v>EPA
United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/7-79-178g
Laboratory December 1979
Research Triangle Park NC 277-11
Technology Assessment
Report for Industrial
Boiler Applications:
NOX Flue Gas Treatment
*
Interagency
Energy/Environment
R&D Program Report
-------
-------
EPA-600/7-79-178g
December 1979
Technology Assessment Report
for Industrial Boiler Applications
NOX Flue Gas Treatment
by
Gary D. Jones and Kevin L Johnson
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
Contract No. 68-02-2608
Task No. 45
Program Element No. INE624
EPA Project Officer: J. David Mobley
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This study assesses the applicability of NO flue gas treatment
X
technology to industrial boilers and is one of a series of technology assess-
ment reports to aid in determining the technological basis for a New Source
Performance Standard for Industrial Boilers. The status of development and
performance of alternative NO flue gas treatment control techniques were
assessed and the cost, energy, and environmental impacts of the most promis-
ing processes were identified. It was found that processes utilizing selec-
tive catalytic reduction (SCR) of N0x with ammonia can achieve 90 percent
reduction of NO emissions, and that these processes are the nearest to com-
X
mercialization in the U.S. In Japan, SCR processes have been successfully
operated on commercial scale gas-and oil-fired sources and are being
installed on coal-fired sources. Cost estimates of applying SCR processes
in the U.S. indicated that the cost effectiveness varies significantly de-
pending on the fuel fired, boiler size, and control level. However, boiler
size is the most significant factor affecting cost effectiveness with the
economy of scale causing control of large sources to be the most effective.
The energy impact of applying SCR processes averaged about 0.5 percent of
boiler capacity. No adverse environmental impacts were apparent although
there are emissions, liquid effluents, and solid wastes that must be con-
trolled. For regulatory purposes this assessment must be viewed as pre-
liminary, pending the results of the more extensive examination of impacts
called for under Section 111 of the Clean Air Act.
-------
PREFACE
The 1977 Amendments to the Clean Air Act required that emission stan-
dards be developed for fossil-fuel-fired steam generators. Accordingly,
the U.S. Environmental Protection Agency (EPA) recently promulgated revisions
to the 1971 new source performance standard (NSPS) for electric utility
steam generating units. Further, EPA has undertaken a study of industrial
boilers with the intent of proposing an NSPS for this category of sources.
The study is being directed by EPA's Office of Air Quality Planning and
Standards, and technical support is being provided by EPA's Office of
Research and Development. As part of this support, the Industrial Environ-
mental Research Laboratory at Research Triangle Park, NC, prepared a series
of technology assessment reports to aid in determining the technological
basis for the NSPS for industrial boilers. This report is part of that
series. The complete report series is listed below:
Title Report No.
The Population and Characteristics of Industrial/ EPA-600/7/79-178a
Commercial Boilers
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178b
Applications: Oil Cleaning
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178c
Applications: Coal Cleaning and Low Sulfur Coal
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178d
Applications: Synthetic Fuels
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178e
Applications: Fluidized-Bed Combustion
Technology Assessment Report for Industrial Boiler EPA-600/7/79-178f
Applications: NO Combustion Modification
iii
-------
Title ' Report No.
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178g
Applications: NC) Flue Gas Treatment
X
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178h
Applications: Particulate Collection
Technology Assessment Report for Industrial Boiler EPA-600/7-79-178i
Applications: Flue Gas Desulfurization
These reports will be integrated along with other information in the
document, "Industrial Boilers - Background Information for Proposed Stan-
dards," which will be issued by the Office of Air Quality Planning and
Standards.
IV
-------
CONTENTS
Abstract , . ii
Preface m
Figures xi
Tables „ xviii
Acknowledgements xxvi
1. EXECUTIVE SUMMARY 1
1.1 Introduction 1-1
1.1.1 Background and Objectives 1-1
1.1.2 Report Organization and Approach. 1-1
1.1.3 Scope of Study 1-4
1.2 Flue Gas Treatment for Control of NOX Only 1-11
1.2.1 System Descriptions 1-11
1.2.2 Economic Impacts 1-15
1.2.3 Energy Impacts 1-31
1.2.4 Environmental Impacts 1-40
1.2.5 Development Status 1-45
1.3 Flue Gas Treatment for Control of NOX and SOX. 1-49
1.3.1 System Description 1-49
1.3.2 Economic Impacts 1-52
1.3.3 Energy Impacts 1-55
1.3.4 Environmental Impacts 1-55
1.3.5 Development Status 1-57
References 1-60
2. EMISSION CONTROL TECHNIQUES 2-1
2.1 Principles of Control 2-2
2. 2 Controls for Coal-Fired Boilers. 2-6
2.2.1 Selective Catalytic Reduction-Fixed Packed Bed
Reactors 2-6
v
-------
CONTENTS (Continued)
2.2.2 Selective Catalytic Reduction-Moving Bed Reactors.. 2-7
2.2.3 Selective Catalytic Reduction-Parallel Flow
Reactor 2-7
2.2.3.1 System Description 2-7
2.2.3.2 System Performance 2-13
2.2.4 Absorption-Oxidation 2-15
2.2.4.1 System Description 2-15
2.2.4.2 System Performance 2-21
2.2.5 Selective Catalytic Reduction-N0x/S02 Removal 2-22
2.2.5.1 System Description 2-22
2.2.5.2 System Performance 2-37
2.2.6 Adsorption 2-37
2.2.6.1 System Description 2-37
2.2 2.2.6.2 System Performance 2-41
2.2.7 Electron Beam Radiation 2-41
2.2.7.1 System Description 2-41
2.2.7.2 System Performance 2-44
2.2.8 Absorption-Reduction..., 2-45
2.2.8.1 System Description 2-45
2.2.8.2 System Performance 2-54
2.2.9 Oxidation-Absorption-Reduction 2-56
2.2.9.1 System Description 2-56
2.2.9.2 System Performance 2-63
2.2.10 Oxidation-Absorption..... 2-64
2.2.10.1 System Description 2-64
2.2.10.2 System Performance 2-68
2. 3 Controls for Gil-Fired Boilers 2-69
2.3.1 Selective Catalytic Reduction-Fixed Packed Bed
Reactors 2-69
2.3.1.1 System Description 2-69
2.3.1.2 System Performance 2-75
2.3.2 Selective Catalytic Reduction-Moving Bed Reactor... 2-81
2.3.2.1 System Description 2-81
VI
-------
CONTENTS (Continued)
2.3.2.2 System Performance 2-87
2.3.3 Selective Catalytic Reduction-Parallel Flow
Reactor 2-93
2.3.3.1 System Description 2-93
2.3.3.2 System Performance 2-100
2. 3. A Absorption-Oxidation 2-108
2.3.4.1 System Description 2-108
2.3.4.2 System Performance 2-114
2.3.5 Selective Catalytic Reduction-N0x/S02 Removal 2-115
2.3.5.1 System Description 2-115
2.3.5.2 System Performance. 2-131
2.3.6 Adsorption. . .. 2-131
2.3.6.1 System Description 2-131
2.3.6.2 System Performance 2-134
2.3.7 Electron Beam Radiation 2-134
2.3.7.1 System Description 2-134
2.3.7.2 System Performance 2-137
2.3.8 Absorption-Reduction 2-139
2.3.8.1 System Description 2-139
2.3.8.2 System Performance 2-149
2.3.9 Oxidation-Absorption-Reduction 2-150
2.3.9.1 System Description 2-150
2.3.9.2 System Performance 2-157
2.3.10 Oxidation-Absorption 2-160
2.3.10.1 System Description 2-160
2. 3.10. 2 System Performance 2-164
2. 4 Controls for Natural Gas-Fired Boilers 2-165
2.4.1 Selective Catalytic Reduction-Fixed Packed Bed
Reactor 2-165
2.4.1.1 System Description 2-165
2.4.1.2 System Performance 2-170
2.4.2 Absorption-Oxidation 2-176
2.4.2.1 System Description 2-176
Vll
-------
CONTENTS (Continued)
2.4.2.2 System Performance 2-181
References 2-183
3. CANDIDATES FOR BEST SYSTEMS OF EMISSION REDUCTION 3-1
3.1 Criteria for Selection 3-1
3.1.1 Factors Considered in Selection of Best Systems... 3-1
3.1.1.1 Performance 3-6
3.1.1.2 Operational and Maintenance Impacts 3-6
3.1.1.3 Preliminary Environmental Impacts 3-6
3.1.1.4 Preliminary Economic Impacts 3-7
3.1.1.5 Preliminary Energy/Material Impacts 3-7
3.1.1.6 Boiler Operation and/or Safety 3-7
3.1.1.7 Reliability.. 3-7
3.1.1.8 Development Status 3-8
3.1.1.9 Adaptability to Existing Sources 3-8
3.1.1.10 Compatibility with Other Control Sources. 3-8
3.1.2 Selection of Control Levels—Moderate, Stringent,
and Intermediate 3-9
3.2 Best Control Systems for Coal-Fired Boilers..... 3-12
3.2.1 Moderate Reduction Controls 3-12
3.2.2 Stringent Reduction Controls 3-18
3.2.3 Intermediate Reduction Controls 3-18
3.3 Best Control Systems for Oil-Fired Boilers 3-18
3.3.1 Moderate Reduction Controls 3-25
3.3.2 Stringent Reduction Controls 3-25
3.3.3 Intermediate Reduction Controls 3-25
3.4 Best Control Systems for Gas-Fired Boilers 3-25
3.4.1 Moderate Reduction Controls 3-31
3.4.2 Stringent Reduction Controls 3-31
3.4.3 Intermediate Reduction Controls 3-32
3.5 Summary 3-33
References 3-35
Vlll
-------
CONTENTS (Continued)
4. COST ANALYSIS OF CANDIDATES FOR BEST EMISSION CONTROL SYSTEMS. 4-1
4.1 N0x-0nly Systems 4-1
4.1.1 Introduction 4-1
4.1.2 Control Costs for Coal-Fired Boilers 4-7
4.1.3 Costs to Control Oil-Fired Boilers 4-17
4.1.4 Control Costs for Natural Gas-Fired Boilers 4-29
4.1.5 Summary 4-35
4.2 NOX/SOX System 4-36
4.2.1 Introduction 4-36
4.2.2 Control Costs for Coal-Fired Boilers 4-37
4.2.3 Control Costs for the Oil-Fired Boiler 4-37
References 4-42
5. ENERGY IMPACT 5-1
5 .1 NOx-Only Systems 5-2
5.1.1 Introduction 5-1
5.1.2 Energy Impact of Controls for Coal-Fired Boilers.. 5-3
5.1.3 Energy Impact of Controls for Oil-Fired Boilers... 5-15
5.1.4 Energy Impact of Controls for Natural Gas-Fired
Boilers 5-25
5.2 NOx/SOx Systems 5-29
5.2.1 Introduction 5-29
5.2.2 Energy Impact of NOX/SOX Controls for Coal-Fired
Boilers 5-31
5.2.3 Energy Impact of NOX/SOX Controls for Oil-Fired
Boilers 5-35
5 . 3 Summary 5-37
References 5-38
6. ENVIRONMENTAL IMPACT OF CANDIDATES FOR BEST EMISSION CONTROL
SYSTEMS 6-1
6 .1 Introduction 6-1
6.2 Environmental Impacts of Controls for Coal-Fired Boilers. 6-6
6.2.1 Air Pollution. 6-6
6.2.2 Water Pollution. 6-16
IX
-------
CONTENTS (Continued)
6.2.3 Solid Waste 6-16
6.2.4 Other Environmental Impacts 6-17
6.2.5 Environmental Impact on Modified and Reconstructed
Facilities 6-17
6.3 Environmental Impacts of Controls for Oil-Fired Boilers.. 6-18
6.3.1 Air Pollution 6-18
6.3.2 Water Pollution 6-24
6.3.3 Solid Waste 6-24
6.3.4 Other Environmental Impacts 6-25
6.3.5 Environmental Impacts on Modified and
Reconstructed Facilities 6-25
6.4 Environmental Impacts of Controls for Gas-Fired Boilers.. 6-25
6.4.1 Air Pollution 6-25
6.4.2 Water Pollution 6-27
6.4.3 Solid Waste . 6-27
6.4.4 Other Environmental Impacts 6-28
6.4.5 Environmental Impacts on Modified and
Reconstructed Facilities 6-28
References . 6-29
7. EMISSION SOURCE TEST DATA 7-1
7.1 Introduction 7-1
7.2 Emission Source Test Data for Coal-Fired Boilers 7-5
7.3 Emission Source Test Data for Oil-Fired Boilers 7-10
7.4 Emission Source Test Data for Gas-Fired Boilers 7-22
References 7-27
APPENDIX 1 - DETAILED SYSTEM EVALUATIONS Al-1
APPENDIX 2 - EXAMPLE OF TECHNIQUE FOR ECONOMIC SCALING A2-1
APPENDIX 3 - MATERIAL BALANCES FOR COAL-FIRED BOILERS A3-1
APPENDIX 4 - MATERIAL BALANCES FOR OIL-FIRED BOILERS A4-1
APPENDIX 5 - MATERIAL BALANCES FOR NATURAL GAS-FIRED BOILERS.. A5-1
APPENDIX 6 - CAPITAL COST BREAKDOWNS A6-1
APPENDIX 7 - ANNUAL COST BREAKDOWNS A7-1
APPENDIX 8 - SAMPLE CALCULATIONS A8-l
-------
FIGURES
Numb er Page
1.2.1-1 Flow diagram for typical N0x~only SCR process 1-13
1.2.2-1 Annual cost of parallel flow SCR N0x FGT systems for
coal-fired boilers 1-17
1.2.2-2 Annual cost comparison of NO FGT systems for residual
oil boilers 1-18
1.2.2-3 Annual cost of fixed packed bed SCR N0x FGT systems for
distillate oil boilers * 1-19
1.2.2-4 Annual cost of fixed packed bed SCR NO FGT systems for
natural gas boilers 1-20
1.2.2-5 Comparison of annual costs of NO FGT systems applied to
150 MBtu/hr boilers
1.2.2-6 Capital cost of parallel flow SCR N0x FGT for coal-fired
boilers 1-23
1.2-2-7 Capital cost comparison of NO FGT systems for residual
oil boilers 1-24
1.2.2-8 Capital cost of fixed packed bed SCR NO FGT systems for
distillate oil boilers 1-25
1.2.2-9 Capital cost of fixed packed bed SCR N0x FGT systems for
natural gas boilers
1.2.2-10 Cost effectiveness of parallel flow NO control systems
for coal-fired boilers 1-27
1.2.2-11 Cost effectiveness of FGT systems applied to residual
oil-fired boilers 1-28
1.2.2-12 Cost effectiveness of FGT systems applied to distillate
oil-fired boilers 1-29
1.2.2-13 Cost effectiveness of FGT systems applied to natural gas-
fired boilers 1-30
1.2.3-1 Energy consumption of parallel flow SCR NO .FGT systems
for coal-fired boilers 1-32
1.2.3-2 Energy consumption of NO FGT systems for residual oil
boilers 1-33
1.2.3-3 Energy consumption of fixed packed bed SCR NO FGT systems
for distillate oil boilers 1-34
XI
-------
FIGURES (Continued)
Number
1.2.3-4
1.2.3-5
1.2.3-6
1.2.3-7
1.2.3-8
1.2.4-1
1.2.4-2
1.2.4-3
1.3.2-1
1.3.2-2
1.3.3-1
2.2. 3-1
2.2.3-2
2.2.3-3
2.2.4-1
2.2.4-2
2.2.5-1
2.2.5-2
2.2.5-3
2.2.5-4
2.2.5-5
2.2.5-6
Energy consumption of fixed packed bed SCR NO FGT
systems for natural gas boilers
Energy usage of NO control systems as percent of boiler
heat input . Coal-fired boilers
Energy usage of NOX control systems as percent of boiler
heat input . Residual oil-fired boilers
Energy usage of NO control systems as a function of boiler
heat input. Distillate oil-fired boilers
Energy usage of NO control systems as percent of boiler
heat input . Natural gas-fired boilers •
NH3 emissions from SCR NO FGT systems for coal-fired
boilers
NH3 emissions from SCR N0x FGT systems for oil-fired
boilers
NH3 emissions from SCR NO FGT systems for natural gas-fired
boilers
Annual cost of parallel flow SCR NO /SO FGT for coal-fired
boilers at intermediate level of control. .
