expected to achieve design performance for up to seven years
and if necessary, the third layer could be added to meet the
3-70
-------
10-year performance guarantee. The system was designed for an
inlet NOX level of 0.27 Ib/MMBtu with an outlet of
0.10 Ib/MMBtu. Overall SCR design removal efficiency is
63 percent with ammonia slip of 5 ppm (corrected to 7 percent
o2).A5
3.6.4.1 Carneys Point Unit 1. Continuous emission
monitoring data were obtained from U.S. Generating Company for
Carneys Point Unit I for the operating period between July 1
and September 30, 1995. This data set includes a total of
92 (24-hour-averaged) NOX emission values. A time plot of the
NOX emission data is shown in figure 3-30. Typical hourly NOjj
emission rates are between 0.118 and 0.147 Ib/MMBtu and
average approximately 0.14 Ib/MMBtu.
Analysis of the data set included assessing the
implications for standard setting resulting from
autocorrelation between successive measurements. Basic
statistics for this data set are summarized in the first data
column of table 3-7 for the 24-hour block averaging time. As
indicated in the table, the mean NOX level was 0.137 Ib/MMBtu.
Based on the absolute values of the skewness estimate of
1.253, the data were judged as not being normally distributed
at the 95 percent confidence level (for a sample size of 92,
the 95 percent confidence limit for skewness and kurtosis are
approximately 0.41 and 3.80, respectively). To compensate
for skewness, the data were transformed using a Box-Cox
transform with a X=7.00. The statistical properties of the
transformed data are presented in the second data column of
table 3-7. As shown in the table, the absolute value of
kurtosis is within the 95 percent confidence limits for a
normal distribution. Based on the autocorrelation analysis of
the transformed data, an AR(1) model was used to estimate the
7- and 30-day rolling average NOX emission limits.
Table 3-8 shows the estimated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-7 and the
3-71
-------
in
a\
in
s
o
i .
co
-------
TABLE 3-7. SUMMARY STATISTICS FOR CARNEYS POINT
UNIT NO. 1
(24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean (x24) t lb/MMBtu]
Standard deviation (824) [lb/MMBtu]
Skewness3
Kurtosis3
Shapiro-Wilk W statistic3
Probability < W
Autocorrelation coefficient^ (p^j
Correction factor (Fi)
Transformed
As measured data
data (X=7.00)
92
0.137
0.007
-1.253
4.639
0.886
0.0001
NEC
NEC
92
9.32E-7
2.71E-7
-0.481
2.831
0.941
0.0007
0.421
1.016
aFor a perfectly normal distribution skewness
kurtosis = 3, and W = 1.
0,
lag 1 is significant.
CNE = not estimated.
3-73
-------
TABLE 3-8. ACHIEVABLE NOX EMISSION LIMITS* FOR
CARNEYS POINT UNIT NO. 1
Period
24-hr Block
7-day Rolling
30-day Rolling
Exceedance Frequency
1%
0.148
0.144
0.141
I/year
0.149
0.145
0.141
1/10 years
0.152
0.147
0.142
alb/MMBtu.
3-74
-------
statistical procedures described in section 3.2.3.2 of the
subpart Da Technical Support Document*2. As indicated in
table 3-8, achievable NOX emission limits based on one
exceedance in 10 years range from 0.152 Ib/MMBtu for a 24-hr
block average to 0.142 Ib/MMBtu for a 30-day rolling average.
3.6.4.2 Carneys Point Unit 2. Continuous emission
monitoring data were obtained from U.S. Generating Company for
Carneys Point Unit 2 for the operating period between July 1
and September 30, 1995. This data set includes a total of 85
(24-hour-averaged) NOX emission values. A time plot of the
NOX emission data is shown in figure 3-31. Typical hourly NOX
emission rates are between 0.061 and 0.143 Ib/MMBtu and
average approximately 0.13 Ib/MMBtu.
Analysis of the data set included assessing the
implications for standard setting resulting from
autocorrelation between successive measurements. Basic
statistics for this data set are summarized in the first data
column of table 3-9 for the 24-hour block averaging time. As
indicated in the table, the mean NOX level was 0.13 Ib/MMBtu.
Based on the absolute values of the skewness estimate of
2.375, the data were judged as not being normally distributed
at the 95 percent confidence level (for a sample size of 85,
the 95 percent confidence limit for skewness and kurtosis are
approximately 0.42 and 3.84, respectively).58 To compensate
for skewness, the data were transformed using a Box-Cox
transform with a X=6.75. The statistical properties of the
transformed data are presented in the second data column of
table 3-9. As shown in the table, the absolute values of
skewness and kurtosis are within the 95 percent confidence
limits for a normal distribution. The absolute values of
skewness is slightly outside the 95 percent confidence limits
for a normal distribution, but is within the 99 percent
confidence limits (0.61) for a normal distribution. Based on
the autocorrelation analysis of the transformed data, an AR(l)
3-75
-------
ID
C\
O\
o
M
13
4J
(X
0)
W
O
•P
0)
c
o
•H
)
o*
o
CO
0)
(n»aww/qi)
3-76
-------
TABLE 3-9. SUMMARY STATISTICS FOR CARNEYS POINT
UNIT NO. 2
(24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean (x24) [ Ib/MMBtu]
Standard deviation (324) [Ib/MMBtu]
Skewnessa
Kurtosisa
Shapiro-Wilk W statistic3
Probability < W
Autocorrelation coefficient*3 (p±)
Correction factor (F^)
Transformed
As measured data
data (X=2.25)
85
0.131
0.015
-2.375
9.629
0.716
0.0001
NEC
NEC
85
1.52E-6
4.32E-7
-0.509
2.859
0.940
0.0005
0.423
1.017
aFor a perfectly normal distribution skewness
kurtosis = 3, and W = 1.
lag 1 is significant.
CNE = not estimated.
3-77
-------
model was used to estimate the 7- and 30-day rolling average
NOX emission limits.
Table 3-10 shows the estimated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-492 and the
statistical procedures described in section 3.2.3.2*of the
42
subpart Da Technical Support Document. As indicated in
table 3-10, achievable emission limits based on one exceedance
in 10 years range from 0.152 Ib/MMBtu for a 24-hr block
average to 0.142 Ib/MMBtu for a 30-day rolling average.
3-78
-------
TABLE 3-10. ACHIEVABLE NOX EMISSION LIMITS3 FOR
CARNEYS POINT UNIT NO. 2
Period
24-hr Block
7 -day Rolling
30-day Rolling
Exceedance Frequency
1%
0.148
0.144
0.141
I/ year
0.150
0.145
0.141
1/10 years
0.152
0.147
0.142
alb/MMBtu.
3-79
-------
3.7 REFERENCES \
1. Vatsky, J, et al. Development of an Ultra-Low NOX
Pulverized Coal Burner. Presented at the 1991 Joint
Symposium on Stationary Combustion NOX Control.
Washington, DC. March 25-28, 1991.
2. Larue, A. D. The XCL Burner - Latest Development and
Operating Experience. In Proceedings: 1989 Joint
Symposium on Stationary Combustion NOX Control. Vol. 1.
U. S. Environmental Protection Agency. Research Triangle
Park, NC. Publication No. EPA-600/9-89-062a. pp. 3-93
through 3-109.
3. Way, K., et al. Results from a Utility-Scale
Installation of ABB-CE Services RO-II Low NOX, Wall-Fired
Burners. Presented at the 1993 Joint Symposium on
Stationary Combustion NOX Control. Miami Beach, FL.
May 24-27, 1993.
4. Briggs, O. G., A Total Combustion Systems Approach
Proves Successful for NOX Control for Two Steam
Generators. Presented at the American Power Conference,
April 1991.
5. Emmel, T. E., Radian Corporation, (letter and J
attachments) to Kosim, Z., U. S. Environmental Protection ^
Agency. July 11, 1991.
6. Donais, R. E., et al. 1989 Update on NOX Emission
Control Technologies at Combustion Engineering. In
Proceedings: 1989 Joint Symposium on Stationary
Combustion NOX Control. Vol. 2. U. S. Environmental
Protection Agency. Research Triangle Park, NC.
Publication No. EPA-600/9-89-062b. pp. 4-37 through
4-56.
7. Hardman, R. R., Tangentially Fired Low-NOx Combustion
System Test Results at Gulf Power Company's Lansing Smith
Unit 2. Presented at the EPRI Conference on NOX Controls
for Utility Boilers Workshop. Cambridge, MA. July 7-9,
1992.
8. State-of-the-art Analysis of NOX/N20 Control for
Fluidized Bed Combustion Power Plants. Acurex. Final
Report. 90-102/ESD. Prepared for EPRI. July 1990.
9. Alternative Control Technique (ACT) Document—NOx
Emissions from Industrial/Commercial/Institutional (ICI)
Boilers. Publication No. EPA-453/R-94-022.
U.S. Environmental Protection Agency. Research Triangle -w
Park, NC. March 1994, pp. 5-23 to 5-25. J
3-80
-------
10. Facsimile from Pat Tortora, Riley Stoker Corporation to
Rachel Adams, Radian Corporation. Low NOX CCV Burner
Experience List. June 30, 1995
11. Ref. 9. Appendices A and B.
12. Hutchinson, B., The Pyroflow Boiler at BFGoodrich Company
- The First Eighteen Months of Steaming at Henry,
Illinois. Presented at the 9th International Conference
on Fluidized Bed Combustion. Boston, MA. May 3*987.
pp. 85-90.
13. NOX RACT Guidance Document, Jan. 1993 Presented at the
CIBO Sixth Annual NOX Control Conference. Houston, TX,
February 8-10, 1993. pp. 57-84.
14. Telecon from Rachel Adams, Radian Corporation to Bob
Bassett, Council of Industrial Boilers. Further
information concerning test data. July 1995.
15. deVolo, N. B., et al, NOX Reduction and Operational
Performance of Two Full-Scale Utility Gas/Oil Burner
Retrofit Installations. Presented at the 1991 Joint
Symposium on Stationary Combustion NOX Control,
Washington, DC. March 25-28, 1991.
16. ROPM Burner for Oil and Gas Wall Fired Generating
Facilities. ABB Combustion Engineering. Publication
PIB 103. 1990.
17. Peabody ISC™ Low NOX Burners. Peabody Engineering.
Bulletin No. ISC-1. Undated.
18. Lisauskas, R. A., and c. A. Penterson. An Advanced Low-
NOX Combustion System for Gas and Oil Firing. Presented
at the 1991 Joint Symposium on Stationary Combustion NOX
Control. Washington, DC. March 25-28, 1991.
19. Ultra Low NOX Rapid Mix Burner Brochure. Todd
Combustion. Undated.
20. PM Burner for Oil and Gas T-Fired Generating Facilities.
ABB Combustion Engineering. Publication PIB 102. 1990.
21. DeHaan T.E. The New QLN - A Quantum Leap in Low NOX
Technology. In Proceedings: Council of Industrial
Boiler Owners Eighth Annual NOX Control Conference.
Philadelphia, PA. March 7-8, 1995.
22. Gamburg, M. Todd Ultra Low Nox Rapid Mix Burner (RMB).
In Proceedings: Council of Industrial Boiler Owners
Eighth Annual NOX Control Conference. Philadelphia, PA.
March 7-8, 1995, pp. 41-48.
3-81
-------
23. Epperly, W. R., et al, Control of Nitrogen Oxides \
Emissions from Stationary Sources. Presented at the l
Annual Meeting of the American Power Conference, April
1988.
24. Letter and attachments from Haas, G.A., Exxon Research
and Engineering Co., to Gundappa, M., Radian Corporation.
May l, 1992. Information concerning Thermal DeNOx.
25. Cato, G. A., Maloney, K. L., and J. G. Setter. * Reference
Guideline for Industrial Boilers Manufacturers to Control
Pollution with Combustion Modification.
U. s. Environmental Protection Agency. Research Triangle
Park, NC. Publication No. EPA-600/8-77-003b. pp. 49-51.
November 1977.
26. Ref. 9. p. 5-75.
27. Technical and Economic Feasibility of Ammonia-Based
Postcombustion NOX Control. Report No. EPRI CS2713.
November 1982. pp. 3-18 to 3-25.
28. Jones, D. G., et al, Preliminary Test Results High Energy
Urea Injection DeNOx on a 215 MW Utility Boiler.
Presented at the 1991 Joint Symposium on Stationary
Combustion NOX Control. Washington, DC. March 25-28,
1991. ^
;
29. Ref. 27. pp. 3-7 to 3-9.
30. Bosch, H. and F. Janssen. Catalytic Reduction of
Nitrogen Oxides, A Review on the Fundamentals and
Technology. Catalysis Today. Vol 2. p. 392-396.
April 1987.
31. Heck, R. M., Bonacci, J. C., and J. M. Chen. Catalytic
Air Pollution Controls Commercial Development of
Selective Catalytic Reduction for NOX. Presented at the
80th Annual meeting of the Air Pollution Control
Association. June 1987.
32. Cichanowicz, J. E., and G. Offen. Applicability of
European SCR Experience to U. S. Utility Operation. In
Proceedings: 1987 Joint Symposium on Stationary
Combustion NOX Control. Vol. 2. U. S. Environmental
Protection Agency. Research Triangle Park, NC.
Publication No. EPA/600/9-88/026b. pp. 28-1 through
28-18.
33. Hjalmarsson, A. K. NOX Control Technologies for Coal
Combustion. IEA Coal Research, p. 40-53. June 1990.
3-82
-------
34. Letter from Wax, M. J., Institute of Clean Air Companies
to Eddinger, J. A., U. s. Environmental Protection
Agency. Research Triangle Park, NC. November 18, 1993.
35. Chen, J. P., Buzanowski, M. A., Yang, R. T. , and
J. E. Cichanowicz. Deactivation of the Vanadia Catalyst
in the Selective Catalytic Reduction Process. Journal of
the Air Waste Management Association, 40:1403-1409,
October 1990.
*
36. Ref. 30. pp. 459 to 462.
37. Rummenhohl, V., Weiler, H. , and W. Ellison. Experience
Sheds Light on SCR O&M issues. Power Magazine.
136:35-36. September 1992.
38. Damon, J. E., et al. Updated Technical and Economic
Review of Selective Catalytic NOX Reduction Systems. In
Proceedings: 1987 Joint Symposium on Stationary
Combustion NOX Control. Vol. 2. U. S. Environmental
Protection Agency. Research Triangle Park, NC.
Publication No. EPA/600/9-88/026b. pp. 32-1 through
32-21.
39. Jung, H. J., et al. Vanadia/Ceria - Alumina Catalyst for
Selective Reduction of Nitric Oxide from Gas Turbine
Exhaust. Johnson Matthey, Catalytic Systems Division.
Wayne, PA. pp. 1 through 14. Undated.
40. Ref. 30. pp. 495-499.
41. Ellerhorst, R. And C.Edvardsson. Experience with NOx
Control at T.B. Simon CFB Boiler at Michigan Sate
University - Case History. In Proceedings: Council of
Industrial Boiler Owners Eighth Annual Nox Control
Conference. Philadelphia, PA. March 7-8, 1995.
pp. 55-67.
42. New Source Performance Standard, Subpart Da Technical
Support for Proposed Revision to NOX Standard, EPA-453/R-
94-012. U.S. Environmental Protection Agency, Research
Triangle Park, NC. October 1995.
43. Bay Area Air Quality Management District. Westinghouse
Electric Corporation Application Number 883, Steam
Boiler.
44. Oxides of Nitrogen (NOX as N02) and Ammonia
Emissions from a Boiler with Selective Catalytic
Reduction, Source Test Report 93-0114 conducted at
Darling Delaware Co. , South Coast Air Quality Management
District. September 23, 1993.
3-83
-------
45. Cho, M., Hannay, D.P., Khan, S., and Taylor, S.R.,
Operating Experience of a selective Catalytic Reduction
System for Flue Gas Denitrification in a Coal-Fired
Cogeneration Plant. Foster Wheeler Energy Corporation.
Paper No. SP95-3, March, 1995.
46. Wallace, A.J., et al, Selective Catalytic Reduction
Performance Project at Public Services Electric and Gas
Company's Mercer Generating Station Unit No. 2.
Presented at EPRI/EPA 1995 Joint Symposium on Stationary
Combustion NOX Control Book 3; Kansas City, Missouri,
May 16-19, 1995.
47. Southern California Edison Research Division, System
Planning and Research Department. Selective Catalytic
Reduction DeNOx Demonstration Test Huntington Beach
Unit 2. June 1988.
48. Cho, S.M., Design Experience of Selective Catalytic
Reduction Systems for Denitrification of Flue Gas.
Presented at the 1992 International Joint Power
Generation Conference Atlanta, Georgia, October 18-22,
1992.
49. Letter and attachments from Dorris M. Bailey, South Coast
Air Quality Management District, to Ben Dow, Darling-
Delaware Company, Incorporated, Permit No. D78848.
December 7, 1993.
50. Trip Report. Trip to the Carneys Point Generating
Station. James E. Eddinger, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina. June 9,
1994.
51. Letter and attachments from Iclal Atay, State of New
Jersey Department of Environmental Protection and Energy
to Jonine G. Kelly, Keystone Energy Service Company,
L.P., Amendment to PSD Permit for Logan (Keystone)
Generating Plant. March 1, 1993.
52. Letter and attachments from Iclal Atay, State of New
Jersey Department of Environmental Protection and Energy
to Richard Ciliberti, Keystone Cogeneration Systems,
Incorporated, PSD Permit for Logan (Keystone) Generating
Plant. September 6, 1991.
53. Letter and attachments from Patrick M. Tobin, U.S.
Environmental Protection Agency, Region IV to G.A.
DeMuth, Orlando Utilities Commission, PSD Permit
Modifications for Stanton Unit 2. March 2, 1993.
3-84
-------
54. Letter and attachments from Carol M. Browner, Florida
Department of Environmental Regulation to Stephen A.
Sorrentino, Indiantown Cogeneration. L.P. PSD Permit
Modifications to Indiantown Cogeneration Project.
July 16, 1992.
55. Letter and attachments from Katherine L. Miller,
Commonwealth of Virginia, Department of Environmental
Quality to Rachel Adams, Radian Corporation.
December 14, 1994. w
56. Trip Report. Plant Visit to Stanton Energy Center, James
A. Eddinger, U.S. Environmental Protection Agency,
Research Triangle Park, NC. December 8, 1994.
57. Letter and attachments from Jones, M.L., U.S. Generating
Company, to Eddinger, J.A., U.S. Environmental Protection
Agency. October 30, 1995.
