United States
      Environmental Protection
      Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-453/R-94-012
June 1997
      Air
EPA  New Source Performance Standards,
       Subpart Da - Technical Support for
       Proposed Revisions to NOx Standard

-------
                                     EPA-453/R-94-012
New Source Performance Standards,
 Subpart Da - Technical Support for
Proposed Revisions to NOx Standard
      U S Environmental Protection Agency
           Combustion Group/ESD
    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 UTILITY BOILERS  	   2-1
                                                        f

     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

     2.2  Utility Boiler Designs  	   2-7

          2.2.1  Fundamentals of Boiler Design and
                 Operation	2-10
          2.2.2  Furnace Configurations and Burner Types  2-13
          2.2.3  Other Boiler Components  	  2-25

     2.3  Fossil Fuel Characteristics 	  2-32

          2.3.1  Coal	2-32
          2.3.2  Oil	2-37
          2.3.3  Natural Gas	2-39

     2.4  Utility Boiler Emissions  	  2-42

          2.4.1  Nitrogen Oxide Formation 	  2-42
          2.4.2  Factors that Affect NOx Emissions  .  .  .  2-47

     2.5  Baseline Emissions  	  2-57

          2.5.1  Baseline Emission Levels 	  2-57
          2.5.2  Other Regulations on the Source Category  2-60

     2.6  References	2-61

3.0  NITROGEN OXIDES EMISSION CONTROL TECHNIQUES  ....   3-1

     3.1  Combustion Controls for Coal-Fired Utility
          Boilers	   3-3

          3.1.1  Low NOx Burners for Conventional Boilers   3-3
          3.1.2  Low NOx Burners and Overfire Air for
                 Conventional Boilers 	  3-15
          3.1.3  Air Staging for Fluidized Bed Combustion
                 Boilers	3-20
                              111

-------
                 TABLE  OF  CONTENTS  (Continued)
                                                          Page
     3.2  Performance of Combustion Controls on Coal-
          11 Fired Boilers	3-21
                                                       t
          3.2.1  Summary of Available Long-Term (CEM)
                 Data	3-21
          3.2.2  Selection of Boilers for Analysis    .  .  3-26
          3.2.3  Analysis of Long-Term Continuous
                 Emission Monitoring Data 	  3-39
          3.2.4  Summary	   3-112

     3.3  Combustion Controls for Natural Gas- and Oil-
          Fired Utility Boilers   	   3-116

          3.3.1  Flue Gas Recirculation	   3-116
          3.3.2  Low NOx Burners	   3-116
          3.3.3  Combinations of Combustion Controls  .   3-128

     3.4  Performance of Combustion Controls for Natural
          Gas- and Oil-Fired Boilers	   3-128

     3.5  Flue Gas Treatment Controls for Coal-, Natural
          Gas- and Oil-Fired Boilers	   3-131

          3.5.1  Selective Noncatalytic Reduction .  .  .   3-132
          3.5.2  Selective Catalytic Reduction  ....   3-142

     3.6  Performance of Flue Gas Treatment Technologies
          On Coal-, Natural Gas-, and Oil-Fired Utility
          Boilers	   3-155

          3.6.1  Selective Noncatalytic Reduction .  .  .   3-155
          3.6.2  Selective Catalytic Reduction  ....   3-164

     3.7  Advanced Clean Coal Technologies  	   3-186

          3.7.1  The SNOX™ Process	   3-186
          3.7.2  The SOx-NOx-ROx Box (SNRB™) Process     3-189
          3.7.3  The NOXSO™ Process   	   3-191

     3.8  References	   3-195

4.0  MODEL BOILERS AND CONTROL OPTIONS  	   4-1

     4.1  Selection of Model Boiler Parameters  	   4-1
                              IV

-------
                 TABLE OF CONTENTS (Continued)
                                                          Page


          4.1.1  Fuel Type	  4-1
          4.1.2  Furnace Type	  4-3
          4.1.3  Boiler Size	  4-4
          4.1.4  Capacity Factor and Heat Rate  ....'.  4-4
          4.1.5  Baseline NOx Emission Rates  	  4-4

     4.2  Model Boilers	  4-6
     4.3  Control Options	  4-6

          4.3.1  Combustion Controls  	  4-6
          4.3.2  Selective Noncatalytic Reduction ....  4-9
          4.3.3  Selective Catalytic Reduction  	  4-9

     4.4  References	4-11

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	  5-4

     5.2  Liquid Waste Impacts  	  5-5
     5.3  Solid Waste Disposal Impact 	  5-7
     5.4  Energy Impacts	  5-7
     5.5  References	5-13

6.0  MODEL BOILERS AND CONTROL OPTION COSTS 	  6-1

     6.1  Costing Methodology	  6-1

          6.1.1  Total Capital Cost	  6-2
          6.1.2  Operating and Maintenance Costs  ....  6-4
          6.1.3  Calculation of Busbar Cost, Cost
                 Effectiveness, and Incremental Cost
                 Effectiveness  	  6-6
          6.1.4  Other Cost Considerations	  6-7

     6.2  Costing Procedures  	  6-8

          6.2.1  Combustion Controls  	  6-8
          6.2.2  Selective Noncatalytic Reduction ....  6-9
          6.2.3  Selective Catalytic Reduction  	 6-11

-------
                 TABLE OF CONTENTS (Continued)
                                                           Page
     6.3  Model Boiler Cost Impacts	6-12

          6.3.1  Combustion Controls	'.  6-13
          6.3.2  Selective Noncatalytic Reduction  ....  6-17
          6.3.3  Selective Catalytic Reduction   	  6-20
     6.4  References

APPENDIX A
                                            6-25
APPENDIX B



APPENDIX C

APPENDIX D
DETERMINATION OF FURANCE VOLUME AND BURNER
ZONE VOLUME FOR CONVENTIONAL SUBPART Da
BOILERS

PRIMARY AND SECONDARY AIR IMPACTS FOR
MODEL BOILERS

COSTING PROCEDURES

MODEL BOILERS EMISSIONS AND COST DATA
                               VI

-------
                        LIST OF TABLES


                                                          Paoe

2-1  Wall- and Tangentially-Fired Boilers Subject to
     Subpart Da	   2-3

2-2  Size Distribution of Known Utility Boilers Subject
     to Subpart Da	   2-5
                                                        *
2-3  Projections of Generating Capacity Growth for
     Boilers Commencing Operations Between 1993 and 2000   2-8

2-4  List of New Utility and NUG Boilers Projected
     to be Built Between 1996-2000  	   2-9

2-5  Classification of Coals By Rank	2-33

2-6  Sources and Typical Analyses of Various Ranks
     of Coal	2-35

2-7  ASTM Standard Specifications for Fuel Oils	2-38

2-8  Typical Analyses and Properties of Fuel Oils ....  2-40

2-9  Characteristics of Selected Samples of Natural
     Gas From United States Fields	2-41

2-10 Typical Fuel Nitrogen Contents of Fossil Fuels .  . ,  2-53

2-11 Baseline Emission Rates (Based on Subpart Da) For
     Nitrogen Oxide Emissions from Fossil Fuel-Fired
     Boilers	2-58

2-12 Fifth-Year Baseline NOX Emissions Estimate for New
     Utility and NUG Boilers	2-59

3-1  NOX Emission Control Technologies for
     New Fossil Fuel Utility Boilers  ..... 	   3-2

3-2  Long-Term Data Summary for Conventional
     Subpart Da Boilers	3-22

3-3  Long-Term Data Summary for Subpart Da FBC
     and Stoker-Fired Boilers 	  3-24

3-4  Summary of Long-Term Data	3-27

3-5  Design Parameters for Subpart Da Conventional
     Boilers	3-28

3-6  Short-Term NOX Emission Data for Subpart Da


                              vii

-------
                  LIST OF TABLES  (Continued)
                                                           Page


     Conventional Boilers 	  3-33

3-7  List of Boilers Selected for Long-Term Data
     Analysis	>.  3-40

3-8  Summary Statistics for AB Brown Unit No. 2
     (24-Hour Block Averaged NOX Data)  [1989 Data Set]   .  3-53

3-9  Achievable NOX Emission Limits for AB Brown
     Unit No. 2 (1989 Data Set)	3-54

3-10 Summary Statistics for AB Brown Unit No. 2
     (24-Hour Block Averaged NOX Data)  [1992 Data Set]   .  3-56

3-11 Achievable NOX Emission Limits for AB Brown Unit
     No.  2 (1992 Data Set)  	3-57

3-12 Summary Statistics for Big Bend Unit No. 4
     (24  Hour Block Averaged NOX Data)  [1985 Data Set]   .  3-67

3-13 Achievable NOX Emission Limits for Big Bend
     Unit No. 4 (1985 Data Set)	3-68

3-14 Summary Statistics for Big Bend Unit No. 4
     (24  Hour Block Averaged NOX Data)  [1992 Data Set]   .  3-69

3-15 Achievable NOX Emission Limits for Big Bend
     Unit No. 4 (1992 Data Set)	3-71

3-16 Summary Statistics for Big Bend Unit No. 4
     (24  Hour Block Averaged NOX Data)
     [1992 Data Subset]	3-72

3-17 Achievable NOX Emission Limits for Big Bend
     Unit No. 4 (1992 Data Subset)  	3-73

3-18 Summary Statistics for Holcomb Unit No. 1
     (24  Hour Block Averaged NOX Data)   	3-79

3-19 Achievable NOX Emission Limits for Holcomb
     Unit No. 1	3-81

3-20 Summary Statistics for Holcomb Unit No. 1
     (Load-Adjusted 24-Hour Block Averaged NOX Data)   .  .  3-83

3-21 Achievable NOX Emission Limits for Holcomb Unit No. 1


                             viii

-------
                  LIST OF TABLES (Continued)


                                                          pace


     Based on Continuous Operation at Load Corresponding
     to Maximum NOX Emission Rate	3-85

3-22 Summary Statistics for WA Parish Unit No.  8
     (24-Hour Block Averaged NOX Data)  	'.  3-90

3-23 Achievable NOX Emission Limits for Parish
     Unit No. 8	3-92

3-24 Summary Statistics for WA Parish Unit No.  8
     (Load-Adjusted 24-Hour Block Averaged NOX Data)   . .  3-94

3-25 Achievable NOX Emission Limits for WA Parish
     Unit No. 8 Based on Continuous Operation at Load
     Corresponding to Maximum NOX Emission Rate 	  3-95

3-26 Summary Statistics for Dolet Hills Unit No. 1
     (24-Hour Block Averaged NOX Data)  	   3-101

3-27 Achievable NOX Emission Limits for Dolet Hills
     Unit No. 1	   3-103

3-28 Summary Statistics for Dolet Hills Unit No. 1
     (Load-Adjusted 24-Hour Block Averaged NOX Data)   .   3-105

3-29 Achievable NOX Emission Limits for Dolet Hills
     Unit No, 1 Based on Continuous Operation at Load
     Corresponding to Maximum NOX Emission Rate ....   3-107

3-30 Summary Statistics for TNP One Unit No. 1
     (24-Hour Block Averaged NOX Data)  	   3-113

3-31 Achievable NOX Emission Limits for TNP One
     Unit No. 1	   3-114

3-32 Summary of Statistical Analysis Results for
     Subpart Da Boilers	   3-115

3-33 Performance of Combinations of Combustion Controls
     On U. S. Natural Gas- and Oil-Fired Utility
     Boilers	   3-129

3-34 Performance of Selective Noncatalytic Reduction on
     Conventional U. S. Utility Boilers 	   3-156
                              ix

-------
                  LIST OF TABLES (Continued)


                                                          Page


3-35 Summary of Carbon Monoxide,  Ammonia Slip,  and
     Nitrous Oxide Emissions from Conventional  Boilers
     With SNCR	   3-rl59

3-36 Performance of U. S.  Fluidized Bed Boilers with NH3*-
     Based Selective Noncatalytic Reduction System  . .   3-163

3-37 Performance of Selective Catalytic Reduction Systems
     on U. S. Utility Plants	   3-165

3-38 Performance of Pilot Scale Selective Catalytic
     Reduction Systems on U.S.  Utility Plants 	   3-168

3-39 List of New Utility Boilers Using Selective
     Catalytic Reduction  	   3-175

3-40 Summary Statistics for Carneys Point Unit  No. 1  .   3-180

3-41 Achievable NOX Emission Limits for Carneys Point
     Unit No. 1	   3-181

3-42 Summary Statistics for Carneys Point Unit  No. 2  .   3-184

3-43 Achievable NOX Emission Limits for Carneys Point
     Unit No. 2	   3-185

3-44 The SNOX™ Process Coal-Fired Utility Boiler
     Demonstration Average Test Results 	   3-188

4-1  Effects of Model Boiler Parameters 	   4-2

4-2  Capacity Factors and Associated Heat Rates 	   4-5

4-3  Model Boilers	   4-7

4-4  Control Technology Performance Levels  	 4-10

5-1  Summary of NOX Emission Reductions for Model
     Boilers	   5-3

5-2  Summary of Secondary Pollutant Emissions for
     Model Boilers	   5-6

5-3  Summary of Energy Impacts for Model Boilers  ....   5-9

6-1  Capital and Operating Cost Components  	   6-3

-------
                  LIST OF TABLES (Continued)


                                                          Page



6-2  Fixed and Variable O&M Unit Costs	  6-5

6-3  Model Boiler Impact Ranges for Incremental Cost
     Effectiveness  	 6-14
                                                        •
                                XI

-------
                        LIST OF FIGURES
2-1  Simplified Boiler Schematic  ............  2-11

2-2  Single, Wall-Fired Boiler  .............  2-14

2-3  Circular-Type Burner for Pulverized Coal,  Oil,
     or Gas ......................  ••  2-16

2-4  Opposed Wall-Fired Boiler  .............  2-17

2-5  Firing Pattern in a Tangentially-Fired Boiler  .  .  .  2-18

2-6  Burner Assembly of a Tangentially-Fired Boiler .  .  .  2-20

2-7  Simplified AFBC Process Flow Diagram ........  2-23

2-8  Spreader Type Stoker-Fired Boiler - Continuous
     Ash Discharge Grate  ................  2-26

2-9  Variation of Flame Temperature with Equivalence
     Ratio  .......................  2-44

2-10 Fuel-Bound Nitrogen-to-Nitrogen Oxide in
     Pulverized-Coal Combustion .............  2-48

2-11 Fuel NOX to Fuel Nitrogen Content-Pulverized
     Coal, Pre-Mixed  ..................  2-49

2-12 Comparative Physical Sizes of Utility Boilers
     Firing Different Fuels .......... .....  2-54

2-13 Effect of Mill Pattern Usage on Nitrogen Oxide
     Emissions  .....................  2-56

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™ Low NOX Burner   .    3-7

3-3  RO-II Low Low NOX Coal Burner  ...........    3-8
                                  TM
3-4  Controlled Combustion Venturi  Low NOX Burner  .  .  .    3-9

3-5a Typical Fuel and Air Compartment Arrangement for
     One of a Tangentially-Fired Boiler .........  3-11

3-5b Plan View of Fuel and Air Streams in a Typical
     Tangentially-Fired Boiler  .............  3-11

                              xii

-------
                  LIST  OF  FIGURES  (Continued)
3-6a Low NOX Concentric Firing System Fuel and Air
     Compartment Arrangement  	  3-12

3-6b Plan View of Low NOX Concentric Firing System  .  .  •.  3-12

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 Systems	3-19

3-11 Comparison Between Furnace Volume and Burner Zone
     Volume for Conventional Subpart Da Boilers 	  3-30

3-12 Maximum Thermal Heat Input as a Function of Furnace
     Volume for Conventional Subpart Da Boilers 	  3-32

3-13 Short-Term NOX Emission as a Function of Furnace
     Heat Release Rate for Conventional Subpart Da
     Boilers	3-35

3-14 Short-Term NOX Emission as a Function of Burner Zone
     Heat Release Rate for Conventional Subpart Da
     Boilers	3-36

3-15 Short-Term NOX Emissions as a Function of Heat Release
     Rate Per Unit Furnace Plan Area for Conventional
     Subpart Da Boilers	3-37

3-16 Time Plot of Hourly NOX Emissions for January 1,  1989
     to December 31, 1989 (AB Brown, Unit No. 2)  ....  3-46

3-17 Time Plot of Hourly Load for January 1, 1989 to
     December 31, 1989 (AB Brown, Unit No. 2)	3-47

3-18 Hourly-Averaged NOX Emissions as a Function of Load
     (1989 Data Set) [AB Brown, Unit No. 2]	3-48
                             Xlll

-------
                  LIST OF FIGURES (Continued)
                                                          Page


3-19 Time Plot of Hourly NOX Emissions for October 1, 1992
     to December 30, 1992 (AB Brown, Unit No. 2)   .... 3-50

3-20 Time Plot of Hourly Load for October 1, 1992 to
     December 30, 1992 (AB Brown, Unit No. 2)	3-51

3-21 Hourly-Averaged NOX Emissions as a Function of Load
     (1992 Data Set) [AB Brown, Unit No. 2]	3-52

3-22 Time Plot of Hourly-Averaged NOX Emissions for
     April 23 to August 19,  1985 (Big Bend, Unit No. 4)  . 3-59

3-23 Time Plot of Hourly Load for April 23 to August 19,
     1985 (Big Bend, Unit No. 4)  	3-61

3-24 Hourly-Averaged NOX Emissions as a Function of Load
     (1985 Data Set) [Big Bend, Unit No. 4]	3-62

3-25 Time Plot of Hourly-Averaged NOX Emissions For the
     July 1 to September 30, 1992.   (Big Bend,
     Unit No. 4)  	3-63

3-26 Time Plot of Hourly Load for July 1 to September 30,
     1992 (Big Bend, Unit No. 4)  	3-64

3-27 Hourly-Averaged NOX Emissions as a Function of Load
     (1992 Data Set) [Big Bend, Unit No. 4]	3-65

3-28 Time Plot of Hourly NOX Emissions for October 1 to
     December 30, 1992 (Holcomb, Unit No. 1)  	3-76

3-29 Time Plot of Hourly Load for October 1 to
     December 30, 1992 (Holcomb, Unit No. 1)  	3-77

3-30 Hourly-Averaged NOX Emissions as a Function of Load
     (Holcomb, Unit No. 1)  	'	3-78

3-31 Time Plot of Hourly NOX Emissions for June 1 to
     July 31, 1992  (WA Parish, Unit No. 8)  	3-87

3-32 Time Plot of Hourly Load for June 1 to  July 31' 1992
     (WA Parish, Unit No. 8)  	3-88

3-33 Hourly-Averaged NOX Emissions as a Function of Load
     (WA Parish, Unit No. 8)  	3-89
                              xiv

-------
                  LIST OF FIGURES  (Continued)
                                                           Page


3-34 Time Plot of Hourly NOX Emissions for October 19 to
     December 15, 1992 (Dolet Hills,  Unit No.  1)   ....  3-98

3-35 Time Plot of Hourly Load for October 19 to
     December 15, 1992 (Dolet Hills,  Unit No.  1)   ....  3-99

3-36 Hourly-Averaged NOX Emissions as a Function of Load
     (Dolet Hills, Unit No.  1)  	   3-100

3-37 Time Plot of Hourly NOX Emissions for September 1,
     1992 to October 30,  1992 (TNP One, Unit No.  1)  .  .   3-109

3-38 Time Plot of Hourly Load for September 1, 1992 to
     October 30, 1992 (TNP One,  Unit No. 1)	   3-110

3-39 Hourly-Averaged NOX Emissions as a Function of Load
     (TNP One, Unit No.  1)  	   3-111

3-40 Flue Gas Recirculation System	   3-117

3-41 ROPM'K Burner for Natural Gas- and Oil-Fired
     Boilers	   3-119

              TM
3-42 Dynaswirl  Low NOX Burner	   3-120

        TM
3-43 ISC  Low NOX Burner	   3-122

                                             TM
3-44 Primary Gas-Dual Register Low NOX Burner   ....   3-123

3-45 XCL™ Natural Gas- and Oil-Fired Low NOX Burner    .   3-124
                                      *nj
3-46 Low NOX Swirl Tertiary Separation  Low NOX
     Burner	   3-125

3-47 Pollution Minimum™ Burner for Natural Gas- and
     Oil-Fired Boilers  	   3-127

3-48 Ammonia-Based Selective Noncatalytic Reduction .  .   3-133

3-49 Urea-Based Selective Noncatalytic Reduction  . .  .   3-134

3-50 High-Energy Selective Noncatalytic Reduction
     Process	   3-136

3-51 General Effects of Temperature on NOX Removal  .  .   3-138
                              xv

-------
                  LIST OF FIGURES (Continued)
                                                          Page


3-52 General Effect of NH3:NOX Mole Ratio on NOX
     Removal  .....................   3-140

3-53 Ammonia Salt Formation as a Function of Temperature"
     and NHs and SO3 Concentration  ..........   3-141
3-54 Relative Effect of Temperature on NOX Reduction  .   3-144

3-55 Possible Configurations for Selective Catalytic
     Reduction  ....................   3-145

3-56 Typical Configuration for a Catalyst Reactor .  .  .   3-147

3-57 Example of Optimum Temperature Range for Different
     Types of Catalysts ................   3-148

3-58 Configuration of Parallel Flow Catalyst  .....   3-150

3-59 Effect of Temperature on Conversion of S02 to SO3   3-152

3-60 NOX Reduction Versus Molar N:NO Ratio for
     Conventional U. S. Coal-Fired Boilers with Selective
     Noncatalytic Reduction ..............   3-161

3-61 NOX Reduction Versus Molar N:NO Ratio for
     Conventional U.S. Natural Gas- and Oil-Fired
     Boilers with Selective Noncatalytic Reduction  .  .   3-162

3-62a Extruded Catalyst NOX Conversion and Residual
      NH3 Versus NH3-to-NOx Ratio ...........   3-169

3-62b Replacement Composite Catalyst NOX Conversion
      and Residual NH3 Versus NH3-to-NOx Ratio  ....   3-169

3-63a Relative changes in V/Ti Activity with Exposure
      Time (TVA, Shawnee Site)  ............   3-171

3-63b Relative Changes in Zeolite Catalyst Activity
      with Exposure Time (TVA, Shawnee Site)  .....   3-171

3-64a Titanium Corrugated Plate Catalyst NOX Conversion
      and Residual NH3 Versus NH3-to-NOx Ratio  ....   3-173

3-64b Vanadium Titanium Extruded Catalyst NOX Conversion
      and Residual NH3 Versus NH3-to-NOx Ratio  ....   3-173
                              xvi

-------
                  LIST OF FIGURES  (Continued)
                                                           Page


3-65 Time Plot of Hourly NOX Emissions for July 1 to
     September 30, 1995 (Carneys Point,  Unit No.  1)  .  .   3-179

3-66 Time Plot of Hourly NOX Emissions for July 1 to
     September 30, 1995 (Carneys Point,  Unit No.  2)  .  .   3-183

3-67 SNOX™ Process Schematic	   3-187

3-68 SNRB™ Process Schematic	   3-190

3-69 Effect of Catalyst Temperature on NOX Removal
     (SNRB Process)	   3-192
          TM
3-68 NOXSO  Process Schematic   	   3-194

6-1  Effect of Fuel  Type and Furnace type on ICE for
     Combustion Controls  	  6-16

6-2  Effect of Boiler Capacity Factor and Size on ICE
     for Combustion  Controls  	  6-18

6-3  Effect of Inlet NOX Emission Rate on ICE for SNCR   .  6-19

6-4  Effect of Boiler Capacity Factor and Size on
     ICE for SNCR	6-21

6-5  Effect of Inlet Emission Rate on ICE for SCR ....  6-22

6-6  Effect of Boiler Capacity Factor and Size on
     ICE for SCR	'	6-24
                             XVI1

-------
                       1.0  INTRODUCTION

     This document supports regulatory development action
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 stearr. generating units covered in this document are
these currently subject tc 4C Cede of Federal Regulations
!CFR) 60, subpart Da.
     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 111(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, D.C., is another source of background information.
Types of information that may be included in the docket are  •
information provided by the affected industry, vendors, or
trade associations 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 Utility Boilers,
describes the affected industry.  The regulatory definition of
the industry and a description of the current and future
utility boiler population are provided.  Chapter 2.0 also
describes various types of utility 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 utility 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  (CC)
and flue gas treatment techniques.  Long-term continuous
emission monitoring  (CEM) data from representative coal-fired
boilers equipped with CC were evaluated to assess the
                              1-2

-------
continuously achievable NOX emission rate for CC techniques.
The performance of flue gas treatment techniques was evaluated
based on available long-term CEM data and short-term test
data.
     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
ir.pacts 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, annualized busbar cost, 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 four appendices.  Appendix A
describes the methodology used to determine the boiler design
parameters for existing subpart Da boilers.  Appendix B
contains tables showing the primary air impacts  (NOX emission
reductions)  and secondary air impacts (emissions of NH^, N20,
and CO)  resulting from the application of the NOX control
                              1-3

-------
options to each model boiler.  Appendix C describes the cost
procedures used for estimating the costs of the NOX control
options considered.  Finally, appendix D presents the costs
impacts associated with each control option for the model
boilers.
                              1-4

-------
           2.0  CHARACTERIZATION OF UTILITY BOILERS

     This chapter presents an overview and characterization of
electric utility boilers.  It is divided into five main
sections: source category description, utility boiler designs,
fossil fuel characteristics, utility boiler NOX emissions, and
baseline emissions.
2.1  SOURCE CATEGORY DESCRIPTION
2.1.1  Source Category Definition
     The revision of subpart Da of the New Source Performance
Standards (NSPS) applies to any fossil fuel-fired electric
utility steam generating unit that is capable of combusting
more than 250 million British Thermal Units per hour
(MKBtu/hr)  heat input,  and for which construction commenced
after the date of proposal.  Electric utility steam generating
units : furnaces, boilers, or ether devices'  are defined as
those that are constructed for the purpose of supplying more
than one-third of their potential electric output capacity and
more than 25 megawatts  (MW) electrical output to any utility
power distribution for sale.
     Besides conventional units (i.e., those that produce
steam for electrical generation only)  and cogeneration units
(i.e., those that produce steam for both electrical generation
and process heat),  the revision to the current standard also
applies to electric utility combined cycle gas turbines that
are capable of combusting more than 250 MMBtu/hr heat input of
fossil fuel in the steam generating unit.  Only emissions
resulting from the combustion of fossil fuels in the steam
generating unit are subject to the revision to the current
standard (the gas turbine emissions are subject to
subpart  GG).
                              2-1

-------
2.1.2  Current and Future Industry Description
     This section describes the population of boilers
currently subject to the existing standard and the population
projected to be subject to the revised standard.   The source
category includes both utility and non-utility owned boilers
(referred to as non-utility generators [NUGs]).   Based on the
Energy Information Administration Energy Outlook Supplement  •
for 1995, the role of NUGs is increasing.  In 1993,  NUG
boilers represented 1.7 percent of the total generating
capacity attributable to boilers.  That percent  is projected
to increase to 2.5 in the year 2000 and to 3.6 in 2010.'  As
discussed later in section 2.1.2.2, the rate of  growth for new
boilers is small.  However, the large size of each new boiler
makes it a major NOX emitter.  Section 2.5,  Baseline
Emissions, provides an estimate of NOX emissions for the
projected boiler population in the absence of control
resulting from the revision to the current standard.
     2.1.2.1  Current Industry Description.   Data are
available for wall-fired, tangentially-fired, stoker-fired,
and fluidized bed combustion :F3C  boilers subject to the
current standard.  In 1995, there were 38 known  utility owned
wall- or tangentially-fired boilers subject  to the current
subpart Da standard,  and they are listed in  table 2-1.   There
are no known oil- or natural gas-fired utility boilers subject
to the current subpart Da standard; all 38 boilers burn coal .
Table 2-2 presents the number of boilers and share of
generating capacity for four size ranges. The majority of
boilers  (76 percent)  and generating capacity (92 percent) are
for boilers greater than 300 MW in size.
     Of the 38 boilers, 25 are wall-fired and 13 are
tangentially-fired.  All 38 boilers have applied some type of
low NOX burner technology to achieve their emission limits.
Of the 25 wall-fired boilers, 21 have low NOX burners  (LNB)
and 4 have LNB plus overfire air (OFA) installed on them.  Of
                              2-2

-------
(0
Q
0
XlH
z§
fs
&• £
IH
01 0>
•H 0.
—i >
O 4J
CO
4J 0) —
•S55
ID O ^*
e O
01
4J
n
w


w
C
03
C

4J
C
(D
*"*
u<












fll
w
(0


^
*j
,
•H

J:











m

^




in
vD

"
5





to
i-H
—1
• P^
^r*

4J
0)
i-H
0
D








U
•*H
IH
4-1
U
01
W

10
c
10
**4
o
1-1
3
0
^

iH
10
IH
4-1
C
8
CO

u


10

CO

n
8











o

10
w
o

c
0
-H
4J
10
•"4
u
0
CO
CO
(^

u
IH
4J
U
0)
i-H
U

0)
D
1
0
t3
10
k
0
i-H
0
u
CO

co


*

o

-
g









10
N
c

c
0
CQ

•
o
o
0
0

•
CO
c
10
Vj
fcn
12

C
o

4J
n
IH
01
01
o

4J
0)
kl
01
CO
01
o
CO
z

u


*

in
t-
ui

CN
O












«c
Q
K
U









>•
4J
iH.
IH
O
4J
3


p
flj
O

U
01
^
•f^
PC

•O
C
10
IH
U
CO

W




o
o
o>

«
z







01
c
\^
3
XI
IH
01
£
W






J>.
C
10
a


U

O
5
Q
IX

0)
4J
10
4J
10

C
kl
01

4-1

0
z
CO
2
M
CO


*

00
CN

CN
£





^,
E

10


£
4J

0
z






^,
c
H







C
0
CO
1— 1
**-!
JJ

m
a








^
c
10
o
E
s
W
u

^i
4J
U
0)
i-H.
H
^
01
>

te.

(j
•H
a
CD
z

CO


«

0
o
n
^"H

-
E
O










\^
0)

g
^H
t4









U

^
4-1
U
ai
w
iH
U

^)

CD
(0
u

4J
n
c
c
— 1
u
c
u
CO
z

CO


g

in
vO

n
M
X









CO
E

10

m
^>

0)
^
0
0,

tj)

4J
£
cr

^J

0

•o
IH
10
0
CQ

0)

10
S

•o
c
ffl
kl
o
CD
Z

ffi


s

CN
t-l
vo

r-l
U.

u
0)

— 1
a

CO
c
£
O
^

•
4J
W



^
4J

Vj
o
£

3


U
l-l
4J
U
01
•-H
U

01

-4
^
c
0
CO
^
u
10
CQ
Z

CO


£

CN
VO

CN
Eb


01

*«H
K

O
C
£
0
^

•
4-1
CO



^,
4J

kl
o
£

3


U

^j
U
01
•-I
u

01
l-l
1— 1
•*4
^
c
0
D
^
U
ID
CO

CO


tO

0
o
a>

i-i
s-



c
.,4
(0
4-1
c
3
O

kl
01

c
M








a.

4

JJ;

o

•
4J
a
0)
Q

n
01
•->
0)
u

(^

CO
0
J
m
z

CO


10

o
o

CN
e-
D



C
.^
n
4J
C
3
O

iH
01
4J
C
M








(X

i3

s

o

•
4J
a

Q

a
Oi

0
^
2
"C

a
a
o
5
CO


£

in
in

r-i
M
X




•
^4
I"}

4J

0
U

•o
c
H
U
C
01
o


kl
0)

0
ft.

i-H
10
4J
c
0)
CJ

£
4J
3
0
to

c
a
o
1-1
£
U
— 1
Z
CO
5
n
CO


' £

t-l
o

-
£




kl
01
c
•o
IH
tfl
O

•a

0)
K














C
10
o
E1
o
CJ

IH
01
0
CU

a
•a
IB

0)
CO
5
9^
CO


«

CN

r-i
Z









£

-H
1 1
C

fci





CO
10


^1

u
— 1
kl
u
0)
l-l
u

0)
4J
a
4J
to

k^
14
O


^
ft)
CQ
D
8
CQ
D
W
O
m
u
a
H
Ex
 I
X
•J
EH

U
D
 I
CM

W
J
CQ
                                          2-3

-------
•o
0)
•o
 o

 o
u
 (0
a
CQ
D
CO


8
co
D
w

w
en
O
CQ
 I

>

.J

J

<
H

H


U

U
D
•z.
 I
CN


U
J
CQ
o
XH
O 4J
Z C
O
u
3
Ll
.28
-r4 J>.
0 4J
03
4J 0) —
•H M S
C__i C«
"" «U
D ID «-
4J .

0)
4->
ID
co

Q)
E
ID
C

+^
u
TO
T*
CM









0)
C
C
^.

*\

•*-(

O











CQ
5


03

r-l
«


vD


iH
,




c
0
4J
c
ID
4J
CO

u






c
0
-H
CO
co
E
Q
cj

>
4J
•H
•H
•H
4J
D

0
•D
C
ID
Ll
0
B+OFA
Z

00

.H
.H
|0
^

o
00
VO

tH
,






0)

0
c

E
0>









o
8
u

u
•H
Li
4J
u
0)

(jQ

0)
•H
O
c
— 1
Si
CO
B+OFA
Z

03

tH
tH
ID


O
00
vfl

CN
fi






01
•H
0
c
•H
£
CO









D
0
O
U

U
•H
Li
4J
U
0)
*— 4
H

0>
i-H
0
c
§
W
0
z


03

iH
i-l
ID


in
vo
CN

CN
Z
IH






C
^
Q
Li
CQ

CQ


U
Li
4J
U
0)
tH
U

n
ID
O

ID
c
ID
•H
•o
C
M

C
Ll
0)
4J
0
CO
B+OFA
z

•J

er
id
EH

O
00

•"
X
EH





01
C
0
4-1
CO
0)
£
J







Ll
u
1
a.

ta

D
c
•H
4J
^p"
O
•H
tj

C
0
4J
O
0
X
B+OFA
Z

-

er
ID
EH

O
00

CM
X
H





01
C
0
«p
o
fl)
£
J







n
ft)
Cu

t»

CF
c
-H
4J

er

>-3

c
0
4J
CO
0
X
B+OFA
Z

-

IT
id
EH

0


O
iH
g

C
0

ID
4J
W

C
0
4J
C
ID
4-1







C
O
4J
ID
•H
U
0
CD
CD
(£

k|
01
3
O
0-

•o
0)
•H
c
D
B+OFA
Z

CO

0^
c
ID


in
vo

oo
X
EH








f
co
-H
ki
ID
O.







^
QJ
O.

i2

CT
C
•H
4J

O
— 1
J

c
0
4J
CD
3
0
X
B+OFA
Z

(0

c
10
EH

iH
in
*

ro
X







01
4J
4-1
0>

ID
In



4-1
— 1
k|
0
s.
3
Ll
0)

•H
K

0
•o
ID
kt
0

0
u

Ll
01

*J
B+OFA
Z

CO

er
a
EH

„


ro
M
X








CO
ID
^
— 1
J=
CO
01
o

ca

o
c
4J
x:
D

H-i
o

•o

ID
O
m

0)
4J
4J
0)
g
Ll
ID
X
B+OFA
Z

CO

Cn
ID
EH

O
in
CN

tH
Z








CO
c
•H
ID
O.





•
O
8
u
EH
+

O

U
— 1
l^
4-1
u
01
iH
gg

CD
C
—1
ID
•H
(X
B+OFA
Z

CO

g
ID
EH

£
CN

tH
8







0>
•D

r"
J
ID
K




>

•^
O

4-1

Ll

J
Q
a,

^
01

•H
a

0)
4J
4J
ID
o!
B+OFA
z

a

O1
a
EH

vD
vD
"^

*
s





0)
C

4J
ID
U
CD
3








4->
O
•J

4jf

^
Q)
j
0
a.

01
c
•H
4J
ID
U
CO
3
X
B+OFA
Z

03

cr
a
EH

ro
n

f-
tH
z
M




01
U-l
r*
(D
_r^
U
CO

a




01
u

Ll
Oi
u

«_f
XI
y
CM

ffl
C
ID

•o

M

4J
0
Z
B+OFA
z

03

C
10
EH

CN
n

00
z
IH




01

r*
10
£
U
CO

g




Q)
0
-H
>
01
CO
o

r-l
XI
3
a.

10
c
ID
•H
•o
c
M

s,
4J
Ll
O
Z
B+OFA
z

00

CT
C
CD
H

O
O
in

CN
O
co









CO
CO
o
o






0>
CJ
^
Ll
01
CO

U
•H
1— 1
o
3
a.

4
c

rH
0
Ll
ID
CJ

CO
B+OFA
Z

03

cr
ID
EH

in
in
^

*
£






•o
c
0)
CQ

a
CO









c
ID
O
E
O
O

u

Ll

U
01
•H
U

ID
D
i
H
(D


0


-H

E
3
4J
OQ


•D


ID
n

0
c
•H
e
                                                                                                      X)
                                                                                                      XI

                                                                                                      3
                                                                                                      D)
                                                                                                      co
 0)
4J
                                                         2-4

-------
  TABLE 2-2.
SIZE DISTRIBUTION OF KNOWN UTILITY BOILERS
        SUBJECT TO  SUBPART Daa

Size (MW)
<100
>100 and <300
>300 and < 600
>600
Totals

Number
boilers
4
5
13
38
Percent of
total
boiler
population
11
13
34
42
100
Capacity
of size
range
(MW)
224
1,244
5,637
12,224
19,329
Percent
of total
capacity
1
7-
29
_£!
100
aWall- and tangentially-fired boilers only.
                            2-5

-------
the 13 tangentially-fired boilers, 1 has LNB and 12 have LNB
plus close-coupled OFA installed on them.
     The precise population of stoker-fired and FBC boilers
subject to the current standard is not known,  however, data
are available for 12 stoker-fired and 15 FBC boilers.  All
12 stoker-fired boilers are in service at cogeneration
facilities.  Of the 12 stoker-fired boilers, 8 employ
selective noncatalytic reduction  (SNCR)  and 4 employ
combustion controls to control NOX emissions.   All 12 are
rated at 37 MW and burn bituminous coal.  Of the 15 FBC
boilers, 11 employ a circulating bed and 4 employ a bubbling
bed.  Typically, FBC boilers are smaller than wall- and
tangentially-fired boilers.  The 15 known FBC boilers range in
size from 25 to 165 MW.  Six FBC boilers are 50 MW or less,
seven are between 50 and IOC MW, and 2 are greater than 100 MW
(165 MW each).   Selective noncatalytic reduction (SNCR) is
used to control NOX emissions from 7 of the 15 FBC boilers and
8 use combustion controls.
     2.1.2.2  Future Industry Description.  Future electric
generating capacity will include steam generating units
(boilers)  and a variety of other devices and sources  -- gas
turbines,  internal combustion engines, hydroelectric, wind,
solar, etc.  The discussion that follows addresses the
projected growth of both utility and non-utility owned and
operated boilers.  The year 2000 is used to represent the
fifth year of the revised standard.  The following trends in
the projected growth of boilers are discussed:  1)  capacity
share of utilities versus NUGs,  2) capacity share of different
fuel types, and 3) boiler size.
     According to the Energy Information Administration (EIA)
Annual Energy Outlook Supplement for 1995, the projection over
a seven-year period (1994-2000)  for new electricity generating
capacity attributable to boilers is 8.1 GW.1  Of  this total,
5.7 GW or 70 percent is attributable to utility boilers and
                              2-6

-------
2.4 GW or 30 percent is attributable to NUG boilers.  These
data are presented in table 2-3.
     According to this projection, all the new boilers
projected to be built by utilities will burn coal.  No new
natural gas- or oil-fired boilers are projected.  For NUG
boilers, 96 percent of the projected capacity will come from
coal-fired boilers and 4 percent from natural gas-fired
boilers.                                                  .
     Data from two sources on the size of boilers likely to be
built were reviewed:  a 1995 EIA inventory2  and  an article
from Power Engineering magazine3.  These  two sources present
data from two different perspectives.  The EIA inventory is
based on information reported by utilities on a federally
required form and includes only utility owned boilers.  The
EIA inventory identified projects planned through the year
20C5.  The Power Engineering data are based on a survey that
targeted utility and NUG base-load boilers larger than 100 MW
in size.  Only the NUG data from the Power Engineering article
were used.  These data were updated through telephone contacts
made with boiler owners."
     Table 2-4 lists the utility- and NUG-owned boilers that
are projected to be built between 1996-2000.  This reflects
the population of boilers that will be affected during the
first 5-year period following the revision of the standard.
In table 2-4, the boilers are arranged by fuel type within
each category (utility and NUG).  As shown in the table,
17 new boilers with a total capacity of 5.2 GW are projected
to come on-line during this period.   This includes 7 utility
owned boilers with a combined capacity of 2.7 GW and 10 NUG
boilers with a combined capacity of 2.5 GW.   Boiler size
ranges between 80 and 830 MW with an average size of 400 MW.
2.2  UTILITY BOILER DESIGNS
     The basic purpose of a utility boiler is to convert the
chemical energy in a fuel into thermal energy that can be used
by a steam turbine.  To achieve this objective,  two
fundamental processes are necessary:  combustion of the fuel
                              2-7

-------
      TABLE 2-3.   PROJECTIONS OF GENERATING CAPACITY
               GROWTH FOR BOILERS COMMENCING OPERATIONS
                        BETWEEN 1993  AND 20001
                                            Generating
	Fuel type	capacity (GW)

 Utility  boilers     Coal                           5.7
 NUG  boilers         Coal                           2.3
                    Natural  Gas                    o.i
 Total                                              8.1
                            2-

-------
w
«
(71

H
•J
t — 1
m
w
m

o
Q
w
£H
u
£5
0
a;

a:
Bi
^
UH ~
CC ^
C
O ™
t— , I
f^ ^£
^ cr-
— a*
2 "
2

r" r;
^ s
' R"
1 — i r ->
M ™
E-^
D
^
Cij
^y

t.
o

H
3
CN
m



W , 0)
(C {-< C
QL) t -H
J>"1 r3


"0
*" a



ri D ^—*
m N £
1-1 CO —


Q
M
4-1
-iH
Q

CD
4-1
rt
W




n
£
(C
2

4J
-H
i — |
-•H
U
1C
u<








r Owner
Boile



ocAvovncnoo u>r-«)ooo>co
oa\cn(ricAOO a>cr\(nc>(TNO>
C^r-(iHrHrHCN(N rHr-tr-ftHrHr-l


rH
•H

JWCQrQtn.3,2 UUOUU^

o r* ^ ^ o in in
O 0^1 tN VC O U) **D
00 rO ^ *3< <*} rH H




rH m (N (N rH in VD



in
r-t
r-
r7



t~-



Z
O CN CN O O O
o rn O oo oo co
U) CO ro -H r}i




1 t rH H rH T-H
rH rH



•3-
r-
«»"
r7



o




fc§Eii ub.a-ags




c"
t".
c ^
r^ ^
Si n. w w
(DO1 S CD  tiCDGC I'CUtC C
TO  C U CD 1' IH 0 .Q £
-H Vi 0 -c Cd U C CD CD 4J E CD
S&irHOJO • • -fHJ/CrH^i;
tHtnUCOQOO CXOW
U CD 0 til IH in
CD 3 ^ U -H 0 0
rH o ^ ft a
K a u w n ^
-H CD ^ 0 0
UJ W U V4 -H U U
i-l CU -H 4-> 4-i IH
CU -H U CJ -H CD ^ in
rH4J4JDrH?CDCD
•H - D -H 0 > (0 (C
4J C C TJ in T3 T3
-H U] 0 'H C (C -H -rH
DHH>oS££





4J
•H
rH
-H
D
rH
(0
0
H


u
c
M W
i — 1 CO
ITS Q)
CD U U
rH -H -iH
(C E >
>H (1) ' r<
Q] £ 0 D U
C • U U 0) C
•H U M
?. G * G >,
M o tn -
rO T3 tn -H >H I-H
W D - 4J 4J CD «
VH 4J S U (0 C 4J
D rc tn 3 IH W --H
rH T3 l-l "O D CU
•H -H D 0 C T3 (C
0 rH C rH CD CD U
CQ 0 til CX O -H
o c u in c/i a, cn
p o a -H s a oi
a u s < u < u





O
rH
(C
0
H










cr\
00
r-l
in



^





























!
(Utili
rH
(C
4J
0
(C
3
 CD 5
3 .Q 0 D
C 4J
C T3 A; -H
O D C C
2 N 3 &l
II TJ CD r-t
-H 4J
u 3 4
2-9

-------
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 nitrogen oxides (NOX)  formation,  are
discussed in section 2.4.  The objective of this section is to
provide background information on the basic physical
components found in utility boilers and how they work together
to produce steam.                                         ,
2.2.1  Fundamentals of Boiler Design and Operation
     A utility boiler consists of several major subassemblies
as shown in figure 2-1.  These subassemblies include the fuel
preparation system, air supply system, burners, the furnace,
and the convective heat transfer system.   The fuel preparation
system, air supply, and burners 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
the superheated steam required to operate the steam turbine
and produce electricity.
     The NOX formation potential of a boiler is determined by
the design and operation of the fuel preparation equipment,
air supply, burner, and furnace subassemblies.  The potential
for reducing NOX after it forms is primarily 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 in the burners and furnace.  Oxygen is typically
supplied in two, and sometimes 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
                             2-10

-------
     Superheaters and Reheaters
                                                   Rue Gas
                                                Fuel
Figure 2-1.   Simplified boiler  schematic
                   2-11

-------
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.
     Utility boiler furnace walls are formed by 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 540 °C  (1,000 °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
1,1C: to 1,3CO °C (2,000 to 2,40C °F; , depending or. fuel type
and boiler design.   At this point,  in the third key process,
the gases enter the convective pass of the boiler, and the
balance of the energy 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, 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 significant amount of radiant
energy, whereas a flame produced from burning gas is
relatively transparent and produces less radiant flux.  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,
                             2-12

-------
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 of these options--SNCR and selective
catalytic reduction (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 utility boilers.  For purposes of presentation, these
configurations have been divided into four groups:  wall-
fired,  tangentially-fired,  FBC,  and stokers.  Wall-fired
boilers are further subdivided based on the design and
location of the burners.
     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.  In contrast
to tangentially-fired boilers that produce a single flame
envelope,  cr fireball, each of the burners in a wall-fired
boiler has a relatively distinct flame zone.  Depending on the
design ana 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 2-2 shows the
burner arrangement of a typical single,  wall-fired boiler.5
     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
                             2-13

-------
   BurnerB
   Burner A
AirA-
AirB-
AirC-
AirD-
FuelA
FuelB
FueIC
FuelD
  Burner D
  Burner C
  Figure 2-2.  Single, wall-fired boiler.5
                    2-14

-------
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 firing is shown in figure 2-3.6  To  burn
fuel oil at the high rates demanded in a modern boiler,
circular burners must be equipped with oil 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.  Atomizers that
use fuel pressure are generally referred to as uniflow or
return flow mechanical atomizers.  Steam- and air-type
atomizers provide efficient atomization over a wide load
range,  and are the most commonly used.
     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
     Unlike tangentially-fired boiler designs, the burners in
wall-fired boilers do not tilt.  Superheated steam
temperatures are instead controlled by excess air levels, heat
input,  flue gas recirculation, and/or steam attemperation
(water spray).   In general,  wall-fired boilers do not
incorporate the twin-furnace design.
     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 as shown in figure 2-4.
     2.2.2.2  Tangentially-Fired.   The tangentially-fired
boiler is based on the concept of a single flame zone within
the furnace.  As shown in figure 2-5, the fuel-air mixture in
a tangentially-fired boiler projects from the four corners of
the furnace along a line tangential to an imaginary cylinder
                             2-15

-------
Gas-fired lighter
       Cool
  Figure  2-3.
Circular-type  burner for pulverized
coal, oil,  or  gas.
                          2-16

-------
                           Burner Zone
Figure 2-4.  Opposed wall-fired boiler.
                   2-17

-------
                                                           •  l-l
                                                             0)
                                                            •O
                                                             0)
                                                             U-l
                                                             I
                                                             r-l
                                                             (0
                                                             •H
                                                             -P
                                                             c
                                                             0)
                                                             D>
                                                             c
                                                             (0
                                                             a

                                                             c
                                                             •H
                                                             -p
                                                             (0
                                                             a
                                                             in
                                                              i
                                                             (N

                                                              0)
2-18

-------
located along the furnace centerline.7  As shown in
figure 2-6, the burners in this furnace design are in a
stacked assembly that includes the windbox, primary fuel
supply nozzles, and secondary air supply nozzles.7
     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.   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 uc and down within the furnace in order tc control the
furnace exit-gas temperature and provide stearr. 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
their original position.  As the fouling of the tubes resumes,
the tilting cycle repeats itself.
                             2-19

-------
                                             D

                                            •i-i
                                            b-
2-20

-------
     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 2-6.
     Approximately 10 percent of the tangentially-fired
boilers are twin-furnace design.  These boilers, which are
generally larger than 400 megawatts (MW),  include separate
identical furnace and convective pass components physically
joined side by side in a single unit.   The flue gas streams
from each furnace remain separate until joined at the stack.
     2.2.2.3  Fluidized Bed Combustion.  Fluidized bed
combustion is an integrated technology for reducing sulfur
dioxide (S02) 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 windbcx
located directly below the combustion floor.  This fluidizea
state provides a large amount of surface contact between the
air and solid particles,  which promotes uniform  and efficient
combustion at lower furnace temperatures between 86C and 900
°C (1,580 and 1,650 °F) compared to 1,370 and 1,540 °C (2,500
and 2,800 °F) for conventional coal-fired boilers.  Furnace
components include fluidizing air nozzles, fuel-feed ports,
secondary air ports,  and waterwalls lined at the bottom with
refractory.  Once the hot gases leave the combustion chamber,
they pass through the convective sections of the boiler which
are similar or identical to components used in conventional
boilers.  Fluidized bed combustion boilers are capable of
burning low grade fuels and sizes range between 25 and 200 MW,
                             2-21

-------
     Fluidized bed combustion technologies based on operation
at atmospheric and pressurized conditions have been developed.
The atmospheric FBC (AFBC)  system shown in figure 2-7 is
similar to a conventional utility boiler in that the furnace
operates at or near atmospheric pressure and depends upon heat
transfer of a working fluid (i.e.,  water)  to recover the heat
released during combustion.8   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.
     2.2.2.3.1  Atmospheric fluidized bed combustion.  There
are two major categories of AFBC boilers:   the bubbling bed,
and the circulating bed designs.  In the bubbling bed design,
coal and limestone are continuously fed into the boiler from
over or under the bed.  The bed materials, consisting of
unreacted, calcined, and sulfated limestone, coal, and ash,
are suspended by the combustion air blowing upwards through
the fluidizing air nozzles.  The desired depth of the
fluidized-bed is maintained by draining material from, the bed.
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 freeboard 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
                              2-22

-------
          Convection
            Paaa

Coal  Llmaatona
             Fraaboard «
            Splaah Zona

                  Bad
                                                          Flu* Gat
                                                 I       Cyclona
                                                 \s
 Tranaport Air
                         Fluldlzlng Air
                                            Racycla
                                                    "• Dlatributor
                                                        Plata
                                                      Planum
    Foread Draft Air
                                      Watta
Waata
       [Compraaaorj
  Figure  2-7.   Simplified AFBC  process  flow diagram.'
                            2-23

-------
0.01 inches.  The bubbling bed also incorporates relatively
low air velocities through the unit,  ranging from 4 to
12 feet per second (ft/sec).8  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.9  As  a  result,  a physically well-defined  bed is  not
formed; instead, solid particles (coal,  limestone,  ash,
sulfated limestone, etc.)  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.8
     Circulating AFBC's are dominating new FBC installations,
in part due to their improved performance and enhanced fuel
flexibility.10  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 S02 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
     •    Reduced solid waste generation rates, because of
          lower limestone requirements.
                             2-24

-------
     2.2.2.3.2  Pressurized fluidized bed combustion.
Pressurized FBC is similar to AFBC with the exception that
combustion occurs under pressure.  By operating at pressure,
it is possible to reduce the size of the combustion chamber
and to develop a combined-cycle or turbocharged boiler capable
of operation at higher efficiencies than atmospheric systems.
The turbocharged boiler approach recovers most of the heat
from the boiler through a conventional steam cycle, leaving
only sufficient energy in the gas to drive a gas turbine to
pressurize the combustion air.  The combined cycle system
extracts most of the system's energy through a gas turbine
followed by a heat recovery steam generator and steam turbine.
     2.2.2.4  Stoker-Fired.  There are several types of
stoker-fired boilers used by utilities.  The most common
sroker type is the spreader stoker.  Spreader stokers are
designed to feed solid fuel onto a grate within the furnace
and remove the ash residue.
     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-S.-i  The thin bed of fuel en the grate
is fuel-burning and responsive to variations in load.
However,  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.  Because of material
handling limitations, the largest stokers used by utilities
are roughly 50 MW or less.
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).
                             2-25

-------
                                                              0}
                                                             JJ
                                                              1C
                                                              ui
                                                              Cn

                                                              0)
                                                              O>
                                                              h
                                                              ro
                                                            ' jC
                                                              U
                                                              (0
                                                             -H
                                                             73

                                                             £
                                                              (0
                                                              c
                                                             •H
                                                             -P
                                                              C
                                                              o
                                                              u
                                                              0)
                                                              o
                                                              15

                                                              T3
                                                              0)
                                                               I
                                                              1-1
                                                              Q)
                                                              ^
                                                              O
                                                              4J
                                                              W


                                                              0)
                                                              a
                                                              CL)
                                                              T3
                                                              ro
                                                              o
                                                              u
                                                              a
                                                              w
                                                              CO
                                                               I
                                                              CM


                                                              (1)
                                                              ^


                                                              CT
                                                              •H
2-26

-------
     2.2.3.1  Pulverizers.  The majority of coal-fired boilers
burn pulverized coal.  The only fuel preparation system
discussed here is the pulverizer.  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.12
     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
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
toiler loads make them well-suited for utility boiler
applications and they comprise the largest number of
medium-speed pulverizers 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.13
                             2-27

-------
     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).14
     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
the pulverizers to 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 fans determine 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-draft  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.
                             2-21

-------
     Primary-air fans are located before or after the fuel
preparation systems, and provide primary air to the burners.
In pulverized coal boilers,  primary air fans are used to
supply air to the pulverizers and then to transport the
coal/air mixture to the burners.  There are two types of
primary air fans:  mill exhauster fans and cold air fans.   A
mill exhauster fan is located between the pulverizer and the •
windbox and pulls preheated combustion air from the secon4ary
air supply duct through the pulverizers.  Cold air fans are
located before the pulverizers and provide ambient air to the
pulverizers through a separate ducting system.  Primary air
fans are used in all boilers.
     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
tc 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.
     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.
     The second part of the air supply system is the windbox.
A 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
                             2-29

-------
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.   To produce electricity,
a steam turbine converts thermal energy (superheated steam)  •
into 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.1?
     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
                             2-30

-------
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.16
     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, ir. some cases, may completely bridge the space
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.16  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.
                             2-31

-------
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.  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) .1?
     FOSSIL FUEL
     This section
fuels used for electric power generation:  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) classified
coals by rank, according to fixed carbon and volatile matter
content,  or heating (calorific) value.  Calorific value is
calculated on a moist, mineral-matter-free basis and shown in
table 2-5.18  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
                             2-32

-------















cc
^H
1

^
CD

~>

<
C

r.
O
J2
o
M
£-^
^
O
!I
. — ,
CC
cr.
^
^



un

rxi

^
ffi
^
E-1










P»


*•>.
to
X) '^
^ to
*•* ^

-* re
QJ "^

~* "*«
to ™
> fc
£ Z
o *jf
"(0 "~
"§

C
• •- L.
(0 *-•
l_ U
§u
(C
O) U
Ol
c
JC


(/)
tf)
*J
c
(O
O 1-
fD O
cr ro
UJ CJ
o

.1 r


* CD
^ O
*~ O
£ £
— ' L-
L. °
0 £
c ^

0 "^
' QJ
Z C
(C
.c

s
— 1


c

0 1-
	 , [_
C CJ
_: zii g" 'c
> >1 5
o
^
^
- c
^ cc £
- 0
y c
*-• u
i ^
—i O
If

OJ
— '


0 C
•£ *! C
CJ C
"c c
X —
•~ s:
u_
>
L.
o
o £
	 , [_
(C CJ
UJ CJ
o
tii



c
0
(D
V
«
"

Ol
c
"^"
ro
i_


Ol
w
ro
|




1





: : :






f\J CO vT







i r\j co







i CO f\J
0 0







CO (Nj ^£
O O- CO





GJ
4^ CJ
U • —
u ra u
••* u QJ ro
u c 'o *£
«] ro ro c
U i l_ ffi
X ro .c —
*rf *j *-• £
C CJ C C

- ^ r^ m

Ol Ok
c c
"~ '^
ffl (0
X l_ U
5 w o>
O O) O>
U (0 «>


, , i O O O
, , • o o o
O O m

^ r^ T^



. i O O O 0
, . o o o o
o o i/> in
^i to •— o






oo •— > * '
r\j Ki • ' '







^* r\j *— '
r- rg N^






 '
S- sO ' • >




(D ffl ro
— ' CJ — O O O
ro — o — •— — ' •— — - —
.? o.£ E.H §§ §s §§
S ii ll is is is
>p- O •— OJ •— -t- •— ••-
pa — i xi xx> i < z to zu

•• •— r\j Kl »j- in
Ol
c
) .
ro
^
e
O
Ol
Ol
ro
c.
o
c


o o o
O O 0
in in in

«- o o-



o o o
o o o
in in Ki
o O- co
«—













i > >
i i i















,
i i i







333
a- O o o
| C C C
E . >— (^ '^ re •"• rt
r j^ o xi o jD o
,Z XI 0 XI u XI U
.5 a 3 ^
TJ 
































1 '








< m
u
• •- tu Oj
ii & ii
»j _i ^

— «— (V


























































QJ
3
U
1
0
•2.
}l
2-33

-------
its metamorphism by pressure and heat, thus raising the
fraction of fixed carbon.  These values are not suitable for
ranking low rank coals.  Lower ranking (younger) coals are
classified by calorific value and caking (agglomerating)
properties which vary little for high rank coals but
appreciably and systematically for low rank coals.
     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:  (I) water or 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.  Ultimate analyses may be giver, or. 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 2-6 presents
sources and analyses of various ranks of as-received
coals.1""0  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
                             2-34

-------
J
<
o
u

fc
o

CO
«
a;

CO
D
O
H
(X

S

[14
O

00
w
CO
>-
J
<
S
<
<
u
M
O4
>^
E-
w
u
ex
t—s
I—'
o
CO
W

J
CQ
.0
1"
0 1
1

«
1
o
|
i
1

S
^ 1




> C
t
1
££

I
~



§
"
c
5
o;





u



o

1C

^
(N
CC

0)

n
ex
X
-


CN



"~





-c

C
1
s




j
c
t)
"•

0
(N

O

n
09


5
— *

o
s



«
o
t»

CD

X


-


c-




1


















o



„
CO
CO

o

c

to
-


«
^




1
?.
X
i-
c

1
£
Q.
I
I
tr


"c
p
1
"~

O
CB
ID



to
CO
CO

(N

^


LC


CC
fs.



i


















o
s



*
B

k£

ta
r*j
tC
C


X







=
£
£


A
1
snouiuj

3
<
^:
«
o
>
^:
o
—

*



^
GO

^

CO


LC


a




2


















o
s



o
CD

n

CE
CC
c
n


IE








C
1
O
£
1
£
§
F
3
i
CC
J-
"c
»
--

o
CO



rv
CD
O
CO

1C

«


0


LC
13



I


















G
CC



o
»

CO

CC
(C
c-
1C


IT



"




Lt
c

c
E
c
(C
si
1
F

£,
o
u
n
O
>
i
—

o
<*>



M
CD
CO

(O

c


«?


(C




5


















o
8



CN
O
to

o

„
tc
«*!
-


n



*~




f-
C
J
=!
to
1


<
S
o
1
^e
u>

o
03
n



1C
o
CD

CN

CC


n


i-




5


















o
ts



0
a
ifi

CT)

*


-


~



"^



u
rt
C

i
w
1


CC

r
E
-S
u.