Capital cost of parallel flow SCR N0x/S0x FGT for coal-fired
boilers at intermediate level of control
Energy consumption of parallel flow SCR N0x/S0x FGT systems
for coal-fired boilers .
Shapes of parallel flow catalysts
Typical reactor used with parallel flow SCR process
Flow diagram for parallel flow SCR process
Gas/liquid contactor options for Absorption-Oxidation
Processes
Process flow diagram for Absorption-Oxidation Process
The SFGT parallel flow reactor
Flow diagram of the SFGT process
SFGT reactor performance versus acceptance time •
Unconverted N0x as a function of catalyst bed length
M) reduction with NH3 over commercial SFGT acceptor
X
S02 removal efficiency vs. cycles
P_age_
1-35
1-36
1-37
1-38
1-39
1-42
1-43
1-44
1-53
1-54
1-56
2-8
2-9
2-9
2-16
2-17
2-23
2-27
2-29
2-30
2-38
Xll
-------
FIGURES (Continued)
Number Page
2.2.6-1
2.2.7-1
2.2.7-2
2.2.7-3
2.2.8-1
2.2.8-2
2.2.8-3
2.2.8-4
2.2.8-5
2.2.9-1
2.2.10-1
2.3.1-1
2.3.1-2
2.3.1-3
2.3.1-4
2.3.1-5
2.3.2-1
2.3.2-2
2.3.2-3
2.3.2-4
2.3.2-5
2.3.2-6
2.3.2-7
2.3.2-8
2.3.3-1
2.3.3-2
Flow diagram of Foster Wheeler-Bergbau Forschung Dry
Adsorption Process
Process flow diagram for Ebara-JAERI electron beam process..
Oil-fired pilot plant test results
Effect of pollutant concentration on removal efficiency
Perforated plate absorber option for Absorption-Reduction
Processes
Normal operation of sieve plate
Other gas dispersers
Process flow diagram of Dureha absorption-reduction process.
EDTA-Fe(II) concentration and NO absorption at 50°C
Process flow diagram for MHI oxidation-absorption-reduction
process
Flow diagram of Kawasaki Heavy Industries process
Example of typical fixed packed bed reactor
Example of catalyst support plate
Process layout for fixed bed SCR process
Performance of experimental calyst of Sumitomo Chemical
Typical example of operation data (oil-fired boiler, 350-
400°C, granular or honeycomb catalyst)
Moving bed reactors of three process vendors
Process flow diagram for moving bed SCR process
SV vs. NO removal and NH3 leak (ring type catalyst, 15 mm
diameter, 350°C NH3/NO 1.0, inlet N0v 250 ppm)
X
Relation between boiler load and denitrification efficiency
(one example)
NH3/NO mole ratio vs. denitrification efficiency and
reactor outlet ammonia concentration
SV value and denitrification efficiency (for small, <1 mm,
diameter particles)
Relationship of NH3/NO ratio to outlet NO , NH3 concen-
trations . ,
At 300°C
Shapes of parallel flow catalysts
Typical reactor used with parallel flow SCR process .........
2-39
2-42
2-44
2-45
2-46
2-47
2-47
2-49
2-50
2-57
2-65
2-70
2-70
2-71
2-77
2-77
2-82
2-83
2-89
2-90
2-91
2-92
2-92
2-92
2-94
2-95
Xlll
-------
FIGURES (Continued)
Number
2.3.3-3
2.3.3-4
2.3.3-5
2.3.3-6
2.3.3-7
2.3.3-8
2.3.3-9
2.3.4-1
2.3.4-2
2.3.5-1
2.3.5-2
2.3.5-3
2.3.5-4
2.3.5-5
2.3.6-1
2.3.7-1
2.3.7-2
2.3.7-3
2.3.8-1
2.3.8-2
2.3.8-3
2.3.8-4
2.3.8-5
2.3.9-1
2.3.9-2
Flow diagram for parallel flow SCR process
Catalyst life test results
Durability of NO removal catalyst for exhaust gas of
high sulfur oil burning boiler
Typical example of operation data (oil-fired boiler, 350-
400°C, granular or honeycomb catalyst)
NHs/NO mole ratio vs. N0x removal (plate calyst; 350°C,
LV 5 . 9 m/sec)
NO removal efficiency, NO concentration, and pressure
loss over 2,000 hr test period for JGC Paranox Process
NH3/NO mole ratio and denitrif ication efficiency and
reactor outlet ammonia concentration
Gas/liquid contactor options for Absorption-Oxidation
Processes
Process flow diagram for absorption-oxidation process
The SFGT parallel flow reactor
Flow diagram of the SFGT process
SFGT reactor performance versus acceptance time
Unconverted NO as a function of catalyst bed length
X
NO reduction with NH3 over commercial SFGT acceptor
Flow diagram of Foster-Wheeler-Bergbau Forschung Dry
Adsorption Process
Process flow diagram for Ebara-JAERI electron beam process.
Oil-fired pilot plant test results
Effect of pollutant concentration on removal efficiency....
Perforated plate absorber option for Absorption-Reduction
Processes
Normal operation of sieve plate
Other gas dispersers
Process flow diagram of Dureha absorption-reduction process
EDTA-Fe(II) concentration and NO absorption at 50°C.
Process flow diagram for MHI oxidation-absorption-
reduction process
Effect of CaCl£ and NaCl concentration on NO removal
efficiency
Page
2-96
2-101
2-102
2-103
2-104
2-105
2-106
2-109
2-110
2-116
2-118
2-121
2-123
2-123
2-132
2-135
2-137
2-138
2-140
2-141
2-141
2-143
2-144
2-151
2-158
XIV
-------
FIGURES (Continued)
Number
2.3.9-3
2.3.10-1
2.4.1-1
2.4.1-2
2.4.1-3
2.4.1-4
2.4.1-5
2.4.1-6
2.4.1-7
2.4.1-8
2.4.2-1
2.4.2-2
4.1.2-1
4.1.2-2
4.1.2-3
4.1.2-4
4.1.2-5
4.1.3-1
4.1.3-2
4.1.3-3
4.1.3-4
4.1.3-5
Effect of pH on SO and NO removal efficiency
X X
Flow diagram of Kawasaki Heavy Industries process
Example of typical fixed packed • bed reactor
Example of catalyst support plate
Process layout for fixed bed SCR process
Test results at gas- and oil-fired boilers
Characteristic curve of the effect of mol ratio of NH3:NO
on NO removal efficiency for Hitachi, Ltd. Process
X
Performance of catalyst MTC-102 (flue gas by LPG burning)..
SV and NO removal (MTC-102) (flue gas by LPG burning)
Relationship among inlet NH3:NO mol ratio, NO removal
efficiency, and exiting NH3 concentration using the Sumi-
tomo Chemical C-l Catalyst
Gas /liquid contactor options for Absorption-Oxidation
Processes
Process flow diagram for absorption-oxidation process
Annual cost of NO control systems applied to underfeed
stoker standard boiler
Annual cost of NO control systems applied to chaingrate
standard boiler
Annual cost of NO control systems applied to spreader
stoker standard boiler
Annual cost of NO control systems applied to pulverized
coal standard- boiler
Cost effectiveness of parallel flow SCR NO control systems
applied to the coal-fired standard boilers
Annual cost of NO control system applied to 4.4 MW distil-
late oil-fired standard boiler
Annual cost of NO control system applied to 44 MW distil-
late oil-fired standard boiler
Annual cost of NO control systems applied to 8.8 MW
residual oil-fired standard boiler.
Annual cost of NO control systems applied to 44 MW
residual oil-fired standard boiler
Cost effectiveness of FGT systems applied to distillate
oil-fired boilers
Page
2-151
2-161
2-166
2-166
2-167
2-172
2-173
2-174
2-174
2-175
2-177
2-178
4-11
4-12
4-13
4-14
4-16
4-21
4-22
4-23
4-24
4-26
XV
-------
FIGURES (Continued)
Number
4.1.3-6
4.1.4-1
4.1.4-2
4.1.4-3
4.2.2-1
5.1.2-1
5.1.2-2
5.1.2-3
5.1.2-4
5.1.2-5
5.1.3-1
5.1.3-2
5.1.3-3
5.1.3-4
5.1.4-1
5.1.4-2
5.2.2-1
6.1-1
6.1-2
6.1-3
Cost effectiveness of FGT systems applied to residual
oil-fired boilers
Annual cost of NO control system applied to 4.4 MW
natural gas-fired standard boiler
Annual cost of N0x control system applied to 44 MW
natural gas-fired standard boiler
Cost effectiveness of FGT systems applied to natural
gas-fired boilers
Annual cost of parallel flow SCR N0x/S0x FGT for coal-
fired boilers
Energy usage of NO control systems for pulverized coal
standard boiler
Energy usage of N0x control systems for spreader stoker
standard boiler
Energy usage of NO control systems for chaingrate standard
boiler
Energy usage of NO control systems for underfeed stoker
standard boiler
Energy usage of NO control systems as percent of boiler
heat input
Energy usage of NO control systems for residual oil-fired
standard boilers
Energy usage of NO control systems for distillate oil-
fired boilers
Energy usage of NO control systems applied to residual
oil-fired boilers as percent of boiler heat input
Energy usage of N0x control systems applied to distillate
oil-fired boilers as percent of boiler heat input
Energy usage of N0x control systems for natural gas-fired
standard boiler
Energy usage of N0x control systems as percent of boiler
heat input -
Energy consumption of parallel flow SCR NO /SO FGT systems
for coal-fired boilers
NH3 emissions - Fixed Packed Bed Reactor
NH3 emissions - Parallel Flow Reactor
NHa emissions - Moving Bed Reactor
Page
4-27
4-31
4-32
4-34
4-39
5-10
5-11
5-12
5-13
5-14
5-20
5-21
5-22
5-23
5-27
5-28
5-34
6-3
6-4
6-5
XVI
-------
FIGURES (Continued)
Number Page
7.1-1 Sampling train, flask valve, and flask 7-3
7.1-2 Measurement system design for stationary gas turbine
tests 7-5
7.2-1 Change of NO removal efficiency and pressure drop
(Kawasaki Heavy Industries process, Takehara power
station, Hiroshima, Japan) 7-6
7.2-2 Pilot plant test of a parallel flow reactor treating a flue
gas from a coal-fired utility boiler (Hitachi, Ltd. process,
unknown location, Japan) 7-7
7.2-3 Pilot plant test of an intermittent-moving bed reactor
treating a flue gas from a coal-fired utility boiler
(Hitachi, Ltd. process, unknown location, Japan) 7-8
7.2-4 Durability test of NO removal catalyst (Kawasaki Heavy
Ind. process, Takehara power station, Hiroshima, Japan).... 7-9
7.3-1 Catalyst life test results (IHI process, Taketoyo power
station, Japan) 7-11
7.3-2 Pilot plant test of a prallel flow reactor treating a flue
gas from a high sulfur heavy oil-fired utility boiler
(Hitachi, Ltd. process, unknown location, Japan) 7-12
7.3-3 Test results of oil-fired boiler (Hitachi, Ltd process,
unknown location, Japan) 7-13
7.3-4 NO removal for the month of May 1977 (Hitachi Zosen
fixed bed process, Shindaikyowa Petrochemical, Yokkaichi,
Japan, chemiluminescence method) 7-19
7.3-5 NO removal for August, 1978 (MHI process, Fuji Oil,
Sodegaura, Japan, PDS/chemiluminescence method) 7-20
7.4-1 Test results of gas-fired boiler (Hitachi, Ltd. process,
unknown location, Japan) 7-23
7.4-2 Characteristic curve of the effect of mole ratio of NH3:
NO on NO removal efficiency for Hitachi, Ltd. process.... 7-24
X X
7.4-3 Performance of catalyst MTC-102 (Mitsui Toatsu process,
unknown location, Japan) 7-25
7.4-4 SV and NO removal (MTC-102) (Mitsui Toatsu process,
unknown location, Japan) 7-25
7.4-5 Relationship among inlet NH3:NOx mol ratio, NO removal
efficiency, and exiting NH3 concentration using the
Sumitomo Chemical C-l Catalyst 7-26
xvii
-------
TABLES
Number Page
1.1.3-1 Characteristics of the Standard Boilers 1-6
1.1.3-2 NOV Emission Rates for the Standard Boilers 1-7
X
1.1.3-3 NO Control Levels <•.. 1-9
X
1.1.3-4 Cases Selected for Detailed Analysis - N0x-0nly FGT
Processes 1-10
1.1.3-5 Cases Selected for Detailed Analysis - S0x/N0x FGT
Processes 1-10
1.2.1-1 Candidates for Best Emission Control System 1-12
1.2.2-1 Annual Cost of NOX Control Systems Applied to Coal-Fired
Boilers 1-16
1.2.2-2 Annual Cost of N0x Control Systems Applied to Oil
Fired-Boilers 1-16
1.2.2-3 Annual Cost of N0x Control Systems Applied to Natural
Gas-Fired Boilers 1-16
1.2.3-1 Areas of Energy Consumption in NOX FGT Systems 1-31
1.2.5-1 Planned FGT Installations of SCR Coal-Fired Utility
Boilers 1-45
1.2.5-2 Existing FGT Installations of SCR Parallel Flow
Systems Oil-Fired Industrial Boilers 1-46
1.2.5-3 Existing FGT Installations of SCR Parallel Flow
Systems Oil-Fired Utility Boilers 1-46
1.2.5-4 Existing FGT Installations of SCR Moving Bed Systems
Oil-Fired Industrial Boilers 1-47
1.2.5-5 Existing FGT Installations of SCR Fixed Bed Systems
Oil-Fired Industrial Boilers 1-48
1.2.5-6 Gas-Fired SCR Plants in Japan 1-49
1.3.1-1 Best NOX/SOX Emission Control System for Coal-Fired
Boilers 1-50
1.3.2-1 Annual Cost of Parallel Flow N0x/S0x Control Systems 1-52