58. Snedecor, G.W., and Cochran, W.G., Statistical Methods,
Sixth Edition. The Iowa State University Press. Ames,
Iowa. p. 552.
59. Wagner, P.A., Cook, G.S., U.S. Generating Company,
"Pulverized Coal-Fired Experience with SCR at the Logan
Generating Plant," presented at the ASME International
Joint Power Generation Conference, October 9-11, 1995.
3-85
-------
-------
4.0 MODEL BOILERS AND CONTROL OPTIONS
This chapter presents the model boilers used to assess
the environmental and cost impacts resulting from a revision
to the current subpart Db standard. Model boilers were
selected to represent the range of boilers potentially
affected. In addition, control options that were examined are
presented.
The selection of model boiler parameters is discussed in
section 4.1. Model boilers are presented in section 4.2, and
the control options considered are presented in section 4.3.
Environmental and energy impacts associated with each control
option are discussed in chapter 5.0. The cost impacts are
discussed in chapter 6.0.
4.1 SELECTION OF MODEL BOILER PARAMETERS
The selected model boiler parameters include fuel type,
furnace type, size, and capacity factor. These factors affect
one or more of the following: 1) baseline NOX emissions,
2) control system applicability, 3) control system
performance, and 4) control system costs. Table 4-1
summarizes the effect of each model boiler parameter. The
selection of baseline emission rates is also discussed in this
section.
4.1.1 Fuel Type
The following three fuel categories were chosen for the
model boilers: coal, residual oil, and distillate oil/natural
gas. Fuel type has a major impact on baseline NOX emissions,
primarily due to differences in combustion characteristics and
fuel nitrogen content. Generally, coal has higher NOX
emissions than residual oil, which has higher emissions than
distillate oil/natural gas. Control system performance and
4-1
-------
CO
c*
W
EH
*S
1
^
04
K
W
. n
*— 1
M
0
m
j
w
D
§
b
0
CO
B
u
h
fc
W
H
i
«*
H
. T
1— 1
59
<
EH
§^
•3eo>;
§5^2
S5.3«
8°1§
«J rj
rH e r-t 3
p | rt w 4J
U * H «@
c 5, o. 8 §
o g1 * -v-
00 2
0)
u
sM
8«1
C r. O
O 5?«w
U ffl H
0)
o
!*
R3
m^
-1 c
s"
M _.
is
•o
• JS-Sg
-5o§2'£
.-1 H C g <"
Sc^Sfl
S o Kgd
« u o g
s
>— *
•o c
Q) O ^
0) rH -H 3
Cr-l W AJ
•rl O ffl 0) 0
rH ^1 -H 4J I
0) 4J 6 «J g
OQ C 0) M -•»
(DO jQ
"M-
H
d
4->
(
II
ft
Pk
X X X X
XXX
XX X
X X
X X X X
X X
0) 0)
Q! N J>i
K 0) -H 4J
H U W -H
id u
rH C 0) 4J Id
o) H a -H a
3 P S C id
fc fc ft D U
4-2
-------
costs are sensitive to baseline NOX emission rates. The
specific effects of baseline emissions on environmental,
energy, and cost impacts are discussed in chapters 5.0 and
6.0.
4.1.2 Furnace Type
Based on available information concerning the types of
industrial boilers currently in use, three coal-fired- model
boiler types were selected.1 These include pulverized coal
(PC), fluidized bed combustion (FBC), and spreader stoker (SS)
boilers. Two model boiler types were selected to represent
residual oil-fired boilers and distillate oil/natural gas-
fired boilers. These include packaged watertube (single-
burner) and field-erected water tube (multiple burners).
Packaged firetube boilers were not considered because these
are typically smaller than the 100 million British thermal
units per hour (MMBtu/hr heat input) cut-off for the
subpart Db standard.
4.1.3 Boiler Size
A wide range of boiler sizes (100, 250, 500, and
1,000 MMBtu/hr heat input) was used to define PC, FBC, and
field-erected oil- and gas-fired model boilers. Due to their
smaller capacities, spreader stoker boilers were modeled in
three sizes (100, 250, and 500 MMBtu/hr), and packaged oil-
and gas-fired boilers were modeled in two sizes (100 and
250 MMBtu/hr). The 500 MMBtu/hr cut-off for spreader stoker
boilers was based on vendor data.2
Boiler size affects annual baseline NOX emissions and
control system costs. Larger boilers emit more NOX per year
and have higher absolute capital costs for control systems
than smaller boilers. However, capital costs per MMBtu of
heat input capacity decrease as boiler size increases due to
economy of scale. In addition, some control system operating
and maintenance costs tend to vary directly with boiler size
(e.g., ammonia for a selective noncatalytic reduction [SNCR]
system), while others are less directly affected (e.g.,
operating labor).
4-3
-------
4.1.4 Capacity Factor
Three capacity factors (10, 30, and 60 percent) were used
to define model boilers. Capacity factor affects annual
baseline NOX emissions, NOX control system performance, and
annual costs. A boiler operated at a low capacity factor will
have lower annual baseline NOX emissions than a similar boiler
operated at full load. Variance from the load for wHich the
control system was designed to perform can affect performance;
however, the cost algorithms may not reflect this. The cost
per ton of NOX removed is higher as the capacity factor
decreases, due to the reduction in the amount of NOX removed.
4.1.5 Baseline NOX Emission Rates
Average annual baseline NOX emission rates were based on
emission limits from the current subpart Db standard. As
discussed in chapter 2.0, the average annual baseline NOX
emission rate is 0.02 Ib/MMBtu less than the applicable NOX
emission limit from the current subpart Db standard. The NOX
emission limits based on the current subpart Db standard are
shown in table 2-10 of chapter 2.0. The average annual
baseline rate for FBC boilers was based on vendor data
supplied in the subpart Da revision BID.3 Field-erected
boilers were assumed to have a low heat release rate and
packaged boilers were assumed to have a high heat release
rate. Based on this methodology, the average annual baseline
emission rates for PC, FBC, spreader stoker, residual oil-
fired field erected, residual oil-fired packaged, distillate
oil/natural gas-fired field-erected, and distillate
oil/natural gas-fired packaged model boilers are 0.68, 0.38,
0.58, 0.28, 0.38, 0.08, and 0.18 Ib/MMBtu, respectively.
4.2 MODEL BOILERS
A total of 22 model boilers reflecting the parameters
discussed in section 4.1 are presented in table 4-2. These
boilers represent four different fuels, five different furnace
types, and four different sizes. Each of these models was
evaluated at three capacity factors (0.10, 0.30, and 0.60) to
4-4
-------
(t
CO
OS
w
rH
H
O
Q
g
(N
W
rH
S3
Jj
(0
V-4
c
-2*3
50 ffl
'H 1
cu"^
•Q
cu
E~"
•H
i-H
CU
CO
n
IH
si
N 3
-H 4J
co pq
5
cu
pt
pi
4J
rl
0)
i-H
-H
O
ca
cu
rH
0)
b.
CO CO CO
VD ro in
O 0 0
00° o o o 9 ooo
in o X o in o ° oino
01 in j^ H CN in ^ I-H 01 in
U
§
B
0
•H
4J
^ CO
U 3
CU S\
rH U rl
10 CU
o TJ j<:
0 JS S
•O to
CU T3
NO) rl
-rl N CU
co -a «s
> -H 0)
rH 3 H
3 rH ft
o< ti< co
10
0
u
CO CO
CN ft
O O
000° o o
o in o ;; o in
H tN in J^ H CN
cu
f^
^
4J
^1
CU CO
4J Q
10 3
S 4J
T3 0)
0) 4-1
4J 10
U >
0)
rl T3
CU CU
' tn
•a <0
CO U
-H (0
1
"o
H
10
•o
•rH
CO
&
CO CO
0 H
0 0
o o o 2 o o
o in o 2 o in
rH CN in 2 "-• M
(1)
3
4->
i-l
O 0)
•u S3
10 3
S 4J
•a eu
CU 4->
U >
>H "O
CU 0)
i tn
2 JS
CU U
-H 10
[T! PJ
CO
8.
.
fl) (d
^j tj
10 3
i-H 4J
rH 10
•H C
CO i-H
-rl -H
a o
o
1H
4J
a
CU
o
X
cu
T)
• 10
~-o
o c
VD 10
• 4J
o m
•O S3
S
4-1
O (0
o 3
CO
*
0 4J
H C
• CU
0 H
— M
CO U
O CU
4J f"J
U 4-1
10
-rl
U co n
(0 CO
O< cu cu
(0 rH U
U C
3 0)
CU 4J rl
CU PQ CO
rl § *W
rj § d)
4J ^^ p^
X!
cs
•O O TJ
0) • CU
4-1 O CO
(0 10
3 COX3
1-H rl
(0 n) cu
> 4J
cu to m
CU rl
CU 4J
XI (0 B
rl 0
rH -rl
i-H C M
-H o tn
£ -H -H
CO £
iH tO CU
CD-rl
I-H g eu
-H fl) B
O -H
S3 XrH
O CO
I-H 2 n
0) 10
"Eg-0
B-H u
rH PQ
rC fl) r*i
u to
10 10 0)
PQ OQ »G
4J
10 XI
4-5
-------
represent intermittent/seasonal use, cycling, and baseload
operation.
4.3 CONTROL OPTIONS
Three control options were considered for the revision to
the current standard: combustion controls (CO, CC with SNCR,
and CC with selective catalytic reduction (SCR). Table 4-3
depicts the control technologies that were evaluated ^for each
model boiler. Because there are no known applications of SCR
on FBC boilers and because of technical concerns that the high
alkalinity of fly ash from FBC boilers may poison the
catalyst, SCR was not considered for FBC boilers. Table 4-4
lists baseline and controlled NOX emission rates for each
control technology. The emission levels associated with each
control technology are discussed in the following sections.
4.3.1 Combustion Controls
Combustion controls were considered for all five model
boiler types. Low-N0x burners (LNB) were considered for PC-,
oil-, and gas-fired boilers. Staged combustion air (SCA) was
considered for FBC and spreader stoker boilers. For coal-
fired boilers equipped with CC, NOX emission levels shown in
table 4-4 were based on analysis of long-term continuous
emission monitoring (CEM) data obtained from utility boiler
owners.4 For oil- and gas-fired boilers, CC levels were based
on a Best Available Control Technology (BACT)/Lowest
Achievable Emission Rate (LAER) search5 and the industrial
boiler Alternative Control Technology (ACT) document.6
4.3.2 Selective Noncatalytic Reduction
This control option is applicable to all the model
boilers. The specific SNCR system considered here is a low-
energy, urea-based system with two rows of wall injectors. A
normalized stoichiometric ratio of 1.0 was used. The injected
urea solution was assumed to be 10 percent urea by weight. A
constant emission reduction of 50 percent over CC levels was
assumed.
4-6
-------
CO
H
O
PQ
Q
g
O
fc,
CO
S3
o
H
H
o
o
u
S3
o
H
CO
CO
H
s
w
X
o
^
w
fCi
CO
CU
tn
o
rH
O
d
.C
u
CU
4J
rH
o
4J
a
o
CJ
CU
pj
4-1
CU
rH
•H
o
PQ
CU
g
rH
CU
Pu
Cd *s^ PH p4 f*J
CO W CO CO W
u + H- cj+ cj + +
0 cj o o cj
0 U CJ CJ U
a
0
•H
4J
^ CD
U 2
o
H Or*
rt cu
o -o ,*
u cu o
•a co
CU TJ
N CU M
•H N CU
$H *H *&
0) T3 ^ (Ci
> -H CJ CU
rH P PQ M
3 rH fri Qj
DJ fe — CO
rH
m
o
CJ
fV fVJ fV^ ^
cj cj y g
S3 CO S3 ^
CJ + + 0 + +
CJ CJ CJ ^
CJ CJ U
0)
Q
s
4J
^1
cu cu
4J ^Q
nj ^3
^ t^
T3 (!)
CU 4J
4J (t!
U >
0
r< T>
CU CU
tn
TJ fO
CU CJ
-H nj
CLI CU
rH
•H
O
rH
fC
•o
-H
CO
CU
PH /^j PH *o
^ cj y u
Oco w o ww
CJ + H- CJ + +
CJ U CJ U
CJ U CJ U
cu
Q
3
j_i
cu cu
Jj _Q
ftS 3
* £
T3 OJ
CU 4->
4-> (t
U >
0)
r) "O
cu cu
' en
"O nS
rH ,^
CU CJ
•H ftJ
Cu PJ
rH
CJ
-H
O
CU
C
CJ O
3-H
T) 4J
CU CJ
rl 3
•o
U Q)
-rt ^1
• 4J
CO >i O
rH rH -H
O (ti 4J
r) 4-> >,
4J (0 rH
a o us
o a u
u o >
4J -H -H
CO J-> 4->
3 O U
5 CU CU
6 rH rH
O 0) CU
cj co co
II II II
CJ tf
CJ S3 CJ
CJ CO CO
(ti
4-7
-------
w
i
0)
JJ
U rH *"" JJ
W rH g Q
+ iJ "H •
o\ CT\
H r-l
0 0
GO GO
M m
0 0
0)
3
JJ
0) 0)
II
^
•a a)
4) JJ
jj id
u >
0)
V4 T3
0) 0)
' 01
•a *
rH A!
0) U
•rt Id
U. ft
-H
0
rH
id
*o
-H
D)
(D
a
03 cs
O 0
O 0
•* VD
O 0
o o
o -o
00 (N
O H
0 O
GO GO
0 H
O O
Q>
3
JJ
0) 0)
JJ XI
id 3
* JJ
y ]
•O 0)
0) JJ
jj (d
u >
Q)
o> a>
| Oi
rH fl^
0) U
-H id
n
Oi
rH
0) Id
It Ij
id 3
rH JJ
rH Id
•H C
CO rH
•H -H
Q 0
vo
V
g
•Q ^
id to
n
a>
•u
ft 0)
S1
8 S
u
-H
g
C
0)
M
0)
Q}
ower emission rate
uggest
O
JJ
O
C
•o
0)
m
id
•O
cu
ca
-------
4.3.3 Selective Catalytic Reduction
This control option was considered for all the model
boiler types except FBC boilers since it has not been
demonstrated on FBC boilers. Catalyst life was assumed to be
3 years for coal-fired boilers and 6 years for natural gas-
and oil-fired boilers. Space velocity was assumed to be
3,200 per hour for coal-fired boilers, 5,000 per hour for oil-
fired boilers, and 14,000 per hour for distillate oil/natural
gas-fired boilers. A normalized stoichiometric ratio of 0.82
was used. A constant NOX emission reduction of 80 percent
over CC levels was assumed.
4-9
-------
4.4 REFERENCES
1. Alternative Control Technique (ACT) Document -- NOX
Emissions from Industrial/Commercial/Institutional (ICI)
Boilers. Publication No. EPA-453/R-94-022. U.S.
Environmental Protection Agency. Research Triangle Park,
North Carolina. March 1994. p. 3-7 to 3-33.
2. Telecon. Rachel Adams, Radian Corporation with traig
Arcari, Riley Stoker Corporation, December 6, 1994.
3. New Source Performance Standard, Subpart Da - Technical
Support for Proposed Revisions to NOX Standard,
Background Information Document (BID). EPA-453/R-94-012.
U.S. Environmental Protection Agency. Research Triangle
Park, NC. October 1995. p. 4-6.
4. Ref. 3, pp. 3-21 to 3-115.
5. Best Available Control Technology (BACT)/Lowest
Achievable Emission Rate (LAER) Search, November 1994.
6. Ref. 1, p. 5-79, 5-80, and 7-3.
4-10
-------
5.0 ENVIRONMENTAL AND ENERGY IMPACTS
Chapter 5.0 presents the environmental and energy impacts
associated with the application of combustion controls (CC),
CC + selective noncatalytic reduction (SNCR), and CC +
selective catalytic reduction (SCR) on the model industrial
boilers. In section 5.1, the incremental air pollution
impacts associated with each control option are discussed.
Sections 5.2 and 5.3 deal with liquid waste and solid waste
impacts, respectively and section 5.4 discusses the energy
impacts associated with each control option.
5.1 AIR POLLUTION IMPACTS
Air pollution impacts include primary impacts associated
directly with each control option (i.e., nitrogen oxides [NOX]
emission reductions) and secondary impacts, which include the
increases or decreases in the emission levels of other
pollutants caused by the application of a specific control
option (e.g., increases in ammonia (NH3) emissions associated
with the application of SNCR and SCR).
5.1.1 Primary Air Impacts
The NOX emission reductions achieved by the application
of each control option to the model boilers are presented in
appendix A.I. For each fuel and furnace type, the annual
baseline NOX emissions and the emission reductions associated
with each control option were calculated based on emission
rates and NOX reduction efficiencies presented in chapter 4.
The primary air impacts for the model industrial boilers
are summarized in table 5-1. The data are presented for
boilers operating at a capacity factor of 0.30. The size
ranges are 250 to 1,000 million British thermal units per hour
(MMBtu/hr) for pulverized coal boilers; 100 to 1,000 MMBtu/hr
for fluidized bed boilers (FBC) and field-erected watertube
5-1
-------
I
in
1
n
§
3
TJ
0)
K
0
CD
n
E
W
^
O
^
C
4-
t
Ll
O
C
e
0
4J
O
Ll
0)
O.
XI
chnolog
0)
E-t
A)
Q
tr
t^
0)
1
^
ik!
s
H
"o
2
00
to
1
CM
o\
8
to
1
00
in
^^
r~
tf
8
^"
8
rH
id
S
TJ
01
N
•H
Vj
0)
J>
•H
£
00
00
1
£
^»
oo
g
^"
o
o
ro
1
CM
O
O
00
to
«*>
1
Jn
«
O
O
e
o
4J
m
3
XI
O
TJ
0)
m
Tl
0)
•H
TJ
•^t
3
iH
h
1jj
O
O
•!