O
CO
o
^


n
n

o

CO


X


LC




1


















i
1C




1C

N

Pi
cc
CC
-


c







_
r
X
c



c


o
2
£
1
£
"•

i
LC
*~


-
-

0

10


«


^




lAJtAJf)


















o
§



CD
CD
o

a

C1
c-

Lf


-^



'



I
1-
-

.
1
§





4.
c
o
—

0
PS
*~


-
en
o

0

CC


1C


X




DMM

















                                         2-35

-------
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.21
Because of its low volatile matter, anthracite is difficult to
ignite and is not commonly burned in utility boilers.
Specific characteristics of anthracitic coals are shown in
tables 2-5 and 2-6.  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.21  Specific characteristics of bituminous  coals are
shown in tables 2-5 and 2-6.  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
relative volatile matter and sulfur content,  and its slagging
and agglomerating characteristics.  As a general rule,  low-
volatile-matter and low-sulfur-content 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
                             2-36

-------
agglomerate.  The nitrogen content typically ranges from 0.6
to 1.4 weight-percent.21  Specific characteristics of
subbituminous coals are shown in tables 2-5 and 2-6.  Because
of the low sulfur content in many subbituminous coals, their
use by electric utilities grew rapidly in the 1970's and
1980's when lower sulfur dioxide (862) emissions were
mandated.  Their higher moisture content and resulting lower-
heating value, however, 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 .:i  Specific characteristics of lignite
are shown in tables 2-5 and 2-6.  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 power plants located near the
producing mine.
2.3.2  Oil
     Fuel oils produced from crude oil are used as fuels in
the electric utility industry.  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.
     To provide commercial standards for petroleum refining,
specifications have been established by the ASTM for several
grades of fuel oil and are shown in table 2-7."  Fuel oils
are graded according to specific gravity and viscosity,  the
lightest being No. 1 and the heaviest No. 6.  Typical
                             2-37

-------
TABLE 2-7. ASTM STANDARD SPECIFICATIONS FOR FUEL OILS22
O.
^ C
n C
* S
o. t
Q. O
0 u
U
I!
o
c
V
u
fl
u
5
c
"c
-C
L-
c
V
o
£
P
§
8
(\ —
•~ P
"c ~
u.
§
n: ~
5 ^
£
s
c
c
H
> o
-c -c £
< 5
liU
0!
f-
a
f E P
a
C ""
i
c
5
X
*
c
1
c
5
a
c
J
c
1
1
1
1
X
a
1
i
C
5

n
I
1C


CN
:




c cc
IT X

C 5
r, —

iC
C
t
0
CO
n
O
c
I

o


CD
O
(N


c-
tc
r,
r:
c c
§ ?


LT
C
C
o ^
1OO (381
1




tD
(S
CD


S
LT



C
C

C
LC
C
(S j_
13O (BBI
n
c
z




LC
CO
§


c
c
c



^

c
c

t£
C
n
u, J
c ?


OD
I
CM
CD
1C
r*.
C
r-
c
LC
r-



C

g

LC
LC
O
No B
(Heavy)


CO
n
CO
a


c
0
LC
c
o
c
cr
c
r





g

199) OS I
ce
o
Z
i
£
£
CC
c
1
0
o.
c
R
o
2-31

-------
properties of the standard grades of fuel oils are given in
table 2-8.23-24
     Compared to coal, fuel oils are relatively easy to burn.
Preheating is not required for the lighter oils, and most
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.                                                    ,
     Because of the relatively low cost of No. 6 residual oil
compared with that of lighter oils, it is the most common fuel
oil burned in the electric utility industry.  Distillate oils
are also burned, but because of higher cost are generally
limited to startup operations, peaking units, or applications
where low PM and SC>2 emissions are required.
     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
ar.d the economics of oil transportation, fuel oil supplies
vsrv i_n compos—tion across tns 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 through 5 and 0.4 to 1.0 weight-percent for grade 6.:i
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 2-9.:5  Prior to
distribution, however, most of the inerts (carbon dioxide
[CC>2] and nitrogen),  sulfur compounds, and liquid petroleum
gas  (LPG) fractions are removed during purification processes.
                             2-39

-------
CJ
M
CO
OJ

W
r-H
QJ
r-'
rH
^'o

O
QJ
••^
rH
.p-j
O

O
s
rH
QJ
O
2
( — j
QJ

M 0
O
2

C'
^rH
O
s








(1)
U-1
rrt
'u












rH
rrt
'U
2
-H
CO
tf
.— 1
4_) 2
X* r2
CD _"j
*** rn
j g


jj
ery ligh
residual
>

jj
rC
•rH
-H
Q
OJ -
4-j QJ
rc d
rH QJ
•H O
J i ^_J
cn QJ
•rH Jv1
Q —




















Q)
a
Jx,
EH

J^d
u
rC
H
OQ


y
U
fC
rH
ffl



Black



SH
QJ
i




.d
Cn
•H




















}H
O
rH
O
U



OJ
H






t~-
H



rH
OJ




m





0







fj_i
0

0
U3

-
^S,
Jj
•rH
£>
rC
SH
Cn

H
a
<
H
oo
cn
•
o


O4
tn
Cn
.
o



cn
cn
0


tn
00
o



tn


-
^
4J
•H
J>
rc
SH
Cn

U
-H
<4H
•H
U
Q)
a
CO
OJ
rH
OJ
•
00


tn
ro
cn
t
[^




OJ


<^D
0
""



C
oo
vo




El«<
O

o
^D

^
C
O
rH
rH
fC
Cn

.
CO
.
D
^^
Q
rH

O
•
0
UD
r^l



0
•
O
tn



o
in
rH



00
OJ




u
H



tn
Q)
_y;
0
^j
cn
•H
JJ
d
Q)
U

»
^,
4J fr.
•HO
CO
O o
U 0
CO rH
•rH
> ©



tn






o
m



0
rH

o

QJ
N
O
QJ
m
0
L,
a
M
O
rH
QJ
CQ










Cij
0

~
Jj
d
-rH
0
a,

SH
rj
O




0
o
H

?
3
tn E
"J -H
d
-H
3


15
(minimum)

u
•rH
QJ
a
tn
o
E
<
L-

QJ
"a
cn
O
E
<



£H
O

^
Cn
d
•H
a
E

Q-

J-l
o
14H

.
DJ
E
QJ
E-



O
o
OJ





o
ro
rH


25
(minimum)

u
•H
QJ
cn
o
E
<
V
"l"
QJ
tn
o
E
<

tL,
0

»
Cn
d
•H
N
•H
E
O
4J
(C

rH
O


.
a
E
QJ
EH

O
•
OJ
rH




O
.
tn




tn
OJ



QJ
u
rC
SH




QJ
U
SH
EH








0\°

^
QJ

T3
•H
W
Q)
SH

d
o
£)
^_i
rC
U
E
3
oo E
• -H
oj X

-
?
3
o E
• -H
oj X
(C
J;


tn
H
o

t-
o

o




rH
O
















<$p

..
in
2
14H
rH
3
CO
tn
•
0
1
H
•
O
in
•
o
i
H
.
O


tn
o
rH
O



rH
0
o
V




rH
0
O
V














o\°

-
d
QJ
CD
O
SH
4_J
•rH
2

in
•
o
rH




f-
.
rH
rH



cn
H
rH



r-
rH




OJ
rH














o\°

-
d
QJ
Cn
O
SH
rrj
JXi
E
0
[-^
.
tn
oo



in
in
.
in
oo



o
rH
VD
CO



«•
OO




in
00
















<&°

.
d
o
_Q
^_j
rC
CJ
E
3
o £
• -H
oj X
rC
E
"e

o E
• -H
H X
rC
E


0.5
(maximum)



Q)
U
rC




0
U
rC
rH




0\°

-
SH
QJ
4-1
(C
J

TD
d
rC

U
d
QJ
E
•H
T3
QJ
CO
00
o

o



tn
o
.
o




OJ
o
0



QJ
U
rC
SH




QJ
CJ
rC
SH



















o\°

-
fj
cn
<
o
o
0
o
tn
rH
o
o
o
t
oo
^*
H
t

O
O
o
rH



O
0
0
rH
rH



C
0
o
r-
rH















d
O
rH
rH
rC
Cn

rj
4J
m
                                        2-40

-------


co n5
< v
0 K
J
^> o oo
00
I— lOi— lOrHr-tOr— IrH
-H O -H O -r-l -H O -H -H
2 -2 -22 -22
o ui in
Oi



CNrOLDr-ILnCNOLnOO
OOroCN-H^CN^rnf-l

HrOO OOr^OO
o>




i— lOr— IO— lOOr-Hi— 1
•H O -H 00 -rH C^ in -H -H
2 -2 -2 • -22
<* ^ 00
CO H




r— iOi — 1 O i — 1 i— 1 O i — 1 i — 1
•H <3< --J CO -H -i- CO -H -M
2 -2 -22 -22
CO r-i







ll
CD T3
13  c o a
O -H
>i C 0) E T3 C C
£1 QJ 0) C 0) 0
mccDoicatnct
o\» O fO i-H C O O O CD C
V4 XI >^rOXJ£! W tn V
D - 'a-ux:x:^rMjj>1T
j m >,a)j-)jJrCfO-iHXh
H aJ KSWWUU2O^
J c
n a)
H 3
M jJ
D CO -H
U) CD aJ
U CO CO i C ^ffi ffi CN
r-l rH O CNW CN CNO O CN CN
rC rOU KUUUUU2O
5 §
I-H in ^t o 1-1 o
-H ao oo o-i «t m
2 • • • • ^
O ^ CN H
CN ^D T-H O

iH 00 vn UD rH O
-H U) CN O -H O
2 • • • 2 ^>
CN cr\ ao
CN U) O


«* O CN VO 00 t^
m CN rH r- in u>

O ro C^ in H
CN >-O C




rH O CN U3 (N ^1
-rJ p(~| ^K ^v. J~O pi-
& .... vi
rO ^ O H
cxi r*^ r
^ N L ^-





rH rn-i 1X1 CN rH V.
•rH tn CN CN -iH r
2 • • • 2 ^
C*~ JT i — '
CN r-












1
}
: c c
J 0) (D
T) i-i tn c tn c
3 JJ 3 O O O 0)
-i 5 UH v-i JD u tn
3 rH T3 i-l -U >-l
x, >, p >, m --H X
C £) CO K U 2 O
0\°

rt i
vl/
JJ
(TS
xi e
CN -H CN CN CN
ffi JJ CO ffi U 2 O
rH
D
** 0
r- VD
a\ H
0
CN
CN «*
O CN
O CD
H rH

3 rH CN
CN




D cn o
•n CN t~-
r r-i H
-" r-l" ^
CN




tn
tn
. — .
rH
. J f-H
1^ >-<
rti -H

O 0
aJ rn

rH US
0)
rH [J-,
— o
0)
>^ P 0 r-
jj r— i VD a
•H ro ;
> > ® ^
rd
>H JJ JJ 14.
tn its iw c
0)
u x: p j
•-H U r
*w in ~^^^
-H CD P
U X! J-) -
0) Cn CQ (
D-l -H
CO K
















tn
c
-H
CO
CO
CD
c
Q
1
rH
Cu
O
JJ
rH
o
-H
>_l
&

^
Tl
w
OJ
jJ

0)
rH
r-H
O
5 v
1 CO
' ?
D m
m

2 CO
-1 J
^ (D
3 ^
5 &
5 1
CO
fC
2-41

-------
As a result, natural gas supplies burned by utilities are
generally in excess of 90 percent methane, with nitrogen
contents typically less than 0.4 percent .26>27'28
2.4  UTILITY BOILER EMISSIONS
     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 (3l "fuel" NOX,  which is  formed frorr
chemical reactions involving nitrogen atorr.s chemically bound
within the fuel.
     2.4.1.1  Thermal Nitrogen Cxide 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)
     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
     aN20  is  not  considered  a  component  of  NOX for regulatory
purposes under the Clean Air Act.
                              2-42

-------
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-R/T  [N2]  [02]1/2 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.  If temperature, oxygen concentrations, or
nitrogen concentrations can be reduced quickly after
combustion, thermal NOX formation is suppressed or  "quenched".
     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
terroerature peaks at equivalence ratios near one as shown in
figure 2 - 9 . :Q  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 flame' may  exist  in  a wide range  of
equivalence ratios, and thus premixed flames have a wide range
of peak temperatures.  However,  a non-premixed flamed will
bEquivalence  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.
dA non-premixed  flame  exists where the  reactants must diffuse
 into each other during chemical reaction.
                              2-43

-------
I
1

I


I
0)

JS
LL
                            1.0
             Fuel-lean
                   Fuel-rich
                         = (F/A)/(F/A)
                                    Stoich
    Figure 2-9.
Variation of flame temperature with
                    29
 equivalence  ratio.
                           2-44

-------
generally react near an equivalence ratio of one, causing high
peak temperatures.
     For utility boilers, the temperature is also related to
the heat release per unit of burner zone volume.  Units with
large heat release rates per unit volume, may 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 quench prompt NOX formation in the
manner by which thermal NOX formation is quenched.  However,
the contribution of prompt NOX to the total NOX emissions of a
system is rarely large.30
     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
hydrocren cyanide  !HC1C'  or CN rs.c.ica.ls  and that the presence
of hydrocarbon species is essential for the reactions to take
place.31  The following reactions are the most likely
initiating steps for prompt NOX:3:
               CH  +  N2 ^ KCN  +  N                      (2-5)
               CH2 +  N2 ^ HCN  +  NK                     (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 .33  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
conditions,  an increasing amount of HCN is formed, creating
more NOX.  However,  above an equivalence ratio of
                             2-45

-------
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) ,34  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
suppressed because the oxygen concentration is lower relative
to fuel-lean conditions.  This promotes the formation of N2
                             2-46

-------
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 2-10 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.35  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 2-11.3b  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 corbusticr. 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.  In utility boilers, 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 utility
boilers.  These include tangential, wall, and FBC designs.
Background information on each of these boiler designs was
presented earlier.  Each configuration has design
                             2-47

-------
S2
1
o
O
O)
2
0)
D
16

14

12

10

 8

 6
                6     10    14     18    22    26
                  Ratio of Coal Oxygen to Coal Nitrogen
                                                  30
32
     Figure 2-10.
               Fuel-bound nitrogen-t.o-nitrp.cjen oxide  in
               pulverized-coal combustion.
                               2-48

-------
    1400
    1200
         - Lignite
a
a.
O

"55
LL
    1000
800
     600
     400
                     Subbit.
                     CCoal
                               High-Vol.
                                BitC
                                High-Vol.
                                  BitB
                                High-Vol.
                                  Bit A
            _J	i	  __ i	1	I	I	I
                                                    Med.-Vol.
                                                       Bit
                                                   Low-Vol.
                                                 Bituminous
           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 2-11.
             Fuel NOX to  fuej.6 nitrogen  content-pulverized
             coal,  premixed.
                                 2-49

-------
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
                             2-50

-------
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
S02 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 860 to 900 °C (1,580 to 1,650 °F)  while a conventional
boiler may operate at 1,370 to 1,540 °C (2,500 to 2,800 °F).
     Another influential factor en the NOX emissions of an FBC
boilers is the quantity of calciurr oxide,  used for SO—
emissions control,  present in the bed material.  Higher
quantities of calcium oxide result in higher base emissions of
NCX.   Therefore, as S02 removal requirements increase, base
NCX production will increase.
     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 have low gas velocities through the
boiler in order to prevent fly ash erosion and 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].
                             2-51

-------
     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 2-10),
significant contributions from the fuel nitrogen mechanism to
total nitrogen oxide occur.37   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.  Natural gas is a vapor, oil is a liquid,
and coal a solid.  The injection methods of the three types of
fuels are fundamentally different due to their different
physical states.  Boilers designed for coal have larger
furnace volumes than boilers designed for oil or gas as shown
in figure 2-12.38
     2.4.2.3  Boiler Operating Conditions.   During the normal
operation of a utility 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 NCX formation:
     •    Operating load,
     •    Excess oxygen,
     •    Burner secondary air register settings,  and
     •    Mill operation.
     All these parameters either directly or indirectly
influence the NOX emissions from utility boilers.   For the
most part, these parameters are within the control of the
boiler operator.  Sometimes they are controlled based on
individual operator preference or operating practices, and at
other times are dictated by boiler operating constraints.
     The effects of changing load on NOX emissions are varied
and complex.  With an increase in load,  furnace temperatures
     eThe nitrogen present in natural gas exists almost
exclusively as elemental nitrogen and not as organic nitrogen
compounds.
                             2-52

-------
       TABLE 2-10.  TYPICAL  FUEL NITROGEN CONTENTS
                      OF  FOSSIL FUELS37
	Fuel	Nitrogen (wt.
Natural  gas                                  0  -  0.2
Light  distillate oils (#1,  2)                0  -  0.4
Heavy  distillate oils (#3 - 5)              0.3  - 1.4
Residual oils                               0.3  - 2.2
Subbituminous  coals                        0.8  - 1.4
Bituminous coals                           l.l  - 1.7
                           2-53

-------
        On°o°n
     0 O  O
       o  o o
Wo

   °0°0°0
                               CO
                               03
                               o

                               "

                               ro
                               in
                               o
                               CC
                                  4J
                                  3
                                  •P
                                  CQ

                                  0)
                                  E
                                  (0
                               TO  "
                               O  0
                               0  X
                               §  S
                               O
                               c
m
                                      w
                                      i-4
                                      Q>
      -P
      C
      a)
      ^
      0)
                                     •H
                                     14-1


                                     U)
                                     U
                                     0)
                                     t-H
                                     -H
                                     O
                                      3

                                      1
      x:
      a

      0)

      -H
      •P
                                     (0
                                     a

                                     o
                                     u
0)
M
3
^
-H
   2-54

-------
increase and excess oxygen levels typically decrease.  While
higher furnace temperatures increase the formation of thermal
NOX, the lower excess oxygen concentrations decrease the
formation of both thermal and fuel NOX.  In wall-fired boilers
an increase in load generates higher turbulence which in turn
increases the local temperatures in the furnace.  This
increases the potential for thermal NOX formation.  Because of
these many varied effects, it is difficult to predict thet
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
ur.it - specif ic parameters.
     A frequently overlooked influence on NOX emissions for
coal units is the mill pattern usage.  Figure 2-13 illustrates
the impact of operating with various mill - out - of - service
patterns on NOX emissions.JG  These data are from a
365 megawatt (MW) single-wall,  coal-fired boiler, operating at
250 MW (68 percent load), and firing subbituminous coal.  The
NOX emission level varies by as much as 25 percent depending
upon which mills are operational.  This is because when
operating at a fixed load and with the top mill out-of-
service,  the lower mills operate at a higher coal-to-air
ratio, creating fuel-rich regions.  The secondary air from the
top mill insures complete combustion.  If the bottom mill is
out-of-service, the advantages of stratified combustion using
overfire air to insure complete combustion are reduced,
resulting in increased NOX formation.  Biasing fuel to the
                             2-55

-------
                                    u
                                    O
                                    O
I
I:
• •
••
•
r
\^j
O
O
•

•
•
•

•
•
   V V
-*• •
  I • •
  !••
                                               0)
                                               I
                                               8
                                               Q)
                                               CO
                                               O
                                 -^1* * *
                                 ^'* • •
            x—-,
            v_/
            O
            O
                                    < CD O D
§
                                                        0
                         0)
                         .u
                         iJ
                         (0
                         •H
                         E

                         U-l
                         O
                         u
                         0)
                                                        w
                                                         I
                                                        CN

                                                        
-------
lower mills can also be used to create a similar combustion
environment.
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 2-11 presents NOX emission rates for each fuel type
as required by subpart Da.  However, as discussed in
section 2.5.2, individual States may set limits on a case-by-
case basis that are more stringent than the levels shown in
table 2-11.  Consequently, to estimate baseline levels for the
year 2000  (assumed 5th year for the revised standard),  owners
of the boilers projected to be built between 1996-2000
(table 2-4; and respective State agencies were contacted to
determine the NOX levels at which these boilers would be
permitted.4  Table  2-12  shows  these  baseline levels based on
the rerrit levels for each boiler.  In cases where permit
levels were r.ot available, the baseline level  was based on the
subpart Da emission rates shown in table 2-11.
     Annual emissions were estimated with the  following
equation:

     tons per year NOX = ER * MW * HR * CF * 0.00438

where:    ER        =    annual average NOX emission rate
                          (Ib/MMBtu); assumed = 0.02 Ib/MMBtu
                         less than permit/baseline emission
                         limit
          MW        =    boiler size (MW)
          HR        =    heat rate (Btu/kWh);  assumed = 9,500
                         for coal and 9,000 for gas
          CF        =    capacity factor;  assumed = 0.65
          0.00438   =  conversion factor
                             2-57

-------
TABLE  2-11.   BASELINE  EMISSION  RATES  (BASED ON SUBPART Da)
                 FOR NITROGEN OXIDE EMISSIONS FROM
                     FOSSIL  FUEL-FIRED  BOILERS


                                                   NO-y  emission
Fuel type
                                                      limitsa
                                                    ,'lb/MMBtu;
                                                        0.50

                                                        0.60

                                                        0.60

                                                        0.50

                                                        b


                                                        0. 80
Solid Fuels

  Subbituminous coal

  Bituminous coal

  Anthracite coal

  Coal-derived fuels

  Any fuel containing more than 25%, by weight,
  coal refuse

  Any fuel containing more than 25%, by weight,
  lignite if the lignite is mined in North
  Dakota, or Montana, and is combusted in a slag
  tap furnace.0

  Any fuel containing more than 25%, by weight,
  lignite, not subject to the C.8C Ib/MMBtu heat
  input emission limit.

  All other fuels
Liquid Fuels

  Coal-derived fuels
  Shale oil
  All other fuels
Gaseous Fuels

  Coal-derived fuels
  All other fuels
                                                        0.6!
                                                        0.50
                                                        0.50
                                                        0.30
                                                        0.50
                                                        0.20
aBased on 30-day rolling averages.

^Exempt from NOX standard and NOX monitoring requirements.

cAny fuel containing less than 25%, by weight, lignite is
 not prorated but its percentages is added to the
 percentage of the predominant fuel.
                            2-58

-------




EH
H

t—4
H
EH


W

OJ
0
Cu
W
EH
^
^",
H

CO
DQ


h_^
12
O CO
1 — i Qi
CO W
CO J
H *H
2 0
w m
^Xrfl

2 D
K
2 Q
i i ^y
1-3 3
w
CO
<^
CO
<
t£
^
p"
L1
b--
M
[T j



*
03
	 i

O3

TABLE



Q)
£2 .
•H a
tn ~"
>s

E .^

Q) t*
a. w

Q)
c
•H
S-J
5 > e
\ 0)^
-H -°
E 	
H
QJ
tx


,

0 ^

£ K
fH




) 1
c ki ^r
1 — ' -o
a. v/-


Q
M

4J
«
^

a>
aJ
(C
w













Q)
(t)
2
rH
•H
U
*
b.



°oootNr~ ojtn r-
'^(N'tCOi-lfirO rHr-COVDCTirO
"^(N^cocNOO ^rnm^fcoro
** <^ ^ ^ h CO 00 fc •* 00 f*1 (^0 ^
2jvDrnHr-i U^SbjIi ^DuQjDjS?











UJ
to
ID £>
tn >-i in
0) V 01 S 0) CD
i-H rH >-l 03 U C S
-H rH QJ >-lWW 4J.QO
O -H CVHd)(l)^i-i(l)V4 'DQi
e > tcaiccaiaiains c
^> (B Q) C tn 3 3 'H QJ 0 fd i — ' 'H
i-*O DlV-l 0 Dt-lEn Ot-^^C r>iC V^rH
•H CO)4->,C CQ UCnOO)iT5
rHC-H>cu • • -io.a.c
-iH-HrHOlTJV-l&3KIUCd) tn
1—^ ^
tn -H 0
i- 0 C
0 .Q -H
4J g
tO C 3
>-, O U
0! -H -H
C J-> J3
a) in ^3
Oi 3 3
XI in
>, e
4J 0 II
1 f 1
•r-l LJ
rH • CO
•H *O C
3 il 0 0)
I C A-*
C T3 A; -H
0 0) C C
Z N 3 Cn
•r-l -r-l
II "O 4) rH
O 3 (C II
Z li. W rJ
(0 J2 U T3
2-59

-------
As shown in table 2-12, baseline NOX emissions in the fifth
year are estimated at 43,600 tons per year.
2.5.2  Other Regulations on the Source Category
     Multiple regulations affect this source category.  In
addition to subpart Da, the primary Federal program affecting
this source category is the New Source Review (NSR)  program.
The NSR program regulates NOX emissions from new utility
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 lowest achievable
emission rate (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 best available control
technology  (BACT).
     Under these two programs, and considering local air
quality conditions, limits more stringent than those
considered in seruing a national standard  (NSPS!  may be set by
individual states on a case-by-case basis.  For example, as
discussed in section 3.6.2.3, recently permitted coal-fired
boilers in New Jersey, Florida, and Virginia are required to
achieve NOX emission limits ranging between 0.10 and
0.25 Ib/MMBtu.
                             2-60

-------
2.6  REFERENCES
1.   Energy Information Administration.  Supplement to the
     Annual Energy Outlook 1995.  DOE/EIA-0554(95).
     Washington, D.C.  February 1995.  pp. 231-233

2.   Energy Information Administration.  Database 1995, Form
     860, Annual Generator Report.  Washington,  D.C.  1995.

3.   Smock, R.  Baseload power plants continue to expand.
     Power Engineering, April 1994.  pp. 23-34.

4.   National Impacts resulting from revision of the NSPS,
     Subpart Da for NOX.  Memorandum from Adams, R., and
     Gundappa, M.,  Radian Corporation to Eddinger, J.,
     U.S. Environmental Protection Agency.  Research Triangle
     Park, North Carolina.  October 11, 1995.

5.   Singer, J. G.  (ed).  Combustion, Fossil Power Systems,
     Third Edition.  Combustion Engineering, Inc.  1981.
     p. 13-3.

6.   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.

7.   Ref. 5, p. 13-4 ar.d 13-5.

8.   White, D. M.,  and M. Maibodi.  "Assessment  of Control
     Technologies for Reducing Emissions of S02  ana NOX from
     Existing Coal-Fired Boilers."  Prepared for the
     U. S. Environmental Protection Agency.  Air and Energy
     Engineering Research Laboratory.  EPA Report No. 600/7-
     90-018.  pp. 3-13 through 3-15.

9.   Makansi, J. and R. Schwieger.  "Fluidized-bed Boilers."
     Power.  May 1987.  pp. S-l through S-16.

10.   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.

11.   Ref. 5, pp. 13-19.

12.   Steam, Its Generation and Use.  Babcock & Wilcox.  New
     York, NY.  1975.  p. 9-1.

13.   Ref. 5, p. 12-7.

14.   Ref. 6, p. 9-11.


                             2-61

-------
15.  Ref. 12,  p. 12-8.

16.  Ref. 6,  p. 9-20.

17.  Ref. 5,  p. 5-11.

18.  Ref. 5.  p. 2-3.

19.  Ref. 6.   p. 7-4.

20.  Energy Information Administration.  Coal Data - A
     Reference.  DOE/E14-0064.   Washington,  D.C.  September,
     1978.  p. 6.

21.  Bartok,  B., Sarofim, A. F. (eds.)  Fossil Fuel Combustion,
     A Source Book.   John Wiley & Sons, Inc. 1991.  p. 239.

22.  Ref. 12,  p. 5-17.

23.  Ref. 5,  p. 2-31.

24.  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.

25.  Ref. 12,  p. 5-20.

26.  Eagle Point Cogeneration Facility West Deptford Township,
     New Jersey Compliance Test Report -  Unit B, Radian
     Corporation.  July 1992.  p.  2-20.

27.  Atlantic Electric Sherman Avenue Generating Station
     Combustion Turbine Unit 1, Vineland,  New Jersey
     Compliance Test Report. Appendix A,  Radian Corporation.
     July 1991.  pp. 2 through 5.

28.  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.

29.  Glassman, I.,  Combustion,  2nd ed., Academic Press,
     Orlando,  Florida (1987). p. 20.

30.  Ref. 21.  p. 231.

31.  Ref. 29,  pp. 330 through 337.

32.  Ref. 29,  p. 331.

33.  Ref. 29,  pp. 333 through 334.
                             2-62

-------
34.   Ref. 5,  p. 4-34.


35.   Ref. 5,  p. 4-35.


36.   Ref. 5,  p. 4-34.


37   Ref  21, pp. 230 through 231.


3B'.  Letter  ana attachments

     *g^r'  £cl^"l'5   '^'-    iscussion  of  NOX HACT.

                 „    r,*  T   L  Smith   Effects  of  Boiler Cycling
39.  Kanary,  D.  A.  and  L_  L.  ^m^n   1990 EPRI Fossil Plant

                                     ,  DC.  Dece^er 4-6,  1990.
      15 pp.
                                2-63

-------
       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 utility 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 (CO
and flue gas treatment controls.  Combustion controls reduce
KCX emissions by suppressing NOX formation during the
combustion process while flue gas treatment controls reduce
NOX emissions after formation.
     Combustion controls are the most widely used method of
~o*~t roll i*"o NCY formation ^:~- utility boilers.   All new utility
          —•   -A.                    -*                          -*
boilers are expected to use some form of CC to reduce NCX
formation.  Although flue gas treatment methods can often
achieve greater NOX control than CC,  they have not been
applied to many utility boilers in the United States.  The
types of NOX controls currently demonstrated or applicable for
new fossil- fuel- fired utility 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
utility boilers.  Section 3.2 presents the demonstrated
performance of these control techniques based on long-term,
continuous emission monitoring  (CEM)  data.  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 deal with flue gas treatment techniques
for all three fuel types (coal, oil,  and gas).   Finally,
section 3.7 describes advanced flue gas treatment technologies
                              3-1

-------
      TABLE 3-1.  NOX EMISSION CONTROL TECHNOLOGIES
                 FOR NEW FOSSIL FUEL UTILITY BOILERS
NOX control options
                             Fuel
                             applicability
Combustion Control Techniques

            Low NOX burners for
            conventional boilers

            Low NOX Burners + overfire
            air for conventional boilers

            Air staging for fluidized
            bed combustion boilers

            Flue Gas Recirculation

Flue Gas Treatment Techniques

            Selective Noncatalytic
            Reduction

            Selective Catalytic
            Reduction
Advanc
                       ean Coal
            Technologies
                  ™
            - SNOX   Process
            - S0x-N0x-Rox Box™ (SNRB)
              Process
            - NOXSO™ Process
                             Coal, Natural
                             Gas, Oil

                             Coal, Natural
                             Gas, Oil

                             Coal
                             Natural Gas, Oil
Coal, Natural
Gas, Oil

Coal, Natural
Gas, Oil

Coal
                            3-2

-------
that have demonstrated NOX emission reductions under the
Department of Energy's (DOE) clean coal technology  (CCT)
program.
3.1  COMBUSTION CONTROLS FOR COAL-FIRED UTILITY 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)
boilers.  Each of these control technologies is discussed ,in
the following sections.
3.1.1  Low NOX Burners for Conventional Boilers
     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.a 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:
     I.   Reduced oxygen  ',02'  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
     "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

-------
turbulent, high-intensity flames.  The longer, less intense
flames produced with delayed combustion LNB inhibit thermal
NOX generation because of lower flame temperatures.
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  (N2) 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 prirr.ary 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 cf 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.1
     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,
                              3-4

-------
                               H 0)
                               Q) C

                               *
                                  x
                               £g
                               0)
                               H
                                I
                               r>

                                0
                                M
                                3
                                r-l
                                a
                                 •0 5
                                 0) O
                                 c
                                 O
                                 U
                                  I
                                 n

                                  0)
                                   0)
                                   e
3-5

-------
except that the split-flame nozzle has been replaced by the
IPS nozzle, which generates a coaxial flame.
     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 of
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.:  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.3   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.3
     The Controlled Combustion Venturi™ (CCV)  burner for
wall-fired boilers is shown in figure 3-4.4   NOX control  is
achieved through the venturi coal nozzle and low swirl coal
spreader located in the center of the burner.  The venturi
                              3-6

-------
                                        Pilot     Fixed
                                       Manifold  Vanes
               Conical
               Diffuser
                                               \    /    Adiustabie
                                                          Vanes
                                                          Som Vanes
Figure  3-2.   Dual register-axial control  flow" low NOX burner/
                                  3-7

-------
                                             0)
                                             c
                                            fO
                                            O
                                            u

                                            X
                                            O
                                            2

                                            5
                                            O
                                            i
                                           O
                                           ro
                                           i
                                           0)
                                           ^

                                           Cn
3-8

-------
                                 (N
                                                            0)
                                                            CJ
                                                             x
                                                           g

                                                            O
                                                             c
                                                             CD



                                                             O
                                                            •i-t
                                                             U
                                                             w
                                                             0
                                                             u
                                                             0)
                                                             r-t
                                                             r-H

                                                             O
                                                              c
                                                              O
                                                             u
                                                              0)
                                                              W

                                                              Cn
                                                              •H
3-9

-------
nozzle concentrates the fuel and air in the center of the coal
nozzle, creating a very fuel-rich mixture.  As this mixture
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.4
     3.1.1.1.2  Tangentially-fired coal boilers.  As discussed
in chapter 2 and shown in figure 3-5a5,  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.  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.5  Each corner  has  its  own
windbcx rhat supplies primary and secondary air through the
air compartments located above ana below each fuel
compartment.
     The low NOX concentric firing technique for
tangentially-fired boilers is shown in figure 3-6a.5 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.5   In  addition,  "flame
attachment" nozzle tips that accelerate the devolitization of
the coal are used.  This configuration suppresses NOX
emissions by providing an 02-richer 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
                             3-10

-------
                                                                                                       0)
                                                                                                      f-t
                                                                                                      •H
                                                                                                       O
 -H
 (C

 •o

 flj  ft)
     U TJ
 r-l -H tt)
 o)  a
 3  >•
 14-1 aJ
                                                                                                O    r-t
                                                                                                    C r-l
                                                                                                £-H  flj
                                                                                                O    -H
                                                                                                •H  10 AJ
                                                                                                >  E  C
                                                                                                    m  a)
                                                                                                C  ft)  Ol
                                                                                                (C  U  C
                                                                                                iH AJ «J
                                                                                                Oi  to AJ
                                                                                               in
                                                                                                 I
                                                                                                0)
a>
c
c
re
o
— Q.

*







§
O

V^




-


(*
\
\



re
0
O



r<



re
o
O



u-



re
O
O


3
I8


LL
l(()^


re
o
O



< c
.g


5
a>
UJ
(C
•o
 C  ft
 (D
 e  '^
 aJ  O
 ^   ^

 Cu V
 e  c
 O  W  r-t
 U  O  -H
    u  o
                                                                                                      O)
 m  c T5
    O (U
T!     k
 C  S^ -H
 (0  O 1 V4 (0
H  m AJ
                                                                                                (IS
                                                                                               in
                                                                                                0)
                                      3-11

-------
w *
-i
                                                    3
                                                    &i
                                                    •H
                                                    Cn
                                                    c
                                                    •H
                                                    X
                                                      0)
                                                   m
 0)
 W

 Cn
-H
3-12

-------
Concentric Firing System™  (LNCFS)  and are  discussed  in
section 3.1.2.
     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.6
     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
wirh conventional burners.   Flarr.e impingement on furr.ace walls
cr superheater tubes can be eliminated or minimized by
adjusting the 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.
                             3-13

-------
             Ef
            Si
                            OJ
                            c
                            6
                            3
                            C
                            o
                            -H
                            O
                            a

                            X
                            o



                            o
                           ro
                            i-i


                            Cn
                           •H
                           tL,
3-14

-------
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
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
                                                          9
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.5   Since  OFA introduces
combustion air at two different levels in the furnace, this
combustion hardware is also called air staging.
     uss o~ OFA reduces tne amount of a —r in ens burner zc^,~
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

                             3-15

-------
             o
             C\J
                                          t- *O 0)
                                                                             •H
                                                                             tO

                                                                             0)
                                                                             i-l
                                                                          T3-H
                                                                           0) <4-l
                                                                           H i-l
                                                                          -H Q)
|.i
§|g
IK
o o
00
                                           o
                                         c to
                                           o
                                           O
                                                                           (C
                                 T)
                                 (I) )-l
                                 U 0)
                                 O -H
                                 OU-H
                                 & o
                                 O £>
1
875
cu —
13
vp
0s <
° LJ
uS ^

N
C
5
t |

k- "o "53
17 C ^

'
                                                                           oo
                                                                           i
                                                                           ro
                                          u_ "C O
                                                                            0)
                                           g
                                         C ^5 05
                                           O
                                           O
                                                                         o
                                                                         a
                                                                            o
                                                                          o -i
                                                                         i— I -H
                                                                          ft K-l
                                                                          U  i
                                                                         •H — i
                                                                          &I-H
                                                                          >, fO
                                ft
                                oo
                                 i
13
                                  3-16
-ts «
    . 3
                                                                          0)
                                                                          ^

                                                                          Cn
                                                                         •H

-------
                                    0)

                                   r-l

                                   •H

                                 C O

                                 O.Q
                                 (C


                                 0)
                                   TJ
                                    0)
                                    (T!
                                 o-d
                                    a>
                                 •o  co
                                 0)  D
                                 u  a
                                 c  ex
                                 ro  o
                                 ro


                                 0)



                                 tr
                                 •H

                                 b-
                                 m£

                                 £•0
                                 £2
                                 ^7

                                 o^

                                 ^ s
                                 re •*•

                                 o 'O

                                 °g
                                 c o
                                 0) g

                                 ^°
                                 o
                                 o ]





                                  (0



                                 CO


                                  0)



                                  Cn
3-17

-------
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,
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 directl-y
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 NCX 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
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.5   These  conflicting
requirements must be considered when designing OFA systems.
                             3-18

-------
  coal
offset air
  coal
   oil
  coal
   air
 LNCFS
 Level 1
                  SOFA
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

-------
     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
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 NC.
This also promotes the formation of CC and  H2 which
catalytically reduces any NO formed to N2 in the presence of
CaO and char catalysts present in the bed.8
     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
bed material.8  NOX  formation  decreases  with  an  increase in the
degree of staging.  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 in the freeboard the secondary
air can be introduced without affecting combustion efficiency,
corrosion, and steam temperature control.
                             3-20

-------
3.2  PERFORMANCE OF COMBUSTION CONTROLS ON COAL-FIRED BOILERS
     The performance of combustion controls on subpart Da
coal-fired boilers was evaluated using detailed statistical
analysis of CEM data obtained from the operators of
conventional and FBC electric utility boilers.  The objective
of the data analysis was to assess long-term NOX emission
levels that can be continuously achieved by these boilers.
For the data analysis, individual boilers were selected to
represent each of the primary coal types and furnace
configurations used in this source category.
     This section describes the procedures used to select
individual boilers for statistical analysis and presents the
results from these analyses.  A preliminary analysis of the
data is presented in section 3.2.1.  Section 3.2.2 discusses
the boiler design parameters examined and the boilers selected
for detailed analysis.  Finally, the statistical analyses of
NCX emissions data from each of the selected boilers is
presented in section 3.2.3.
3.2.1  Surr.r-.ary of Available long-Term  'CEM,  Data
     All of the CEK data obtained from operators of subpart Da
conventional,  FBC, and stoker-fired electric utility boilers
were reviewed to identify the parameters (NOX, load, etc.)
reported,  the averaging period for the data, and the typical
NOX emission values.  The results of this review are
summarized in table 3-2 for conventional boilers and in
table 3-3 for FBC and stoker-fired boilers.   The typical
long-term NOX emissions levels shown are based in most cases
on ninety days of CEM data provided for each boiler.
     The data in table 3-2 are sorted by coal rank  (lignite,
subbituminous,  or bituminous) and boiler design (wall or
tangential).   All wall-fired boilers are equipped with low NOX
burners for NOX control.  A few are also equipped with
overfire air.   The tangentially-fired units are all equipped
with low NOX burners and overfire air.
     The data for FBC units shown in table 3-3 are also sorted
by coal rank and boiler design  (bubbling bed or circulating
                             3-21

-------
IX
(0
o
CO
D
CO
o
M
EH

U
>

O
u

IX
o
IX
<
£
£

cn
Q



(X
W
EH


O

O
m
<
EH


X

«
o
'1
»—





t
4*
I
Vi
u
01
4>
U
ic
a.










«<
g
•*
H-
CK
*+*
»2
7; ff
*l
v*

•g
^
ft
^

•g
o
_*
»
O
z
"o
s"s
o
u
If
-\
o B
CO ^
o
Nl
en aj
tl
c ^
z>
*J ,
5§
c
4S
c
re
a

o


*
o
>»
o

_
»
*•/
*
(J
a>
o
z
3=
CD
Z

_J
o
"re
3
s
-o


*
z
4->
O
a
o

«
«
in
-*
O

R
&.
a
Tf.
C
ID

!
a
«.
u.
O
•f
z
-J
_J
cr
c
re
t—
s

"
anton
vi


*
*
«
in
o

K
&
in
3

'•
0
O
r-l
<
u.
O
+
Z
_J
_J
O>
r
a
»—
r>
o
CO


mcstone
_j


*
«
in
o

8
*
3

'
O
o
fl
<
u.
O
+
Z
-J
OI
r
ro
*—
f>
O
00


mestone
~j
rvi

J
3
o

s
X
o
g

o
o
o
K)
CD
z

(/)
.
5
2
IN!


1
"re
>
X
w
t.
z
K)

*
£
e

i\j
o-
£
3

•'
o
CD
Z

VI
o
ro
3
CO
>»
>T


O
(0
CJ
%»

«
*
&
o

CM
1
=

!
o
o
ro
CO
z

in
O
IB
3
O
O
O-


erburne
in
in

«
o
o

(M
1
3

;
0
CD
Z

bO
O
5
O
Kl
-st


L.
0>
^*
§
z
•c


*0
o

rvi
1
g

z
z
CO
z

t/1
0
5
CO
xT
^o


1
u
z
N.

*
in
o

IV
i!
Q
g

O
O
CO
z

(/)
o
re
3
o
vl
•sj


re
N
re
g
CD
00


m
o

g
*
ik
c
a

X
X
CO
z



«
*
in
0

s
1.
a
X
c
a

:
0
o
Kl
CO
Z

in
o
(0
3
0
ro
h-


§
1
O
0


in
o

81
"3
•^
§

z
*
<
U.
O
+
2
_j
trt
F1
re
^-
in
•O


f
VI
1_
a.



o
o

S
&
V)
*3

Z
3:
<
u_
O
•f
Z
_1
en
?
re
in
-*


&
4-»
4^
&
X
u.
fM

*
«
M
O

IM
ii
O
4J
^

•'
o
o
Kl
<
u.
O
^
z
_J
in
01
C
10
t—
o
u^
r\i


8
a
a.
Kl


in
fw
0

10
u
g

z
-
<
U.
0
Z
_j
cn
P
re
f—
o-
r^.
rvi


u
•o
i
DC
rj


in
o

»
&
vi
*

z
-
CD
1

CO
.
re
3
o
3


(A
H
U
O
£
I/I
C
o
•g
re
CO
3


&
o

r\<
&
v>
g

-
X
CD
Z

CO
,
5
o
in
_


«
o
o

f\J
2
O
g

I
o
CD
Z

CO
o
5
r\j
-o


i_
>
or
(A
C
.r
o
-^
M
fs.


*
O
O

rvi
Z
0
g

;
o
CO
z

CO
o
5
ro
S


L.
>
oc
(A
c
-^
CO
CO


«
irt
0

CM
&
i/i
3

O
0
CD
Z

CD
O
n
3
ru
s


.c
en
i
^
R;


&
o

(M
£
g

z
z
m
z

CD
O
re
3
in
a


o
L.
co
<
o

*
&
o

«M
1
g

0
o
CO
z

CO
o
§
o
NX
v»


g
Ul
3
0


*
&
o

rvj
i!
o
X

a
o
CO
z

CO
C)
§
3
N3-


Stanton
U
i
(NJ

*
*
&
o

N
1
3

;
o
o
Kl
«C
u.
O
+
Z
_J
CD
o
5
o
S


c
1
i
V)
                                          3-22

-------
•d
 0)
T5
 3
(H
 U
 c
 o
u
w
a:
O
m

(0
D
CO
O
M
H

U
>

O
u

c:
o
t-
£
w
H
 i
U
2
O
CM

 I

ro
CQ
<
EH

*
Z
u
'!





reportedc
Parameters









Reference


v ««.

"8
&
•^

I
>
o
Z

o
X L.
* i

n
EC "
c
3
*- Z
C ^
D
I8
C
n
o.
~
«
o
in
o

e>
&
3

'•
0
o
K^

u.
0
CD
Z

CO
o
re
3
O
CO
r\j
Seminole
Kl
Kl

o
o

8!
&



o
Q

m
-1

CD
O
re
3
o
§
-
c
re
C
R

S
o

e>
fr
V)
3

O
o

CO
_J

CO
o
re
0
o
o
rg
C
re
t_
01
C
M

o
in
o

fv
V
to


!
o

o
CO
Z

CO
CT)
C
re
o-
t
Z
01
re
u
in
ft

o
in
o

rvj
&
»


;
o

o
m

CD
u>
c.
c
INJ
O
K)
CO
Z
ce
u
*•-
re
JC
u
£

O
in
o

r\j
&
1/1
3

O
O

o
CO
Z

EC
!
1
o-
Muscatine
a

*
in
0

PO
8-
3


O

o
CO
Z

CD
cn
C
re
O
0
Kl
tfi
0
L.
U
Kl

s
0

IM
&
3

X
-

O
CO
Z

CD
re
in
-
1
CD
O>
CO


1

'^
X
"S
'^
c
c
tr.
C
n

O)
C
re




&
•-

^
re
3
1
a
o

n
o
_^


3
•g
u
c
-1-
re

Cl

1

«i
n
_»

n
3
re
























tr
3
g
|
•-

II

cc
(A


c
1
•r-
s
V)
n
en


0
4-»

i


n


JD
provided.

Cl
en
re
o
re
X
3
O


II
7.


C
o



Cl
re
o
re
X

're
M
o
, fc


§
V!
c>
Dl
re
ti
a

c"
'Z
~o

X
n
T3
O
Kl
II
0

O

u
ay rolling averages
•D
O

g
"2
S
-°
H


*
*
•g
re



0
en
re
t_
CJ
re
X
._
re
•D
C
o
"g
re
.0

n
*


VI
en
re
i.
OJ
CD
X
l_
1


g

"8
v>
re
m

ii

*









































^
*
u
o
fr
L.

8

II


•
                                           3-23

-------
w

u
J
H
O
CQ

O
u
w

D
U
CQ
(C
D
Cu
CQ
D
W


K
O
t-
s:


w


p

Q
W
H
 I
O
n
 I
o
CQ
«^
c
cc
X
s.
H-
(lb/MMBtu)*
&
,—
1
!
u
re
0.
»
O
T3
O
N
U
C
o
u
ri
CD ^
Capacity
3
z
4-< t
I
c
re
et











)N BOILERS
^>
—
CO

eo
o

rsi
a
>
o
3C
y
i/>



X

u

_J



o
r\j



8
o.
z
»—
0-

rw
o
o

»
i!
o
8



z

u

m



o
<



U)
-C
t—
o

r—
o
0

CM
O
Cl
0
s



z

u

CD



O
CD



Thames
o

(M
O


3
^


z

u

Of
(_}
•v.
CO



in
(M



I
re
u


*
*
S
o

CM
O-
I
t
C/)
K)
X
ro
X
or
2
l/l
X
Z
CO

CO

r\i
<


*
Stocktc


*
*
s
o

C\J
o
I
«•
tfl
ro
X
K1
X
(X
Z
l/J
X
z
CD

CO

00
CD


e
Stocktc
CM


O


|
S



X
cc
z
I/)
X
z
CD

U

O
<

4-<
C
a.
Barber:
Kl

S
o


*J
8
*



X
cc
z
l/l
X
z
CD

U

in
<
Z
neration
1
S
StockK
«r

S
o

rvi
|
4^
t
(/)


x
Of
z
CO
z
z
CQ
U
u

m
<
z
neration
u
S
o
0
a.
«^
z
in


O

nj
&
(/I
1
3
~*


X
QC
Z

(C
-s
5
u
m
0
QC
IA


0

(M
fr
to
3
"*


X
QC
Z
(/)
X
z
CO
u
u

Kl

0
^
CM
|
1
**
a-
««


z

u

u
s
u
CD

(M
CM
CM
*
4->
i
Steam 1





















r^

ro
in
O

R
S
a.



o

u

CO
01
o
I/)

*•*
<
«>
o
z
X
.*
o
o
Of
X
c
01
en
o
u
h.

eo
-j
o

R
S
a



o

u

CO
01
o
LO

fO
CO
*->
z
>.
J£
ot
X
Cogent
N.

O
in
o

R
§
s-



o



CO
01
o
CO

Kl
<
M
«•<
Z
X
.*
o
DC
X
c
Oi
en
0
u
N.

K
o

R
§
s-



o



CO
01
o
en

fO
CO
CM
4->
Z
X
-K
O
cc
X
£
en
o
u
N.

S
O

R
fr
V)
3



O

ce
(/>
CO
Oi
o
CO

fO
<
1
cc
X
£
Ol
o
u
r—

S
o

R
I
(A
3



O

CC
in
CO
u
o
CO

fO
CD
chmond
cc
X
£
Cl
o
u
r>.

R
0

Ov
&
l/>
i
3



a

ce
CO
CO
01
o
CO

Kl
<
CM
chmond
ce
X
£
en
o
LJ
r-.

a
o

R
*«
&
in
3



O

ce
CO
CD
O>
O
CO

Kl
CD
CM
.C
oc
X
c
Oi
en
0
u
r>-

s
o

R
1
3



O

ce
LJ
CO
CD
OI
O
CO

Kl
<
Kl
I
DC
X
C
OJ
OI
o
u
r>-

Kl
O

R
4-*
fr
«A
3



O

O£
U
C/)
CO
0>
o
VJ

fO
CD
K)
?
x
o
ee
X
1
i_>
                                            3-24

-------
a
u
.4
H
O
CD

D
U
a
 I
a
u
x,
o
E-"
u
CO
fe-

re
D
CO
D
W

a
U-
 QJ
T3
 3
^-l
 O
   u
u
H
 I
O
z
s
n
 I
n
CO



ox
z
a
u
ft





Ci
L.
8.
4*

L.
O
4V
l~
IB
a




u
u
tl
ee
*
si
* ^.
- £

^8
&
i
'^
re
o




>
o
2


U
O
p~h


z c
u
fl
CO *"




X
u 3
re z
i. t
TO
_>
1 §
















i
c
4J
c
IB
CL








"*

~
0

K>
O-
&
«n
3






o



or
tj
w


m





c

o

l/l



^_


«t


•g
I
u
DC
fc
X
^

c
&
o



.
•*

R
o

K!
O
&
f
•3






o



cc
o
trt


CO





tl

o

u^



^


CO


•g
x:

O£
^
X
l_




S













u
•g
cc

Uu
o
a:
"i
.§ 5
*- >
3 L-
•i -o
S
U ft)
C/l
"P 3
.£) *£)
U
•o
O
*j "D
3
Z>
^ 0
Ol ft>
C u
re x
"3 'I
u
1. II

O (J
ex
u 3

CD C*
u. 01
U 13
c" !!
o

tr O
3 u
u
4,
•° s
"S 8
i I
5 —
01
C CO
n 4j
2 4-1
3 C
O)
It —

u
CD II
u.
CD — J

                                        O
                                        L-
                                        o.
                                          -c
                                           c.
                                          •c
                                           o

                                           a

                                           V.
                                           CI

                                           re
                                       o

                                       CL
                                               I   5
                                               •s   ~
                                               in
                                               o
                                               CO
                                                        3
                                                        O
                                                        "8
                                                      3-25

-------
bed).   The NOX control technologies that have been applied to
these units include CC, such as air staging in the combustion
zone,  and selective noncatalytic reduction (SNCR).   Table 3-3
also includes data for subpart Da stoker-fired boilers.  All
of the boilers burn bituminous coal.  The NOX control
technologies that have been applied include combustion
controls  (e.g., air staging)  and SNCR.
     To highlight the differences in the NOX emissions as a.
function of coal rank and boiler furnace type, all of the data
in tables 3-2 and 3-3 are summarized in table 3-4.   The table
lists the average NOX emission rate, the range of  NOX emission
values,  and the number of boilers within each coal rank and
boiler furnace type category.
3.2.2  Selection of Boilers for Analysis
     Based on the differences in NOX emissions as  a function
of coal rank and furnace type, representative boilers were
selected within each fuel and furnace category for detailed
analysis.  To ensure the selected boilers are representative
of the units within each category,  key design parameters for
each cf the boilers  furnace volumes,  heat release rates,
etc.)  were compared.
     3.2.2.1  Analysis of Boiler Design Data
     3.2.2.1.1  Conventional boilers.   To determine furnace or
burner zone heat release rates, information received on the
conventional subpart Da boilers and the PowerPlants database45
were used as explained in appendix A.   The total furnace and
burner zone volumes obtained from these sources are shown in
table 3-5.  The maximum furnace and burner zone heat release
rates (Btu/hr-ft-3) computed from these values are  also
presented in table 3-5.
     A graph of the total furnace volume versus the burner
zone volume is shown in figure 3-11.  The plot includes only
those boilers for which adequate data were available from the
subpart Da boiler operators to calculate both volumes.  As
seen in the figure, most boilers are designed with similar
                             3-26

-------


















<
H
<
*"•*•<
D

TERM
1
o
S
c
J
&J
o

><
C£
1

p
cc


^

n"
to
!-J
ffi
<
H





























tn
rH
0)
p
0)
rH
c
o
-rH
tn
en
-H
g
td "^
HJ -J
^•^
0@
2 g
6 >
1-
H
Cn
C
O
J
r— 1
rc

-^-
£1,
rr^


























0)
5-i
f-t

0)
u
IB
lH
3
Cu

















JH Pi
0) U
i! 2
o to
4J ^ —
M S



^
0)
^
0
4J
W




Pi
U2
s|
3




U
CQ
fc,


r-l
IS
-H
4J
C
0)
Cn
c
^
c-1


r— 1
r-H
rd






4J ro >i cnC
C D 4J ro C g
CD -H T3 -H ~
M CO E rH '^
vj a -H o ^ J
3 CO r4 M 0 S
U 2 - *H -

a
c
C
£

r
(

L























00
rH
^°IB
0^

O
in
co o
•=>• i Si
• in m
o ^<

c



c
^ rd
rH
O



CO
4J
0) -rl
Cn C
rd 0) 3
in Cn
0' C
> (C C
< Pi 2


1 ^
o
i

J
1,
>1
-t
CD
H 4J
U -H
3 C
ij cn
-H
,4


























in
ro
° 0-,
m , TJ
' in ^
0 CN

f-^

in
<#
£°U
1 c ^
0 £

c

tn
4-1
0) -rl
Cn C
to 0) 3
'H Cn
0' C
> re 0
< B. 2


in
0

cn
3
0
C
-H

•3
4J
•H
£1
X!
3
CO
rH
m

5 o _e
" , it
• CO
°s

o

m
in
£rd
°°/

G>
LTt
U~l
IS OUH
^ 4-
0°^
0 «•

(^

IP
LO
S0<1)
o ^

o

CD
4->
0! -H
Cn C
m o) 3
,H Cn
0) C •
> fC 0
< a 2


^
o



tn
3
0
C
•H


4J
•rH
03



Cn
Cn C
c ••j
-H rH
rH rH
•H S
O rH
. ^ • >,
4J >i J-1 !§
-H 18 -n T3
§ * 30
o ^
mi m rH *
PH ™1 ' ' 	 .
T3
n -o ^ g . S
£ S ° « -H
^ 4J C
tn tn en tn -H 3
0) 4J CD 4J g
Si -H cn -H 3m
* c * c
SH 3 -H 3 "* n
0) 5 ,. °
> CN > CT\ n
* ^ * ^ ^ S
>, 0 Cn 0 J>
rH UH C M-i w Pi
7, -rH 0) <0
3 CO rH U) Cn l-i
00) rH 0) IS £
£ Cn 0 Cn VH >
rd -H rO |U *
CO) >i Q) (8 >i
10 5 -S 5 ^ •§
S >, 0 >, -H |
-H rH m r-l IS J=
C -H -n T3 _
3 * T3 rfl -0
•a c -o 'C c
rsi rd S
rc
c: v- t-
C i- 4- X- ^ 4-
- •? "5 c c S
tn 3 33 33
pi rn m m rH CN
? 0 0 0 00
CD O U U , ,X
> U-l MH 4H 4-1 4H
rr [T LO tD CO CO '
i s i ii ill
I.! rH H rl 1H V, in 18
^^S g) 0. £££
rS 2 m * « * * >
ID H 'O "-• v —
>,>,>, >, >, ^ M S,
i-H (B rH r-HrH T'SZ?
Lj'Tll-l M H >-l O rH
^ M M _r" -H
Psl.rr.£--S
^ffltn
g § g- g g.^ g B §
•g 'S "S§ ^ 'S3 "S « "S
- coSSrocotncNCQtntn
cflfCro * S , Snn
(DXiXl-H^il^-0-0-0
o o
CD CD tn4H tn tn 4H tn into
rHi-HrH rH rH c~trH'7'1
oifli aicn 0) ai w o> cu cp
>>>i>>fl)>>>
rj £ tiicntt) (DCno) H
X X X
0 0 0 > 0 0 > 0 0 0
a^^rtss*2^?
IS J (J -0 (U 4HO)£
3-27

-------
CO
O
CO
§
M
EH
O
u

(0
Q
w

(X
o
CL,
w
EH
CU

Z
o
IT)
 I
.J
CO
1_
u
»»-
(I
ee
u ^»
U C, »J
I-""-
2|L S
•= 01 i
E ei <" "-
3 £ re x:
.5 5 « ^
X N t> 4J
« k. CD
Z v»
^*
I?
~ 3
«_ ~
4> ..
U
CB "S
^V
4-- 0* -M
10 4-« *-
c at »
Kl

§
•»
fV
f\J
r\j
in
s
r-.

>o
s
o
o
in
o
-o
N.
O
re
3
.J
-
en
X
+•*
u
"o
a
o


s
eo

o

?
TO
_J
O
Stanton
*—


0
•-

o-
3
in
«»
CM
rsi
o
>»
eo

«
iTi
K^
O
o-
N.
^»
>O
N.
Ol
C
(C
»—
_J
-
Limestone
t«-


M
O>
O

I
in
a
N.
U1
Kl
CO

*
m
Kl
O
O-
>c
C
.C
N.
Cl
C
(Q
H-
_J
rj
Limestone
rvj


§
*—

*
§
S
o-
K1
•-

«
e
m
-C
fVJ
l/^
-o
m
ro

re
3
M
(M
r
n
>
JC
u
L.
O
•x
o-


Kl
•O
CO

s
IA
rS
I
fO

g
CO
r\j
CM
-o
m
0-
rvj

(D
3
c
o-
^J
o
re
3
VI
M
CD
a
u
u
^*


ro
ft

-£
«
O
s
•o
-*
rj
0-

2
o
o-
o
o
>o
co
0
re
3
(/)
K»
Sherburne Co.
m


z


<
z

«
z


«
z


o
o
0-
Kl
O
re
3
VI
Kl
Hunter
•o


&
S

0
in
s

-
Holcomb
K.


s
o-

s
in
s
s
''I
»o

i
CO
Kt
o
M
^J
^1
0
5
in
-
Bonanza
o>


s
Kl

•*
-»
in
s
IM
1
O

O
•C
in
f»
>O
o
§
•o
o
a
3

-
Ok I aim ion
co


£
<5

,
i
r\i
&•
*-

§
^
v»
«»
8
Kl
in
0
re
3
M
rj
|
u
o



ft

8
»—
T
0

1
o-
in
eo
fsj
r^
m
o
C
n>
^~
U)
CO
Parish



•o
^

CO
(M
i
o
r\j
m
Kl

§
CO
C-
K)
O
O
ro
~±

>—
in
Kl
Fayette
0


z


5

<
z


«t
z


CO
in

Ol
C
(D
^~
I/I
Kl
Shi ras
Kl


rS
co

R
•o
«f
in
s
M

§
O
s
CM
O
r*.
•0
fVJ
r"
C
t—
V)
'-
Rawhide
r\j


§
tx

§
5
(NJ
rj
Kl

o
§
s

g
^»
nj
w
c

n
i_
CO
%»


co
Kl

.
s
*—
s
o-
Kl

g
m
-*»
c
>o

re
3
CD
(M
tfl
01
L.
O
5
L.
CD



^
in
CO

0
z
K
CO
o
in

*
-o
o-
^
o
in
in


re
3
CD
-
L.
"3
*
«-*
+rf
0
u
1
w
in


»
S

g
Oj.
fc
v»
N»

NO
eo

§

3
»
•~
o
n
3
CD
-
U
CM
f>-


8
S

«'
i
|
R
o-

i
CO

ee
1
o
-j
+j
v>
f~


o
IM
tit

*
§
|
R
0-

S
co
•&
~f
*r
•o
p
s
CD
(M
1
QC
-^
4^
V*
s


in
•^
Is.