1.3.2-2 Costs of Parallel Flow N0x/S0x Control System |. . . 1-52
XVlll
-------
TABLES (Continued)
Number Page
1.3.3-1 Energy Consumption of NO /SO Control Processes
Applied to Coal Fired Boilers 1-55
1.3.3-2 Energy Consumption of Parallel Flow N0x/S0x Control
System 1-55
1.3.5-1 Shell/UOP Process, Pilot and Demonstration Unit 1-58
1.3.5-2 Shell/UOP Process Commercial Applications 1-59
2.1-1 Characteristics of the Standard Boiler Considered
for Analysis in this Report 2-3
2.1-2 NO Emission Rates for the Standard Boilers 2-5
2.2.3-1 Reaction Rate Data for Two Catalyst Formulations 2-12
2.2.3-2 Catalyst Design Variables for Various Catalyst Shapes ..... 2-12
2.2.3-3 Planned FGT Installations of SCR Coal-Fired
Utility Boilers 2-14
2.2.3-4 Process Vendors of Parallel Flow SCR Processes 2-14
2.2. 4-1 Nitrogen Oxides Characteristics 2-19
2.2.4-2 System Design Considerations 2-20
2.2.4-3 Process Vendors of Absorption-Oxidation Processes 2-21
2.2.5-1 Design and Operating Variables for SFGT System 2-30
2.2.5-2 SFGT Process, Pilot and Demonstration Units 2-32
2.2.5-3 SFGT Process, Commercial Units 2-33
2.2.5-4 Economics of SFGT System 2-34
2.2.5-5 Economics of"SFGT System Estimated Chemicals and
Utility Requirements 2-35
2.2.5-6 Economics of SFGT System Estimated Capital and
Operating Cost 2-36
2.2.5-7 Summary of Base Operating Conditions on the SFGT Pilot
Plant at TECO 2-38
2.2.7-1 System Variables 2-43
2.2.8-1 System Design Considerations 2-53
2.2.8-2 Typical Values for Process Variables of Absorption-
Reduction Processes 2-53
2.2.8-3 Process Vendors of Absorption-Reduction Processes 2-54
xix
-------
TABLES (Continued)
Number Page
2.2.9-1 System Design Considerations 2-62
2.2.9-2 Typical Ranges of Operating Variables for
Oxidation-Absorption-Reduction Processes 2-62
2.2.9-3 Process Vendors of Oxidation-Absorption-Reduction
Processes 2-63
2.2.10-1 Process Vendors of Oxidation-Absorption Processes 2-68
2.3.1-1 Reaction Rate Data for Two Catalyst Formulations 2-74
2.3.1-2 Design and Operating Variables for Fixed Packed
Bed Systems 2-74
2.3.1-3 Vendors of SCR Fixed Bed Systems for Oil-Fired
Applications 2-75
2.3.1-4 Existing FGT Installations of SCR Fixed Bed Systems
Oil-Fired Industrial Boilers 2-76
2.3.1-5 Operation Parameters of Major Plants Constructed by
Hitachi Zosen 2-78
2.3.1-6 SCR Plant by Mitsui Engineering & Shipbuilding Co. 2-79
2.3.1-7 Operation Data of SCR Plants for Dirty Gas 2-80
2.3.2-1 Design and Operating Variables for Moving Bed
SCR Systems 2-85
2.3.2-2 Vendors of SCR Moving Bed Systems for Oil-Fired
Applications 2-88
2.3.2-3 Existing FGT Installations of SCR Moving Bed Systems
Oil-Fired Industrial Boilers 2-88
2.3.2-4 Operation Data of a Commercial SCR Plant for Dirty Gas 2-93
2.3.3-1 Catalyst Design Variables for Various Catalyst Shapes 2-98
2.3.3-2 Vendors of SCR Parallel Flow Systems for Oil-Fired
Applications 2-99
2.3.3-3 Existing FGT Installations of SCR Parallel Flow
Sy-tems Oil-Fired Industrial Boilers 2-99
2.3.3-4 Existing FGT Installations of SCR Parallel Flow
Systems Oil-Fired Utility Boilers 2-100
2.3.3-5 SCR Plant by Mitsui Engineering and Shipbuilding Co 2-107
2.3.4-1 Nitrogen Oxides Characteristics 2-112
xx
-------
TABLES (Continued)
Number Page
2. 3.4-2 System Design Considerations 2-113
2.3.4-3 Process Vendors of Absorption-Oxidation Processes 2-114
2.3.5-1 Design and Operating Variables for SFGT System 2-124
2.3.5-2 SFGT Process, Pilot and Demonstration Units 2-126
2.3.5-3 SFGT Process, Commercial Units 2-127
2.3.5-4 Economics of SFGT System 2-128
2.3.5-5 Economics of SFGT System Estimated Chemicals and
Utility Requirements 2-129
2.3.5-6 Economics of SFGT System Estimated Capital and
Operating Cost 2-130
2.3.7-1 System Variables 2-136
2.3.8-1 System Design Considerations 2-147
2.3.8-2 Typical Values for Process Variables of Absorption-
Reduction Processes 2-147
2.3.8-3 Process Vendors of Absorption-Reduction Processes 2-148
2.3.9-1 System Design Considerations
2-156
2.3.9-2 Typical Ranges of Operating Variables for Oxidation-
Absorption-Reduction Processes 2-156
2.3.9-3 Process Vendors of Oxidation Absorption-Reduction
Processes 2-157
2.3.10-1 Process Vendors of Oxidation—Absorption Processes 2-164
2.4.1-1 Reaction Rate Data for Two Catalyst Formulations 2-169
2.4.1-2 Design and Operating Variables for Fixed Packed
Bed Systems 2-169
2.4.1-3 Vendors of SCR Fixed Bed Systems for Gas-Fired
Applications 2-171
2.4.1-4 Existing FGT Installations of SCR Fixed Bed Systems
Gas-Fired Industrial Boilers 2-171
2.4.1-5 Existing FGT Installations of SCR Fixed Bed Systems
Gas-Fired Utility Boilers 2-171
2.4.2-1 Nitrogen Oxides Characteristics 2-180
2.4.2-2 System Design Considerations 2-180
2.4.2-3 Process Vendors of Absorption-Oxidation Processes 2-181
xxi
-------
TABLES (Continued)
Number Page
3.1.1-1 Rating Criteria and Weighting Factors 3-2
3.1.1-2 Specific Point Values Associated with Selection
Factors 3-3
3.1.2-1 Controlled Emission Levels in This Study 3-11
3.2-1 Comparison Information of N0x~0nly Systems for Coal-
Fired Boilers 3-13
3.2-2 Comparison Information of Simultaneous NOX/SOX Systems
for Coal-Fired Boilers 3-15
3.2.1-1 Candidate Systems Selection: Coal-Fired Boilers -
Moderate Control 3-17
3.2.2-1 Candidate Systems Selection: Coal-Fired Boilers -
Stringent Control 3-19
3.2.3-1 Candidate Systems Selection: Coal-Fired Boilers -
Intermediate Control 3-20
3.3-1 Comparison Information of NO -Only Systems for
Oil-Fired Boilers * 3-21
3.3-2 Comparison Information of Simultaneous NOX/SO Systems
or Oil-Fired Boilers 3-23
3.3.1-1 Candidate Systems Selection: Oil-Fired Boilers -
Moderate Control 3-26
3.3.2-1 Candidate Systems Selection: Oil-Fired Boilers -
Stringent Control 3-27
3.3.3-1 Candidate Systems Selection: Oil-Fired Boilers -
Intermediate Control 3-28
3.4-1 Comparison Information of N0x~0nly Systems for
Gas-Fired Boilers 3-29
3.4.1-1 Candidate Systems Selection: Gas-Fired Boilers -
Moderate Control 3-31
3.4.2-1 Candidate Systems Selection: Gas-Fired Boilers -
Stringent Control 3-32
3.4.3-1 Candidate Systems Selection: Gas-Fired Boilers -
Intermediate Control 3-32
3.5-1 Summary of Candidate Systems: All Levels of Control 3-33
3.5-2 Major Performance Characteristics of Candidate
Systems 3-34
4.1.1-1 Purchased Equipment for NO FGT Systems 4_2
xxii
-------
TABLES (Continued)
Number Page
4.1.1-3 Annual Cost Factors 4-3
4.1.1-4 Load Factors for the Standard Boilers 4-3
4.1.1-5 Sources of Costs for Specific Equipment Items 4-4
4.1.1-6 Capital Cost Factors 4-4
4.1.1-7 Chemical Engineering Cost Indices 4-5
4.1.2-1 Costs of NOX FGT Control Techniques for Coal-Fired
Boilers 4-8
4.1.2-2 Costs of N0x FGT Control Techniques for Coal-Fired
Boilers 4-9
4.1.2-3 Costs of NOX FGT Control Techniques for Coal-Fired
Boilers 4-10
4.1.2-4 Costs of NOX FGT Control Techniques for Coal-Fired
Boilers " 4-10
4.1.2-5 Cost Effectiveness of NO FGT 4-15
X
4.1.2-6 Relative Costs of Retrofit SCR Systems 4-17
4.1.3-1 Costs of N0x FGT Control Techniques for Oil-Fired
Boilers 4-19
4.1.3-2 Costs of NOX FGT Control Techniques for Oil-Fired
Boilers 4-20
4.1.3-3 Cost Effectiveness of N0v FGT 4-25
X
4.1.3-4 Relative Costs of Retrofit SCR Systems 4-28
4.1.4-1 Costs of NOX FGT Control Techniques for Natural
Gas-Fired Boilers 4-30
4.1.4-2 Cost Effectiveness of NO^ FGT 4-33
X
4.1.4-3 Relative Costs of Retrofit SCR Systems 4-35
4.2.1-1 Purchased Equipment for NOX FGT Systems 4-36
4.2.2-1 Costs of N0x/S0x FGT Control Techniques for Coal-Fired
Boilers 4-38
4.2.3-1 Costs of the Dry N0x/S0x Control Technique for the
Residual Oil-Fired Boiler 4-41
5.1.1-1 Areas of Energy Consumption in NO FGT Systems 5-2
X
5.1.1-2 Range of Design Parameters Used for Energy Impact
Calculations 5-2
5.1.2-2 SIP Control Levels 5-4
5.1.2-1 Relative Significance of Parameters Considered in
Energy Analysis 5-5
XXlll
-------
TABLES (Continued)
Number
5.1.2-3 Energy Consumption for NO FGT Control Techniques
for Coal-Fired Boilers 5-6
5.1.2-4 Energy Consumption for N0x FGT Control Techniques
for Coal-Fired Boilers 5-6
5.1.2-5 Energy Consumption for NOX FGT Control Techniques
for Coal-Fired Boilers 5-7
5.1.2-6 Energy Consumption for N0x FGT Control Techniques
for Coal-Fired Boilers 5-7
5.1.2-7 Summary of Energy Requirements for Coal-Fired
Industrial Boilers -. 5-9
5.1.3-1 Energy Consumption for NOX FGT Control Techniques
for Residual Oil-Fired Boilers 5-17
5.1.3-2 Energy Consumption for NOX FGT Control Techniques
for Distillate Oil-Fired Boilers 5-18
5.1.3-3 Summary of Energy Requirements for Oil-Fired
Industrial Boilers 5-19
5.1.4-1 Energy Consumption for NO FGT Control Techniques
for Natural Gas-Fired Boilers 5-26
5.1.4-2 Summary of Energy Requirements for Natural
Gas-Fired Boilers 5-26
5.2.1-1 NOX/SOX FGT/Boiler Combinations Analyzed for
Energy Impact I. 5-30
5.2.1-2 Areas of Energy Utilization in the NO /SO FGT System 5-30
X X
5.2.2-3 Summary of Enersy Usage of NOX/SOX Systems
Applied to Coal-Fired Boilers 5-31
5.2.2-1 Energy Consumption for N0x/S0x FGT Control Techniques
for Coal-Fired Boilers ..* 5-32
5.2.2-2 Energy Consumption for N0x/S0x FGT Control Techniques
For Coal-Fired Boilers ,. 5-33
5.2.3-1 Energy Consumption for N0x/S0x FGT Control Techniques
f r r Oil-Fired Boilers 5-36
6.2.1-1 Aii Pollution Impacts from Best NOX FGT Control
Techniques for Coal-Fired Boilers 6-7
6.2.1-2 Air Pollution Impacts from Best N0x FGT Control
Techniques for Coal-Fired Boilers 6_y
6.2.1-3 Air Pollution Impacts from Best N0x FGT Control
Techniques for Coal-Fired Boilers 5_g
xxiv
-------
TABLES (Continued)
Number page
6.2.1-4 Air Pollution Impacts from Best NO FGT
Control Techniques for Coal-Fired Boilers 6-8
6.2.1-5 Air Pollution Impacts from Best NO FGT
Control Techniques for Coal-Fired Boilers 6-9
6.2.1-6 Air Pollution Impacts from Best NO FGT
Control Techniques for Coal-Fired Boilers 6-9
6.2.1-7 Air Pollution Impacts from Best NO FGT
Control Techniques for Coal-Fired Boilers 6-10
6.2.1-8 Air Pollution Impacts from Best NO FGT
Control Techniques for Coal-Fired Boilers 6-10
6.2.1-9 Air Pollution Impacts from Best NO /SO
Control Techniques for Coal-Fired Boilers 6-11
6.2.1-10 Air Pollution Impacts from Best NOX/SOX
Control Techniques for Coal-Fired Boilers 6-11
6.2.1-11 Air Pollution Impacts from Best NO /SO
FGT Control Techniques for Coal-Fired Boilers 6-12
6.2.1-12 Air Pollution Impacts from Best N0x/S0x
FGT Control Techniques for Coal-Fired Boilers 6-12
6.2.1-13 N0x Emission Levels and SIP Control Levels 6-14
6.3.1-1 Air Pollution Impacts from Best N0x FGT
Control Techniques for Oil-Fired Boilers 6-19
6.3.1-2 Air Pollution Impacts from Best N0x/S0x
FGT Control Techniques for Oil-Fired Boilers 6-20
6.4.1-1 Air Pollution Impacts from Best N0x FGT
Control Techniques for Gas-Fired Boilers 6-26
7.3-1 Operation Parameters of Major Plants Constructed
by Hitachi Zosen 7-14
7.3-2 SCR Plants by Mitsui Engineering & Shipbuilding Co 7-15
7.3-3 Operation Data of SCR Plants for Dirty Gas 7-16
7.3-4 Oil-Fired Industrial SCR Plants 7-17
7.3-5 Oil-Fired Utility SCR Plants 7-18
7.3-6 NO Removal Levels at Several Japanese Industrial
Boxlers with N0x Control by SCR 7-21
7.4-1 Gas-Fired SCR Plants 7-22
XXV
-------
ACKNOWLEDGEMENTS
This report would not have been possible without the assistance of
several people. The authors would like to express their appreciation to
the following people for their support in the preparation of this report.