^*
6
V)
8
in
I
r-l
.H
O
o
CM
1
m
in
1
8
Ll
0)
O
4-1
Ll
0)
TJ
id
0)
Ll
a
(0
in
tH
1
en
to
00
e
w
8
00
f-l
1
CM
i-l
CM
n
O
o
to
CM
1
CM
to
to
1
w
o
o
0)
X)
3
4J
Ll
0)
4J
Id
S
TJ
0)
4J
O
0)
Ll
0)
1
TJ
rH
0)
00
tH
1
CM
CO
\o
00
o
ca
O
o
CM
VO
1
in
CM
in
O
O
p-i
•H
0
r^
«
3
TJ
•^
a
a>
i
n
in
OS
o
V)
8
0)
XI
3
4J
Li
01
4J
id
S
*D
Q)
o
a
jV
U
£
CM
tH
1
in
a>
o:
O
U)
8
O
8
in
i
in
in
1
(0
8
Jg
3
4J
Ll
fl)
4J
1
TJ
0)
4J
O
0)
Ll
0)
1
TJ
01
fe
00
1
00
00
g
o
o
o
CM
1
00
n
8
a
id
>H
Ll
3
i
•»..
rH
^4
O
0)
rH
rH
4J
(Q
D
»
1
to
«H
•^
r^
to
en
(A
O
O
0)
XI
3
Ll
0)
4J
a
TJ
0)
Oi
id
y
u
id
cu
r-t
in
1
o
CM
p*.
00
o
8
o
o >K>
10 rH-
CM K)B
00
050
•HLi
•• TJ H *> fli
« 0) U>
§4J TJ 0>0
O 0) «-l
-H 0> OiQ>«
-I U * WU
O fl) X 2
Vl I O II tO
m >-i ft«+
(BO) U
-I — WO
01 «n Li O
M £ TJ
. CLi
C S MO
a « 4J *O
ro ^Li
• TJ LI «Q)
« •- n) -H a
H-t 1-1 0) 4J*
£ Ll OXl
*J 3 B
•H *J fl)4J
O 00 — M
«B £ U 0)
as & M >
« — , M
U O 3 «*J
o -u Ort
mom Z-H
0)0)
•>£
^H 2
I O -1C df
0) O OO) O
MOO M> 00
« in >4>v4 •
CSH CO»B) 0)
TJ O -^ ^
at H i oc ja
•U O O *
C -I o C^ O O
0) « o o-y ~4
U O.»-t -^O U ^H
0) U -UD 0)
flJTJ >
TJ 0) Ll
> 01 II O +
Q)
-------
boilers; 100 to 500 MMBtu/hr for spreader stoker boilers; and
100 to 250 MMBtu/hr for packaged boilers. The NOX controls
considered for these boilers are CC, CC + SNCR, and CC + SCR.
However, as indicated in section 4.3, SCR was not considered
as a control option for FBC boilers.
The percent NOX reduction in table 5-1 is given relative
to baseline emissions. Because the NOX emission levels
associated with CC on field-erected, distillate oil-/natural
gas-fired boilers are the same as baseline levels, the percent
NOX reduction is zero. As seen in table 5-1, NOX reductions
for CC range from 0 to 50 percent, and from 0 to 368 tons/yr
depending on furnace type and size. For CC + SNCR, NOX
emission reductions range from 50 to 75 percent, and from 5 to
631 tons/yr. Finally for CC + SCR, NOX emissions reductions
range from 80 to 90 percent, and from 8 to 788 tons/yr.
5.1.2 Secondary Air Impacts
Secondary air emissions associated with NOX control
technologies include NH3 slip, nitrous oxide (N2O), and carbon
monoxide (CO). For pulverized coal-fired and stoker boilers,
baseline emissions of NH3, N2O, and CO were estimated at 0
parts per million (ppm), 5 ppm, and 20 ppm, respectively (each
at 3 percent oxygen [02]). For FBC boilers, baseline N2O
emissions were estimated at 100 ppm (at 3 percent ©2) based on
the measurements of N20 emissions from these boilers.
Emissions of NH3 and CO were assumed to be the same as for
pulverized coal-fired and stoker boilers. For oil- and gas-
fired boilers, baseline emissions of NH3, N2O, and CO were
estimated at 0 ppm, 5 ppm, and 5 ppm, respectively (each at 3
percent ©2).
Available data from utility boilers retrofit with CC have
shown increases in CO emissions. However, CC is not expected
to increase CO emissions from new boilers that are designed to
minimize combustible emissions such as CO. Emissions of NH3
and N2O were also assumed equal to baseline levels. Hence,
5-3
-------
secondary air impacts for CC were assumed to be zero for all
model boilers.
The addition of urea-based SNCR control to CC can result
in increases of CO, NH3, and N2O. Secondary impacts from
industrial boilers are expected to be similar to those from
utility boilers and based on information in the Subpart Da
2
Technical Support Document (TSD) , increases in NH3,»N2O, and
CO emissions were estimated at 25 ppm, 10 ppm, and 25 ppm,
respectively (each at 3 percent 02) for all model boilers.
However, if NH3 is used as the SNCR reagent, increases in CO
and N20 emissions are expected to be much lower.
For CC + SCR, increases in NH3 emissions leaving the SCR
reactor were estimated at 2 ppm (at 3 percent ©2) over CC
levels based on information in the Subpart Da TSD . Emissions
of N2O and CO were assumed equal to CC levels.
Based on these values, the estimated annual emissions of
NH3, N2O, and CO resulting from the application of each NOX
control option to each model boiler are presented in
appendix A.2. The annual emissions were calculated using the
F-factors contained in EPA Method 19 (40 CFR Part 60).3
Table 5-2 presents ranges of annual tons of NH3, N2O, and CO
emissions for model boilers operating at a capacity factor of
0.30. As seen in the table, the annual tons of pollutant
emissions for CC are identical to baseline: the NH3 emissions
are zero, N2O emissions range from 1 to 171 tons/yr, and CO
emissions range from 0 to 22 tons/yr in each case. For CC +
urea-based SNCR, NH3 emissions range from 1 to 17 tons/yr; N2O
emissions range from 2 to 188 tons/yr; and CO emissions range
from 3 to 49 tons/yr. The elevated level of N2© emissions is
the result of inherently high baseline and CC levels for FBC
boilers. The addition of SNCR to FBC boilers has little
effect on N20 emissions from these boilers. For CC +.SCR, NH3
emissions range from 0 to 1 tons/yr; ^O emissions range from
1 to 9 tons/yr; and CO emissions range from 0 to 22 tons/yr.
5-4
-------
(0
CO
H
O
CQ
CO
D
D
2;
Q
O
CO
^
U
s
CM
I
in
CQ
10
c
o
»«J
•^1
CQ
W -^
*^H ^
CO
^ o
10 *J
o
Q)
CO
f"
c
}.
-4-
c
c
c.
o
CJ
O
CN
r
a
1 ^^
> 0
< «H
> O
: c
> .c
> o
0)
6-1
0)
D,
Jx,
EH
Q)
U
(0
c
,4
3
tn
0)
>«
E"1
H
(U
3
CM
CM
CM
1
rH
rH
«
1
CM
0
0)
c
•H
rH
<0
CQ
10
CQ
CM
CM
1
H
rH
cn
1
CM
O
U
CJ
f
(
'
*
r
i
TT
1
CM
rH
VO
CM
1
VO
rH
1
*
K
CJ
S5
CO
^.
CJ
o
H
(0
0
J
0
0)
N
H
tl
1)
H
3
ll
CM
CM
1
in
»
i
CM
rH
1
O
u
CO
+
0
CJ
CM
CM
|
CM
1 H
t> J^
O
Q)
C
-H
fB^
0)
CQ
(0
CQ
CM
CM
1
CM
' rH
r-1"
o
o
u
•O
0)
CQ
•O
(D-
N
•H
•o
•H
iH
fe
r"
n
C
C.
o\
1
in
1 CO
o\ "
H
1
CM
8
25
CO
+
CJ
CJ
c
o
H
p
(Q
3
a
e
o
u
H
3
)
>
2
s:
0
5Jj
CJ
CO
^-
CJ
o
H
H
1
CM
«
1
H
O
0)
C
•H
rH
0)
(0
10
CQ
H
H
1
CM
*
1
rH
O
u
u
in
CM
i
in
n
H
1
CO
CO
1
CM
1
CO
-f-
CJ
CJ
J_l
a)
*
o
P
CO
j_i
0)
73
(0
0)
>_i
a
CO
rH
1
CM
*
1
H
H
1
O
I
K
CJ
CO
.)_
CJ
CJ
in
1
H
CO
1
rH
O
0)
c
-H
rH
0)
(0
(0
CQ
in
>
rH
CO
1
H
0
CJ
CJ
•o
0)
4J
O
a)
01
1
rH
0)
•H
tn
H
CO
1
CO
CM
1
CM
VO
rH
1
CM
1
CO
-J.
CJ
CJ
Q)
jj
+J
a)
4J
(0
in
1
H
CO
1
H
H
1
0
a
CJ
CO
4.
CJ
CJ
r
1
i
i
H
CM
1
H
O
Q)
C
•H
rH
(U
(0
10
CQ
H
H
O
H
(0
3
O
H
CQ
Q)
X
H
CM
1
H
O
CJ
CJ
•
4
I
r
*
i
CO
1
CO
VO
1
CM
*
1
CM
1
CO
4.
rj
W
0)
a
tj
u
P
10
£
o
1)
10
*
u
(0
X
*
e*
i
rH
V
0$
U
CQ
+•
U
O
5-5
-------
D
I
in
H
r?
I
CO
G
0
•ri
CO
to —.
l£
CO
)H O
<0 4->
•O «—
u
0)
CO
8
o
2
r
2
rH Cp
o o
U-3
c c
O X!
O O
Q)
**
Q)
a
EH
Q)
U
(0
C
>H
s
0)
a
EH
a
fa
in
i
o
00
rH
O
Q)
-H
0)
(0
fl)
03
in
l
o
CO
H
o
u
u
o\
1
c*i
CO
OJ
1
in
rH
1
H
8
2
CO
4.
u
0
in
•
0
00
rH
H
1
O
8
CO
+
u
CJ
•o
(U
4J
0
Q> Q>
M42
0) 3
\ 4->
•O M
rH 0)
0) 4->
•H (C
fa S
H
1
0
ra
rH
O
Q)
•H
rH
0)
CO
<0
CQ
rH
1
o
CM
1
H
O
U
u
I
n
vo
CM
•«*
1
*"*
1
CO
H-
u
o
rH
1
O
CM
l
rH
V
tf
O
CO
4.
o
CJ
A
3
4J
CO
4J
(0
T5
Q)
cn
(0
J^
0
(0
rH
(0
(0
z
-^
2 w
4J°
(0
rH
rH
-H
4J
(0
-H
Q
•o
.. CD
CQ 4J
» o
O Q)
«J Ij
rH Q)
O 1
«WO
_J
CO 0)
<0 *H
MH *^
0) V4
MTJXJ
«J C ^.
10 9
CQ 4J
OJ C CQ
N O £
CQ 4J
CQ O
k 3 O
a» JQ m
•H E
•H O 1
0 0
m o
o
. *° H
0 X) II
CO
• *CJ ^4
0 Q) (U
N ,&£
H O
O TJ 4J
-H IQ
IH 3
O rH M
4-^ (M CD
O *O
(0 •- (0
«M p 0)
•C JH
^^~« D<
4-> 3 W
-H 4J
O CD •*
* S r)
agx;
(0 -^ •
O 0 3 >H
0 4J X!
(0 O CQ *^
^ S rJ
4J rH g 4J
<0 CQ
1 0 E
0) 0 g
V4 0 0
o
CM H in
•d CM
CP
•O 0) r< (0
> Q) ^
Q) rH 4J O
X! 3 <0 (0
EH CU SB &
-H
4J
O
0)
rH
Q)
CO
H
05
U
CO
•a
c
10
• »
c
o
•H
4J
o
3
T3
0)
a
o
-H
4J
rH
(0
4-1
(0
U
C
o
2
0)
^
•H
4J
0
CO
rH
Q)
CO
II
Oi
C,,)
55
CO
• ^
rH
O
M
4J •
C C
0 0
4J
C U
O 3
•fH *O
4J Q)
n a
Xt U
0 *J
U >i
rH
II (0
4->
U (0
u u
*
a>
rH
<7
(0
u
,— i
a
04
4J
0
II
rtj
2
(0
O
5-6
-------
5.2 LIQUID WASTE IMPACTS
The use of NH3 for SNCR and SCR systems on boilers
equipped with flue gas desulfurization (FGD) units can
increase the ammonia content of the aqueous discharge streams
exiting the FGD unit. This problem can be significant at high
NH3 slip levels. Because the expected NH3 slip levels are
relatively low for the SCR and SNCR control options .being
considered, the water pollution impact resulting from the
application of these control options on the model boilers is
minimal.
5.3 SOLID WASTE DISPOSAL IMPACT
The application of CC for NOX control is not expected to
increase the quantities of solid wastes generated. However,
the potential exists for changes in the characteristics of
solid waste that can impact its disposal. Data on retrofit
applications of CC on coal-fired boilers have shown increases
in unburned carbon (UBC) levels in the fly ash and bottom ash.
However, for new boilers, CC can be designed to ensure
complete combustion of fuel, thus minimizing the UBC content
in ash. Therefore, no incremental impact on solid waste
disposal is expected as a result of using CC.
Besides the direct emission of NHs from SNCR or SCR
systems, NH3 can absorb onto the fly ash. Absorbed NH3 could
present ash handling and disposal problems or prevent the sale
of fly ash to cement producers that may specify upper limits
on the amount of NH3 in ash. This problem is more significant
with SNCR than with SCR because of higher unreacted NH3 levels
in flue gas.
For SCR systems, disposal of the spent catalysts can be
of concern. Some catalysts contain vanadium pentoxide (V^Os),
which may be considered a hazardous chemical in some States.
However, catalyst manufacturers are working on improved
methods for reactivation and recycling of spent catalyst as
well as recoating the catalyst support. In Japan, where there
are a number of SCR applications in which raw materials such
5-7
-------
as titanium dioxide (TiC>2) are being recovered from the spent
catalysts.
5.4 ENERGY IMPACTS
The energy impacts for each NOX control option on the
model boilers are presented in appendix A.3. The impacts
reflect the additional energy requirements for each NOX
control option over baseline levels. Energy impact ranges are
expressed in units of MMBtu/yr and as a percentage df boiler
energy output. The equations used to estimate the energy
impacts are discussed later in this section. The results for
the model industrial boilers operating at a capacity factor of
0.30 are summarized in table 5-3.
As shown in the table, the additional energy requirements
for CC are zero. Energy impacts resulting from CC + SNCR on
pulverized coal-fired boilers range from 1,590 to
6,350 MMBtu/yr corresponding to 0.28 percent of boiler output.
For spreader stoker boilers, energy impacts are much lower and
range from 740 to 3,690 MMBtu/yr (0.33 percent of boiler
output). For FBC boilers, the energy impacts are 430 to
4,300 MMBtu/yr corresponding to 0.19 percent of boiler output.
Additional energy requirements for residual oil-fired
boilers range from 420 to 4,190 MMBtu/yr (0.19 percent of
boiler output) for field-erected watertube boilers and 420 to
1,050 MMBtu/yr (0.19 percent of boiler output) for packaged
watertube boilers. For distillate oil- and natural gas-fired
boilers, the additional energy requirements range from 310 to
3,060 MMBtu/yr for field-erected watertube boilers
(0.14 percent of boiler output) and 350 to 870 MMBtu/yr for
packaged watertube boilers.
Energy requirements for SCR systems are higher than for
SNCR. As shown in table 5-1, the energy requirements range
from 2,310 to 9,250 MMBtu/yr (0.41 percent of boiler output)
for pulverized coal-fired boilers. For spreader stoker
boilers, the energy impacts are much lower and range from 920
to 4,620 MMBtu/yr, corresponding to 0.41 percent of boiler
5-8
-------
(0
O
n
H
a
EH
Q
O
at
H
u
•<
cu
s
H
X
tn
CQ
o
4J
o
id
§
tr
Ml
e
C
t
!l
^
• ,
M
^**
3
nrt
H
5
*•
lontrol
:hnologyb
01
H
S
£
01
id
u
fkj
S
H
^H
01
3
m
o
o
8
oo
CM
O
o
in
ro
vo
O
ff>
in
^
i-H
6
&
01
+
o
rH
*
O
0
in
CM
ta
9*
1
O
rH
ro
«
CM
e
01
+
O
U
r-l
id
0
O
•o
0)
N
^H
^
0)
rH
3
(X
O
o
8
rH
O
0
o
ro
^
*
1
0
ro
^
01
+
U
O
g
U
jg
e
01
+
o
o
•o
Q)
(Q
C
IS
N 41
•H m
•O 3
3 "I
«H O
fc U
o
o
8
ro
ro
O
O
9>
*
ro
1
O
^»
f**
1
01
+
O
O
rH
*
o
o
CM
VO
«
«
1
o
CM
91
6
01
+
O
O
01
0
4J
01
0)
•o
«
0)
o<
u
id
o
0
o
o
o
o
rH
O
o
rH
—
<*
1
O
CM
^*
1
01
+
O
u
in
m.
O
o
o
00
*
**
1
o
00
r*
6
01
+
U
U
0)
3
U
0)
4J
id
s
TJ
0)
4J
o
0)
Ll
0)
1
•o
rH
01
t,
o
o
o
*t
o
o
m
o
ta
rH
1
O
CM
^
U
01
+
O
u
in
ro
O
O
in
9)
—
rH
1
O
ao
r-
6
01
+
8
.§
3
U
0)
-p
id
^
•c
0)
o>
id
u
a
OU
o
id
3
•o
O
0)
K
o
o
o
u
"^
0
o
vO
0
ro
1
O
rH
ro
OS
U
0}
+
O
u
in
n.
O
O
O
00
*,
^
1
o
00
r»
6
0]
+
8
0)
3
4-1
01
4J
id
S
*o
Q)
4J
U
2
0)
I
•o
rH
0)
fc,
0
o
8
vO
rH
O
o
ao
o
in
ro
1
01
+
U
u
in
ro
O
O
in
O>
rH
1
O
oo
t^
6
0)
+
8
0)
3
0)
-P
id
^
•o
0)
D»
J^
U
a)
CL.
a
id
rH
a
id
^J
•H
o
01
4J
id
rH
rH
e>
Q
H
0)
3
4J
C
o
4>
U
I
a
4J
•••ox
a> oi
S 41 0
5 u m
rH 0) CM
rH rl
O 01 1
•o o
B rH O
Id 0) rH
~4
0)
Ll O •»
« in LI
CM.C
•a -^
0) II 3
4J 4J
CrHCQ
01 * X
max
0) U
Lt O
ftTJ O
oio
« N -
4J -rl rH
id LI
TJ 0) 1
>
0) rHO
J3 3 O
H CUrH
4J
U
0)
rH
01
n
»
05
y
01
TJ
id
«fc
c
o
4J
u
3
O
K
U
4J
>i
id
4J
«
U
e
o
at
>
^j
u
r-l
at
*
05
CJ
z
01
•^
rH
o
Ll
4J
C
8
c
^
^
a
3
jQ
£
Q
^
n
o
.0°
•
e
^-»
^
u
•«4
rH
a,
cw
^
4>
0
s
It
<
z
5-9
-------
output. As discussed previously, SCR was not considered an
option for FBC boilers. \
For residual oil- and natural gas-fired boilers, the
additional energy requirements range from 780 to
7,800 MMBtu/yr for field-erected watertube boilers
(0.35 percent of boiler output). Additional energy
requirements for residual oil- and natural gas-fired boilers
range from 780 to 7,800 MMBtu/hr for field-erected w'atertube
boilers and 780 to 1,950 MMBtu/yr for packaged watertube
boilers (0.35 percent of boiler output).