S
r\j
3
rj
£
o>

Kl
3
•«
!c
Kl
in
CM
~O
O
ID
3
CO
^»
Kintigh Station
R


rsj
o
w

8
5
(\t
S
(\f

O
S
o3

0
ro
in
CM
O
re
3
CD
CM
L.
CD
CD
<
                                               3-28

-------
TJ
 0)
T3
 U

 o
u
w
(J
H
o
CO
o
2
o
u

(0
D
CU
CQ
Di
O
w
H
W
s:
<
a
2

O
M
 i
r*>
tJ
CO
Cl
£
L.
Cl
«*•
Cl
oc
u ^»
Cl Cl +>
C ^ *••••-
il=?
E u <" c
5 ? a f
.1 o *>
•£ w-J 2
£1- CD
^
V *"
(. w
Al .,
Ij
~ Cl *•
TO 4J ^
c ci re •
S -C i- 3
E (-1
.= 0 Cl .
± 
•5 i
re *-*
.O
|
01
'o
m
_, re
*- ,
c
\
a

s
R
••
in
CO
§
1
—


3
g
Kl

ee

o
re
3
CO
-
c
o
3
CO

Kl
2

o
3
5
1C
in
(M


o
CO
(NJ
KV

"

O
re
3
CO
-
CH Stanton

Kl
8
in

*
in
ex
o
o
rj


in
•C
in

r\j
(X'

O
re
3
CO
-
Seminole

Kl
s

in
s
o
f\J


in
m

R:
•o
o
re
3
CO
rv
Scminole

Kl
I
OJ
o
in
in
Kl
rvi
r^


tM
in
CM
CM

u~
to
00
o
5
m
-
C
re
L_
C

Kl
O
CM
0
in
in
§
Kl
(M
Kl
r-


CM
IA
CM
CM

£
00
o
re
3
m
r\j
Intermountain

Kt
§

«
»•
CO
s
•-


in
Kl
Kl

-C

Dl
r

CO
t
I.
Cl
re
u
z
o:

Kl
|

~
5
CO
•-


in
Kl
Kl
Kl

$

Ol
C

CO
00
RM Schahfer

Kl
00
O
Kl
CM
CO
co
Kl
O
o


o
o
5



Ol
c
ro
CD
0
Muscat ine

Kt
z

<

<



z


c
PJ
in
Ol
c
re
CO
CM
VI
O
U

Kl
g
Kl

J*
CM
O
Kl
O
«-


fc
Kl
O-
Kl

Kl

C
CC
CO
-
1
CO
Ol
CO
                                                                    c   <
                                                                    re
                                                                    4S    X
     t   5
     u    T3

     •C    £
5   -i
O   X

1   "8
                                                                         O
                                                                         l_
                                                                         a
^    >*-    Q

1    T    =
                                                               c
                                                               d
     ci    re
     ^    ^
     o>    a
     C.    T>

     °»    Cl
          (/)
     II    Cl
                                                          3-29

-------
                                  8
                                                   Q)
                                                   C
                                                   o
                                                   N
                                                   OJ  •
                                                   C w
                                                   ^ i-i
                                                   3 0)
                                                  J3 rH
                                                     •H
                                                  T3 O
                                                   C J3
                                                   rO
                                                     (0
                                                   0) Q

                                                   3 4J
                                                  .H H
                                                   O fO
                                                   > U
                                                     ja
                                                   0) 3
                                                   u w
                                                  m
                                                     c
                                                     O
                                                  a)
                                                  OJ
                                                  0)
                                                  c  ^
                                                  o  o
                                                  >-l  QJ
                                                  m  E
                                                  e-3
                                                  E rH
                                                  O  O
                                                 U  >
                                                 ro

                                                 a;
                                                 ^

                                                 en
               8
               ca
8
3-30

-------
proportions between the burner zone volume and the total
furnace volume.  Exceptions include the Brandon Shores Units 1
and 2 and the Zimmer Unit 1.  The Brandon Shores boilers were
originally designed for oil firing, but the fuel was switched
to coal prior to start-up.24  Therefore, this boiler design may
not be representative of a typical coal-fired boiler.  Zimmer
personnel52 indicated that before construction, the burner
arrangement for this boiler was redesigned and changed from
three rows to four rows which resulted in an increase in
burner zone volume.  This is reflected in figure 3-11 which
shows that the boiler has a disproportionately larger burner
zone volume when compared to the other boilers.
     Figure 3-12 compares maximum thermal heat input
(MMBtu/hr) to the boiler to total furnace volume (ft3) .   With
the exception of the Brandon Shores Units 1 & 2 and Zimmer
Unit 1, the data show that the maximum thermal heat input
correlates with furnace volume.
     To assess the effect of the furnace and burner zone heat
release rates or. NTY emissions, the reported short-tern" NOV
 ~                 .A.           '       -                   .A.
emission data, measured at the maximum tested load for each
boiler, are shown in table 3-6.  The furnace and burner zone
heat release rates shown in the table do not correspond to the
maximum thermal heat input to the boiler, but reflect the heat
inputb  corresponding  to  the  maximum load  at  which the
emissions data were measured.  Only boilers with available NOX
emissions, load, and design data are included in the table.
     The short-term NOX emissions data for each boiler in
table 3-6 are plotted in figures 3-13,  3-14, and 3-15 as a
function of the furnace heat release rate (Btu/hr-ft3),  the
burner zone heat release rate  (Btu/hr-ft3),  and the heat
release per furnace plan area  (Btu/hr-ft2),  respectively.
Short-term NOX emissions data were used instead of long-term
(CEM) data because the variations in the measured NOX and load
     b
     The  thermal  heat  input was  calculated based  on  an assumed
heat rate = 10,000 Btu/kWh for all units.

                             3-31

-------
                         i



T3
«
O)
JS






0)
O)
ID
•
en
3
1
E

15
•s
to
-*


*
^
.5
E
.ts
m
«
                                      ir
*-  H

                                          §
                                          O
                                                                             8    fe
                                                                             co
                                                                                  I
                                                                                  
                                                            O rH
                                                           •H -rl
                                                            4J  O
                                                            U^2
                                                            C
                                                            3  m
                                                            "W  Q

                                                            OS  4J
                                                               M
                                                            w  m
                                                            a
                                                            C rH
                                                           •H  tO
                                                               c
                                                            4J  O
                                                            (0 -rl

                                                            XI  C
                                                               0)
                                                            fO
                                                               O
                                                               U
                                                                                            0)
                                                                                           XI
                                                            3 0)

                                                            •H 3
                                                            X rH
                                                            ro O
                                                            S >
                                                                                            (N
                                                                                             I
                                                                                            m
CNJ
      r-    O
                      yidui
                                           in
                                                              PJ    »-
                                               3-32

-------
§
o
u

ro
D

EH
Pi
02
D
CO

a
o
u.

<
EH
<
a

2
O
M
LO
W
M

U

 X
o
a
w
EH
 I
EH
a
o
 I
n
CO
<
EH





•8
o
T?
4-*
(/>
o
4-'
4V
(0
i t i ons
s
o










41
|
4>
<^
C
EL
E ^
fc S3
«* v— CD
r fi 1
I fS
« *»
4^ •
C V
L. tl H-
U « •
ti a ii cr
l/> C L. 10
8t « •
•z1^ g^
:n-i
i i
8 *
"K
"- u *
E - •-
fc CO
-1
^i
^£
4i a/
8S
t u
S£
^u
fl_
^
^
"~ D
"i
x|
E "D
g OJ TD ••-
fe *j re 3
J « 0 Z
ra £- ^
z
^
o|
CD
(0
S|
12
c
«rf
c
10
a.
0-

»
to
o
in

•o
o

ro
r^

o
in
o-
•0
s
•o
o
re
3
_i
-
U)
X
*-
41
"o
0


CO
-»
o
in

o
CM

o
eo

o
ec
o
CO
00
o
CO
en
c
CO
H-
_J
-
Limestone


»
0
in

00
r-

»
o
CO

in
N.

o>-
CM

O
IT
(X
•J
in
CM
>j
.
CC
3
t/>
-
Bonanza
o

-»
-»
0
•0

CM
>»

v»
O

o
o
o
r-
o
e
o
re
3
iy>
-
Oklaunion

CO
Parish


«M
•»
O

ry
CM
r-

~»
to

O
m
CM
VI
in
r\j
*»
CD
c
CO
h—
in
Kl
Fayette
to

a
o
•0

fO
CM

O
rO

O
>c
CC
no
£
CM
01
c
<6
(/>
-
Rawhide
CM

CM
K>
o
r>-
«—
o
IM

00
CO

O
CO
LO
r\j
CO
in
(Ni
f
to
*—
in
-
Plains
S

0
in
o

CM
»»
*

XI
in

o
s
-o
IS
>o
_,
re
3
CC
-
£
(3
£
l/>
CO
t.
CO
s

CM
in
o

CM
in
•O

to
in

o
0
oo
•0
CO
>o
_
re
3
CO
rj
w
CJ
u
o
f.
v>
l_
CO
in

h.
>»
O
f\J
T-
in
r5

to
*»

o
o
«o

o
0
	 ,
5
CC
w
1
I/I
CO
-a
O

0
•*
o
in
rsi
in
C\l

eo
-»

e
o-
fo
hw
o
H^
o
CC
3
CO
-
L.
r-J
ft

ft
0
«»
«—
CM
«-

CM
CM

O
rvj
%r
INJ
CM
^J
CM
O
ra
3
CO
CM
L.
CO
CO
<
e

0
•»
o
«
v
r<~
oo


^j-
0
5
CO
-
§
V)
3
CD
O


CM
«»
O
r
CO


o

0
>»

o
o
-o
»»
§
^3-
0
5
CO
-
CH Stanton
m

«
O

•

in

IV
tv

o
0
»<1
o
0
to

-------

W
Di
U
•J
M
O
OQ
|
O
M
EH

U
^
8


PH **^
rt; "O
CU 01
eo -o
D 3
10 iH
O
&. c
0 0
It, U
*— '
^5
Q
*^

O

w
CO

M
£
X
o
£
1

a

a
CO

\£>
1
ro

tQ

pa
EH














o
T
Vi
Of
4.,
1C
c
Q

.t^
1
U































tl
i.
tl
«4-
tl
ee
* *•- »

0 I J2
tl *»
4V •
« 4V
U tl •«-
u a •
ti fi ti O"
vi C i- at
*> t- a i

?:|f
ti i
x --
c 3>
o j^

*- "^ ~
4> (Q
5 4V «-
ffi c
^
4V-D
*o ^
<^ m
-<= 2
t> jj
s s
u.


zL~

'- 3
4- ^


Z S
.1 i re 3
1 ^
_^
CO **


|\


18



a
1
CL




KJ


Kl
O








Kl
rsj






>»
"^



0
Kl

00*



Kl

O
1


00


-



c
ermountai

c


Kl


Kl
0








fNJ
in






«-
^



o

^
Kl



£
Kl

f
re
•"


CO


r>-




Schahfer

x
IX

3


«_
o








*»
5






fNJ
^



o

^
Kl



Kl

?
fr-


ee


"




Schahfer

x
K

in
Kl


to
Kl
0








O
in






fsi
rvj



o
rvj

,—



rv

?
re
*~


CO


0




catine
w

1

r-.
Kl


,_
0








N-
™






•-
^



O

K>^
^y



Kl

1
H-


00


-




1
CD
O)

m




















•p
L.
H-
X
"(D
4-»
C
C
cn
C
c

n

O)
c
<0
»—
l_
3 (0 •
.— -n S
§ w —
"° o "§ m
If J* C

CD 3
O 4"* ifl
in" — *° 0
c 3 o j C
§"~ •- O* H- E
4- -g ° =; 1
5 •- " V c
2 >*- ti i- o
3 1 4V .C .-
(A — ' CD **. 4V
II ^0 4V tl
3 4V CD VI
vt a
ti u o in
•» — ' .C O VI
tl (71 O O
4V C 0 L.
— .- »- O
O> O II U
— II O
- - "8 2 g


— 1 3 IE 2 U.
re ja o ~C ti
3-34

-------












































;


!s
= i
* c
II D
£ H








.^
>
4



























4

*
4
^
«
*
* H ^
1 t-
X
H
X
I
•£
«
^
> >
* > 4 4
I X


f-
44 r-
1- »

?
1









F—
1
i-




*
1
i- ^ ^ •» ^ r jl

0)
-£ u
(U
V
w
u
-- ft)
"^
^^
C!
o
_f, i—
«5.
3
m



_N '•^
"re

o
re
~ "w
CC
re
o
*»^
•~"
= °
~ re
c
_j
f T



^^


I
g

£

^^
^H

minous
3
IE
x;
t/5
)K





o
"E
ot
_;




0)
U
(0
c
14-1

o

c
o
•H
4J
u
c
ll (
*^-l
fC
tn
m
c
o
•H
w
CO
•H
E
QJ
X
O
55

E
L
0)
1
£J
0
CO
— a:
I
m
r^ O
i-t
i 3













'
^
0)
,— 1
•H
O
fO
Q

4-1
TO
a
3
w

iH
(C
C
0
•H
4J
C
OJ
>
c
o
u

l_l
o
U-l
a>
4J
U




 XON
3-35

-------












































To
c
"3*
•? C
sP £
H
I' II
*H


*
4
































H ^
^ I
I
£
s
£
4 *
4
*
I

H
*
^
i

f-
4
j_ > i_
i- t- # i~
4 1 * 4 »


£

I


* =5
4 4
* *
I
H ^
I *
H
I
*
I



£
4

J> J. J. . i J, 1

•»
<-7i


fSi
-K

-R
-5?

-5?

-5



1^1
~^T




-t


-£


-S
'V
~T^,
f*^
~rr


c
^p^.

-s


55 C
O
6 u
m LJ
AJ O
1 <4-l
4J
i-l 0)
O 4->
jr m
CO t-i
T
r- 1
1
ro
Is a)
XON
 3-36

-------





|
= I £
^ £ *
*£•



H
i > *
4 «






















=
c
'1


>
*w-
»— 4
^ £
, * !: ^ ^t 3=
c
I
£
3

-H




V
'c
•"
j!
•

— * > ,
* * *
r-
• -;..
H >
I •

^
4


f-
4
4




L 1 1 1 1 1
1 £ s * s £ *"< i


"H
£T
t
^C *
s
~S *5
cs
u
_S ZS
N ^
re
c
3
U.
«-. .ts
-- c
b
c.
"co
c ^
""—; S^
re
—
^*
CM
<__.
CC
4/- +±+
^~




g
4
Be°b0e°d
0}
y CQ
"5 ^,
W 0)
rM
_— *^>
m °
0) -0
£S
ro (B
s SL
* JQ
**
^1
n
-?
«§
a u
re j_,
K °
C^
°s
M g
-LH^
§c
X1"1
§
e U
5 ^
(U g
^
1 ^
4J ^
1-
w §

in
,«(
i
ro
0)
g)
•H
tu
3-37

-------
caused by changes in boiler operating conditions could alter
any correlation that may exist between these quantities.  The
data in figures 3-13 and 3-14 do not indicate a correlation in
NOX emissions with either furnace or burner zone heat release
rates for these subpart Da boilers.  There also appears to be
no correlation among boilers burning the same fuel (lignite,
subbituminous, or bituminous) or of the same firing type (wall
or tangential).   The data in figure 3-15 also shows no
correlation between NOX emission and heat release rates per
furnace plan area.
     3.2.2.1.2  FBC boilers.  NOX emissions from FBC boilers
are influenced by several design and process parameters.
These variables include bed temperature, ash properties,
recirculation rate,  and coal properties such as nitrogen
content and reactivity.  As discussed in section 3.1.3, the
most widely used CC technique for NOX control is air staging
which includes secondary air ports located in the freeboard
section of the boiler.  In addition to air staging, some FBC
boilers are also equipped with SNCR for additional NOX
control.
     Among the boilers equipped only with CC for NOX control
(see table 3-3), inadequate data are available to correlate
the measured NOX emissions with the above-mentioned design or
process parameters.   The Cambria Units 1 and 2 burn coal
refuse as the primary fuel while Tacoma Public Utilities,
Units 1 and 2 burn a mixture of fuels including wood, refuse
derived fuel  (RDF),  and coal.  Consequently, these boilers
were not considered.
     Unlike typical FBC boilers, the Thames units are equipped
with an external heat exchanger.  The ash removed by the
cyclones is routed through this heat exchanger before
returning to the combustor.  By controlling the flow of ash,
lower bed temperatures  (about 75 °F lower than design) and NOX
emissions are achieved.  Although NOX emissions typically
decrease with lower bed temperatures, it is not known whether
this is the only factor contributing to the lower NOX
                              3-38

-------
emissions.  Additionally, the unburned carbon emissions from
this unit is high (about 12 percent).   Consequently,  these
boilers were also not considered for detailed analysis.
     3.2.2.2  Boiler selection.  The analysis presented in the
previous section indicates that most conventional subpart Da
boilers have similar furnace designs and burner zone heat
release rates.  Further, the NOX emissions from these boilers
do not appear to correlate with either the furnace or burner
zone heat release rates.  As a result, the observed
differences in NOX emissions from these boilers cannot be
attributed to differences in boiler design but may be due to
differences in boiler operation practices.  This is based on
information provided by plant personnel and is discussed in
sections 3.2.3.1 and 3.2.3.2.  Based on this analysis and the
observed differences in the measured long-term NOX emissions
values shown in table 3-4, representative boilers within each
fuel category were selected for detailed statistical analysis
of long-term  (CEM) data.  Except for Brandon Shores Units 1
ar.d 2, ana Zirrmer Unit 1, the other conventional boilers
appear re be acceptable fcr long-term dara analysis.
     The selection of 6 boilers is shown in table 3-7 and
represents the range of fuels fired and furnace types for the
boilers shown in table 3-4.  The selection includes three
wall-fired boilers (one each burning lignite, subbituminous,
and bituminous coal), two tangentially-fired boilers  (one each
burning subbituminous and bituminous coal), and one FBC boiler
(burning lignite).  Neither the analysis in section 3.2.2.1
nor the obtained data reveal anything unusual about the design
and operation of these boilers.  The selection of these
boilers was also based on the availability of hourly-average
NOX emissions and load data for these units.
3.2.3  Analysis of Long-Term Continuous Emission Monitoring
       Data
     This section describes the detailed statistical analysis
of CEM data obtained for each of the selected boilers.  The
objective of the data analysis was to assess the long-term NOX

                             3-39

-------
H
CO
Q

2
tt!
W
E-
O
o
fa

o
Cd
H
U
fa
J
EC
CC
Ct!
fa
O
CQ

fa
O

EH
CO
r-

ro

Ed

CQ

H
u
contro]
^
o
a
0)
N
•H *—.
03 Q)
s
-u S
C
D
jj
-H X
a H
D c
0)
jj
rC
jj
CO

Q,
E
rC

jJ
C
rC
QJ
t-M



urnace type'3/
manufacturer
fa
"a! ^
sir

CQ






in
VD
(N



CN


M






g
O
V-l
CQ

CQ
J




in
H
U)



co


X
EH







X!
tn
-H
^_l
rti

Tang/ABB-CE |


CO

CQ
S






in
_«
•H -H
rC
LO r— |
3
^; u
u w
0 -H
u u
X3
fO li
CQ
U
1' CC
'T
Is u
<-8
CQ •-
tn
•- C
rH -H
rC SH
•H 0)
jJ 0)
C C
QJ -H
tn tn
C C
rC fa
f-H
C
II O
-H
tn jJ
C tn
rC 3
H "I
•- O
T3 U
0)
CO
o •
a
a
O -H
^_j
II QJ
a o
am
o
&












.
^-i
-H
rC
0
•H
I4H
^
CU
O
+
^1
0)
a
Vj
^j
12

X
C
^

3
O
n

II

*3l
fa
0

+

CQ
S


• -
^_i
QJ
C
^_j
3

|«(
O
X

5
O
tJ

II
PQ
2
,_5
U
                                               3-40

-------
emission levels that can be continuously achieved by these
boilers.  Each data set contained hourly-averaged values of
NOX emissions and boiler load that in most cases were measured
over a three-month operating period.
     In each data set, NOX emissions measured at boiler loads
lower than 30 percent of rated capacity, which included
startups and shutdowns of the unit, were excluded from the
analysis.  The remaining data were then averaged in 24-hour
blocks  (one for each day).
     The statistical analysis of each data set focused on the
24-hour block averaged NOX values calculated for each day with
18 or more hours of operating data.  The SAS15 PROC UNIVARIATE
procedure (with normal option) was used to calculate the mean
(X24),  standard deviation(824), and distributional form
(Shapiro-Wilk W-statistic,  skewness, and kurtosis) of the
24-hour block averages.  The distributional form of the data
was examined to determine if the data could be judged as being
normally distributed at the 95 percent confidence level.
Based or. this determination, different statistical procedures
were used for normal ar.d nor.-ncrmal data sets.  These
procedures are discussed in the following sections.
     3.2.3.1  Analysis of Normal Data Sets.  If the data were
determined to be normally distributed at the 95 percent
confidence level,  the autocorrelation structure of the data
was analyzed using the SAS~ PROC AUTOREG procedure to
determine the autocorrelation coefficients (p-±	p^) .
Previous analyses of time series emissions data were based on
an autoregressive model with lag 1  (AR[1]) .53-54-55-5tl  The analysis
presented here assumes a general AR model that includes the
effects of other statistically significant lags in addition to
lag 1.
     Because of the effects of autocorrelation,  the standard
deviation of the data as estimated by the SAS® PROC UNIVARIATE
procedure could be biased.   To correct for this bias, the
standard deviation for the 24-hr block averages was multiplied
                             3-41

-------
by a correction factor F]_.  This factor was calculated from

the following equation:57
                          Fl =
N - 1
N - C
(3-1)
where:
     N    =    no. of data points  (i.e., no. of 24-hr block
               averages in the data set)

     C    =    1 + 2[(N-l)/N]p1 + 2  [(N-2)/N]p2 +	+
               (2/N)pN-i

     P]_....PN-1 = expected autocorrelation  functions with  lags
                  1,....N-l

     The achievable emission limit for the  24-hr block

averaging time was estimated using the following equation:

                    Limit = x"   - z * 8   * F-              (3-2)
For the 7- and 30-day rolling averages, the emission limit was

calculated from:

                 Limit = x24  - z * s24  ~ FI * F2             (3-3;
where:
     X24  =    mean NOX value for the 24-hr block averaging
               period

     z    =    one-tailed standard normal deviate

     324  =    standard deviation for the 24-hr block
               averaging period

     FI   =    correction factor to correct for the bias  in
               the standard deviation  (calculated using
               equation 3-1)

     F2   =    adjustment factor for a general AR model
                (calculated using equation 3-4).
                              3-42

-------
     In equation 3-3, the adjustment factor ?2 for the general
AR model was calculated from the following equation:57
           F^ = - [n + 2 (n-l)p, -2 (n-2) p9 + . . . + 2^ ,]1/2     (3-4)
            
-------
emissions data were examined to determine if the NOX emissions
were correlated (either positively or negatively)  with load.
A regression equation was used to estimate the influence of
boiler load on the hourly-averaged NOX emission rate.   Next,
the residual NOX variation (i.e., the measured emission rate
minus the predicted emission rate based on the regression
equation) was calculated.  These residual values were then
added to a mean NOX value equal to the maximum NOX emission
rate predicted by the regression equation.  This new set of
hourly NOX emissions data were averaged into 24-hour blocks
and tested for normality.  If normal, the achievable emission
limits were calculated using equations 3-2 and 3-3.  For non-
normal data sets,  the 24-hr block-averages were transformed
and analyzed as explained under section 3.2.3.2.
     The following sections describe the results of the
statistical analysis performed on the CEM data from each of
the six representative boilers.  For each boiler,  a brief
description of the boiler is followed by a characterization of
the data set and the results of the statistical analysis.  The
data for each boiler reflects recent operating experience.  In
addition, for two of the boilers (AB Brown Unit No. 2 and Big
Bend Unit No. 4),  an additional data set reflecting previous
boiler operating performance was also available.  Analyses of
these two data sets are also presented.
     3.2.3.4  AB Brown Unit No. 2
     3.2.3.4.1  Boiler description.  AB Brown Unit No. 2 is a
265-MW, opposed-wall, bituminous coal-fired boiler located in
Mt. Vernon, Indiana.  The boiler is operated by Southern
Indiana Gas and Electric Company (SIGECO).  It is typically
base-loaded and in 1991 it was operated at an annual capacity
factor of 58 percent.  It is equipped with 24 Babcock and
Wilcox dual register burners.  The burners are arranged on two
opposite walls in three rows on each wall.  There are four
burners per row.  Each row of burners on each wall is serviced
by a single mill  (total of six mills).  No physical
modifications have been made since the boiler began operation
                             3-44

-------
in 1986.  The boiler does not have OFA ports.
     The boiler is subject to the NSPS, subpart Da NOX
emission limit for bituminous coal of 0.6 Ib/MMBtu.   A short-
term test conducted in 1992 indicated that the NOX emissions
averaged 0.39 Ib/MMBtu at a load of 242 MW.
     3.2.3.4.2  Data summary.  Continuous emission monitoring
data were obtained from the unit for two operating periods.
The first data set was collected between January 1 and
December 31, 1989.  The second data set was collected between
October 1 and December 30, 1992 and reflects more recent
operating experience.  Both data sets include hourly average
data on boiler load and NOX emission rate.
     The 1989 data set includes a total of 7,458 hourly-
averaged NOX emission values.  Time plots of these data for
each of the four quarters of 1989 are shown in figure 3-16.
Note that there are a number of gaps in this data set due to
outages of either the boiler or the CEM system.  The data show
higher hourly-averaged NOX emissions between December 5 and
Decerx.er 7.  Since nothing unusual was apparent in either the
load or C02 data during the period,  the increase is thought to
be due to an equipment malfunction.   Consequently,  these data
were not used in further analyses.  Typical hourly NOX
emission rates are between 0.40 and 0.60 Ib/MMBtu and average
approximately 0.50 lb/MM5tu.  As shown in figure 3-17,  the
duty cycle of the boiler on most days during the year
consisted of operation at near full load during the day
followed by operation at approximately 40 percent of full load
during the night.  Although some variation in the daily duty
cycle is apparent, there does not appear to be a significant
seasonal trend.  Figure 3-18 shows hourly-averaged NOX
emission values plotted as a function of load.  The graphs
suggest that over the normal operating load range,  NOX
emissions are independent of load.  However, there is
significant scatter in the data at a given load.
                             3-45

-------

•
•
•
\
*^e
^-
f
1?
DOOOOOOOOC
3OOOOOOOOC
(njgiMW/qi) XQN

Ji
!"
^
l|?
^
5-
j^
DOOOOOOOO*
68/>Z/9
68/^/9
68/01/9
68/C/9
68/I2/S
68/OZ/S $
ts
68/CI./S °
68/9/S
68/62/fr
68/22/fr
68/Sl/fr
68/8/V
I cc
3 C

68/92/e
68/6 UC
68/2 L/C
68/S/E
i- 68/92/2
68/61/2 «
68/2^/2 0
68/S/2
r 68/62/t
68/22/t
68/Sl/l
68/8/t
. en/ 1 / 1
* 68/1./1.
i §
_^L 68/IC/21.
^^T 68/fr2/2l
~^* i 68//l/2t
-^ t 68/01/21
^1 68/E/21-
^_ 68/92/1- 1
~^r 68/6J./U £
•^Z. 68/21/U °
3^- 68/9/U
"^ 68/62/01
68/22/01-
^ 68/St/OI-
J6"" | 68/8/0 1
JOOOOOOOOO
30OOOOOOOO

"^ * 68/OC/6
^ - 68/C2/6
•^ 	 J- 68/91/6
-4 	
-^» t 68/6/6
"^_ : 68/2/6
^ • 68/92/8
f: 68/6 1-/8 £
- 68/2 k/9 °
• 68/S/8
-^ - 68/62/Z
^ • 68/22/^
"^_ * 68/St/Z
^=1 7 68/8/Z
JOOOOOOOOO



0
CTi
CO
H
-
> H
m
C
(t ~
("3 (N
-( CTi
2 CO
O en
X! rH
H-l -
O rH
ro
0-)
O ^
r-t 0)
QJ 0)
E U
•H OJ
h 0
vo
1
n
0)
§,
•H
CM

o bdododo'd
                                 oooooooooo
                                3-46

-------
                                                                       : 68/lCftl.
                                  o
                                  o
        (MW)
                                              o    o
                                              o    o
                                              CO    CM
                                                (MW)
                                                                o
                                                                o
                                                                        69/01/31-
                                                                        68/Cftl.
                                                                        68/93/U
                                                                        68/6 l/U
                                                                        68ft k/U
                                                                        68/S/U
                                                                        68/62/01
                                                                        68/33/01-
                                                                        68/SlVOl.
                                                                        68/8/01-
                                                                        68/lVOt
                                                                                CO
                                                                                CTi
                                                                                H
                                                                                rO
                                                                                (U
                                                                                0)
                                                                                O
                                                                                0)
                                                                                O
                                                                                O
                                                                                Jj
                                                                                CO
                                                                                CTi
                                                                                H
8
o
s
I
o
o
                           68/61/3  *
                           68/31/3 O
                          | 68/S/3
                           68/63/1
                           68/33/t
                           68/SL/L
                           68/8/1
                           68/1-/1.
                                                                   68ft 1/8 0
        (MW)
                                                                                       03
                                                                                       O —
                                                                                       <-< fN
                                                                           3
                                                                           O 4J
                                                                          XI -H
                                                                             c
                                                                          iw D
                                                                           O
                                                                          JJ C
                                                                           O S
                                                                          r-H O
                                                                           a w
                                                                             PQ
                                                                           0)
                                                                           e ffl
                                                                          •H <
                                                                          H "
                                                                                  I
                                                                                 n
                                                                                 0)
                                                                                 l-i
                                                (AAW)
                                             3-47

-------
                              o>
                              I
                              0>

                      -  §


ooooooooo
b  b b b  b  b b b  b
ooooooooo
b  b b b  b  b b b  b
                                                                    CO


8
rr>
1


***v* **.
* *^bS5£ ~ ™
yj^^g^^-
X*=5^^r~
•
'

^ ££•'«»•*
K X






O
a
o 3L
«o
T- -0
I
o
o


o
10
^^
3OOOOOOOO
Oh*»CDIO^COCMvO



o
o
o

^:5%.


o
oo
o>
o
03
i
^
eg
CO





"SPfsfer
^j^^T^X ^.
^^*fe*^r —
^g^5^t fM *"

^5^^. M*
•vfe





1 !
*"' 0>
•o S.
ra 5j
o w
m^ I
§2?
3
~*




OOOOOOOOO
E —
QJ (N

X •
O 0
T3 4J
01 -H
cn c
n3 D
M
OJ >
%g
i O
XH
^4 ffl
V-i
§§
K —

CO
H
i
^
-H
OOOOOOOOO
               XON
ooooooooo
              'XON
                                                                     bu
                                     3-48

-------
     The 1992 data set consists of 1,860 observations of
hourly-averaged NOX emission values.  These data are plotted
in Figure 3-19.  There are no significant gaps in the data.
Typical NOX emission rates range from 0.30 to 0.50 Ib/MMBtu
and average approximately 0.40 Ib/MMBtu.  Figures 3-20 and
3-21 show the duty cycle of the boiler and NOX emission values
plotted as a function of load, respectively.  As seen in the
figures, both the duty cycle of the boiler and the N0x-load
relationship are similar to the 1989 data set.
     3.2.3.4.3  Statistical_analysis results.  Analysis of
each data set included assessing the implications for standard
setting resulting from autocorrelation between successive
measurements.  Because the NOX emissions are independent of
load, the effect of the N0x-load relationship on the
achievable NOX emission limits was not analyzed.
     1989 data set--Basic statistics for this data set are
summarized in table 3-8 for the 24-hour block averaging time.
As indicated in the table, the mean NOX level was
C.5C lb/MM3tu.  Based on the absolute values of the skewness
estimate of O.C16 and the kurtosis estimate of 2.51, the data
were judged to be normally distributed at the 95 percent
confidence level (for a sample size of 323, the 95 percent
confidence limit for skewness and kurtosis are approximately
0.22 and 2.60, respectively.)50  Although, the kurtosis
estimate is lower than the confidence limit, the assumption of
normality is conservative.  Based on an autocorrelation
analysis of the 24-hour block averaged data, an AR  (1,5) model
was used to evaluate the 7-,  and 30-day rolling average NOX
emission limits.
     Table 3-9 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3 and the statistical data in table 3-8.
As indicated by the table, achievable emission limits based on
one exceedance in 10 years range from 0.64 Ib/MMBtu for a
                             3-49

-------
                               26/13/31
                              26/2721
                          +  26/OC/U
                              26/C2/U
                              26/9i./U S
                                       re
                                       Q
                          —  26/6/U
                         —  26/2/14
                         T  26/92/01
                         -f-  26/61/01
                        +  26/21701.
                            26/SAH
                                              a\
                                              O 
   CTi
                                              O  -
                                             s: o
                                             14-1
                                             O
                                                0)
  (1)
  u
  0)
  Q
                                             O-i
                                             a)
                                             V-i
                                            tu
3-50

-------
8
CM
o
to
                                                       +  26/12/31
                                                       4-  26/n/zi
                                                       4-  26/OC/U
                                                           26/6/U
                                                        —  26/2/U
                                                            26/92/01
                                                        —  26/6 I/O t
                                                        -f  26/21/01.
                                                            26/S/OI.
                                                                        CN
                                                  O
                                                  n

                                                  M
                                                  

                                                        -  26/9l/U |  g

                                                                    0   ^
                                                                         O
                                                  •o
                                                  m
                                                  o
                                                                           (N
                                                  rH  O
                                                  W 2
                                                  3
                                                  O 4J
                                                  x: -H
                                                                        4J C
                                                                         O 3
                                                                        M O
                                                                         a>
                                                                           QQ
                                                                         i
                                                                         ro
01
•H
                  (MW)
                               3-51

-------
   xxx  xx x
X
                                         8
                                         n
                                         o
                                         ru
                                         o
                                         o

                                                   0)
                                                   CO

                                                   (0
                                                   Jj
                                                   (0
                                                   •o
                                                  m
                                                  H
                                                 -0
                                                 m
                                                 0
                                                  o

                                                  o
                                                 o
                                                 c
                                                 3
                                                 w
























3 C
3 I"
3 C
>
x'
^
*
,>


















,
1
- ° U
•a d c
' *fe<£x^Cx*f
^^^^xXx^X-^
'v v*>-X^ '^- \^*y*-j v
• •'i X. F<" * i •* *^v V •* >
y *^ ' -^ " >" '- ""*" • " •
X -",> "^ "x'A?xvAAv';'"' <>
^* X
X x


V












,
1
3 O C
DOC
•\ x
<.y^ x
~-
/'•>
*



















3 C
0 C
3 C
























3 C
^ T
3 C















—








3 C
- £
3 C
§ g
-1 o
_J
•^1
CO
•H
E —
0) (N

- S 0*0
73 4J
0) -H
cn c
ro D
0) -
> c
fO S
,' ^
- S ^m
y CQ
o <
w —
rH
(N
1
ro
0 0)
5 *2
3 &*
(niaww/Qi) XON
                                                 Cn
            3-52

-------
  TABLE 3-8.  SUMMARY STATISTICS FOR AB BROWN UNIT NO. 2
           (24-HOUR BLOCK AVERAGED NOX DATA)  [1989 DATA SET]

	Statistic	Value	
 No. of observations                            323
 Mean  (X24) [Ib/MMBtu]                            0.498
 Standard deviation  (s24)   [Ib/MMBtu]             0.041
 Skewnessa                                       -0.018  '
 Kurtosisa                                       2.507
 Shapiro-Wilk W statistic3                       0.977
 Probability < W                                 0.064
 Autocorrelation coefficients^3
             P!                                   0.642
             P5                                   0.063
 Correction factor (F^)                           1.008
 Adjustment factor (F2)
             7-day rolling  average                0.706
	50-day rolling average	C .432	
aFcr a perfectly normal distribution skewness =  0,
 kurtosis = 3,  and W statistic = 1.
konly  lags 1 and 5 are significant.
                            3-53

-------
  TABLE  3-9.   ACHIEVABLE NOX EMISSION LIMITSa FOR AB BROWN
                      UNIT NO. 2  (1989 DATA  SET)

period
24-hr Block
7 -day Rolling
30-day Rolling
Exceedance frequency
1%
0.595
0.566
0.540
I/year
0.613
0.580
0.548
1/10
0.
0.
0.
years
642
600
560
alb/MMBtu.
                            3-54

-------
24-hr block average to 0.56 Ib/MMBtu for a 30-day rolling
average.
     1992 data set--Basic statistics for the 1992 data set are
summarized in table 3-10 for the 24-hour block averaging time.
As indicated in the table,  the mean NOX level was
0.39 Ib/MMBtu with a standard deviation of 0.041 Ib/MMBtu.
Based on the absolute values of the skewness estimate of 0.075
and the kurtosis estimate of 2.39, the data were judged to, be
normally distributed at the 95 percent confidence level  (for a
sample size of 79, the 95 percent confidence limit for
skewness and kurtosis are approximately 0.43 and 2.28,
respectively) .5"  Based on an autocorrelation analysis of the
24-hour block averaged data, an AR(1) model was used to
evaluate the 7- and 30-day rolling average NOX emission
limits.
     Table 3-11 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3 and the statistical data in table 3-10.
J-.s indicated by the table,  achievable emission limits based or.
one exceedance frequency in 1C years range from C.53 lb/MM3tu
for a 24-hour block average to 0.43 Ib/MMBtu for a 30-day
rolling average.
     3.2.3.4.4  Conclusions.  Both the 1989 and 1992 data sets
show that the NOX emissions from the unit are independent of
load.  The NOX emissions reflected in the 1992 data set are
lower than in the 1989 data set.  Plant personnel60 indicated
that this may be due to the better control of boiler operating
parameters in 1992 than in 1989.
     3.2.3.5  Big Bend Unit No. 4
     3.2.3.5.1  Boiler description.  Big Bend, Unit No. 4, is
a 455 MW tangentially-fired boiler burning eastern bituminous
 (Illinois) coal.  The boiler is operated by Tampa Electric
Company  (TECO).  It is typically base-loaded and operated at
an annual capacity factor of 71.2 percent in 1991.  It is
equipped with a total of 20 burners  (five at each corner) and
                              3-55

-------
 TABLE 3-10.  SUMMARY  STATISTICS  FOR AB BROWN UNIT NO.  2
          (24-HOUR BLOCK AVERAGED  NOX DATA)  [1992 DATA SET]

              Statistic	Value	
No. of observations                             79
Mean  (X24)  [Ib/MMBtu]                            0.388
Standard deviation  (824)   [Ib/MMBtu]              0.041
Skewnessa                                       -0.075
Kurtosisa                                        2.389
Shapiro-Wilk W statistic3                        0.977
Probability < W                                  0.467
Autocorrelation coefficient53  (PD                0.521
Correction factor (F]_)                            1.003
Adjustment factors  (F2)
            7-day rolling  average                0.602
            30-day rolling average               0.318
For a perfectly normal distribution  skewness  =  0,
kurtcsis = 3, and W  statistic  = 1.
   y  lag 1 is significant.
                           3-56

-------
      TABLE 3-11. ACHIEVABLE NOX EMISSION LIMITSa FOR
                  AB BROWN UNIT NO. 2  (1992 DATA SET)
                  	Exceedance Frequency	
 Averaging Period	1%	I/year	1/10 years
    24-hr Block        0.483        0.501         0.529
   7-day Rolling       0.445        0.456         0.473 '
  30-day Rolling	0.418	0.424	0.433
alb/MMBtu.
                           3-57

-------
8 OFA ports (two each located immediately above the top burner
in each corner).   The burners were manufactured by Combustion
Engineering.  The OFA ports share the same windbox as the
burners and can be manually tilted.  However,  unlike current
state-of-the-art designs,  the OFA ports cannot be yawed to
offset the air away from the primary combustion zone.
Typically, 20 percent of the total combustion air is
introduced as overfire air.  No physical modifications hav,e
been made to the boiler since it came on-line in 1985.
     Plant personnel indicated that the unit can typically
maintain full load when four of the five burners located in
each corner are in service.61  One burner at each corner is
usually rotated out of service for routine maintenance.  There
is some variation in NOX emission rate depending in which
burners are in service.  Additionally, the NOX emission rate
could increase if the burner tile is not optimized.
     The boiler is currently subject to the NSPS, subpart Da
with a NOX emissions limit of 0.6 Ib/MMBtu.  A short-term test
conducted in 1991 indicated that the NOX emissions averaged
C.41 Ib/MMBtu at a load cf 435 KW.  Plant personnel indicated
that they do not follow any special boiler operation
procedures other than continuously monitoring emissions to
ensure compliance within the regulatory limits.
     3.2.3.5.2  Data summary.  Continuous emission monitoring
data were obtained from the unit for two operating periods.
The first data set was collected between April 23 to
August 19, 1985 and reflects boiler operation fairly soon
after startup of the unit.  The second data set was collected
between July 1 to September 30, 1992 and reflects more recent
operating experience.  Both data sets include hourly average
data on boiler load and NOX emission rate.
     The 1985 data set includes a total of 1,949 hourly-
averaged NOX emission values.  A time plot of these data are
shown in figure 3-22.  Note that there are a number of gaps in
this data set due to outages of either the boiler or the CEM
                             3-58

-------
                                                   o
                                                   4J
                                                   CM

                                                   •-I
                                                   •H
                                                   W

                                                   §•
                                                   O
                                                   tw

                                                   to
                                                   C
                                                   O
                                                   •H
                                                   CO
                                                   m
                                                   •H
                                                   e
                                                   (1)
                                                   o c
                                                   2 P

                                                   •o  -
                                                   OJ T>
                                                   tnc
                                                   m QJ
                                                   l-i CQ
                                                   0)
                                                   > en
                                                   (8 -H
                                                    i  CQ
                                                    O IT)

                                                   £ CO
                                                    4-) CTi

                                                    O r-l
                                                    0)  3
                                                    E  01
                                                   •H  3
                                                   H <
                                                    (N

                                                    (N
                                                    I
                                                    (1)
                                                    i-i
XON
  3-59

-------
system.  Typical hourly NOX emission rates are between 0.25
and 0.70 lb/MMBtu and average approximately 0.48 Ib/MMBtu.  As
shown in figure 3-23, the duty cycle of the boiler on most
days consisted of operating at near full load during the day
and at a reduced load (about 75 percent of rated capacity) for
approximately 6 hours each night.  Figure 3-24 shows hourly-
averaged NOX emission values plotted as a function of load.
No apparent relationship exists between these two parameters.
     The 1992 data sets consists of 1,991 observations of
hourly-averaged NOX emission values.  These data are plotted
in figure 3-25.  Typical hourly NOX emission rates range from
0.20 to 0.55 lb/MMBtu and average approximately 0.40 lb/MMBtu.
Figure 3-26 shows the duty cycle the boiler which is similar
to the duty cycle reflected in the 1985 data set.  Figure 3-27
shows the relationship between the hourly-averaged NOX and
load data.   The graph indicates that at loads of 275 MW
(60 percent of design) and grearer, there is no clear
relationship between NOX and load.  At loads less than 275 MW;
however, there appears tc be a positive correlation between
NCX and load (i.e.,  NOX decreases with decreasing load,. .
These reduced load and NOX emission levels generally
correspond to a nightly load reduction between approximately
midnight and 6:00 a.m.
     Based upon an initial review of the 1992 data set and
discussions with plant personnel61, it was concluded that this
data set is representative of normal operating conditions for
this boiler.
     3.2.3.5.3  Statistical analysis results.  Analysis of
each data set included assessing the implications for standard
setting resulting from autocorrelation between successive
measurements.  For the 1985 data set, because the NOX
emissions are independent of load, the effect of the N0x-load
relationship on the achievable NOX emission levels was not
analyzed.  For the 1992 data set; however, the NOX emissions
are independent of load only for loads greater than about
                             3-60

-------
                                              in
                                              oo
                                             o\
                                              4J

                                              W
                                              O

                                              4J


                                              CO

                                              (N
                                              J-l

                                              O
                                             TD
                                              (0
                                              O
                                                O

                                                2
                                             o n
                                             iH (1)
                                             acn

                                             a; en
                                             E -H
                                             •H DQ
                                             fS
                                              I
                                             0)
                                             •H
                                             CL,
3-61

-------
»,  » **'  #    *1*J **
 *              \  *'   *
              * )
     o
     
O
CO
                                 -8
                                                           -8
                                  o
                                 -in
                                  n
                                                                8
                                  o
                                  • in
                                                           -S
                                  o
                                 -to
O'


O
                                                                         4J
                                                                         0)
                                                                         (0

                                                                         (C
                                                                         4J
                                                                         m
                                                                         •a

                                                                         in
                                                                         oo
                                                                        •a
                                                                         <0
                                                                         O
                                                                         O
                                                                        •H
                                                                         iJ
                                                                         O
                                                                         c
                                               (T5

                                               CO
                                               m

                                               CO
                                               c
                                               o
                                              •H
                                               to
                                               CO
                                    0)  -a
                                    (T3 C
                                     i  0)
                                                                         ^ en
                                                                         3-H
                                                                         o aa
                                                                         (N
                                                                          I
 0)

 I
-H
                          3-62

-------
                                                           _
                                                            CM JO
                                                            CM s:
                                                            CM

                                                          ^§
                                                            CM
                                                               0)
                                                               in
                                                            CM
                                                              O
                                                             •£<
                                                              w *
                                                             •g*
                                                              oh
                                                          	  Si    o  .
                                                                       c
                                                                     l£
                                                              >s
                                                              m
                                                               00
o
O)
o
00
to     in
od
^;
d
o
CO
               o
               CM
                                                               CM
                                                            CM
                                                            N: c^J
                                                            ^ o
                                                            ~  CO
                                                            CM O
                                                            S> ^
o
i—
o


                                                                     o

                                                                     4J
                                                                     o
                                                                       0)
                                                                     in
                                                                     CM
                                                                     0)
                                                                     >-i
                                                                     3
                                                                     Cn
                                                                     •H
                                                                     Du
                                     3-63

-------
                                                               «N
                                                               o
                                                               m
                                                               0)

                                                               I
                                                               .u
                                                               0)

                                                               o
                                                               l-i
                                                               O
                                                              •o
                                                               (B
                                                               O
                                                                 •o
                                                                 c
                                                                 0)
                                                                4CQ

                                                                 Ol
in
8      8       8      §
                 CM      CM
                                                              U)
                                                              OJ
                                                               I
 Q)


 i
•H
Ct4
               3-64

-------
                                                m

                                                to
                                                oj
                                                (0
                                                •O
                                               Ch
                                               o\
                                               •o
                                                10
                                                o
                                                C
                                                o
                                               •H
                                               Jj
                                                u
                                                C

                                               14-1

                                                m

                                                CO
                                                (0

                                                to

                                                o
                                               •H
                                               10
                                               CQ
                                               -H
                                               E.
                                               01
                                               tt)
                                               D
                                               i
                                               ro

                                               
-------
275 MW.  At loads below 275 MW,  the NOX emissions are
positively correlated with load (i.e.,  NOX emissions decrease
with decreasing load).   Consequently,  this data set was
examined in two ways.  The first analysis examined all of the
data, while the second analysis examined a subset of this data
which only included data collected at  an operating load of
275 MW and greater.
     1985 data set--Basic statistics for the 1985 data set; are
summarized in table 3-12 for the 24-hour block averaging time.
As indicated in the table, the mean NOX level was
0.48 Ib/MMBtu.  Based on the absolute  values of the skewness
estimate of 0.195 and the kurtosis estimate of 2.61, the data
were judged to be normally distributed at the 95 percent
confidence level (for a sample size of 82, the 95 percent
confidence limit for skewness and kurtosis are approximately
0.43 and 2.29, respectively).50  Based  on an autocorrelation
analysis of the 24-hour block averaged data, an AR(1,3) model
was used to estimate the 7- and 30-day rolling average NOX
emission limits.
     Table 3-13 shows the calculated NCX emission limits as a
function of averaging time and exceedance frequency based or
equations 3-2 and 3-3 and the statistical data in table 3-12.
As indicated by the table, achievable  emission limits based on
one exceedance in 10 years range from  0.67 Ib/MMBtu for a
24-hr block average to 0.60 Ib/MMBtu for a 30-day rolling
average.
     1992 data set--Basic statistics for the entire 1992 data
set are summarized in table 3-14 for the 24-hour block
averaging time.  As indicated in the table, the mean NOX level
was 0.40 Ib/MMBtu with a standard deviation of 0.025 Ib/MMBtu.
Based on the absolute values of the skewness estimate of 0.085
and the kurtosis estimate of 2.60, the data were judged to be
normally distributed at the 95 percent confidence level  (for a
sample size of 82, the 95 percent confidence limit for
skewness and kurtosis are approximately 0.43 and 2.29,
                             3-66

-------
  TABLE 3 - 12 .  SUMMARY STATISTICS  FOR BIG BEND UNIT NO .  4
          (24 -HOUR BLOCK AVERAGED  NOX DATA)  [1985 DATA  SET]

_ Statistic _ Value _
 No.  of  observations                              82
 Mean (X24)  [Ib/MMBtu]                             0.479
 Standard  deviation (824)  [Ib/MMBtu]               0.051
 Skewnessa                                        0.195
 Kurtosisa                                        2.608
 Shapiro-Wilk W statistic3                         0.966
 Probability < W                                  0.107
 Autocorrelation  coefficients^3
             Pl                                   0 . 591
             P3                                   0.250
 Correction  factor  (F]_)                            1.092
 Adjustment  factor  (F2)
7
i
- dav
0 - day
f~cl
ro
-\
]_
" "^ 3"
ling
average
average
C.
c.
8
6
-^
£
o
aFor a perfectly normal distribution skewness = 0,
 kurtosis = 3, and W statistic = 1.
    y lags 1 and 3 are significant.
                            3-67

-------
 TABLE 3-13.  ACHIEVABLE NOx EMISSION LIMITS3- FOR  BIG  BEND
                      UNIT NO. 4  (1985 DATA SET)

Aver
Pei
24-hr
7 -day
3 0 - day

riod
Block
Rolling
Rolling
Exceedance Frequency

0
0
0
1%
.609
.587
.558
I/year
0
0
0
.635
.608
.574
1/10
0.
0.
0.
years
673
639
597
r
alb/MMBtu.
                            3-68

-------
  TABLE 3 -14.  SUMMARY STATISTICS FOR BIG BEND UNIT NO. 4
          (24-HOUR BLOCK AVERAGED NOX DATA)  [1992 DATA SET]

	Statistic	Value	
 No.  of observations                             82
 Mean (X24)   [lb/MMBtu]                            0.398
 Standard deviation (s24)  [Ib/MMBtu]             0.025
 Skewnessa                                       0.085
 Kurtosisa                                       2.595
 Shapiro-Wilk W statistica                       0.979
 Probability < W                                 0.521
 Autocorrelation coefficients*3
             Pi                                  0.286
             P5                                  -0.255
 Correction  factor (F]_)                           1.000
 Adjustment  factor (F2)
             7-day rolling average                0.444
             30-day rolling  average               0.181
aFor a perfectly normal distribution skewness =  0,
 kurtosis =  3,  and W statistic = I.
      lags 1 and 5 are significant.
                            3-69

-------
respectively) .5"  Based on an  autocorrelation analysis of the
24-hour block averaged data an AR(1,5)  model was used to
estimate the 7- and 30-day rolling average NOX emission
limits.
     Table 3-15 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3 and the statistical data in table 3-14.
As indicated by the table, achievable emission limits base,d on
one exceedance in 10 years range from 0.50 Ib/MMBtu for a
24-hour block average to 0.42 Ib/MMBtu for a 30-day rolling
average.
     1992 data subset--Basic statistics for the 1992 data
subset (i.e.,  for loads > 275 MW)  are summarized in table 3-16
for the 24-hour block averaging time.  As indicated in the
first data column of the table,  the mean NOX level was
0.41 Ib/MMBtu with a standard deviation of 0.022 Ib/MMBtu.
Based on the absolute values of the skewness estimate of
0.752, the data were judged as not being normally distributed
at the 95 percent confidence level (for a sample size of 70,
the 95 percent confidence limit for skewness and kurtosis are
approximately 0.46 and 3.90, respectively).50  To compensate
for skewness,  the data were transformed using a Box-Cox
transform with a A of -5.25.  The statistical properties of
the transformed data are presented in the second data column
of table 3-16.   As shown in the table,  the absolute values of
skewness and kurtosis for the transformed data are within the
95 percent confidence limits for a normal distribution.  Based
on an autocorrelation analysis of the transformed data, an
AR(1,2) model was used to estimate the 7- and 30-day rolling
average NOX emission limits.
     Table 3-17 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-16 and the
statistical procedures described in section 3.2.3.2.  As
indicated in table 3-17, achievable emission limits based on
                             3-70

-------
      TABLE 3-15. ACHIEVABLE NOX EMISSION LIMITSa FOR
                  BIG BEND UNIT NO. 4  (1992 DATA SET)
                  	Exceedance Frequency	
 Averaging Period	1%	I/year	1/10 years
    24-hr Block        0.464        0.477         0.496
   7-day Rolling       0.427        0.433         0.442 ,
  30-day Rolling	0.410	0.412	0.416
alb/MMBtU.
                            3-71

-------
  TABLE 3-16.
SUMMARY STATISTICS FOR BIG BEND UNIT NO. 4
    (24-HOUR BLOCK AVERAGED NOX DATA)
           [1992 DATA SUBSET]
Statistic
No. of observations
Mean (X24> [lb/MMBtu]
Standard deviation (824) [lb/MMBtu]
Skewnessa
Kurtosisa
Shapiro-Wilk W statistica
Probability < W
Autocorrelation coefficients*3
Pi
P2
Correction factor (F]_)
Transformed
As measured data
data (X=-5.25)
70
0.406
0.022
0.752
3.072
0.939
0.003

NEC
NEC
NEC
70
'117
30
-0
2
0
0

0
-0
1

.97
.40
.048
.511
. 971
.289

.634
.169
.008
aFor a perfectly normal distribution skewness
 kurtosis = 3, and W statistic = 1.

bOnly lags 1 and 2 are significant.

CNE =  not estimated.
                               = C,
                            3-72

-------
      TABLE 3-17. ACHIEVABLE NOX EMISSION LIMITS3- FOR
                 BIG BEND UNIT NO.  4 (1992  DATA SUBSET)
                             Exceedance Frequency
 Averaging Period	1%	I/year	1/10 years
    24-hr Block        0.481        0.514         0.604
   7-day Rolling       0.444        0.457         0.486 •
  30-day Rolling	0.424	0.429	0.439
alb/MMBtu.
                           3-73

-------
one exceedance in 10 years range from 0.60 Ib/MMBtu for a
24-hour block average to 0.44 lb/MMBtu for a 30-day rolling
average.
     3.2.3.5.4  Conclusions.  The 1985 data set shows that the
NOX emissions from the unit did not vary with load.  However,
the 1992 data set indicated that NOX emission levels decreased
with decreasing loads for loads less than 275 MW.  To study
the effect on the achievable NOX emission limits of boiler
operation at loads corresponding to the maximum NOX emission
rates, a subset of the 1992 data set which included loads
^275 MW was analyzed.
     The 1985 data set,  collected soon after startup of the
unit,  showed both a higher mean NOX emission rate and greater
variability than in 1992.  Based on discussions with the plant
engineer"1, the lower NOX emissions reflected in the 1992 data
may be due to the plant  personnel being more familiar with the
operation of the unit and consequently having better control
over the boiler operating parameters now than when the unit
came on-line in 1985.
     3.2.3.6  Hoi comb Ur.it No. 1
     3.2.3.6.1  Boiler description.  Holcomb Unit No. 1 is a
348-MW, opposed-wall, subbituminous coal-fired boiler located
in Holcomb, Kansas.  The boiler is operated by Sunflower
Electric Power Corporation.  It is typically base-loaded and
in 1991 operated at an annual capacity factor of 73.4 percent.
It is equipped with 25 Babcock and Wilcox dual register
burners.   The burners are arranged on two opposite walls in
three rows on the front wall and two rows on the rear wall.
There are five burners per row.  Each row of burners on each
wall is serviced by a single mill  (total of five mills).  No
physical modifications have been made since the boiler began
operation in 1983.  The boiler does not have overfire air
ports.  During an outage in 1990, the coal nozzles and the
burner nozzle tips were replaced.
                             3-74

-------
     The boiler is subject to the NSPS, subpart Da NOX
emission limit for subbituminous coal of 0.5 Ib/MMBtu.
Compliance tests  (short-term) conducted in 1984 indicated that
the NOX emissions averaged 0.17 Ib/MMBtu.  Additionally, the
results of long-term measurements made to demonstrate initial
compliance indicated a 30-day average of NOX emissions at
0.23 Ib/MMBtu.
     3.2.3.6.2  Data summary.  Continuous emission monitoring
data were obtained from Sunflower Electric Power Corporation
for the operating period between October 1 and December 30,
1992.  This data set includes a total of 2,018 hourly-averaged
NOX emission and load values.  A time plot of the NOX emission
data is shown in figure 3-28.  Typical hourly NOX emission
rates are between 0.20 and 0.40 Ib/MMBtu and average
approximately 0.30 Ib/MMBtu.  As shown in figure 3-29, the
duty cycle of the boiler on most days during the year
consisted of operation at full load  (348 MW) during the day
followed by operation at approximately 60 percent of  full load
•'2CC MW} during the night.  Figure 3-3C shows hourly-averaged
NOX emission values plotted as a function of load.  The graph
shows a positive correlation between these quantities.
Specifically, as the load increases from about 200 MW
(60 percent of design)  to the unit's design load of 348 MW,
hourly NOX emission values increase from an average of
approximately 0.20 to 0.35 Ib/MMBtu.  However, there  is
significant scatter in the data at a given load.  The reduced
NOX emission values generally correspond to a nightly load
reduction as shown in Figure 3-29.
     3.2.3.6.3  Statistical analysis results.  Analysis of the
data set included assessing the implications for standard
setting resulting from autocorrelation between successive
measurements.  In addition, the effect of the N0x-load
relationship on achievable NOX emission limits was analyzed.
                             3-75

-------
                                                                                     CN
                                                                                     O\
                                                                                    o
                                                                                    m
                                                                                     0)
                                                                                     U
                                                                                     0)
                                                                                     Q

                                                                                     O
                                                                                     4J
                                                                                     0)
                                                                                    J3
                                                                                     O
                                                                                    4.)
                                                                                     U
                                                                                    O
                                                                                     en

                                                                                     O
                                                                                    •H
                                                                                     CO
                                                                                     CO
                                                                                    •H

                                                                                     0)
                                                                                    o
                                                                                    JS
ooooododoooooooooo
                                                                                    ex o
                                                                                       u
                                                                                    0) rH
                                                                                    £  O
00
(N
 I
                                        XQN
 &
•H
fc
                                          3-76

-------
8L
I
                                               a\
                                               en
                                               H
                                               (U
        CVJ    r-    r-    i-
                                               o
                                               
          -H
          (X4
3-77

-------






























5 L




**ll
: 3K *
lx



*



















1 C




1)1
II
*l
p
*
* X
* 5K *

X»
I *l
*




^












D 1




IK
I
i
I I
W ^
fill
1C
i
i
i* i

)**•?'
K!
> *














f) C

«


Hi.'
IISx!

1
^
i

i
I*:
H|
9
1

I
'
* 1 *
>
11
•i|














3 I





*
*
11 I
^ ^
, |
* 3
: * *
i * ;
*. 3f
||

i*






7* .