The process vendors who supplied much of the data used for
the analyses.
Dr. Jumpei Ando for supplying information concerning NO flue
gas treatment systems in Japan.
C. B. Sedman and L. D. Broz for their coordination efforts
throughout the program.
• J. D. Mobley for his guidance and assistance.
M. Harris, J. C. Fischer, and C. K. Holcomb for their work
in typing this report.
xxvi
-------
SECTION 1
EXECUTIVE SUMMARY
1.1 INTRODUCTION
1.1.1 Background and Objectives
The Clean Air Act Amendments of 1977 require the Environmental
Protection Agency to coordinate and lead the development and implementation
of regulations on air pollution. These include standards of performance
for new and modified sources of pollution. Fossil fired steam generators
are specifically mentioned in the act and EPA has undertaken a study of
industrial boilers with intent to propose emission control levels based upon
the results of this and other studies.
This specific report examines the impacts of application of flue gas
treatment (FGT) for NOx control on industrial boilers. The overall objective
is to provide a background document that quantifies the economic, energy and
environmental impacts as well as establish whether or not the technology
is demonstrated and available to the U.S. market. All potential FGT tech-
nologies are considered and detailed analyses are performed on those which
are the most promising.
1.1.2 Report Organization and Approach
Several boiler/FGT combinations are considered in the detailed analyses
that follow. In Section 2, all NOx control processes that have been devel-
oped to treat boiler flue gas are discussed in moderate detail. The section
1-1
-------
is divided into three subsections based on fuel—coal, oil, and gas. This
is done for two reasons: 1) to make this report consistent in format with
other Individual Technology Assessment Reports (ITAR's, see preface), and
2) to examine the effect of fuel type on the various technologies considered.
In the case of FGT, the majority of the technologies can be applied to the
majority of the fuels. As a result, much of the material in the three sub-
sections is very similar, especially with regard to the technical descrip-
tions of the systems.
A decision was made early in the development of this report to produce
essentially three stand-alone sections; one for each fuel type. This allows
one or more fuel types to be eliminated from consideration without impacting
the quality of the data in the remaining sections. As a result, there is a
significant amount of repetition in the three subsections. The subsection
dealing with applications to oil-fired boilers (Section 2.3) contains
descriptions of all of the FGT technologies considered and has the greatest
amount of information on specific systems. Therefore, for most readers, it
is necessary to read only subsections 2.1 and 2.3 for a complete description
of all FGT technologies considered. Subsections 2.2 and 2.4, dealing with
coal- and gas-fired applications, do contain unique information on the status
of development and number of applications and can be consulted if this
specific information is desired. The Executive Summary is organized dif-
ferently than the body of the report in that all fuel types are discussed
together instead of separately. This is done to allow the reader to
directly examine the effect of fuel type on the economic, energy, and en-
vironmental impacts. The summary discusses each of the impacts separately
and also separates NO -only systems from NO /SO systems.
X XX
The large number of potential fuel/boiler/FGT system combinations
requires the combinations be reduced to those systems with a high potential
for commercial application and successful operation. This is done in
Section 3. The data used to make these preliminary evaluations is derived
from Section 2. The combinations selected in Section 3 are then analyzed
in detail in subsequent sections.
1-2
-------
Section 4 presents the economic impacts of these FGT processes on ten
industrial boilers. Standard costing techniques are used to develop annual-
ized costs which are plotted to show the effect of several parameters on the
total system costs. The process specifications used in the economic analyses
are developed in Sections 5 and 6. These two Sections present the results of
material and energy balances performed for each case to quantify the energy
and environmental impacts of FGT systems. These balances are also used to
size most of the individual pieces of equipment.
In Sections 4, 5 and 6 results are not presented for all possible
control levels. That is, while some systems have data presented for three
levels of control, others have data presented for only two or one control
level(s). The curves, however, are plotted over the range of the three
control levels (70, 80 and 90 percent). When the individual Sections were
initially prepared, data was calculated for all three control levels.
During the interim period prior to the compilation of these Sections into
the final report, new cases were added and several economic premises were
changed. In order to meet budget and time constraints, it was necessary to
reduce the number of analyses.
It was observed that curve shapes were all very similar and that new
curves could be drawn accurately without a complete set of points. In
cases with a lot of similarity, i.e. among the coal-fired boilers, a curve
shape was established for one case by using three points. For the other
cases, an analysis was made to determine the midpoint of the curve. Curves,
similar in shape to that developed by a three-point analysis, were then
passed through these midpoints. In cases where there was little similarity,
a two-point analysis was performed to determine the end-points of the curve.
A curve was then drawn through these points using the original curve
(determined by three-point analysis in the initial case analysis) to
determine the shape. It should be noted that, even if straight lines were
used, the interpolated and extrapolated results would not be changed
1-3
-------
significantly. This is why a limited set of analyses are used to determine
a complete set of cost and energy data.
The final section, Section 7, deals with test data that have been
determined for operating FGT systems. These types of data do not exist
for U.S. applications since FGT has yet to be applied in this country.
A Japanese consultant with contacts among FGT system users was retained
to obtain test data from industrial boilers in Japan. The test data pre-
sented in Section 7 represents the most complete set of data of this type
available.
For the reader interested in the details involved with the analyses
presented in this report, the Appendices present an example calculation as
well as complete sets of process selection criteria material and energy
balances, and cost breakdowns.
1.1.3 Scope of Study
Several variables are considered in order to make the study as compre-
hensive as possible, these being
Fuel
Boiler, type and size
• Control level
• FGT process type
Pollutants controlled.
As mentioned previously, three fuel types are considered: coal, oil and
natural gas. Coal and oil are further divided as shown below.
1-4
-------
Coal
High Sulfur Eastern (3.5% S)
Low Sulfur Eastern (0.9% S)
Low Sulfur Western (0.6% S)
Oil
Distillate
Residual
One boiler type is considered for natural gas, distillate oil and residual
oil. However, four boiler types are considered for the coal fuels. The
combinations of fuel and boiler type considered at the beginning of the
study are shown in Table 1.1.3-1. These boilers are termed "standard
boilers" because they apply to all of the ITAR's. The NOX emissions from
these boilers are shown in Table 1.1.3-2.
In the ensuing discussion of emission control technologies, candidate
technologies were compared using three emission control levels labeled
"moderate, intermediate, and stringent." These control levels were chosen
only to encompass all candidate technologies and form bases for comparison
of technologies for control of specific pollutants considering performance,
costs, energy, and non-air environmental effects.
From these comparisons, candidate "best" technologies for control of
individual pollutants are recommended by the contractor for consideration
in subsequent industrial boiler studies. These "best technology" recommenda-
tions do not consider combinations of technologies to remove all pollutants
and have not undergone the detailed environmental, cost, and energy impact
assessments necessary for regulatory action. Therefore, the levels of
"moderate, intermediate, and stringent" and the recommendation of "best
technology" for individual pollutants are not to be construed as indicative
of the regulations that will be developed for industrial boilers. EPA will
1-5
-------
TABLE 1.1.3-1. CHARACTERISTICS OF THE STANDARD BOILERS
Type
Package, Firetube
Package, Firetube
Package, Watertube
Package, Watertube
Underfeed Stoker
Package, Watertube
Chaingrate Stoker
Package, Watertube
Package Watertube
Package, Watertube
Field Erected, Watertube
Spreader Stoker
Field Erected, Watertube
Pulverized Coal
Fuel*
Distillate Oil
Natural Gas
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
Natural Gas
Distillate Oil
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
Rating
MWt(MBtu/hr)
4.4
4.4
8.8
8.8
8.8
8.8
22
22
22
44
44
44
44
44
44
58.6
58.6
58.6
(15)
(15)
(30)
(30)
(30)
(30)
(75)
(75)
(75)
(150)
(150)
(150)
(150)
(150)
(150)
(200)
(200)
(200)
Gas Flow Rate
Nm3/hr
5,400
5,600
9,500
12,500
12,600
12,900
31,300
31,000
32,400
52,800
51,900
47,800
62,900
62,700
64,700
72,600
72,800
75,500
*Coal types: HSE = High Sulfur Eastern (3.5% S)
LSE = Low Sulfur Eastern (0.9% S)
LSW = Low Sulfur Western (0.6% S)
1-6
-------
TABLE 1.1.3-2. NO^ EMISSION RATES FOR THE STANDARD BOILERS
NOX Emissions
Boiler
Package, Firetube
Package, Firetube
Package, Watertube
Package, Watertube
Underfeed Stoker
Package, Watertube
Chaingrate
Package, Watertube
Package, Watertube
Package, Watertube
Field Erected, Watertube
Spreader Stoker
Field Erected, Watertube
Pulverized Coal
Fuel*
Distillate Oil
Natural Gas
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
Natural Gas
Distillate Oil
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
g/s
0.300
0.332
2.02
2.40
2.06
2.95
6.02
5.15
7.40
3.31
2.99
7.47
12.0
10.3
14.8
19.2
16.5
23.7
(Ib/hr)
(2.38)
(2.63)
(16.0)
(10.05)
(16.35)
(23.40)
(47.70)
(40.80)
(58.65)
(26.26)
(23.76)
(60.00)
(95.40)
(81.45)
(117.15)
(152.46)
(130.50)
(187.56)
ng/J
68.8
77.4
228
275
237
335
275
232
335
75.3
68.0
172
275
232
335
327
280
404
(lb/10e Btu)
(0.16)
(0.18)
(0.53)
(0.64)
(0.55)
(0.78)
(0.64)
(0.54)
(0.78)
(0.18)
(0.16)
(0.40)
(0.64)
(0.54)
(0.78)
(0.76)
(0.65)
(0.94)
ppm
97
104
373
335
288
402
336
290
401
110
101
292
337
288
400
466
396
550
-'Coal types:
HSE = High Sulfur Eastern (3.5% S)
LSE = Low Sulfur Eastern (0.9% S)
LSW = Low Sulfur Western (0.6% S)
-------
perform rigorous examination of several comprehensive regulatory options
before any decisions are made regarding the standards for emission from
industrial boilers. The' control levels are defined in Table 1.1.1-3.
The types of FGT systems considered are different for each fuel type
and these are discussed in subsequent sections for each specific fuel.
The project schedule required that the number of potential combinations of
boiler, fuel, and control level be reduced in order to keep the number of
required analyses manageable. Detailed analyses were performed on the
cases shown in Table 1.1.3-4. Note that these are the cases for FGT pro-
cesses which remove only NO . For FGT processes which remove both NO and
X X
SO a separate set of cases was developed and is shown in Table 1.1.3-5.
Only one coal is considered for the NO -only cases. This is due to
the fact that FGT process designs and impacts are not significantly affected
by fuel sulfur content and therefore analyzing each coal type would not
yield any additional information. The flue gas flow rates and NO concen-
X
trations vary somewhat among the coal types considered, but not enough to
cause much difference in the size of the necessary FGT process. With all
of the FGT systems analyzed, the equipment size is primarily a function of
the flue gas flow rate and secondarily a function of NO concentration.
However, since coal sulfur level can affect the environmental impact, two
coal types are considered in this section.
For the processes which remove both NO and SO , two coals are analyzed
X X
to show the effect of coal sulfur level on the various impacts. High sulfur
eastern and low sulfur western were selected in order to have the widest
range of coal sulfur levels. Also, NO /SO processes for oil-fired boiler
X X
application are considered only for the case of residual oil since this oil
has the most significant S0x emissions. N0x/S0 processes are examined for
application to these boilers to enable comparison between a simultaneous
N0x/S0x system and a combination of a NO -only system and an FGD system.