The equations used in estimating energy impacts for each
NOX control are presented in the following paragraphs. There
are no additional energy impacts associated with CC. For SNCR
systems, boiler efficiency can be affected, due primarily to
the heat loss associated with the vaporization of water
injected with the reagent. This results in a net loss of
useable energy, although a portion of this penalty is offset
in urea-based SNCR systems by the exothermic oxidation of the
reagent. For a 10-percent solution of urea in water, the 1
useful thermal energy loss is calculated with the following
equation:
Energyv (MMBtu/yr) = Urea * [-4654 + 9 * (1183)] *
8760* CF * (0.85)/106 (5.1)
where:
Urea = Pure urea injection rate (Ib/hr)
calculated from equation 5.2.
-4654 = Heat of urea oxidation reaction (Btu/lb
urea)
9 = Ib H2O/lb pure urea
1183 = Heat of vaporization of water (Btu/lb H20)
CF = Capacity factor (decimal fraction)
8760 = Conversion factor; hr/yr
0.85 = Boiler efficiency
106 = Conversion factor; Btu/MMBtu
The urea injection rate is calculated from the following
equation: ->
5-10
-------
Urea (Ib/hr) - UncNOx * NSR * 0.5 * (60/46)
* size (5.2)
where:
UncNOx - NOX emission rate at inlet to SNCR system,
i.e., after combustion controls (Ib/MMBtu)
NSR = Normalized stoichiometric ratio (NH2:NOX);
assumed to be 1.0 for 50 percent NOX
reduction k
0.5 = Stoichiometric ratio (urea:NH2)
60 = Molecular weight of urea
46 = Molecular weight of NOX
size = Boiler capacity (MMBtu/hr)
For urea-based SNCR systems, there are additional
electricity requirements associated with injection of reagent
in the flue gas. Based on the SNCR case studies performed for
utility boilers, this power requirement was estimated at 1
kilowatt (kW) (electrical) per megawatt (MW) of boiler
electrical output. Assuming a heat rate of 10,000 Btu/kWh,
this is equivalent to 0.1 kW per MMBtu/hr of heat input.
Based on this, the following equation is used to estimate the
thermal equivalent of this energy requirement expressed as
MMBtu/yr of boiler thermal output:
Energyc (MMBtu/yr) = 0.1 * size * CF * 8760/0.40
* 0.003412 (5.3)
where:
0.1 = Conversion factor, kW per MMBtu/hr
size = Boiler capacity (MMBtu/hr)
CF = Capacity factor (decimal fraction)
0.40 = Conversion efficiency of thermal energy to
electrical energy
0.003412 = Conversion factor; MMBtu/kWh (thermal)
For urea-based SNCR systems, the total useful thermal energy
loss is calculated with the following equation:
5-11
-------
EnergySNCR " Energyv + Energyc (5.4)
For SCR, energy is required to inject reagent into the
flue gas and can also be estimated with equation 5.3. Also,
additional fan horsepower is required to overcome the pressure
drop across the catalyst. The following equation is used to
estimate the thermal equivalent of this energy requirement:
t
Energyf (MMBtu/yr) = AP * 0.0361 * 144 * Q * CF *
8760/737.56/EF/O.40 * 0.003412 (5.5)
where:
AP = Pressure drop across the catalyst (assumed
to be 5 in. H2O)
0.0361 = Conversion factor; lbf/in2 per in. 1*20
144 = Conversion factor; in2/ft2
Q = Flue gas flow rate at inlet to catalyst
reactor (actual ft3/sec); calculated using
equation (5.6)
CF = Capacity factor (decimal fraction)
8760 = Conversion factor; hr/yr
737.56 = Conversion factor; [(Ibf _ ft/sec)/kW]
EF = Fan efficiency (decimal fraction = 0.85)
0.40 = Conversion efficiency of thermal energy to
electrical energy
0.003412' = Conversion factor; MMBtu/kWh (thermal)
Flow rate (Q) is calculated based on the following equation:
Q (ft3/s) = flow * Tcor * size * 100 / 3600 (5.6)
where:
flow = Flue gas flowrate at inlet to catalyst at
normal conditions (126 Nft3/kWh for coal-fired
and 100 Nft3/kWh for oil- and gas-fired
boilers)*
Tcor = Temperature correction = 1060 R / 492 R
(assuming catalyst operating temperature =
600°F)
5-12
-------
size — Boiler capacity (MMBtu/hr)
100 = Conversion factor; kW per MMBtu/hr (assuming
heat rate = 10,000 Btu/kWh)
3600 = Conversion factor; sec/hr
The following equation is used to calculate the total
useful thermal energy loss for SCR systems:
t
EnergyscR = Energyc + Energyf (5.7)
5-13
-------
5.5 REFERENCES
1. Sage, P. W. Nitrous Oxide Emissions from Coal-Fired
Plant - An Update on the Joule Collaborative Project.
Proceedings of the 5th International Workshop on Nitrous
Oxide Emissions. Tsukuba, Japan. July 1-3, 1992. pp.
2-4-1 to 2-4-4.
2. New Source Performance Standard, Subpart Da-Technical
Support for Proposed Revisions to NOX Standard k(TSD) .
Publication No. EPA-453/R-94-012. U.S. Environmental
Protection Agency, Research Triangle Park, North
Carolina, pp. 5-4 to 5-5. October 1995.
3. Method 19 - Determination of Sulfur Dioxide Removal
Efficiency and Particulate Matter, Sulfur Dioxide, and
Nitrogen Oxides Emission Rates. Federal Register
40 CRF Part 60. Appendix A. p. 47853.
4. Ref. 2, p. C-19.
5-14
-------
6.0 MODEL BOILERS AND CONTROL OPTIONS COSTS
This chapter presents the estimated cost and cos-t
effectiveness of the identified control options applied to the
model boilers described in chapter 4.0. The chapter includes
estimated total capital cost, annualized cost, cost
effectiveness, and incremental cost effectiveness (ICE) for
the application of the control options to the model boilers.
Section 6.1 discusses the costing methodology. Section 6.2
presents the cost algorithms for combustion controls (CC),
selective noncatalytic reduction (SNCR), and selective
catalytic reduction (SCR). Section 6.3 presents the cost
results for application of the three control options to the
model boilers and discusses the sensitivity of ICE to
variations in model boiler parameters.
6.1 COSTING METHODOLOGY
This section describes the procedures used to estimate
the capital costs, operating and maintenance (O&M) costs, and
cost effectiveness values for the three control options.
Appendix B.I presents example calculations for the costing
procedures described in this section. This section also
discusses cost considerations associated with solid and liquid
waste, emissions monitoring and compliance testing, and
regulatory and enforcement agency activities. Cost procedures
follow the general methodology contained in the Electric Power
Research Institute (EPRI) Technical Assessment Guide (TAG™)1
and the Office of Air Quality Planning & Standards (OAQPS)
Costing Manual.2 The major components of capital and O&M costs
are shown in table 6-1. All costs are based on 1995 dollars.
6-1
-------
tQ
I
CO
fd
jj £j
rd O
U -H
\ 4J
CQ fd
rH N
4J rd -H
a o g
-H
•rl 4-> 4J M 4J
CO -H CD fd CO
fd d d 4J JJ
o ena a
fd a o o
iw -H u o
^
rH Q) 4-1 CO
fd tt) O CO
^ d a) cn
0 0
EH U
CO
1— 1
a)
-•H
M ^1 i-l
O O 0)
to J3 A 4J
o fd fd fd
X) i — I i — I g
fd
rH 0) >1 0)
o w u
end 0 c
G fd CQ fd
•H a -H a
4-> 0) > 0)
(tf 4J H 4J
M d i o
d ^- 4-1 Pi
0) CQ -H CO
ftrH U -H
OVrl 4J K <1)
M g CJ Q) 4J
0) (U CU 4J CO
d id rH fC) fd
W O W ^ ^
4J
CQ
O
U
jr
c3
O
(1)
rH
Q
fd
•H
fd
4J
CQ
0
U
o
iH
rd
0
EH
6-2
-------
6.1.1 Total Capital Cost
This section describes the procedures for estimating
direct and indirect costs that comprise total capital cost.
6.1.1.1 Direct Cost. Direct cost includes purchase and
installation of system hardware directly associated with the
control technology. Initial chemical or catalyst costs and
start-up/optimization tests are also included. The purchase
and installation of continuous emission monitoring (CEM)
equipment is already required by the current subpart Db
standard; as a result, there are not additional costs
incurred.
The methodology used to determine the direct cost of each
NOX control option is similar to that used to evaluate the
subpart Da model boilers.3 Additionally, the final cost
algorithms for CC and CC + SCR are the same as those derived
for the subpart Da boilers. For CC + SNCR, however, the
capital cost equations were revised to reflect the assumption
that lances, required to improve mixing between reagent and
gas in utility boilers, are not required for industrial
boilers because of their smaller size. Details of the costing
methodology and the cost algorithms used are discussed in
appendix B.
The direct cost for installing each NOX control option on
the model boilers was calculated using a functional
relationship of the form:
Direct Cost ($/MMBtu/hr) = a * sizeb (6-1)
where:
a,b = Constants, specific to each NO, control
option
size = Boiler size (MMBtu/hr)
6.1.1.2 Indirect Costs. Indirect costs include general
facilities, engineering expenses, royalty fees, and
contingencies. Indirect costs are estimated by multiplying
the total direct cost by an indirect cost factor.
6-3
-------
General facilities include offices, laboratories, storage \
areas, or other facilities required for installation or
operation of the control option. Engineering expenses include
the industrial boiler owner's internal engineering efforts and
those of the architect/engineering (A&E) contractor.
Engineering costs incurred by the control option vendor are
included in the equipment cost and are considered direct
costs.
There are two contingency costs: project contingency and
process contingency. Project contingency covers minor support
equipment and materials not included in the direct cost
estimate. The size of the project contingency for a given
application depends on the level of detail included in the
direct cost estimate. Generally, the more detailed the cost
estimate, the smaller the project contingency required.
Process contingency is based on the maturity of the technology
and the number of previous installations. Process contingency
covers unforeseen expenses incurred because of inexperience |
with newer technologies. Generally, for older and more mature
technologies, less process contingency is required. The
specific project and process contingencies vary with control
option and are included in appendix B.
6.1.2 Operating and Maintenance Costs
Operating and maintenance costs include fixed and
variable components. Fixed O&M costs include operating,
maintenance, and supervisory labor; and maintenance materials.
Fixed O&M costs are independent of capacity factor. Variable
O&M costs include any energy penalty resulting from efficiency
losses associated with a given control option, disposal of
additional wastes, and use of additional chemicals and
electricity. Variable O&M costs are dependent on capacity
factor.
Cost rates for labor and materials included in the cost
estimates are shown.in table 6-2. The prices listed for coal,
residual oil, distillate oil, and natural gas are the \
estimated national average prices (in 1995 dollars) for the
6-4
-------
(
C/D
H
00
O
U
S
_>j
<*>
O
w
_J
ARIAB
>
D
$
^^
P
S
/S
E
cs
VO
«
_!
§
H
8
c
0
l~
rn
1
OS
•*-»
'c
^
to
O
O
n
-— ^-^cococococsmCN
8
5^
e^- ?
U.....3333 ^^
J3j3J3*-''t-J<-'-»-> ,,
•Y^r-^QgQQ *g g
1 1 1 1 1 1 1 4i ° §' =
C C C ^< ^i ^- ^ (£H -*-J •*-* *T*
V^€^€^€^€^€^Vi€^€^€^^
«Q
000
o o o o o o
ooor-r--^*-- ,— ;/-o<-s
d d d •*. ^ °. «. 8 o o
CN(NCSr^ *j^ *™i L« ^^ s
s. | | i 1 :i l 1 I 1
oSc^uc±;p^;o<^
m
CS •— ' •— ^i
"* ^s
oo £
-COO
c S 0 O
§322
< ^ C C
&"J V) &=) V)
oo
cs >^i o o
^ o ^r »n
^- O O co
g
p"
S
S
CU
ex
o
in
R]
U.
c2
T3
CU
tt
•- .S3
_t 5/5
TO ;fs
S Jo
P b
t-4 t->
«2 "^
^ CO
•*^
S c
8 °
s g
»*^ UH
t? 3
13 c
45 O
4»* ^^
CU O
Is
1
«
6-5
-------
year 2000, using the reference case analysis of the Department
of Energy's (DOE's) 1997 Annual Energy Outlook.4 These prices
were used to best represent the average price across the
20 year expected life of the control option equipment. The
prices listed for ammonia and urea were obtained from vendors.
Prices for labor, solid waste, electricity, and water are in
1989 dollars. The items do not significantly affect *the total
O&M costs, and therefore, more recent values were not
obtained.
6.1.3 Calculation of Total Annual Cost. Cost Effectiveness.
and Incremental Cost Effectiveness
Total annual cost is the sum of annualized capital costs
and total O&M costs. To convert total capital cost to an
annualized capital cost, the total capital cost is multiplied
by an annual capital recovery factor (CRF). The CRF is based
on the economic life over which the capital investment is
amortized and the cost of capital (i.e., interest rate), and
is calculated using the following equation:
CRF = i(l+i)n/[(l+i)n -1] (6-2)
where:
i = interest rate as a decimal fraction (assumed to
be 0.07 -- 7 percent)
n = the economic life of the equipment in years
(assumed to be 20 years)
Cost effectiveness indicates the total cost of a control
option per unit of NOX removed. Cost effectiveness is
calculated by dividing total annual costs associated with the
control option, expressed in dollars per year, by the annual
emission reductions, expressed in tons of NOX per year.
Incremental cost effectiveness is calculated by dividing the
incremental total annual costs (change in costs between two
control options) by the incremental emission reductions.
Example calculations of CRF, cost effectiveness, and ICE
are provided in appendix B.I.
6-6
-------
6.1.4 Other Cost Considerations
Potential costs associated with disposal of solid and
liquid wastes, emission monitoring and compliance testing, and
regulatory and enforcement agency activities were considered.
Only incremental costs attributable to the revision of the
standard are evaluated.
6.1.4.1 Solid & Liquid Waste. The combustion control
NOX control option being considered here is the same as is
currently being used to meet the current subpart Db standard.
Therefore, there is no change in potential solid and liquid
waste and no incremental cost associated with waste disposal.
The cost of catalyst disposal for SCR systems is included
in the estimates of variable O&M costs. There is a potential
for ammonia contamination of the ash for the SNCR and SCR
control options which could result in an increase in solid
waste disposal costs.
6.1.4.2 Emission Monitoring and Compliance Testing. The
current subpart Db standard already requires NOX and oxygen
monitoring equipment. As a result, there are no incremental
emission monitoring costs associated with any of the control
options for monitoring NOX or oxygen. Ammonia monitoring
costs could potentially be incurred for the SNCR and SCR
control options. However, at this time it is not known
whether ammonia monitoring would be required and no costs have
been included for these monitors. There are no incremental
compliance testing costs associated with SNCR or SCR systems
since outlet NOX concentration monitoring similar to that
already required will be used.
6.1.4.3 Regulatory and Enforcement Agency Activities.
There are no incremental costs to regulatory and enforcement
agencies due to the NOX control options being considered.
6.2 COSTING PROCEDURES
This section presents the cost procedures used for
estimating total capital cost, and total O&M cost for each
control option applied to the possible combinations of fuel
type and furnace type for the model boilers.
6-7
-------
6.2.1 Combustion Controls
For pulverized coal-fired boilers, the CC option was low
NOX burners (LNB). A review of the NOX emission limits being
set for industrial boilers that are required to comply with
the current NSPS, subpart Db standard, indicates that these
levels are similar to the NOX control levels associated with
combustion controls (CC). As a result, no additional capital
costs will be incurred for new boilers to meet these levels.
However, an incremental O&M cost equal to 3 percent of the
direct cost for the combustion control equipment was assumed
to reflect potential costs associated with improved boiler and
burner operation that may be required to achieve the lower NOX
emission limits.
The cost equation, is the same as was used for wall-fired
subpart Da (utility) model boilers and is as follows:3
O&M ($/yr) = 0.03 * [60,600 * (size)0-56] (6-3)
where:
size = boiler capacity (MMBtu/hr)
As discussed in chapters 3.0 and 4.0, LNB are not
applicable to fluidized bed combustion (FBC) boilers.
Combustion controls for FBC boilers include controlling bed
temperature and the use of air staging. Based on information
from FBC boiler vendors, there is no additional cost
associated with achieving the lower NOX emission levels
associated with combustion controls.15'16'17 As a result, there
are no combustion control costing procedures for FBC boilers.
For spreader stoker boilers, combustion controls include
air staging. As shown in chapter 3.0, spreader stokers are
already achieving the NOX emission levels associated with CC.
Therefore, no CC costing procedures are presented for SS
boilers.
6-8
-------
For residual oil-fired boilers, no additional hardware
requirements were assumed to achieve the lower NOX emission
levels associated with combustion controls. However, an
incremental O&M cost equal to 3 percent of the direct cost of
combustion controls was assumed to reflect potential costs
associated with improved boiler and burner operation. A
comparison between natural gas-/oil-fired boilers and* coal-
fired boilers indicated that capital costs for CC are
similar.18 Consequently, for residual oil-fired boilers, O&M
costs were evaluated using equation (6-3) .
For field-erected, distillate oil- and gas-fired boilers,
the CC levels are assumed to be the same as baseline levels
(table 4-4). Therefore, no costs were assumed for combustion
controls. For packaged boilers, however, equation (6-3) was
used to estimate potential costs associated with improved
boiler and burner performance to achieve the lower NOX levels
compared to baseline.