*




0 C







*
*

*:
3K
p
i * ¥
V
. | * * *
ijjr
* * s













^ i















:














•) C






























D I






























D C
Zf
U
O
aa
-i I
1-1
' |ii
o
o
•H
-| s
J3
IW
re
CO
?:
-8^|
'""S »
o »
-1 E
0)
X
o
§ ^
"w g)
re
(U ^~
> r-l
re
1
is 0
-H Z,
k-i
0 3 jj
-m O-H
"" W D
o
ro
i

-------
     Basic statistics for this data set are summarized in the
first data column of table 3-18 for the 24-hour block
averaging time.  As indicated in the table,  the mean NOX level
was 0.29 Ib/MMBtu.  Based on the absolute values of the
skewness estimate of 0.522,  the data were judged as not being
normally distributed at the 95 percent confidence level (for a
sample size of 86, the 95 percent confidence limit for
skewness and kurtosis are approximately 0.42 and 3.84,
respectively).59  To compensate for skewness, the data were
transformed using a Box-Cox transform with a X=2.25.  The
statistical properties of the transformed data are presented
in the second data column of table 3-18.  As shown in the
table, the absolute values of skewness and kurtosis are 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 3-19 shows the estimated NOX emission limits as a
function of averaging time ar.d exceedance frequency based on
the statistics of the transformed data in table 3-18 and the
statistical procedures described in section 3.2.3.2.  As
indicated in table 3-19, achievable emission limits based on
one exceedance in 10 years range from 0.40 Ib/MMBtu for a
24-hr block average to 0.34 Ib/MMBtu for a 30-day rolling
average.
     Effect of load on achievable NOX emission limits - - The
results presented in the previous section provides an
assessment of NOX emission rates for the boiler load
characteristics experienced during October to December 1992.
As shown in figure 3-30; however, there is a general increase
in NOX emission rates as load increases.  Consequently, the
NOX emission rate from the boiler will be higher if the boiler
is operated at near full load for a longer period of time than
was reflected by the data for this operating period.
                             3-79

-------
  TABLE 3-18.  SUMMARY STATISTICS FOR HOLCOMB UNIT NO. 1
                   (24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean (x24) [Ib/MMBtu]
Standard deviation (824) [Ib/MMBtu]
Skewnessa
Kurtosisa
Shapiro-Wilk W statistic3
Probability < W
Autocorrelation coefficient13 (PD
Correction factor (F^)
As measured
data
86
0
0
-0
3
0
0
NEC
NEC

.294
.038
.522
.464
.970
.186


Transformed
data
(A-2.25)
86
0
0
-0
2
0
0
0
1

.065
.018
.035
.698
.981
.626
.537
.034
aFor a perfectly normal distribution skewness = 0,
 kurtosis = 3, and W = 1.
    y lag 1 is significant.

 NE = no~ estimated.
                            3-80

-------
     TABLE 3-19   ACHIEVABLE NOX EMISSION LIMITS* FOR
     TABLE 3 iy-  A       HOLCOMB UNIT NO.  1

Aver
Pei
24-hr
7 -day
3 0 - day

riod
Block
Rolling
Rolling
Exceedance Frequency

0
0
0
1%
.373
.345
.323
I/year
0
0
0
.385
.354
.328
1/10
0.
0.
0.
years
404
366
335
9
aib/MMBtu.
                           3-81

-------
     To assess the potential NOX emission rate when operating
the boiler at other duty cycles, linear regression was first
used to estimate the influence of boiler load on the hourly-
averaged NOX emission rate.  The resulting regression equation
was:
                  NOX = 0.00089  * MW + 0.04210             (3-5)

where:
     NOX  =    NOX emission rate in Ib/MMBtu;  and
     MW   =    boiler load in MW.

The as-measured hourly-averaged NOX emission rate data were
then adjusted based on this regression equation to estimate
the variation in the NOX emission rate that is not explained
by changes in boiler load.  The residual variation (i.e., the
measured emission rate minus the predicted emission rate based
on equation 3-5) ranged from -0.14 to 0.17 Ib/MMBtu.   These
residual values were then added to a mean value equal to the
maximum hourly NOX emission rate predicted by equation 3-5.
(As shown in figure 3-30, the highest expected NOX emission
rate corresponds to the design operating load of 348 MW).  The
new set of load-adjusted hourly NOX emissions data were then
averaged into 24-hr blocks and tested for normality.
     The results of the statistical analysis on the load-
adjusted data set are presented in table 3-20.  Based on the?
absolute values of the skewness estimate of 0.18 and kurtosis
of 2.84, the data were judged as being normally distributed at
the 95 percent confidence level  (for a sample size of 86, the
95 percent confidence limit for skewness and kurtosis are
approximately 0.42 and 3.84, respectively).5"  Based on the
autocorrelation analysis of the data, an AR(1,3) model was
used to estimate the 7- and 30-day rolling average NOX
emission limits.
                             3-82

-------
   TABLE 3-20.   SUMMARY STATISTICS FOR HOLCOMB UNIT NO.  1
            (LOAD-ADJUSTED 24-HOUR BLOCK AVERAGED NOX DATA)

	Statistic	Value
 No.  of  observations                                86
 Mean (X24)  [Ib/MMBtu]                               0.352
 Standard deviation  (824)  [Ib/MMBtu]                 0.029
 Skewnessa                                         -0.182
 Kurtosisa                                          2.844
 Shapiro-Wilk W statistica                           0.973
 Probability <  W                                    0.278
 Autocorrelation coefficientsb
             Pl                                     0.604
             p3                                     0.241
 Correction  factor (F^                              1.090
 Adjustment  factor (F2)
             7 - day rolling  average                  0.830
	30-day  rolling  average	0.615
aFor a perfectly normal distribution, skewness = 0,
 kurtosis = 3,  and W statistic = 1.
      lags 1 and 3 are significant.
                            3-83

-------
     Table 3-21 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3,  and on the statistical data in
table 3-20.  As indicated in table 3-21,  the NOX emission rate
corresponding to a 30-day rolling averaging period and a one
in 10 years exceedance frequency is 0.42  Ib/MMBtu.  As
expected, this value is higher than the corresponding emission
rate based on the actual operating load cycle for the unit,
presented in table 3-19, of 0.34 Ib/MMBtu.
     3.2.3.6.4  Conclusions.  As indicated  in table 3-19,
achievable NOX emission levels calculated from the Holcomb
data for October to December, 1992 and a one-in-ten-years
exceedance frequency, range from 0.40 Ib/MMBtu based on a 24-
hour block averaging period to 0.34 Ib/MMBtu based on a 30-day
rolling averaging period.  Because of the relationship between
NOX emission rate and load; however, higher NOX emission
levels are expected to occur during periods of sustained
operation at high load levels.  As shown in table 3-21,
achievable NOX emission levels based on continuous operation
of the boiler at full load  (348 MW) and a one-in-ten-years
exceedance frequency are estimated at 0.46  Ib/MMBtu for a 24-
hour block average and 0.42 Ib/MMBtu for a  30-day rolling
average.
     3.2.3.7  WA Parish Unit No. 8
     3.2.3.7.1  Boiler description.  WA Parish Unit No. 8 is a
615-MW, subbituminous coal, tangentially-fired boiler located
in Thompsons, Texas.  The boiler is operated by Houston
Lighting and Power Company  (HL&P).  The boiler is typically
base-loaded and in 1991, it was operated at an annual capacity
factor of 72 percent.  In 1987, the boiler  was retrofitted
with 24  (6 at each corner) Combustion Engineering concentric-
fired tilting tangential burners.  At each  corner, two
overfire  (OFA) air ports are located immediately above the top
burner.  The OFA ports share the same windbox as the burners
and can be manually tilted.  However, unlike current state-of-
                              3-84

-------
  TABLE 3-21. ACHIEVABLE NOx EMISSION LIMITS51 FOR HOLCOMB
              UNIT NO.  1 BASED ON CONTINUOUS OPERATION AT
            LOAD CORRESPONDING TO MAXIMUM NOX EMISSION RATE


                   	Exceedance Frequency	

 Averaging  Period	1%	I/year	1/10  years

    24-hr Block        0.425         0.439         0.461
                                                        v
   7-day Rolling       0.413         0.425         0.442

  30-day Rolling	0.397	0.406	0.419

alb/MMBtu.
                           3-85

-------
the-art designs, the OFA ports cannot be yawed to offset the
air away from the primary combustion zone.  Typically, about
15 percent of the combustion air is introduced as OFA.  Each
row of burners is serviced by a single mill (total of six
mills).  No physical modifications (other than the burner
retrofit) have been made since the boiler began operation in
1982.
     The boiler is subject to the NSPS, subpart Da NOX
emission limit for subbituminous coal of 0.5 Ib/MMBtu.  A
short-term test conducted in 1991 indicated that the NOX
emissions averaged 0.34 Ib/MMBtu at a load of 604 MW.
     3.2.3.7.2  Data summary.  Continuous emission monitoring
(CEM) data were obtained from HL&P for the operating period of
June 1 through July 31, 1992.  This data set includes a total
of 1,371 hourly-averaged NOX emission and load values.  A time
plot of the NOX emissions data is shown in figure 3-31.  Note
that there are several gaps in this data set due to outages of
either the boiler or the CEM system.   Typical hourly NOX
emission rates are between 0.20 and 0.30 Ib/MMBtu, and average
approximately 0.25 Ib/MMBtu.  As shown in figure 3-32, the
duty cycle of the boiler on most days consisted of operating
at above 90 percent of full load during the day followed by
operation at approximately 50-75 percent of full load during
the night.  Figure 3-33 shows hourly-averaged NOX emission
values plotted as a function of load.  The graphs suggest that
over the normal operating load range, NOX emissions increase
with load.  However, there is significant scatter in the data
at a given load.
     3.2.3.7.3  Statistical analysis results.   Analysis of the
data set included assessing the implications for standard
setting resulting from autocorrelation between successive NOX
measurements.  In addition, the effect of the N0x-load
relationship on achievable NOX emission limits was analyzed.
     Basic statistics for this data set are summarized in
table 3-22 for the 24-hour block averaging time.  As indicated
in the first data column of the table, the mean NOX level was
                             3-86

-------
                                       (U
                                       c
                                        CO
                                        c
                                        o
                                       •H
                                        CQ
                                        CO
                                       •H

                                        0}
                                        T-l
                                        ro
                                         i
3-87

-------
   t
(O
O
in
IT)
U)
                   in

CO
s
CM
                                                            CM
                                                              8
                                                          -co
                                                            CM
                                                                 „.
                                                              CM
                                                              O>
                                                          — cvT
                                                              CM
                                                            s
                                                              to
                                                            to
                                                                01

                                                                       0)
                                                                       e
                                                           I
                                                          ro


                                                          0)
                                                          M
                    (MW)
                                  3-88

-------
in
•«*

d
                                  .
                            **  **
                                    *: *
co      co
6      d
in
cvj

o
                             ** f
                                          * * *
        in
eg      *-
6      6
in
q
o

                                                           o
                                                           m
                                                           co
                                                          .8
                                                                          I
                                                                                    •o
                                                                                    (0
                                                                                    o
                                                                                    O

                                                                                    •H

                                                                                    AJ

                                                                                    U
                                                                    (t


                                                                    (0

                                                                    (6


                                                                    CO

                                                                    C

                                                                    O

                                                                   •H

                                                                    CO
                                                                                    0)
                                                                                    en.
                                                                                    to '
                                                                                    n
                                                                                    0)
                                                                                       (T3
                                                                                    ro
                                                                                    m
                                                                                     i

                                                                                    m


                                                                                    0)
                                                                                    •H

                                                                                    [14
                                          *ON
                                      3-89

-------
  TABLE  3-22.   SUMMARY STATISTICS FOR WA PARISH UNIT NO. 8
                   (24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean (£24) [Ib/MMBtu]
Standard deviation (824) [Ib/MMBtu]
Skewnessa
Kurtosisa
Shapiro-Wilk W statistic3
Probability < W
Autocorrelation coefficient*3 (PI)
Correction factor (F]_)
As measured
data
56
0
0
-0
3
0
0
NEC
NEC

.260
.026
.618
.837
.971
.347


Transformed
data
(X=2.75)
56
0
0
-0
3
0
0
0
1

.025
.007
.033
.060
.988
.953
.457
.015
aFor a perfectly normal distribution skewness = 0,
 kurtosis = 3, and W statistic = 1.

bOnly lag 1 is significant.

CNE = not estimated.
                            3-90

-------
0.26 Ib/MMBtu.  Based on the absolute value of the skewness
estimate of 0.618, the data were judged as not being normally
distributed at the 95 percent confidence level (for a sample
size of 56, the 95 percent confidence limits for skewness and
kurtosis are approximately 0.51 and 3.96,  respectively)59  To
compensate for skewness, the data were transformed using a
Box-Cox transform with a X of 2.75.  The statistical
properties of the transformed data are also presented in t;he
second data column of table 3-22.  As shown in the table, the
absolute values of skewness and kurtosis for the transformed
data are within the 95 percent confidence limits for a normal
distribution.  An autocorrelation analysis of the transformed
data indicated that an AR(1)  model was most suitable to
estimate the 7- and 30-day rolling average NOX emission
limits.
     Table 3-23 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-22 and the
procedures described in section 3.2.3.2.  As indicated in
table 3-23, achievable emission limits based on one exceedance
in 10 years,  range from 0.33 Ib/MMBtu for a 24-hr block
average to 0.29 Ib/MMBtu for a 30-day rolling average.
     Effect of load on achievable NOX emission limits--The
results presented in the previous section provides an
assessment of NOX emission rates for the boiler load
characteristics experienced during June and July of 1992.  As
shown in figure 3-33; however, there is a general increase in
NOX emission rates as load increases.  Consequently, the NOX
emission rate from the boiler will be higher if the boiler is
operated at near full load for a longer period of time than
was reflected for this operating period.
     To assess the potential NOX emission rate when operating
the boiler at other duty cycles, linear regression was first
used to estimate the influence of boiler load on the hourly-
                             3-91

-------
     TABLE 3-23.  ACHIEVABLE NOX EMISSION LIMITS51 FOR
                         PARISH UNIT NO. 8

                  	Exceedance Frequency	
 Averaging Period       1%          I/year      1/10  years
    24-hr Block        0.313        0.320         0.333
   7-day Rolling       0.292        0.298         0.305
                                                        *
  30-day Rolling	0.278	0.281	0.285
alb/MMBtu.
                            3-92

-------
averaged NOX emission rate.  The resulting regression equation
was:
                  NOX = 0.000364 * MW + 0.07075            (3-6)

where:
     NOX  =    NOX emission rate in Ib/MMBtu;
     MW   =    boiler load in MW.
                                                          9
The as-measured hourly-averaged NOX emission rate data were
then adjusted based on this regression equation to estimate
the variation in the NOX emission rate that is not explained
by changes in boiler load.  The residual variation (i.e., the
measured emission rate minus the predicted emission rate based
on equation 3-6) ranged from -0.12 to +0.11 lb/MMBtu.  These
residual values were then added to a mean value equal to the
maximum hourly NOX emission rate predicted by equation 3-6.
(As shown in figure 3-33,  the highest expected NOX emission
rate corresponds to the design operating load of 615 MW).  The
new set of load-adjusted hourly NOX emissions data were then
averaged into 24-hr blocks and tested for normality.
     The results of the statistical analysis on the data set
are presented in table 3-24.  Based on the absolute values of
the skewness and kurtosis estimates of 0.163 and 2.164,
respectively,  the data were judged as being normally
distributed at the 95 percent confidence level (for a sample
size of 56 the 95 percent confidence limit for skewness and
kurtosis are approximately 0.51 and 3.96, respectively) .5q  An
autocorrelation analysis of the data indicated that an AR(1)
model was most suitable to estimate the 7- and 30-day rolling
average NOX emission limits.
     Table 3-25 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3, and on the statistical data in
table 3-24.  As indicated in table 3-25, the NOX emission rate
corresponding to a 30-day rolling averaging period and a one-
                             3-93

-------
  TABLE 3-24.   SUMMARY STATISTICS FOR WA PARISH UNIT NO. 8
            (LOAD-ADJUSTED 24-HOUR BLOCK AVERAGED  NOX DATA)

	Statistic	Value	
 No.  of observations                               56
 Mean (X24)  [Ib/MMBtu]                              0.295
 Standard deviation (824)  [Ib/MMBtu]                0.016
 Skevmess3-                                         0.163'
 Kurtosisa                                         2.164
 Shapiro-Wilk  W statistic3                         0.966
 Probability < W                                   0.222
 Autocorrelation coefficient*3 (p^)                  0.475
 Correction  factor  (F]_)                             1.016
 Adjustment  factor  (F2)
            7-day  rolling average                 0.575
            30-day rolling average                 0.299
aFor a perfectly normal distribution, skewness  =  0,
 kurtosis = 3, and W statistic = 1.
      lag 1 is significant.
                            3-94

-------
 TABLE  3-25.   ACHIEVABLE NOX EMISSION LIMITSa FOR WA PARISH
           UNIT NO. 8 BASED ON CONTINUOUS OPERATION AT LOAD
              CORRESPONDING TO MAXIMUM NOX EMISSION RATE.


                 	Exceedance Frequency	

 Averaging Period	1%	I/year	1/10  years

   24-hr Block        0.333         0.340         0.351

   7-day Rolling       0.316         0.321         0.327

  30-day Rolling	0.306	0.308	0.311

alb/MMBtu.
                           3-95

-------
in-ten-years exceedance frequency is 0.31 Ib/MMBtu.  As
expected, this value is higher than the corresponding emission
rate based on the actual operating load cycle for the unit,
presented in table 3-23, of 0.29 Ib/MMBtu.
     3.2.3.7.4  Conclusions.  As indicated in table 3-23,
achievable NOX emission levels calculated from the WA Parish
data for June and July, 1992 and a one-in-ten-years exceedance
frequency range from 0.33 Ib/MMBtu based on a 24-hour
averaging period to 0.29 Ib/MMBtu based on a 30-day averaging
period.  Because of the relationship between NOX emission rate
and load; however, higher NOX emission levels are expected to
occur during periods of sustained operation at high load
levels.  As shown in table 3-25, achievable NOX emission
levels based on continuous operation of the boiler at full
load (615 MW)  and a one-in-ten-years exceedance frequency are
estimated at 0.35 Ib/MMBtu for a 24-hour block average and
0.31 Ib/MMBtu for a 30-day rolling averaging period.
     3.2.3.8  Dolet Hills Unit No. I
     3.2.3.8.1  Boiler description.  Dolet Hills Unit No. 1 is
a 695-MW, opposed-wall, lignite-fired boiler located near
Mansfield, Louisiana.   The boiler is operated by Central
Louisiana Electric Company  (CLECO).  It is typically base-
loaded and in 1991 operated at an annual capacity factor of
73.4 percent.   It is equipped with 48 Babcock and Wilcox dual
register burners.  The burners are arranged on two opposite
walls in three rows with eight burners per row.  Each row of
burners on each wall is serviced by a single mill  (total of
six mills).  All burners are original equipment and have not
been physically modified since the boiler began operation in
1986.  The boiler does not have overfire air ports.
     The boiler is subject to the NSPS, subpart Da NOX
emission limit for lignite of 0.6 Ib/MMBtu.  Compliance tests
(short-term) conducted in 1986 indicated that the NOX
emissions were 0.385 Ib/MMBtu at 678 MW (98 percent load).
     3.2.3.8.2  Data summary.  Continuous emission monitoring
data were obtained from CLECO for the operating period between
                             3-96

-------
October 19 and December 15, 1992.  This data set includes
1,334 hourly-averaged NOX emission and boiler load values.  A
time plot of the NOX emission data is shown in figure 3-34.
There are no significant gaps in the data.  Typical hourly NOX
emission rates are between 0.3 and 0.50 Ib/MMBtu and average
approximately 0.40 Ib/MMBtu.  As shown in figure 3-35,  the
duty cycle of the boiler on most days during the year
consisted of operation at near full load during the day
followed by operation at approximately 40 percent of full load
during the night.  Figure 3-36 shows hourly-averaged NOX
emission values plotted as a function of load.  These data
indicate a general increase in average NOX emission rates as
load decreases from the unit's design load of 695 MW to
approximately 450 MW.  The increase in NOX is especially
pronounced at loads less than 600 MW and is believed to be
caused by the increase in 02 and decrease in air-staging
benefits attributable to the boiler's dual register burners.
Note, however there is significant scatter in the data in this
load range.  At loads less than 450 MW the NOX emissions
decrease with decreasing load to approximately 20C MW.
However, data between boiler loads of 200 to 450 MW are
limited.  The higher NOX levels at loads below 200 MW occur
during startups and shutdowns and as indicated earlier in
section 3.2.3,  these data were not included in the analysis.
     3.2.3.8.3  Statistical analysis results.  Analysis of the
data set included assessing the implications for standard
setting resulting from autocorrelation between successive NOX
measurements.  In addition, the effect of the N0x-load
relationship on achievable NOX emission limits was analyzed.
     Basic statistics for this data set are presented in the
first data column of table 3-26 for the 24-hour block
averaging time.  As indicated in the table,  the mean NOX level
was 0.39 Ib/MMBtu.  Based on the absolute value of the
skewness estimate of 0.866, the data were judged as not being
normally distributed at the 95 percent confidence level (for a
sample size of 55, the 95 percent confidence limit for
                             3-97

-------
                                              (N
                                              Oi
                                              (f,
                                              in
                                              (D
                                              0)
                                              u
                                              0)
                                              0

                                              o
                                              4J
                                              0)

                                              •8
                                              4->
                                              u
                                              o
                                              CO
                                              c
                                              o
                                             -H
                                              (0
                                              CO
                                             -H
                                              g
                                              4J

                                              O
                                              0)
XON
                                             EH 8
ro
 I
m

0)

!
   3-98

-------
                                   CN
                                   in
                                   0)
                                   0)
                                   U
                                   0)
                                   Q
                                   CTl
                                   t-l


                                   0)

                                   o

                                   o
                                   o


                                   o
                                   14-1



                                   s*
                                   o  .

                                   * o
                                   U-l

                                   O
                                   0)
                                     0)
                                   -H *
                                   H£
                                   in
                                   n
                                   i
                                   a;
                                  •H
3-99

-------
3-100

-------
 TABLE 3-26.   SUMMARY STATISTICS FOR DOLET HILLS UNIT NO.  1
                    (24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean 1X24) [Ib/MMBtu]
Standard deviation (624) [Ib/MMBtu]
Skewness3-
Kurtosisa
Shapiro-Wilk W statistic9-
Probability < W
Autocorrelation coefficient13 (p]_)
Correction factor (F]_)
As measured
data
55
0
0
0
3
0
0
NEC
NEC

.393
.034
.866
.660
.923
.002


Transformed
data
A= - 2.50
55
10
2
0
3
0
0
0
1

.659
.149
.040
.485
.966
.224
.573
.025
aFor a perfectly normal distribution skewness = 0,
 kurtosis = 3, and W = 1.
      lag 1 is significant.

CNE = not estimated.
                           3-101

-------
skewness and kurtosis are approximately 0.51 and 3.96,
respectively) .5"  To compensate for skewness, the data were
transformed using a Box-Cox transform with a X of -2.50.  The
statistical properties of the transformed data are presented
in the second data column of table 3-26.   The skewness and
kurtosis values for the transformed data are within the
95 percent confidence limits for a normal distribution.  An
autocorrelation analysis of the transformed data indicated
that an AR(1)  model was most suitable for use in estimating
the 7- and 30-day rolling average NOX emission limits.
     Table 3-27 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-26 and the
statistical procedures described in section 3.2.3.2.  As
indicated in table 3-27, achievable emission limits based on-
one exceedance in 10 years range from 0.64 Ib/MMBtu for a
24-hr block average to 0.45 Ib/MMBtu for a 30-day rolling
average.
     Effect of load on NOX emission limits--The results
presented in the previous section provides an assessment of
NOX emission rates for the boiler load characteristics
experienced between October to December,  1992.  As shown in
figure 3-36; however, there is a general increase in NOX
emission rates as load decreases.  The maximum NOX emissions
are at a load of approximately 450 MW.  Consequently, the NOX
emission rate from the boiler will be higher if the boiler is
operated at this load for a longer period of time than was
reflected for this operating period.
     To assess the potential NOX emission rate when operating
the boiler at other duty cycles, a curve-fit of the NOX
emissions data between loads of 200 MW (30 percent load) and
615 MW  (full load) was first used to estimate the influence of
boiler load on the hourly-averaged NOX emission rate.  The
resulting equation was:
                             3-102

-------
        TABLE  3-27.   ACHIEVABLE  NOX EMISSION LIMITSa
                         DOLET HILLS UNIT NO. 1

                  	Exceedance Frequency	
 Averaging Period	1%	I/year	1/10 years
    24-hr Block        0.505        0.546         0.638
   7-day Rolling       0.458        0.477         0.514
  30-day Rolling	0.425	0.433	0.447

alb/MMBtu.
                           3-103

-------
        NOX =  -9.98 * 10"7 *  (MW)2 + 0.00098  * MW * 0.1809   (3-7)

where:
     NOX  =    NOX emission rate in Ib/MMBtu;  and
     MW   =    boiler load in MW.
The as-measured hourly-averaged NOX emission rate data were
                                                          ?
then adjusted based on this regression equation to estimate
the variation in the NOX emission rate that is not explained
by changes in boiler load.  The residual variation (i.e., the
measured emission rate minus the predicted emission rate based
on equation 3-7) ranged from -0.20 to +0.34 lb/MMBtu.  These
residual values were then added to a mean value equal to the
maximum hourly NOX emission rate predicted by equation 3-7.
(As shown in figure 3-36, the highest expected NOX emission
rate corresponds to an operating load of 450 MW).   The new set
of load-adjusted hourly NOX emissions data were then averaged
in 24 hour blocks and tested for normality.
     The results of the statistical analysis on the data set
are presented in the first data column of table 3-28.  Based
on the absolute values of the skewness and kurtosis estimates
of 0.619 and 3.747, respectively, the data were judged as not
being normally distributed at the 95 percent confidence level
(for a sample size of 55, the 95 percent confidence limit for
skewness and kurtosis are approximately 0.51 and 3.96,
respectively).50  To compensate for skewness a Box-Cox
transform with a A = -1.0 was used.  The statistical
properties of the transformed data are presented in the second
data column of table 3-28.  The skewness and kurtosis values
for the transformed data are within the 95 percent confidence
limits for a normal distribution.  An autocorrelation analysis
of the transformed data indicated that an AR(l) model was most
suitable for use in estimating 7- and 30-day rolling average
NOX emission limits.
                             3-104

-------
 TABLE 3-28.   SUMMARY STATISTICS FOR DOLET HILLS UNIT NO. 1
            (LOAD-ADJUSTED 24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
No. of observations
Mean (X24) [Ib/MMBtu]
Standard deviation (824) [Ib/MMBtu]
Skewnessa
Kurtosisa
Shapiro Wilk W statistic3
Probability < W
Autocorrelation coefficient (PI)
Correction factor (F^)
As measured
data
55
0
0
0
3
0
0
NEC
NEC

.420
.035
.619
.747
.942
.018


Transformed
data
(X=-1.00)
55
2
r
0
-0
3
0
0
0
1

.394
.193
.009
.932
.960
.124
.599
.027
aFor a perfectly normal distribution skewness = 0,
 kurtosis = 3, and W statistic = 1.

bOnly lag 1 is significant.

CNE = not estimated.
                           3-105

-------
     Table 3-29 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
the statistics of the transformed data in table 3-28 and the
statistical procedures described in section 3.2.3.2.  As shown
in table 3-29, the NOX emission rate corresponding to a 30-day
averaging period and a one-in-ten-years exceedance frequency
is 0.47 Ib/MMBtu.  As expected, this value is higher than fthe
corresponding emission rate based on the actual operating load
cycle for the unit, presented in table 3-27,  of 0.45 Ib/MMBtu.
     3.2.3.5.5  Conclusions.   As indicated in table 3-27,
achievable NOX emission levels calculated from the Dolet Hills
data for October to December, 1992 and a one-in-ten-years
exceedance frequency range from 0.64 Ib/MMBtu based on a 24-
hc.;r averaging period to 0.45 Ib/MMBtu based on a 30-day
averaging period.  Because of the relationship between NOX
emission rate and load, higher NOX emission levels are
expected to occur during periods of sustained operation at low
load levels.  As shown in table 3-29,  achievable NOX emission
levels based on continuous operation of the boiler at part
load (450 MW) and a one-in-ten-years exceedance frequency are
estimated at 0.58 Ib/MMBtu for a 24-hour block average and
0.47 Ib/MMBtu for a 30-day rolling average.
     3.2.3.9  TNP One. Unit No. 1
     3.2.3.9.1  Boiler description.  TNP One Unit No. 1 is a
165-MW, lignite-fired, atmospheric pressure,  circulating FBC
boiler located in Bremond, Texas.  The boiler is operated by
Texas New Mexico Power Company.  It was manufactured by
Combustion Engineering and began operation in 1990.  The bed
is composed of ash, fuel, and limestone.  Bed temperatures
range between 1,500 to 1,600 °F.  Primary combustion air is
introduced through orifices located in the bottom of the
boiler.  Secondary air is introduced through three rows of
secondary air ports.  Approximately 35 percent of the total
combustion air is secondary air.  The boiler is typically
base-loaded and operates at an annual capacity factor of
                             3-106

-------
   TABLE  3-29.   ACHIEVABLE  NOX EMISSION LIMITSa FOR DOLET
                   HILLS UNIT NO. 1 BASED ON CONTINUOUS
                OPERATION AT LOAD CORRESPONDING TO MAXIMUM
                         NOX EMISSION RATE


                             Exceedance Frequency

  Averaging  Period	1%	I/year	1/10 years

    24-hr Block        0.518         0.543         0.584'

   7-day Rolling       0.480         0.495         0.517

   30-day Rolling	0.454	0.459	0.469

alb/MMBtu.
                           3-107

-------
90 percent.  No physical modifications have been made to the
boiler since it came online.
     The boiler is subject to the NSPS,  subpart Da NOX
emission limit for lignite of 0.6 lb/MMBtu.   A short-term test
conducted in 1990 indicated that the NOX emissions averaged
0.22 lb/MMBtu at a load of 163 MW.
     3.2.3.9.2  Data summary.  Continuous emission monitoring
(CEM) data were obtained for the operating period of
September 1 through October 30,  1992.  This data set includes
a total of 1,313 hourly-averaged NOX emission and load values.
A time plot of the NOX emissions data is shown in figure 3-37.
Note that there are several gaps in the data set due to
outages of either the boiler or the CEM system.  Typical
hourly NOX emission rates were between 0.10 and 0.15 lb/MMBtu,
and average approximately 0.13 lb/MMBtu.  Data collected
during a two-week period between October 9 and 22, 1992, were
eliminated from the statistical analysis because a mixture of
petroleum coke and lignite may have been fired.e   As  shown  in
figure 3-38,  the boiler was typically base-loaded and operated
at above 90 percent of full load during the day and night.
Figure 3-39 shows hourly-averaged NOX emission values plotted
as a function of load.  Because the boiler usually operates at
near full load, there are only a few data points reflecting
operation at lower loads.  At near full load, the graph shows
significant scatter in the data.
     3.2.3.9.3  Statistical analysis results.  Analysis of the
data set included assessing the implications for standard
setting resulting from autocorrelation between successive NOX
measurements.  Because the boiler is base-loaded, data were
insufficient to analyze the effect of the N0x-load
relationship on achievable NOX emission limits.
     'Under normal operating conditions, only lignite  is used
as fuel for the boiler; the addition of petroleum coke reduces
NOX emissions and consequently the estimated NOX emission
limits could be biased low if  these data are included.
                            3-108

-------
CsJ
5
_ N
cvi
0
CT)
^^
cB

SI
5i
pf*
- ijj
*^^
o
'"CNJ
5>
— co
o
CM*1"
o>
-8 1
sg°
__§

o
4J

CM
rH
W
r-t
r
                                                                        O)
                                                                      CM<=>
                                                                      o>
                                                                    _ in

                                                                        CVJ
                                                                        o>
                                                                        00

                                                                        i
                                                                      CNJ


                                                                      S


                                                                      §
9
O
o
^t
O
to
CO
o
CO
O
in
CVJ
o

o
in
P
o
                                                3 (N

                                                O ON
                                                                                 14-1
                                                                                 O  -
                                                                                   O
                                                Q, QJ
                                                  X>
                                                0) O
                                                e 4J
                                               •H O
                                               H o
                                                                                 n
                                                                                  I
                                                0)
                                                U
                                                3
                                                CP
                                               •H
                                       XQN
                                         3-109

-------
                                                                                       fN









i
J~
4
|
t
v
1
$
4
f
?
p







































































































































i
-5
c
T
C
c
T
^
c
T

c

c
'8

§
O
 8
 CD
                                                                        >    Q
                                                                        'CM


                                                                        CJ
                                                                        CM
                                                                    _  in
                                                                         i
                                                                      1
                                                                      i
                                                                                       0)
                                                                                      .Q
                                                                                       O
                                                                                      4J
                                                                                       CJ
                                                                                      O

                                                                                       O
                                                                                      J-l
                                                                                      CTl
                                                                                       V-l
                                                                                       0)
                                                                                       0)
                                                                                       -u
                                                                                       a
                                                                                       0)
                                                                                       CO
                                                                                      O
                                                                                      U-l
                                                                                      m
                                                                                      o
                                                                                      £ 4J

                                                                                      •sl
                                                                                      O  0)
                                                                                      •-H  C
              0) Q<
              CO
              m
               i
              m

              0)
                            (MIM)
                                            3-110
klj-237

-------
in

d
TT

d
m

d
o
CO
in
CVJ
o
CNJ
        in
                                                    8
                                                   -8
in
o
                                                    o
                                                    00
                                                        03
                                                    8
                                                   -8
                                                    8
                                                           1-1


                                                           6
                                                                    0)
                                                                    c
                                                                    o

                                                                    O4
                                                                   •o
                                                                    (C
                                                                    o
                                                                    o
                                                                    •rl
                                                                    4J
                                                                    u
                                                            (0

                                                            CO
                                                            rt

                                                            10
                                                            c
                                                            o
                                                            •H
                                                            to
                                                            w
                                                            -H


                                                            0)
                                                                    §

                                                                    •o
                                                                     (1)
                                                                     en
                                                                     (t
                                                                     i-l
                                                                     Q)

                                                                     rt
                                                                     §
                                                                     ffi
                                                                     ro
                                                                      i
                                                                     0)
                                                                     &
                                                                     •H
                    3-111

-------
     Basic statistics for this data set are summarized in
table 3-30.  As indicated in the table, the mean NOX level was
0.14 Ib/MMBtu.  Based on the absolute value of the skewness
estimate of 0.176, the data were judged to be normally
distributed at the 95 percent confidence level (for a sample
size of 44, the 95 percent confidence limits for skewness and
kurtosis are approximately 0.55 and 4.0, respectively).59  An
autocorrelation analysis of the data indicated that an AR(1)
model is most suitable for use in estimating the 7- and 30-day
rolling average NOX emission limits.
     Table 3-31 shows the calculated NOX emission limits as a
function of averaging time and exceedance frequency based on
equations 3-2 and 3-3 and the statistical data in table 3-30.
As indicated in the table, achievable emission limits based on
one exceedance in 10 years range from 0.18 Ib/MMBtu for a
24-hr block average to 0.15 Ib/MMBtu for a 30-day rolling
average.
     3.2.3.9.4  Conclusions.  The data from TNP One Unit 1
indicate that the boiler was base-loaded during the period of
September and October 1992.  As indicated in table 3-31,
achievable NOX emission levels calculated from the TNP data
based on one exceedance in 10 years range from 0.18 Ib/MMBtu
for a 24-hour block average to 0.15 Ib/MMBtu for a 30-day
rolling average.
3.2.4  Summary
     The results of the statistical analysis on data sets
reflecting recent boiler operating experience for the six
representative boilers are summarized in table 3-32.  For each
boiler, the hourly-average NOX emission level and the
estimated 30-day rolling average NOX emission limit are
presented.  The current subpart Da standard for NOX emissions
are also included for comparison.  The results indicate that
the estimated NOX emission limits from each boiler are
significantly lower than the current standard.
                             3-112

-------
  TABLE 3-30.  SUMMARY STATISTICS FOR TNP ONE UNIT NO.  1
                   (24-HOUR BLOCK AVERAGED NOX DATA)

	Statistic	Value
 No.  of  observations                                44
 Mean (X24)  [Ib/MMBtu]                               0.137
 Standard  deviation (824)  [Ib/MMBtu]                 0.011
 Skewnessa                                          0.17&
 Kurtosisa                                          3.022
 Shapiro-Wilk W statistica                          0.988
 Probability < W                                    0.969
 Autocorrelation coefficient13 (PD                   0.482
 Correction  factor (F^)                              1.021
 Adjustment  factors (F2)
             7-day rolling average                   0.579
             30-day rolling average                  0.302
aFor a perfectly normal distribution skewness =  0,
 kurtosis = 3, ana W = 1.
      lag 1 is significant.
                           3-113

-------
     TABLE 3-31.  ACHIEVABLE NOx EMISSION LIMITS3 FOR
                          TNP  ONE UNIT NO.  1

Period
24-hr Block
7 -day Rolling
30 -day Rolling
Exceedance Frequency
1%
0.163
0.152
0.145
I/year
0.169
0.155
0.147
1/10 years
0.176
0.160
0.149
alb/MMBtu.
                           3-114

-------
CO
O
m
Q

EH
CM
CQ
D
Cti
o
U)
EH
J
D
C/3
CO
H
C/!
CO

W
J
CQ



3
1
^\
XI
r-t
•«_ -
CO
C
o

M
to
-r-t
E
0)
O


0) jj rd
(? C Q 4J
0 ^ 2 ~ e
> ><_ 3^ M
C"**-^ C
_] M

rH ^
VH *j ij O
6 (U NO}

jrt *" | 	 | (*
•0 H rd -O

o <2
ro


>> a
l-i *
£E
E _
SsS*
01 •§ 0) 0
o £-S

4J 0)~
c "* i
0)
id
c

4J
C
rd
r-H
a

in
0)
13
8$
(0 T3
C^J
c
3 i
[14
X)
Q)
'a! <"
4J

U)
o





ro
•^

o





co
co
ro
O
en
u
0)
Q
4-1
U
0
in
s


C

0
'H
m

CQ
<



5
CQ
a

CQ


VD
O





CN
^

0





CO
en
ro
o
en
a
0)
CO
rH
^
in
in


T5
C
0)
PQ

Oi
-rl
PQ



W
U
ea
C
H
CQ


in
o





«HH
ro

o





en
CN
o
CN
U
0)
Q
4J
U
o
CO
ro



X)
E
0
U
rH
0
X



5
a
a
o

CO


in
o





CTl
CN
•
O





0
CN
0
CN
en
r-H
7
C
^
in
rH




r|
tn
-rl
^J
rd




w
u
PQ
s
Oi
EH
CO


vo
o





in
^

o





ro
en
ro
o
CN
U
CU
Q
JJ
U
0
in
en
r-t
r-t
-rl
a

4J
0)
rH
0
Q



S
ea
a

^


vo
o





in
rH

O





o
ro
rH
O
CN
en
O
&
a
0)
CO
in
vc



0)
C
o

CD
S
H



W
U
u
CQ
CN
U


























0)
4J
-rl
C
tn
-H
rH

1!

^
'O
c
rd
,fc
in
rj
o
c
•rl
e
3
4-1
-H
f~)
a
3
tn
II

CO

• ..
tn
3
O
C
-rl
g
3
4-1
•H
X)
II
CQ
rd
T3
0)
N
"H
T!
3
m
X C
0-H
U 4->
-rl r-t
S 3
U
i-S iH
•H
Ai U
U
0 II
U
X) U
ns ca
CQ b
u
II
S tn
u£J £^
PQ -H
1^
•- Q)
T) (1)
0) C
rl -H
-H en
M-l C
' U
i-H C
rH O
rd -ri
-rl 4J
C 3
0) X3
Oi E
C 0
rd U
^j
i
II
•rl
Ol 'H
C Q)
rd >
H 0
PQ
T3 C
CD S
^ O
-rl i-|
>H CQ •
C
T) rfl O
CD CD -rl
tn ro 4J
o <( tn
a 3
a n X)
0 6
w o
II U U
am TJ
aca a)
O 
-------
3.3  COMBUSTION CONTROLS FOR NATURAL GAS- AND OIL-FIRED
     UTILITY 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 3-40, FGR involves extracting a portion of the flue gas
from the economizer or air heater outlet and readmitting it to
the furnace through either the furnace hopper, or the burner
windbox, or both.  To reduce NOX, the flue gas is injected
into the windbox.62  The degree of FGR varies between
10-20 percent of the combustion air.  Windbox FGR is primarily
effective at 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.  To maximize NOX
reduction, FGR is used to suppress temperature within the
flame.  The effectiveness of the 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 NOv 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.

                             3-116

-------
                     CVJ
.*
o
2
C/3
CVJ,

J2
0)
(0
I
D
















Sf
(/)
flj
O
0)
D
U-









v>
a
O
n^ .
3!
E 1




16
0
V
ct
5


i



u.
2
LL
£












/
























l_
<
/ ,










g
I
r«t


//^~
r /
/
/
/
I/
/
£
f
1

v
V
^-*^z 	 •
0) r?\
                          d

                          £
                          a:
                          g
               ^^^
               ^8^
                                                    I
                                                    JJ
                                                    in
                                                    >1
                                                    K

                                                    c
                                                    o
                                                    •H
                                                    u
                                                    i-i
                                                    •H
                                                    U
                                                    0)
                                                    w
                                                    (T3
                                                    cn

                                                    
-------
     3.3.2.1.1  Wall-fired boilers.   As with 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 3-41.63  Combustion in a ROPM™
burner is internally staged,  and takes place in two different
zones; one under fuel-rich conditions and the other undert
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.63
     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 Dynaswirl™  burner divides  combustion  air  into
several component streams and controls injection of fuel into
the air streams at selected points to maintain stable flames
with low NOX generation.  Figure 3-42 illustrates the internal
configuration of the burner.62  For natural gas-firing, fuel  is
introduced through six pipes, or pokers,  fed from an external
manifold.  The pokers have skewed, flat tips that are
perforated with multiple holes directed inward toward the
burner centerline.  Primary air enters through the burner
                             3-118

-------
       Ji
.v*ir    ~*   j
                                    •i*
                                    WO
ill    il
                l
                                                        W
                                                        ^
                                                        0)
                                                        rH
                                                        •H
                                                        O
                                                        £1

                                                        •o
                                                        0)
                                                        ^1
                                                        •H
                                                        U-l
                                                        I
                                                        tH
                                                        •H
                                                        O

                                                        T>
                                                        C
                                                        re
                                                          re
                                                          tn
                                                          re
                                                          re
                                                          c
                                                          (U
                                                          c
                                                          i-l

                                                          .Q
                                                         2
                                                         &4
                                                         O
                                                          i
                                                         ro


                                                         0)
                                                         t-i
                           3-119

-------
                                    0)
                                    c
                                     X
                                    o
                                    O
                                    3
                                    W
                                    (0
                                    c
                                    0)
                                    ^
                                    3
3-120

-------
venturi surrounding the center gas gun and mixes with this gas
to form a stable flame.  Gas fed through the pokers mixes with
secondary air flowing along the outer walls of the venturi and
is ignited by the center flame.62
     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 3-43, can fire natural gas or
oil."
     The wall-fired Primary Gas - Dual Register Burner™
(PG-DRB), shown in figure 3-44, was developed to improve the
NOX reduction capabilities of the standard DRB.65   "Primary
gas" is recirculated flue gas that is routed directly to each
PG-DRB and introduced in a dedicated zone surrounding the
primary air zone in the center of the burner.  The
recirculated gas inhibits the formation of thermal and fuel
NOX by reducing peak flame temperature and oxygen
concentration in the core of the flame.  The dual air zones
surrounding the PG zone provide secondary air to control fuel
and air mixing, and regulate flame shape.  The system usually
includes FGR to the burner and to the windbox, with OFA ports
installed above the top burner row.
     In addition to the XCL™ burner  for  coal-fired boilers,
the XCL™  as  shown  in  figure  3-45  is  also available for wall-
fired boilers burning natural gas and oil.2   This  design
enables the use of an open windbox (compartmental 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
gas- and oil-fired retrofits is shown in figure 3-46.66  In
this design,  the internal staging of primary and secondary air
can be adjusted depending on required NOX control and overall
                             3-121

-------
                                             COOLER OXYGEN RICH ZONE I
                                              REDUCES THERMAL NO>  I
                                             TM
Figure 3-43.   Internal staged combustion1"  low NOX  burner.
                              3-122

-------
                                              0)
                                              c
                                              M
                                              X
                                              o
                                              S-i
                                              0)
                                              .p
                                              0)
                                              (0
                                              3
                                              13
                                               I
                                              t/1
                                              m
                                               (0
                                               e
                                               0!
                                               )-i
                                               3
3-123

-------
                                                      14
                                                      0)
                                                      c
                                                      M
                                                      3
                                                     .Q
                                                     o

                                                     8
                                                     0)
                                                     ^
                                                     •H
                                                     14-1
                                                     I
                                                     rH
                                                     •H
                                                     O

                                                     T3

                                                     ra

                                                     i
                                                     CO
                                                     to
                                                     ro
                                                     S-i
                                                     (C
                                                     C
                                                    O
                                                    X
                                                    IT)
                                                    ^r
                                                    I
                                                    0)
                                                    14
                                                    D
3-124

-------
                                               0)
                                               c
                                                X
                                               §
                                               I
                                               1-1
                                               o
                                               •rl
                                               AJ
                                               ro
                                               a
                                               
-------
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.66
     3.3.2.1.2  Tangentially-fired boilers.   The
tangentially-fired Pollution Minimum™  (PM) burner is  shown in
figure 3-4V.67  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 "GM"
(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.0
     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
flame zones, thereby maintaining the NOX reducing
characteristics of both flames.67
     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
                             3-126

-------
   O


   I
o
1
A
hi
3
                                        Ik

                                        O

                                        *
                                                        I
                                                        0
                                                        V)
CO
14
0)
i-l
-rl
o
jQ

•O
0)
•o

10

 I
ID
to
01
                                                                to
                                               JJ
                                               m
     I  ii
     o  •>

     i!
    I
    T)
    C
    i
                                o
                                                                o
                                                               4-1

                                                                V-l
                                                                0)
                                                                c
i
Comparti
i
>


A
^
It
§tt
O
A.
t1
L



»
^_
y
,, 	 , ^_


, i
,r-7l.Lr
1
f
Jl
\
U=
i
1
i


=^O*H^+-
1 , '>>
1 H H
1 I i ;
, ' i
— < \
•— ,hr- i i-i 1 1—
=n np
1^
AtJlfff
Jiii^
i i' ;•.
1
!

" i
o i
~Y i
x! 0


J
r

!' ''''"lig
O
Ni
l=Jt^
*

-^




\
||:; j
^ o
1 1
IL. 1?
^ Ik
3
a
a
0)
c
*c
iZ
M
5
"5
£
|
1

g
nimum
•H
2
C
O
-H
4J
P
O
a

T
ro
C;
P
Cr
                         3-127

-------
successful NOX reductions are on natural gas and light fuel
oil firing on boilers initially designed for specific fuel use
patterns.  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
and fuel type.  The process descriptions for FGR and LNB are
in sections 3.3.1.1 and 3.3.2.1,  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
     New natural gas- and oil-fired utility boilers are
expected to be built with combinations of LNB,  FGR and OFA to
achieve current subpart Da NOX emission limits.  Although
there are no known subpart Da natural gas- and  oil-fired
utility boilers,  there are subpart D and pre-NSPS utility
boilers burning these fuels that have been retrofit with
combinations of these combustion controls.  For these
applications long-term (CEM) data are unavailable.
Consequently, only the available short-term test data
reflecting the performance of these technologies are presented
in table 3-33.  Results are given for several combinations of
controls on two wall boilers firing fuel oil, and on four wall
and one tangential boiler firing natural gas.
     For the oil-fired Kahe 6,  the combination  of LNB and FGR
achieved NOX emissions of 0.43 Ib/MMBtu.  The combination of

                             3-128

-------
w
  8
i- u
§0
w
D
(Q
D
§W
8£
go

2^
c] 2
  0
si
gs
o

a

PI
n
I
CQ

EH


c
o
**-
QC
!*
|
u
3
Uncontro


|S
"U
gS
NOX
emissio
(Ib/MMB
'o i rj
re « C,
Q. CJ
re «^
l°l


•o
oH, £

J2
§
U


ro
^ ro
4-1
i

o
WALL-FIRED BOILERS, FUEL







«
*
K1
O
00
o
«

L.
o
jr
00
oc
U. ^^
+ 3
m
CO *-*
3
CO
>o


•0
OJ ^^
ro ^"


o
C u
to ••-
to u
3 CJ
(0 —
Z UJ
-
is
CO
ru
o
CO
o
fM
0-

L-
O
u.
+ 3
cc
CO ^^
Z
3
u:
m
•0


>o
|s


0
5.S
•»- *^
ID U
3 OJ
ID —
X UJ
S
3
L5.o
00_

in N- .^
in «- *^
00°
g o in
~ in r\j

o o o
JZ JZ -C
oo in in
 ^
re 5
4^
C
o
05
I/) O
rc LJ
u
H- 4^
O OJ
(0 —
0. U-l
3
-o
o-
o
00
o
fxj
0-

L.
o
in
u.
o ce ^
o 3
+ u. oa
CO
CO + ^
Z
3
CO
-0


-0
-§
ID


0
c u
 V
a

08
in o
ra u
u
u
u •—
H- 4-1
— O
o o>
1C —
Q. UJ
o-
•o
$
in
~»
o
in
in
o
o
o


L_
O
r-
Irt
^
u.
O
+
Of
o
u.
3
od
CO
in
vf

a
01
° "t
L. *~*
4-1
§
U
0«
V) O
a u
13
U
O "-
— u
»4- 4-
"~ fli
ro — •
a. uj

PJ

*
0
«^-
o
o-
o
o
o


4->
o
in
ce
if £
+ 5
)__
5
3

m
in
O-

vO
«
£ 7
1 ~
<



a •
ccS

OJ O C
•£ "*- °
O fo ^
I/I U UJ

CM

»» N. ^«-
t> o- o
i i i
o o o
•*- *4- *4-
t> o. o-
i xO -0
p^- o o
co ^ m


U U L.
O O O
(/) in in
C£
" 9
+ "§
CO *~
Z "^
_J
3

CO
o
o
CO
r\l
JI
u
(C
OJ ^
CO U
•n o.
i
o


CD •
ccS
l_ I.
cu o c
f. «4- O
O CO "D
in o uj
                         3-129

-------
TABLE 3-33. PERFORMANCE OF COMBINATIONS OF COMBUSTION CONTROLS ON
U. S. NATURAL GAS- AND OIL-FIRED UTILITY BOILERS (Concluded)
Reference
Reduction
in NOX
emissions
(X)
Controlled
NOX
emissions
(Ib/MMBtu)
Uncontrol led
NOX
emissions
(Ib/MMBtu)
£-0
o2s
E.Z-
te 4->
u
~ "*.
pS
•D
°°
XI
Z
g
>.
£'G 3
H-
L)
Unit
(standard)3
x
4-<
4^
3


VI
*t
ENTIALLY-FIRED BOILERS, NATURAL (
o
x
<
H-



S
S
83E5
O •« M
•-00
000
in r\j ro
o* -* ro
o o o'
goo
~ in ro
i_
o
.c
en
<
U.
O
•f
or
(J
u.
UJ
u
CO
CD
*I
LT»
f£
Pittsburg 7
(Pre)
Pacific Gas &
Electric
aStandard: D = Subpart 0 and Pre = Pre-NSPS
bOEM = Original Equipment Manufacturer; ABB-CE = Asea Brown Boveri -Combust ion Engineering; and B&U = Babcock & Wilcox.
cType Control: Flue Gas Recirculation; LNB = Low NOX Burners; and OFA = Overfire Air
**
so
x
_j
L.
'5
M
o
1
I
u
h-
II
?
O
t—
•g
ro
i_
OJ
c
1_
D
CD
t_
(U
4-*
ifl
Ol
(U
a:
to
D
O
W
(C
o
X
(_
(_
CL
x
o
u
3
06
J*
U
o
jj
1C
OB
II
3
o£
m
CO
c.
o
•g
o
•f
(/)
L.
i
jr
OJ
«T
1
4^
10
0)

4^
4->
b.
O
JZ
in
II
4-^
i_
O
-C
(/)
a>
&
u
c
&
u
a>
«4-
OJ
(_
a*
o
W5
CO
4J
re
T)
ID
U
L.
O
4-<
cn
JZ
c
o
•g
C4
(I!
XI
*
01
4^
4^
U!
U
C
a
en
o
3

t-
3-130

-------
LNB and OFA on Kahe 6 achieved NOX emissions of 0.28 Ib/MMBtu
and LNB + OFA + FGR achieved NOX emissions of 0.19 Ib/MMBtu.
For the other oil-fired wall boiler  (Contra Costa 6), FGR +
OFA achieved NOX emissions of 0.19 Ib/MMBtu at full-load.
     For two natural gas-fired wall boilers (Pittsburg 6 and
Contra Costa 6), FGR + OFA achieved NOX emissions of 0.16 and
0.24 Ib/MMBtu.  For the other two natural gas-fired wall
boilers  (Alamitos 6 and Ormond Beach 2),  combining LNB and FGR
reduced NOX to approximately 0.1 Ib/MMBtu.  Finally, the
combination of FGR + OFA on the natural gas-fired tangential
boiler achieved NOX emissions of 0.1 Ib/MMBtu at full load.
Similar controlled NOX emissions were obtained at reduced
loads.
     The results presented in table 3-33  reflect the
performance of the NOX control technologies retrofit on
existing subpart D and pre NSPS utility boilers that were
built with small furnace volumes.  Because new boilers of the
same capacity are designed with larger furnaces volumes, the
NOX emissions from new boilers are expected to be lower.
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
utility boilers are selective noncatalytic reduction (SNCR)
and selective catalytic reduction (SCR).
     Selective noncatalytic reduction involves injecting
ammonia or urea into the flue gas of a utility boiler to yield
N2 and water.  The ammonia or urea must be injected into
specific temperature zones in the boiler's upper furnace or
convective pass for this method to be effective.70  The other
commercially available flue gas treatment method,  SCR,
involves injecting ammonia into the flue  gas in the presence
of a catalyst.  The catalyst promotes reactions that converts
NOX to N2 and water at lower temperatures than required for
SNCR.
                             3-131

-------
     Both SNCR and SCR technologies have been applied to
commercial-scale, conventional coal, natural gas- and oil-
fired boilers, in the U.S.  In addition, SNCR has been applied
to a number of stoker and FBC coal-fired boilers in the U.S.
3.5.1  Selective Noncatalytic Reduction
     3.5.1.1  Process Description.  The SNCR process involves
injecting a nitrogen-bearing chemical  (usually ammonia or
urea) into boiler flue gas at temperatures of 870 to 1,04Q °C
(1,600 to 1,900 °F).   The ammonia or urea reacts with NOX in
the flue gas to produce N2 and water.
     As shown in figure 3-48, for the ammonia-based SNCR
process, ammonia is injected into the flue gas where the
temperature is 950 ± 30 °C (1,750 ± 50 °F).71  The primary
reaction equation is:
               2NO + 2NH3 +  1/2 02 -» 2N2 + 3H20          (3-10)
Competing reactions that use some of the NH3 are:
                    4NH3 + 502 -» 4NO +  6H20              (3-11)
                    4NH3 + 302 -* 2N2 +  6H20              (3-12)
To maximize reaction 3-10, NH3 must be injected into the
optimum temperature zone, and the ammonia must be effectively
mixed with the flue gas.  When the temperature exceeds the
optimum range, reaction 3-11 becomes significant, NH3 is
oxidized to NOX,  and the net NOX reduction decreases.72  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  (not enough to appreciably raise the
temperature) can be injected with the NH3 to lower the
effective temperature range to approximately 1,300 °F.
     As shown in figure 3-49 for the urea-based SNCR process,
an aqueous solution of urea  [CO(NH2)2]  is injected into the

                             3-132

-------
C
O
   E
   0)
                              CVJ
                                                     o
                                                    •H
                                                     4-)
                                                     0)
                                                     u
                                                    -H
                                                     fO
                                                     U
                                                     to
                                                     u
                                                     c
                                                     o
                                                     c

                                                     Q)

                                                     •H
                                                     JJ
                                                     U
                                                     0)
                                                     tH
                                                     0)
                                                     to
                                                      0)
                                                     X!
                                                      i
                                                      C
                                                      O
                                                      oo
                                                      
-------
          NOxOUT Process
                                      Boiler
Figure 3-49.   Urea-based selective  noncatalytic reduction.
                                                              70
                             3-134

-------
flue gas at one or more locations in the convective pass.70
The urea reacts with NOX in the flue gas to form N2 , water,
and carbon dioxide  (CC>2) •   Aqueous urea has a maximum NOX
reduction activity at approximately 930 to 1,040 °C  (1,700
to 1,900 °F) .   Proprietary chemical enhancers may 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 °F) .70
     The exact reaction mechanism is not well understood
because of the complexity of urea pyrolysis and the subsequent
free radical reactions.  However, the overall reaction
mechanism is:
           CO(NH2)2 + 2NO + 1/202 -» 2N2 + C°2 + 2H20     (3-13
     Based on the above chemical reaction, one mole of urea
reacts with two moles of NO.  However, results from previous
research indicate that more than stoichiometric quantities of
urea must be injected to achieve the desired level of NOX
removal ."^  Excess urea can result in emissions of nitrous
oxide  (N20) ,  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 3-50.73  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
system improves flue gas mixing.74
     Hardware requirements for SNCR processes include reagent
storage tanks,  air compressors, reagent injection grids, and
an ammonia vaporizer (ammonia -based SNCR) .  Injection

                             3-135

-------
                 0)
                 16
                 0>
                 .c
                 o5
                 Q.
                 3
                 CO
        0)
       .N
        o

       I



I
CO
                       sjeujng
                                                        I
                                                       ro

                                                       Q)
                                                       M
                                     3-136

-------
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 injection is used.  A
control system consisting of a NOX monitor and a controller/
processor (to receive NOX and boiler data and to control t;he
amount of reagent injected) is also required.
     3.5.1.2  Factors Affecting Performance.  Six factors
influence the performance of urea- or ammonia-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
reduce the reaction kinetics and hence the achievable NOX
emissions reductions.
     As shown in figure 3-51,  the gas temperature can greatly
affect NOX removal and NH3 slip.''''  At temperatures below the
desired operating range of 930 to 1,090 °C  (1,700
to 2,000 °F),  the desired NOX reduction reactions begin to
diminish and unreacted NH3 emissions (slip)  increase.  Above
the desired temperature range, NH3 is oxidized to NOX,
resulting in low NOX reduction efficiency and low reactant
utilization.73
     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 1040 °C ± 70 °C  (1,900 °F
± 125 °F) at half load.  At similar loads, temperatures can
increase by as much as 30-60 °C (55 to 110 °F)  depending on
the extent of ash deposition on heat transfer surfaces.  Due
                             3-137

-------
400-


350-



	 \
\
\
\
\ NH3 Slip
300-1 *
t  250J
|

8  200-


Z  150-


o  100-

      j
   50 -'
            1000
                 1200
                                  NO
                                       Initial
                                       NH3,NOx = 2.1
                             1400


                          Temperature F
                                                1600
1800
Figure 3-51.   General effects  of temperature on NOX removal.