This comparison will be made during a future phase of the industrial boiler
evaluation, but not in this report.
I-1
-------
TABLE 1.1.1-3. NOX CONTROL LEVELS
Baseline Level of Control
NOX Emissions Moderate, 70% Intermediate, 80% Stringent, 90%
Fuel ng/J (Ib/MBtu) ng/J (Ib/MBtu) ng/J (Ib/MBtu) ng/J (Ib/MBtu)
Pulverized 404 (0.94) 121 (0.28) 80.8 (0.19) 40.4 (0.094)
Coal
Stoker 335 (0.78) 101 (0.23) 67.0 (0.16) 33.5 (0.078)
Coal
Residual 172 (0.40) 51.6 (0.12) 34.4 (0.080) 17.2 (0.040
Oil
Distillate 68 (0.16) 20.4 (0.047) 13.6 (0.032) 6.8 (0.016)
Oil
Natural 75 (0.18) 22.6 (0.053) 15.1 (0.035) 7.5 (0.018)
Gas
Where emissions are dependent on boiler size, the largest boiler is shown.
-------
TABLE 1.1.3-4. CASES SELECTED FOR DETAILED ANALYSIS -'NOX-ONLY FGT PROCESSES
Boiler
Package, Firetube
Package, Firetube
Package, Watertube
Package, Watertube
Underfeed Stoker
Package, Watertube
Chaingrate Stoker
Package, Watertube
Package, Watertube
Package, Watertube
Field Erected, Watertube
Spreader Stoker
Field Erected, Watertube
Pulverized Coal
Fuel*
Distillate Oil
Natural Gas
Residual Oil
LSW
LSW
Natural Gas
Distillate Oil
Residual Oil
LSW
LSW
Size Control Level
MWt %
4.4 70, 90
4.4 70, 90
8.8 70, 90
8.8 80
22 70, 80, 90
44 70, 90
44 70, 90
44 70, 90
44 80
58.6 70, 90
*LSW = Low Sulfur Western Coal (0.6%S)
TABLE 1.1.3-5. CASES SELECTED FOR DETAILED ANALYSIS -
SOX/NOX FGT PROCESSES
Boiler
Package, Watertube
Package, Watertube
Underfeed Scoker
Field Erected, Watertube
Pulverized Coal
Boiler Size,
Fuel* MWt
Residual Oil 44
HSE 8.8
LSW
HSE 58.6
LSW
Control
% NOX
80
80
80
Level
% SOX
85
85
85
*HSE = High Sulfur Eastern Coal (3.5% S)
LSW = Low Sulfur Western Coal (0.6% S)
1-10
-------
It should be noted that FGT technology for NO control has not yet been
X
commercially applied to coal-fired boilers. However, pilot units have been
tested and two full scale systems are scheduled. Coal-fired applications
are considered here since they are currently being offered in the U.S. and
it is felt that they will be demonstrated commercially in the near future.
1.2 FLUE GAS TREATMENT FOR CONTROL OF N0x ONLY
The systems of emission reduction considered in this study for applica-
tions to coal-fired boilers are divided into two general categories: those
which remove only NO and those which remove both NO and SO . Here and
J\ A X
throughout the study these two types of systems are considered separately to
avoid confusion.
1.2.1 System Descriptions
The N0x~only systems considered are as follows:
• Fixed Packed Bed Selective Catalytic Reduction (SCR)
• Moving Bed SCR
• Parallel Flow SCR
Absorption-Oxidation
From the comparison evaluation of these systems, the candidates for
"best" emission control systems were selected. These candidate systems
are shown, along with a brief description, in Table 1.2.1-1.
SCR systems utilize ammonia to selectively reduce nitrogen oxides. The
chemical mechanisms can be summarized by the following gas-phase reactions.
4NO + 4NH3 + 02 t 4N2 + 6H20 (1-1)
2N02 + 4NH3 + 02 -? 3N2 + 6H20 (1-2)
1-11
-------
TABLE 1.2.1-1. CANDIDATES FOR BEST EMISSION CONTROL SYSTEM
Process Description Fuel Application
Moving Bed SCR Utilizes NH3 to selectively reduce NOX to N2; Residual Oil
capable of achieving stringent NOx control level;
catalyst (rings or pellets) gravity-bed, mechani-
cally-screened, and returned to reactor.
Parallel Flow SCR Utilizes NH3 to selectively reduce NOX to N2; Coal
capable of achieving stringent NO control level; Residual Oil
special catalyst arrangement (honeycomb, parallel
plate or tubes) greatly reduces particulate impac-
tion as gas flow is parallel to catalyst surface.
V Fixed Packed Bed SCR Utilizes NH3 to selectively reduce NOX to N2; Distillate Oil
K capable of achieving stringent NOX control level; Natural Gas
ring shaped catalyst pellets packed in fixed bed.
-------
The first reaction predominates as flue gas NOX consists primarily of NO.
Oxygen is in large excess in the flue gas and does not limit the extent of
reaction. A process flow diagram is shown for an SCR system in Figure
1.2.1-1. Flue gas is taken from the boiler between the economizer and air
preheater. Ammonia, taken from a liquid storage tank and vaporized, is
injected and mixed with the flue gas prior to the reactor. The flue gas
passes through the catalyst bed where NO is reduced to N2. The flue gas
then exits the reactor and is sent to the air preheater and, if necessary,
further treatment equipment.
Particulate Re-
moval to FGD
and/or Stack
Air
Figure 1.2.1-1. Flow diagram for typical NO -only SCR process.5
X
With this and all SCR systems it is desirable to treat flue gas exiting
the economizer at 350-400°C prior to any air preheater since it is at this
temperature range than the catalysts show the optimum combination of activity
and selectivity. The analyses conducted in this study assumed that the
boilers were operated constantly at full load and, therefore, had constant
flue gas temperatures. However, it is possible that the boiler may ex-
perience large and frequent load swings which result in a variable flue gas
temperature. FGT systems in this service will require flue gas heating in
order to maintain sufficiently high temperatures. Temperature control can
be accomplished by either a heater or a slipstream around the economizer.
The heater will effectively decouple the FGT system from the boiler and
does not require flow control of a flue gas slipstream. The economizer
1-13
-------
bypass will not derate the boiler since it will only be required during low
load situations. In each of these approaches, much of the heat added to the
flue gas will be captured in the air preheater. Both alternatives do,
however, present an additional economic impact.
SCR systems can generally be applied to all boiler sizes and types,
although with existing boilers there may be problems with spacial limita-
tions. All of the catalysts considered here for'Use in treating flue gas
containing SOz and 80s are resistant to poisoning by these compounds. Long
term tests of these catalysts in the presence of SO have shown very little
or no decrease in activity or selectivity. Reactor size is proportional to
flue gas flow rate, and this will determine the size and cost of the SCR
system while the particulate concentration will determine the necessary
catalyst/reactor combination.
The reactor itself is usually sized on the basis of a space velocity
which is defined as the gas flow rate divided by the catalyst volume. A
typical space velocity for a parallel flow system is about 6000 hr l com-
pared to 8000 hr * for a moving bed or fixed, packed bed SCR system. The
pressure drop through parallel flow systems is typically on the order of
100 mm H20, which is somewhat higher than moving or fixed, packed bed sys-
tems. The pressure drop is being reduced as this technology develops.
Parallel flow, moving bed and fixed, packed bed SCR systems are all
capable of attaining the stringent level of N0x control. Greater than 90
percent NOX reduction is achieved at NH3:NOX mole ratios of 1:1 on commer-
cial systems applied to industrial boilers in Japan. All of these systems
have been applied to a variety of oil-fired industrial boilers in Japan and
appear to I
-------
or ammonium sulfate or 2) the NHs can enter the downstream equipment un-
reacted. The bisulfate has been shown to cause air preheater pluggage and
this is the subject of ongoing research both at the EPA and the Electric
Power Research Institute (EPRI). Both the bisulfate and sulfate exist as a
particulate, but may be difficult to collect if the particles are submicron
in size. Unreacted NHs is not likely to present any operational problems.
A recent study has shown that if an ESP exists downstream, then most of the
NHs will exit with the ash. NHa can actually improve the performance of an
FGD system.16
1.2.2 Economic Impacts
The costs of NO., FGT systems applied to the industrial boilers are
X
presented in this section. Two types of data are presented. First the
capital and annual costs are shown as a function of boiler size. Then the
cost effectiveness in terms of $/kg NO removed is evaluated. Tables
X
1.2.2-1 through 1.2.2-3 show the range of annual cost for the moderate to
stringent level of control for the various boiler/size/control system
combinations.
The annual costs in terms of $/MBtu/hr are plotted against boiler size
in Figures 1.2.2-1 through 1.2.2-4. In all cases, there is clearly an
economy of scale with the larger units. An interesting result is that for
the small residual oil-fired boiler, the parallel flow system is somewhat
less expensive, but with the larger boiler, the moving bed system is less
expensive. This is a result of the labor cost, which is a fixed cost, and
is higher for moving bed systems than for parallel flow systems. Therefore,
with small systems, the labor component has a significant effect on the
annual cost of these systems. This result is the primary reason why it is
not possible to choose a best system for residual oil applications. The
effect of fuel type on annual cost is shown in Figure 1.2.2-5 when costs
for the 44 MWt (150 MBtu/hr) boilers are compared for each fuel type. Sys-
tems applied to coal-fired boilers are the most expensive while those
1-15
-------
TABLE 1.2.2-1. ANNUAL COST OF NOX CONTROL SYSTEMS APPLIED
TO COAL-FIRED BOILERS
Boiler
Underfeed Stoker
Chaingrate
Spreader Stoker
Pulverized Coal
Size,
MBtu/hr
30
75
150
200
Annual Cost, $1000/yr
Control System
Parallel Flow SCR
Parallel Flow SCR
Parallel Flow SCR
Parallel Flow SCR
Moderate
108
153
221
254
Stringent
130
197
291
351
TABLE 1.2.2-2. ANNUAL COST OF NOX CONTROL SYSTEMS APPLIED
TO OIL FIRED-BOILERS
Boiler
Distillate Oil
Distillate Oil
Residual Oil
Residual Oil
Residual Oil
Residual Oil
Size,
MBtu/hr
15
150
30
30
150
150
Annual Cost, $1000/yr
Control System
Fixed Packed Bed SCR
Fixed Packed Bed SCR
Parallel Flow SCR '
Moving Bed SCR
Parallel Flow SCR
Moving Bed SCR
Moderate
64
137
96
120
181
168
Stringent
67
176
108
130
223
204
TABLE 1.2.2-3. ANNUAL COST OF NOx CONTROL SYSTEMS APPLIED
TO NATURAL GAS-FIRED BOILERS
si^e Annual Cost, $1000/yr
Boiler MBtu/hr Control System Moderate
Package, Firetube 15 Fixed Packed Bed SCR 64.4
Package, Watertube 150 Fixed Packed Bed SCR 129
Stringent
67.6
175
1-16
-------
7000-
6000-
5000-
u 4000-
3000-
2000-
1000-
—I"
100
Stringent (90%)
Intermediate (80%)
Moderate (70%)
150
200
250
Boiler Size (MBtu/hr)
Figure 1.2.2-1. Annual cost of parallel flow SCR NO FGT systems for coal-fired boilers.
-------
7000-1
I
I-1
00
g
0>
o
u
5000
4000-
3000-
2000"
1000"
Moving Bed SCR
Parallel Flow SCR
Stringent (90%)
Stringent C90%)
Moderate (70%)
Moderate (70%)
50
100
Boiler Size (MBtu/hr)
ibo
2(10
Figure 1.2.2-2. Annioal cost comparison of NO FGT systems for residual oil boilers.
-------
3000-
2000-
1000-
100
Boiler Size (MBtu/hr)
Stringent (90%)
Moderate (70%)
200
Figure 1.2.2-3. Annual cost of fixed packed bed SCR NOX FGT systems for distillate oil boilers.
-------
to
3000-
2000-
1000-
Stringent (90%)
Moderate (70%)
100
Boiler Size (MBtu/hr)
200
Figure 1.2.2-4. Annual cost of fixed packed bed SCR NOX FGT systems for natural gas boilers.
-------
300 -
200
o
o
100
ho
Basis: 44 MW. (150 MBtu/hr) boilers
80% N0x Control
Coal
Residual
Oil
Distillate
Oil
Natural
Gas
Fuel Type
Figure 1.2.2-5.
Comparison of annual costs of NOX FGT systems applied to 150 MBtu/hr boilers,
(Average costs used where two systems apply.)
-------
applied to distillate oil- and natural gas-fired boilers are the least ex-
pensive. Annual costs for residual oil-fired boilers lie in between these
two extremes. The higher costs for systems which treat flue gas from coal-
and residual oil-fired boilers are a result of the use of systems that will
handle high particulate loadings, higher baseline N0x emissions with these
fuels and, in the case of coal, higher flue gas flow rates.
Capital costs in terms of $/MBtu/hr are presented as a function of
boiler size in Figures 1.2.2-6 through 1.2.2-9. These figures also show
larger systems to be less expensive in terms of cost per unit of capacity.
This is due to the fact that the equipment costs used in this study were
either constant for all sizes or varied exponentially with size.
In addition to determining the annual and capital costs, the study also
examines the cost effectiveness of the various combinations. Cost effec-
tiveness is defined as $/kg NO., removed. Comparing the systems in this man-
X
ner shows which combinations provide the largest environmental benefit for
the lowest cost. Cost effectiveness if plotted against the level of NO
control in Figures 1.2.2-10 through 1.2.2-13.
For coal-fired boilers, the increased annual cost over an uncontrolled
boiler for N0x~only parallel flow SCR systems ranges approximately 6-12
percent, depending on the boiler and level of control. The figure plainly
shows economy of scale as the largest coal-fired standard boiler, pulverized
coal, has the most cost effective N0x control system. Annual costs for the
small boilers are labor cost-dominant, hence the maximum cost effectiveness
at 90 percent NOX control. The large boiler's costs are catalyst cost-
dominant, hence the maximum cost effectiveness at 70 percent NO control
X
(additional cttalyst is required to remove the additional NO ). Similar
X
effects occur with the other fuels as well. In all cases it is apparent
that the system size has a significantly larger effect on the cost effec-
tiveness than does the control level.