6.2.2 Selective Noncatalytic Reduction
Costing procedures for SNCR are based on a low-energy,
urea-based system. This control option is applicable to all
the model boilers presented in chapter 4.0. To accommodate
fluctuations in flue gas temperature, each boiler was assumed
to have two levels of wall injectors. The injected urea
solution was assumed to be 10 percent urea by weight,
90 percent dilution water. The normalized stoichiometric
ratio (NSR) was assumed to be 1.0. The procedures used to
develop the cost equation are discussed in appendix B. The
following equation was derived for all three coal-fired boiler
types (PC, FBC, and stoker):
Direct Cost ($/[MMBtu/hr]) - 26,877 * (size)'0-55 (6-4)
For natural gas- or oil-fired boilers;
Direct Cost ($/[MMBtu/hr]) = 32,695 * (size)'0-59 (6-5)
As discussed in appendix B.3, the indirect cost factor
was assumed to be 1.3.
6-9
-------
Fixed O&M costs include operating, maintenance, and
supervisory labor; and maintenance materials. Fixed O&M costs
were estimated using the following equation for all boilers:
FO&M ($/yr) = 53,027 * (size)0-21 (6-6)
Variable O&M costs include the cost of the urea ^solution
(chemical costs); energy penalties associated with
vaporization of the urea solution in the boiler; and
electricity costs for equipment operation. The urea solution
is the primary (90 percent) variable O&M cost and was
estimated by determining the amount of injected urea as a
function of the baseline NOX emission levels and the assumed
NSR of 1.0. The amount of injected urea was multiplied by
solution unit cost to determine the total chemical cost. The
amount of injected urea was also used to determine the energy
penalties. The total energy penalty was converted to an
annual cost based on the fuel per million Btu. Electricity
costs were determined based on unit size and urea injection
rate. Appendix B.3 presents the equation for calculating urea
cost and the basis for other variable O&M costs.
6.2.3 Selective Catalytic Reduction
The SCR cost algorithms developed for utility boilers
were used to estimate costs for industrial boilers. The
algorithms are based on hot-side SCR technology (i.e., the
catalyst is located between the boiler economizer and air
preheater).
This control option has not been demonstrated on FBC
boilers and may not be technically feasible due to the
likelihood of catalyst poisoning resulting from high levels of
calcium oxide present in the fly ash.19 Therefore, this
control option was not considered for FBC boilers.
Catalyst price, which has a significant impact on total
cost, was assumed to be $350/ft3 for coal-, natural gas-, and
oil-fired boilers. Catalyst life was assumed to be 3 years
6-10
-------
for coal-fired boilers and 6 years for natural gas- and oil-
fired boilers. Catalysts space velocities were assumed to be
3,200/hr for coal-fired boilers 5,000/hr for residual oil-
fired boilers, and 14,000 for distillate oil-/natural gas-
fired boilers. An NH3/NOX molar ratio of 0.82 was used for
achieving 80 percent NOX reduction efficiency.
Based on the available data, simplified algorithms in the
form of equation 6-1 were developed to estimate process
capital costs for SCR systems. The coefficients for the
coal-, oil-, and gas-fired boilers are:
Fuel
Coal
Oil/Gas
Furnace type
Pulverized Coal
Fluidized Bed Combustion
Spreader Stoker
Field Erected
Packaged
a
34,718
35,276
b
-0.30
-0.33
The equation for estimating initial catalyst charge costs is
presented in appendix B, section B.4.1.
Indirect cost factors (ICF) were applied to process
capital and initial catalyst charge costs. Indirect costs
were assumed to be 45 percent of the process capital (i.e.,
ICF = 1.45). For the application of SCR to boilers burning
medium- to high-sulfur coals, indirect costs may be greater
than 45 percent of the process capital, due to factors
discussed in chapter 3.0. Variable indirect cost factors were
used for estimating indirect costs associated with the initial
catalyst charge (1.25 for coal-fired boilers, 1.20 for oil-
fired boilers, and 1.15 for gas-fired boilers).
Fixed O&M costs for an SCR system include operating,
maintenance, and supervisory labor; and maintenance materials.
Variable O&M costs are ammonia, catalyst replacement,
electricity, water, steam, and catalyst disposal. Equations
for estimating these costs are presented in appendix B,
sections B.4.3 and B.4.4.
6-11
-------
6.3 MODEL BOILER COST IMPACTS
Model boiler impacts have been developed for three
control options: CC, CC with SNCR, and CC with SCR.
Performance of the control options is discussed in
chapter 3.0. The emission rate associated with CC was assumed
to be a function of fuel and furnace type. Selective
noncatalytic reduction and SCR systems are assumed trf achieve
a constant emission reduction of 50 and 80 percent,
respectively, over CC levels.
Model boiler impacts are presented in two tables included
as appendix C: one for coal (PC, FBC, and SS) and another for
oil/gas. Each table consists of two parts. The first part
presents emissions and emission reductions, and the second
part presents capital costs, annualized costs, cost
effectiveness, and incremental cost effectiveness values.
Emissions and emission reductions are discussed in
chapter 5.0. The impacts for each boiler were estimated at
capacity factors of 0.10, 0.30, and 0.60 with the exception of
PC-fired boilers which were evaluated at capacity factors of
0.30 and 0.60, only.
The annualized costs presented in appendix C for each NOX
control option are summarized in table 6-3. The data are for
model boilers operating at a capacity factor of 0.3. In each
case, the annualized costs are expressed in $/yr and as a
percentage of the annual cost of the steam generated by the
boiler.
As seen in the table, the annualized costs for CC range
from $0 to $87,000, and are less than 2 percent of steam costs
for the model boilers presented. With the addition of SNCR to
CC, annualized costs vary from $169,000 to $539,000 and from 3
to 16 percent of steam costs. For CC + SCR, annualized costs
vary from $418,000 to $2.8 million and from 10 to 48 percent
of steam costs.
Table 6-4 presents ICE values for the three control
options applied to the model boilers at a capacity factor of
0.30. Because CC + SCR was assumed not applicable to FBC
6-12
-------
rn
i
vo
s
(_>
u
oc
z
o
o
u
4
r>
0
X
£
x
L.
£
^
^
D "i1
•1 j>
.Q
Ol
|
u
CO
c
3
0,
f
3
u_
R
0
fj
r—
8:
o
o
ro
in
1.
01
TJ
CO
01
n
•o
CM
o
PJ
in
in
•0
o
in
ro
o
o
o
o
in
00
ro
1
5
•o
g
S
-
o
PJ
o
o
in
o
o
in
*
i_
11
CO
3
ro
1
PJ
PJ
in
^
o
PJ
ro
ro
in
•*
-
o
o
co
o
o
o
X
1
5
-o
s
o
£
PJ
t~t
o
K
o
o
PJ
PJ
0
s
1
o
o
o
•o
in
-
o
o
o
o
in
PJ
§
u
l_
01
4_*
ID
S
0)
CD
^
O.
in
ro
§
eo
2
o
o
CO
>o
o
o
o
o
CO
(_
4-*
CO
z
o
01
CO
4-*
o
o
o
CO
in
^
o
in
CM
o
o
o
in
PJ
"8
OJ
•n
CJ
u.
in
CD
ts
+
§
CM
CO
00
^
O
s
CM
0
o
o
o
in
*
i_
01
4^
CD
3
O
§
in
CM
ro
s
ro
o
o
o
o
o
p.
ro
o
PJ
3
C
«l O
o
— 4^ 1!
co » N
30 •—
co Ti
I 4-1 CO
. « u.
CO
o
u
a ci
o>
CD CD
13 OJ
U
01 I-
e&
-
."8
II
u
6-13
-------
id
co
N
O
CO
CO
H
H
H
fa
fa
H
H
CO
O
CJ
1
H
O
CO
Q
§
l
vo
I
"c
CO t^
rH w
i CO
§c
i>
w ^
«M
w
^1
0*
o
° ^
5 o
o "*
s
H
01
3
fa
O
1
O
OJ
Q)
•H
0)
CO
•
CO
>
u
u
o
in
n
l
o
OJ
r-
H
y
O
«
U
>
1
CO
+
u
(J
o
H
t-l
m
i
o
in
r>
VO
8
•
CO
8
CO
+
t_>
o
0
CO
H
1
o
r-
o
«
1
CO
+
%
•
CO
8
to
+
o
CJ
tj
&
0
Q)
C
•H
iH
OJ
CO
(0
•
CO
u
u
o
CO
H
r>
•"'
1
o
CO
%
CJ
u
•
to
>
1
CO
+
u
u
•0
$
u
o
•
CO
8
CO
+
u
CJ
^
2
8
2
CO
+
0
U
*
CO
8
CO
+
u
u
0
m
fa
o
Q)
C
•H
rH
0)
CO
•
to
>
u
u
o
VO
in
i
0
o
H
OJ
u
o
•
CO
>
1
CO
+
8
o
o
H
H
1
O
0
en
in
u
U
*
CO
8
CO
+
o
u
o
01
0
01
1
0
in
OJ
OJ
H
O
to
+
0
o
•
CO
a
CJ
CO
+
8
^
OJ
X
o
4J
to
^1
Q)
•O
(C
0)
a
CO
^
«j
o
CJ
o
n
o
01
1
o
0)
c
•r^
rH
Q)
CO
«S
•
CO
>
o
CJ
o
CO
r>
^j
^^
1
o
n
en
oT
CJ
u
•
to
^
8
2
CO
+
u
u
o
01
rH
Ol
1
o
H
*
t~
U
O
•
CO
g
+
u
tj
o
in
n
in
n
l
o
CO
01
H
CJ
2
to
+
£
•
CO
8
CO
+
8
•o
Q)
U
•H m
fa 5
o
VO
1
0
VO
a>
c
•H
rH
0)
(0
(0
•
CO
>
u
u
0
00
n
l
o
*
r-
CJ
O
•
10
j>
a
CO
+
CJ
CJ
o
01
rH*
01
1
O
o
VO
Ol
rH
CJ
U
•
CO
tf
CJ
CO
+
CJ
CJ
o
in
in*
i
0
•«*
rH
a
CJ
2
to
+
CJ
CJ
•
CO
8
CO
H-
CJ
CJ
0)
.p
a)
4J
A3
T3
Q)
(0
_*^
O
(0
a<
rH
•H
0
rH
R)
3
•C
CO
0)
a
6-14
-------
I
VO
I
sS
1
rementa]
:iveness
0 o
• 1 1^1
W
^f
tT
O
rHo
o £
C Q
0.2
® l_l
E"t jg
Q
2°
4J
C
O
O
•o
o
*
EH
Furnace
OJ
o
rH
OJ
3
fa
0
QJ
C
•H
rH
OJ
CO
J8
•
CQ
CJ
CJ
O
rH
*
1
O
H
VO
U
u
CO
>
«
g
CO
J,
{_)
CJ
TJ
d-erecte
rH
OJ
'H
fa
O
O
CO
o\
\
o
rH
CO
rH
CJ
CJ
•
CO
>
8
CO
„!_
CJ
CJ
rtube
0)
-P
id
o
in
CM
(j\
1
O
rH
O\
CM
8
55
CO
_f_
CJ
o
•
CO
>
8
CO
^
CJ
u
OJ
4.
id
rH
H
•H
•P
CO
-H
Q
O
O
«
1
O
o
CM
0)
C
•H
rH
Q)
CQ
3
•
CQ
CJ
CJ
CO
id
id
^4
•P
id
c
r-t
•H
O
0
VO
H
CM
O
rH
H
rH
rH
U
U
CO*
>
tt
O
CO
j.
CJ
u
.
4
f
*
1
o
n
n
1
o
VO
CO
H
O
U
•
CO
>
8
CO
±
o
u
Q)
Q
3
J
1)
P
s
O
1)
id
M
CJ
id
a<
o
0
VO
cs
in
I
o
o
r>
8
ss
CO
_J_
CJ
u
•
CQ
>
8
CO
J_
CJ
CJ
0 3
M 0 4J
O rH CQ
O •»»
id co
irH in
4J *H
•H O TJ
CJ 42 C •
id id co
D-CJ 14
Id O4 - Q)
O O rH
id o CM o
*W 4^
•P ••
id IH o TJ
43 0 0)
CO ^H O1
CQ 3 Id
OJ +J —X
C CO CO O
0) £ rl Id
> S 0) a
•H rH
4J 0-H M
o o o o
0) 0 J2 O O
C fl) **>
OJ43 rH (0
CO 4J
OJ TJ M
r4 M C 0)
a o id TJ
i
rH
ll id
4J
cj id
CJ CJ
o
0)
rH
43
id
0
•H
rH
rtj
•g
55
II
S
TJ
6-15
-------
boilers, ICE values were not calculated for these cases. ICE
values were not calculated for CC on field-erected, distillate
oil/natural gas-fired boilers because the baseline and
combustion control NOX emission rates were assumed to be the
same for these boilers. Also, ICE values for CC on FBC and
spreader stoker boilers are zero because the annualized cost
is zero. k
6.3.1 Combustion Controls
Figure 6-1 illustrates the effect of fuel and furnace
type on ICE over baseline levels. The data presented are for
a capacity factor of 0.30. As discussed in section 6.2.1,
capital costs were assumed to be zero for combustion controls.
However, an incremental O&M cost equal to 3 percent of the
direct cost of the combustion control equipment was assumed
for PC-fired, residual oil-fired, and distillate oil-/ natural
gas-fired packaged boilers. This O&M cost was not assumed for
FBC or spreader stoker boilers because these boilers are
already achieving the CC levels. The O&M cost for distillate
oil-/natural gas-fired, field-erected boilers with CC is
assumed to be zero as there is NOX emission reduction for
these boilers.
Operating and maintenance cost is independent of boiler
type and is only a function of boiler size. As a result, all
boilers of the same size have the same annualized costs for
CC. Therefore, differences seen in ICE are due to differences
in NOX emission reductions. For example, distillate
oil-/natural gas-fired packaged boilers have the largest ICE,
because they have the smallest emission reductions.
Conversely, ICE for PC-fired boilers are the smallest because
of the large NOX emission reductions associated with these
model boilers. The shape of the individual curves results
from changing boiler size and indicates that ICE decreases as
boiler size increases. This is because of the effect of
economy of scale on capital costs and consequently on O&M
costs per the assumed relationship between O&M and capital
costs described earlier.
6-16
-------
O. B K d
MM
0)
CO
Q)
O
OQ
CO
01
w
<0
4J
O
0)
CO
O
(I)
6
0
b
O
O
O
I
VO
0)
M
a
n
6-17
-------
Because the NOX emission levels associated with CC on
field-erected distillate oil-/natural gas-fired boilers are
the same as baseline levels, the ICE for these model boiler
types is not presented. Also, because the incremental cost
effectiveness for FBC and spreader stoker boilers is zero,
they are not presented in figure 6-1.
Figure 6-2 presents the effect of capacity factorr and
boiler size on ICE for each fuel. The range of ICE values for
coal-, residual oil-, and distillate oil-/natural gas-fired
boilers are presented. Coal-fired boilers have the lowest
ICE. FBC and spreader stokers have zero ICE for combustion
controls because their annualized costs are zero. Of the
coal-fired boilers, the 250 MMBtu/hr PC-fired boiler has the
largest ICE.
Distillate oil-/natural gas-fired boilers have the
largest ICE. Because the field-erected boilers have zero
emission reductions, only packaged distillate oil-/natural
gas-fired boilers are presented.
The ICE values for residual oil-fired boilers have
intermediate ICE. In this group, the 100 MMBtu/hr, field-
erected boilers had the largest ICE, while the 250 MMBtu/hr
packaged boilers had the smallest ICE.
As indicated earlier, differences seen in ICE for CC are
due to differences in NOX emission reductions. The ICE for
combustion controls decreases as capacity factor increases
because of increasing NOX emission reductions.
6.3.2 Selective Noncatalytic Reduction
This section discusses ICE for the application of CC +
SNCR over CC for the model boilers. Figure 6-3 shows the
effect of boiler size, fuel, and furnace type on ICE. The
curves presented are at a capacity factor of 0.30. As boiler
size increases, ICE decreases. Differences in fuel and
furnace type are reflected by the differences in inlet NOX
emission rates. This effect is seen in the relative position
of each curve in figure 6-3. These curves show that as the
inlet NOX emission rate increases, ICE values decrease. This
6-18
-------
Q?g«
§. i. i. §. §. §. §. §. §. i
O
JJ
O
«
fa
>t
O
A)
a
h
o
c
o
u
I
vo
Q)
W 10
(uovt) nauMipajja JSOQ pnuauiu
6-19
-------
\ H H H
Q)
N
-H
CO
0)
o
n
s
w
-------
results from changes in both annualized cost and NOX emission
reductions. While annualized costs increase as inlet NOX
emission rates become larger, NOX emission reductions increase
at a faster rate. For a 500 MMBtu/hr boiler with a capacity
factor of 0.30, when the inlet NOX emission rate increases
from 0.08 Ib/MMBtu (distillate oil/natural gas, field-erected)
to 0.50 Ib/MMBtu (spreader stoker), annualized costs^increase
32 percent while NOX emission reductions increase by
740 percent. The net result is a decrease in ICE of
41 percent.
It should also be noted that the ICE for both field-
erected and packaged residual oil-fired boilers are the same.
This is because they have the same annualized costs and the
same NOX emission reductions.
Figure 6-4 presents the effect of boiler size and
capacity factor on ICE for CC + SNCR. Curves are for PC-fired
boilers, spreader stokers, and distillate oil-/natural gas-
fired field-erected boilers are plotted to present the range
of cases. Spreader stokers represent the largest emission
reduction while distillate oil/natural gas field-erected
boilers represent the smallest NOX emission reductions.
Again, ICE decreases as boiler size and capacity factor
increase for the same reasons discussed in the combustion
control section.
Among model boiler types not shown in the figure, ICE
values are smaller for residual oil-fired boilers when
compared to similar sized distillate oil-/natural gas-fired
boilers. For FBC boilers, ICE is less than residual oil-fired
boilers of the same size but greater than PC boilers of the
same size.