                                3-138
                                                                      73

-------
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.73
     The third factor affecting SNCR performance is mixing of
the reagent with the flue gas.  The zone surrounding each
reagent injection nozzle is 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.73
     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,
there will be insufficient time for completion of the desired
reactions between NOX and NH3.
     The fifth factor in SNCR performance is the ratio of
reagent to NOX.  Figure 3-52 shows that at an ammonia-to-NOX
ratio of 1.0, NOX reductions of less than 40 percent are
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 NH3:NOX 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.  Excess NH3 can react
with sulfur compounds in the flue gas to form ammonium sulfate
salt compounds that deposit on downstream equipment.  Higher
NH3 feed rates also can add costs.
     The sixth factor in SNCR performance is the sulfur
content of the fuel.  Sulfur compounds in the fuel can react
with NH3 and form liquid or solid particles that deposit on
downstream equipment.  As shown in figure 3-53, depending on
the concentrations of NH3 and 803 in the flue gas, ammonium
bisulfate  (NJ^HSC^) or ammonium sulfate [(NI^^SC^]  will form
                             3-139

-------
                             Performance of
                             Actual Commercial
                             Installation
          1          2         3

               Initial Mole Ratio of NH3 to NOx
Figure 3-52.
General  effect  of NH3 :?NOX mole
 ratio on NOX removal.
                        3-140

-------
   500
   100




a  50


g

1

    10  —
     5  —
                         10
50    100
500
                        SO .Concentration, ppm
                          3
      Figure 3-53.   Ammonia salt formation as a function of

                    temperature and NK3 and 803 concentration.
                        76
                              3-141

-------
at temperatures below 260 oc (500 oF)-76  For example/ ammonium
bisulfate and ammonium sulfate can form and plug and corrode
air heaters which typically operate at temperatures of less
than 260 °C (500 °F).
     Most SNCR experience has been on boilers less than 200 MW
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, 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 and more engineering and physical/
mathematical modeling of the process as part of system design.
     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.
3.5.2  Selective Catalytic Reduction
     3.5.2.1  Process Description.  Selective catalytic
reduction involves injecting ammonia 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:77
                 4NH3 + 4NO + 02 -» 4N2  + 6H20            (3-14)
                   8NH3  +  6N02  -» 7N2  + 12H20             (3-15)
There are also undesirable reactions that can occur in an SCR
system, including the oxidation of NH3 and S02 and the
                             3-142

-------
formation of sulfate salts.  Potential oxidation reactions
are:78
                    4NH3 + 502 -* 4NO + 6H2°              (3-16)
                    4NH3 + 302 ~* 2N2 + 6H2°              (3-17)
                    2NH3 + 202 •* N20 + 3H20              (3-18)
                        2S02  +  02 -»  2S03                  (3-19)
                                                          r
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 3-54.78
     Figure 3-55 shows several SCR configurations that have
been applied to power plants in Europe or Japan.70  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 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.79
     The SCR system hardware includes the catalyst material;
the ammonia system-- including  a vaporizer, storage tank,
blower or compressor, and various valves, indicators, and
controls,  the ammonia injection grid; the SCR reactor housing
(containing layers of catalyst); transition ductwork; and a
continuous emission monitoring system.  Anhydrous or dilute
aqueous ammonia can be used; however, aqueous ammonia 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 ammonia feed rate), or a
combination of the two.   The individual catalyst honeycombs or
                             3-143

-------
   NOx
CONVERSION
                          COMPOSITE
                          OF SCR NOx
                          AND NH3 OXIDATION
                          REACTIONS
                             MAXIMUM
                             CONVERSION
                             OPERATING

                            OPERATING
                           — WINDOW	
                           TEMPERATURE
           Figure  3-54.
Relative effect ofg temperature
on NOX reduction.
                                3-144

-------
08)
                L
                   NH,

                !    Waste Waste
               Ash   Water Solids
                 I

                Ash
         L
            NH,
Waste Waste
Water Solids
(1C)
                L
                   NH,
                                  Ash Dry Scrubber Waste
                   NH,
                            Ash       Dry Scrubber Waste
                           Waste Waste     I
                           Water RnllHc     '—*•
                      Ash  Water Solids
                              NH,
             SCR          ESP or   WetSO2  Spray
                   Heater    FF    Scrubber  Dryer
Figure 3-55.
Possible configurations for  selective
catalytic reduction.
                          3-145

-------
plates are combined into modules,  and the modules are applied
in layers.  Figure 3-56 shows a typical configuration for a
catalyst reactor.80
     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, and 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
surface area for the active material, thus enhancing the
contact of the flue gas with the active material.  Figure 3-57
shows examples of relative optimum temperature ranges for
precious metal, base metal, and zeolite catalysts.78
     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.S!
     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.78
     The most common commercially available base metal oxide
                             3-146

-------
                                           flow directing layer
        catalyst element
         catalyst module
                                                 catalyst layers
Figure  3-56.  Typical  configuration for  a catalyst reactor.
                                                                 80
                                 3-147

-------
                                          •§


                                          -8


                                          -S
8
5        5"
                              o
                              LL)
                              DC
                             tr
                             UJ
                             Q.


                             I
                    1-8


                      2
                                                                   w
                                                                   oJ
                                                                   CO
                                                                   (Q
                                                                   4J
                                                                   (0
                                                                   o
                                              (0
                                              0)

                                              &
                                              4->

                                              4J
                                              c
                                              0)
                                              •o

                                              IM
                                              o
                                                                  0)
                                                                  en
                                                                  c
                                                                  rt
                                                                  0)
                                                                  M
                                                                  d
                                                                  iJ
                                                                  rC
                                                                  ^
                                                                  (P
                                                                  cu
                                                                  4J
                                                                  i
                                                                  •H
o

OJ
     o
     o
                                                                  X
                                                                  w
                                                                  r^-
                                                                  in
                                                                  i
                                                                  m


                                                                  0)
                                                                  M


                                                                  &
                                                                  •H
                                                                  EL,
                      3-148

-------
catalysts are vanadium/titanium based, with vanadium pentoxide
("^205) used as the active material and titanium dioxide  (TiC>2)
or a titanium oxide-silicon dioxide  (SiC>2) as the support
material.82  Vanadium 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.83
     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,  SC>2,  and 02 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.
     The SCR catalyst is usually offered in extruded honeycomb
or plate configurations as shown in figure 3-5S.84  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
                             3-149

-------
                 yyyy
                 vvvv
 honeycomb
plate
Figure 3-58. Configuration of parallel flow catalyst
             3-150

-------
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 1-4 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 802 to S03 conversion rate, or
burn a low sulfur coal.  As shown in figure 3-59, the
conversion of 802 to 803 is temperature dependent, with higher
conversion rates at the higher temperatures.85  The temperature
sensitive nature of 802 to ^03 conversion is especially
important for boilers operating at temperatures greater than
370 °C (700 °F)  at the economizer outlet.85  Potential reaction
equations for ammonium sulfate salts are:86
       NH3(gas) +  S03(gas) + H20(gas)  -> NH4HS04 (liquid)  (3-20)
         NK4HS04 (liquid)  + NH3 (gas)  -*  (NH4 ) 2S04 (solid)    (3-21)
      2NH3(gas) +  S03(gas) + H20(gas)  -»  (NH4) 2S04 (solid) (3-22)
With the use of medium- to high-sulfur coals, the
concentration of 803 will likely be higher than experienced in
most SCR applications to date.   This increase in 803
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
                             3-151

-------
0)
o

0)~
T3
£2
I
o
u
cT
05
      2-
1 ~
      o-i
          482
                 572           662

                    Temperature, (°F)
752
     Figure 3-59.  Effect  of temperature  on conversion
                   of  SC>2  to 803.
                            3-152
                                  85

-------
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.87  However, at inlet concentrations below
150 ppm, the reduction efficiencies decrease with decreasing
NOX concentrations.87
     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.88  As space velocity increases, the
contact time between the gas and the catalyst decreases.  As
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.88  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
"poisoned" 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
                             3-153

-------
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
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 the
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.
                             3-154

-------
3.6  PERFORMANCE OF FLUE GAS TREATMENT TECHNOLOGIES ON
     COAL-, NATURAL GAS-, AND OIL-FIRED UTILITY BOILERS
     This section presents the demonstrated performance of the
gas treatment technologies  (SNCR and SCR) applied to
conventional and FBC utility boilers.  Because long-term CEM
data are not available for these applications  (except FBC
boilers with SNCR), the performance levels are based on
reported short-term test data.
3.6.1  Selective Noncatalytic Reduction
     3.6.1.1  Performance of SNCR on Conventional Utility
Boilers.  The results of SNCR applied to conventional fossil
fuel utility boilers are shown in table 3-34.  There are
2 coal-fired, 2 oil-fired, and 10 natural gas-fired
applications represented on the table.  One application is
NH3-based SNCR with the remainder being urea-based.  Available
data on NH3 slip and N20 emissions during these tests are
presented in table 3-35.
     For the Valley 4 coal-fired boiler, the NOX reductions
varied from a 60-percent reduction at full load to about a
50-percent reduction at minimum load.  The higher NOX
emissions at the 34-percent load are attributed to a different
burner pattern being used.89  For the Arapahoe 4 coal-fired
boiler, the NOX was reduced 30 percent at full-load and
10 percent at low load.  This low NOX reduction performance
was attributed to low flue gas temperatures during injection;
however,  the system is still being optimized and tested."0
     For the Port Jefferson oil-fired boiler, the NOX
emissions at a molar N:NO ratio of 1.0 were 0.17 Ib/MMBtu at
full-load and 0.21 Ib/MMBtu at minimum load,  corresponding to
NOX reductions of 48 and 36 percent,  respectively.  Higher
molar ratios of 1.5 to 2.0 resulted in NOX removals of up to
56 percent across the load range.   The NH3 slip at an NSR of
1.0 was 20 to 40 parts per million (ppm).  Further
experimentation to reduce the NH3  slip at this site is
planned.91
                             3-155

-------
2
O
O
H

e
ss
M
U
u
J
u
LO
u
z
o
li,
Oi
tn
 I
W


CQ


EH
S'iS
            l
:ontrolled

Emissions

b/MMBtu)e
          O *J
          CO «
          O, OinK>o«M>»r--*>»'0'-in
oooooooooooo
-.--.-(MrvjcMCM^vj-^s*
0000

oooo^^^^^^^^



oooooooooooo
OOOOOOOOt/lOOOOOOOOlrtUOOO


10 8
1 «

r^OMN-OKl^OOMI^O

CO


*?
^ o.
CD


1
o
c
— o
„._
o ^

M OJ •
.- _ o
3 UJ U






_,
o
00
s
*i


UJ

m
o
cc
u.
8
DC
















§
o o
Kl «-
1 1




2°



0 0
00 00


CO ^
= I



o
o
^
o» ^
•C j_
Q. a-
L.


u
>
c o
U) U
o •*-
— o

"§ 0
ft LJ






o

LLJ
u.
DC
UJ

CD
UJ

^
_J

z
UJ
z
















5
t>- co 
CM
o
o
in
0
o
N-
o




v
u
o
oo


CO
I,
3

o

0
fM

2*

UJ


(0 O
c,.-
O 4-
01 o
i, a,
O liJ

00 CO



I/)
, NATURAL G/
BOILERS
o
oe
t—
z
UJ
z



"
N Kl O
9
Y"~ *& t^
o o o
0 O O
CM  in CM

t- *> V
0 0 O
OO 00 OO
Q
Cl
u
ID
I
Kl
Kl
Kl
Kl
1? J>
UJ
& .
cS
w c
JZ O
II
V) UJ
                                  3-156

-------
n
 I
M
CO


H
          4)



          ee
trolled

mission

u)
ncontrolled

OX Emissions

Btu)c
          u ^
          re » :
          a. 01
          ro i-
          •£  -o
          § •- i
          u   >
          o ^ ro
          I z or
          H "S
          S^5





1
—
§
u
V)

i °-
(D ^^
*

Southern Ca
Edison Co.
in
O
•O ro eo

§gg
o o o
o in in
•- o o
o o o


CO O O
O*- vj rvj

o o o
jr j= j=
CO CO CO

ro
o •- o
ru %* •»

§•0 M
0 0
O 0 0
•- -»
d°o


JI in o>

000
j: .c ^
CO CO CO

= 5



o
in
^.
1«
S1 "•
CO ^
LU
.
Southern Ca
Edison Co.
5
o o co

N. in m
o o o
0 O 0
,_ O- eo
"^00
0 C3 0


in ro h-
oo -o ro

o o o
JZ -C JZ
CO CO CO

(0
n



•o
in
PJ
1«
o ^
CO ^^
LU
.
Southern Cal
Edison Co.
•
I\J fM CM K) fM CM N
9






Illfsss

o o o o o o o
^ f j: f f f j=
CO CO CO CO CO CO CO

li

co o r\i -j o in o
o •- •- PJ «- «- rj
S

X

-------





S5
O
£ |
« °
U ""
H w

ij pj
r, tH
l§
a .>,
o ^

S M
>£
1-1 D
U W
u D
J
^

°0
r i ^^


<"
1|
« °
u

•^
CO
i
n

U
CD
4^
p..













u
s
H-
V
ee
c j2
JJ,j
?j=
tit
5g<^

'g-iU
— "" 2
|l|
3 2 *^


X
as-
 00 (/)(/) C/l (/) OO



E —
o o
c o
E z
< -^
^ O fNJ CO O IA O


Kl



ro

(D ^^
CO &
0 0.
^
o
X




cd
{/>
ID
o
u
o •—
•— 0
u o>
(D — -
Q. UJ


















X
1
—
I
CO
II
J^

M
X

u c
- 0
I I \
* " JI
.? I K
? Isi § =
on z -s "* ° ""
I/) ^ O *
z ffl cj -- co
at 'c i/i a* ci o0
a. S < z .J <
n I n n - n
10
O 10 4^
<0 — 0
OJ &> Irt O *^ *^ '
i- t. r> o « a c
Q. D < Z «> <» —
4J O JO
n

E o -^
Q> O CO
i a

a o o
X £ X C
.• )_ C/> O
•D •• — 'ID
t- — u; u — 4-1
-S 2 o ~ > -S
C 4^ •§ 1-
10 C C O t- II
*-. o m £ o
V) (J > (/? U- i
(0 ,O U "D OJ I
3-158

-------





W
Q
H
X
o
(-^
C/J r \
t— j LJ


D«
F1 ^
H £
^

^
Q
§ W
> W
QJ

J °
' CD
C/D

^ J
^ <£
H a
a «
o u
y H
i E-1

^ rn

- ^
MONOXIDE
FROM CO!
CARBON
ISSIONS


0 H
OH
^^
fcj
g
W
^
LD
m
i
ro
W

CQ
^

















ox


to
L)
C
1_
111
OJ
g

"*s
i"
3f

3 QJ /-N
0 T) E
t_
4-»

Z

• 2

o

*



CJ
.^
X
o
C '"*'
o E
jj
ro
CJ





>
u
ro
Q-
ro

X D-
0 ^


^^

— g_
« S


o

^
O
CJ


o

c
0
L)
c
II
ct
T3
O
t—
i2 >-
TJ
fD
C£






£„
o ^
re *"*
CJ
re
•-I
— 1 ro
4-J
CO

i




























	 1
o
CJ
CO
o
•z.
•— '

2
CO
ce
UJ
	 i
O
cc
0
LU
ct
	 1
	 1





























Ov
CO



Ov 0
in in











1 <
'




• o
r^- m
«—




o' ro




re
QJ
L.

CO O
O  _
Sg
'i Z,
re
^




_J
re
cj •
o
C CJ

QJ £
j= O
4-J C/)

o -5
CO LU


in
0



NO ro co
to r\j f\i



.







^ D- ^-





; ; :





I ! !




to
QJ
L,

CO O O
o- -j- ro
CD
3 en
4^ CD
ft!
^
f\J
xr



tO
0 ^
	 ~o o
cn*~"
OJ
CO




^
re
u •
o
C CJ

(D C
.c o
4-- y)

O-D
CO LU


in o
ro ro
v3- r\i
2°
^
o
o
r\j in
0 <—
0
o
o
PO in
i r\j
0 •-
0
ro
E
o
1
Is

re



re


Ln
-j-

x
CO CJ

0 Q.
t- **s
t_
0


X
C
re


y) O
re cj
o
u
o •—
.— t_

•— u
o QJ
re — •
O. LU


N- ro
r\j r\J
2-
°0
^ CO
0 0
0 0
oj r\i
0 0
O 0
O 0
0 0
CM f\J

O 0
0 0
ro
CD
ZJ
Is

ro

u to
4-1 (J
re


m


ro '••»•
CD OJ
L.
O Q-
t_ N-/
U.
O


X
c
re


(/i O
re cj
u
o .-
H- 4-1
•^ (J
O O
re — •
Q. UJ
in
0
sO «— O

: : ;
in ro co


i r i
' ' (







re
3
^1 m o

re
L. t/)
^ re

re


vO
m

0 ^N


m 13 Q.
O) -w

to






ro

0
C CJ
L.
cu c
JZ O
4^ yi
3 —
O TT
CO LU



















O
II
o
4-J
ro
L.

U
L,
GJ
E
O
JZ
u

0
4J
— > QJ
ro N
o •—
CJ —
ro


O 0
to c c
IX "-
co E ro
Z 3
QJ •— O
!_ CO M-
CL.
ll C
II 3t
4-- O
OJ •- JC
i- CQ y}
CL
to
(• 4-J
• i OJ — *
-nan
i- >- yi
re H- QJ
T) L_
C — '
ro 
-------
     For the tangentially-fired natural gas boilers with urea-
based SNCR, the NOX emissions at full-load range from
0.06 Ib/MMBtu to 0.08 lb/MMBtu at loads of 80 to 100 percent.
At lower loads, the NOX emissions range from 0.03 lb/MMBtu to
0.05 lb/MMBtu.  The NOX reduction for these boilers ranged
from approximately 10 to 40 percent.   While the results varied
from station-to-station for the same  boiler type, sister units
at the same station generally achieved a similar reduction.
                                                          *
Ammonia slip for these units ranged from 6-17 ppm.
     The results were similar for the wall-fired natural gas
boilers.  Both urea-based and NH3-based SNCR systems were
tested at Morro Bay 3.   Both systems  reduced NOX emissions
levels by approximately 30 percent with an NH3 slip of 50 to
110 ppm.  The relatively high NH3 slip levels are thought to
be due to the relatively short residence times in the
convection section.
     The effect of increasing the molar N:NO ratio on percent
NOX reduction is shown in figure 3-60 and 3-61 for coal-fired
and for natural gas- or oil-fired boilers, respectively.  As
shown in the figures, percent NOX reduction increases with
increasing molar N:NO ratio.  However,  as molar ratio is
increased the amount of NH3 slip will also increase.  Further,
above a molar ratio of approximately  1.0 to 1.5, only slight
increases in NOX reduction are generally seen. Thus,
applications of SNCR must be optimized for effective reagent
use.
     3.6.1.2  Performance of SNCR on  Fluidized Bed Combustion
Boilers.  Short-term results of SNCR  on seven fluidized bed
boilers are given in table 3-36.  Two of the boilers are
bubbling bed and five are circulating bed.  All of these
boilers utilize NH3-based SNCR systems.  The NOX emissions
from the Stockton A and B bubbling fluidized bed boilers were
0.03 lb/MMBtu at full-load.  The NOX  emissions from the
circulating fluidized bed boilers ranged from 0.03 lb/MMBtu to
0.1 lb/MMBtu at full-load conditions.  The average NOX
emissions from these five boilers were 0.08 lb/MMBtu.
                             3-160

-------
o
o
o
00
                   I
                  o
            I
            o
I
o
in
I
o
I
o
            o
            (M
1
O
                                                     o
                                                     •H
                                                     4J


                                                     Ll


                                                     O
                                                                 0
                                                               cc

                                                               c
                                                               1C

                                                               o
                                                                                    O
                                                                                   •H

                                                                                    U
                                                                                    3
                                                                                   •O
                                                                                 -H  *
                                                                                 
-------
XON 3U3OJSd
                                                            c
                                                            o
                                                           •H
                                                           iJ
                                                            u
                                                         C  (V
                                                         O  H
                                                        •H
                                                        JJ  U
                                                         C-H
                                                         0) 4J
                                                         >  >,
                                                         C rH
                                                         o  rt
                                                         t> JJ
                                                            (0
                                                         M  U
                                                         o  c
                                                          >
                                                         03 -H
                                                         >-! 4J
                                                            U
                                                        O   £1

                                                        C T3
                                                        O 0)
                                                        •H M
                                                        0-) -H
                                                        U iw
                                                        3  i
                                                        •a -H
                                                         x-o
                                                        o c
                                                        S m
                                                           w
                                                           
-------
Q
CQ

 ro
ffin
   W

   H
   UD
   >H
   03
W
O
m

Q
w
   o
   H
   EH
   u
Q
J
fc
   H

   EH
  U
  X
  o
  w
  W
ro


W
J

CQ
<
EH
(U
- L.
•M 73 -C
•- OJ ^-
u *-• 3
ro :/* 4-1
Q. (D CO
to t~ z:
o z:
L. O
ro o •-
o •*-, ro
Z: Z Ct:
~>- t-
73 .^ ^


^ D. CO
* ro z:





E i-





^

^

t;





U




1



ro
o
o



o
c/5



o
00
r\j



,



0
CO



c/)



0



OJ

o
o.
u>
UJ
o •
t/) Q.
o •
CL _l

~*




|




O
o



L-
o
SL
in



o
CO
f\j







o
CO
r\j

m

o o
•w CO
-^ u_
0 CO
o ^^
I/)



o
o


01
3
o
u_
UJ
o •
W Q.
o •
O. —i
in
>»




1



CO
0
o



0
(/)



o
CO


o
^!



o-
co
K1

c
'!
tD

CJ
O CO

t_ ^^
CO
o

OJ
CJ


C c
O C)
• r— (/)

cj ro D)
3 — C
o — < -^
CL OJ -t-1
ro +^ co
t- c/) t-
•M C CD
— 0 D.
DUO
in
•*




\



CO
o
0



c_
o
x:
 CO
ID U-
CO ^^
0
C£
V)
CJ
u


C 1-
O fl>
•F- t/>

tu to en
3 — c
0 ~ —
CL CJ ^
to ^ to
t- v? u
•M C OJ
— • O Q.
=300
r*1




ro
s



fO
o
o



•4-J
o
x:
CO



o


r\j
ro



0
r\i
•0

C
at
en
o
o /->

C CD
0 u-
o




tf)
E
QJ
4-*

CO


L.
OJ
c
UJ

"•*




1




*-
o



o
.c




CO



,
1




CO

V
c




0) ^
XI
f_
cc
CD





Dl
U


UJ
*O OJ
o> o

_rf ^>
Q. t_
Q. 0>
< CO
CM
"*




!




«~
d



L.
o
-C
(0



§
CO



,



ro

CO

CD
JjV-J
c




l_
TO
CD





en

0)
c
UJ
TJ 

JI >
&t_
Ot
< V)
                                                           T3
                                                           
m
U-
co



CJ
a
>-

jj
1
o
JC
-



V
o
-C
C/)

II



o
JL:
t/)
o









CJ

JD
CO




to

o
c

ro
4^
ro
o

II


1
                                      3-163

-------
     Long-term CEM data are also available on these units.  A
summary of the long-term data was presented earlier in
table 3-3.
3.6.2  Selective Catalytic Reduction
     3.6.2.1  Commercial Scale SCR Installations.  Table 3-37
presents the performance data on commercial size installations
of SCR systems on three coal-, one coal-/gas-, and one oil-
/gas-fired boilers.  The data on two of the coal-fired 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 plant equipped with an SCR.  The
plant has two coal-fired boilers nominally rated at 285 MW.
The boilers burn a 2 percent sulfur, eastern bituminous coal.
Besides the SCR system, each boiler is equipped with LNB and
an AGFA system to control 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 V2C>5, TiC>2,  and WC>3.  Aqueous NH3,
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 levels) with an NH3 slip of 5 ppm (at
7 percent 62)•   Permitted NOX levels are not to exceed
0.17 Ib/MMBtu (3-hour rolling average) .g7-98
     As shown in table 3-37, for both boilers, NOX emissions
(4-hour averages)  at full load were below 0.15 Ib/MMBtu with
NH3 slip levels below 0.3 ppm  (at 7 percent 02).  No adverse
effects on the catalyst or on downstream equipment were
reported during inspections following the acceptance tests.98
     The Logan Generating Plant is a 202 MW cogeneration
plant.  The boiler burns low-sulfur (1.5%), eastern bituminoxas
coal.  The combined use of LNB, OFA, and SCR minimizes NOx
emissions from the unit.  The SCR, located between the
economizer and the air preheater, consists of a plate-type
catalyst.  The catalyst is made of V205 and Ti02.  The plant
utilizes agueous NH3 with an NH3 content less than 27 percent
                             3-164

-------
o
H
E-<
U
t>
Q
W
O
H
s
O

w
>
H CO
H H
U S
W <
J J
W &j
CO
   >>
W E-i
J H
< J
U H
a D
W
o
U CO
O


W
   CO

   >H
   CO
o
CQ
OJ
u
c
OJ
a
OJ
ex
c
o
O L) S-S
Z -0 ^
01
0 "5
Z 4J
00
4J ^
QJ £
— ' ^-^
4^ ja
3 —
o ^
ro
§ -^ §L
i in 5-
0
-C y, "^

C "^ *"*
OJ
	 1
4->
5 en P-
c/) re r"
c "
L.
C' QJ
3 4-J «-^
C j£ ™~o i
3 * — \
£ '^£
*- -o
"~ ^ y-v
CO '/ ^
Q. GJ ^
CO *-•
OJ
CL x *—
X J2 *J

c/1 D)

QJ f£ CJ
3 *-^ 3
u.

X
£'o S
m n

co.


^.

L.
4-* ro
•- T3
C C
=> ro
«
^

X

.,_


=




co co co co eo co co




:',:;;;!



ro ro in ro T- o ro
o o o o o o o



•«O >O *O in --J" ro in
ro ro c\j ro ro ro o
o o o o o o o

*J -J- ^3- --T --J -^- 
ro o c z
[_) CL ^ ^-^
Dl
C

4-- X
ro c
i- ro
• oj Q.
0 CJ U



00 CO CO CO CO CO
o &• o o o o




, 	



c r^ t^- *o ro ru
to r-o ho ro ro -J"



o o o o o o
0 0 0 0 O 0

zi
o
c
• r-



•— O (Nl
CO O N- '


m

ru

*~
ru ^^
ro
« o
X Z -
Q) 4-J t/3
C C 4- Q.
ro o c z
U Q- Z3 ^
O)
c

•*-" X
ro c
l- (0
• (U Q.
Crt C E
• QJ O
ZD CJ O



^




5



5
o



0
o

!


eo
o ro
OJ 3:
3 z
cr
*


o
1


o
o

3 in
o
C —



4J re
••- D
CO o




ro
o
ro


• re
O Q
z *
c e/>
ro 4J o-
01-- e/>
o c 2

en
c

4J X
ro c
i_ ro
• o» ex
wee
. 0) O
ID 0 U



e> e> e>
o o o




u u "o
§eo -o
S- e>-



0 0 TJ
O O •£>
O O O



U o T3
O ^ ^


1


3
O ro
OJ 3;
3 Z
cr
"




;

O 0 O
o o o

to
o
c
'.—


4-j re '
• — O i
co e_j N->



ro


ro
ro ^

• o.
cow
I- O Z Z
OJ • —
 C D.
Z W D ^
o«

u
QJ •- X
out- c
•— •— 4~i ro
-- > U D.
O t_ Qj (/J E
ZJ QJ — ' (0 O
Q. t/) UJ O LJ



8:8:8:




U U TJ
ru oo ^*
eo m o



u u TJ
o o CD



0 U TJ
O y^ ^


i


zi
o ro
QJ x
Z3 Z
cr
*




1

0 0 O
0 0 O







t/3
ro




ro


fO
ru *-*
(^}
• Q.
C O (/>
c- o z z
QJ -^
O 4-1 4-J Q)
^ ro •-- L_
OJ *^ C D.
Z O? Z3 ^->
ad

(J
 (J Q.
^3 L QJ y) E
zi o — • ro o
CX. C/) UJ t3 U



o
o




o
o*



,
1


g_
0


!


z
TJ

^_
-

-J-
0



o
o


o

ro
o m
•^ ru *^

•— O "-
o *** o




in

ru
C ru ^^
O t/)
*-> • CL
Ol O {/)
c z z
4-J O 4-* QJ
C CD •- (-
D QJ c a.
Z CO Z3 ^
ro

c c
L- C-
QJ O C
JZ <*- O
o ro T3
OO O UJ
ro
a
ro
u

u
.2!

o
JD

H-
O

C
GJ

D
t/3
CD
QJ
E
a

(/i

ro
z
QJ"
c
o

ro
CL
CJ
00
•4-J
U
D


C
O

QJ

ro
jr


,,_
ro

4_j
CD
-Q
QJ

Z3

ro
1
a


Cfi
ro
^r~



oc
CJ
V)

L.
QJ

flj
JC.
L_

ro

+
Q£
CJ
to
+-J
u

TJ

c
                                                  3-165

-------
by volume.  The system is designed for 63 percent NOX removal
(0.1 Ib/MMBtu outlet NOX level)  with an NH3 slip of 5 ppm or
less (at 7 percent 02).   Permitted NOX levels are not to
exceed 0.17 Ib/MMBtu.  As shown in table 3-37,  typical NOX
emissions at full load are about 0.13 Ib/MMbtu with no
detected ammonia slip.117
     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 alone, 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 level of 0.2 Ib/MMBtu (approximately 88 percent reduction
for coal) .QC<
     Results shown in table 3-37 indicate that the in-duct SCR
system achieved NOX emissions of 0.20 Ib/MMBtu for coal-firing
and 0.11 Ib/MMBtu for gas-firing at NH3 slip levels of 10 ppm
(measured at the inlet to the air heater for both fuels).
Higher NOX emissions of 0.30 and 0.27 Ib/MMBtu for coal and
gas were measured at NH3 slip levels of 5 ppm at the air
heater inlet.  The addition of the air heater SCR decreased
NOX emission to 0.04 to 0.06 Ib/MMBtu at less than 5 ppm NH3
slip at the air heater outlet for coal- and gas-firing.
     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
                             3-166

-------
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-firing.  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).   The remainder was on
natural gas.100
     The results indicate that during the initial 2,000-  ,
7,000 hours of operation, 90 percent NOX reduction was
achieved at about 14 ppm NH3 slip.  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.
     3.6.2.2  Pilot Scale SCR Installations.  Data on pilot-
scale installations of SCR on two coal-fired boilers101'102 and
one oil-fired boiler103  have  also  been  reported  and  are  shown  in
table 3-38.  The SCR pilot plants process a slip-stream of
flue gas from the boiler  (equivalent to 1 to 2 MW).  Each
pilot plant contained two different catalysts that were
evaluated simultaneously.  The data shown in table 3-38
reflect the performance of these SCR systems as of March 1993
(approximately 1-3 years of operation).
     Results from the coal-fired SCR demonstration projects
indicate that 75-80 percent NOX reduction was achieved with
NH3 slip of less than 20 ppm.  The lower NOX reduction and
higher NH3 slip for the oil-fired demonstration at the Oswego
site were measured at higher-than-design space velocities.
Note that these results are on pilot facilities in which
operating and process parameters can be carefully controlled.
     The effects of catalyst exposure time and space velocity
on catalyst performance were also examined for each of the
pilot-scale demonstrations.   Figures 3-62a and 3-62b show NOX
removal and NH3 slip as a function of NH3:NOX ratio for two
catalysts in a cold-side, post-FGD SCR demonstration at the
Kintigh site.101  This pilot  demonstration  began in  December
                             3-167

-------













r )
*• — •*
H
E-H
J
s
u
Cd ^
»> ^
^
^ >H
W r,

SCALE i
. UTILI
L_i U)

O r- J
r-l '
H g.

C
tu ,r
°2
ryi W
r i E*"1
^ CO
s K^J
S rn
g UJ
rt g
t^-* , 	 j
DH Pj
W r~^
Q_i
1-4 j--,
Q
. w
m

rn

["T"]
*1
m







jj
O
E





, — i
rd
0
u




f\\
Shawnee







I
tt)
J^j
1C
0) >,
0) JJ
W -H
rn LJ
Tennei
Autho:


n
o
rH








co co
o o



Cnx!
0 O
CN in
0 0
•J— ' 4-J
a a
D D



oix;
0 O
co co
in o
r- VD






jj
O
E






rH
•n
0





o
01
0)
s
CD
o









•^
<0
o
(0

n)
Oi
-H
tt)
O
MH
tt)
X3

^
a
u
CO

^
w
T)

01
-rl
tt)
T>
-H
tn

jj
o
E
II
JJ
O
E

T)
C
(C
tt)
1 1
(0

JJ
tt)
^
tt
-H
10
rl
0)
JJ
4H
n)

K
U
CO

Q
O
fc
jj
tn
O
tt
~
tt)
•O
•H
tn
i
•o
rH
0
U
II rl
D
T3 JJ
rH rtj
O 0)
U £
tt)
^H
•• tt
D
H *
10
































n
jj
tn
j>,
rH
fO
JJ
rd
U
JJ
C
tt)
^
tt)
IM
M-l
-rl
•c
O
3
JJ

rl
O
MH
C
D
. ,-j
Di
tt)
rl
tfl
tn
jj
rH
D
tn
tt)
a:
X!























tn
^
o
o
0
CO
I~-

II

tt)
E
-ri
JJ
0)
iH
3
tn
0
Qj
^
d)
jj
tn
j>,
rH
(0
JJ
(0
u
rH
I
rl
f.

o

n n

JJ JJ
-rl -rl
U U
0 0
rH rH

tt) tt)
U U
10 n)
ft ft
CO CO
Oi £
3-168

-------
  100


   90



1  8°


   70
X
O
   60
   50
                           Ttmenbotl
                          Pvtotmtnse  •'
                                      NOx Removal
                                                   3.0
                                                   2.S
                                                   2.0
                                                      I
                                                   1.0  9


                                                   0.5  §
                                                   0.0
    0.5    0.6    0.7    0.8    0.9    1.0    1.1    1.2
                      NH3-to-NOx Ratio
     Figure  3-62a.
                     Extruded catalyst NOX
                     conversion  and residual NH3
                     versus NH3~to-NOv Ratio
                      (Kintigh site). *
  100
   90 •
«  80 h
 x
O
   70
   60
   50 L-
                                     NOx Removal
                                    SfpmStpO WpfXTiNC*..
        Thtontxal
  3.0


  2.5


  2.0


  1.5


  1.0


  0.5
                                                       w
                                                       CO
                                                   0.0
     0.5
           0.6   0.7    0.8    0.9    1.0    1.1
1.2
                      NH3-to.NOx Ratio
      Figure 3-62b.   Replacement composite
                       catalyst  NOX conversion
                       and residual NH3  versus
                       NH3-to-NOx Ratio
                       (Kintigh  site) .m
                       3-169

-------
1991.  The data shown in figure 3-62 reflect the performance
of the SCR system as of March 1993.  The results show no
significant loss in the activity of either the extruded
catalyst  (figure 3-62a) after 7,800 hours of operation or the
replacement composite catalyst (figure 3-62b)  after
2,400 hours of operation.  Each catalyst reduced NOX emissions
by 80 percent at a near stoichiometric NH3:NOX ratio with an
NH3 slip of less than 1 ppm.101
     The original composite catalyst was replaced after
5,200 hours of operation because of significant deterioration
in catalyst performance and excessive NH3 slip.  An
examination of the catalyst revealed that catalyst poisons had
penetrated throughout the active catalyst surface layer.  The
catalyst vendor indicated that the catalyst was manufactured
using titanium with an incorrect size distribution, which was
believed to have contributed to the catalyst poisoning.
     The original extruded V/Ti catalyst has been exchanged
with a new extruded V/Ti formulation designed for higher
catalyst activity.  The reformulated higher-activity catalyst:
is undergoing baseline testing.  The original catalyst was
moved below the new catalyst layers to allow for continued
exposure beyond 1.5 years.1M
     The hot-side/high-dust pilot at the Shawnee site was
operated between May 1990 and May 1994.  During this period,
the SCR unit operated for approximately 22,000 hours.  Two
catalyst formulations, an extruded V/Ti and a zeolite catalyst
were tested.  Activity curves for both catalysts are shown in
figure 3-63.   (A 20 percent decrease in catalyst activity
[K/Ko = 0.8] translates to about a 10 percent decrease in NOX
reduction potential for an initial NOX reduction of
80 percent.)  Over time, both catalysts exhibited significant
deactivation from pluggage of catalyst channels and masking of
the surface by sulfated ash.  Analysis of ash samples
indicated that the ash was becoming enriched with sulfur in
the presence of ambient moisture during the frequent pilot
                             3-170

-------
  100
               3600 Hours
               7700 Hour*
               14000 Hours
    50
     0.5
        0.6
0.7
 O.B    0.9    1.0

NH3-to-NOx Ratio
1.1
       Figure 3-63a,
                     Relative changes in V/Ti
                     activity with  exposure  time
                     (TVA,  Shawnee  site)10*
   100
    90
O
O
oc
X
o
z
    80
70
    60
           1500 Hours
           4500 Hours
           5500 Hours
           7700 Hours
                             Thtontcal
                             Ptrtomnanct
    50
      1.5    0.6    0.7    0.8    0.9    1.0

                        NH3-to-NOj< Ratio
                                         1.1
                                                      J2
                                                      •o
                                       a
                                       5
                                       &
                                                          O
                                                         M
    Figure 3-63b.   Relative changes  in zeolite catalyst
                     activity with exposure time
                     (TVA,  Shawnee site)10*
                        3-171

-------
shutdowns.  Reaction between this moisture and the alkaline
constituents in the ash resulted in the formation of hard
deposits that plugged catalyst channels causing a loss in
catalyst activity.10*
     The original V/Ti catalyst (figure 3-63a) was tested for
the entire pilot operating duration of 22,000 hours.  Samples
taken from the first catalyst bed exhibited a higher activity
than those from the third bed.  This is believed to be due to
                                                          ?
sootblowers located above the first layer.  A replacement V/Ti
catalyst with a different hardness (softer)  value than the
original catalyst was tested near the end of the test program.
Although, the replacement catalyst shows a lesser rate of
deactivation, the results are not conclusive because a lower
sulfur coal was fired during this test period.104
     The original zeolite catalyst failed to meet its
performance specifications after only 5,000 hours of operation
and was replaced after about 12,000 hours with a reformulated
zeolite design from the same vendor.   As shown in
figure 3-63b, the reformulated design showed less deactivation
than the original catalyst.   Figure 3-63b also shows that
water-washing the original zeolite catalyst resulted in a
significant improvement in its activity.
     Figures 3-64a and 3-64b show the performance results for
a composite V/Ti and a plate V/Ti catalyst evaluated on the
oil-fired boiler at the Oswego plant."  The Oswego SCR pilot
unit was operated between October 1991 and October 1993.
During this period, the SCR unit operated for approximately
4,800 hours.  The data shown in figure 3-64 reflect the
performance as of March 1993.  The curves show the effect of
space velocity on NOX reduction and NH3 slip as a function of
NH3:NOX ratio.  The results show the expected decrease in NOX
reduction and increase in NH3 slip at the higher space
velocity for both catalysts.103  Catalysts  activity  changes were
measured by both catalyst vendors on samples taken after
2,400 and 4,100 hours.  For the corrugated plate catalyst, the
                             3-172

-------
     120
     100
! Space Velocity
  17,400 1/hr  O
- 4,350 1/hr  •
       0.00
      0.20
0.40
        0.80
        1.00
        1.20
 Figure  3-64a.  Titanium corrugated plate catalyst
                 NOX  conversion  and residual  NH3 versus
                 NH3~to-NOx ratio.   (Oswego Site)
           27,400 1/hr O
            6,900 1/hr •
       0.00
      0.20
0.40
0.60
0.80
1.00
1.20
Figure  3-64b.
        Vanadium titanium extruded catalyst
        NOX conversion and residual NH3ioyersus
        NH3-to-NOx ratio.  (Oswego site)
                          3-173

-------
relative activity increased during both intervals and exceeded
the original activity by 24 percent by the end of the test
program.  The increase in activity was attributed to the
deposition of vanadium from the boiler on the catalyst
surface.  The vanadium content of the fuel oil was 55 to
170 ppm during the testing.104
     The activity of the top layer of the composite V/Ti
catalyst decreased about 20 percent after 4,100 hours.  The
deactivation was attributed to masking by a thin layer of
solids on the catalyst surface and pluggage of the catalyst
channels (approximately 50 percent overall pluggage) .1CM
     3.6.2.3  Recent 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, subpart Da.  Only those decisions that involve
the installation of SCR systems are discussed in this section.
     Table 3-39 lists new utility boilers subject to the NSPS,
subpart Da that are either already operating with or are
expected to be built with SCR for NOX emissions control.  The
list  includes six coal-fired boilers, two each in New Jersey,
Florida, and Virginia.  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.
     The two PC-fired plants in New Jersey (table 3-39)  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 C>2  (3-hour average).  However,  the
permit conditions require that the SCR systems be designed for
outlet NOX emission levels of 0.10 lb/MMBtu  (3-hour average).
                             3-174

-------
o
H
E-"
U
C3
O
w
U
H
EH
>H
J


s
U

w
U
W
J
w
U)
H
CO
D


CO
o
m
H
J
H
o

H
w
H
(Ti
m








(/I
c
o
4-J
(J
4->
1
CL










c




































(U
11

Oi
01
on
.£• ~
^ E
0> Q
2 ^

'01 °
ro £
I"

-Q ~
•— C 4-1

~" QJ -^
XT3 -D
0 ^ —
Z "-'
QJ
L- C
CO • —
QJ —
>- C
o
0 i- C
QJ 3 QJ -—^
X— L. -C
t— 3 CJ O)
U^ Cj •—
1- QJ
CJ — - - "3.
— • O 4-J
- 3 - X
0 u. = jQ
CD ^- —
*J QJ ^-C
C •- E OJ
Z3 C/5 v-' Qj
•M C/)
C ^
3 O
z
CJ
ro
to




QJ x-v
E CJ
co E
Z CO
x*
CO C

o o.
LJ -**s












in
o




u
o


o
.c



^
r^ o
o o



o
o





(_> ^~
d ,J




in
CM

oa



z

"D
QJ

"~ 4-J
E C
_l O
O-
c
O tf)
4^ QJ
ro c
U L.
QJ CD
C LJ
Qj \-s
0)
o a
U "-

1/5 s/l
L- L-

~Q C

CO t-
^ ra
CJ a.

so h*
0 0
*— «—



o
o


L_
3
O
r^


^^
h- 0
o o


•J-
o
o-





o i"
°- ^





r\j
0
rvj

«~



^



C
ro
E~
O -M
u c
ro
OJ — -
0 0-
> 01
1- C
C/l 4J
ro
X t-
Ol 01
1- C

C CO
LU
C
c ro
ro Ol
01 o
O -1
_J ^^

CO
0
•—



•D
o


•8-1
M I-


^^
N- 0
0 0



o
o





0°
^ r\j





sc

r\j



	 i


C
O

l/>
o
o

5

II—


c
o o
T> 4J
C C
ro ro

o ~^

o
o
*~



o


3 ^
0 0
£ O
ro



|v-
o



o
o





u °
11 rxj





o

*—



	 i

c
ro


c
>~ o
c •-
fto
01
CJ C
OJ
O) Ol
C 0
•- u

ro c.
c. 3
OJ O
C 4-<
01 C
CD ro


vi c
='C

o
*—
*~



in


If
° °


^^
m o
o o



§
o





CJ ~
CL ,J





O
ro
f\j

*~



g

-o
OJ
4->
ro
u
I/! O
i- a
C 0
•M (J
L. C
ro —
a.

OJ O
3 O
O 3
CL -C
0
T) L.
O •-
O CO

_£; k
U ^-
[_ UJ
.- in

o
T—
r~



in
fVJ


X
•8
o



U-l
o


0)







CO
en




u>
fc
CO




§





01
ro
o
a.
i_
o
u
c



X



c
Ol

o
LJ










Ol
C
o
L.


^^
0
o










^
^

















1
5
T3
C

°

Qj
X *->
- CO
L. l_


Q> t—
Dl O
O 0
U C
                                                    OJ

                                                    CL
                                                        TJ
                                                        OJ
                                                        C
                                                        o-
CO
0
u

T3
QJ

.,_
£_
CJ
>

i

n

u
CL
(0

f
L)
.

OJ
3


Q_
Q.

O

ro
w


c

TJ
4^


O
c

c
o

4^
o
i.
4->

C
o
a
0)
                                     3-175

-------
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
NOX levels that are demonstrated to be continuously achievable
over the first 5-year operating period.105'106'107              ,
     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) .10S   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.111  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.109
     Finally,  Virginia has issued permits for one PC-fired
boiler and eight SS boilers with perir.it 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.110
     The SCR systems for the eight SS boilers in Virginia must
be designed to achieve a NOX emission level of 0.10 Ib/MMBtu
                             3-176

-------
 (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
 slip level is limited to 25 ppmv and fuel sulfur is limited to
 1.1 percent by weight per shipment.110
     3.6.2.4  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.m  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 continued 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
 lease 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 S02  by  90 percent or
more and a reverse-air type fabric filter baghouse removes the
particulate matter.158
     The SCR unit is located between the economizer and air
                             3-177

-------
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
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 02) .98   •
     3.6.2.4.1   Carneys Point Unit 1.  Continuous emission
monitoring data were obtained from U.S. Generating Company for
Carneys Point Unit 1 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 3-65.  Typical hourly NOX
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 3-40 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).50  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 3-40.  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.
                             3-178

-------
                                                                                        in




                                                                       in
                                                                       CM
                                                                       o>
                                                                       35
                                                                       I

                                                                       I
                                                                       CM



                                                                       i
                                     O)



                                     CO



                                     in



                                     (N



                                     00
                                                                       L?5

                                                                       O


                                                                       03
                                     92
                                     in
o
o
                                                                                        o
                                                                                        m
                                                                                        o>
                                                                                        0)
                                                                                        CO
                                                      >i
                                                      H


                                                      I'D




                                                      O
                                                                                         (0
                                                                                         C
                                                                                         o
                                                                                        •H
                                                                                         0)

                                                                                         xo
                                                                                         O 2
                                                                                         3
                                                                                         O
                                                                                         M-t
                                                                                         o  u
                                                                                         rH  >i

                                                                                         a c
                                                                                         0)  M

                                                                                         J53
                                                                                         EH —
                                                                                         in
                                                                                         vo
                                                                                         i
0)
^
3
                                           3-179

-------
     TABLE 3-40.
SUMMARY STATISTICS FOR CARNEYS POINT
      UNIT NO. I112
             (24-HOUR BLOCK AVERAGED NOX DATA)
               Statistic
                      As measured
                         data
Transfer
    data
  (X=7.0C
No.  of  observations
Mean (x24)  [Ib/MMBtu]
Standard  deviation  (324)  [Ib/MMBtu]
Skewness3
Kurtosis3
Shapiro-Wilk W  statistic9
Probability < W
Autocorrelation coefficient*3 (PI)
Correction  factor
                          92
                           0.137
                           0.007
                          -1.253
                           4.639
                           0.886
                           0.0001
                         NEC
                         NEC
aFor a perfectly normal distribution skewness
 kurtosis = 3, and W = 1.
                             =  0,
      lag 1 is significant.
CNE = not estimated.
                           3-180

-------
     Table 3-41 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 3-40 and the
statistical procedures described in section 3.2.3.2.  As
indicated in table 3-41, 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.                                                  •
     3.6.2.4.2  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 3-66.  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 3-42 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).5"  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 3-42.  As shown in the table, the absolute values of
skewness and kurtosis are within the 95 percent confidence
limits for 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
                             3-181

-------
     TABLE 3-41.  ACHIEVABLE NOX  EMISSION  LIMITS3 FOR
                      CARNEYS POINT UNIT NO. I112

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 -

-------
                                               in
                                               o\
                                    re
                                    O
                                               O
                                               n
                                               0)
                                               Q)
                                               CO
                                               >s
                                               rH
                                               3
                                               in
                                               c
                                               o
•r-t ^»
 g (N
 0)

 xd
o *.
                                               H E)
                                               3
                                               O  -
   (X
                                              «M

                                               0
                                              .p
                                               o
                                               a o>

                                               0) >~i
                                               E (0
                                              •H U
                                              El ^
                                              V0
                                              vo
                                               I
                                               Q)
                                               b
                                               3
                                               Cn
                                              •H
3-183

-------
    TABLE 3-42.   SUMMARY STATISTICS FOR CARNEYS POINT
                        UNIT NO.  2
             (24-HOUR BLOCK AVERAGED NOX DATA)
Statistic
                                        As  measured
                                           data
No. of observations
Mean  (x24)  [Ib/MMBtu]
Standard  deviation  (524)  [Ib/MMBtu]
Skewness3
Kurtosis3
Shapiro-Wilk W  statistic3
Probability < W
Autocorrelation coefficient*5  (PI)
Correction  factor
                             85
                              0.131
                              0.015
                             -2.375
                              9.629
                              0.716
                              0.0001
                            NEC
                            NEC
3For a perfectly normal distribution skewness
 kurtosis = 3, and W = 1.
                                =  0,
      lag 1 is significant.
CNE = not estimated.
                           3-184

-------
confidence limits (0.61) 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 3-43 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 3-42 and the
statistical procedures described in section 3.2.3.2.  As ,
indicated in table 3-43, 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.
3.7  ADVANCED CLEAN COAL TECHNOLOGIES
     Several advanced technologies are being demonstrated
through DOE's Clean Coal Technology Program (CCT).   Of these,
three technologies that include NOX reduction are described in
the following sections.
3.7.1  The SNOX™  Process
     The SNOX™  technology  includes  five process elements:
particulate collection, NOX reduction, S02 oxidation, sulfuric
acid condensation, and acid conditioning.113  Heat addition,
transfer, and recovery also represent a significant portion of
the SNOX™  system.
     As shown in figure 3-67,  flue gas leaving the air
preheater is treated in a particulate control device and
passed through the cold side of a gas/gas heat exchanger (GGH)
raising the flue gas temperature to above 700 °F.114  A mixture
of NH3 and air is added to the flue gas prior to the SCR where
NOX is reduced to N2 and water.  The flue gas leaves the SCR
and, after a slight temperature increase, enters the S02
converter which oxidizes S02 to 803.  The 803-laden flue gas
then cools as it passes through the hot side of the GGH.  Flue
gas exists the hot side of the GGH and enters a falling film
condenser where 803 and water in the cooled flue gas condenses
to form marketable sulfuric acid.
                             3-185

-------
     TABLE 3-43.  ACHIEVABLE NOX EMISSION  LIMITS* FOR
                      CARNEYS POINT UNIT NO. 2112

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.
                            3-18 6

-------
                                        u
                                        
-------
     The SNOX™  Process was  demonstrated  on  a  35 MW flue  gas
slipstream from Unit 2 of Ohio Edison's Niles Generating Plant
in Niles, Ohio.   The treated flue gas was extracted prior to
the electrostatic precipitator (ESP).  Initial results from
the demonstration were compiled after two months of operation
of the SNOX™ plant.   As  such,  the  results are considered
preliminary and only provide an indication of long-term
performance.
     As shown in table 3-44, the results indicate a NOX
removal efficiency of 94  percent and a S02 removal efficiency
of 96 percent.112  Early indications showed that the SNOX® plant
used approximately I percent of the unit's power production
for complete operation.   The operator expects NOX removal
efficiency can be increased to 95 percent by adjusting the NH3
distribution into the system and that some enhancement of S02
removal will be achieved  by other process adjustments."1
3.7.2  The SO^-NO^-RO^ Box  (SNRB™) Process
     The SNRB™  process consists  of injecting  NH3  and  either a
calcium- or sodium-based  sorbent upstream of a
high-temperature baghouse, which contains woven
high-temperature bags and a SCR catalyst.115  The selection  of
either a calcium- or sodium-based sorbent determines the
optimum operating temperature and,  therefore,  the arrangement
of the system relative to the boiler.
     A schematic representation of one proposed commercial
arrangement of the SNRB™ process is  depicted  in figure 3-68.115
In this version, which features a calcium-based sorbent,
commercial hydrated lime  is injected into the convection pass
of a boiler upstream of the economizer, where the flue gas
temperature ranges from 900 to 1,100  °F.   Simultaneous
dehydration and sulfation of the sorbent begins immediately
upon injection and continues as the flue gas passes through
the economizer,  fluework, and into the baghouse.   The reaction
products (CaS03 and CaSO^J along with the fly ash and
                             3-188

-------
 TABLE  3-40.   THE  SNOX™ PROCESS COAL-FIRED UTILITY BOILER
                  DEMONSTRATION AVERAGE TEST RESULTS112
System Load
Inlet NOX
Inlet S02
Outlet NOX
Outlet S02
H2504 Produced
5,680 Ib/min
  616 ppm
 2,056 ppm
   35 ppm
   88 ppm
28 tons/day
                          3-189

-------
                                                      u
                                                      m
                                                      e
                                                      Q)
                                                     .c
                                                      u
                                                     w

                                                      M
                                                      (0
                                                      0)
                                                      u
                                                      o
                                                     CO
                                                     «

                                                     CO
                                                     CO

                                                     VO
                                                     0)
3-190

-------
unreacted sorbent are collected as a filter cake in the
high-temperature baghouse.
     The baghouse operates in the temperature range of 700 to
850 °F and employs high-temperature ceramic or fiberglass
bags.   S02 reacts quickly with the hydrated lime following
sorbent injection into the flue gas and continues to react as
the flue gas passes through the filter cake.  The overall SC>2
removal is designed to exceed 70 percent.
     The SNRB™  process  utilizes  SCR for  post-combustion  NOX
control.  A vanadium-free variant of a commercially available
zeolite catalyst is being used in the CCT project.  Use of a
zeolite catalyst and incorporation of the catalyst downstream
of both SOX and particulate removal is intended to minimize
(1) catalyst deactivation by heavy metals and sulfur species;
(2) ammonium bisulfate formation and subsequent deposition;
(3) catalytic oxidation of SC>2 to 803 and subsequent equipment
corrosion due to the 803 generation; and (4) erosion or
pluggage of the catalyst by fly ash.  NOX removal upstream of
the combustion air preheater eliminates the need for a flue
gas reheat system to provide the appropriate gas temperature
for NOX reduction.
     The SNRBTM process was demonstrated at the R. E. Burger
plant near Shadyside, Ohio.  The demonstration facility was a
5 MW slipstream pilot employing commercial - scale filter
bag/catalyst assemblies.  Approximately 700 hours of operation
at baghouse temperatures of approximately 700 to 900 °F were
completed during the first few months of operation.  Most of
the testing was completed at an air-to-cloth ratio of 4  (the
ratio of the flue gas volumetric flow rate to the filter
fabric surface area), a NH3:NOX molar ratio of 0.8, and a Ca/S
stoichiometry of 2.
     As shown in figure 3-69,  initial NOX reduction
performance ranged from 90 to 97 percent for the majority of
the test runs."4  Ammonia  slip was  less than 10  ppm with  most
of the measurements below 5 ppm.   The performance data are
                             3-191

-------
                                             8
                                             o>
         03
         O
        CD

         o
         Si,
        'o
        (L
                                          to
                                          (A
                                          0)
                                          U
                                          O
                                          CO
        O
        O>
                                                0>
                                                A)
                                          o

                                          Q)
                                                80)
                                                fc
                                                s
                                                0)

                                                0)
                                                M
                                                X
                                          O

                                          i
                                          r-l
                                          (0
                                                                   o
                                                                  +J
                                                                  o
                                                                  0)
               \r>
               CO
o
eo
                     o

                     (O
[•4]  (iAOiuaa XON
                                                                  vo
                                                                   i
                                          0)
                                          M
                     3-192

-------
preliminary, but provide an indication of future performance
potential.
3.7.3  The NOXSO™ Process
     The NOXSO™ process is  a  dry,  post-combustion flue gas
treatment technology that uses a regenerable sorbent to
simultaneously adsorb S02 and NOX from flue gas.  In the
process, the S02 is reduced to elemental sulfur and the NOX is
reduced to N2 and ©2 -
     The NOXSO™ process flow  diagram is  shown  in
figure 3-70.116   Flue gas from  downstream  of  the ESP  is  ducted
to two flue gas booster fans.   Downstream of the booster fans,
the flue gas is cooled by spraying water directly into the
ductwork to maintain the adsorber inlet temperature at 300 °F.
After cooling the flue gas is passed through two parallel,
two-stage,  fluidized bed adsorbers where S02 and NOX are
simultaneously removed using a high surface area ^-alumina
sorbent impregnated with an alkali material.  Tail gas from
the sulfur plant is injected between the two adsorber stages
to increase the ratio of S02:NOX and increase NOX removal
efficiency in the second (upper)  bed.  The cleaned flue gas
passes through a cyclone separator, returns to the plant
ductwork,  and exits through the chimney.   The cyclone returns
entrained sorbent back to the adsorber.
     The NOXSO™ process is  scheduled to  be  demonstrated in
1997 on a 150 MW pulverized coal-fired boiler at Alcoa
Generating Company's Warrick Power Plant, Unit 2.118
                             3-193

-------
£ 1 I
It fc
? u.
E I!

b
*.,. xf

                                                o
                                               •H
                                               4J
                                                o
                                                M

                                                (0
                                                w
                                                a>
                                                o
                                                o
                                                ui
                                                a
                                                o
                                                w
                                                x
                                                o
                                                o
                                                r>
                                                I
                                                n


                                                0)
                                                l-i


                                                tn
                                                •H
3-194

-------
3.8  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, 0. G., A Total Combustion Systems Approach
     Proves Successful for NOX Control for Two Steam
     Generators.   Presented at the American Power Conference,
     April 1991.

5.   Letter and attachments from Emmel, T. E., Radian
     Corporation, to Kcsim, 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-N0x 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.   Questionnaire response from Jeansonne, D., Central LA
     Electric  Company.   Dolet Hills 1.  Undated.
                             3-195

-------
10.  Questionnaire response from Maloney,  J.,  United Power
     Association - Stanton 10.   Undated

11.  Questionnaire response from Smith, J.  R.,  Houston
     Lighting & Power Company.   Limestone  1 and 2.   Undated.

12.  Questionnaire response from Dawes, S., Sierra Pacific
     Power Company.  North Valmy 2.   Undated.

13.  Letter and attachments from Lewis, P.  E.,  Colorado -  Ute
     Electric Association,  to Kanary,  D. A.,  Ohio Edison
     Company.  December 8,  1992.  Response  to NOX information
     request for Craig 3.

14.  Letter and attachments from Boyce, P.  L.,  Northern States
     Power Company to Kanary,  D. A.,  Ohio  Edison Company.
     December 1,  1992.  Response to  NOX Information Collection
     Request of November 5, 1992 for Sherburne 3.

15.  Letter and attachments from Marshall,  G.,  Pacific
     Corporation,  to Harrison,  C. S.,  Hunton and Williams.
     December 14,  1992.  Information Collection Request for
     Hunter 3.

16.  Letter and attachments from Linville,  C.,  Sunflower
     Electric Power Corporation, to  Harrison,  C., Hunton and
     Williams.   February 25,  1993.   NOX Information Collection
     Request for Holcomb 1.

17.  Questionnaire response from Vubas H.,  Deseret Generation
     and Transmission Cooperative.   Bonanza 1.   Undated.

18.  Letter and attachments from Barney C., Grand River Dam
     Authority to Jordan B. C.,  U.S.  Environmental Protection
     Agency.  June 28, 1993.   Information  Collection Request
     for GRDA 2.

19.  Letter and attachments from Norton,  C.,  West Texas
     Utilities Company to Jordan, B.C., U.S.  Environmental
     Protection Agency.  July 28, 1993.  Information
     Collection Request for Oklaunion 1.

20.  Questionnaire response from Smith, J.  R.,  Houston
     Lighting & Power Company.   W. A.  Parrish 8.  Undated.

21.  Questionnaire response from Bentley,  J.,  Lower Colorado
     River Authority.  Fayette 3. Undated.

22.  Questionnaire response from Griego B.   Plains Escalante
     Generating Station.  Plains Escalante  1.   Undated.
                             3-196

-------
23.  Questionnaire response from Schulz, P., Platte River
     Power Authority to Jordan, B. C. ,  U. S. Environmental
     Protection Agency.  Response to Information Collection
     Request for Rawhide Energy Station Unit 1.  Undated.

24.  Letter and attachments from Brownell, W. F., Hunter and
     Williams,  to Eddinger, J. A., U. S. Environmental
     Protection Agency.  December 18, 1992.  Response to
     Information Collection Request for Brandon Shore Units 1
     and 2 .

25.  Letter and attachments from Brownell, W.F., Hunton and
     Williams,  to Eddinger, J.A., U.S.  Environmental
     Protection Agency.  March 1, 1993.  Response to
     Information Collection Request for JB Sims 3.

26.  Letter and attachments from Huff,  B. L., Cincinnati Gas &
     Electric Company,  to Harrison, C.  S., Hunton and
     Williams.   December 7, 1992.  Response to Information
     Collection Request for Zimmer 1.

27.  Questionnaire response from Kappelman R. L., Jacksonville
     Electric Authority, St. Johns River 1 and 2.  Undated.

28.  Letter and attachments from Chadwick, A.,  New York State
     Electric and Gas Corporation to Harrison,  C., Hunton and
     Williams,  December 4,  1992.  Information Collection
     Request for Kintigh Station 1.

29.  Letter and attachments from Sandefur, M. L., Southern
     Indiana Gas and Electric Company,  to Kanary, D., Ohio
     Edison.  March 8,  1993.  NOX information collection
     request for A. B.  Brown 2.

30.  Letter and attachments from Todd,  D. L., Big Rivers
     Electric Corporation,  to Harrison, C. S.,  Hunton and
     Williams.   March 5, 1993.  NOX information collection
     request for D. B.  Wilson 1.

31.  Questionnaire response from Hicks, R. F.,  Orlando Utility
     Commission - C. H. Stanton 1.  Undated.

32.  Questionnaire response from Steinlen, J.,  Seminole
     Electric Coop, Inc. -  Seminole 1 and 2.  Undated.

33.  Questionnaire response from Giese, J., Los Angeles Dept.
     of Water & Power.   Intermountain 1 and 2.   Undated.

34.  Letter and attachments from Smith, A. E.,  Northern
     Indiana Public Service Company to Harrison, C., Hunton &
     Williams.   December 1, 1992.  Response to NOX Information
     Request of October 29, 1992 for R. M. Schahfer 17 and 18.
                             3-197

-------
35.  Letter and attachments from Scherrer,  C. R.,  Muscatine
     Power and Water,  to Kanary, D.  A,  Ohio Edison Company.
     December 2, 1992.  Response to NOX Information Collection
     Request of November 5, 1992 for Muscatine 9.