1-22
-------
7000-
6000-
5000-
4000-
3000-
2000-
1000-
Stringent (90%)
Intermediate (80%)
Moderate (70%)
—T
50
100 150
Boiler Size (MBtu/hr)
200
250
Figure 1.2.2-6. Capital cost of parallel flow SCR NOX FGT for coal-fired boilers.
-------
I
ho
7000
6000 -
5000 -
4000 -
o 3000 -
2000 ~
1000 -
'- Stringent (90%)
~ Moderate (70%)
Stringent (90%)
Moderate (70%)
Moving Bed
Parallel Flow
—T
50
100
Boiler Size (MBtu/hr)
150
~T
200
Figure 1.2.2-7. Capital cost comparison of NOX FGT systems for residual oil boilers.
-------
I
M
Ul
3000-
2000-
1000-
Stringent (90%)
Moderate (70%)
50
I
100
Boiler Size (MBto/hr)
150
I
200
Figure 1.2.2-8. Capital cost of fixed packed bed SCR NO FGT systems for distillate oil boilers.
-------
I
N3
P.
3
3000-
2000-
1000-
Stringent (90%)
Moderate (70%)
I
50
100
Boiler Size (MBtu/hr)
I
150
I
200
Figure 1.2.2-9. Capital cost of fixed packed bed SCR N0x FGT systems for natural gas boilers.
-------
3.0
o
z
2.0
1.0
70
80
Percent NO Control
Underfeed Stoker
Chaingrate
Spreader Stoker
Pulverized Coal
90
Figure 1.2.2-10. Cost effectiveness of parallel flow NOX control systems
for coal-fired boilers.
1-27
-------
10
ox
z
o
u
8.8 MWt Boiler
Moving Bed SCR
8.8 MWt Boiler
Parallel ?low SCR
44 MWt Boiler
Parallel Flow SCR
44 MWt Boiler
Moving Bed SCR
70
80
90
Percent NOy Control
Figure 1.2.2-11. Cost effectiveness of FGT systems applied
to residual oil-fired boilers.
1-28
-------
o
z
22
20
18
16
12
10
4.4 MWt Boiler
Fixed Packed Bed SCR
44 MWt Boiler
Fixed Packed Bed SCR
70
80
90
Percent NOX Control
Figure 1.2.2-12. Cost effectiveness of FGT systems applied
to distillate oil-fired boilers.
1-29
-------
x
o
z
20
18
16
14
12
10
II
u^
*M
[I]
o
o
4.4 MWt Boiler
"Fixed Packed Bed
SCR
44 MWt Boiler
"Fixed Packed Bed
70 80
Percent NO* Control
90
Figure 1.2.2-13.
Cost effectiveness of FGT systems applied
to natural gas-fired boilers.
1-30
-------
The costs of SCR applications to modified or reconstructed facilities
will be higher than those shown here. It is estimated that these costs will
range from 25 to 120 percent more than applications to new boilers.
1.2.3 Energy Impacts
In calculating energy usage for each of the cases, all sources of
energy consumption were considered. These sources are shown in Table 1.2.3-1,
TABLE 1.2.3-1. AREAS OF ENERGY CONSUMPTION IN NOV FGT SYSTEMS
NOV FGT System
Energy Consumption Step
(equipment)
Type of Energy
Consumed
Parallel Flow SCR
Moving Bed SCR
Fixed Packed
Bed SCR
Reactor Draft Loss (Fan) Electrical
Liquid NH3 Transfer (Pump) Electrical
NH3 Vaporization (Vaporizer) Steam
NH3 Dilution Steam
Reactor Draft Loss (Fan) Electrical
Liquid NH3 Transfer (Pump) Electrical
Catalyst Screening & Transfer (Elevator) Electrical
Baghouse Draft Loss (Blower) Electrical
NH3 Vaporization (Vaporizer) Steam
NH3 Dilution Steam
Reactor Draft (Fan) Electrical
Liquid NH3 Transfer (Pump) Electrical
NH3 Vaporization (Vaporizer) Steam
NH3 Dilution Steam
Soot Blowing-Distillate Oil Boiler Only Steam
The energy impacts are presented in two forms. In the first, energy
consumption in terms of MBtu/hr is plotted as a function of boiler size.
These data are shown in Figures 1.2.3-1 through 1.2.3-4. Essentially,
in all of the cases energy consumption is less than 1 MBtu/hr and represents
a small amount of energy. The relative amount of energy consumed is shown
in Figures 1.2.3-5 through 1.2.3-8 where usage is shown as a percent of
the boiler heat input.
1-31
-------
1.5
I
U>
Stringent (90%)
Intermediate (80%)
50
100
Boiler Size (MBtu/hr)
Figure 1.2.3-1. Energy consumption of parallel flow SCR NOX FGT systems for coal-fired boilers.
-------
1.0-H
I
(jJ
U>
ex
I
o
0.5-
Level of Control
A Stringent (90%)
D Moderate (70Z)
Parallel Flow SCR
Moving Bed SCR
Parallel Flow SCR
200
Boiler Size (MBtu/hr)
Figure 1.2.3-2. Energy consumption of N0y FGT systems for residual oil boilers.
-------
i.o-
I
OJ
I
.3 0.5-
-------
i
U)
Ln
p,
e
3
0.5-
Stringent (90%)
Moderate (70%)
50
100
Boiler Size (MBtu/hr)
~T~
150
200
Figure 1.2.3-4. Energy consumption of fixed packed bed SCR NO FGT systems for natural gas boilers,
-------
3
O.
C
rt
0)
PC
•H
O
PQ
C
CU
O
0.6-
0.5-
0.4-
0.3-
Pulverized Coal
Spreader Stoker
Chaingrate
Underfeed Stoker
01
oo
a
in
00
l-i
OJ
c
w
0.2-
0.1-
_L
50
70
Percent NO
80
Removal
90
100
Figure 1.2.3-5.
Energy usage of NOX control systems
as percent of boiler heat input.
Coal-fired boilers.
1-36
-------
0.6
0.5
3
cx
c
ctf
HI
EC
•H
O
CO
ctf
0)
60
cfl
cn
60
t-l
(1)
c
w
0.4
0.3
0.2
0.1
Parallel Flow SCR
Moving Bed SCR
44 MWt Boiler
8.8 MWt Boiler
60
70 80
Percent NOX Control
90
100
Figure 1.2.3-6. Energy usage of N0x control systems as percent of
boiler heat input.
Residual oil-fired boilers.
1-37
-------
0.6
0.5
P.
e
M
4J
n)
01
ra
•H
O
PQ
o
M
HI
CO
cd
0)
00
cd
ta
00
H
-------
0.5 I
3
a,
-------
For the coal-fired boilers, the systems have a range of 0.27 to 0.64
percent. Energy usage is more a function of N0x control level than boiler
size. In this analysis, energy usage is higher for larger boilers as a
result of the reactor design method which allowed the reactor pressure drop
to vary. Also, for the pulverized coal case, higher inlet N0x concentra-
tions lead to higher energy usage.
With the oil-fired boilers, the parallel flow SCR systems have a range
of 0.20 to 0.38 percent (from moderate to stringent level of control) and
the moving bed systems from 0.19 to 0.29. One can see that the moving bed
systems require less energy than the parallel flow systems. This is due to
the greater pressure drop across a parallel flow reactor, which is larger
than the AP across a moving bed reactor. Again, as in economics, the two
candidate systems are considered to have similar energy impacts. For the
distillate oil-fired boiler, the energy consumption ranged from 0.33 to
0.62 percent of the boiler heat input. For NOX FGT applied to gas-fired
boilers, the small fixed packed bed SCR system has a range of 0.27 to 0.42
percent, whereas the large system varies from 0.30 to 0.49 percent.
In all cases, energy usage was less than 0.64 percent of the boiler
heat input, and in most cases it was less than half of this amount.
1.2.4 Environmental Impacts
There are some potential adverse environmental impacts of SCR systems.
The use of NHs as the gaseous reducing agent introduces the possibility of
ammonia emissions. The level of NHs emissions experienced by commercial
SCR operations range from 1 to 10 ppm depending on the control level. Even
at elevated NH3:NOx ratios (>1.0), the NHs emissions are reported to be less
than 20 ppm. It is possible that NH3 emissions will increase as the catalyst
ages; however, commercial applications have not operated long enough to show
this effect. Ten ppm of NH3 may be an optimistic value, especially consider-
ing that currently there is no continuous monitoring technique for measuring
1-40
-------
in the presence of SOX. The data, therefore, represent spot measurements
and not continuous data. It seems reasonable to assume that 10 ppm repre-
sents a minimum level of NHs emissions. NH3 emissions in terms of pounds
per MBtu are presented in Figures 1.2.4-1 through 1.2.4-3 as a function of
boiler size. While there is some variation, emission levels are essentially
constant for all boiler sizes.
Another potential environmental problem is the formation of ammonium
bisulfate, NHijHSOit, or ammonium sulfate, (NHit)2SOit. The presence of NH3,
SOs, and H20 in the hot flue gas leads to the formation of liquid N
upon cooling to approximately 180-220° C by the following reaction.
NH3tg) + S03(g) + H20(g) J NH^HS04(1) (1-3)
This can create a plugging and corrosion problem in heat exchange equipment,
particularly when medium- or high-sulfur fuels are fired. A beneficial
effect is obtained by the tying up of S03 which is more hazardous than SOa
and difficult to catch with FGD.12 Further cooling to about 190°C precipi-
tates the formation of solid ammonium sulfate by the following reaction.
+ NH3(g) (NH^SCMs) (1-4)
It is speculated that minor, if any, amounts of these sulfates will be
emitted to the atmosphere in situations where particulate control equipment
exists downstream of the NCv control system. Sulfate formation is not a
X
problem with gas-fired boilers since there is no sulfur present in the fuel
Disposal of spent catalyst is the final environmental concern of the
parallel flow SCR systems. Catalysts such as titanium dioxide (TiCh) and
vanadium pentoxide (V20s) are probably recycled due to their high cost.
This question is currently unanswered since all applications of this tech-
nology are very recent and none have yet required a catalyst change.
1-41
-------
5.0-1
4.0-
3.0-
Stringent (90%)
2.0-
N>
Intermediate (80%)
.0-
Moderate (70%)
100 150
Boiler Size (MBtu/hr)
200
25C
Figure 1.2.4-1. NH3 emissions from SCR NOX FGT systems for coal-fired boilers.
-------
3.0-
2.0-
i.o-
50
100
Boiler Size (MBtu/hr)
Moving Bed SCR
Parallel Flow SCR
Fixed, Packed Bed SCR
20U
Figure 1.2.4-2. NH3 emissions from SCR NO FGT systems for oil-fired boilers,
-------
3.0-1
2.0-
1.0-
All control levels
1
50
I I
100 150
Boiler Size (MBtu/hr)
I
200
Figure 1.2.4-3. NHa emissions from SCR NO FGT systems for natural gas-fired boilers.
-------
Summarizing, FGT processes are relatively clean, possess minor potential
air pollution and waste problems, and have no water, thermal, or noise
pollution.
1.2.5 Development Status
Parallel flow SCR processes have been applied in Japan to several
residual oil-fired industrial boilers. Oil-fired utility boilers and other
sources with high particulate concentrations are also being treated. SCR
processes have not yet been demonstrated commercially on coal-fired boilers.
However, pilot units have been operated and some U.S. firms are offering
SCR processes for use on coal-fired boilers. Two applications to coal-
fired utility boilers are planned for 1980 (Table 1.2.5-1) although none
exist at the present time. A coal-fired pilot unit demonstration of one
parallel flow design is currently underway in the U.S. under EPA sponsor-
ship14 and several have been conducted in Japan. The EPA facility should be
operational by early 1980. Another U.S. demonstration of a N0x-only SCR
process will be performed in 1980 by the Electric Power Research Institute.
TABLE 1.2.5-1. PLANNED FGT INSTALLATIONS OF SCR COAL-FIRED UTILITY BOILERS13
Location
Takehara
Tomato
User
Electric Power
Development Co.
Hokkaido
Electric
Process
Developer
Has not been
selected
Hitachi, Ltd.
Capacity
Fuel (Nm3/hr)
Coal 800,000
Coal 280,000
Completion
Date
July 1981
October 1980
Parallel flow and moving bed SCR processes have been applied in Japan
to several oil-fired industrial and utility boilers. These operations are
summarized in Tables 1.2.5-2 through 1.2.5-5. SCR systems are considered
commercially available for oil-fired boilers at this time.
1-45
-------
TABLE 1.2.5-2.
EXISTING FGT INSTALLATIONS OF SCR PARALLEL FLOW
SYSTEMS OIL-FIRED INDUSTRIAL BOILERS17
Location
(Japan)
Sodegaura
Kawasaki
Chiba
User
Fuji Oil
Aj inomoto
Ukishima
Pet. Chem.
Process
Developer Fuel
Mitsubishi Resid
H.I.
Ishikawaj ima Resid
H.I.
Mitsui Resid
Engineering
Capacity
(NmVhr)
200,000
180,000
220,000
Completion
Date
January 1978
January 1978
April 1978
TABLE 1.2.5-3. EXISTING FGT INSTALLATIONS OF SCR PARALLEL FLOW
SYSTEMS OIL-FIRED UTILITY BOILERS18
Location
(Japan) User
Process
Developer
Capacity Completion
Fuel (Nm /hr)
Date
Yokosuka
Tokyo
Electric
Mitsubishi H.I. Resid 40,000 March 1977
Chita
Chubu
Electric
Mitsubishi H.I. Resid 1,920,000 February 1980
Kudamatsu Chugoku Ishikawajima Resid 1,900,000 July 1979
Electric H.I.
Niigata Tohoku Ishikawajima Resid 1,660,000 August 1981
Electric H.I.
1-46
-------
TABLE 1.2.5-4. EXISTING FGT INSTALLATIONS OF SCR MOVING BED SYSTEMS
OIL-FIRED INDUSTRIAL BOILERS18
Location
(Japan)
Process
Capacity Completion
User
Developer Fuel (Nm /hr)
Date
Kaizuka
Chiyoda Kenzai Hitachi, Ltd. Resid 15,000 October 1977
Amagasaki Nippon Oils & Hitachi, Ltd. Resid 20,000 April 1978
Fats
Sodegaura
Sumitomo
Chemical
Mitsubishi
H.I.