6.3.3 Selective Catalytic Reduction
This section discusses ICE for the application of '-CC +
SCR over CC + SNCR for the model boilers. Figure 6-5 shows
the effect of boiler size, fuel, and furnace type on ICE. SCR
was not considered applicable for FBC boilers. Curves are
presented for the remaining model boiler types each with
6-21
-------
- „ . -
M t M H I
14
o
JJ
u
(0
•H
U
8
to
CO
Q)
c
•H
4J
O
0)
H
4J
(0
O
U
(0
4J
i
0)
V4
o
u
W
+
u
a)
I
(uoj/t)
6-22
-------
H HH
£ S
si X
12 £ «•£
5 §1* c
E-a 5-3-1
2 S S m 8.«.
7.f | S STo
« | .2 Stt »
o « P 3 o U-
2 Q. « 5=,W O
N
•H
W
h
Q>
r-»
•H
tt
0)
c
0>
•H
4J
O
0)
«M
4
O
H
u
CO
m
l
to
0
I
•H
fa
6-23
-------
different inlet NOX emission rates (except field-erected and
packaged residual oil-fired boilers which have the same inlet
NOX emission rates). The effect of inlet NOX emissions is
seen in the relative position of the curves. As inlet NOX
increases, ICE decreases. This is the result of changes in
both annualized costs and emission reductions, with emission
reductions increasing at a faster rate. k
Figure 6-6 presents the effect of boiler size and
capacity factor on ICE for CC + SCR. As was done in the SNCR
section, PC boiler, spreader stoker, and field-erected,
distillate oil-/natural gas-fired boiler curves are presented.
Spreader stoker emission reductions are the largest while
distillate oil-/natural gas-fired boiler emission reductions
are the smallest. As for SNCR, the ICE for residual oil-fired
boilers is less than distillate oil-/natural gas-fired boilers
of the same size, but greater than coal-fired boilers of the
same size. Again, it is seen that ICE decreases as boiler
size and capacity factor increase for the same reasons given
for combustion controls and CC + SNCR.
6-24
-------
M
M
O
•P
U
U
(0
u
w
to
(0
0)
c
•H
4J
O
01
«M
a
o
o
(0
I
M
U
8
to
vo
I
0)
I
•H
CM
(uovt) nauaAipajja isoo
6-25
-------
6.4 REFERENCES
l. EPRI (Electric Power Research Institute). TAG™
Technical Assessment Guide. EPRI P-6587-L, Volume 1:
Electricity Supply - 1989. Revision 6. Technical
Evaluation Center, Palo Alto, CA. November 1989. pp.
3-1 through 3-14.
t
2. Vatavuk, William M. OAQPS Control Cost Manual. Fourth
Edition. Chapters 1 and 2. EPA 450/3-90-006.
U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park,
NC. January 1990. pp. 1-1 through 2-32.
3. New Source Performance Standard, Subpart Da Technical
Support for Proposed Revision to NOX Standard,
EPA-453/R-94-012, U.S. Environmental Protection Agency,
Research Triangle Park, NC. October 1995.
4. U.S. Department of Energy, Office of Integrated Analysis
and Forecasting. Annual Energy Outlook 1997. DOE/EIA-
0383(92). Washington, DC. November 1996. table 3.
5. Integrated Air Pollution Control System, Version 4.0,
Volume 2: Technical Documentation Manual. EPA-600/7-90-
022b. U.S. Environmental Protection Agency, Air and
Energy Engineering Research Laboratory, Research Triangle
Park, NC. December 1990. pp. 4-77 through 4-97.
6. Telecon. Martina, T., LaRoche, Industries to Illig, C.,
Radian Corporation. SNCR Chemical Cost. January 4,
1993.
7. Telecon. Millard, D., National Ammonia Company with
Illig, C., Radian Corporation. SNCR Chemical Costs
(Urea, Anhydrous, and Aqueous Ammonia). January 4, 1993.
8. Fax. Vereah, J., LaRoche Industries with Illig C.,
Radian Corporation. SNCR Chemical Cost. January 7,
1993.
9. Telecon. Vereah, J., LaRoche Industries with Illig, C.,
Radian Corporation. SNCR Costing - Aqua + Anhydrous
Ammonia. January 4, 1993.
10. Telecon. Kellog, G., Nalco Fuel with Illig, C., Radian
Corporation. SNCR Chemical Cost for Enhancers.
January 5, 1993.
6-26
-------
11. Fax. Miskus, J., Cargill, Inc. to Illig, C., Radian
Corporation. SNCR Chemical Cost for Urea + N0x0ut A.
January 13, 1993.
12. Telecon. Moredyke, D., UNOCAL Corp. with Illig C.,
Radian Corporation. SNCR Chemical Cost, N0x0ut A
Costing. January 4, 1993.
13. Letter and attachments from Poole, M. F., W. H. Shurtleff
Company to Illig, C., Radian Corporation. SNCR *Chemical
Cost for Urea + N0x0ut A. January 6, 1993.
14. Alternative Control Techniques Document - NOX Emissions
from Stationary Reciprocating Internal Combustion
Engines. EPA-453/R-93-032. U.S. Environmental
Protection Agency, Research Triangle Park, NC.
July 1993. p. 6-5.
15. Telecon. Stone, J., Ahlstrom Pyropower, Inc. with King,
B., Radian Corporation. Fluidized Bed Combustion
Boilers. August 18, 1993.
16. Telecon. Czarnecki, T., ABB-CE with King, B., Radian
Corporation. Fluidized Bed Combustion Boilers.
August 18, 1993.
17. Telecon. Edvardsson, C., Tempalla Power with King, B.,
Radian Corporation. Fluidized Bed Combustion Boilers.
August 17, 1993.
18. Alternative Control Techniques Document - NOX Emissions
from Utility Boilers. EPA-453/R-94-023.
U.S. Environmental Protection Agency, Research Triangle
Park, NC. March 1994.
19. Telecon. Evardsson, C., Tempalla Power with King, B.,
Radian Corporation. Application of Selective Catalytic
Reduction to Fluidized Bed Combustion Boilers.
September 27, 1993.
6-27
-------
-------
APPENDIX A
ENVIRONMENTAL AND ENERGY IMPACTS FROM MODEL INDUSTRIAL BOILERS
-------
-------
A.I. PRIMARY EMISSIONS FROM COAL-FIRED MODEL INDUSTRIAL BOILERS
-------
-------
'o
CD
"ai
•D
o
•o
0)
re
o
O
o
c
o
LU
(0
E
eductions (a
Emission R
Control Data
1 Combustion
Baseline
0)
Baselin
f
5 I
w
o
0 *
* f
s &
tn
B g
f
^
o ^
#
I Annual
Emissions
(ton/yr)
Rate
(Ib/MMBtu)
Annual
Emissions
(ton/yr)
0 ="
« 13 5
EKI
LU ^
>,
•50
CO t>
S- <°
o "•
*-
TT *
U
Size
(MMBtu/hr)
8 .
£ £
3 *-
LL
a>
Q.
><
t-
"5
o
u.
a>
CO
00
CO
00
tO
0
CO
o
in
o
8
V-
r-
£
oo
CO
r-
op
CM
f-
S
h-
in
0
0
h».
oo
r«.
S
0
to
o
to
<
s
CM
T*
V
^
00
o>
0
**
oo
CO
o
Y—
0
r~-
o
o
1—
Fluidized
*
*
h*
CO
S
Si
s
0
s
oo
CO
0
CO
o
00
1
CD
2
£
s
*»•
t^
t^
V
CO
in
o
|
CO
CO
o
to
o
o>
§
1
£t
CO
^
V
0
CO
oo
oo
CO
o
*-
o
co
g
in
2
2
S
^
00
»-
CO
0
s
CM
S
0
CO
o
••o-
<
i
i
f^
h>
R
CO
s
o
o>
o>
V
00
CO
o
~
CO
CM
CO
S
O
1
oo
CO
0
CO
o
f--
t
*
%
p~
z
n
t~-
v
S
in
o
o>
s
oo
CO
o
ID
O
OO
N
CO
oo
V
^
n
*
Ri
O
a
OO
in
o
T-
0
O)
I
0>
•o
ra
£
D.
0)
CO
(O
CO
co
co
<»•
in
T"
g
O
s
oo
in
o
CO
o
o
1
o
55
§
CO
oo
r-
oo
in
^
CM
*-
CO
^
o
CM
in
oo
tn
o
to
o
S
CO
00
s
in
01
in
in
o
2
S
0
T-
o
pj
o
a
8
CO
oo
1
in
8
i
o
5>
T-
co
in
o
CO
o
PI
in
co
CO
00
r~
CM
in
CO
in
O)
P(
0
eo
CO
S
o
to
0
s
g
s
M
U)
00
o
o
r>-
CM
S
o
^
o
in
s
in
n
CO
00
r*.
N
in
CO
in
§
o
00
p>
CO
in
o
CO
o
s
*>
-------
2
_«
m
•c
T
O
1
il
v>
o
TJ
C
O
E
c
c
'J
E
u
c
I
0
CM
j
1
to
1-
n
5
s
Emission
5
S
5
c
Combustio
1
Baseline
1
|V C
§ -
0 *
5 I
M
+
8 g
f
I
o
u
M
l||
III
Ul
to 2
tx J|
UJ
Emissior
Rate
(Ib/MMBtu
•« S
HI
« u.
w
5-2
° 1
Size
(MMBtu/hr)
S .
I-6
u.
0}
"S
u.
.
00
s
00
o>
o
N
B
o
0
oo
o
o
1
u.
(Residual
CO
s
CN
CD
N
in
M
0>
o
"
oc
o
CO
o
O)
erected
6
S
8
»
o
IN
O)
0
s
oc
o
CO
o
o
'
1
1
CD
CN
.
"
£
3
N
0>
o
CO
oc
o
o
rn
O
s
D
CO
00
D
CO
O
CO
N
o>
o
Si
OC
o
o
CM
S
.
s
CO
.
"
in
CM
T-
o>
o
s
oc
0
CO
o
CO
10
.
r
CO
N
N
O>
0
CO
oc
o
o
^-
o
5
.
s
CO
"
^
o»
o
s
T-
on
0
CO
o
in
0
.
o
CO
o
CO
o
n
CN
o>
o
1
oc
0
CD
O
CO
8
0
s
.
CO
CO
CO
o>
o
?!
CO
0
o
8
a
CO
s
N
CO
0
CO
N
O>
O
CC
cc
o
CO
o
CO
§
.
,
CD
s
CO
o>
3>
CO
O
c
on
o
CD
o
O)
m
§
00
00
CO
o
T-
*
o
0
1
"S
c
s
o
s.
a
g
o
in
CN
"
in
CN
a>
o
o
in
cc
o
CO
o
1-
«
> s
1
o
CO
.
o
.
CO
o
o
o
r-
a
o
CO
o
c>
o
s
'"
N
"
N
CO
O
CN
00
O
o
CO
01
CM
.
S
CD
0
5!
£
o
fO
o
T
m
a
§
3
S!
'-
o
C)
o
00
d
CO
o
in
o
00
CM
O
O
o
00
0
0
CO
g
•&
•3
il
1
oo
g
in
S
o
o
s
o
T-
g
0
CO
o
r>.
3
!
6
g
8
o
0
N
S
d
c>
§
o
CO
e>
00
Watertube
J
s
s
o
o
Co
S
o
CO
§
0
o
CO
s
CN
*
g
O
s
o
0
S
d
8
§
o
r
o
o
M
o
oo
a
o
o
o
s
o
.
s
o
CO
o
o
CO
CO
s
o
o
o
OC
00
00
o
o
a
1
M
o
oo
CM
s
o
o
CO
m
s
o
8
§
o
CO
o
CO
s
g
m
o
o
0
in
o
T-
S
o"
s
s
o
CO
o
oo
9
S
o
o
3
0
58
CO
o
o
in
o
§
S
s
n
S
0
o
£
00
o
o
T-
g
0
to
o
ffi
00
>
g
g
o
o
3
N
5
0
0
CN
OO
O
CO
O
^
CN
' OO
> CO
9 GO
38
n co
N CN
3 0
O f
CN
0 CO
3 0
i- CO
ri o
OO O)
o
o
t—
•o «
0 JC
en r
u ^
co n
0. S
"*
D
co
N-
O
n
8
CN
0
*
OO
0
CO
o
s
0
0
N-
"
f
CN
5
CO
0
o
IT
CM
5
O
CO
<0
l>-
"
•:
Jo
CN
S
00
o
CO
o
s
CM
3
O
n
h~
8
CO
CO
3>
CM
d
cc
00
0
CO
o
CO
#
o
00
JJ
o
o
in
tion
d for SCR ov
3 «
^#
(K 0
0 »
§.-*
^0
SO
r> i.
: Selective (
SNCR ove
«, .5
jction, SCF
reduction
cc. S
0
o 2
?«
5 •>
|5
a: -5
n-:
stion Control
reduction give
I 1
o o
U S
II 0.
8-
A. 1-2
-------
A.2 SECONDARY EMISSIONS FROM COAL-FIRED MODEL INDUSTRIAL BOILERS
-------
-------
Secondary Emissions (tons/yr) (b) I
a
u
CQ
+
8
6
2
n
•f
u
o
u
o
Baseline
8
0
CM
"Z.
ro
as
8
0
CN
2
ro
X
8
0
CN
•z.
ro
z
8
0
(N
n
X
Z
Capacity
Factor
Model
Number
Vi
£
o» •*•.
N 3
•* +J
M CQ
Z
Furnace
Type
a
>
a
6u
in
tM
0
CN
i-l
VO
«*
in
tN
0
in
CM
o
n
0
i-i
I-H
^
1-1
in
CN
ro
•-I
oo
•-I
fH
^
O
fH
i-l
Tt
O
VD
O
i-(
tH
*r
1-1
in
CN
n
1-1
oo
.-!
tH
^
O
1-1
fH
TT
O
ro
o
CN
CN
0>
1-1
o»
^
vo
CN
r-
1-1
CN
CN
ff>
O
CN
CN
O
VO
0
CN
01
cn
1-1
o>
^
VO
CN
r-
t-i
CN
CN
ff>
O
CN
CN
O>
O
ro
0
«*
^«
r*
1-1
ro
CO
O\
i-l
in
n
n
«*
«»
r>-
i-S
0
*»
f
r^
rH
o
VO
0
i-i JN n *r in vo
o
in
tN
o
o
in
o
0
o
tH
^
Pulverized
Coal
<
<
z
u
^
CN
VO
i-(
i-l
VO
O
i-l
VO
O
•H
O
4
z
z
rt
in
o»
1-1
CN
CN
r-
•H
O
CN
P-
i-H
O
fO
o
1
g
g
o
tH
03
f)
n
'S-
rf
ro
O
V
^r
n
o
VO
0
r^ eo o>
O
o
•H
g
g
g
^
VO
fH
•-I
CN
«»
r-(
0
CN
00
i-H
rH
VO
00
O
f-l
i-H
VO
00
o
vO
o
<
z
<
z
g
oo
ft
ro
ro
«*
a*
CN
o
^
o>
CN
O
1-1
o
<
Z
g
g
in
CN
*t
a>
00
1-1
i-t
vO
00
0
1-1
1-1
vO
co
o
ro
0
g
<
Z
g
cr.
^1
00
ao
i-l
r-
iH
CN
CN
f-l
r~
1-1
O
CN
CN
i-l
r~
i-i
0
vO
O
<
Z
<
VO
f-l
ro
VO
vO
r-
t--
in
O
r~
r«
in
O
r-i
o
g
g
g
CTt
t
ao
ao
f-i
f»
f-i
CN
CN
tH
r-
fH
O
CN
CN
iH
r»
rH
O
ro
O
g
g
g
00
cn
r~
r*
ro
ro
ro
«*
Tt
ro
«*
ro
O
^t
f
ro
»»
ro
O
vO
O
O'-lCNro'ij'invDr-oo
O
in
CN
o
o
in
o
o
o
iH
•a o
0) -H
N *J
•-I CO
•0 3
"3 T3 6
^H 0) 0
b m u
t-1
0
o
CN
f-l
fH
iH
O
0
fH
O
o
1-1
0
CT>
CN
f-»
o
in
ro
CN
CN
f-l
o
CM
fH
O
ro
O
r>-
CN
O
O
fH
in
ro
«*
CN
O
•<*
CN
0
vO
O
CN
rt
O
^
CN
iH
CN
fH
O
CN
iH
O
fH
0
in
CN
o
CN
tH
vO
^«
in
CN
o
in
CN
0
ro
o
fH
fH
5."
fH
in
CN
ro
fH
00
fH
rH
5*
O
rH
fH
*t
O
VO
o
^J-
fH
0
00
*t
ro
t
fH
0
**
fH
O
r-«
0
fH
fH
^1
fH
in
CN
ro
rH
00
fH
fH
*»
O
fH
rH
T
O
ro
O
CN
CN
01
fH
CT>
*!•
vO
CN
r~
iH
CN
CN
0>
O
CN
CN
a>
o
vO
0
OfHCNro^invOI--
CNCNCNCNCNCNCNCN
O
O
i-l
o
in
CN
o
o
m
Vj
Q)
•a V4
cd cu
01 .*
U 0
a^j
to co
r-t
-------
CO
u
0)
o
m
ed
•H
LI
a
3
01
73
•a
u
t,
i
n
ed
T3
C
ed
O
C
O
•H
n
n
M
(0
Second
CC
0)
•*
"a
a
K
to
8
N
z
en
X
2
8
8
o
(N
Z
en
S
8
o
z
en
a
Capacity
Factor
Model
Nunnber
V
«•-«
N g
~4 w
(0 0
a
Furnace
Tvoe
i
j
EH
"o
h
o
o
o
o
o
o
o
o
0
•H
•H
O
i-l
*
O
-«
o
en
0
iH
N
O
r*
CM
O
CM
O
4>
O
oo •
o
VO
O
i-« CM «*»
en en ro
O
in
CM
»
iH
r-t
O
(-(
rH
0
^
O
o
•«*
en
en
t
iH
n
^
o
t
o
n
o
in
CO
i-i
in
00
o
00
o
VO
o
in vo
o
o
m
CM
n
o
o
tH
CM
O
rH
CM
O
CM
O
VO
O
O «-l CM
o
o
«H
O
.-t
0
o
iH
o
fm
O
0
•-•
CM
O
iH
CM
O
CM
O
"t
0
«n
*•
r-t
en
^1
o
^1
o
VO
O
n r» in
O
in
CM
Packaged
Watertube
r-l
«
•a
~4
0 rH
0) "4
«0
0
o
o
o
o
o
0
o
o
0
iH
O
O
^
o
-^
o
en
O
iH
CM
0
iH
CM
O
CN
O
VO
O
vo r~ oo
o
o
i-4
O
i-l
O
o
r-4
o
*
o
0
tr>
i-i
CM
O
rH
CM
0
CM
O
en
o
CM
*
1-1
CM
^
0
4 CD
Bu 0 *
O
O
o
o
o
o
o
o
o
O
iH
o
o
-
0
-
0
en
O
•-«
CM
0
H
CM
O
CN
O
VO
C
oo o» o
m in vo
O
o
iH
O
1-t
o
^
o
r*
o
^
o
o
!-<
CM
O
t
H
CM
O
CM
O
en
o
CS
•*
rH
in
i-4
«-4
iH
c-
CM
^
o
CM
^l
O
VO
O
1-4 CN en
vo vo vo
O
in
CM
•oJS
O 9
O>4>
ed U
X 9
U 4J
« «
Q. 3
m
0) (I)
-P O
nt
r-l iH
•H *
-4 fc
*J-^ 3
ffl r-t 4J
-4 -4 efl
Q O Z
U -1
« 41
*
• V
u «
z o
u u
& i
-o
u • c
o u o
f e a
•-
C 4-> O
• O O
e 5 .