36.  Questionnaire response from Chaplin,  M. C., South
     Carolina Public Service Authority.  Cross 2.   Undated.

37.  Questionnaire response from Hunter,  J., Tampa Electric
     Company.  Big Bend 4.  Undated.

38.  Letter and attachments from Sanburg,  L.L, Texas New •
     Mexico Power Company to Jordan B.C.,  U.S. Environmental
     Protection Agency.  December 21, 1992.  Information
     request for TNP Units 1 and 2.

39.  Questionnaire response from Pierce K., Applied Energy
     Services Thames,  Inc., AES Thames  Units 1A and IB.
     Undated.

40.  Questionnaire response from Hess,  T.,  Cambria Cogen
     Company.  Cambria Units 1 and 2.  Undated.

41.  Questionnaire response from Recor, R.  A., POSDEF Power
     Co.,  L.P. Stockton A and B.  Undated.

42.  Letter and attachments from Cooper,  T., AES Barbers
     Point, Inc., to Jordan B. C.,  U. S.  Environmental
     Protection Agency.  December 23, 1992.  Information
     Collection Request from Barbers Point A and B.

43.  Questionnaire response from Hess,  T.,  Stockton Cogen -
     Stockton Cogen.  Undated.

44.  Questionnaire response from Neal,  M.,  Pyro-Pacific
     Operating Company.  Mt. Poso Cogeneration Plant.
     Undated.

45.  Letter and attachments from Barber,  D. E., Ultrapower
     Constellation Operating Services to Jordan, B. C., U.  S.
     Environmental Protection Agency.  December 17, 1992.
     Information Collection Request for Rio Bravo Jasmin and
     Rio Bravo Poso.

46.  Letter and attachments from Gamble,  M., Tacoma Public
     Utilities to Jordan, B. C., U.  S.  Environmental
     Protection Agency.  December 10, 1992.  Information
     Collection Request for Steam Plant No. 2.

47.  Letter and attachments from Welsh, M.A. Electric
     Generation Association to Eddinger,  J.A.,
     U. S. Environmental Protection Agency.  November 18,
     1993.


                             3-198

-------
48.  PowerPlants Database, Details of the Equipment and
     Systems in Utility and Industrial Power Plants, 1950-
     1984, Thomas C. Elliott, Editor, McGraw Hill, Inc. New
     York, New York.  1985.

49.  Questionnaire response from Ewing, D., Neveda Power
     Company.  Reid Gardner 4.  Undated.

50.  Questionnaire response from Pyle, W. C., Marquette Board
     of Lighting and Power.  Shiras 3.  Undated.

51.  Questionnaire response from Michigan South Central Power
     Agency, Endicott Jr. Unit 1.

52.  Personal communication, M. Gundappa, Radian Corporation
     with J. Roebel, Cincinnati Gas and Electric Company,
     May 28, 1993.

53.  DuBose, D.A.,  Kwapil, W.D.,  and E. F. Aul,  Jr.,
     "Statistical Analysis of Wet Flue Gas Desulfurization
     Systems and Coal Sulfur Content, Volume I:   Statistical
     Analysis."  Prepared for Office of Air Quality Planning
     and Standards, U. S. Environmental Protection Agency,
     1983.

54.  Dickey, D. A., "Effect of Autocorrelation on Statistical
     Analyses" Prepared for the Cadmus Group, July 1992.

55.  Pitts,  W. S.,  Smith, L. L.,  Flora, T. 0. and R. E. Rush,
     "Analysis of NOX Emissions Data for Prediction of
     Compliance with NOX Emissions Standards" presented at Air
     and Waste Management Association Conference, Seattle,
     March 1989.

56.  Giguere, G.  C. and Margerum, S. C., "Determination of
     Mean SC>2 Emission Levels Required to Meet a 1.2 Ib
     SC>2/Million Btu Emission Standard for Various Averaging
     Times and Compliance Policies," Technical Note, Prepared
     for U.  S. Environmental Protection Agency,  March 1985.

57.  Dickey, D. A.   Report to W.  S. Pitts Consulting Inc.
     January 1994.

58.  Box, G.E., and D. R. Cox.  An Analysis of
     Transformations,  Journal of  the Royal Statistical
     Society, Volume 26,  No. 2, 1964.

59.  Snedecor, G. W.,  and Cochran,  W. G.,  Statistical
     Methods, Sixth Edition.  The Iowa State University Press.
     Ames, Iowa.   p. 552.
                             3-199

-------
60.   Personal Communication.   M.  Gundappa,  Radian Corporation
     with J. Cron,  Southern Indiana Gas and Electric Company.
     July 9, 1993.

61.   Personal Communication.   M.  Gundappa,  Radian Corporation,
     with J. Hunter,  Tampa Electric Company.   March 26,  1993
     and March 31,  1993 .

62.   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.

63.   ROPM Burner for Oil  and Gas  Wall Fired Generating
     Facilities.  ABB Combustion  Engineering.   Publication
     PIB 103.  1990.

64.   Peabody ISC™ Low NOX Burners.  Peabody Engineering.
     Bulletin No. ISC-1.   Undated.

65.   Yee,  J. L.  B.,  Giovanni,  D.  V.,  and M. W. McElroy.
     Retrofit of an Advanced Low-N0x Combustion System at
     Hawaiian Electric's  Oil-Fired Kahe Generating Station.
     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.  9-1 through 9-18.

66.   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,

67.   PM Burner for Oil and Gas T-Fired Generating Facilities.
     ABB Combustion Engineering.   Publication PIB 102.  1990.

68.   Letter and attachments from  Strehlitz, F. W. , Pacific Gas
     & Electric Co.,  to Neuffer,  W. J.,  U.  S.  Environmental
     Protection Agency.  March 26, 1993.  Response to Section
     114 information collection request --  Pittsburgh 6 and 7,
     Contra Costa 6,  Moss Landing 7,  and Morro Bay 3.

69.   Bisonett, G. L., and M.  McElroy.  Comparative Assessment
     of NOX Reduction Techniques  for Gas- and Oil-Fired
     Utility Boilers.  Presented  at the 1991 Joint Symposium
     on Stationary Combustion NOX Control.   Washington,  DC.
     March 25-28, 1991.
                             3-200

-------
70.   Epperly,  W.  R.,  et al,  Control of Nitrogen Oxides
     Emissions from Stationary Sources.  Presented at the
     Annual Meeting of the American Power Conference, April
     1988.

71.   Letter and attachments from Haas, G. A.,  Exxon Research
     and Engineering Co.,  to Gundappa, M.,  Radian Corporation.
     May 1, 1992.  Information concerning Thermal DeNOx.

72.   Cato, G.  A., Maloney, K. L.,  and J.  G. Sotter.  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.

73.   Technical and Economic Feasibility of Ammonia-Based
     Postcombustion NOX Control.   Electric Power Research
     Institute.  Report No.  EPRI CS2713.   November 1982.
     pp. 3-18 to 3-25.

74.   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.

75.   Ref. 73,  p.  3-7.

76.   Ref. 73,  p.  3-9.

77.   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.

78.   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.

79.   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.

80.   Hjalmarsson, A.  K.  NOX Control Technologies for Coal
     Combustion.   IEA Coal Research, p. 44.  June 1990.
                             3-201

-------
81.  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.

82.  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.

83.  Ref. 77,  pp.  459-462.
                                                          »
84.  Ref. 80,  pp.  40-53.

85.  Rummenhohl,  V.,  Weiler,  H.,  and W. Ellison.  Experience
     Sheds  Light on SCR O&M issues.   Power Magazine.
     136:35-36.  September 1992.

86.  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.

87.  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.

88.  Ref. 77,  pp.  495-499.

89.  SNCR NOX Control Demonstration,  Wisconsin Electric Power
     Company.   Valley Power Plant, Unit 4.   March 1992.  Nalco
     Fuel Tech.

90.  Hunt,  T.   SNCR Demonstration at Public Service Company of
     Colorado's Arapahoe  Station.  Presented at the 1992 EPRI
     NOX Control for Utility Boilers Workshop.  Cambridge, MA.
     July 7-9, 1992.

91.  Teetz, R. D., Stallings,  J.  W.,  O'Sullivan, R. C.,
     Shore, D. E., Sun, W.H.,  and L.J. Muzio.   Urea SNCR
     Demonstration at Long Island Lighting Company's Port
     Jefferson Unit 3.  Presented at the 1992 EPRI NOX Control
     for Utility Boilers  Workshop.  Cambridge, MA.  July 7-9,
     1992.
                             3-202

-------
92.  Mansour,  M. N. ,  Nahas, S. N., Quartucy, G. C.,
     Nylander, J. H.,  Kerry, H. A., Radak, L. J.,  Eskinazi,
     D.,  and T. S. Behrens.  Full-Scale Evaluation of Urea
     Injection for NO Removal.  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. 43-1
     through 43-23.

93.  Springer, B.  Southern California Edison's Experience
     with SNCR with SNCR for NOX Control.   Presented at the
     1992 EPRI NOX Control for Utility Boilers Workshop.  •
     Cambridge, MA.  July 7-9, 1992.

94.  Letter and attachments from Brownell, F. W.,  and C. S.
     Harrison, Hunton and Williams, to Neuffer, W. J.,
     U. S. Environmental Protection Agency.  February 10,
     1993.  Information Collection Request - Alamitos 4.

95.  Letter and attachments from Brownell, F. W.,  and C. S.
     Harrison, Hunton and Williams, to Neuffer, W. J.,
     U. S. Environmental Protection Agency.  February 10,
     1993.  Information Collection Request - El Segundo 1
     and 3 .

96.  Teixeira, D. P.,  Lin, C. I., Jones, D. G.,
     Steinberger, J.,  Himes, R. M., Smith, R. A.,
     Muzio,  L. J., and S. Okazaki.  Full-Scale Tests of SNCR
     Technology on a Gas-Fired Boiler.  Presented at the 1992
     EPRI NOX Controls for Utility Boilers Workshop.
     Cambridge, MA.  July 7-9, 1992.

97.  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 .

98.  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.

99.  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.
                             3-203

-------
100. Southern California Edison Research Division, System
     Planning and Research Department.  Selective Catalytic
     Reduction DeNOx Demonstration Test Huntington Beach
     Unit 2.  June 1988.

101. Janik,  G.,  Mechtenberg,  A.,  Zammit, K.,  and E.
     Cichanowicz.  Status of Post-FGR SCR Pilot Plant Tests on
     Medium Sulfur Coal at the New York State Electric and Gas
     Kintigh Station.  Presented at the 1993  Joint Symposium
     on Stationary Combustion NOX Control.   Miami, FL.  May
     24-27,  1993.
                                                         t
102. Huang,  C. M.,  et.  al.   Status of SCR Pilot Plant Tests on
     High Sulfur Coal at Tennessee Valley Authority's Shawnee
     Station.  Presented at the 1993 Joint Symposium on
     Stationary Combustion NOX Control.  Miami, FL.
     May 24-27,  1993.

103. Guest,  et.  al.   Status of SCR Pilot Plant Tests on high
     Sulfur Fuel Oil at Niagara Mohawk's Oswego Station.
     Presented at the 1993  Joint Symposium on Stationary
     Combustion NOX Control.   Miami, FL.  May 24-27, 1993.

104. Mechtenberg, A.J.,  et  al.   Status of Testing of SCR
     Pilots:  A Review of Current EPRI Sponsored Results.
     Presented at EPRI/EPA 1995 Joint Symposium on Stationary
     Combustion NOX Control,  Book 3.  Kansas  City, Missouri.
     May 16-19,  1995.

105. 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.

106. 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.

107. 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.

108. 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.
                             3-204

-------
109. 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.

110. Letter and attachments from Katherine L. Miller,
     Commonwealth of Virginia,  Department of Environmental
     Quality to Rachel Adams,  Radian Corporation.
     December 14, 1994.

111. Trip Report.  Plant Visit to Stanton Energy Center, James
     A.  Eddinger, U.S. Environmental Protection AGency,
     Research Triangle Park, NC.  December 8, 1994.

112. Letter and attachments from Jones, M.L., U.S. Generating
     Company,  to Eddinger,  J.A., U.S. Environmental Protection
     Agency.  October 30, 1995.

113. Cassell,  T. D. et al.   NOX/SC>2 Removal with No Waste -
     The SNOX Process.  Presented at the First Annual Clean
     Coal Technology Conference.  Cleveland, OH.
     September 22-24, 1992.

114. Kingston,  W. H., Cunninghis, S., Evans, R. J., and C. H.
     Speth.  Demonstrating the WSA-SNOX Process Through the
     CCT Program, ASME Paper No. 90 -JPGC/FACT-17.

115. Radinger,  K. E. et al.  SNRB - S02,  NOX, and Particulate
     Emissions Control with a High Temperature Baghouse.
     Presented at the First Annual Clean Coal Technology
     Conference.  Cleveland, OH.  September 22-24, 1992.

116. Black, J.  B., et al.  The NOXSO Clean Coal Technology
     Project:   A 115 MW Demonstration Unit.  Presented at the
     First Annual Clean Coal Technology Conference.
     Cleveland, OH.  September 22-24, 1992.

117. 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.

118. Clean Coal Technology Demonstration Program, Program
     Update 1995.  U.S. Department of Energy.  April 1996.
                             3-205

-------
            4.0  MODEL BOILERS AND CONTROL OPTIONS
     This chapter presents the model boilers used to assess
                                                          9
the environmental and cost impacts resulting from a revision
to the current standard.  Model boilers have been selected to
represent the range of boilers potentially affected.  In
addition, control options considered for the revision to the
current standard 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, capacity factor, and heat rate.  These
factors affect one or more of the following:  1) uncontrolled
NOX emissions, 2) control system applicability, 3)  control
system performance, and 4)  control system costs.  Table 4-1
summarizes the effect of each model plant parameter.  The
selection of baseline emission rates is also discussed in this
section.
4.1.1  Fuel Type
     Three types of coal (bituminous, subbituminous, and
lignite), oil, and natural gas were used to define model
boilers.  Fuel type has a major impact on uncontrolled and
baseline NOX emissions, primarily due to differences in
combustion characteristics and fuel nitrogen content.
Generally, coal has higher NOX emissions than oil,  which has
higher emissions than natural gas.  Control system performance

                              4-1

-------



en
rt
W
H
w
<
PH
rf^
a
("V|
w
H
O
«
J
w
D
O
2
O
CO
EH
U
H
Cti
[j_i
W

•
H
^

W
m
r
E~*












Control
system
annual O&M
cost ($/yr)
r-l c r-l
o fd ^
jj -h1 -H Q ;*
E >^ Q* u v>
O ,>, (D
u w u



0)
u
7! 6 S
O g IT)

) 1 "*"^ ^_l
C ? 0
0 £lH
u ^
Q)
a


g >,
jj _-^
? r—l
ro"H
o 5
^ g.
u *



•0
aj-^ w
r—i fu |»H
rH 3 0 "C
0 C -H >,
ll § g^
o x"^ d
c § ^
D

T3 C
Oi O -^
i— 1 -H D
O W tt) CD
^ -H 4-1 2
4J G fO S
C 0) i-i ^
o &
U X rH
CO —
0)
4-1
01
e
rS
rd
ft
X

X







X









X









X








X




0)
k1
f*1
C"1
r- 1
OJ
X!
       X
X
              X
X
       X
                    X
01
u
fd
C  (D
       0)
       N
       -H
       CO

       JJ
       •H


       D
u  o
(0  4J
a u
*  (6
U  fo
                    (S
                    0)
                    4-2

-------
and 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.
     Boilers burning oil and natural gas are not expected to
represent a large share of the boilers subject to the revised
standard.  However, model boilers burning these fuels are
included to fully represent the possible range of new boilers.
4.1.2  Furnace Type
     Three furnace types--wall, tangential, and fluidized bed
combustion (FBC)--were used to define model boilers.  Other
furnace types (e.g., cyclone or cell-burner) are not expected
to be built in the future due to their high NOX emission
characteristics and are not included in the analysis.
     Furnace type can have an influence on uncontrolled NOX
emissions.  Of the three furnace types selected, wall-fired
boilers can have higher uncontrolled levels than tangentially-
fired and FBC boilers, because of the potential for rapid fuel
and air mixing which can increase fuel and thermal NOX.
Tangentially-fired boilers employ air staging which inhibits
the formation of NOX and typically results in lower
uncontrolled NOX emissions than wall-fired boilers.  As
discussed in chapter 3.0,  however,  NOX emissions for wall- and
tangentially-fired boilers applying combustion controls are
generally similar.  Fluidized bed combustion boilers have low
furnace temperatures  (1600 °F versus 2500 to 3000 °F for
conventional boilers), which greatly reduces the formation of
thermal NOX.   As a result,  FBC boilers generally have lower
NOX emissions than either wall- or tangentially-fired boilers.
     Furnace type also affects the applicability, performance,
and cost of NOX control systems.  For example,  wall- and
tangentially-fired boilers require different types and
configurations of low NOX burners (LNB)  and overfire air
(OFA).   Combustion controls for FBC boilers are limited to OFA
and optimization of process parameters (e.g.,  bed
temperature).
                              4-3

-------
4.1.3  Boiler Size
     A wide range of boiler sizes (100, 300, 600, and 1000 MW)
was used to define wall- and tangentially-fired model boilers.
Fluidized bed combustion model boilers are represented by two
smaller boilers  (25 and 100 MW)  to reflect the size
differences between FBC and conventional boilers.
     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 kilowatt of
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).
4.1.4  Capacity Factor and Heat Rate
     Similar to boiler size, a wide range of capacity factors
(30, 65, and 80 percent) was used to define model boilers.
Heat rate is partially dependent on capacity factor,  with
higher heat rates associated with lower capacity factors due
to lower system efficiency at partial load.  Table 4-2
presents the heat rates associated with the selected capacity
factors for each fuel type.
     Capacity factor affects annual baseline NOX emissions,
NOX control system performance,  and annual costs.  A boiler
operated at a lower capacity factor will have lower annual
baseline NOX emissions than a similar boiler operated at full
load.  Heat rate affects annual baseline NOX emissions and
control system annual costs.  A boiler with a high heat rate
(low efficiency)  will burn more fuel and have higher NOX
emissions per unit of electricity produced and per year.
4.1.5  Baseline NOX Emission Rates
     Average annual baseline NOX emission rates for wall- and
tangentially-fired model boilers were based on emission limits
from the current subpart Da standard.  As discussed in
                              4-4

-------
   TABLE 4-2.   CAPACITY FACTORS AND ASSOCIATED HEAT RATES
Fuel type	Capacity factor  (%)  Heat  rate (Btu/kWh)

Coal                       30                  11,000
                           65                   9,500
                           80                   9,000

Oil                        30                  10,500
                           65                   9,000
                           80                   8,500

Natural Gas                30                  10,500
                           65                   9,000
                           80                   8,500
                            4-5

-------
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 Da standard.  The NOX emission limits
based on the current subpart Da standard are shown in
table 2-11 of chapter 2.0.  The baseline rate for FBC boilers
was based on information from FBC boiler vendors.K2'3
     Average annual baseline emission rates for wall- or
tangentially-fired boilers burning lignite, bituminous coal,
subbituminous coal, oil, and natural gas are 0.58, 0.58, 0.48,
0.28, and 0.18 Ib/MMBtu, respectively.  The baseline rate for
FBC boilers is 0.38 Ib/MMBtu.
4.2  MODEL BOILERS
     A total of 38 model boilers reflecting the parameters
discussed in section 4.1 are presented in table 4-3.  These
boilers represent five different fuels,  three different
furnace types, and five different sizes.  Each of these models
was evaluated at three capacity factors (0.30, 0.65, and 0.80)
to represent peaking, cycling,  and base-load operation.
4.3  CONTROL OPTIONS
     Three control options were considered for the revision to
the current standard:  combustion controls, SNCR,  and
selective catalytic reduction  (SCR).   Each of these is
discussed in a separate section.
4.3.1  Combustion Controls
     As discussed in chapter 3.0, this control option is
comprised of the same equipment utility boilers use to meet
the current standard.  The lower NOX emission rates are
attributable to improved operation of the equipment.
     The specific combustion controls that were considered are
LNB for wall-fired boilers, LNB with close-coupled OFA for
tangential-fired boilers, and air stating for FBC boilers.
Combustion controls are applicable to any of the model
boilers.  However, the information presented in chapter 3.0
suggests that there are insufficient data on combustion
controls demonstrating lower NOX emission rates than the
                              4-6

-------
               TABLE 4-3.  MODEL BOILERS3
Fuel type
Furnace type
Size (MW)
 Baseline NOX
emission rate*3
  (Ib/MMBtu)
Bituminous Wall
coal
Tangential
FBCC
Subbituminous Wall
coal
Tangential
FBCC
Lignite Wall
Tangential
FBCC
100
300
600
1000
100
300
600
1000
25
100
100
300
600
1000
100
300
600
1000
25
100
100
300
600
1000
100
300
600
1000
25
100
0.58
0.58
0.38
0.48
0.48
0.38
0.58
0.58
0.38
                            4-7

-------
            TABLE 4-3.  MODEL BOILERS3 (Continued)
                                                Baseline  NOX
                                               emission rate*3
Fuel type
Oil
Natural gas
Furnace type
Wall and
tangential
Wall and
tangential
Size (MW)
100
300
600
1000
100
300
600
1000
(Ib/MMBtu)
0.28
r
0.18
aEach model boiler was evaluated at three capacity factors
 (30, 65, and 80 percent) .   Heat rate will vary with capacity
 factor and fuel type.  For coal,  the following heat rates
 were used: 30 percent - 11,000 Btu/kWh; 65 percent -
 9,500 Btu/kWh; 80 percent  - 9,000 Btu/kWh.  For oil or
 natural gas, the following heat rates are used:
 30 percent - 10,500 Btu/kWh; 65 percent - 9,000 Btu/kWh;
 80 percent - 8,500 Btu/kWh.
    determine average annual baseline NOX emission rates for
 wall- and tangentially-f ired boilers,  an emission rate equal
 to 0.02 Ib/MMBtu less than the current standard was used.
 Baseline rates for FBC boilers were based on information from
 three vendors. ' '

CFBC = fluidized bed combustion.
                             4-8

-------
current Da standard for boilers burning natural gas or oil.
Therefore, combustion controls with a lower NOX emission rate
were not evaluated for boilers burning natural gas or oil.
     Table 4-4 presents baseline NOX emission rates and
projected emission rates achieved with the application of
combustion controls.  These emission rates are based on the
data presented in chapter 3.0.
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 levels of wall injectors
and one level of lance injectors.  A normalized stoichiometric
ratio of 1.0 was used, and the urea solution as injected was
assumed to be 10 percent urea by weight.  A constant emission
reduction of 50 percent was assumed.
4.3.3  Selective Catalytic Reduction
     This control option was considered for wall- and
tangentially-fired boilers but was not considered for 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 natural gas-fired boilers.  A normalized
stoichiometric ratio of 0.82 was used.  A constant NOX
emission reduction of  80 percent was assumed.
                              4-9

-------
     TABLE 4-4.  CONTROL TECHNOLOGY PERFORMANCE LEVELS
                     Combustion Controls3
Baseline
emission rate
Coal type
Bituminous

Subbituminous

Lignite
Furnace type
Wall
Tangential
FBCC
Wall
Tangential
FBCC
Wall
Tangential
FBCC
(Ib/MMBtu)
0
0
0
0
0
0
0
0
0
.58
.58
.38
.48
.48
.38
.58
.58
.38
Controlled
emission rate13
(Ib/MMBtu)
0
0
0
0
0
0
0
0
0
.40
.40
.20'
.30
.30
.20
.40
.40
.20
aA lower emission rate for the application of combustion
 control to boilers burning oil or natural gas was not
 evaluated.  Baseline emission rates for boilers burning
 these fuels are 0.28 and 0.18 Ib/MMBtu, respectively.

bRates are based on long-term data presented in
 chapter 3.0.

CF3C = fluidized bed combustion.
                            4-10

-------
4.4  REFERENCES
1.   Telecon.  Stone, J.,  Ahlstrom Pyropower, Inc., with
     King, B., Radian Corporation.  August 18, 1993.
     Fluidized Bed Combustion Boilers.

2.   Telecon.  Czarnecki,  T.,  ABB-CE, with King,  B., Radian
     Corporation.  August 18,  1993.  Fluidized Bed Combustion
     Boilers.

3.   Telecon.  Edvardsson, C., Tempella Power, with King, »B.,
     Radian Corporation.  August 17, 1993.  Fluidized Bed
     Combustion Boilers.
                             4-11

-------
             5.0  ENVIRONMENTAL AND ENERGY IMPACTS

     This chapter presents an analysis of the environmental
and energy impacts associated with the application of each
control option to the model boilers presented in chapter 4.0.
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.  Finally section 5.4 discusses the
incremental energy impacts.
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
induced effects of each control option.  These include
increases or decreases in the emission levels of other
pollutants caused by the application of a specific control
option.  Primary and secondary impacts are presented in the
following sections.
5.1.1  Primary Air Impacts
     The NOX emission reductions achieved by the application
of each control option to each of the model boilers are
presented in appendix B-l.  The table includes model boilers
representing each fuel (bituminous and subbituminous coal,
lignite, oil, and natural gas) and furnace type (wall,
tangential, and fluidized bed combustion  [FBC] ) .   For each
boiler, the annual baseline NOX emissions and emission
reductions associated with each control option were calculated
based on emission rates and NOX reduction efficiencies
presented in tables 4-3 and 4-4.  As noted earlier
(chapter 4.0),  selective catalytic reduction  (SCR) was not
considered as a control option for FBC boilers.
                              5-1

-------
     Table 5-1 presents the percent reductions and ranges of
annual NOX emission reductions for the three control options
applied to the model boilers operating at a capacity factor of
0.65.  As shown in the table,  the NOX emission reductions for
wall- and tangentially-fired boilers are the same because the
estimated baseline and controlled NOX emission rates are the
same for these boiler types.
     Lignite and bituminous coal-fired conventional boilers
have the same baseline and controlled NOX emission rates for
each control option.  Therefore, the percent NOX emission
reductions and the range of annual tons of NOX removed are the
same for these boilers.  For conventional, subbituminous coal-
fired boilers with combustion controls (CC),  the range in
annual tons of NOX removed is also the same as for bituminous
coal- and lignite-fired boilers because the difference between
baseline and controlled NOX emission rates is the same for all
three coal types.  However, because the baseline NOX emission
rate is lower for subbituminous coal, the percent NOX
reductions are higher.  The lower baseline NOX emission rate
for subbituminous coal also results in lower annual tonnages
of NOX reduction for CC + selective noncatalytic reduction
(SNCR) and CC + SCR control options when compared to the
reductions for bituminous coal- and lignite-fired boilers.
     There are no differences in NOX emission reductions among
FBC boilers burning the various coal types.   This is because
no distinction was made in either baseline or controlled NOX
emission rates based on fuel type.
     Because sufficient data were not available for natural
gas- and oil-fired boilers to demonstrate a lower NOX emission
rate with CC, the estimated NOX emission reductions are zero
for all natural gas- and oil-fired model boilers with CC.
Because of the lower baseline NOX emission rates for natural
gas-fired boilers, the range of annual tonnage of NOX removed
with CC + SNCR and CC + SCR are lower than the corresponding
ranges for oil-fired boilers.
                              5-2

-------
in

W
J
CQ
CO
G
O
•H
4J
U
3
•o
(D
VH
CO
d
o
•rH
CO
CO
.,_!
e
CD
X
O
rH

^
>"1
CO
o
4J





4J
0)
u
H
CD

Cn
O O
rH rH
4J O
d d
O XI
u u






























0)
4J


CD
a
nj
\L>
u
rd
d
•H
fc






CD
a
i
•^

i — i
CD
r^
W





O
c-
00
^r
i
o
cn






H
CO








U
U
o
00
CN
0
H
O
ro
O
H





VC
U)




1
cn
,

o
u
o
CN
IT>
ro
rH
0
in
m
H





oo




Pi
U
cn
,

u
u



rH
rd
T3 -H
d 4J
rd d
CD
rH rjl
rH d
nj rd
5 -U

rH
rd
O
U
CD
CO 4J
3 -H
o d
d cn
-H -H
6 rH
3
4J tS
-H d
CQ rd

o
00

d
o
u

U-l
O

0)
Cn
d •
rd S
rH 2

(D 0
X! o
4J H
1
rH IT!
O CN
14-1
M-1
in o
U5
• m
0 0)
Nl
IJ-J -H
o co

V i ^j
O 0)
4J rH
CJ -H
rd O
u-i X3
4-) CQ
-H Cu
U
as T!
pi C
OS OS
U
rd S

4J O
rd o
o
(D -
rH rH
rd '
o
T3 O
(D H
4J
d 4H
CD O
CO
CD CO
rH (1)
a N
•H
OS CO
4J
OS rH
T3 CD
rH
CD -H
Xi O
E"1 .Q
rd
•d
d
OS

. ^
d
o
-H
4J
O
3
T3
0)
oi
,
o
-rH
4J
^1
rH
OS
4J
OS
U

•- CD
rH >
O -H
^ 4J
4J U
d CD
O rH
O 0)
cn
d
O II
-H
4J Pi
CO U
3 cn
n
5 *-
o d
u o
-H
II U
U
U 3
O T3
CD
•-Pi
d
0 U
.^ 	 i .^
1 1 rl
4-) 4->
CO >1
3 rH
Q OS
C 1 ^
O OS
u u
d •
T3 O CD
CD S rH
CQ X)
CD 05
"T) b* O
CD -H -H
N 4J rH
•H U Si,
T3 0) Oi
-H rH <
3 CD
rH CO 4J
Ct-i O
II ^
II
Pi II
U U
CQ S 
-------
5.1.2  Secondary Air Impacts
     Secondary air emissions associated with NOX control
technologies include ammonia (NH3)  slip, nitrous oxide  (N2O)
and carbon monoxide (CO).  For conventional coal-fired
boilers, baseline emissions of NH3,  N20, and CO were estimated
at 0 parts per million (ppm),  5 ppm,  and 20 ppm, respectively
(each at 3 percent oxygen  [©2!)   For FBC boilers, baseline N20
emissions were estimated at 100 ppm (at 3 percent 02) based on
measurements of N20 emissions from these boilers.1   Emissions
of NH3 and CO were assumed to be the same as for conventional
coal-fired boilers.  For oil-  and gas-fired boilers, baseline
emissions of NH3,  ^0,  and CO were  estimated at 0 ppm, 5 ppm,
and 5 ppm, respectively  (each at 3  percent ©2).
     Available data on retrofit experiences with CC on utility
boilers have shown increases in CO  emissions.  However, CC are
not expected to increase CO emissions from new boilers because
new boilers can be designed to improve the combustion air and
furnace gas mixing process for maximum control of combustible
emissions such as CO.2  Emissions of NH3  and  N20 were also
assumed equal to baseline levels.   Hence, secondary air
impacts for CC were assumed equal to baseline levels for all
model boilers.
     The addition of the urea-based SNCR control option to CC
can result in increases of CO,  NH3,  and N20.  Based on data
presented in table 3-35,  and on more recent operating
experience3'4'5 increases in NH3,  CO,  and N20 emissions were
estimated at 25 ppm, 25 ppm, and 10 ppm; respectively  (each at
3 percent 02) for all model boilers.   If NH3 is used as a
reagent in SNCR systems,  CO and N20 emissions are expected to
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 data presented in table 3-37.  Emissions of
N20 and CO were assumed equal to CC levels.
                              5-4

-------
     Based on these values, the estimated annual emissions of
NH3,  N20, and CO resulting from the application of each NOX
control option to each model boiler are presented in
appendix B-2.  The annual emissions were calculated using EPA
Method 196  (40  CFR  Part  60).   Table  5-2  presents  ranges  of
annual tons of NH3, N20, and CO emissions for model boilers
operating at a capacity of 0.65.  Because no increase (over
baseline) in secondary air emissions were associated with CC,
the annual tons of pollutant emissions for CC are identical to
baseline.  For conventional coal-fired boilers equipped with
CC + SNCR,  NH3 emissions range from 35 tons/yr for a 100 MW
boiler to 340 tons/yr for a 1,000 MW boiler.  Corresponding
ranges for N20 and CO emissions are 50-530 tons/yr and
100-1,020 tons/yr,  respectively.  The incremental secondary
air impacts of SNCR over CC can be inferred from table 5-2 by
subtracting the range for CC from CC + SNCR.  For coal-fired
boilers, the incremental emission ranges are 35-340 tons/yr,
30-350 tons/yr, and 55-560 tons/yr for NH3,  N20 and CO,
respectively.
     The ranges of NH3 and CO emissions for FBC boilers are
smaller because of the smaller range in boiler size.  The
relatively higher range in ^0 emissions for the FBC model
boilers reflect the higher N20 emissions measured from these
boiler types.  For natural gas- and oil-fired boilers, the
calculated secondary pollutant emissions are lower than for
conventional coal-fired boilers because of differences in heat
rates and F-factors between these fuel types.
5.2  LIQUID WASTE IMPACTS
     The use of large quantities 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 is
significant at NH3/NOX molar ratios of about 3 or greater.
Because the NH3/NOX molar ratios are less than one for the SCR
and SNCR control options being considered, the water pollution
                              5-5

-------
CN
 I
in

M
>4
CQ

•£
CO
d
o
4J
CO
d
o
-H
CO
tn
-H
£
CD
Li
econda:
co


O
CJ






o
fN
S



ro
K
S

cn
o
rH i— |
0 O
i_j r^
JJ XI
d u
O CD
CJ JJ
QJ
a
jj

0)
u
03
d
M
rj
Cn







CD
p.
^N
JJ

rH
CD

Cn

0
in
in





0
oo
H
0
CN


o



a)
d
•H
rH
0
cn
05
m

o
m
«*
i
in





0
03
H
O
CN


O









U
U
o
CN
O
t
O
o
H



0
ro
in
o
m


o
ro
in
ro

CJ
25
CO

+

CJ
u

o
in
i
in





o
oo
H
o



0
ro
00


Od
CJ
CO

+

CJ
CJ


03
T) -H
d i ;
o3 d
QJ
I-H cn
rH d
03 05
|5 JJ

in
o
H





o
m
i
o



o



QJ
d
-H
, — 1
Q)
cn
05
CQ

in
o
H





o
ro
O

d
o
u

H-l
o
CD
tn
d •
n3 Is
M §

(D o
XI 0
d
O
-H
JJ
U
rj
•a
0)


u
-H
JJ
^
•
r-l
.U
03
CJ

CD
>
-H
                                                                                                 JJ  rH  .^  JJ
                                                                                                      i  f	|  CJ
                                                                                                 ^H  m  o  CD
                                                                                                 O  CN  -i
                                                                                                 03  03  o
                                                                                                 U      H  U
                                                                                                     § J_)  'H
                                                                                                 03  S  co  JJ

                                                                                                 JJ  O _Q  rH
                                                                                                 05  o td  rO
                                                                                                     o  5  jj
                                                                                                 0   - CJ  05
                                                                                                  05
                                                                                                     O pQ
0
JJ
d

cn
CD  cn
M  0
Q4 N

05  CO
jj
n3  M
T3  0
                                                                                                        TS

p
r~i  ij
 0  o
CO  2i
                                                                                                  0 -H CJ  CJ
                                                                                                 XI  O CQ  £ <
                                                                                                 E-i X! Cn  CO S
                                                                                                 03    Xi      U
                                                      5-6

-------
impact resulting from the application of these control options
on the model boilers is minimal.
5.3  SOLID WASTE DISPOSAL IMPACT
     The application of the NOX control options are 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 are
usually 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.
     As discussed in section 5.1, the addition of SNCR or SCR
to CC generates NH3 emissions.  The NH3 slip can either be
emitted to the atmosphere or can be absorbed onto the fly ash.
This could present ash handling and disposal problems or
prevent the sale of fly ash to cement producers that may have
upper limits on NH3-in-ash.  This problem is more significant
with SNCR than with SCR because of the larger NH3 emissions
associated with SNCR.
     For SCR systems disposal of the spent catalysts can be a
problem.  Some catalysts contain vanadium pentoxide (V"205)
which is a hazardous chemical.  However, catalyst
manufacturers are working on improved methods for reactivation
and recycling of spent catalysts as well as recoating the
catalyst support.7   In  Japan,  where  there  are  several  SCR
applications, raw materials such as titanium dioxide  (Ti02)
are recovered from the spent catalysts.7
5.4  ENERGY IMPACTS
     The energy impacts for each NOX control option on the
model boilers are presented in appendix B.3.  The impacts
reflect the additional energy requirements for each NOX
control option over baseline levels.  Energy impact ranges are
                              5-7

-------
expressed in units of MW-hr/yr and as a percentage of boiler
electrical output.  The equations used to estimate the energy
impacts are discussed later in this section.   The results for
the model boilers operating at a capacity factor of 0.65 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
conventional, bituminous coal- and lignite-fired boilers range
from 1,460 to 14,590 MW-hr/yr corresponding to 0.26 percent of
boiler electrical output.   For conventional,  subbituminous
coal-fired boilers energy impacts are slightly lower and range
from 1,240 to 12,370 MW-hr/yr (0.22 percent of boiler
electrical output).   The lower energy requirements is due to
reduced reagent usage which results in a decrease in the
energy loss associated with the vaporization of the reagent -
water solution in the boiler  (Eq. 5-1).   To achieve the same
NOX reduction,  less reagent is required for subbituminous
coal-fired boilers because the NOX emission levels at the
inlet to the SNCR system are lower than for bituminous coal-
or lignite-fired boilers.
     For FBC boilers, energy impacts range from 250 to
1,010 MW-hr/yr  (0.18 percent of boiler electrical output).
The lower levels reflect the smaller boiler sizes (maximum
100 MW) and lower NOX levels at the inlet to the SNCR system.
     Additional energy requirements range from 970 to
9,700 MW-hr/yr  (0.17 percent of boiler electrical output) for
natural gas-fired boilers and from 1,190 to 11,920 MW-hr/yr
(0.21 percent of boiler electrical output) for oil-fired
boilers equipped with SNCR controls.
     Energy requirements for SCR systems are higher than for
SNCR.  As shown in table 5-1, the energy requirements range
from 2,350 to 23,500 MW-hr/yr (0.41 percent of boiler output)
for conventional coal-fired boilers.  As discussed previously,
SCR was not considered applicable to FBC boilers.  For natural
gas- and oil-fired boilers, energy requirements are lower and
range from 1,980 to 19,820 MW-hr/yr  (0.35 percent of boiler
                              5-8

-------







rd
CO
rV
MH
W
i-q
H
O
m
l-q
w
Q
O
S
Pi
O
Cn

CO
H
u
<
(X
s
H
>H
O
Pi
u
s
w
Cn
O

>-l
Pi

g
p»l
CO

•
ro
in
H
J
PQ
<
H




03
JJ
C
ireme:
^
D1
QJ
Pi
w
?-t
Cn
W
0)
C
W
rH
(d
C
O
-H
JJ
-H
T3
T3
•<



rH
rd
O 4J
l*H -H 3
O M ft
4J 4J
oV U 3
QJ O
rH
w




TH
>l
rH
x:
i





X!
>i
rH Cn
O O
rH rH
U O
C C
0 X3
U U
QJ
H



QJ
ft
>-,
H

QJ
U
rd
£
^
d
Cn




QJ
p.
H
rH
QJ
3
Cn





0








Q











U
U




VD
CN
O






O
CT\
in
^
H

i
O
^
<*
•»
H




Pi
U
E5
CO

+
L)
U



H
**
O






O
o
in
m
CN

i
o
in
ro
^
CN




Pi
U
co

+
U
u








1 — 1
rd
rrt -iH
^ n
C -U
rrt r1
I U *— i
fl)
w
rH tD
. 	 j r- 1
( ' r-H
rd rd
"^ f-H
t^* L

T3
£
rd
03 "77
•-( UJ
M g 4J
rd 2-H
o 5 c
U d en
§ J
4J U
-H
m




o








o











U
u



CN
CN
O






O
r>
ro
CN
H

1
0
^
CN
*•
H




Pi
U
E5
CO

+
U
U



H
<*
O






O
0
in
m
CN

1
O
in
m
*•
CN




Pi
CJ
co

4-
u
u







.
I~~l
rd
T3 -H
a 4J
rd C
QJ
H Cn
rH E3
rd rd
J£ &-1




o








o











u
u



00
H
O






O
H
O
H

i
0
in
CN






Pi
u
&
CO

4-
u
u



g







u

v
o^

i
o
t>
i
UD rH
• U rd
o pQ jj
Cn rd
UH rj
O T3 d
C O
in rd £5
0
JJ S QJ
U 2 >
rd -H
MH O JJ
0 U
>,C QJ •
JJ -rH C
•H i — i QJ O
U i CO -H
rd o aJ
ft 0 II U
rd rH ^
u pi 73
UH U QJ
rd 0 S Pi
CO
JJ 03 U
rd QJ ---H
N rH JJ
QJ -H O >i
SH 0) SH H
rd JJ rd •
>H C JJ QJ
T3 Q) O rd iH
QJ rH U CJ X!
JJ -H rd
C O d QJ U
QJ XI O > -H
03 -H -H rH
QJ rH JJ 4J ft
JH rd 03 U ft
ft C 3 QJ  II II
0) C Pi
x: o cj u <
H U U CO <5
rd XI U
5-9

-------
electrical output).   These values are lower than for coal-
fired boilers because exhaust flow rates from natural gas- and
oil-fired boilers are lower than from similar size coal-fired
boilers, and consequently, the additional fan horsepower
requirements are also lower (Eqs. 5.5 and 5.6).
     The equations used in the estimation of energy impacts
for each NOX control option 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 by the exothermic oxidation of the reagent
in urea-based SNCR systems.  For a 10-percent solution of urea
in water,  the energy loss is calculated in terms of boiler
electrical output with the following equation:
     Energyv (MW-hr/yr)  = Urea * [-4654 + 9 *  (1183)] /
                          HR * 8760 * CF/1000             (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 H20 per Ib pure urea
     1183      =    Heat of vaporization of water  (Btu/lb H20)
     HR        =    Heat rate (Btu/kWh)
     CF        =    capacity factor (decimal fraction)
     8760      =    conversion factor; hr per yr
     1000      =    conversion factor; kW per MW

The urea injection rate is calculated from the following
equation:

        Urea  (Ib/hr)  = UncNOx *  NSR *  0.5  *  (60/46)  *
                      MW *  1000  * HR/106                  (5.2)
                             5-10

-------
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
     0.5       =    stoichiometric ratio (urea:NH2)
     60        =    molecular weight of urea
     46        =    molecular weight of NOX
     MW        =    Boiler capacity  (MW)

     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 kW
(electrical) per MW of boiler electrical output.  Based on
this, the following equation is used to estimate the energy
requirement:

     Energyc (MW-hr/yr)  = 1.0 * MW * CF * 8760/1000       (5.3)
where:
     MW   =    Boiler capacity (MW)
     CF   =    capacity factor (decimal fraction)
     8760 =    conversion factor; hr per yr
     1000 = conversion factor; kW per MW

For urea-based SNCR systems, the additional energy requirement
is calculated with the following equation:

                EnergySNCR  = Energyv +  Energyc            (5.4)

     For SCR,  energy is required to inject reagent into the
flue gas and this is estimated using equation 5.3.  Also,
additional fan horsepower is required to overcome the pressure
drop across the catalyst.  The following equation is used to
estimate this energy requirement:
                             5-11

-------
     Energyf (MW-hr/yr) = AP * 0.0361 * 144 * Q * CF *
                          8760/737.56/EF/1000             (5.5)
where:
     AP     =  pressure drop across the catalyst  (assumed to
               be 5 in. H20)
     0.0361 =  conversion factor; lbf/in2 per in. H20
        144 =  conversion factor; in2 per ft2
          Q =  flue gas flow rate at inlet to catalyst reactor
               (actual ft^/s); calculated using equation  (5.6)
         CF =  capacity factor (decimal fraction)
       8760 =  conversion factor; hr per yr
     737.56 =  conversion factor;  [(Ibf * ft/s)/kW]
         EF =  fan efficiency (decimal fraction = 0.85)
     1000   =  conversion factor; kW per MW
Flow rate (Q) is calculated based on the following equation:
     Q (ft3/s) = flow * Tcor * MW * 1000 / 3600
where:
     flow
                                                     (5.6;
          flue gas flowrate at inlet to catalyst at
          normal conditions (126 Nft^/kW-hr for coal-
          fired and 100 Nft3/kW-hr for oil- and gas-fired
          boilers)4
Tcor  =   temperature correction = 1060 R / 492 R
          (assuming catalyst operating temperature =
          600°F)
  MW  =   Boiler capacity  (MW)
 1000 =   conversion factor; kW per MW
  3600 =  conversion factor; seconds per hr
     The following equation is used to calculate the
additional energy requirement for SCR systems:
                           = Energyc + Energyf
                                                     ;s.7;
                              5-12

-------
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.   Marion, J. L., et.  al.   Development of ABB C-E's
     Tangential Firing System 2000 (TFS 2000™  System).
     Presented at the EPRI/EPA Joint Symposium on Stationary
     Combustion NOX Control.  Miami,  FL.  May 24-27,  1993.

3.   Shore,  D. E.,  et al.  Urea SNCR Demonstration at Long
     Island Lighting Company's Port Jefferson Station, Unit 3.
     Presented at the EPRI/EPA Joint Symposium on Stationary
     Combustion NOX Control.  Miami,  FL.  May 24-27,  1993.

4.   Hunt,  T., et.  al.   Selective Noncatalytic Operating
     Experience Using both Urea and Ammonia.  Presented at the
     EPRI/EPA Joint Symposium on Stationary Combustion NOX
     Control.  Miami, FL.   May 24-27,  1993.

5.   Hofmann, J. E., et. al.  Post Combustion NOX Control for
     Coal-Fired Utility Boilers.  Presented at the EPRI/EPA
     Joint Symposium on Stationary Combustion NOX Control.
     Miami,  FL.  May 24-27,  1993.

6.   Method 19 - Determination of Sulfur Dioxide Removal
     Efficiency and Particulate Matter,  Sulfur Dioxide, and
     Nitrogen Oxides Emission Rates.   Federal Register 40 CFR
     Part 60.  Appendix A.  p. 47853.

7.   Selective Catalytic Reduction (SCR) Controls to Abate NOX
     Emissions, White Paper.  Prepared by SCR Committee,
     Industrial Gas Cleaning Institute,  Inc.  November 1991.
                             5-13

-------
          6.0  MODEL BOILERS AND CONTROL OPTION COSTS

     This chapter presents the estimated cost and cost
effectiveness of the identified control options applied tq the
model boilers described in chapter 4.0.  The chapter includes
estimated total capital cost, annualized busbar cost
(hereafter referred to as busbar 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, 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 a discussion of 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,
busbar costs, and cost effectiveness values for the three
control options.  Appendix C.I presents an example calculation
illustrating the use of the costing procedures described in
this section.  This section also discusses other 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
                              6-1

-------
Manual.2   The major  components  of  capital  and  O&M costs  are
shown in table 6-1.   All costs are based on 1995 dollars.
6.1.1  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 Da
standard; as a result,  these costs are not included.
     The data used to estimate direct cost for each control
option were obtained from utility questionnaire responses,
vendor information,  published literature,  and other sources.
These data sources are specifically identified in the
appropriate section of appendix C.  These cost data were then
compiled into a data base, examined for general trends in
direct cost versus boiler size (i.e., megawatt [MW]), and
statistically analyzed using linear regression to obtain a
functional relationship of the form:
                 Direct Cost  ($/kW)  = a * MWb             (6-1)
     where:
          a    = Constant derived from regression analysis
          MW   = Boiler size (MW)
          b    = Constant derived from regression analysis
The direct cost for installing each NOX control option on the
model boilers was then calculated using equation 6-1 and the
specific values of "a" and "b."
     6.1.1.2  Indirect Costs.   Indirect costs include general
facilities,  engineering expenses,  royalty fees, and
contingencies.   General facilities include offices,
laboratories,  storage areas, or other facilities required for
installation or operation of the control option.  Engineering
expenses include the utility's internal engineering efforts
                              6-2

-------
EH
2;
o
CU
S
o
u

EH
CO
O
U
H

EH
O

Q
EH
H
&

~l ^1
cn o o
o> d d
-H 0)0)
jj cn cn
-H d d
rH -H -H
-H JJ JJ
o cn d d
rd d o o
HH -H U U
j_i
H CD jj cn
rd cu o cn
in d CU 0)
0) -H -n CJ
d cn o o
cu d M M
rh f T"{ p| p|




JJ
cn
o
u

jj
u
0)
M
-H
•o
d
M

rH
rd
j
rn
ft
u
rH
rd JJ
jj cn
o o
EH U



cn
rH
rd
-H
ll l_i tl
O O OJ
V-t O O jJ
O rd rd rd
£5 rH rH g
rd
rH 0) >1 0)
O M O
cn d o d
d rd cn rd
-H d -H d
JJ 0) > 0)
rd JJ SH JJ
in d 0) d
0) -H ft-H
ft rd 3 rd
O S cn S



jj
cn
o
u

S

o

T3
CU
^
•H
fa






jJ
CO
^s,
rH
>i rd rH
jj jj rd
rH rd cn
rd o >, O
d "~^ JJ ft
d) tn -H ro
ftrH O -H
rd -H T3
>i U H
tJl-rH JJ M 0)
r4 £ O 0) JJ
0) cu 0) JJ cn
d 45 rH rd rd
w u w S ^
jj
cn
o
o

S

O

0)
rH
Q
rd
-rH
J_|
rd
>
JJ
cn
o
CJ
§
O
rH
rd
jJ
O
E-1
                                           6-3

-------
and those of the utility's 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 is intended to cover
miscellaneous equipment and materials not included in the
direct cost estimate.   Project contingencies depend on the
level of detail included in the direct cost estimate.
Generally, the more detailed the cost estimate,  the less 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, the older and more mature the technology,  the less
process contingency required.  The specific project and
process contingencies vary with control option and are
included in appendix C.
     Indirect costs are estimated by multiplying the total
direct cost by an indirect cost factor.  The indirect cost
factor accounts for the indirect costs as a percentage of the
direct cost.
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 assumed to be independent of capacity
factor.  Variable O&M costs include any energy penalty
resulting from efficiency losses associated with a given
control option, costs associated with the disposal of
additional waste, and costs for the 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
                              6-4

-------
o
u
o

UJ
  1

as
<

oL
<

>
p



X


u.



fN
 i
O

rjj

_1

CQ
                                                                    m    (N


                                                                    oo"    —'


                                                                          
-------
estimated national average prices (in 1995 dollars) for the
year 2000, using the reference case analysis of the Department
of Energy's (DOE's) 1997 Annual Energy Outlook.3   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, water, and high
pressure steam are listed 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 Busbar Cost. Cost Effectiveness, and
       Incremental Cost Effectiveness
     Busbar cost is the sum of annualized capital costs and
total O&M costs divided by the annual electrical output of the
boiler (see equation C.6 in appendix C.I).  Busbar cost is
commonly expressed in mills/kWh  (1 mill = $0.001) and is a
direct indicator of the cost of the control technology to the
utility and its customers.  To convert total capital cost to
an annualized capital charge, the total capital cost is
multiplied by an annual capital recovery  factor  (CRF)  (see
equation C.7 in appendix C.I).  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-l]               (6-2)
     where:
          i  = interest rate as a decimal fraction (assumed to
               be 0.07)
          n  = the economic life of the equipment  (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 the sum of total annualized capital
costs and total O&M costs  (together referred to as total
annual costs),  expressed in dollars per year, by the annual
emission reductions, expressed in tons of NOX per year.

                              6-6

-------
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 C.I.
6.1.4  Other Cost Considerations
     Potential costs associated with solid and liquid wastes,
emission monitoring and compliance testing, and regulatory, and
enforcement agency activities were evaluated.  Only
incremental costs attributable to the revision of the standard
are considered.
     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 Da standard.
Therefore, there will be no change in potential solid and
liquid waste and no incremental cost associated with its
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.  However, any increase would be system
specific and cannot be predicted.
     6.1.4.2  Emission Monitoring and Compliance Testing.
There are no incremental emission monitoring costs associated
with any of the control options for monitoring NOX or oxygen.
The current subpart Da standard already requires NOX and
oxygen monitoring equipment.  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 an
outlet NOX concentration test similar to that already required
will be used.
                              6-7

-------
     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 for the
components of 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.
6.2.1  Combustion Controls
     The specific combustion controls considered under this
control option are low NOX burners (LNB) for wall-fired
boilers and LNB with close-coupled overfire air (OFA)  for
tangentially-fired boilers.  As discussed in chapter 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.13'14-15  As a result,  there
are no combustion control costing procedures presented for FBC
boilers.
     As discussed in chapter 3.0 and 4.0,  limited data are
available on the NOX control performance of combustion
controls applied to boilers burning natural gas or oil.
Therefore, combustion control costing procedures for natural
gas- and oil-fired boilers are not presented.
     As shown in chapter 3.0, combustion controls currently
being installed in coal-fired utility boilers are capable of
achieving lower NOX emission levels when carefully operated.
Consequently, because the implementation of this control
option requires no additional hardware, no incremental capital
costs were assumed.  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.  Typical O&M costs
                              6-

-------
for combustion systems burning solid fuels are 3 to 6 percent
of direct costs.16
     Direct costs for LNB applied to wall-fired boilers
burning coal were based on data obtained from 10 units,
ranging in size from 130 to 800 MW.  These data included seven
cost estimates and three actual installation costs.  These
data are summarized in appendix C.2.  Converting the units of
the direct cost equation (see section C.2.2 of appendix C),
from $/kW to $/yr and multiplying by 3 percent yields the O&M
cost equation:
             O&M  ($/yr) = 3% *  (220,000 * MW°-56)         (6-3)
     No direct cost data were available for LNB with close-
coupled OFA applied to tangentially-fired boilers.  Therefore,
vendor information on the relative cost of LNB with close-
coupled OFA and LNB with close-coupled and separated OFA was
used to develop the direct cost algorithm for LNB with close-
coupled OFA.  This information indicates that LNB with close-
coupled OFA direct costs are approximately 55 percent of
direct costs for LNB with close-coupled and separated OFA.1"
Converting the units of the direct cost equation  (see
section C.3.2 of appendix C)  from $/kW to $/yr and multiplying
by 3 percent yields the O&M cost equation:
              O&M ($/yr)  =  3% * (80,000 *  MW0-60)          (6-4)
6.2.2  Selective Noncatalytic Reduction
     Costing procedures presented reflect a low-energy, urea-
based SNCR system.  Because the cost estimates for a low-
energy, urea-based SNCR system were found to be comparable to
a high-energy SNCR system,  results are only presented for the
low-energy,  urea-based SNCR system.  This control option is
applicable to all three furnace types  (wall, tangential, and
FBC) represented by the model boilers.  Each boiler applying
SNCR was assumed to have two levels of wall injectors and one
level of lance injectors.   Because FBC units are typically
smaller and have different operating characteristics than
wall- or tangentially-fired boilers, these units have a
                              6-9

-------
greater likelihood of needing less than three levels of
injectors.  If two levels of injectors were eliminated on the
FBC units, cursory analysis indicates that levelized
technology costs could decrease 40 percent.
     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.  Vendor cost
estimates were used to develop the direct cost equations
below:18
     For coal-fired boilers;
               Direct Cost ($/kW)  = 32 *  MW°-24           (6-5)
     For natural gas- or oil-fired boilers;
               Direct Cost ($/kW)  = 31 *  MW°-25           (6-6)
     As discussed in appendix C.6, the indirect cost factor
was assumed to be 1.3.
     Vendor cost estimates were also used to estimate fixed
O&M costs.  These costs include operating, maintenance, and
supervisory labor; and maintenance materials.  Fixed O&M costs
were found to be independent of fuel type.  Simplified
algorithms presented below were developed from the vendor
estimates:]q
                 FO&M ($/yr)  = 86,000 * MW°-21            (6-7)
     Variable O&M costs include the cost  of the urea solution
(chemical costs); energy penalties associated with mixing air
gas loss, and vaporization of the urea solution in the boiler;
and electricity costs for equipment operation.  Cost of the
urea solution is the majority (90%) of variable O&M costs 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  multiplied by the
fuel cost to determine the annual cost.  Electricity costs
were determined as a function of unit size and urea injection
                             6-10

-------
rate.  Appendix C.4 presents the equation for calculating urea
cost and the basis for other variable O&M costs.
6.2.3  Selective Catalytic Reduction
     Costing procedures for SCR are based on the SCR module in
version 4.0 of EPA's IAPCS,4 published SCR cost  information,20'21
utility questionnaire responses,22'23 and a  draft  cost report
prepared by EPA's Acid Rain Division25.
     This control option has not been demonstrated on FBC,
boilers and may not be technically feasible due to the high
likelihood of catalyst poisoning resulting from high levels of
calcium oxide present in the fly ash.  Further,  the relatively
higher cost of SCR compared to SNCR makes the application of
SCR economically prohibitive.24  Therefore, this control option
was not considered for FBC boilers.
     Catalyst price, which has a significant impact on costs,
was assumed to be $350/ft3 for coal-, natural gas-, and oil-
fired boilers.  Catalyst life was assumed to be 3 years 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 oil-fired
boilers, and 14,000 for 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 four
fuel and furnace type combinations are:
Fuel
Coal
Oil/Gas
Boiler type
Wall
Tangential
Wall
Tangential
a
174
165
165
156
b
-0.30
-0.30
-0.329
-0.324
The equation for estimating initial catalyst charge costs was
also based on IAPCS model estimates and is presented in
appendix C, section C.5.2.
                              6-11

-------
     Indirect cost factors 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.  A
range of 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 C,
sections C.5.4 and C.5.5.
6.3  MODEL BOILER COST IMPACTS
     Model boiler impacts have been developed for three
control options:  combustion controls,  combustion controls
with SNCR, and combustion controls with selective catalytic
reduction.  Performance of the control options is discussed in
chapter 3.0.  For costing purposes, combustion controls are
assumed to achieve a fixed emission rate depending on the fuel
and furnace type.  Selective noncatalytic reduction and SCR
systems are assumed to achieve a constant emission reduction
of 50 and 80 percent,  respectively.
     Model boiler impacts are presented in five tables
included as appendix D:  one for each coal type  (bituminous,
subbituminous, and lignite), natural gas, and oil.  Each table
consists of two parts.  The first part presents emissions and
emission reductions, and the second part presents capital and
busbar costs and cost effectiveness values.  Emissions and
emission reductions are discussed in chapter 5.0.
     As discussed previously, the specific data presented in
appendix D include capital costs, busbar costs, cost
                             6-12

-------
effectiveness values, and ICE values.  Observations on the
data focus on ICE.
     Table 6-3 presents ranges for ICE for the three control
technologies applied to the model boilers at a capacity factor
of 0.65.  These data are discussed in more detail in the
following sections.  The combustion control cost of zero for
fluidized bed combustion (FBC) boilers reflects the ability of
existing FBC boilers to achieve the lower emission rate   ,
without additional combustion controls.  The cost of
combustion control is also zero for boilers burning natural
gas or oil because an emission rate lower than the existing
standard was not evaluated.
     In the following three sections, model boiler impacts for
each control option are evaluated to determine the effect of
boiler size, boiler capacity factor,  furnace type, and fuel
type on ICE.
6.3.1  Combustion Controls
     Because the estimated cost of applying combustion
controls to FBC boilers and to all boilers burning natural gas
or oil is zero,  the discussion in this section is limited to
coal-fired wall and tangential boilers.
     Figure 6-1 presents the effect of fuel type and furnace
type on ICE for combustion controls.   The data presented are
for a capacity factor 0.65.  As discussed previously,
annualized costs for combustion controls were limited to
3 percent of direct costs to account for additional burner and
boiler O&M costs.  As a result, there are no elements of total
annual cost sensitive to coal type.  In addition, the
estimated emission reductions for each combination of furnace
type and coal type are identical (e.g., baseline NOX rate of
0.58 or 0.48 Ib/MMBtu and controlled NOX rate of 0.40 or
0.30 Ib/MMBtu is a change of 0.18 Ib/MMBtu).   As a result, ICE
for the three coal types is the same for a given furnace type.
     Busbar costs change as furnace type changes.  This
results from the difference in the annualized costs for
combustion controls applied to each furnace type.  The busbar
                             6-13

-------
u
  CO

  2
O
  EH
  CO
M
O
CO
o
o
n
 I
VD
•J
CQ
^"^ I)
tA T*
83
•H w
"c ®
1 >
n •*•*
C °
H »4H
0)
. Cn
02
4J g
C ,-;
O (j
U Q)
4-J


0)
a
4-)
a)
u
(0
c

3







(U
a

4->

rH
0)
3






0
CO
r- i
I
0











u
u

O
CM
t-l
1
O
CO
t^.