Resid 300,000 September 1976
Sodegaura
Sumitomo
Chemical
Sumitomo
Chemical,
Mitsubishi
H.I.
Resid 300,000 October 1976
Hirakatu Kurabo
Kurabo
Resid 30,000 August 1975
1-47
-------
TABLE 1.2.5-5. EXISTING FGT INSTALLATIONS OF SCR FIXED BED SYSTEMS OIL-FIRED INDUSTRIAL BOILERS
I 7
Location
(Japan)
User
Process Developer
Fuel
Capacity
(NmVhr)
Completion
Date
i
4>
OO
Amagasaki
Amagasaki
Amagasaki
Sakai
Hokkaichi
Sodegaura
Sodegaura
Sorami
Sorami
Sorami
Sorami
Kawasaki
Kawasaki
Chita
Kansai Paint
Nisshin Steel
Nisshin Steel
Nisshin Steel
Shindaikyowa P.C.
Sumitomo Chemical
Sumitomo Chemical
Toho Gas
Toho Gas
Toho Gas
Toho Gas
Nippon Yakin
Toho Gas
Toho Gas
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi Zosen
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Mitsubishi Kakoki
Mitsubishi Kakoki
Mitsubishi Kakoki
Distillate
Resid
Re s id
Distillate
Resid
Resid
Resid
Distillate
Distillate
Distillate
Distillate
Resid
Distillate
Distillate
16,000
20,000
19,000
30,000
440,000
30,000
240,000
62,000
23,000
23,000
19,000
14,000
30,000
30,000
October 1977
August 1977
July 1977
December 1978
November 1975
July 1973
March 1976
October 1977
December 1977
June 1978
July 1978
July 1978
October 1977
October 1977
-------
Table 1.2.5-6 shows the numerous industrial fixed packed bed SCR
applications. Although gas-fired boilers, both industrial and utility,
are numerous in Japan, few have been equipped with SCR units so far. This
is due to the fact that less expensive NOX reduction by combustion modifica-
tions on these boilers has been adequate to meet environmental regulations.
Fixed packed bed SCR systems are considered to be commercially available
for natural gas-fired boilers at this time.
TABLE 1.2.5-6. GAS-FIRED SCR PLANTS IN JAPAN17
Company
Osaka Gas
Chubu Electric
Kyushu Electric
Chubu Electric
Hyushu Electric
Site
Takaishi
Chita
Kokura
Chita
Kokura
Capacitj'
(Nm3 /hr)
15,000x2
1,910,000
1,610,000
1,910,000
1,610,000
Reactor
type
FPB
FPB
FPB
FPB
FPB
Completion
date
December 1976
April 1978
July 1978
September 1978
December 1978
1.3 FLUE GAS TREATMENT FOR CONTROL OF N0x AND S0x
Some FGT processes have the capability of removing SO in addition to
N0x. These processes are typically more complex and costly than those which
remove just N0x; however, this is offset by the simultaneous dual pollutant
control capability. For this reason, these processes are considered
separately from the NO -only processes.
x
1.3.1 System Description
The following NOX/SOX systems are considered for application to the
coal-fired boilers:
1-49
-------
• Parallel Flow SCR
Adsorption
Electron Beam Radiation
Absorption-Reduction
Oxidation-Absorption-Reduction
Oxidation-Absorption.
Parallel flow SCR is selected as the only candidate for "best" NO /S()
X X
system. The choice here is a combination of no serious secondary environ-
mental impacts, system performance, system reliability, and status of
development. The process is described briefly in Table 1.3.1-1.
TABLE 1.3.1-1. BEST NOX/SOX EMISSION CONTROL SYSTEM
FOR COAL-FIRED BOILERS
Process Description
Parallel Flow SCR Utilizes NHs to catalytically reduce NOX
after SOX is adsorbed by and reacted with
catalyst; capable of achieving >90 percent
NOX and SOX reduction; SOx saturated
catalyst is regenerated while flue gas is
diverted to alternate reactor.
The NOy reduction reactions occurring in this process are the same as
those described by reactions (1-1) and (1-2). The process utilizes an
acceptor material to adsorb SOa and the product of the adsorption reaction
then acts as an NOX reduction catalyst. Elemental copper is converted to
oxide form by flue gas oxygen.3
CU(H) + '.O? (g) -> CuO(s) (1-5)
Sulfur dioxide reacts with the copper oxi.dc, as described by:3
S02(g) + li02(PJ + CuO(s) + CuSO.,(s) (!-())
1-30
-------
80s in the flue gas is also removed:
S03 (g) + CuO(s) ->- CuS04(s) (1-7)
It is this final copper sulfate (CuSOO reaction product that acts as the
primary catalyst for NOx reduction by ammonia in the parallel flow SCR
NOx/SOx system. After the spent reactor is isolated from flue gas flow,
the reactor is purged with steam. A reducing gas, usually hydrogen, is
then added which reacts with the copper sulfate in the following manner:4
CuSCH(s) + 2H2(g) -»• Cu(s) + S02(g) + 2H20(g) (1-8)
The off-gas of this reaction is cooled to condense out the steam, reducing
the gas volume and thus concentrating the S02. The concentrated S02 is
compressed and sent to a workup section to produce either elemental sulfur,
liquid S02, or sulfuric acid. (Sulfuric acid is produced in the cases
studied in detail for this report.)
Besides the catalyst regeneration and sulfur treatment sections of the
parallel flow SCR simultaneous NOX/SOX system, the N0x-only and NOX/SOX
parallel flow SCR systems are quite similar and the technical description
contained in the previous section (1.2.1) applies here also.
No continuous coal-fired NO removal test data for NO /SO systems are
X XX
available. Continuous oil-fired M3 removal test data for the parallel flow
X
SCR NO /SO system have been obtained from a 40 MW unit in Japan. These
data shox
-------
1.3.2 Economic Impacts
The annual and capital costs of this system applied to two coal-fired
v
boilers and a residual oil-fired boiler are presented in this section. The
costs are based on using a sulfuric acid plant for S02 workup and a com-
pressor/gasholder for flow smoothing. Table 1.3.2-1 shows the annual costs
of applications to coal-fired boilers. Two boiler types and two coals are
presented and these data are plotted in Figure 1.3.2-1. Once again, economy
of scale with large systems is evident; however, the effect is most signifi-
cant for the high sulfur coal cases.
TABLE 1.3.2-1. ANNUAL COST OF PARALLEL FLOW NOX/SOX CONTROL SYSTEMS
Boiler Fuel Annual Cost, $1000/yr
Pulverized Coal High Sulfur Eastern 1805
Low Sulfur Western 894
Underfeed Stoker High Sulfur Eastern 711
Coal Low Sulfur Western 462
Capital costs are shown in Figure 1.3.2-2 and the significant effect of
boiler size on costs can again be seen.
Table 1.3.2-2 shows the annual cost of the single case studied. Only
one case was analyzed for reasons described earlier and, as a result, it is
not possible to plot the results or show trends.
TABLE 1.3.2-2. COSTS OF PARALLEL FLOW NOX/SOX CONTROL SYSTEM
Annual Cost, Capital Cost,
Boiler Fuel $1000/yr $1000
Package, Watertube Residual Oil 1092 3801
1-52
-------
I
Ln
Lo
w
5
80-1
70-
60-
50'
40-
30'
20-
10-
High Sulfur Eastern Coal
Low Sulfur Western Coal
~T
50
100 150
Boiler Size (MBtu/hr)
200
~T
250
Figure 1.3.2-1. Annual cost of parallel flow SCR N0x/S0x FGT for coal-fired boilers at inter-
mediate level of control.
-------
I
Ln
-P-
80"
70-
60"
50-
40-
30"
20-
10"
High Sulfur Eastern Coal
Low Sulfur Western Coal
50
~l 1
100 150
Boiler Size (MBtu/hr)
200
~T
250
Figure 1.3.2-2.
Capital cost of parallel flow SCR NOX/SOX FGT for coal-fired boilers at inter-
mediate level of control.
-------
1.3.3 Energy Impacts
Energy usage for these cases is summarized in Table 1.3.3-1 and plotted
in Figure 1.3.3-1 where energy consumption is plotted against flue gas flow
rate. The curves are essentially linear with the high sulfur case having a
significantly greater impact. With high sulfur coal, the energy usage is
7.7 percent of the boiler heat input for both boiler types. With the low
sulfur coal, this figure drops to 2.2 percent of the boiler heat input.
TABLE 1.3.3-1. ENERGY CONSUMPTION OF NOX/SOX CONTROL PROCESSES
APPLIED TO COAL FIRED BOILERS
Boiler Fuel Energy Consumption, MWt (MBtu/hr)
Pulverized Coal High Sulfur Eastern 4.5 (15)
Low Sulfur Western 1.2 (4.1)
Underfeed Stoker High Sulfur Eastern 0.68 (2.3)
Low Sulfur Western 0.20 (0.68)
Energy use for the oil-fired boiler is shown in Table 1.3.3-2. Here
again, it is not possible to plot the result.
TABLE 1.3.3-2. ENERGY CONSUMPTION OF PARALLEL FLOW NOX/SOX CONTROL SYSTEM
Energy Consumption
Boiler Fuel MWt (MBtu/hr) % of Boiler Heat Input
Package, Watertube Residual Oil 2.0 (6.6) 4.4
1.3.4 Environmental Impacts
The environmental impacts of this NOX/SOX process are similar to those
of the N0x-only processes. The primary adverse environmental impact is from
emissions. The process developers claim that these emissions are low
1-55
-------
20-)
I
Ui
cy.
15-
c
o 10-
4-1
P-
§
CO
a
o
o
01
c
w
5-
High Sulfur Eastern Coal
Low Sulfur Western Coal
I
50
I
100
\
150
200
I
250
Boiler Size (MBtu/hr)
Figure 1.3.3-1. Energy consumption of parallel flow SCR NOX/SOX FGT systems for coal-fired boilers,
-------
(<10 ppm); however, a continuous NHs monitoring method for use with gases
containing sulfur oxides will be necessary before NHa emissions can be
accurately quantified. The potential problems with ammonium bisulfate and
sulfate formation should be much less with the NOX/SOX process since much
of the SOX has been removed from the flue gas.
1.3.5 Development Status
The integrated process has been tested on oil but not coal; however,
the NO and SO removal capabilities have been demonstrated separately.
X A
The S02 capabilities have been demonstrated on a pilot unit treating coal-
fired flue gas. An EPA-sponsored U.S. test of the integrated process on
flue gas from a coal-fired boiler is scheduled for 1980. Pilot and demon-
stration units of Shell/UOP process are summarized in Table 1.3.5-1 and
commercial applications are summarized in Table 1.3.5-2.
1-57
-------
TABLE 1.3.5-1. SHELL/UOP PROCESS, PILOT AND DEMONSTRATION UNIT
Location/
Company Designed By
Shell Ref. Shell
at Pernis
Rotterdam Shell
Utility
Tampa Elec . UOP
Big Bend
i — •
i
t-n
JGC Yokohama JGC
Yokohama
Nippon JGC
Steel
Fuel/
Application
Residual
Fuel Oil-
Proc. Heater
Coal-
Steam Boiler
Coal-
Wet -Bottom
Utility Boiler
Fuel Oil
Sintering
Furnace
Size, Type of
Nm3/hr Operation
600-1000 SO -only
Heavy Fly Ash
Loading
1200-2000 S0x-only
SOX-NOX
S imul t aneo u s
250-700 NO -only
X
2000 M) -only
X
Dates
1967-1972
1971
1974-1976
1974-
1976-1978
JGC
Coke Oven
400
N0v-only
1976-1977
-------
TABLE 1.3.5-2. SHELL/UOP PROCESS COMMERCIAL APPLICATIONS
Unit
SYS*
Yokkaichi
Kashima Oil
Co. Ltd.
Fuji Oil
Co. Ltd.
Nippon
Steel Corp.
Fuel/
Designed By Application
Shell Residual
Fuel Oil-
Ref. Boiler
JGC Fuel Oil-
Process Unit
Heater
JGC CO Boiler
JGC Sintering
Furnace
Size, Type of
Nm3/hr Operation
125,000 S0x-only;
N0x-S0x
Simultaneous
50,000 N0x-only
70,000 N0x-only
150,000 N0x-only
Dates
1973-1975
1975-
1975-
1976-
1978-
-------
REFERENCES
1. U.S. Environmental Protection Agency, "Task 2 Summary Report -
Preliminary Summary of the Industrial Boiler Population."
2. Matsuda, S., et al. Selective Reduction of Nitrogen Oxides in
Combustion Flue Gases. Journal of the Air Pollution Control
Association. April 1978. pp. 350-353.
3. Faucett, H.L., et al. Technical Assessment of N0x Removal Processes
for Utility Application. EPA-600/7-77-127. November 1977. pp. 350.
4. Ibid., p. 352.
5. Ibid., p. 217.
6. Ando, Jumpei. NO Abatement from Stationary Sources in Japan. EPA
Report in Preparation. October 1978. pp. 3-67 - 3-82.
7. Ibid., p. 3-7.
8. Ibid., p. 4-41.
9. Ibid., p. 4-95
10. Wong-Woo, Harmon. "Observation of FGD and Denitrification Systems in
Japan." State of California Air Resources Board - SS-78-004. March 7,
1978. Appendix IV. p. 30.
11. Ibid., p. 32.
12. Faucett, H.L., op oit. , p. 5.
13. Ando, J., op cit., p. 1-35.
14. N0x Control Review." Vol. 3, No. 4. Fall 1978. p. 3.
15. Ando, J., op ait., pp. 4-1 - 4-133.
1-60
-------
16. Noblett, J.G., et al. "Impact of NOX Selective Catalytic Reduction
Processes on Flue Gas Desulfurization Processes," Draft Report. EPA
Contract 68-02-2608, J.D. Mobley Project Officer, Radian Corporation.
September 19, 1979.