one
« TJ
a y «i
• "
one
~4 3 o
e 4 6
i '
5 8
A.2-2
-------
A.3 ENERGY IMPACTS OF NOX CONTROL OPTIONS ON COAL-FIRED MODEL
INDUSTRIAL BOILERS
-------
-------
n
u
0
o
oa
a
•a
•o
0
K)
O
u
c
0
n
c
0
0
Vi
4J
0
o
X
i
o
o
u
a
E
0
U
I
ro
.a
n
«
o
JH
0
W
U
trs_
S*|
w
u m
Baseline
MMBtu/yr
Capacity
Factor
"3 0
£
N 9
•H *J
to a
Furnace
Type
I
H
rH
0
9
h
o
o
rH
ro
CN
00
CN
O
o
o*
in
rH
O
0
ro
O
o
o
CN
VO
CO
CN
O
O
rH
O
0
vO
O
O
o
CN
VO
00
CN
O
0
r-
rH
ro
O
o
ro
O
O
o
in
CN
CO
CN
o
0
in
ro
vo
o
O
vO
o
o
0
in
CN
cn
00
CN
o
o
in
fO
vo
o
o
ro
o
rH
O
o
o
in
CO
rH
00
CN
o
0
o
cs
rH
O
o
vO
o
rH r\j ro ^ in vo
0
in
CN
0
0
in
o
o
o
rH
Pulverized
Coal
g
•5
g
»
o
o
rH
O
o
rH
O
z
Z
«
o
o
0
o
ro
O
z
g
*
o
o
VO
CO
o
o
vo
O
r-coo
o
o
rH
Z
Z
cn
O
o
VO
ro
O
O
rH
0
g
g
«
O
o
t-
o
rH
O
o
ro
O
g
g
*
O
o
m
rH
CN
O
O
VO
O
rt
z
z
CTI
O
O
CN
o
o
rH
O
g
g
»
O
0
in
rH
CN
O
o
ro
O
g
g
c*
o
o
o
ro
O
O
vO
O
g
g
«
O
0
rH
O
0
rH
O
g
g
«
O
0
o
o
o
ro
O
g
g
on
O
o
IT
oo
o
o
vO
o
0 rH CM ro * in VO r- 00
o
in
CN
0
0
in
0
O
O
rH
•O 0
0 -^
N 4>
-i m
TJ 9
9 TJ E
rH 0 0
fa oa o
rH
O
O
rH
ro
O
O
m
CN
o
o
rH
O
O\
rH
O
o
CN
ro
ro
O
O
O
o
ro
O
rH
O
O
in
CO
rH
rn
O
O
oo
rH
O
O
VO
o
rH
O
o
ro
ro
O
O
rH
VO
O
O
rH
O
rH
O
O
rH
ro
CN
ro
ro
O
0
CO
rH
O
o
ro
0
rH
O
4,620
ro
ro
O
o
VO
ro
O
O
vO
O
rH
O
o
in
rH
ro
ro
0
O
ro
CN
rH
O
o
rH
O
rH
O
o
CN
VO
"*
ro
ro
0
O
Oft
vo
r*l
O
O
ro
O
rH
O
O
in
CN
Ch
ro
ro
O
O
CO
«O
O
O
vO
O
CNCNCNCNCNCNCNCN
O
o
rH
O
in
CN
0
o
in
IH
0
TJ K
n) 0
0 Jt
Ij O
Q.JJ
(0 (0
m
o
u
lytic Reduction, SCR • Selective Catalytic Reduction,
ted with the assumed CC levels
of 0.85
Selective Noncata
juirements associa
toiler efficiency
N 0
t«l *O
6*8
Z O> 9
w M a
- c «
rH 0
o c
U rH O
4J «
C 4J TJ
0 C 0
u 0 a
Combustion
t no increm
f output ba
Dplicable
H 3 U
n 0 4J
unco
o « u z
A.3-1
-------
h
O
ed
C
0
o
c
0
4J
c
X
o
o
n
u
21
0)
u
CN
m
H
CC+SCR (C)
(% of boiler
output )
CC+SCR
(MMBtu/vr)
5^«
il;°
o !L
b-
w f
if
i
a
Baseline
MMBtu/yr
Capacity
Factor
^H *••
o> ?
•o 7
o !
N :
U) t
Furnace
Type
1
i
i
t
t in
3 O
3 O
D 00
M r-
r> ON
• .
3 O
S O
D O
d o
oo o>
CM CN
in
O
c
in
H
ON
00
O
0
vO
O
O
o
o
r-l
in
0
o
in
VO
o
r
o
o
o
in
o
o
m
ON
o.
0
r-
O
0
en
c
in
o
i\
en
ON
o
0
O)
o
0
vO
o
<-l CM en
0
in
in
o
n
t-i
ON
O
O
o
o
o
in
O
3
IN
en
ON
O
0
CM
O
o
en
0
in
o
3
0
ON
0
o
0
0
vO
o
•o- in vo
0
o
in
in
en
o
D
CM
ON
0
O
o
o
0
in
en
o
3
0
ON
0
o
*
o
o
en
0
in
en
O
0
vO
in
r-t
ON
0
o
r*
00
0
0
vO
c
r> 00 ON
O
O
O
01
•0 3
1 01 P
T3 ^j ^4
r-l O 0)
0) 0) 4i
— i vi id
(M 0 »
in
en
o
D
CM
ON
O
o
o
f
o
in
en
o
a
ON
o
o
0
0
en
o
in
en
o
0
in
H
ON
0
o
00
o
0
VO
o
O •-! CM
0
o
1-t
in
en
0
o
m
vo
ON
o
o
o
o
o
in
en
O
n
H
0
O
o
0
0
o
in
en
o
3
r
en
ON
o
c
0
0
vO
o
en ^ in
o
in
CM
Packaged
Hatertube
id
•0
0) -H
0) •*
0.0
r>
O
D
CM
*
O
:
o
o
r
o
in
en
0
D
5>
O
C
ff,
o
o
en
O
vot-
in
en
O
O
in
O
O
en
o
m
en
o
3
r>
en
*•
o
0
en
If
o
o
VO
o
ON O •-!
« in in
O
in
CM
in
en
o
3
n
rH
*
O
IT
O
O
r
o
in
en
o
3
JN
en
*
o
c
in
f
o
o
en
o
in
en
o
3
a
o
0
vC
O
en
o
o
VO
O
CM en «t
in in in
O
0
in
in
en
o
3
0
*
?N
r*
o
o
o
in
en
o
i
*
o
o
vO
O
en
0
o
en
o
in
en
o
c
vO
I/I
*
o
en
VD
o
o
vO
o
in vo r~
in in in
o
O
o
1-1
01
•-* o «
0 0 4>
-4 U Id
fix « K
in
en
o
0
CM
vO
O
(N
o
o
O
in
en
0
X
VO
ir
en
o
o
en
o
in
en
o
£
in
vo
d
o
vt
O
O
vO
O
00 ON O
in in vo
o
0
iH
in
en
o
n
VO
vO
ft
O
7
O
o
o
in
en
o
r
ON
vO
0
r-
00
o
o
en
o
in
n
O
3
r
en
vo
*
o
^
^
o
vD
O
»H CM en
vO vO vo
O
in
CM
Packaged
watertube
B
« , e
Z O» 3
(0 u n
.88
•3-c
U -H O
•p «
C -p TJ
•H O +J
4J C 3
§•"£•
5 o 3
E e o
N 3 if
One
O < w
1
A.3-2
-------
APPENDIX B
COSTING PROCEDURES
-------
-------
APPENDIX B
B.O COSTING PROCEDURES
This section presents the methodology used to determine NOX
control option costs, cost effectiveness, and incremental cost-
effectiveness (ICE).
An example calculation applying the presented methodology is
also included. Sections B.2 through B.4 present the specific
cost procedures for the various control options. Cost procedures
and the basis for them are presented for direct and indirect
costs, as well as fixed and variable operation and maintenance
(O&M) costs.
B.1 Methodology
B.I.I Direct Costs
The equation to calculate direct cost (DC) for each control
option is:
DC [$/(MMBtu/hr)] = a * (size)b (B.I)
where:
a and b = Constants derived from regression analysis
size = Boiler capacity (MMBtu/hr)
The exception to this is SCR systems. The initial catalyst cost
for an SCR system is considered a direct cost but is not included
in this equation.
For a 100 MMBtu/hr boiler installing a given control option,
where "a" and "b" were determined to be 26,877 and -0.55,
respectively, the calculation is:
DC [$/(MMBtu/hr)] = 26,877 * 100 -0.55
= 2,135
B-l
-------
B.I.2 Indirect Cost Factors \
The equation to calculate an indirect cost factor (ICF) is: /
ICF = 1 + (IC/DC) (B.2)
where:
1C = Indirect cost
For the same 100 MMBtu/hr boiler with a direct cost of
$2,135/(MMBtu/hr), and indirect costs of $641/(MMBtu/hr), the
indirect cost factor is:
ICF = 1 + (641) / (2,135)
= 1 + 0.30
= 1.30
B.I.3 Total Capital Costs
The equation to calculate total capital cost (TCC) is:
TCC [$/(MMBtu/hr)] = DC * ICF (B.3)
The exception to this is SCR systems. As discussed in
B.I.I, the initial catalyst cost is a direct cost, however, in
determining total capital costs as discussed in section B.4.3,
separate ICF are used for the catalyst and the remaining DC
items.
For the example 100 MMBtu/hr boiler with a direct cost of
$2,135/(MMBtu/hr) and an indirect cost factor of 1.3, the total
capital cost is:
TCC [$/(MMBtu/hr)] = 2,135 * 1.3
= 2,776
*
B.I.4 Operation and Maintenance Costs
Operating and maintenance costs include fixed and variable
components. Fixed O&M (FO&M) costs are independent of capacity
factor and are estimated by either: %
B-2
-------
FO&M ($/yr) = c * (size)d (B.4)
where:
c and d = Constants derived from regression analysis
size = Boiler capacity (MMBtu/hr)
or
FO&M = e + (f * size) (B.5)
where:
e and f. = Constants derived from regression analysis
size = Boiler capacity (MMBtu/hr)
Variable O&M (VO&M) costs are expressed in $/yr. These cost
^equations are specific for each technology and are presented in
•the respective sections of this attachment.
B.I.5 Annualized Costs
The equation for calculating total annual costs (TAG) is:
TAG ($/yr) = ACC + FO&M + VO&M (B.6)
ACC = Annualized capital cost ($/yr)
FO&M = Fixed O&M costs ($/yr)
VO&M = Variable O&M costs ($/yr)
In the above equation, ACC is calculated from:
ACC ($/yr) = TCC * size * CRF (B.7)
*
where:
TCC = Total capital cost [$/MMBtu/hr)]
size = Boiler capacity (MMBtu/hr)
B-3
-------
CRF = i (1 + i)n / [(i+i)n -i] (B.8)
where:
i = Interest rate (decimal fraction)
n = Economic life of the equipment (years)
Assuming an interest rate of 0.07 and economic life of 20 years:
CRF = 0.07 (1 + 0.07)20 / [ (i + 0.07)20 . 3.]
= 0.27/ 2.86
= 0.094
For the example 100 MMBtu/hr boiler with a total capital cost of
$2,776/(MMBtu/hr) and a capital recovery factor of 0.094,
annualized capital costs would be:
ACC ($/yr) = $2,776/(MMBtu/hr) * 100 (MMBtu/hr) * 0.094
= 26,094
The TAG (equation B.6) is also expressed on a percentage of
total annual cost of steam generated. The total annual steam
cost (AST) is calculated from:
AST ($/yr) = size * CF / 1300 * 106 * 8760 * 6 / 1000 (B.9)
where:
size = Boiler capacity (MMBtu/hr)
CF = Capacity factor (decimal fraction)
1300 = Btu of heat input required to generate a pound
^
of steam
106 = Btu's per MMBtu
8760 = number of hours per year
6/1000 = cost of steam; 6 dollars per 1000 Ibs of steam
B-4
-------
B.I.6 Cost Effectiveness
The equation for calculating cost effectiveness (CE) is:
CE ($/ton NOX removed) = TAG / (Tons NOX) (B.10)
where:
TAG = Total annual cost ($/yr)
Tons NOX = Tons of NOX removed per year (tpy)
In equation B.10, tons NOX is calculated from:
Tons NOX = RemNOx * size * CF * 4.38 (B.ll)
where:
RemNOx = Difference between baseline and controlled
t NOX emission rate (Ib/MMBtu)
size = Boiler capacity (MMBtu/hr)
CF = Capacity factor (decimal fraction)
4.38 = Conversion factor, Ib/hr to ton/yr
For the example 100 MMBtu/hr boiler with a baseline NOX
emission rate of 0.68 Ib/MMBtu, a controlled emission rate of
0.40 Ib/MMBtu, and a capacity factor of 0.6, the tons of NOX
removed per year are:
Tons NOX (ton/yr) = (0.68 - 0.40) Ib/MMBtu *
100 MMBtu/hr * 0.6 * 4.38
= 74 tons N0x/yr
^
For the example 100 MMBtu/hr boiler with annualized capital
cost of $26,094/yr, negligible O&M costs, and 74 tons of NOX
removed per year, the cost effectiveness is:
I
B-5
-------
CE ($/ton NOX removed) = ($26,094/yr +0) / 74 ton/yr
= $353/(ton NOX removed)
B.I.7 Incremental Cost Effectiveness
Incremental cost effectiveness (ICE) is calculated by
determining the incremental change in both total annual costs and
tons of NOX removed between a less stringent control option (COp)
and a more stringent control option (e.g., combustion controls to
combustion controls plus SNCR).
The equation for calculating ICE is:
ICE = (TAG for COp #2 - TAC for COP #1) [B.12)
(Tons NOX for COp #2 - Tons NOX for COp #1)
where:
COp #1 = less stringent control option
COp #2 = more stringent control option
Other variables are as previously defined.
A set of hypothetical vales for costs and tons of NOX
removed are presented for use in an example calculation:
250,000 = AC for COp #1
450,000 = AC for COp #2
400 = Tons NOX for COp #1
1,000 = Tons NOX for COp #2
B-6
-------
Using equation 1.11, ICE is calculated to be:
ICE ($/tonNOx removed) - ($4?°'°°0/yr " "°'' 0°°{yr)
x ($1000 tpy - 400 tpy)
= 250 $/ton NOX removed
B.2 Combustion Controls
A review of the NOX emission limits for recently permitted
industrial boilers subject to the current NSPS, subpart Db
standard, indicates that these boilers are, at a minimum, meeting
NOX emission levels associated with combustion controls. As a
result, no additional capital costs will be incurred for new
boilers to meet these levels. However, an incremental O&M cost
equal to 3 percent of the direct cost for the combustion control
^equipment was assumed to reflect potential costs associated with
improved boiler and burner operation that may be required to
achieve the lower NOX emission limits.
The cost equation, is the same as was used for wall-fired
Subpart Da (utility) model boilers and is as follows:
O&M ($/yr) = 0.03 * [60,600 * (size)0-56] (B.13)
where:
60,600 and 0.56 = Regression coefficient
size = boiler capacity (MMBtu/hr)
B.3 Selective Noncatalytic Reduction
The costs for SNCR are based on a low-energy, urea-based
system with two levels of wall injectors.
B.3.1 Direct Costs
Direct costs include the urea storage system, the reagent
injection system, air compressors, and installation costs.
B-7
-------
Because industrial boiler are smaller than utility boilers, it
was assumed that lances (typically used to improve mixing between
reagent and flue gas) are not necessary. Also, since these are
new boilers, they can be designed for the appropriate residence
time. Specifically, the assumption of using two levels of wall
injectors and no lances lowers the costs of the injection
systems, compressors, and installation. Based on this
assumption, the case studies developed for utility boilers were
re-evaluated.
The results of these case studies were used to develop
simplified costing algorithms for industrial boilers. The
algorithm coefficients were derived by non-linear regression of
the cost data obtained from the case studies. The following
equation was derived for all three coal-fired boiler types (PC,
FBC, and spreader stoker).
DC [$/(MMBtu/hr>] = 26,877 * (size)"0-55 (B.14)
where:
26,877 and - 0.55 = Regression coefficients
size = Boiler capacity (MMBtu/hr)
The cost coefficients for oil- and natural-gas fired boilers were
nearly identical. The following was used to characterize costs
for both:
DC [$/(MMBtu/hr)] = 32,695 * (size) -0.59 (B.15)
where:
32,695 and -0.59 = Regression coefficients
size = Boiler capacity (MMBtu/hr)
B.3.2 Indirect Cost
Two categories of indirect costs were calculated by the case
studies: V contingency factor and engineering support costs.
The engineering cost is determined as a function of the unit
size, whereas the contingency is calculated as a percentage of
direct capital costs. The indirect costs typically ranged
between 20 to 30 percent of the total direct costs. An overall
B-8
-------
indirect cost factor of 1.3 was assumed for the calculation of
total capital costs.
B.3.3 Fixed O&M Costs
Fixed O&M Costs for SNCR include operation, supervisory, and
maintenance labor; and maintenance materials. Fixed O&M costs
estimated by the case studies were found to be independent of
furnace type and fuel type. Therefore, the following equation,
determined by the methods in section B.I, was used to estimate
fixed O&M costs for all boilers.
FO&M ($/yr) = 53,027 * (size)0-21 (B.16)
where:
53,027 and 0.21 = Regression coefficients"
Size = Boiler capacity (MMBtu/hr)
x_
This is the same equation used for subpart Da boilers.