«
u

co
+
U
u

o
n
0
I
o
CO
o
«
VO





PH
u
CO
+
u
u










r-H
rH
(0
3

0
CO
1
o
CO











u
u

O
1
o
CO
f-





PJ
u

to
+
u
u

o
o
1
o
CM
o
w
VO





tf
u
co
+
u
o




1— 1
(0
•H
4J
C
0)
C
flj
E-1


o












u
u

o
in
n
I
o
a*
CO
».
H




a
u
2
CO
+
U
u


2








Oi
U
to
+
u
u










u
en
fan


rH
(0
0
u

(0
3
O
C
•H
g
3
•H
CD

0
CO
1
O
VO











U
U

o
VO
t-l
1
o
in
CO





«
U
2
to
+
U
u

o
o
VO
cn
1
O
n
>.
CO





OH
U
CO
+
u
u









	 i
f~n
rH
fO
&

o
CO
1
o











u
u

O
VO
t-H
1
0
in
CO





«
U
2
to
+
U
u

o
1
o
in
CM
>.
CO





OH
U
to
+
u
u




rH
(Q
•H
4J
C
0)
C
(0
EH


o












u
u

0
in
i
o
CTi
CO
«
H




a
u
2
to
+
U
u



-------
w

W ^
KrrT
w
SB
ss
a
g
W
iJ
CQ
<
E"
"cT
Incremental cost
f ectiveness ($/to
«M
<
. IT
0°
Contr
techno]



a)
a
>i
-p
HI
UJ
o
rC
C
t.
^
0
Cn



Q)
a
>.
4J
rH
CD
3
[n





0
00
H
1
O
vo




u
u


O
•0-
CM
V
H
1
O
CO
r~



04
CJ
Z
to
+
a
u

O
n
0
r>
i
o
CO
0
VO



04
u
co
1
T
CJ
CJ









i*H
(H
m


o
CO
1
0
CO




u
CJ

o
f
CM
^
H
1
O
00
r»



04
U
Z
CO
u
CJ

0
0
0\
VO
1
o
CM
o
«.
vo



04
u
CO
CJ
CJ





rH
ro
•H
.p
c
0)
rr\
CP
C
a
EH

o




u
CJ

0
in
r*
w
m
1
o
cri
00
^
H



04
U
Z
CO
CJ
u

<
Z



04
u
CO
CJ
u









u
CQ
fn





Q)
-P
fc_J
^1
C
IT
•H
J

0




u
CJ

o
OJ
H
*.
CM
1
O
VD
O
«
rH



(X
CJ
Z
CO
CJ
CJ

o
r>
CM
VO
1
o
o
f*)
^r



tf
CJ
CO
_l-
CJ
CJ
1— 1
m
•H
•p
c
a)
CT
(0
4-)
T3
C
(0
,_-)
rH
rt
S


in
re
CT

rH
<0
tl
W
3
4J
re
Z

o




o
CJ

0
r-
m
V
.H
1
0
o\
CO



K
U
Z
CO
+
u
CJ

o
en
vo
*»
rf
1
o
in
*r
•»
ro



04
U
CO
+
U
CJ
rH
to
•H
4-J
C
0)
?Tl
O1
C
CO
4->
TS
C
fO
rH
i
4J rH
C flJ
O 4J
CJ (0
U
C
O Q)
•H >
4-> -H
U 4->
3 O
£ 0)
6 rH
O .
iJ rH
e(T)
*w
O 4->
CJ  rH
N -H Oi
-H 4-> a
T3 U <
•H Q)
3 rH 4->
rH Q) O
(t, CO Z


II II II
04
rj cj
03 Z <
fe CO Z

».
0 U
O CQ
VO [3-1
- h
0 O
O (M
f)
^ ^^
O
0 0 '
r-l O
^•H

cn -o
0) C
N (0
•H
w in
CM
to
CU TJ
rH C
•H (0
O
.Q W
1 ,
M

CJ C  4J
Q) U
W T3 (0
CU C i
0) 1 4->
to rH -H
•H U
•a (0 ro
Q) > a
4-> fC
C >-i U
0) O
W 
-------
u.
i
*
                                                              E
                                                              .2

                                                              I
                                                                          r-l
                                                                          o
                                                                          o
                                                                          u

                                                                          c
                                                                          o
                                                                          •H
                                                                          4J
                                                                          CO
                                                                          o
                                                                          u
                                                                          w
                                                                          u
                                                                          o

                                                                          QJ
                                                                          OJ
                                                                          u
                                                                          to
                                                                          a
            cu

            a
            4J
                                                                          OJ
                                                                          4J
                                                                          o
                                                                          Q)
                                                                          w
(uoj/S) s»QaAij33jj3
                             6-16

-------
costs of combustion controls range from 0.05 to 0.33 mills per
kilowatt-hour (mills/kWh) for wall-fired boilers, and from
0.02 to 0.14 mills/kWh for tangentially-fired boilers.  The
estimated emission reductions for each furnace type is the
same (0.18 Ib/MMBtu)  regardless of coal type.  The net effect
is a higher ICE for wall-fired boilers when compared to
tangentially-fired boilers as shown in figure 6-1.  The shape
of individual curves is the result of changing boiler size.
     Figure 6-2 presents the effect of capacity factor and
boiler size on ICE for combustion controls.  Curves are
presented for both wall- and tangentially-fired boilers to
indicate the similar nature of the effect of capacity factor
and boiler size on ICE.  Incremental cost effectiveness
decreases as capacity factor and boiler size increase.  The
ICE for combustion controls decreases as capacity factor
increases because of the combined effects of decreasing busbar
costs (mills/kWh)  resulting from increased electricity
production,  and increasing NOX emission reductions.  The ICE
for combustion controls decreases as boiler size increases
because of the effect of economy of scale on capital costs.
6.3.2  Selective Noncatalytic Reduction
     This section discusses ICE for the application of SNCR to
all furnace types (wall- and tangentially-fired and PBC)  and
all fuel types (coal, oil, and natural gas).  Figure 6-3 shows
the effect of inlet NOX emission rate on ICE for SNCR.  Curves
are presented for five different inlet NOX emission rates
representing the controlled emission rates achieved by
combustion controls for the possible combinations of furnace
type and fuel type.   The effect of inlet NOX emission rate is
seen in the relative position of each curve on the figure,
while the shape of individual curves is the result of changing
boiler size.
     These curves show that as the inlet NOX emission rate
increases, ICE decreases.  This is the result of changes in
both busbar costs and NOX emission reductions.  While busbar
costs increase as inlet NOX emission rates become larger, NOX
                             6-17

-------

                                                      w
                                                      1-1
                                                      o
                                                      c
                                                      o
                                                      u

                                                      c
                                                      o
                                                      •H
                                                      4J
                                                      to
                                                      o
                                                      u
                                                      H
                                                      u
                                                      c
                                                      o

                                                      0)
                                                      N
                                                      •H
                                                      CO

                                                      •a
                                                      c
                                                      o
                                                      AJ
                                                      u
                                                      (0
                                                      14-1
                                                      4J
                                                      •H
                                                      u
                                                      fO
                                                      u

                                                      i-l
                                                      •H
                                                      O
                                                      X!

                                                      U-l
                                                      o

                                                      .u
                                                      u
                                                      0)
                                                      IM
                                                      u-i
                                                      W
                                                      fN
                                                       I
                                                      OJ
6-18

-------
                                               OJ
                                                c
                                                0)
                                                U
                                                M
                                                0)
                                                o
                                                in
                                                c
                                                o
                                               •H
                                                AJ
                                                U
                                                3
                                                •o
                                                o
                                                •H
                                                CO
                                                CO
                                                •H
                                                6
                                                0)
                                                c
                                                rt
                                                JJ
                                                CO
                                                c
                                                o
                                                U
                                                 CO
                                                 (U

                                                 0)
                                                •H
                                                £
                                                 U
                                                CO
                                                 cu
                                                 JJ

                                                 £
                                                     oc
                                                     CO
                                                     W
                                                     U
d
o

0)
4J
ns
 c
 o
•H
 W
 CO
•H
 e
 x
o
 4J
 0)
 i—l
 c
 •H
 JJ
 U
 <1J
 U-l
 ro
  i
 vc

 0)
6-19

-------
emission reductions increase at a faster rate.   For a 300 MW
boiler with a capacity factor of 0.65,  when the inlet NOX
emission rate increases from 0.18 to 0.40 Ib/MMBtu, busbar
costs increase by 55 percent while NOX emission reductions
increase by 134 percent.  The net result is a decrease in ICE
of 34 percent.  The large ICE for the 25 MW FBC boiler is a
result of the higher busbar costs for a small boiler.
     Figure 6-4 presents the effect of boiler size and
capacity factor on ICE for SNCR.  Two sets of curves are
presented,  one for boilers with an inlet NOX emission rate of
0.40 Ib/MMBtu (representing the post-combustion control
emission rate for wall- and tangentially-fired boilers burning
either bituminous coal or lignite) and one for boilers with an
inlet NOX emission rate of 0.18 Ib/MMBtu (representing the
post-combustion control emission rate for natural gas-fired
boilers).   Incremental cost effectiveness decreases as boiler
size and capacity factor increase for the same reasons given
for combustion controls.  Except for small boilers  (<300 MW) ,
the effect  of boiler size and capacity factor on ICE is
relatively small as seen in the relative flatness of the
curves for these boilers on figure 6-4.  This behavior
reflects the relatively low capital cost and high variable
operating cost of SNCR.  For example, a change in size from
1,000 MW to 300 MW for the coal-fired boiler increases ICE by
only 20 percent.  Further, a decrease in capacity factor from
80 percent  to 30 percent for the 300 MW coal-fired boiler
increases ICE by only 40 percent.
6.3.3  Selective Catalytic Reduction
     This section discusses ICE for the application of SCR to
the applicable furnace types (wall and tangential) and all
fuel types (coal, oil, and natural gas).  As discussed in
chapter 3.0,  SCR was not considered as a control option for
FBC boilers.   Figure 6-5 presents the effect of inlet NOX
emission rate and fuel type on ICE for SCR.  Curves are
presented for four different inlet NOX emission rates
representing the controlled emission rates achieved by
                             6-20

-------
£

00

o
                                                                        £

                                                                        o
                                                        0)
                                                        u
                                                        M
                                                        0)
(uO}/$)
JSCQ
                                                        C
                                                        O
                                                        •H
                                                        jj
                                                        u
                                                        3
                                                        •O
                                                        0)
                                                        C
                                                        O
                                                        -H
                                                        CO
                                                        CO
                                                        •H


                                                        0)


                                                        .U

                                                        C
                                                        (C
                                                        c
                                                        O
                                                        u

                                                        m

                                                        to

                                                        §1
                                                        (U
                                                        •H
                                                        .C
                                                        U
                                                        ro



                                                        1
        0)
        4J

        §
                                                             CO

                                                              ^
                                                              o
                                                             *w

                                                             u
                                                             u
                                                                                      0)
                                                                                      N


                                                                                      CQ


                                                                                     T5
                                                                                      C
                                                                                      rt
                                                                                      o
                                                                                      .u
                                                                                      u
                                                                                      (0
                                                                                     •H
                                                                                      u
                                                                                      (C
                                                                                      u
                                                                                      Q)
                                                                                      .H
                                                                                      •H
                                                                                      O
                                                                                      U-l

                                                                                      O
                                                                                      U
                                                                                      0)
                                                                                      U-l
                                                                                      0)
                                                                                      •H
        6-21

-------



"u
a
Ifc
8
ra
c
U.




u
*-
n
*
* u
U k U
M ' '
00 00 O
** n f.
odd

u
'c
00
J
Q
V?
3
O
c
|
Q
•d

.t
i
o
d

                                                                ©

                                                                D
                                                                                   c/

                                                                                   _w

                                                                                   "o
                                                                                   m
                                                                              l-l
o
fS
                                                                                                  JJ
                                                                                                   G
                                                                                                   0)
                                                                                                   O

                                                                                                   0)
                                                                                                   CU

                                                                                                  O
                                                                                                  oo
 c
 0
 u

•O
 0)
                                                                                                   c
                                                                                                   o
                                                                                                  •H
                                                                                                   CO
                                                                                                   CO
                                                                                                  •H
                                                                                                   E
 (C
JJ
 CO
 c
 o
 u
 CO
 d>

 0)
-H
.C
 U
 rt
u
w
                                                                                                        u
                                                                                                        w

                                                                                                        M
                                                                                                        O
                                                                                                        U
      C
      o

      cu
      JJ
      m
 c
 o
•H
 CO
 CO
•H

 CU


ox
      0)
      rH

      C
      •H
                                                                                                        JJ
                                                                                                        U
                                                                                                        0)
                                                                                                        w
in
 i
U)

(U
                                                                                                        •H
                                                                                                        Cu
                                          JSO3
 CD
 JJ
 O
 12
                                                   6-22

-------
combustion controls on wall-fired boilers for each fuel type.
The effect of inlet NOX emission rate is seen in the relative
position of the two curves for the coal-fired boilers in the
figure.  These curves show that as inlet NOX increases, ICE
decreases.  This is the result of changes in both busbar costs
and emission reductions, with emission reductions increasing
at a faster rate.
     Although the inlet NOX levels for natural gas- and oil-
fired boilers are lower than for coal-fired boilers, the ICE
values reflected in figure 6-5 are lower because of the lower
capital and fixed O&M costs for these fuel types.  The
differences in flue gas contaminants and their impacts on SCR
system design result in SCR systems for natural gas-fired
boilers costing less than systems for oil-fired boilers and
systems for coal-fired boilers being the most expensive.
     Figure 6-6 presents the effect of boiler size and
capacity factor on ICE for SCR.  Two sets of curves are
presented, one set for boilers with an inlet NOX emission rate
of 0.40 Ib/MMBtu (representing the post-combustion control
emission rate for wall-fired boilers burning bituminous coal
or lignite) and the other set for boilers with an inlet NOX
emission rate of 0.18 Ib/MMBtu (representing the post-
combustion control emission rate for natural gas-fired
boilers).  Incremental cost effectiveness decreases as boiler
size and capacity factor increase for the same reasons given
for combustion controls and SNCR.   Because of the higher
capital cost for SCR versus combustion controls and SNCR, the
impact of capacity factor on ICE is more pronounced.  For
example, a decrease in capacity factor from 80 percent to
30 percent results in an increase in ICE of over 120 percent
for all sizes of the coal-fired model boiler.  The set of
curves representing the natural gas-fired boiler have lower
ICE values due to the lower capital and fixed O&M costs
mentioned earlier.
                             6-23

-------

I ' «

1
6 V

1
:> T

1
t f
2 2 H
i i i i i L i i I
^ O ofc ~f> V C

, ' J
                       ^
                       \»^
                        u


                       [2
                       £>
                       'o
                        eg
                        a.
                              5
                              CQ
oo

O
                             ,£

                             O

                             O
                                             cO

                                             CO
                                             0)

                                             0)
                                             •H
                                             .£
                                             U
                                             fO
                                             U
                                             CO
                                                  u
                                                  CO
                                                  U


4J
C
0}
u
0)
tt
0
00
U-l
o
C
O
•H
4-)
U
T3
CD

C
O
•H
CO
CO
•H
e
0)
JJ
cs
tC
CQ
c
o
0
c
o
size
•0
c
fO
.
o
CJ
m
U-l
.
»Y
•H
U
ro
m
u

0)
rH
-H
O
•Q
14-4
o
4J
U
cu
M-l
14-4
w
                     I
                     vo

                     (D
                                             0)
                                             4J
                                             o
6-24

-------
6.4  REFERENCES


1.    EPRI (Electric Power Research Institute).  TAG™
     Technical Assessment Guide,  volume 1:  Electricity
     Supply - 1989 (Revision 6).  EPRI P-6587-L, Palo Alto,
     CA.  November 1989.  pp. 3-1 through 3-14.

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.    U. S. Department of Energy, Office of Integrated Analysis
     and Forecasting.  Annual Energy Outlook 1997.  DOE/EIA-
     0383(97).  Washington, DC.  December 1996.

4.    U. S. Environmental Protection Agency, Integrated Air
     Pollution Control System,  Version 4.0, Volume 2:
     Technical Documentation Manual.  EPA-600/7-90-022b.   Air
     and Energy Engineering Research Laboratory, Research
     Triangle Park, NC.  December 1990.  pp. 4-77 through
     4-97.

5.    Telecon.  Illig, C.,  Radian Corporation,  with Millard,
     D., National Ammonia Company.  SNCR Chemical Costs  (Urea,
     Anhydrous, and Aqueous Ammonia).  January 4, 1993.

6.    Fax.  Vereah, J., LaRoche, Industries to Illig, C.,
     Radian Corporation.  SNCR Chemical Cost.   January 7,
     1993.

7.    Telecon.  Illig C., Radian Corporation, with Martina, T.,
     LaRoche Industries.  SNCR Chemical Cost.   January 4,
     1993.

8.    Telecon.  Illig, C.,  Radian Corporation to Vereah, J.,
     LaRoche Industries.  SNCR Costing - Aqua + Anhydrous
     Ammonia.  January 4,  1993.

9.    Telecon.  Illig C., Radian Corporation, with
     Moredyke, D., UNOCAL Corp. SNCR Chemical Cost, N0x0ut A
     Costing.   January 4,  1993.

10.  Letter from Poole, M.  F.,  W. H. Shurtleff Company to
     Illig,  C., Radian Corporation.  SNCR Chemical Cost for
     Urea + N0x0ut A.  January 6, 1993.

11.  Telecon. Illig,  C., Radian Corporation with Kellog,  G.,
     Nalco Fuel.   SNCR Chemical Cost for Enhancers.
     January 5, 1993.
                             6-25

-------
12.  Fax.   Miskus,  J.,  Cargill,  Inc. to Illig,  C., Radian
     Corporation.  SNCR Chemical Cost for Urea + NOxoutA.
     January 13,  1993.

13.  Telecon.   Stone,  J.,  Ahlstrom Pyropower,  Inc. with
     King,  B.,  Radian Corporation.  Fluidized Bed Combustion
     Boilers.   August 18,  1993.

14.  Telecon.   Czarnecki,  T.,  ABB-CE, with King, B., Radian
     Corporation.  Fluidized Bed Combustion Boilers.
     August 18, 1993.
                                                          9
15.  Telecon.   Edvardsson,  C., Tempalla Power with King, B.,
     Radian Corporation.   Fluidized Bed Combustion Boilers.
     August 17, 1993.

16.  Ref.  1.  p.  3-14.

17.  Grusha, J. and McCartney, M.  S.  Development and
     Evolution of the ABB Combustion Engineering Low NOX
     Concentric Firing System.  TIS 8551.  ABB Combustion
     Engineering Service,  Inc.  Windsor,  CT.  1991.  p. 9.

18.  Letter and attachments from R. D. Pickens,  Nalco Fuel
     Tech,  to E.  Soderberg, Radian Corporation.   February 8,
     1992.

19.  Letter and attachments from R. D. Pickens,  Nalco Fuel
     Tech,  to N.  Kaplan,  U. S. Environmental Protection
     Agency.  January 20,  1992.

20.  Cochran,  J.  R., et al.  Selective Catalytic Reduction for
     a 460  MW Coal  Fueled Unit:   Overview of a NOX Reduction
     System Selection.   Presented at the 1993 Joint Symposium
     on Stationary  Combustion NOX Control.  Bal  Harbour, FL.
     May 24-27, 1993.

21.  Electric Power Research Institute.  Technical Feasibility
     and Cost of Selective Catalytic Reduction  (SCR) NOX
     Control.   EPRI GS-7266.  Palo Alto,  CA.  May 1991.

22.  Fax.   J.  Klueger,  Los Angeles Department of Water and
     Power, to M. J. Stucky, Radian Corporation.  September 9,
     1993.

23.  Letter and attachments concerning SCR costs.
     Confidential Business Information.  Reference
     Number 931970111-01.

24.  Telecon.   Evardsson,  C.,  Tempalla Power with King, B.,
     Radian Corporation.   Application of Selective Catalytic
     Reduction to Fluidized Bed Combustion Boilers.
     September 27,  1993.


                              6-26

-------
25.  U. S. Environmental Protection Agency,  Cost Estimates for
     Selected Applications of NOX Control Technologies on
     Stationary Combustion Boilers, Draft Report.  Acid Rain
     Division, Washington, DC.  March 1996.
                             6-27

-------
                  Appendix A

Determination of Furnace Volume and Burner Zone
  Volume  for Conventional Subpart  Da Boilers

-------
                           APPENDIX  A

Determination of Furnace Volume and Burner Zone Volume for
Conventional Subpart Da Boilers
   This section describes  the procedures used to determine the
missing data on furnace and burner zone volumes for the
conventional subpart Da boilers.  These data were requested in
the Utility Air Regulatory Group  (UARG) questionnaires* sent
to all operators of conventional  subpart Da boilers.  However,
some respondents did not provide  this information.
   Besides the  questionnaire data, the PowerPlants database was
used as an additional source to determine the missing
information.  The PowerPlants database included values for the
total furnace volume for some subpart Da boilers.  The boiler
dimensions used for this computation were not reported.  To
compare the PowerPlants database  with the questionnaire data,
the furnace volume from the questionnaire data was calculated
from:

        Vf   =     H  * W  *  D                               (A-l)

where   Vf   =     furnace  volume  (ft3)
        H     =     height of boiler (ft)  as measured from  the
                  top of the hopper  to the roof  of the boiler
        W     =     width  of  the boiler (ft) and
        D     =     depth  of  the boiler (ft).

   In  figure  A-l,  the calculated  furnace volumes using
questionnaire data (Vf)  are plotted as a function of the total
furnace volume obtained from the  PowerPlants database (v^) for
10 boilers for which furnace volumes could be determined from
both sources.  A regression analysis (with zero intercept) of
     "Hereafter  referred  to  as  questionnaires.
                             A-l

-------
                                                                                                                                                       e
                                                                                                                                                       O QJ
                                                                                                                                                       i-l W
                                                                                                                                                      •O  (0
                                                                                                                                                       0)  -U
                                                                                                                                                       c  ro
                                                                                                                                                      •H  T3
                                                                                                                                                       (0
                                                                                                                                                      JJ  W
                                                                                                                                                      .0  4J
                                                                                                                                                       O  C
                                                                                                                                                          (T3
                                                                                                                                                       W  rH
                                                                                                                                                       0}  Ou
                                                                                                                                                       E  H
                                                                                                                                                       3  0).
                                                                                                                                                        .
                                                                                                                                                       C
                                                                                                                                                       ^ -o
                                                                                                                                                       3  C
                                                                                                                                                      «w  rt ,

                                                                                                                                                       C  tC (
                                                                                                                                                       O 4->
                                                                                                                                                       Q)  fO r

                                                                                                                                                      4J
                                                                                                                                                       0)  0)

                                                                                                                                                         -H .
                                                                                                                                                       c m
                                                                                                                                                       o ci
                                                                                                                                                       to c
                                                                                                                                                      •H O
                                                                                                                                                       i-l -H

                                                                                                                                                       D< CO
                                                                                                                                                       E a)
                                                                                                                                                       O p
                                                                                                                                                      U O1'
                                                                                                                                                       i
                                                                                                                                                      <
                                                                                                                                                      0)
                                                                                                                                                      i-l
                                                                                                                                                      g,
                                                                                                                                                      •H
8
                                                                                           A-2

-------
these data resulted in the following equation (correlation
coefficient, r2 = 0.98)
                        Vf = 1.007 * Vd                   (A-2)

   Equation A-2  was  used to calculate furnace volumes  (Vf)  for
boilers where the furnace volumes (V
-------
                                                             1
                                                     8
                                                                      Q)

                                                                      O
                                                                      N
                                                                      0)   .
                                                                      C  u
                                                                      V-l  M
                                                                      3  0)
                                                                     JQ^H
                                                                        •H
                                                                     •o  o
                                                                      C JQ
                                                                      tO
                                                                         «
                                                                      0) Q

                                                                      3 4J
                                                                     >H  ^
                                                                      O  (t
                                                                      >  a
                                                                        XI
                                                                      0)  3
                                                                      u  w
                                                                     u-i
                                       (C
                                       C
                                       o
                                      •nt
                                                                      <1J  C
                                                                      Q)  (U
                                                                      S  >
                                                                      4J  C
                                                                      0)  O
                                                                      C  H
                                                                      O  O
                                                                      (0 u-i
                                                                     -H
                                                                      k  0)
                                                                      to  E
                                                                         O
                                                                         >
                 8
8      8
CM      i-
                                                                     (N
                                                                      I
                                                                      0)

                                                                      i.
'A 'eujnioA
                 A-4

-------
personnel2 indicated that the burners arrangement was changed
from three rows to four rows which also resulted in an
increase in burner zone volume.  Thus, as seen in figure A-2,
the boiler has a proportionately larger burner zone volume
when compared to the remaining boilers.
   A regression analysis (with zero intercept)  of the data for
the remaining 18 boilers  (except Brandon Shores Units 1 and 2
and Zimmer Unit 1) resulted in the following equation (r2 «
0.97) for total furnace volume as a function of burner zone
volume.

        Vf =  (3.70)  VBZ                                   (A-4)

   Equation A-4  was  used to determine total  furnace  volume (Vf)
from calculated burner zone volume (VBZ) for boilers with
burner zone dimensions reported in the questionnaire
responses.  Additionally, when only total furnace volume was
known, this equation was used to determine the corresponding
burner zone volume.
   Thus  using  equations A-2  and A-4,  the total  furnace and/or
burner zone volumes were calculated for boilers for which
these data were not provided in the questionnaire responses.
                              A-5

-------
REFERENCES


1.      Letter  and  attachments  from Brownell, W.  F.,  Hunton and
        Williams, to  Eddinger,  J. A., U.  S.  Environmental
        Protection  Agency.  December  18,  1992.  Response to NOX
        information request - Brandon Shore  Units l  and 2.

2.      Personal  communication, M. Gundappa, Radian  Corporation
        with J. Roebel, Cincinnati Gas  and Electric  Company,
        May 28, 1993.
                              A-6

-------
                     Appendix B



Primary and Secondary Air Impacts from Model Boilers

-------
                          APPENDIX B.I

                NOX  EMISSIONS  FROM MODEL BOILERS
Table B.l-1         NOX Emissions from Bituminous Coal-Fired
                    Model Boilers

Table B.l-2         NOX Emissions from Subbituminous Coal-Fired
                    Model Boilers

Table B.l-3         NOX Emissions from Lignite-Fired Model
                    Boilers

Table B.l-4         NOX Emissions from Natural Gas-Fired Model
                    Boilers

Table B.l-5         NOX Emissions from Oil-Fired Model Boilers

-------
0)
I
o
CD
0)
CO
o
o
(0
o
ffi
o
co
E
UJ
x
O
CO
n


-.
I/I
o
s
o
D
0)
n
Emissi




re
re
Control
jstion
Comb
01
Ol
in
(0
D







Tt"
cr
O
rn
U
U
"G"
tr
O

O
0

C)
u

"re
c


"5
c
£•
T)
re
Q.
re
0
"01
•c
2.

Furnace

^

g
£
m
I

g
^
C
o

si
c •£-
11
LU *"
0) CO
ro 2
cr 5
S
i/>
11
LU
O
u
(0
u.
"i
3
Z
0) 5"
N >
CO £,
o>
D.
h-
O>
O.
1-
"5J
LL
g

eg
g
in



8
CM


CO
O
CO
in

O
0
o
CO


o
CO


o
o
"re
| Bituminous
g
CO

s
§



^-


CO
I

o
0
o
.n

vO
in
CD
CM



13
o
O
§

eg
8



o
r-
in


CO
8
CM

O
°
o
CO
CO

^0
O
CO
CO




g
CM
ID
CO
g
(O



o
CO


CO
o
CO

0
o
o
in
CM

CO


O
O
CO


8
*
ID
CO
g
o
TO


8
r~

CO
g
CM
CO
O
o
o
r-

vO
in
CO
in




8
*
CO
CO
1



8
*~

CO
o
CD
CO
o
°
8
in
v«
O
CO
ID




O
CO

eg
8
CO
CO


8
^™

CO
o
CO
o
o
o
CO
o
un
v.C
O
CO


1


o
CO
eg
g
tO

§ o CO to CO o to to U) o S! CM CO 8 CD O o o in CD o S O) § 8 to to o CO CO o in o 0 o O) o o CO o g 8 in to CO 8 CM CO o CO in o o o CO CO CO o CO CM O in CO (O CO o CO CM O ID to O CO * CO 0 CM CO 0 C c 8 CD in in CD CO O m to CO o g ^ to to § in CO o CD CM O C 8 CM CO O CO Z Z o o o N 0 o 0 o TT O CO in in CM O CO LL Z Z 8 £ ^ o 0 0 8 CM in CO CD Z Z o CM CM 0


-------
M

Ji
"5
CD

75
TJ
O


•o

SI
il

15
o
O
tn
3
O
C
3
o
I/)
c
_o
"w
.2


UJ

X
O
m
o>
S
CB


D_
g
0
1

Emissic




ro
ra
Control
Combustion
a-

TO
D







rr
cc
0
ro
+
O
O
"o"
CC
o
A.
o
o

o
u

"(5
<

1
c
<

(0
&
O
"5
T3


Furnace

1
O
£
s
I

g
1
o
**

g
in
S^
VI ~>
•i I
c * —
UJ
Rate
(Ib/MMBtu)
m
1 £
11
LU
k.
S
LL
k_
O)
JD
•z.
0 5*
•* €
V) 1,
0)
Q.
01
Q.
"5
3
LL
0
CO

CO
co
8
^-



8
CM

cn
CO
0
CO
TT

O
CO
o
8
CO


6*!
o
CO


8
"S
| Subbitummous
o
v-

co
CO
8
CO



8
T

CO
CO
o
CO

o
CO
d
o
o
CO


0*
m
CO
CM
CO



"re
0
0
8
CO

CO
CO
o
§


CD
o
f»
in

CO
CO
o
in
cn

o
CO
o
0
in


0*
§
CO




8
CO

CO
co
8
*•



8
r^

CO
CO
o
o
CO

o
CO
d
o
CO
o
CM

e*
O
CO
"3-

O
O
CO


0
TC
CO
g
i
CM

to
8
v
*~
CO
CO
8
rr
CM
s
d
8
CO
CO

s*
in
CO
in




g
o>
CO
CO
CO
§
CO
O)
CD
g
T—
CO
CO
o
T
CO
CM
O
CO
o
0
TT
in
•*r

e«
g
co
CO




o
S
CO
CO
CO
8
CO
CM


8
CO
*~
CO
CO
§
CM
o
CO
0
8
rr

6s;
o
n
t~~

§


s
CO
CD
§
8
CO
in


§
CM
CO
CO
O
("-
CO
«•
O
CO
d
8
r^.
r*-

6*1
S
CO




o
8
r--
co
CO
o
•«•
CM
CD


0
5
co
CO
CO
o
CO
CD
in
0
CO
o
o
CO
o
cn

&?
o
co
0
CO




o
0
co
cS
0
r^.
r^.
•*
rft
CD
S
CM
CO
CO
O
*]•
CO
f
o
CO
0
o
5
CD

e*
O
CO
o

8
o


s
CO
^
CO
CO
§
CO
ni
CD
O
r^
CO
v
CO
CO
o
CO
o
CO
o
0
CO
Ol
CM

0*
in
CO





o
in
CM
CO
CO
CO
o
v
o


1
in
CO
fO
8
f
0)
o
ro
O
o


o
CO
d
o
in


6*
§
un




1

CO
CO
0
CO
V
»—


o
CO
r^

CO
CO
o
o
CO

0
CO
d
o
cc
o
CM

6«
O
CO
CD

o
o
CO


o
V
CO
CO
CO
§
CM


8
•<»•
~
CO
CO
o
CO
•c
CM
o
CO
o
8
CO
CO

c*
in
CO
f~




g
O)
CO
CO
CO
8
CO
CJ)
CD
8
r«
"
CO
W
o
^r
CO
CM
O
CO
o
«T
m
T

S«
o
co
CO




o
s
CO
CO
CO
s
CO
CM


8
in
*~
CO
CO
o
8
CM
O
CO
d
8
T

0*
o
CO
CT>

§


o
CM
CO
CO
CO
CO
s
CO
in
O)
CO
§
CM
CO
CO
o
N-
co
*T
0
CO
o
§
f~-
r^

^
in
CO
o




o
s
f«-
CO
CO
o
a
(D

CD
0
5
CO
CO
CO
o
CO
CD
in
0
CO
o
o
CD
o
01

e£
o
co





o
r^
o
CD
g
o
r^
s.
T


§
CM
cn
CO
o
•*r
CO
•<»•
o
CO
o
0
s
CO

^
o
CO
CM

o


1
»—
CO
CO
o
8
CO
O)
CD
o
r^.
CO
•«t
CO
CO
o
CO
o
CO
o
o
CO
0>
CM

6*
in
CO
co
uo




8
CM
CO
CO
CO
o
f
0


o
CO
CD
in
cn
CO
8
^
O)
o
CO
o
o
^>
m

i*
o
CO
*T




<

5
<
-z.
8


r^
0
h-

(V.
•




<
Z

<
§
CO


N.
s
in

r>»
v
O
s

s
d
8


s*
o
CO
8




s For FBC boilers, an annual emission rate of
5
£
Z
1
£
"3
'•5
g
?
c
•c
c
e

^s
I
0
ift
c
o
in
.£
V
.1
'S
V)
c
^
"e
3
c
e
c
"m
1
£
i
m
3
J»
2
S
z
£
00
V
d
"o
muni emission rate
t
c
c
<
S
j:
S
9
#
S
£
§
m
a'
#
m

'e
u
1
ti
£
£
ar
*
"8
•ft
3
•
C
|
'e
Ti
•>
£
a
c
J"~
^
5
|

1
?
i/MMBtu Is used, f
JJ

o>
o

ed Bed Combustion.
.M
1
^
II
O
ffi
U-
c
o
73
i
cc
.0
•5,
e
"Z
(J
c.
>
"u
tj
&
\t
cc
o
W
c*
.g
TJ
i
cc
.a
"*s
•I
«
0
"Z.
c
i
i!
•i
(0
II
cc
o
in
"o
Combustion Conti

11
o
u
s

o
.a
0
u
c
0
cc
%

0
i
0
^
3F
O
if)
tfi
U
U
^
%>
>
o
ec
U
CO
£
c
o

O
3
1
c
u
«>
o.
1
^
c
•5
in
0
a
0
it
i
'e
t>
£
nt reduction given
^
u
4)
CL
J




































Not Applicable

ii
<
Z
5
                                                B.l-2

-------
 CO


 tn
 o
CD
 0)

il
c
O)
E
o
c
o
'w
(/)

E
UJ
CO


m
a


— .
0
o
01
c
Emissic




re
re
D
Control
jstion
o
J
I
o>
tn
re
D







DC
O
t73
o
o
•~
rr
o
ID
o
o

C)
o

"re
c
c


3
C
£•
•n
re
D.
re
O
"o>
T3
O

Furnace

£
0
E
•£
o

£
"in
o

&
v>
in ">
II
UJ
3"
Oi CO
re P
cr 5
.2 "£
II
UJ
o
%
LL
Number
CD >
•— ^
01
a
a>
a.
£
"5J
U.
8
N.

CD
CO
s
in



8
CM


CO
§
in

o
o
o
rr
CD


O
CO
CD
O
O
~ra
a>
c
Jj
8
CO

s
o
s



8


CO
o
CO
o

o
0
o
.n


in
CD
CM



O
CO
in

CD
CO
8
CM


8
in


CO
8
CM

O
0
0
CO
CO


o
co
s



g
CM
CD
CO
8
co


CD
o
CO


CO
0
CO

o
o
o
in
CM

O
CO
s
o
o
CO


s
*»•
co
CO
o
CO
o
CO


v
'"

CO
o
in
CM
CO
o
o
o


in
CD
in
CD



o
CO

s
I
co

to
8
"T~"

CO
o
CO
CO
o
o
1
in

o
CO
8



0
CO

CD
CO
8
CO
CO


s
in
"

CO
o
f..
Tt-
CO
0
o
o
CO
o
in

o
CO
£
§


O
CO
CD
CO
o
(O

co
o
CM
CM

CO
8
CD
o
o
o
en

in
8



8
O)
to
CO
8


CD
0
CO

CO
o
in

o
o
o
a>
o

o
CO
en
CD



o

CD
CO
8
in
CO
8
8
CM

CO
1
in
o
0
o
CO
CO
CO

o
CO
0
1


o
CM
in
CO
CD
CO
1
o


o
r-
CO
*

CO
o
CM
CO
O
O
O
o
C)
CD
in

in
CD
-



o
in
8
O
^
CD
CD
I
in

CO
o
CM
O
o
8
CM
CO

o
CO
CM



§

8
O
in
in

CD
(8
8
CM


CO
O
CO
in

o
0
o
T3-
CO


o
CO
CO
8
Tangential

CO

CD
CO
o
S



8


CO
o
CO
o

o
o
0
in


in
ID
5



un

CD
CO
1

CD
IS
o
in


CO
8
CM

O
0
0
CO
CO


o
CO
in
tN.



o
CM
co
CO
§



8
f-


co
o
CO

o
o
o
in
CM

o
co
CD
CO


s

s
§
CO
CD
co
8
~

CO
o
in
CM
CO
0
o
o


in
CD
-



O

CO
CO
I
CO
CO
CO
8
"

CO
o
CO
CO
o
o
I
in

o
CO
CO



o
CO

s
8
CO
CD
IS
8
in
"

CO
o
Tf
CO
o
o
o
CO
o
in

o
CO
CD
1


O
CO
co
CO
g
CD


o
CM

CO
8
V
CD
o
0
o
O)

in
CD
O
CO



8
O)
CD
CO
8

to

o
CO

CO
o
in

o
0
o
o

0
CO
CO



s

CD
CO
8
LD

cS
I
CM

CO
o
CO
r--
in
o
0
o
CO
CO
CD

O
CO
CM
CO
O
8


o
CM
in
CO
CO
CO
o
CO
CM
O


o
CO
•t

CO
0
CM
CO
O
o
0
8
ID
in

in
CD
CO
CO



o
in
CD
CO
O
^

S
o
CO
CD
in

CO
o
5
CM
o
0
8
CM
CO

o
CO
co



z

z
8



o



o

o
o
o


o
CO
in
CO
in
CM
0
CD
u.

Z

Z
8



8



o

o
o
8
CM


in
CD
CD
CO



Z

Z
CM



o



8

°
0
CO


o
CO
CO



z

Z
8



8
CM

^

8
CM


CM
O
O
in
in


o
CO
CO
CO
8


Z

Z
S



8

^

0
in


CM
O
o
CO
0


m
CO
0)
CO



Z

Z
1



g
in

N

O
s

o
o
CM


O
CO
8



                                                                                             c 2
                                            B.l-3

-------
 o
m

"55
^
 o
O

"5
E
o
v>
c
_o
,
I/! •>
•ll
UJ
"5
£ 1
« 1
J3
a
" f
ll
UJ
j5
t>
o
u.
t
XI
Z
• ?
•* £
co 5
«i
a
>*
>-
t>
a
>.
>-
1
LL
g

£
o
CM
^
s
o
o
0
in

CO
o
o
in

#
o
CO
O>

O
o
TB
1 Natural Gas
o
K
co

g
o
n
M
o
m
0
o
g

03
O
O





o
eo
v

g
g
CM
O
m
o
0
0
«

CO
o
0
^

#
o
CO
co
0)




o
o





0
o

o
CO
m
s
o
in
o
0
0
o

CO
o
§

#
o
o
r^
O)

o
o
10


0
CM
CM
g
s
n
o
in
o
o
0
r^
N
CO
o
0
N
CM
#
to
(O
flO




s
m
CM
o
CO
o
£>_
o
m
o
0
0
CM
n
CO
o
0
r\j
CO
#
0
CO
01




o
O)
OJ

o
CO
o
s
o
in
o
o
0
CO
CM
CO
O
O
CO
CM
e*
O
CO
O
o

o
o
o


o
o>

o

0
o
Tangential

o
r^
CM

0
CO
o
n
CM
g
o
0
o
(C

CO
o
o
1C

#
in
1C
S




0
o
V

o
CO
o
r^
CM
O
in
o
o
o
T

CO
o
o
V

#
o
CO
in
o




8

o
CO
o
r-v
n
o
in
o
0
in
v

00
o
in
T

#
O
n
1C
o

o
0
o


0

g
o
S
o
in
0
o
o
CO

cc
o
o
cc

#
in
1C
K
o




0
s

o
CO
g
co
o
in
o
o
o
"
CO
o
o

§
CO
CO
0




0
en

S
in
S
o
m
0
0
0
o>
"'
CO
o
o
o>

#
o
ro
o>
o

o
0
1C


0
CM
CM"
g
o
CO
n
0
in
0
0
o
r-
CM
cc
o
o
r~
CM
#
in

Ol

o
CO
0
^
CM
o
m
o
o
0
CO
CM
CO
c
o
CO
CM
#
O
o
CM

O
o
o


o
O)

-------







M
C

o
^
5
1


UJ




e
O
"o
c
0
c
o
t'

t
c
u






















IT
E
O
(0
U
u

^.

E
0
2
M
•f
U
u



o
o
Annual








I 5
* C
Vi r=
£ <
.t:
u
o.
B
U
T
0
S

t>
u













*
&
~£
O



rf



1

*
missions
(ton/yr)
m

3"
* t
r? 5

—

E -
o i


LU
O
0
c

t
XI
E
Z
S ?
«7> 5

&
t-


«j
Q.
t-
S
U.
O
n




g

O
a




8



o


o
o
01
CO



CO
o



o
o>
CO




o
CO

in


8




5


6

o
in




o
CO

§
F)




O
in



o


0
o
CM
r-



co
o



O
CM



r*
ID

CO












0
f»
CO




o
CO

o
CM




8



o


o
o
ro
CO



CO
o



0
CO
CO



#
o
CO














o
CO
a




0
CO

s
in




o
m



o


o
o
(0



CO
o



o












8
CM



O
CO

o
in
CM
^



8



0


0
8
OJ



cc
C



O
O
OJ


#
0
CO

CM
CO











o
in
CO
*•



o
CO

o
1C
^



8



0


o
o
CM
CO
CM"



CO
r_



o
CM
ro
CM



o
CO

CO
CO

o
o
1C









o
CO



s

o
in
CM



8



0


o
o
CO



CD
c



o
CO
rr



in
CO

*r
o











8
•*•



o
CO

8
in
CM



o
in



o


o
o
8
in



cc
o



8
tn



0
CO


CO











o
g
CO



o
CO

o
CO
0>
^



o
in



o


o
0
1C
CO
co'



CC
o



g
CO
n


*
0
CO

1C
ro

8
o









o
in



o
CO

o
CO
in
n



o
in



o


o
0



CO
c



o
f~


*
in
(C


CO











o
CO
ID



O
CO

0
^



o
in



o


o
0
n
co'



CO
o



o
n
CO


*
o
CO

CO
CO











                             * ;  c E

                             I I  1 «
                             ^ «  I o
                             £ f  X -B
                            1
B.l-5

-------
                           APPENDIX B.2

              SECONDARY EMISSIONS FROM MODEL BOILERS
Table B.2-1         Secondary Emissions from Bituminous Coal-
                    Fired Model Boilers

Table B.2-2         Secondary Emissions from Subbituminous Coal-
                    Fired Model Boilers

Table B.2-3         Secondary Emissions from Lignite-Fired Model
                    Boilers

Table B.2-4         Secondary Emissions from Natural Gas-Fired
                    Model Boilers

Table B.2-5         Secondary Emissions from Oil-Fired Model
                    Boilers

-------
o
m
"i
•o
o
S
•o
£
cs
o
O
a
3
O
C

I


m
E
UJ
o
•o

o
u
o
CM
0)
c





g
CA
C
o
i Emission"
Secondan




Capacity






1L
8
4-
O
O

cr
t_>
+
8


0
u


Baseline

Factor
a>
•R

«
N
(0
Furnace

8
o
CM
n
Z
8
o
CM
1
8
o
CM
CO
Z
O
O
0
CNJ
Z
n
I
g
5
^>
F
3
g
e
D.
| Fuel Type
v
CM
O>

7,
8
co
s
0)
o
Si
O)
o
8


8
o
| Bituminous
in
»
CO
CO
o
CO
in
v
n
in
<»
CO
o
in
n
CO
o
in





r-
in
CM
1C
O
CO
8
8
o>
r-.
in
s
0
r»
in
CM
10
O
o
CO
ID




V
^>
r-
in
0)
v
CM
CO
I
8
f
V
r*
in
0
^-
v
^
m
o
o
o
r~

O
g


CM
§
CO
1
I-.
CO
I
O)

CM
0
o
CD
O>




O
si
3
in
i
8
CM
i
0
8
X
o
o
V
CM
*T
O)
0
o
co
0

8
o


01
V
V
g
f^
CM
O
o
O)
CM
in
3
01
^
V
ID
f^
O
O)
tr
•»
1C
r^
0
in
1C
_




Si
in
8
CM
CM
CO
CO
N.
r»
(0
8
Si
in
8
CM
O
a
in
(O
o
CM
0
g
CM




3
O)

tr
in
CO
CM
CO
Si
o>
o
Si
m
0
o
n
CO

8
Tangential

in
^
CD
n
o
co
in
v
n
in
v
CO
o
in
*
CO
o
in



CM
S
10
1
h.
n
I
en
1C
CM
$
o
o>
s
(C
o
0
s
o




v
CO
CO
CM
01
1
O
h.
n
8
*
CO
s
o
^•
CO
n
OJ
0
s
^




o
^1
CXI
s
in
§
in
S
CM
i
o
V
CM
S
o
o
V
CM
s
o
o
o
CM

!


1
g
(«.
CM
0
5
CM
to
»-
3
O)
TT
^
g
o
1
(D
N.
**•
O
S
n




s
in
1
CM
CO
R
r»
to
S
Si
in
S
CM
o
S
(O
o
CM
0
g
»




<
Z
g
<
m
^
CM
U)
U)
(D
S.
»
O
ID
r^
•»
0
g
in

CM
2
O
ffl
LL

<
Z
<
in
CM
h-
o>
O)
»-
S
o
^
CO
CO
0
IS
CO




<
Z
<
CM
CO
o
CO
8
0
CO
CO
o
0
g
^




<
Z
<
»
in
rv
o
CM
CO
3
CO
co
o
V
CM
. r) < "- •
                                                           "5 ,- o>' . . ii '
                                                           2 S  • > ' ,,

                                                           |3£8?S:
                                                           — "• <° Z. 2 „ !
                                   B.2-1

-------
o
m


•o
o
o
U
o
c

I
.c

(/)


!
09
E
LU


a
•o
c
o
u
a
en
CM

CO
.0
a












n
5
I
.2
W)

r-
LJJ
t
10
8
&













2-
u
n
a
in
O












CC
o
05
"*"
0
u




CC
•f
o
u






o
u







c
"n>
m
m



0
o
n
u

-
•D

S
CO
a
|
LL





8
8
7


CO
T
Z
8

S


c,
Z
n
o


o



CO
X
Z
n
o

eg
Z


I
Z
SP


1

z
i
s
i-
8.
£*
1
u.

M
CD



-


S

s


D

5



0>


o



5

Ol



o


o





8
_
1
c
F

.O
co
in
w
00



co


o

8


S

in



00


o



in

CO



o


in
ID



CM





a
u
U


CM
in
CM



CO


oo

S


o

CM
in



CM


o



CM
in

eg



O


§



CO









CM
8



»


s

IS


18

CM



CO
CM


O



CM

CO
eg



o


o
CO



n

8
CO







in
CO
s



CO


i

s


i

s



CO
in


o



in
CO

CO
in



O


in
10



CO









in
(M
ID



o


8

S


o>

in



CM
ID


o



in

eg
ID



O


o
CO



CO









5
h.
in



Ol


S
eo

o


S

?



in


o



?

in



o


o
r>





i







en
ID
CM
8



CD


1

CO


3
CM

Ol
ID
CM



8


o



1

8
*"


o


in
ID



00









CO
s



Ol


1

o
CO


oo
CO
CM

CO



CO


o



?

0>J
*~


o


§



en









o
CM
S



in


o

S
CM


i

o
CM



S


o



o

en



o


o
eo



O

I







01
s



CM


O
5

s
in


S

Ol



ID


0



Ol


CM

n


o

co



8


o



eo

CO
o
*~


0


o
CD













Z
i



<


s

s


00

CM



CO
00


o



CM

CO
CO
•~


o


o
CO



OD

8







z
z



z


o

oo
00
CO


CO

in



S


o



in

in
CO


0


in
ID



en









Z
Z



z


oo

CM
in


§

CM
in



£


o



CM
in

-
^


o


o
CO



o









S ^



E o
5 n

^ E
^ £
n o
•D o





It
01 a
w O

£ -D
U CO
5 75


tl


II
                                                             a f

                                                             - "2
                                                             c S
                                                             O 3
     €
     o
     CC
      a

      To
      U

      o>
                                                                   o>
                                                                   co
  CC
  u
  CO

  I
  o

  1!

  II

  11
  -2 «
  z u
                                                                   o O
                                                                   a o
5 S

I s
a 7c
5 2
B ^
U Ol

s I

•i I
| o


!l
• _•
i I
CD «  x
O CC  °

§ Z  I
Of g  |

rf o'  Z
O £  •

" 9  °
i O  CM
5 c  2
^ £  «
a §  c
m 5  2



IM
# O  £2
S2 o  i

  e.
o
tJ _o

  ^3
  C
                                                                      CD
                                                                      5 &
                                                                      If

                                                                      •!
                                                                      «^
                                                                      £z
                                    B.2-2

-------
I


"o
o

'•
•o
o


•o
c
e
O)
tu
8
«
V)
CM

CO





*n"
v>
|
r Emission:
Secondary




Capacity







1C
O

Z
£
1
3
1
a
a.
>*
t-
1
a
3
u.
^

o

S
Ol
CM
CO
5
o
0
s
o
o
o
o


8
'S
9
'c
Ol
3
m

CO
o
i
s
3
in
V
co
0
in
v
CO
o
in
10
CM



*IO
o
O
co

CM
CO
a>
8
o
»
CO
in
CM
0
CO
in
CM
O
§
8




P

a>
CM
v
s
*-
s
in
U)
to
K
o>
CM
o
co
f^
01
CM
O
o
m
S

g
CO



0
CO
m
CO
I
i
8
10
(0
S3
0
ID
CO
(*)
in
o
S
in




CO

S
0
.
CO
v
CM
01
CO
£
R
co
1
r*.
D
5
O
N.
O
^
CNJ
O
s
o>




CM

3
m
i
g
CM
S
^
CM
IT
CM
in
o>
0
CM
*r
CM
in
OJ
o
o
CO
o

!


CO
»
g
^
CM
a>
o
co
CO
in
CO
^
CO
CO
in
v
CO
r^
o
CO
in
v
CO
rs.
O
in
(0
^




CO
in
r«*
o
CM
CM
CO
§
CM
3
8
(O
CM
in
r^
o
ca
o
CO
CM
in
r-
O
CM
O
8
CM




«

O

s
o>
CM
CO
S
o
o
V
CM
0
o
o
CO
CO

8
Tangential

in

00
n
i
co
in
s
m
v
00
o
in
»
CO
o
8
v




to

CM
CO
O>
T-
OJ
»•
s

CM
O
o
co
(O

o
o
CO



in
o
in
ID
r-




CO

CM
ID
O
3
r-
co
8
CD
in
S
o
CO
in
CM
(O
O
o
CO
p




in

f«
in
0)
N
g
£
o
in
«T
^
in
o
in
v
r^
in
o
o
o
01

1


CM
CM
S
CO
^
CO
8
CO
8
CM
R
CM
f.
O
0
CM
r»
CM
N
o
o
£
o




h*
CO
T
CM
01
CO
K
R
co
o
v
CM
r^
CO
£
o
f-
co
f
rv
o
o
CO
^




CM
CM
S
in
5
§
§
CM
^
CM
S
O
CM
V
CM
in
o>
o
8
S

§


CO
»
t
r^
CM
o>
o
CO
S
5
CO

^
s
o
*-
0>
o
o
in
(O
to




<

<
<
o
CO
V
o
CO
3
o
CO
5
o
o
CO
fv




<

Z
Z
S
i
CD
S
§
O
^
o
a>
o
o
CO
CO

8


<

<
<
s
o>
CO
^
CO
in
v
(O
in
CO
0
in
»
ID
in
CO
o
S
Ol




<

<
<
Ol
(D
in
v
0
V
2
in
v
o
CO
in
in
v
O
8
g




following F-factor 9,860 (dscf/MMBtu) for lignite
o
£
C
O
•D
0
VI
10
.D
al emissions calclated
3
C
C
'«T
1
£
^
o
CO
en
£
t>
JO
5-
0
cl

s
m

o
at
9,500 Btu/kWh: 80% -
»?
£

ctlva Noncatafytfc Reduction. SCR = Selective Catalytic Redu<
CP
&
i
5
z
CO
' Combustion Control,
i
8
a
O = Carbon monoxide.
o
o>*
TJ
X
o
VI
2
o
i
- Ammonia, N2O - Nl
H
CO
Z




c
JO
Tfl
3
- Fluldlzed Bed Comb
U
CD
U.
s






Not Applicable
«
<
g
                                           B.2-3

-------
o
ID
o

I


z

E
ui
CO



CJ
o





^
o
c.
Emissloni
Second an




Capacity






(C
8



8


8


Baseline

Factor
•8
2
1
Furnace

8
8

I
o
u
I
1
8
0
CM
|
0
o
o
04
z
CO
e
1
2
2
1
| Fuel Type
in
B


n
S
m
in
B
o
in
00
o
0
n


8
~e
| Natural
at
in


fe
in
8
o»
in
o
o>
in
o
8
CM



5
-
••-


«
S
8
~
£
o
^
—
o
8
CO




in
CM


8!
K
B
in
CM
o
in
CM
o
0
CO
s

8
to


8
U)


S
5
s
8
in
o
B
CM
in
o
in

in
o

s




-
^


S
8
3
^
K
O
-
-
0
o
B
S




in
S


8!
R
B
in
S
o
in
CM
O
g
8

8
CO


8
*


s
5
S
8
in
•r
O
8
in
o
S
K




CO
CO
K


s
s
8
S
S
o
"
CM
in
o
8
S




n
B


2
*
8
n
B
0
CO
CO
o
s
8

8
1C


in
S


s
§
R
in
S
o
K
in
Ol
CO
o
•
o




s
1


i
CO
8
CM
S
3
0
S
S
o
8
*-




in
8


i
o
in
in
in
8
0
in
o
CO
o
o
CO
CM

!


18
O)

CM
1
B
B
CM
$
0
O
in
Ol
Ol
o
•
CO




0
B

CM
S
o>
U)
8
o
B
o
o
CO
o
8
,




a =  «



IS  1
co  o

O*7  "S

2 S  ^
CD CO
    5Q_
 M  f v
                                    O    •




                                    1  f i

                                      5 i
                                   I*  I?


                                   !H15
                                   sini
                                   ^ s c 
-------
o
D



1
O

S
o


2

S
o
•o
up

CM

m















o
o
V)
•>
F
LU
condar
o>
10














apaclty
O













u
u



or
u

o
u






o
u








0;
c
a>
K
ffi




£
u
B

1

s
(0
•
03
C
3
LL




b
S
z


c,
Z
8

8
7


1

O



O
CM



CO
T
Z

O
0
CM
Z


r>
I
Z
*

_o
F
z
?
*

1
1

a
U.
in

CO



-

CM
O

in
CM



«o
m




00


0



in

00



o



0
CO

in

R


~m

O

0

CD



CM

8

S



o
n
o





10




CM


O



(0

in
CM



O



o
CO

00

3
CO





s

5



f-

8

;



9
0




*


0



8

f-



o



s

0>








in
CO

S



«

1

s



s
in




S


o



8

in
in



0



8

o
CM








S

in



CO

3

3



S
CM




in


o



8

in



o



o

CM

3






0

ct



in

S
CO

g



s
0




5


o



8

S



o



s

CM
CM








S

s



t

00

8



CM
g




8


o



S

O)
o



o



8

CO
CM








5

S



CO

CO
CM
CO

2



s
m




8


0



S

in
00



O



o
CO

CM

g
*-"





8

in



CM

1

$



CO
8




in


o



8

in



O



8

in
CM








1C

CM
00



OD
CM

O>
ID

S



CM
in
CO
10
^



i


o



ID

CM
00



O



8

(0
CM








in

00



-

CM
CO

a



-



O

•o



CM

8

*



o
o





o


o



g

g



o



o
00

in
CO








S

in
00



CO

r>
CM
CO

CM



s
s




s


o



in

in
oo



o



o
CO

ID
CO

g
'-"





8

in



CM

§

£



i
8
»—



in


o



8

in



O



18

CO









fi o   n
CD CO


§ E   5
t. o   w

"So   c-


«*   I
                                             2 -c  f. o
                                             o «  5- P
                                                 S o
                                              g  s -e

                                              5  IS
                                             •e i i t; o
                                             j * g « u


                                             1 8 I " -S
                                             O 3 £ t "x
                                             V • „ /T- O
                                             « c o r '
                                             5 1 • S o
                                             « „ i" O CM

                                             !?U"-
                                             ||l|]


                                             1*1^
                                             i I" ' A

                                             IT    S
                              B.2-5

-------
                           APPENDIX B.3

                 ENERGY IMPACTS ON MODEL BOILERS
Table B.3-1


Table B.3-2


Table B.3-3


Table B.3-4


Table B. 3-5
Energy Impacts of NOX Control Options on
Bituminous Coal-Fired Boilers

Energy Impacts of NOX Control Options on
Subbituminous Coal-Fired Boilers

Energy Impacts of NOX Control Options on Lignite-
Fired Boilers

Energy Impacts of NOX Control Options on Natural
Gas-Fired Boilers

Energy Impacts of NOX Control Options on Oil-Fired
Boilers

-------
*
•§
n
o
O
in

O



I
CD
in
O
'•s.
O
o
O
(C
CL
LLJ



CO

CO
e



^"*
••oH
&
E
«

S
(T
>,
|
111
re
.2
re
•o
5

























of =•
\l\
fill
09 f
+ J
81


o = ^
(0 -° o
* "o 'S
fi S
w ^
4. >
o 1
o *
if
II


- =.
"1
S ^5
£• *
«J u.
O
>-
•D -^
• >
OT 1
8 •>
« Q.
E >>
a H-

•
ii.^
r~
1


0

i
yj



8
0


o
to




o
o
d


1



r-

0





IA
3
O
C
F
i

0

CO
CM



8
0


o
*-



0
o
d


'O



CM











I

O

§
00
CM



8
0


I
*-



o
o
d


30



CO













o

s
CO



8
o


o
S
CM



O
O
o


i
CO



•V

1











0

i
r*»



8
o


i
•*



0
o
o


'O



in













o

§
00



8
o


1
in



o
o
o


30



to













o

o
IA

*~


8
0


s
CO



0
o
o


(0



00













o

?:
fs»
*~


8
o


s
o



o
o
o


30



o>













o

1
o
*"


8
o


g
(O



0
o
0


1



o

1











o

8
CO
CM


8
0


§
in
J



0
o
o


ID



*-













O

§
eo
CM


8
0


§
£



0
o
o


30



CM













o

i
^



in
CM
o


o
r*.
(O




o
o
o


CO



CO

1

1
4)
J*
(2






o

8
CM



8
o


1
'-



0
o
o


'O



•»













o

8
00
CM



8
0


1
*-



0
o
o


1



in













o

§
CO



8
o


8
o
CM



0
o
o


CO



-



8
o


§
CO
*



0
o
o


!£>



f^













O

i
oo



8
o


1
tn



o
o
d


00



00













o

o
s
(D



8
o


1
*



o
o
d


CO



o>

§
*.
*""


8
o


I
o



o
o
d


00

















o

o
s
o
'F"


8
o


h-
(0



o
o
d


1



PS

I











o

1
CO
CM


8
O


i
;



0
o
d


•£>



CO













O

§
OO

-------
CD
O


"8
CO
o
O

in

O


E
3

lo
.a
3
CO
«a

O
'•s.
O

e

o
o
co
o.
E
>,


1
LU


s
CD
•


CD

E
O
s
0
HI
CO
o
•o
TJ

























f{|
OM t
CO £
B!
or ^
z o §
?"5 H
8*°
g?
z •£
+ ^
o i
o *
si
O 2
6|
^ 2»
6 5
c ^
f o
• S.
- $
"S |
I!
8 9
« 5.
E >»

^i


1
»-
^6
ijj



o
o
s
*•

Pi
0


0
in




o

o

o
1

CO

2


^
m

M
o
E
£
A
•^
to

0
S
CN

Pi
0


O
CN




O

O
ro
O
m
(O

S












o
O

o
i
CN

Pi
O


in
T-



o

0
n
o
30

CO
CO















0
o
CO

Pi
0


o
T-



o

o

o
1

a

n













O
s
N.