B.3.4 Variable O&M Costs
Variable O&M costs for SNCR include costs of urea and
electricity. In addition, energy penalties associated with the
vaporization of the urea solution in the boiler, mixing air, and
dry gas loss are also included. The urea cost is a major
component of the total VO&M cost and was determined from the
following equation derived for subpart Da model boilers:
Urea Cost ($/yr) = UncNOx * Cost * NSR * size * CF * 0.5 *
60 / 46 / 2000 * 8760 * 2 (B.17)
where:
UncNOx = NOX emission rate at inlet to SNCR system,
i.e., after combustion controls (Ib/MMBtu)
Cost = Reagent cost (assumed to be $200/ton)
NSR = Normalized stoichiometric ratio (assumed to be
1.0 mole NH2/mole NOX)
size = Boiler capacity (MMBtu/hr)
B-9
-------
CF = Capacity factor (decimal fraction) \
0.5 = Mole urea/mole NH2 .J
60 = Molecular weight of urea
46 = Molecular weight of NOX
2000 = Ib/ton
8760 = hr/yr
2 = Ib urea mixture/lb pure urea
Based upon the case studies, the other variable O&M costs were
estimated to be 11 percent of the annual urea cost. Therefore
the total VO&M costs for SNCR were calculated from:
VO&M ($/yr) = 1.11 * Urea Cost (B.18)
B.4 Selective Catalytic Reduction
The SCR cos^t algorithms developed for utility boilers were
used to estimate costs for industrial boilers. The algorithms
are based on hot-side SCR technology (i.e., the catalyst is \
located between the boiler economizer and air preheater).
Catalyst life was assumed to be three years for coal-fired
boilers and six years for natural gas-and oil-fired boilers. A
normalized stoichiometric ratio of 0.82 and a NOX reduction of
80 percent was assumed for all cases. At this NOX reduction,
catalyst space velocities were assumed to be 3,200/hr for coal-
fired boilers, 5,000/hr for oil-fired boilers, and 14,000/hr for
natural gas-fired boilers. The cost of catalyst was assumed to
be $350/ft3.
B.4.1 Direct Cost
Direct cost for SCR includes both process capital and the
initial catalyst charge:
DC [$/(MMBtu/hr)] = process capital + initial
catalyst charge (B.19)
B-10
-------
Process capital is calculated by an equation of the form:
Process capital [$/(MMBtu/hr)] = a * (size)b (B.20)
Process capital includes ammonia handling, storage, and
injection; catalyst reactor housing; flue gas handling, air
preheater modifications; and process control. Cost equations
derived for wall, coal-fired boilers were used to cost all coal-
fired industrial boilers. A similar approach was used for oil-
and gas-fired boilers. The equations are listed below.
For coal-fired boilers:
Process Capital [$/(MMBtu/hr)] = 34,718*(size) -°-3 (B.21)
For residual oil-, distillate oil-, and gas-fired boilers:
Process capital [$/(MMBtu/hr)] = 35,276*(size)"°-33 (B.22)
The equation for estimating the cost of the initial catalyst
charge is calculated from:
Catalyst cost [$/(MMBtu/hr)] =
Flow * Cat$ / {SVf * [ln(0.20) / In (1 - NOxRed)]} *
100 (B.23J
where:
Flow = Fuel-specific flue gas flow rate in normal cubic
feet per kilowatt -hour(Nft3/kWh); (126
Nft3/kWh for coal, 100 Nft3/kWh for natural
gas and oil)
Cat$ = Catalyst cost ($/ft3) (assumed to be $400/ft3
for all fuel types)
SVf = Fuel-specific space velocity (hr"1)
(3,200/hr for coal, 5,000/hr for oil, and
B-ll
-------
14,000/hr for distillate oil/natural gas)
NOxRed = Target NOX reduction efficiency
(decimal fraction)
100 = Conversion factor (assumed 100 kW/MMBtu per hr)
B.4.2 Indirect Cost
Separate indirect cost factors were used for the process
capital and the catalyst cost. Indirect costs for the process
capital were estimated at 45 percent (ICF = 1.45). Indirect cost
factors for the catalyst were estimated at 1.25 for coal-fired
boilers, 1.20 for oil-fired boilers, and 1.15 for gas-fired
boilers.
Total capital cost is calculated by multiplying the process
capital by the process capital indirect cost factor, multiplying
the initial catalyst charge by the catalyst indirect cost factor,
and adding these two products together.
B.4.3 Fixed O&M Costs
Fixed O&M costs for SCR include operating, supervisory, and
maintenance labor; and maintenance materials. Fixed O&M cost
equations used here are the same as those derived for the subpart
Da model boilers. The equations derived for wall, coal-fired
boilers and wall, oil-and gas-fired boilers are shown below.
For coal-fired boilers, the FO&M costs estimated from:
FO&M ($/yr) = 284,600 + 514 * (size) (B.24)
where:
284,600 and 514 = Regression coefficients
size = Boiler capacity, MMBtu/hr
and for oil- and gas-fired boilers the FO&M costs were estimated
from:
FO&M ($/yr) = 264,800 + 326 * (size) (B.25)
where:
264,800 and 326 = Regression coefficients
size = Boiler capacity, MMBtu/hr
B-12
J
-------
B.4.4 Variable O&M Costs
Variable O&M costs for SCR include catalyst replacement and
disposal, ammonia, electricity, and steam. Cost for these
elements are the same as those derived for utility boilers.
Catalyst replacement and disposal cost ($/yr) =
Flow*(Cat$ + 160) / {SVf* [ln(0.20) / In (l-NOxRed) ] } /
CL * (size) * 100 (B.26)
where:
Cat $ = Catalyst cost ($/ft3) (assumed to be $350/ft3)
160 = Cost to cover installation and disposal of
replacement catalyst ($/ft3)
SVf = Fuel-specific space velocity (hr~l)
(3,200/hr for coal, 5,000/hr for oil, and
14,000/hr for distillate oil/natural gas)
NOxRed = Target NOX reduction efficiency
(decimal fraction)
CL = Catalyst life (years)
Size = Boiler capacity, MMBtu/hr
100 = conversion factor
The equation for estimating costs for the other variable O&M
components is:
Other VO&M Cost ($/yr) = [1.88 + (4.3 * UncNOx * NOxRed)] *
CF * (size) * 100 (B.27)
where:
1.88 and 4.3 = Regression coefficients
UncNOx = NOX emission rate at inlet to SCR system
(Ib/MMBtu)
NOxRed = Target NOX reduction efficiency
(decimal fraction)
CF = Capacity factor (decimal fraction)
size = Boiler capacity (MMBtu/hr)
100 = Conversion factors
B-13
-------
The total variable O&M costs are then calculated from:
VO&M ($/yr) = Catalyst replacement and disposal
cost ($/yr) + other VO&M cost ($/yr) (B.28)
B-14
-------
APPENDIX C
MODEL INDUSTRIAL BOILER EMISSIONS AND COST DATA
-------
-------
O
O
T3
ra
10
HI
.0)
'5
CD
ra
o
O
O
ra
"w
Reductions
.1
8
p
UJ
TO
s
1
o
O
c
Combustio
Baseline
o>
Baselin
CC
o
(0
+
8
cr
£
(O
+
0
o
8
15
c
Annual
c
u
'8
E
UJ
£•
u
(0
Q.
3
0)
1
.§
U)
Furnace
9)
CL
£•
0)
u.
I
1.
#
1
1
#
I
I
g
Emissions
(ton/yr)
,— .
t) CD
«I
""I
Emissions
(ton/yr)
3"
£ CD
1
%
ta
LL
O
i»
5 '
2
&
£"
N.
o>
00
00
s
r-.
3
CO
o
CO
CM
CM
00
fM
O
r>-
oo
co
O
T-
o
f^
8
Fluidized
<
i
t^
CO
I-.
3
8
CM
O
o
in
8
o
CO
o
oo
S
CO
i
<
S
r>-
5
S
CN
O
O
O
T-
S
O
CO
o
o>
Combustion
<
I
CO
h.
O
CM
S
CM
O
CM
8
o
t—
o
o
«
<
i
3
t^
O)
u>
$
CN
O
in
CM
i^
S
o
CO
o
r-
i
I
i
r-
00
*—
co
CM
O
S
CM
%
O
(O
o
CM
I
s
S
t^
o>
CO
?
CM
O
S
oo
co
o
T-
o
co
8
<
<
i
r-~
00
^
CO
^
CM
O
o
in
CM
00
CO
o
CO
o
•*•
i
S
8
r*.
f*
CO
CM
CO
8
CM
O
%
*r
oo
CO
0
CO
o
in
i
2
§
r^
o>
r«.
oo
CO
CM
o
£
oo
CO
o
*-
o
CO
*—
i
1
8
1^
h~
CO
CM
CO
8
CM
O
i
00
CO
o
co
o
N.
i
i
%
t*.
!•*
CO
5
8
in
CM
o
8
oo
CO
o
CO
o
co
CM
CO
oo
T
1—
in
^-
S
in
o
in
CM
oo
in
o
»-
o
o>
8
Spreader
S
CO
00
CO
in
^
S
in
o
E
8
o
CO
o
o
1
55
8
CO
oo
£5
in
CM
CO
T-
m
o
S
8
o
CO
o
s
£3
%
in
o>
m
in
in
o
S
8
o
s
o
CO
o
CM
in
T™
co
S
f>-
CM
in
CO
in
O)
S
in
0
S
S
o
CO
o
S
in
o
T-
CO
oo
ff
m
00
*—
o
T*
T-
m
o
S
00
m
o
*—
o
CM
8
in
CO
CO
oo
f»
CM
fe
S
§
in
o
s
s?
o
CO
o
8
^
CO
CO
CO
00
S
v
t^
in
u
* Z
co co
I fi
Is.
P
I i
(0 c
•II
U
o «
n Q-
c-i
-------
T?
>ata. (Continue
u
i/
o
o
•o
e
<5
Emissions
1
S2
O
PQ
^
"S
.S
"c
U
U
o
X
H
«
V5
J
1
~?
1^
^^ ^ — '
~a
5
5
J
CO
I
& ^
W
55
/^
S §
tif j>
O w
^1
N °^
^2 w
2 "*
c ""
u«
cx
^
tn
O
U
1^
•3 12,
3
"53
O >™N
u •!
C8 d
^ rvs
ca 5
u e
« s
0 &
(-
rV ^
" 2
¥ g|
J +
" s
li
(W ui O
+ > O
o
u
£
%*%
+ > u
^
+ BS
U C.
U »:
+ g
u ¥
u&
u
u
+ a
u u
U OQ
+ U
u 5
u£
u
u
+ tf
u u
u w
+ g
U 7
U c«
U
U
+ pj
u u
U V3
i|
CJ ^;
^
c
"w
"85
S^
s
vu~
o
<•)
oo
n
ro
n
$
rs
40,040
m
—
o
00
VO
o
1
"^
O
fN
1^-
0
(N
3
1
fN
•n
^^
N
(N
•n
•*
u-i
—
VO
"*
m
m
ro
o\
VO
m
<*
CN)
m
•^
o
•"3-
O
f^
—
V.
VO
o
(N
O
g
VO
VO
§
Tt
ts
fo
VO
r^
s
(N
^-«
s
ro
vO
f>
n
«
00
^^
o
CO
O
tN
n
1,480,670
(N
in
S
•
00
c*%
<
o
f*l
o
-
0
r^
<
<
T
|
ro
<:
o
•«!•
O
n
o
<
2
180,770
0
r^
•^
§
r^
O
OS
<
<
T
S
^
Os
<
O
o
""*
o
<•
2"
rs
<
2
o
oc
•*
r-
O
f
—
OC
VO
0
o
<
<
*7
S
OS
VO
<
O
•«*
ts
o
in
OS
i/-
O
t»l
<
<
o
1
•*
<
y
VO
n
*~^
o
*t
•*i
o
<
2
287,410
o
fi
i/-
Os
ir
O
rt
<
<
O
rs
<
V
«
o
*f
2'
rs
<
2
•*
•«*
in
rs
m
O
O
m
m
OS
O
in
o
m
3
<
s
•«t
^
<
in
vo
rs
O
*f
2
o
00
<;
2
325,070
o
vD
0
00
g
r-
o
vO
<
<
1
rs
<
*7;
rs
o
"~^
o
sf
2
r*S
r~s
<;
2
00
r-
m
t^
ro
O
S
O
00
oc
r-
o
t~-
<:
•<
s
^~
<
•7
o
0
^t
2
O\
<
2
o
r;
OC
in
^
o
o
vn
o
00
0
r^
c
oc
o
fl
VO,
n
3
o
O
Os
'n.
^
•n
o
r^
0
r-
'S-
$
t~-
o
Os
O
7^
•—
O
s
vo
>n
o
oo
vO
rs
•
i
in
m
vo
m
CN
O
oo
00
rs
^r
CO
I—I
o
r~
VO_
m"
t--
oo
256,000
0
o
m
—
«
VD
O
r*~j
O
VO
*
m
m.
m
n
O
<•<•>
in
o
3s
OO
m
O
•*
$
Os
fi
r^
Os
5
m
O
r^
O
oc
rs
O
r"->
m
Os
\r
0
in
x>
oo.
D
1
m
8
rs
3
^i
f
§
in
1—1
O
.
rs
rs
r~
in
§
r~-
r^
m
344,450
0
o
r*>
in
o
i/-
o
SO
>n
rs
in
O
rs
O
f>
"
D
N
rS
_
*-"*
0
00
—
vq
m
r-~
o\
f>
•<»•
r^
in
ON
CO
t
O
O
f)
in
Os
team cos
So
II ^
g s
%-x
"ig
11
o *-- •
u tf
c o
2 ^
tS -0
J §
0= §
u i
vJ o>
o
-------
D
0
O
c
to
w
UJ
1
E
'o
CO
V
•o
o
•a
CO
6
6
£
"n
Reductions
o
J
UJ
4
a
"o
S
Combust!
Baseline
4)
1
to
m
o t
w S,
0 ?
o g
8 *
8 ^
I
Annual
Emissions
(ton/yr)
3
0) CO
"I
- c ->
CO O ^
§ #!
< E&
UJ
e -g-
o *•*
.E ^ S
uj 5
$£
"5 .
•6 o
1Z
Size
(MMBtu/hr)
4)
U 0)
c :£
S.
"S
if
*~
8
oo
8
8
oo
en
o
CM
00
CM
O
T-
o
00
CM
8
~&
u.
[Residual
8
8
"
8
CM
8
en
o
ro
a
o
CO
o
s
erected
O
3
8
en
8
S
8
S
en
o
^
oo
CM
O
co
in
CM
S
CM
en
0
v
CO
o
T-
o
co
»
CM
O
O)
3
in
CM
(O
S
en
0
in
CM
i-
n
o
CO
o
3
'-:
.
oo
in
»
«
in
CM
O)
O
o
in
CM
co
o
(O
o
in
co
S
CM
S
O
O
8
o
oo
o
o
T"
0
*
I
1
u.
(Distillate
oo
S
in
m
o
o
8
o
5
oo
o
o
co
o
*
erected
Oil/Natural
*~
o
oo
o
in
o
0
CM
8
o
CM
8
o
to
o
*
Watertube
,
s
o
in
o
0
en
8
o
en
oo
o
o
T-
o
en
(M
CM
§
CO
10
o
o
8
8
0
8
8
o
CO
o
in
CM
O
OO
8
in
o
o
CO
in
8
o
in
8
o
to
o
in
*"
o
00
en
o
in
o
o
oo
8
o
00
oo
o
o
^
o
in
8
v
S
8
in
o
o
CO
in
oo
o
o
CO
in
8
o
co
o
CO
in
S
§
"
»
o
o
§
8
o
U)
o
8
o
(O
o
a
8
§
oo
T-
in
0
o
in
CO
8
o
m
CO
8
o
T-
o
"
'
s
s
8
in
o
0
i
8
o
s
00
o
o
CO
o
8
§
s
i
in
0
0
o
CM
8
o
o
CM
00
O
o
(O
o
in
oo
in
(O
CO
CO
CO
in
CM
o
00
oo
o
T-
0
«
§
ct
o
CM
00
CO
(O
00
CO
co
CM
O
CM
oo
o
CO
o
•n
watertube
"*
oo
8
(0
<0
CO
CM
CO
CM
O
*
00
O
(O
o
8
*~
co
CO
(0
8
CO
CM
0
8
oo
o
T-
o
<0
CM
in
oo
CO
03
CM
CO
CM
O
en
in
oo
o
CO
o
s
s
oo
en
to
co
CO
CO
en
CM
o
oo
00
o
-------
UJ -s
1
O
8 8
O
CM
CM
CM
CO
CM
2
o>
8
CM
CM-
CO
8
CM
CM
CM
CM
8
OJ
CM
CM
(N
CD
oo
UJ
+ Of.
88
8
CM
(D
O>
CO
^^ W*
28
«D
i
I
to
*
f
§
II
CO
O)
0>
(D
CM
co
co
in
in
CD
in
in
in
co
8
CM
(D
m
CM
oo
o>
CO
GO
in
in
co
CM
(O
co
co
m
co
co
in
in
co
CM
in
in
CM
CM
CM
CO
CM
00
O
0>
CO
o>
m
CM
a>
co
CM
CM
CM
in
CM
O
1
o
•o
§
UJ
+ o:
o o
O V)
8
CM
CM
8
oo
8
8
o>
co
CM
CM
8
co
O)
8
CM
co
a>
O)
CM
S
i
•5
S
P'a
II
o --
O S.
I3S
o
•o
3
a
a
8
8
in
CM
O)
m
CM
a»
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-453/R-95-012
2.
4. TITLE AND SUBTITLE
New Source Performance Standards, Subpart Db - Technical
Support for Proposed Revisions to NOx Standard
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards Division
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
June 1997
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
H. CONTRACT/GRANT NO.
68-D1-0117
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report characterizes nitrogen oxide (NOx) emissions from industrial steam generating units having
capacities greater than 100 million Btu hour heat input and assesses the performance and costs associated
with controlling NOx emissions. Available data on NOx emissions from industrial and utility steam
generating units are summarized. Various control technologies for reducing NOx emissions, both
combustion modifications and add-on controls, are reviewed. Performance data and operational
experience for NOx controls which have been applied to industrial steam generating units are presented.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Control Costs
Industrial Boiler
Nitrogen Oxide
Steam Generating Unit
1 8. DISTRIBUTION STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
19. SECURITY CLASS (Report)
Unclassified
20. SECURITY CLASS (Page)
Unclassified
c. COSAT1 Field/Group
21. NO. OF PAGES
250
22. PRICE
EPA Form 22ZO-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
|