Pi
O


o
N.
CO



o

o
CO
o
n
(O

8















o
o
g
CO

Pi
o


o
m
^r



o

o
ro
o
ao

8















O
0
in
co

Pi
o


o
CN
CO



o

o
ro
o
s

CO

§













o
I
?

Pi
o


V




0

o
ro
O
10

3















0
CO
£

Pi
o


T-
o>



0

o
ro
o
s

o>















o
1
o

CN
CN
0


o
in



0

o
ro
o
CO

o

8
o
T"













0
i
CO
CN

CN
CN
O


0
h-
co
CN



O

o
ro
O
1

5















O
g
co
CN

Pi
0


o
in



0

0
ro
o
00

CN















O
i
*~

CN
CN
0


O
r—
in




0

o
ro
O
1

5

8
"3
f
0)
O)
c
n








0
CO
CN

Pi
o


o
1-



0

o
ro
o
(O

5















O
i
CN

Pi
0


O
T-



0

o
ro
o
1

$















0
s
CO

Pi
o


o
T-



o

0
ro
O
1

$

8




0

o
ro
o
CO

5















o
o
0

Pi
o


0
in



o

o
ro
o
1

CN

T-













o
§
CO
CN

Pi
0


CO
CN



0

o
ro
O
in
CO

CO
in














T"
O
O
3

Pi
o


§
in



o

o
CO
o
i

s














«t
z
_o
<
Z

co
o


o
CN




0

o
CN
O
1
CO

in

m
CN


.
w
J.









Z


co
o


S




o

o
CN
O
(O

s
















z


CO
o


0
CO




o

o
CN
O
1

in














1f
i
z


co
o


0
*




o

o
CN
O
i

s

8












«r
Z
z


CO
O


o
0




o


CN
O
in
CO

S















Z
i


co
o


g
CN




O

o
CN
O
CO

8














                                                               o
                                                               u
                                                               ta.





                                                               fi



                                                               a>'

                                                             « a>"
lu. «
•£i
§£ S
^lf
aSf
£ g *
    s
y s I
«• i
(0 Ji-|
£ • |

2 "• o>
* 1 1




§1|
^ S « i

fill
  _

U 0
o o
                              B.3-2

-------
"5
CD
"8
 O)
 VI

 o
 o
o
 o
 M
•c
 09
 Q.
ui

CO

^
CD




t~*
CO
••^
42
§

a
'5
SF
cc
>.
*>
LU
TO
O
*:
•o
5























a: *
O =5 C-
w .2 !
+ "5 -5
B*°
g£

O if
o e


o 3= .-,
2 g -g
W f o
0 # °
O —
o?
w %
4. >
<•> 2
O =•
si
0^
0 1
5|

o 5
c ^
t*
CD T5
«" a
o ""
«l
1 i
I!
2
^1
3 r-

1

U.

o


o
*•



8
o


S
-*.

1








o


8
04



04
O


§
5




o


o
o

m
(0

£










o


1
8



8
o


§
£




o


o
o

0
00

P!










o


i
*"



in
O4
o


o
(0





o


o
o

#
CO

CO
r—

°
-
|
c*
,5




o


§
04



04
0


1
*~




0


o
o

*
(O

5










o


i
04



04
O


8
oo
«-




o


o
o

o
00

in










o


s
«o



04
O


1
O4




O


o
o

#
CO

s

8








0


i
^



04
0


§
«
•»




o


o
o

in
(O

fc










o


I
00



04
0


CO
in




o


o
o

o
00

s










o


o
in
to



04
0


o
s
*




o


o
o

o
CO

en

*








o


§
?



04
0


s
00




0


o
0

in
(O

§










o


g
t



04
0


8
0




o


0
o

#
o
ao

00










o


i
o



04
O


g
(O




o


o
o

o
CO

00

V-








o


i
8



04
0


in
J




0


o
0

#
in
(O

00










o


§
04



04
O


i
^




o


o
o

1
00

s













<
2



oo
o


o
O4





O


o
fN
O

0
«

in
00

in
01


O
X







i




00
o


S





0


o
0

in
(O

S













i




ao
o


o
CO





o


o
(N
O

#
00

f-
00













i




00
o


o
r-





a


o
(N
O

0
«

g

8








-------
12
£.
"5
CO
"55
o
cc
O

1
1


o
in
o
1.
O

S

o
O
x
o
v>
0)
CO

CD
05




s.
£
0)


It
E

UJ
CD
0
ss
<£


























OL 9
"If
8e*
tt£
0) fc
* ^
Q £
o e
O s *-+
z g «

O >
if
u 5
tl
9- «
CB li
0
w
"1 1
S?
w^
It
§ H


1
1-
0)
U.


$
o

o
5




0




o
m





o


oo
o
o
co



5

0
o


—
CO
5
M
CO
O
e
3
n
Z
m
CO
0

o
s
^_



0




a>





o


co
O
o



s













co
0

1
CM



O




o
o>
«-




o


oo
o
3
00



0>













CO
0

o
CM



O




O
3
*-




O


eo
o
o
CO



s

o
s











m
CO
o

s
e»
in



o




0
O)
CM




O


00
o
in
CO



in













in
co
o

1
r-



o




§
m
co




o


eo
O
o
00



s













in
co
o

s
in



0




i
CD
CM




O


00
o
1



3»

O











co
o

i
*_
*~


o




co
m




o


oo
O
in
CD



oo
o>













m
co
o

o
s
^
*~


o




o
CO
»-




o


00
o
g
00



O)
o>













co
o

o
m
o>



o




o
oo
•*




o


oo
o
o
co



8

1











m
CO
o

§
at
*~


o




1
o>




o


00
o
in
CO



o













co
o

g
^J-
CM


O




o
s
«-




o


co
o
o
00



s













CO
o

o
s




o




s





o


00
o
o
co



co
o

8
T-

i
CB
c
03






3
o

i
T-



o




o





o


00
o
in
(D



s













CO
o

o
CM



o




i
*-




0


CO
o
go



in
0













co
o

§
f-
CM



O




o
«-




o


CO
o
o
CO



8

o











in
CO
o

Oft
in



o




o
o>
CM




O


00
o
m
CO



o













Jo*
o

o
CM
CO
r-



0




o
S
co




o


eo
o
o
00



co
o













m
co
o

§
in



o




§
CO
CM




O


eo
o
o
CO



g

o
s











in
co
o

o
o>
oo
T-
*~


o




o
CM
00
in




o


eo
o
fl
CO



o













in
CO
o

1
^
*~


o




i
»*•




o


eo
o
o
00



*-













in
co
o

o
in
o>



o




g
•*




o


co
o
1



CM

8
o











in
co
o

§
o>



0




1
o>




o


00
o
m
CO



CO













in
CO
o

|
^
CM


O




O





0


oo
o
o
00



•*













                                                                                     i?
                                                                                     o
                                                                                     H
                                                                                     U.
                                                                                     O.
                                        B.3-4

-------
'o
m
 Q)

|




I
il
 U)
_o
•s.
O
 o
O
 o
 V)
•G
 CO
 Q.
9)
in

"
m
0)
s
CO




15"

(A
1

S
3
6
>,
E

c
HI
CO
.2
«




















K 0
CO 1 ?
+>**
# °
0?
CO jj
- 1
o *
o e

£ fc
zf^
to £ 5.
+ O 3
0 £ °
o e.
a: .p,
Z t:
CO *r
8^
£?

u i
il
o 5
tfe
g."5
CO Q.
O
^
i|
« s~
N S»
55 S
8 «
u.

1
if

in
co
o


o
a>





CM
d




S



0

oo
CM
O
0


m

3



'oo
§


3
m
co
o


g
as
,_





0




g
*"


o

oo
CM
O
n
(D


(O










in
CO
o


o
CM




CM
0




*•
*~


o

00
CM
O
0
x>


r-










m
CO
o


g
CM




CM
0




S
CO
*~


0

CO
CM
O
0
co


00

ro








in
CO
o


o
m





o




g
in
co


o

oo
CM
O
in
(O


S










in
co
o


§
f»





o




§
*


o

oo
CM
O
b
co


o
CM










in
co
0


1
m




CM
o




§
CO
CO


0

00
CM
0
to


CM

§








m
co
o


i
v-
^™




0




o
m
^


0

oo
CM
O
(0


a










in
CO
o


o
s.
^
V"



CM
O




§
oo
00


o

oo
CM
O
b
CO


co
CM










m
CO
o


8
o>





0




o
s
in


0

00
CM
0
co


CM

O
*~







m
CO
o


§
O)
^"



CM
0




o
SI
T-


o

CO
CM
O
in
«o


in
CM










3
O


|
^
CM



CM
O




o
(O
-


o

oo
CM
O
g


S










3
o


o






o




o
m
m



o

00
CM
O
O
co


fv.
CM

§


i
§>
*



m
co
o


o
s
T-




CM
O




o
o>
*~


o

00
CM
O
(D


CM










in
co
o


o
CM




CM
O




O
*•
*~


o

00
CM
O
b
CO


CM










in
co
O


g
CM





0




s
(0
*~


o

CO
CM
O
O
co


o
CO

o








co
o


at
in"





o




g
m
co


O

00
CM
O
(O


co










So
o


§
r-





o




g
"*


o

co
CM
O
o
00


co










in
o


§
m





0




g
CO
CO


o

00
CM
O
o
co


co
co

o
g








in
CO
o


o
OJ
T-
^"




o




s
^


o

00
CM
O
(O


s










in
CO
o


o
^
^"



CM
O




g
00
00


o

00
CM
O
o
eo


in
CO










in
co
o


o
in
o>




CM
O




g
m
in


o

CO
CM
O
O
co


CO

1
T™







in
CO
o


§
o>
^™



CM
0




i
V-


0

00
CM
O
ID


co










in
co
o


o
o
rt
CM




O




o
(O
?


o

00
CM
O
b
oo


CO
CO










                                                                                     u
                                                                                     u.
                                                                                     OL
                                                                                     u
                                                B.3-5

-------
                     APPENDIX C

                 COSTING PROCEDURES
C.I  Methodology
C.2  Low NOX Burners (LNB) Applied to Coal-Fired Wall
     Boilers
C.3  LNB's with Close-Coupled Overfire Air Applied to
     Coal-Fired Tangential Boilers
C.4  Selective Noncatalytic Reduction (SNCR)
C.5  Selective Catalytic Reduction (SCR)
C.6  Combination Controls - LNB + SNCR and LNB + SCR
C.7  References

-------
C.I  Methodology
     The basic methodology used to determine NOX control
option costs, cost effectiveness, and incremental cost
effectiveness (ICE) is presented in this section.  This
section also includes an example calculation applying the
presented methodology.
     The data used in this appendix to determine the NOX
control option costs were obtained from retrofit projects.  To
obtain the cost associated with installing the control option
on a new boiler, the additional cost associated with
installing the control option on an existing boiler has been
subtracted from the total capital cost.
     Section C.2 through C.6 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 operating and maintenance
(O&M)  costs.
C.I.I     Direct Costs
     The equation to calculate direct cost (DC) is:

                      DC  ($/kW; = a x MWb                 (C.I}
where:
          a = Constant derived from regression analysis
         MW = Boiler size  (MW)
          b = Constant derived from regression analysis

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 MW boiler installing a given control option,
where "a" and "b" were determined to be 220 and -0.44,
respectively, the calculation is:
                              C-l

-------
                   DC ($/kW)  = 220 * 100~°'44
                             = $29/kW
C.I.2     Indirect Cost Factors
The equation to calculate an indirect cost  (ICF) factor is:
                       ICF = 1 * [IC/DC ]                  (C.2)

where:
                       DC = Direct cost
                       1C  = Indirect cost

For the same 100 MW boiler with a direct cost of $29/kW, and
indirect costs of $9/kW, the indirect cost factor is:
                   ICF = 1 + ($9/kW) / ($29/kW)
                       = 1 + 0.26
                       = 1.26

C.I.3     Total Capital Costs
     The equation to calculate total  capital cost (TCC) is:

                      TCC  i$/kW) = DC - ICF                (C.3)


Other variables are as previously defined.
     The exception to this is SCR systems.  As discussed in
C.I.I,  the initial catalyst cose is a direct cost, however, it
is not multiplied by the ICF in determining total capital
costs.
                              C-2

-------
     For the example 100 MW boiler with a direct cost of
$29/kW and an indirect cost factor of 1.3, the total capital
cost is:
                    TCC  ($/kW) = $29/kW * 1.3
                              = $38/kW

C.l. 4     Operating 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:
                      FO&M ($/yr) = c * MV^                (C.4)
where:
          c = Constant derived from regression analysis
          d = Constant derived from regression analysis

other variables are as previously defined, or
                       FO&M = e - (f * MW)                 (C.5)
where :
          e = Constant derived from regression analysis
          f = Constant derived from regression analysis
Other variables are as previously defined.
     Variable O&M  (VO&M) costs are expressed in $/yr, except
for SCR VO&M costs which are in $/kW-yr.  These cost equations
are specific for each technology and are presented in
individual sections of this appendix.
                              C-3

-------
C.I.5     Busbar Costs

     The equation for calculating busbar costs  (BC)  is:
       BC (mills/kWh) .  (ACC + FO&M + VO&M) * 1000 mills/$  (c>fi)
                                      AEO


where:

             AEO = Annual Electrical Output (kWh/yr)
             ACC = Annualized Capital Costs ($/yr)
            VO&M = Expressed in $/yr


Other variables are as previously defined.

     In the above equation, ACC is calculated from:

                ACC = TCC * MW *  CRF * 1000 kW/MW           (C.7)


where:

                  CRF = Capital Recovery Factor


Other variables are as previously defined.

     In equation C.7, CRF is calculated from:


                 CRF = i(l-i)n/[d-i)n-l]             (C.8)


where:

           i = Interest rate (decimal fraction)
           n = Economic life of the equipment (years)


Assuming an interest rate of 0.07 and a economic life  of

20 years:


           CRF = 0.07 (1 + 0.07)20 / [ (1 * 0.07)2C - 1]
                = 0.27/2.86
                = 0.094


For the example 100 MW boiler with a total capital cost of
                              C-4

-------
$38/kW and a capital recovery factor of 0.094, annualized
capital costs would be:
        ACC  ($/yr) = $38/kW * 100 MW * 0.094 * 1000 kW/MW
                   = $357,200/yr

     In equation C.6, AEO is calculated from:

             AEO  (kWh/yr) = MW * CF * 8,760,000 kWh        (C.9)
                                             MW-yr
Variables are as previously defined.
For the example 100 MW boiler with a capacity factor of 0.65,
the annual electric output is:
         AEO  (kWh/yr) = 100 MW * 0.65 *8,760,000
                                                MW-yr
                      = 569,400, 000 Mil
                                    yr

     For the example 100 MW boiler with annualized capital
costs of $357,200 per year, negligible O&M costs, and an
annual electrical output of 569,400,000 kWh/yr, the busbar
cost is:
   Busbar Cost j       = (($357,200/yr-0)  *1000mills/$) /
               \  kWh  /
                        (569,400, 000 kWh/yr)
                      =0.63 mills/kWh
                              C-5

-------
C.I. 6     Cost Effectiveness
     The equation for calculating cost effectiveness is:
          CE ,$/tonNOx removed) .  'ACC                  ,c.l<»
where:

            Tons NOX = Tons NOX removed per year(tpy)
               VO&M = Expressed in $/yr

Other variables are as previously defined.
     In equation C.10, tons NOX is calculated from:

              Tons NOx = UncNOx * NOX Reduction * KR
                                                        (C.ll)
                        * MW  * CF * 0 .00438


where:

        UncNOx = Uncontrolled NOX emission rate (Ib/MMBtu)
 NOX Reduction = NOX control performance (decimal  fraction)
           HR = Boiler net heat rate (Btu/kWh)
           CF = Average annual capacity factor (decimal fraction)
      0.00438 = factor to convert Ib NOv/kWh to tens NO /MW-yr
                                      J\-               J\.

Other variables are as previously defined.
     For the example 100 MW boiler with a baseline NOX level
of 0.9  Ib/MMBtu,  a NOX reduction of 0.40, a  net heat rate of
12,500  Btu/kWh,  and a capacity factor of 0.65, the tons of NOX
removed per year are:

    Tons NO  (tpy) =0.90 Ib/MMBtu * 0.40  * 12,500 Btu/kWh *
                    100 MW *  0.65 * 0.00438
                  = 1,281 tons N0x/yr

     For the example 100 MW boiler with annualized capital
costs of $357,200 per year, negligible O&M costs, and 1,281
tons of NOX removed per year, the cost effectiveness is:

      CE  ($/ton NO  removed) = ($357,200/yr  +  0) /1,281 tpy
                            = $278/ton NOX removed
                              C-6

-------
C.I.7     Incremental Cost Effectiveness
     Incremental cost effectiveness  (ICE)  is  calculated  by
determining the incremental change in both total annual  costs
 (TAC) 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:
             (Total TAC for COp #2 - Total TAC  for COp #1)      (C-
               (Tons NOX for COp #2 - Tons NOX for COp #1)   12)

where:
            Total TAC = ACC - FO&M •*- VO&M ($/yr)
               COp #1 = less stringent control  option
               COp #2 = more stringent control  option

Other variables are as previously defined.
     A set of hypothetical values for costs and tons  of  NOX
removed are presented for use in an  example calculation:
     250,000   =    Total AC for COp #1
     450,COC   =    Total AC for COp #2
         4CO   =    Tons NOX for CCp #1
       1,COO   =    Tons NOX for COp #2
Using equation C.12, ICE is calculated to  be:

     ICE ($'tonNOx removed) =  ($400,000/yr -  $250,000/yr)
                  x                (1000 tpy - 400 tpy)
                            = 250 $ 'ton NO,, renoved
                              C-7

-------
C.2  LNB'S APPLIED TO COAL-FIRED WALL BOILERS
C.2.1     Data Summary
     The data used to develop cost equations for applying LNB
to wall-fired boilers are shown in table C-l.  Presented in
the table are utility and plant name or code, boiler size,
direct costs, indirect costs, total capital cost, and fixed
and variable O&M costs.  Fixed O&M costs were available for
only one boiler, and variable O&M costs were not available for
any boilers.
     The data for three of the boilers were obtained from
questionnaire responses and are actual installation costs.1-23
The data for the other seven boilers were obtained from the
EPA's "Analysis of Low NOX Burner  Technology Costs"  report  and
represent cost estimates for installing LNB's,  rather than
actual installations.4
C.2.2     Direct Cost
     As discussed in section 6.2 of chapter 6.0, there are no
incremental capital costs attributable to this control option,
and O&M costs were assumed to be 3 percent of direct costs.
Cost equations for direct cost are presented only to provide
the basis for O&M costs.
     Based on linear regression analysis of the natural
logarithms of direct cost ($/kW) and boiler size (MW) data,
the cost coefficients for equation C.I were calculated to be
a = 220 and b = -0.44.  Therefore, the direct cost algorithm
for LNB's is:

                   DC  ($/kW) = 220 * MW"0-44
Figure B-l presents the plot of the data and the curve
representing the above equation.
C.2.3     Indirect Cost
     Indirect cost factors were not required since there were
no direct costs attributable to this control option.
                              C-8

-------
i
3g
E?
-8
I
u

u
J
m
<
EH
11
g
OP
«»-
01
or
ll
g|
u. ^
a «/>
u ^
«rf 4^
IB «!
0 ^,
*-2
o _
U 3
o
1
N
+•*
i
at
4^
c
(0
o.
X
4-


1
1
(M
O
«M
*~
0-
rj
o
0
o

K,
1
U
II
|| Consumer's Power


:

•
o
K>
PJ

Kl
PO






•



Kl
„
-0
O
•o
fVJ
o
Kl


GO



i


to
o-
rg
in
in
N.
KI
fNJ
o
in


u



i
i



Kl

to
s
o


o



:



s
to
in
Kl
in
(M
in


IU



'•


to
to
r\i
in
in
Kl
Kl
in
in


u.



i

i
o>
•o
r\j
o-

o
o
(M
0
O


o



I


-o
Kl
CM
„

fNJ
O*
g


-



1



CM
Kl

5
o
0


-

                           D.

                           *J
                           O
                           C-9

-------
                                    s
                                    I
                                       s
                                                  E
O
cn
iH
(0

JJ
en
O
u


01
jj
in
                                                  i
                                                 U

                                                  0)
C-10

-------
C.2.4     O&M Costs
     As discussed in section 6.2 of chapter 6.0, O&M costs
were assumed to be 3 percent of the direct cost.  Converting
the units of the direct cost equation from $/kW to $/yr, the
equation for O&M costs is:

               O&M  ($/yr) = 3% *  (220,000 *MW0>55)
                             C-ll

-------
C.3  LNB WITH CLOSE-COUPLED OVERFIRE AIR APPLIED TO COAL-FIRED
     TANGENTIAL BOILERS
C.3.1     Data Summary
     There were no available cost data for installing LNB with
close-coupled overfire air (CCOFA)  on tangentially-fired
boilers.  As a result, the direct cost algorithm was developed
based on the relative price differentials between LNB with
CCOFA and LNB with CCOFA and separated OFA (SOFA).   Based on
information presented by ABB-Combustion Engineering, the ratio
of LNB with CCOFA and SOFA to LNB with only CCOFA is $9/kW to
$5/kW, or 80 percent.5 This difference  corresponds  generally
to the price differential between LNB and LNB with advanced
OFA applied to wall-fired boilers.  The economy of scale was
assumed to be 0.60 for LNB with CCOFA (corresponding to b =  -
0.40).  This economy of scale is similar to that for LNB
applied to wall-fired boilers (b = -0.44).
C.3.2     Direct Cost
     As discussed in section 6.2 of chapter 6.0, there are no
incremental capital costs attributable to this control option,
and O&K costs were assumed to be 3 percent of the direct cost .
Cost  equations for direct cost are presented only to provide
the basis for O&M costs.
     Using the relative price differential for LNB with CC and
SOFA to LNB with CCOFA of 1.8, an algorithm for direct cost
for LNB with CC and SOFA was modified to develop the algorithm
for LNB with CCOFA.
     Dividing the LNB with CC and SOFA algorithm applied to  a
400 MW reference plant by 1.8 yields the basic system cost:

                DC  ($/kW) = 247 * 400~°'49 /I.8
                          = $7.3/kW
                             C-12

-------
Then, using b = -0.40, the coefficient  "a" was determined:

                      $7.3/kW = a *  400~°-4
                           a = 80
From this, the direct cost algorithm for LNB with CCOFA is:

                     DC  ($/kW) = 80  * MW~°'4
C.3.3     Indirect Cost
     Indirect cost factors were not required since there were
no direct costs attributable to this control option.
C.3.4     Fixed O&M Cost
     As discussed in section 6.2 of chapter 6.0, O&M costs
were assumed to be 3 percent of the direct cost.  Converting
the units of the direct cost equation from $/kW to $/yr, the
0&K equation is:

               O&M ($/yr) = 3% *  (80,000  * MW0'60)
                             C-13

-------
C.4  SNCR
C.4.1     Data Summary
     To estimate the cost of urea-based SNCR systems, a
detailed engineering model was used.  The detailed model was
developed based upon vendor supplied information on direct and
indirect costs6-7 and on system operating parameters.8  No data
were available on the cost of actual installations.  This
model requires 25 inputs, which are listed in table C-2.
     A total of 15 case studies were evaluated.  The results
for these case studies were used to develop simplified costing
algorithms for use in this study.
     For the case studies, the SNCR system operated at a
normalized stoichiometric ratio  (NSR)  of 1.0, and contained
two levels of wall injectors and one level of injectors in the
convective pass.   No enhancer was assumed to be injected with
the urea solution.  Urea costs were assumed to be $200/ton for
a 50 percent solution.
C.4.2     Direct Cost
     Direct costs include the urea storage system, the reagent
injection system,  air compressors, and installation costs.
The algorithm coefficients were derived by a non-linear
regression of cost data from the 15 case studies using the
methodology described in section C.I.   The coefficients were
nearly identical for the three coal-fired boiler types.
Therefore, the following algorithm was used to characterize
the costs for all three:

                    DC ($/kW) = 32 * MW"°'24
Similarly, the cost coefficients were nearly identical for
both natural gas- and oil-fired boiler types, and the
following algorithm was used to characterize costs for both:

                    DC ($/kW) = 31 * MW"°'25
                             C-14

-------
TABLE C-2.  SNCR MODEL INPUTS

Boiler Size
Maximum Heat Input
Uncontrolled NOX
Annual Capacity Factor
Exhaust Flow Rate
NSR
Number of Wall Injector Levels
Number of Convective Pass Injector Levels
Fuel Price
Urea Solution Price
Purchased Solution Urea
Injected Solution Urea
Enhancer Solution Price
Ratio of Enhancer Solution to Injected Urea
So" ution

Required Compressor Power
Compressor Air Volumetric Flow Rate
Compressor Air Pressure
Miscellaneous Power Requirement
Cost of Electricity
Cost of City Water
Maintenance Labor Rate
Operational Labor Rate
Maintenance Labor
Operational Labor
Maintenance Material
Units
MWe
MMBtu/hr
ppm dry
%
dscfm
N/NOX
#
#
$/MMBtu
$/ton
% by weight
% by weight
$/ton
by weight

kW
dscfm
psia
kW
$/kW-hr
$/1000 gallons
$/hr
$/hr
man hr/yr
man hr/yr
$/yr
            C-15

-------
C.4.3     Indirect Cost
     The SNCR model calculated two categories of indirect
costs:  a 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 indirect cost factor of 1.3 was assumed for the
calculation of total capital costs.
C.4.4     Fixed O&M Costs
     Fixed O&M costs for SNCR include operating, supervisory,
and maintenance labor;  and maintenance materials.  Fixed O&M
costs were estimated for each of the five boiler types using
the SNCR model and were found to be independent of furnace
type and fuel type.  Therefore, the following equation,
determined by the methods described in section C.I, estimated
fixed O&M costs for all five types of boilers:

                  FO&M  ($/yr) = 86,000  * MW0'21
C.4.5     Variable O&M Costs
     Variable O&M costs for SNCR include costs of urea and
electricity.  In addition, energy penalties associated with
the vaporization of the urea solution in the boiler, mixing
air, and dry gas loss are also included.  The urea cost is a
major component of the total VO&M cost and was determined from
the following equation:
 Urea Cost ($/yr)  = UncNOx  * HR * Cost * NSR *  5.71 x 10'3 * MW * CF
                             C-16

-------
 where:


     Unc NO  = Uncontrolled NO  emission rate (Ib/MMBtu)
         HR = Boiler net heafTate (Btu/kWh)
       Cost = Purchase price of the urea solution  ($/ton)
        NSE = Normalized Stoichiometric  Ratio  (N/NO)
 5.71 x 10   = Conversion Factor
         MW = Boiler size  (MW)
         CF = Average annual capacity factor (decimal fraction)


     Based upon the 15 case studies,  the other variable O&M

costs were estimated to be 11 percent of the annual urea cost.
                             C-17

-------
C.5  SCR
C.5.1     Data Summary
     The SCR cost estimates are based upon the SCR module in
version 4.0 of EPA's IAPCS9, published  SCR  cost  information10-11,
and utility questionnaire12-13, and the draft cost report
prepared by EPA's Acid Rain Division14.   The IAPCS cost
estimates were compared to a more recent estimate for SCR
systems."  The comparison revealed that the costs for SCR
reactor housing, process control equipment, and fans were
high.  Consequently, the IAPCS reactor housing and process
control equipment costs were reduced 71 percent and
80 percent, respectively.  Because actual fan modification
costs were low, these costs were excluded from the estimates.
The existing IAPCS algorithms were used to estimate ammonia
handling and storage, flue gas handling, air heater
modifications, and catalyst costs.
     A total of 15 case studies were developed using the
modified IAPCS output.  These case studies were for boilers of
100 MW, 300 MW, and 600 MW, for each of four boiler types
iwall and tangential coal-fired boilers, plus wall ana
tangential natural gas- and oil-fired boilers).  The results
from these case studies were then used to develop simplified
costing algorithms for use in this study.
     The IAPCS algorithms are based on hot-side SCR technology
(i.e., the catalyst is located between the boiler economizer
and air preheater).   For the case studies,  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 case studies.  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.  Cost of catalyst was
assumed to be $350/ft3.
                              C-18

-------
C.5.2     Direct Cost
     Direct cost for SCR includes both process capital and the
initial catalyst charge:
     DC ($/kW)  = process capital + initial catalyst charge
     Process capital is calculated by an equation of the form:
                Process  capital  ($/kW) = a  * Mwb
     Process capital includes ammonia handling, storage, and
injection; catalyst reactor housing; flue gas handling; air
preheater modifications; and process control.  The cost
coefficients for process capital were determined by the
methods described in section C.I based on cost data from the
15 case studies.  The coefficients for each combination of
furnace type and fuel type are presented below:
Fuel
Coal
Oil/Gas
Boiler Type
Wall
Tangential
Wall
Tangential
a
174
165
165
" R £*
.*- D —
b
-0.30
-0.30
-0.329
-0 . 324
     The equation for estimating the cost of the initial
catalyst charge is based on the IAPCS documentation and EPA
method 19 for estimating flue gas flow rates:
  Catalyst  ($/kW)  = Flow*Cat$/{SVf * [In(0.20)/In(l-NOxRed)]}
where:
     Flow =    Fuel - specific flue gas flowrate in normal cubic
               feet per kilowatt-hour (Nft3/kWh)
               (126 Nft3/kWh for coal,  100 Nft3/kWh for gas
               and oil)
     Cat$ =    Catalyst cost  ($/ft3)
     SVf  =    Fuel- specific space velocity
               (3,200/hr for coal, 5,000/hr for oil, and
               14,000/hr for gas)
     NOxRed =  Target NOX reduction  (in decimal fraction form)
                             C-19

-------
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.
C.5.3     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.
C.5.4     Fixed O&M Costs
     Fixed O&M costs for SCR include operating, supervisory,
and maintenance labor; and maintenance materials.  Fixed O&M
costs in $/yr were estimated for each of the four boiler types
using the IAPCS model.9   The  resulting  data  were  then  used  to
develop a cost algorithm as discussed in section C.I.   The
coefficients for each combination of furnace and fuel type for
use in equation C-5 are presented below:
Fuel
Coal
Oil/Gas
Boiler Type
Wall
Tangential
Wall
Tangential
e
284,600
276,400
264, 800
256, 600
f
5,141
5, 103
3,260
3,219
C.5.5     Variable O&M Costs
     Variable O&M costs for SCR include catalyst replacement,
ammonia, electricity, steam, and catalyst disposal.  Cost for
these elements were derived from the IAPCS model.'5   The
equation for estimating catalyst replacement cost in $/kW-yr
was based on the case studies and the IAPCS documentation:
            Catalyst replacement  cost  ($/kW-yr)  =
  Flow * (Cat$ + 160)  / {SVf *  [ln(0.20) / In (l-NOxRed) ] } / CL
                             C-20

-------
where:
     CL   =    Catalyst life  (years).
     160  =    Cost to cover  installation and disposal of
               replacement catalyst  ($/ft3)

Other variables are as previously defined.
The equation for estimating costs in $/kW-yr for the other
three variable O&M components  (ammonia, electricity, and
steam)  was also based on the  case study data and the IAPCS
documentation:

     VO&M ($/kW-yr)  = [1.88 +  (4.3 * UncNOx  * NOxRed) ]  * CF
where:
     UncNOx     =    Uncontrolled NOX emission rate  (lb/MMBtu;
     CF        =    Capacity  factor  (decimal fraction)
     1.88 and 4.3 = Regression coefficients

Other variables are as previously defined.
                             C-21

-------
C.6  COMBINATION CONTROLS - COMBUSTION CONTROLS WITH SNCR OR
     SCR
     Costs for combined control options were determined by
combining individual cost algorithms for each control option.
For example,  to calculate the costs of CC + SNCR,  the
individual capital,  variable O&M,  and fixed O&M cost
algorithms for CC were combined with those for SNCR.
Similarly, the CC cost algorithms were combined with these for
SCR to calculate the CC + SCR costs.  Refer to each individual
section for the specific cost procedures.
                             C-22

-------
C.7  REFERENCES
1.   Letter and attachments from Brownell, F. W.,  Hunton and
     Williams, to Neuffer, W. J.,  U. S. Environmental
     Protection Agency.  December 1, 1992.  Information
     collection request from Consumers Power - J.  H.
     Campbell 3.

2.   Questionnaire Response - Plant D.

3.   Questionnaire Response - Plant G.

4.   Analysis of Low NOX Burner Technology Costs (Draft).
     U.S. Environmental Protection Agency, Office of
     Atmospheric Programs.  Washington, DC  20460.   February,
     1993 .

5.   Grusha, J. and M. S. McCartney.  Development and
     Evolution of the ABB Combustion Engineering Low NOX
     Concentric Firing System.  TIS 8551.  ABB Combustion
     Engineering Service, Inc.  Windsor,  CT.  1991.

6.   Letter and attachments from R. D. Pickens,  Nalco Fuel
     Tech,  to N.  Kaplan,  U. S. Environmental Protection
     Agency.  January 20, 1992.

7.   NOX Emission Controls for Utility Boilers.   Utility Air
     Regulatory Group.  January 1993.   Table 12.

8.   Letter and attachments from R. D. Pickens,  Nalco Fuel
     Tech,  to E.  W. Soderberg, Radian Corporation.
     February 8,  1993.

9.   Integrated Air Pollution Control System, Version 4.0.
     EPA-600/7-90-022b.  U. S. Environmental Protection Agency
     Air and Energy Engineering Research Laboratory, Research
     Triangle Park, NC, 1990.  Section 4.8.

10.   Cochran, J.  R.,  et al.  Selective Catalytic Reduction for
     a 460  MW Coal Fueled Unit:  Overview of a NOX Reduction
     System Selection.  Presented at the 1993 Joint Symposium
     on Stationary Combustion NOX Control.  Bal Harbour,  FL.
     May 24-27, 1993.

11.   Technical Feasibility and Cost of Selective Catalytic
     Reduction (SCR)  NOX Control.   Electric Power Research
     Institute.  EPRI GS-7266.  Palo Alto, CA.  May 1991.

12.   Fax.  J. Klueger, Los Angeles Department of Water and
     Power,  to M.  J.  Stucky,  Radian Corporation.  September 9,
     1993 .
                             C-23

-------
13.   Letter and attachments concerning SCR costs.
     Confidential Business Information.  Reference
     Number 93197-111-01.

14.   U.S. Environmental Protection Agency, Cost Estimates for
     Selected Applications of NOx Control Technologies on
     Stationary Combustion Boilers,  Draft Report.  Acid Rain
     Division, Washington, DC.  March 1996.
                             C-24

-------
                       APPENDIX D



         MODEL BOILERS EMISSIONS AND COST DATA








Table D-l.  Bituminous Coal Emissions and Cost Data



Table D-2.  Subbituminous Coal Emissions and Cost Data



Table D-3.  Lignite Emissions and Cost Data



Table D-4.  Natural Gas Emissions and Cost Data



Table D-5.  Oil Emissions and Cost Data

-------
^


5
(5
•o

«
w
c
o
M
3
w
3
O
C

I
s
o
V



8
o
i
Emission





in

0
o
co





8

8
1
in

-------
 o
 c
 c
 w
 K
C
 C
U
•a
 9S
 s
 K

 if.
 C
 C
I
 w
£
C

x ^C
0 «£,
2 S
Incrcmcr
Iftectivcrx




•f
u
.0 Z
C^ (y
"y;
~



-r
'""'' -^
X ^
1 s
<


>
_u:
^
J3 ~

u



g "
O -7
? £ ft
3 8
a:
u
£ sb
'O
o

•*• o£
s-> u

"V
-^

O

^ X
*v/
y Z
-' £
U
'^i

'-> u
_• :/•-
a:
Hz
-J £
u
o
5l
2I
§


oe


fN
^


-

"


r j


rf


X
X





c
r~
<"


5c
DC



rj

C
ri
>/">
-T

^T


OC
n
(N


rr-
r-


r\


ri
\D
*O
O



-T

0
(S
oc
TT

•^T


-^r
w,


i/",
•n

™"

r-

•*

^


r-i
•^c,



'A

O
ff,
^T

^


m


>/-




oc
3C

3C


n
O





C
TT
or.
oc
r-
S


ft
A
^T

t~-

r '1
oc
^

^


'A
^





c
1-
vC
OC

^


C5



r^

ly_


rr
r-



r-l

O
r
oc





r-
r-
X
1-*",



»/"l

O
3
o
o
*r>
rs


C^
sC-


PC



r-
f-





O
^





\s
r^-
^3-

^


(N
t/"i


(N
ir>

•/",

^






5
o



r—

c
r^
O
•^r

3*


TT
tv.


"T

rr

r.

X




5
^c



X

o
»n
r~
oc
c
>£>
c


r)
•r,
(N

>/•,

r-

r^-

J




rj
»/~i
C
^N





C
o
^c
•q-

s


r*-,
•q-


T



"
rrj





r^
'/•,
=





-^r
c^
r*',

^


^C
rN


T
-t



oc
r j
Ti





ri
'/-.
0





be
»r,
oc

•c


f,
^r


«/"!



r-
ri


~


r-
TT
ON
r-



r j

O
fN
'n
-3-

oc


oc
^


rf
-T

r«-

-
r-





r^
t—





s
oc
ff,

•o


'r,


"7

rr

r j



rj


r-
0
r-



T

<

0
IT)
C^

<


r~)
r*-,
r-i



<
Z
DC
'




<
Z
fN
w^



Tl

<
V

o
w-1
1—
f*~,

<
Z


T
rf.



<
Z
oc





<
Z
ri
o



^C

<


o
rr,
1^1

<
Z






<;
Z
X:





<
i~l
3*,



t^

<
r

0
S
f*~j

<
Z


-



<
Z
r^





<
Z
sc



oc

<
r

c
^
oc


<
Z


X



<
Z






<
21
oc
r*-.





<


r^


<


C:



<
Z
r^-





<
Z
DC





                                                     D-2

-------
s
Ta
Q
 o
u

•o
 c
 a
 CO
 c
 o
"35
 w

I
UJ

"n
 o
O
 in
 3
 O
 C

 E
 s
V)


e\i

Q
 a



£
o
s
i
IJOISSIUJ3




ro
(5
Control
usiion
u
O
J
CD
01
Vi
TO
0






£*^
QC
O
fO
+
0
O
A
tr
o
(/i
8

o
o

"re
<


To
c
<
Capacity
0)
•O


Furnace

c~
,>,
I
e
c-
°c
5.
£
S
g

g.
(A
S *:
w ">•
ii
LU
a £
'TO ^
*2
.D

Q.
t-
.
"5
LL
O
5

00
00
§

8
8

CD

0
TO
V

O
TO
O
8
ID

o«
O
TO
^

8
"re
I Subbituminous
o

S
§

8
8

00

o
CO

o
TO
O
O
o
TO

B?
in
CO
CM



"TO
o
O
8
TO

OO
eo
0
S

8
S
in

CO

o
in
O)

0
TO
O
0
UC)

B?
CO
TO




8

00
00
o
5

s
§
N.

00

8
CO

0
CO
o
0
CO
o
CM
B«
o
CO
•V

8
CO


o
5
CO
s
§
CM
8
8
*~
CO

O
CO
V
CM
o
CO
o
8
CO
CO
s?
in
ID
in




o
£
TO
8
§
TO
S
1
^~
CO

o
s
CM
0
CO
o
o
«•
in
*}•
B«
O
CO
CD




1
CO
s
§
CM
$
§
*"
CO

§
CM
%
O
s
Tj-
s«
o
CO
r*.

§


8
CO
to
s
8
in
8
§
eg
03

o
r>-
CO
TT
O
CO
o
8
(^
r^
S*
in
(O
CD




S
r>»
00
oo
o
cti
(O
S
0
?
m
00

o
§
in
o
CO
o
o
00
o
0)
s?
o
CO
O)




0
r«-
o
ID
S
O
f-
r*.
<
S
§
OJ
00

o
*
CO
^•
o
CO
o
o
3
(D
1
O

§


8
CO
»•••
00
oo
o
Si
oo
s
o
t~-
00
T
CD

o
CO
o
CO
o
o
§
oo
s?
in
(O
^




o
in
CM
TO
CO
00
0
5
o
s
o
s
in
CO

S
TT
O)
o
CO
o
0
rr
in
S?
o
CO
CM




O
CO

00
00
§

8
8

CO

o
CO
V

0
CO
o
s
co

g
CO
CO

8
Tangential

0

9
00
8
CO

8
8
*

CO

o
CD

O
CO
o
o
o
CO

B?
in
co
Tf




s
TO

£
o
s
^-
8
o
N.
in

CO

s
O)

o
CO
o
o
in

S?
o
CO
in




§

00
00
o
?

8
§
r^

CO

o
o
CO

o
CO
o
o
CO
o
CM
s?
0
CO
ID

8
TO


O
?
TO
00
00
O
8
CM
8
8
v
~
CO

o
CO
T
CM
O
TO
O
S
CO
TO
0?
in
(O
f-




O
r«.
0)
TO
S
O
CM
TO
S
1
^
CD

o
*T
CO
CM
O
TO
O
O
•«•
in
^•
6?
o
CO
CD




O
S
TO
S
8
CM
8
8
in
~
CO

o
8
CM
O
TO
O
8
T
6?
O
TO
0)

§


S
CD
(O
S
8
TO
in
8
|
CM
CO

O
f^
CO
^r
o
TO
O
S
r^.
f^
S?
in
CD
O




S
r^.
oo
00
o
Si
(O
8
o
5
TO
CO

1
in
O
TO
0
o
§
O)
6?
o
00





o
r>»
o
(O
oo
00
g
r~
•*
8
§
CM
CO

0
t
TO
••9-
O
TO
O
O
s
CD
S?
O
TO
CM

1


8
TO
^
S
O
SI
oo
8
o
f^
CD
•*
CO

o
CO
o
TO
O
0
§
CM
s?
in
(D
TO




o
in
CM
TO
^~
0§
0
V
0
8
§
in
CO

S
V
O)
o
TO
O
O
T
in
8«
O
CO
S




<

<
8

F-.
o
r^

r^

o
f»

O
CM
O
o
1-

6?
o
TO
in

in
CO
O
m
u.

<

i
8

S
§

r*.

O
^p

O
CM
0
g
CM

S?
in
(D
(D




<

<
Z
§

r-
o
^~

r~

S

O
CM
O
8
TO

S?
§
r^




<

<
8

S
8
CM

f-

g

O
CM
0
o
in
in

S?
o
TO
CO

O
O


<

<
8
N.

JJ
8

r>-

o
3

o
CM
O
O
TO
O

s?
IS
05




<
Z

<
Z
§

r^
o
Z

t*-

o
s

o
CM
0
o
8

B?
o
00
g




I
                                                      D-3

-------
•B
 V
•O

"5
 e
 e


 ee
••*

C
•o
 c
 K
 v.
 S
 v;
 v.
fi
^ £
5 «£,
5 -S
Incrcmci
EfTcclivcn



v
o
.a •£
s: v^
~




^
— 5
o -*
i K
E 2
<



>"
-*.
(/;
H 0
"re
c.
CJ



CO SCR
\s
CCfSNCR
a
CJ
£ ssH
o
'_;

Pp

'V'
^

O
- 2£
"_- V.
*S_
P g
S"' v;
LJ


y ^

_ si
n
\j
'^>
il
5I
o
fN
"*l
r*~,
r$
(N
CN

-?
rr

-


—

-

«•


C:







O
O

»A


^

'/"i
r-

rj

rj


c

ac
rr


C^,

^T
(S
O
oc
"=T


»/-»
^


OC
r-

r]
«r

ri

D


ri


-


T

S
*C,
vC
C
O
c


^\
t


'^




I/-,
oc

CS


ri


—


»^

o
r~
vC
irT
§


r*-.


»?)


OC

l>

X


ri


~


o

o
oc
e>
c
(N
(N


5e
Tf


T
^

IA
_

£

i/",


~T


C^


r^-

c
GO
r^
\c
rs


"5


*c
"7

X
•c

rv
r-

t


n


c^-


oc

oc
T.
>A)
f^


(N


-r

r-^
3C
X

Ti

\c


•^r


c^


^

c
tn
r-;
I


^


l^l

n
ri
oc

ri

r i


^


r-




s
/",


~


r-


ri

o
oc
(N
»^-,
(N
r,

T.
=

5c




0

s£



r-

vC
c
?


S
rj


t^-


;~

r^




—


2


oc

C
TT
n
T
H

0
v,

r-
—

a:




rj


^




C
r«-,
^C
C
fS


oc


rr
T


3C

^




ri


^v




O
(N
•V,
t*n
0
r-


/~l


r*~.


-?


r i

r*-
O

r*-


r-


r-'


r*-:

f *
rr,
""i
O


r--


oc


r~-
r J
r-
>r,

rj


r-


r-


•T

<
»/")
vC

<


fN


<

ic

-------
a
O
T»
C
CO

m
v>


UJ

15
o
U
c
O)
CO
a
o



E
o
1
c
in
E
jj




(0
CO
Control
0
«l
D
E
j
i
o>
I/)
re
E






^
d
O
fO
+
O
o
S"
cc
o
+
o
o

C)
0

"5
<


"5
c
<
Capacity
0)
•o


Furnace

1
I
?
1
SI1
2
&
1
S

?
V)
is
11
UJ ""
2"
01 CD
f 5
cc 5
^
VI
1?
II
LU
0
15
1C
LL
.
01
a.
>.
"5
LL
S
N.

$
S

18
§


co
DO
if>

o
0
o
S

«£
o
to


8
~n
£
c
D)
J
S?
m

8
8

S
S


CO
§

o
o
o
f>

•^
m
(0
8




|

£
§

$
£
in


co
S

o
o
0
co
CO

S«
O
CO
S




o
r«.
CM
£
§

S
S
N.


CO
o
co
t^

o
d
0
in
CM
CO
S

8
CO


§
V
8
o
8
CO
8
8
*"

CO
o
in
CM
CO
o
o
o
r^
•t
S?
in
to
S




8
5»-
8
§
CO
8
1
"

CO
1
CO
0
o
8
in
S?
8
S




1
v
S
8
CO
co
S
|
^

co
g
CO
o
0
0
CO
o
m
CO
!S

§


o
eo
!§
o
fs.
(O
S
cf
CM

CO
8
f
(D
0
O
O
?
O)
5?
in
(0
S




8
o>
S
8
r>>
S
o
?
CO

CO
o
r^
in
t^
o
o
o
r-»
O)
o
i
S




|
r^
S
1
m
S
§


CO
o
CM
CD
O
O
O
8
(D
in
8?
in
(0
_




S
h-
in
S
§
^-
S
I
in

n
o
to
CM
O
O
8
C\J
CO
00
CM




8
r»

£
S
m

S
S


CO
o
00
in

0
o
o
^r
CO

CO
CO

8
*—
Tangential

S
co

S
§

S
8


co
o
§
T-
0
O
0
r^
in

55?
in
(O
v




i

£
|

£
§


co
8

$
o
0
CO
00

i
in




R
CM
S
§

$
S
r-.


CO
0
CO
h-

o
o
o
in
CM
CO
(O

8
CO


§
•«•
£
§
eo
18
8
"

CO
S
CM
CO
O
O
o
r-I
*
s«
S
r^




R
*•
£
§
eo
$
1
"

CO
§
f»
CO
o
o
8
V
in
S
00
00




§
*•
(O
eo
8
co
CO
S
§
"

CO
o
r^
t
CO
0
o
o
§
in
eo
O)

1


0
»•-
CO
£
R
(0
18
§
CM

CO
8
^1
(0
o
o
o
•»
o>
«
(O
8




8
a>
S
8
^»
h*
S
o
*"
•»
CO

CO
o
h.
in
r^
o
0
o
N.
O)
o
i





1
r~-
£
8
in


CO
S
CO
o
0
o
8
in
8?
£
S




g
h-
m
s
§
^
CO
25
i
in

CO
o
5
CM
O
O
8
00
g
GO
S




i

<
8

5
o
^



g

r?,
O
o
V-

S?
o
eo
in

in
CM
O
m
u.

i

<
8

R
8
^~



o
>«•

SI
0
8
CNJ

55*
in
CO

GO



i

<
§

R
|



§

o
CM
0
§

I
r*




<

<
8

r>-
8



8
CM

SI
0
8
in

g
eo
CO

8


<

<
8
r>-

s
§



o
5

r?,
0
O
8

5?
in
CO

eo



<

<
§

R
§



§

ft
0
8
CM

g
00
8




!H
3 s g
*^ II 5%
    ^x

    U
                          D-5

-------
•o
_s
~
 c
 o
 cs
c
•c
 e
 re


U
 o
 I

C

'x ,C
5 *£,
3 S
Incrcmei
liffcelivcn



y
O
b £
c: n


t^
O





>r,


—


0

X



r*-.
^C
g
O^
un
n


(V,
r~
(N

r-

rj



X
r,


*r,
>/-.





£


=:


r\

^C



'/^
vC
O
f*',
TT
O
C>


i/~i
rn


«/^





DC
oc


—


ri

^



^
^C
C
2
oc
i


v^
'/•l
*N

\C





r-i
rv


—




	 	 	



r-
\c
*?>
*c
^
c
T


l/^
TT


OC

3C

»/•

•*c
oc


c


TX

	



oc
vC
5c
0s
r*",



r^
rsi


>r.
rr

r*^
r)


oc
r-


o


-T

	 ,



C^
O
P
O
oc
o
"sC


5
T
(N

ir.



3C

oc


—


I-

c^



p
§
>n
T
C
oc


•5


•4





X


c;


-

o



r-
Ov
m
O
o


rv,
rt


r^i
T



X

ri
t^


=


~

ON



CN
r-
o
fN
OC
Ov
C
oc


C:


fj



DC

2


—


r^

oc



r<,
r~
c
oe
i/~i
c
•^p
fS


t—
r-


^c

oc
oc
l^t

ri


=


r-

5C



-^-
r~
c
^
i^-
v
c
TT


C^
'/~,


-^

r^-

>r

X


-


r~

oc



'/"l
r-
c
2
o>
>?i
(N


O


r^



~T

r-
-T


=


0

^C



-C
r-
o
}S
^r
o
'^>


n
«/^


(N
>/•.



C

r*"





O

O



r^
r-
fC
c
^r
c
o


^


a:



c

rr
3C


=


-

^C



X
r-
c
>n
r-
oc
o
\£


rs
»n
r<

c^
«r,

r-~



T
/•>
vO

<
Z



-------
 a

 O


 O
 U
 •O
 C
 «
 II
 c

 ~v>


 I
 ill
a
a
i
o
0



£
g
S
*
Emission 1


£
e
Control C
c
o
w
iCombi








2°
S
CO
•»•
8

S
4-
8
o
u
_
<


Baseline
Annual
Capacity
v
t)
o


Furnace


S
4*
*.
£
4*
f
#
c —
o >.
11
c &
LJ
•?
i 1
= S.
.C
Emissions
(ton/yr)
Factor
£
£>
3
s r
W S
t>
a
>s
(-
D Fuel Type
8
CM

O
CO
S

S
o
0
o
in
CM

oo
o
0
in
CM
#
O
CO
5>

o
o
I
j Natural Gas
o
r»
n

o
CO
o
8

o
m
o
o
o
r

o
CO
o
r-
CM

O
in
o
0
0
V
m

03
O
O
V
tn
#
o
00
n
01




g
«O

O
CO
o
K
n

g
o
0
n
v
f-

00
o
in
v
N.
#
o
<•>
^-
O)

o
o
n


g

o
00
0
3

g
o
o
0
CO
n

CO
0
o
00
o
#
in
<£>
tn
o>




o
S

S
g
CO

o
en
0
o
o
CD

CO
o
o
to
#
o
CO


§
ID


O
^
CM
N
S
S
n

S
o
o
o
r*
r~
CM
CO
o
o
r^
N.
CM"
#
in
(C
CO
0>




S
n
CM
O
co
o
CO

o
V)
o
o
o
CM
CM
r3
cc
o
o
CSJ
CM
P5
#
O
00
o>
en




g

o
CO
o
V
CM

O
m
o
0
0
CO
V
CM
CO
o
0
00
^1
CM"
#
o
n
o
o

o
o
o


o
3
n
S
g
n
CM
o
in
o
o
o
tc
•V
CO
o
o
5
V
4*
in
CD
o




1
^
S
o
S
CM
O
n
o
0
g
o
in
03
O
0
1C
r>_
in*
#
o
CO
CM
O




g
tJ

S
o
CM

O
in
o
o
o
in
CM

00
o
o
in
CM
4*
o
n
r>
O

o
o
Tangential

g
CM

O
CO
o
n
CM

0
in
o
o
0

4*
in
CO
r^
o




g

o
CO
o
S

S
0
o
0
(0

CO
o
0


o
CO
in
*»
r^

g
o
0
0
o>
V

00
0
o
Ol
•V
4*
o
n
g

§
(C


0
CM
CM
O
CD
O
CO
n

0
in
o
o
0
r^
K
CM
CO
o
0
f~
r^
CM"
4*
in
(0
o




0
h-
m
CM
S
1

g
o
o
o
CM
CM
n
CO
o
0
CM
CM
n"
•4*
O
CO
._




0
O)
o>

§
o
^
CM

O
in
o
o
o
CO
"W
CM
CO
o
0
00
V
CM"
4*
o
n
CM

O
o
o


o
S
CO
o
CO
g
r>
n
o
in
o
o
o
to
^
CO
o
o
CO
V
4*
in
1C
n




g
V
S
o
CO
CO
CM
O
in
o
o
o
ID
r>
m
CO
o
o
1C

m"
4*
o
CO
»




                                                 D-7

-------
•o
 u
•a
^
~
 e
 c
 if.
 e
U

•o
 c
 re
 IT
c
 I/

jl
*5 w
Increinci
EtTectivcn


./•
S _






- -s
5 -^



^,
"2 ~
~ 15
U



s 1
3 s
a:
'j
r s£
'—> u






-v
s^ '-S.
ii

w1
- ^

, tt
y£


"8 "=

^

c

c
oc

-------
O

V
O
U

•o
c
«

m

_o
71
jo


UJ
in

6

£
zi
a








s
^
s>
•p


M
»
E
UJ




Te
O
'o
c
o
c
°
3
.c
r-
0
J


























&
C
U
(0
4-
8

—•.
£
E
o
z

»
^



o
u
"5
c









|i

m ^

$
o.
I)
U


0
S


c



3
U.






^
o




V*
.
"^
o




2^


1
£

•
V)
f I
UJ

3-
0 B
•g 5
a 5



I 5
S c
•S o
E «.
UJ
6
«•

t
^
E
Z
.§ g


c
a

*~

a


u.
o
0





s

o
o>




s




o


o
o
a>
CO




cc
CM
o



o
o>
CO



*
CO


m
*-

o
o





"5


sr
O

g
in





S

s
r>




0
m




o


o
o
CM




CO
CM
O



f^



*
in
10


CO














o
f-
to





s

o




o
in




0


o
o




CO
CM
0



o
CO
CO



*
CO


^
*-













o
CO
a





S

o
CD
n




0
m




0


o
o
(O
'



CO
w
0



g



#
CO


CO
•-

o
o
CO










g
"




s

o
^.



o




0


o
o
in
CM



CO
CM
O



0
in
CM"



*
in
(0


o>














|
CM




S

s
,_



g




0


o
o
o
CM



CO
CM
O



§
in
CM"



*
CO


0
CM













O
in
CO
*-




s

o
CO
^



o
m




0


0
o
CM
CO
CM



CO
CM
O



0
CM
CO
CM"



*
CO


^
CM

O
O
1C










o
t
n




S

o
CM



O
m




o


o
o
CO
*



CO
CM
O



O
CO




in
10


CM
CM













|
«




O
CD

g
n
CM



o
in




o


0
8
o
in



CO
CM
o



g
o^
in



#
o
CO


CO
CM













i
CO




o
CO

o
CO
at
^



o




o


0
S
CO
CO



CO
CM
O



o
1C
CO
co"



r*
O
CO


^
CM

8











o
in




S

o
n



g




o


o
o
r^
f^



CO
CM
0



g
K



#
in
10


in
CM













g
co
co




s

g
^f



o
in




o


o
0
V
CO
CO



cc
CM
o



o
CO
co"




£


CO
CM













o
co





o
CO

o
a




o
m




0


o
o
s




CO
CM
O



0
o>
CO




o
CO


ps.
CM

s


0

r
to
CJI
c
i-




o
h.
n





o
CO

s
n




g




o


o
o
CM




CO
CM
O



O
CM




in
10


CO
CM













o
o





s

o
CM




0
in




o


0
o
to
CO




CO
CM
o



o
CO
CO




o
CO


Ol
CM













s
a





s

s
in




0
in




o


0
o
1C
'



CO
CM
o



o
ID




o
CO


o
CO

0
CO










o
CM
»•




S

1
^



0
in




o


0
o
in
CM



CO
CM
o



o
in
CM"



#
in
10


^
CO













1
CM




O
CO

o
in
CM
*B



O
m




o


o
8
in
CM



CO
CM
O



o
o
in
CM"



*
o
cc


CM
CO













O
g
•-




o
CD

g
v.



o
in




0


o
o
CM
CO
CM



CO
CM
o



o
CM
CO
CM"



*
O
D


CO
n

R
1C










o
CO




s

o
in
CM



g




0


o
o
CO
*



CO
CM
o



o
CO




in
(0


^
CO













§
*




s

m
CM



g




o


o
8
o
in



CO
CM
O



in"



#
o
CO


in
CO













o
s
p>




o
CO

o
CO
0)
^.



o




o


o
§
CO
CO



CO
CM
0



o
10
CO
CO"




o
CO


CO
CO

g
~










0
K
in




S

s
m
n



o
in




o


0
o
r*



CO
CM
o



o




in
to


fs.
CO













g
to
«




s

o
(If



g




0


o
0
CO
CO



cc
CM
o



o
n
co"



#
o
CO


CD
IO













                                                 D-9

-------
 e
 o
 If.
 o
U
•B
 5
 e
 tr
 "
.^
«
5 ~-
c ^
"re
'E,
u



•v "•
D 4
* % &
* 't
&
•-j
S  u
L)
O
og

^
a*
u>
Sw>
_ -v*
:-' o

jV
5g
o
s_;

u - v-
+ &
'^
0 &.
•~J
u
5l
2I
o
oc
t^
!>e
(N
T
(N
§
>n
T

r\
"^
(N


a

X




_

X





§
vO
T
O
r-
»r^

O
*ri
O
(N

r-
ir~,


r~-

,
—





oc
r^.



\C

0
£
»»
o
m
•q-

o
t
f,
rs

•^


r i

'A
^-





oc
fr,



r-

c
(N
I~-
O
(^
in

O
>^-
>n
r^i

*^1


£







•rr
\C
O



ex:

§
oc
r^i
O
fS

i
c
n
ro
CO
0
V"l
O

C
in
00


*n
C


r^

n
—




rr
sC
o



rj

(^
*
o
o
PO

c
r-~


r?;


r-i
r-


"





C
f>



r-)

O
O
o
r*"*
O
r^
OS

oe
oc


r^
o^


X


"^





r>



TI
oc


-








r-



^c

o
o
r*,
00
C
(N
•

O
O
^
»N

1.


C-








oc
r*-,



^

O
»n
O
(^
o
ON
m

C
oc
T
r*"j

>n


=






T

^



-

OC
f^i
c
CN

0
in
O


c
r*-j
r*^

r*-,


tJC






O

rr>



^

o
in
m
fi
o
r~
CN

C
\c
oc


r^
a-


r^-
3C






^

c>



T

O
r^t
O
O

s
^c


3s


X
T






"O

r>



>/-i

o
m
r'1,
^
0
>n

O
rr
0
(N

«r-i


\c

*Ai




ri

r~



^

O
^
r*%
O
ffi

o
vC
r-


oc


r^
r-~






r4

O^
I—



r~

o
*

-------