17. Ando, J., op cit., pp. 3-4, 3-5.
18. Ibid., P. 3-7-
1-61
-------
SECTION 2
EMISSION CONTROL TECHNIQUES
This section presents descriptions of all control techniques for NO
control by flue gas treatment (FGT). Each control technique is described
separately, however, there may be several vendors offering processes that
are similar. Where this occurs, an effort has been made to generalize the
various processes into a single technique within a single category. This
is usually, but not always, possible. Where significant differences exist,
they are discussed separately.
A distinction has been made between those processes which remove only
NO and those which remove both NO,, and SOa- This is necessary because when
X X
final process comparisons are made it will be necessary to compare the cost
of a NO only process plus an FGD system versus the cost of a N0x/S0 pro-
cess. In the subsections which follow, all NO^ only processes are grouped
X
together and presented first and the NO /SO process are presented second.
Economics for the various NO control processes are presented only for
comparison and use in Section III for process selection. These economic
figures do not necessarily represent costs for application of these systems
to industrial boilers, in fact, most were developed with utility applications
in mind. However, at this time they are the only published economic data
available. Detailed cost estimates for several systems as applied to indus-
trial boilers will be developed for this study in Section IV.
2-1
-------
2.1 Principles of Control
The FGT systems are examined on the basis of application to industrial
boilers. These boilers are generally smaller than those useti for utility
applications and produce steam for purposes such as electrical power genera-
tion, process heating and space heating. They range in size from small
package units to large field erected units. The demand on the boilers may
be constant, such as with process heating, or cyclic, such as with space
heating.1 Industrial boilers generally have fewer burners than utility
boilers and, therefore, taking just one burner out of service can have
a significant effect on the flue gas characteristics. Also, the stoker
units typically run with higher excess air. These characteristics of indus-
trial boilers indicate that typical flue gases can have a wide variety of
characteristics.
This study considers seven standard boilers as selected for a variety
of reasons in a separate study.3 These boilers are described in Table 2.1-1.
Three coals, two oils and natural gas are included as well as four sizes of
coal-fired boilers. The coals considered are low sulfur western (0.6%S),
low sulfur eastern (0.9%S) and high sulfur eastern (3.5%S). The two oils
are distillate oil (#2) and residual oil (#6) .
NOX is formed in boilers by two mechanisms. In one mechanism, thermal
fixation, Nz and 02 present in the combustion air react to form NO. This
reaction requires the high temperatures that are present in the burner flame
and is dependent also on the Oz concentration in the flame. The reaction
does not reach equilibrium and therefore the amount of NO formed by this
mechanism is governed by reaction kinetics.* The second mechanism, fuel
nitrogen conversion, involves the reaction of nitrogen contained in the
molecular structure of the fuel with 02 in the combustion air. The rate of
reaction is a function of fuel nitrogen conversion and 02 concentration. A
more detailed description of the NOX formation mechanisms is contained in the
Technology Assessment Report on NOX control by combustion modifications.
2-2
-------
TABLE 2.1-1.
CHARACTERISTICS OF THE STANDARD BOILERS CONSIDERED
FOR ANALYSIS IN THIS REPORT
Boiler
Package, Firetube
Package, Firetube
Package, Watertube
Package, Watertube
Underfeed Stoker
Package, Watertube
Chaingrate Stoker
Package, Watertube
Package, Watertube
Package, Watertube
Field Erected, Watertube
Spreader Stoker
Field Erected, Watertube
Pulverized Coal
Package, Watertube
Package, Watertube
Underfeed Stoker
Field Erected, Watertube
Pulverized Coal
Fuel*
N0x-0nly FGT Systems
Distillate Oil
Natural Gas
Residual Oil
LSW
LSW
Natural Gas
Distillate Oil
Residual Oil
LSW
LSW
NOX/SOX FGT Systems
Residual Oil
HSE
LSW
HSE
LSW
Size
4.4
4.4
8.8
8.8
22
44
44
44
44
58.6
44
8.8
58.6
Control Level
70, 90
70, 90
70, 90
80
70, 80, 90
70, 90
70, 90
70, 90
80
70, 90
80 NOX, 85 SOX
80 NOX, 85 SOX
80 NOX, 85 SOX
*HSE = High Sulfur Eastern Coal (3.5% S)
LSW = Low Sulfur Western Coal (0.6% S)
2-3
-------
The NOX emissions for the various coals considered are different,
presumably due to different fuel nitrogen concentrations. However, the
emissions from the stoker boilers, on a ppm and mass per energy input basis,
do not change from boiler to boiler. The mass rates do change due to
differences in the flue gas flow rates for the various boilers. Emission
rates for the standard boilers are shown in Table 2.1-2. The emission rates
are based on AP-42 calculations.
In the sections which follow Section II, it is shown that NOX FGT
system designs are not significantly affected by NOX concentration. The
most significant design variables are flue gas flow -rate and control level.
For this reason, it is possible to generate information over the entire
boiler size range while considering only one coal type. The coal chosen
for analysis is low sulfur western since this coal has both the highest
flue gas flow rates and NOX emissions.
FGT systems utilize either a gas phase reaction or liquid absorption
to treat the flue gas. In most cases the gas phase reaction is between NO
and NHs in the presence of a solid phase catalyst. The catalyst is contained
within a reactor and may be either fixed or moving bed. The NOx is converted
to NZ which exits with the flue gas.
Systems utilizing a liquid absorption technique contact flue gas and
absorbent in conventional scrubbers. The absorbed NO either remains in the
scrubbing liquor and is treated in the liquid phase or reacts with a solute
to form Na which degasses and leaves with the flue gas.
The N0x FGT systems discussed in the following subsections are divided
into two categories. Those which remove only NO are presented first and
the simultaneous N0x/S0x processes are discussed second. The distinction
is made since the two process types cannot be accurately compared unless FGD
flue gas desulfurization (FGD) is included with the N0x-only processes. This
comparison will be made, but only in the Comprehensive Technology Assessment
2-4
-------
TABLE 2.1-2. NOX EMISSION RATES FOR THE STANDARD BOILERS
ro
I
NOx Emissions
Boiler
Package, Firetube
Package, Firetube
Package, Watertube
Package, Watertube
Underfeed Stoker
Package, Watertube
Chaingrate
Package, Watertube
Package, Watertube
Package, Watertube
Field Erected, Watertube
Spreader Stoker
Field Erected, Watertube
Pulverized Coal
Fuel*
Distillate Oil
Natural Gas
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
Natural Gas
Distillate Oil
Residual Oil
HSE
LSE
LSW
HSE
LSE
LSW
g/s
0.300
0.332
2.02
2.40
2.06
2.95
6.02
5.15
7.40
3.31
2.99
7.47
12.0
10.3
14.8
19.2
16.5
23.7
(lb/hr)
(2.38)
(2.63)
(16.0)
(19.05)
(16.35)
(23.40)
(47.70)
(40.80)
(58.65)
(26.26)
(23.76)-
(60.00)
(95.40)
(81.45)
(117.15)
(152.46)
(130.50)
(187.56)
g/MJ
0.0688
0.0774
0.228
0.275
0.237
0.335
0.275
0.232
0.335
0.0753
0.0680
0.172
0.275
0.232
0.335
0.327
0.280
0.404
(lb/106Btu)
(0.16)
(0.18)
(0.53)
(0.64)
(0.55)
(0.78)
(0.64)
(0.54)
(0.78)
(0.175)
(0.158)
(0.40)
(0.64)
(0.54)
(0.78)
(0.76)
(0.65)
(0.94)
ppm
97
104
373
335
288
402
336
290
401
110.
101
292
337
288
400
466
396
550
*Coal types: HSE = High Sulfur Eastern (3.5%S)
LSE = Low Sulfur Eastern (0.9%S)
LSW = Low Sulfur Western (0.6%S)
-------
Report (CTAR) which follows completion of the Individual Technology
Assessment Reports (ITAR's). Therefore, in Section III of this ITAR, NOX-
only processes will only be compared with other N0x-only processes and NOX/
SOX processes will only be compared with other NOX/SOX processes. This
distinction will be maintained throughout the other sections of the ITAR
also. The N0x-only processes described are:
Fixed Packed Bed Selective Catalytic Reduction (SCR)
Moving Bed SCR
Parallel Flow SCR
Absorption-Oxidation
The NCL/SO.. processes described are:
X X
Parallel Flow SCR
• Adsorption
• Electron Beam Radiation
• Absorption-Reduction
Oxidation-Absorption-Reduction
• Oxidation-Absorption
2.2 CONTROLS FOR COAL-FIRED BOILERS
2.2.1 Selective Catalytic Reduction-Fixed Packed Bed Reactors
Fixed packed bed systems for selective catalytic reduction of NO are
applicable only to flue gas streams containing particulate emissions of less
than 20 mg/Nm3. Particulate emissions for all coals are higher, on the
order of 1-5 grams per Nm3. Although it is possible to install a hot ESP
to reduce the particulate level to 20 mg/Nm3 this is expensive and not always
effective. For these reasons fixed packed bed SCR systems are not considered
for application to coal-fired boilers by process vendors.5
2-6
-------
2.2.2 Selective Catalytic Reduction-Moving Bed Reactors
Moving bed systems for selective catalytic reduction of NO are
X
applicable only to flue gas streams containing less than 1 g/Nm3. Particu-
late emissions for all coals are higher, on the order of 1-5 grams per Nm3.
Although it is possible to install a hot ESP to reduce the particulate level
to 1 g/Nm3 this is expensive and not always effective. For these reasons
moving bed SCR systems are not considered for application to coal-fired
boilers in this report.
2.2.3 Selective Catalytic Reduction-Parallel Flow Reactor
2.2.3.1 System Description—
The distinguishing aspect of. this process is the catalyst shape which
is produced in a variety of shapes. The catalysts are produced in either a
honeycomb, pipe, or plate shape. Both metal and ceramic supports are em-
ployed. Several shapes are illustrated in Figure 2.2.3-1. The catalyst
shapes allow particulate laden flue gas to pass through the reactor with no
inertial impaction of the particles while the NO is transported to the
catalyst surfaces by basic diffusion. The catalysts can handle all of the
particulate levels emitted by the standard boilers. All of the catalysts
considered here for use in treating flue gas containing S02 and S03 are
resistant to poisoning by these compounds. Long term tests of these cata-
lysts in the presence of SO have shown very little or no decrease in
activity or selectivity.
The reactors used are similar to standard fixed packed bed units and
an example is shown in Figure 2.2.3-2. The catalyst is usually prepared in
small modules and manually stacked within the reactor. The specific arrange-
ment will depend on the particular process under consideration.
A typical flow diagram for a parallel flow SCR system is shown in
Figure 2.2.3-3. The arrangement is similar to the other SCR processes in
that hot flue gas leaving the boiler economizer is injected with NH3 and
2-7
-------
Ml N II
Honeycomb
(Ceramic)
(Grid Type)
?0°0°6
Honeycomb
(Ceramic)
(Hexagonal)
Honeycomb
(Metal)•
(Wave Type)
Plate (Ceramic)
Plate (Metal)
Tube (Ceramic)
Parallel Passage
Figure 2.2.3-1. Shapes of parallel flow catalysts.
22
2-8
-------
CATALYST LAYER
Figure 2.2.3-2. Typical reactor used with parallel flow SCR process.
2 3
PARTICULATE REMOVAL,
TO FQD
AND/OR STACK
A!R
Figure 2.2.3-3. Flow diagram for parallel flow SCR process.
2-9
-------
passed through a catalyst bed. Temperature control is important and can
be accomplished with either a fired heater or an economizer bypass. NH3
can be controlled using boiler operating condition inputs to conventional
control components.
Within the reactor, NOX reacts with NH3 to form N2 and H20 according
1 ?
to the following reactions.
4NH + 4NH3 + 02 £ 4N2 + 6H20 (2-1)
2N02 + 4NH3 + 02 £ 3N2 + 6H20 (2-2)
Reaction (2-1) is the primary reaction since flue gas N0x is typically 90-
95 percent NO. 02 is necessary for both reactions and is present in suffi-
cient quantities (>3 percent) in all of the flue gases from the standard
boilers.
The catalyst volume for a desired NO removal can be determined by the
X
fundamental design equation for a plug flow reactor.13
?= r —
Jo r
(2-3)
The reaction rate, r, can be expressed as
r = k[NH3]a [N0]b [02]° (2-4)
The variables presented here have the same definitions as those presented
in equations 2-3 and 2-4 of Section 2.3.2. Catalyst volume can also be
determined by knowing the space velocity for a given catalyst and NO con-
version level. The space velocity is defined as the flue gas flow rate
divided by the catalyst volume.
2-10
-------
The reaction rate is different for each catalyst formulation since
different catalysts will lower the activation energy by different amounts.
The activation energy affects the reaction rate constant, k, according to
the Arrhenius equation..
k = Ae
E
RT (2-5)
Example values of k, a, b, and c for two catalyst formulations are shown
in Table 2.2.3-1.
An important design variable with catalytic systems is the space
velocity which expresses the volume of catalyst required to treat one
volume per hour of flue gas. Space velocity varies with catalyst formula-
tion, catalyst shape, and control level. Typical values of space velocity
for various catalyst shapes are shown in Table 2.2.3-2. Also shown are
other catalyst design variables such as catalyst dimensions, gas velocities,
bed depth and pressure drop. Ranges of values are used since specific values
are different for each catalyst. The values shown pertain to 90 percent NO
A
removal and an NH3/NO mole ratio of 1:1.
Both NH3/NO ratio and space velocity will change with removal level.
X
The NHs/NO mole ratio will range from 0.7-1.0 and the space velocity will
X
range approximately as shown in the table for control levels of 70 to 90
percent.l5
Variables associated with the boiler can also affect the performance
of these systems. These are
flue gas flow rate
NO concentration
X
boiler load variability
2-11
-------
TABLE 2.2.3-1. REACTION RATE DATA FOR TWO
CATALYST FORMULATIONS11
Catalyst: VzOs on
_ 9650
k = 2.05 x 103e RT
a = 0.30
b = 0.22
c = 0.05
Catalyst: Fe-Cr on
k = 3.25 x 103e
a = 0.45
b = 0.10
c = 0.15
10,860
RT
TABLE 2.2.3-2. CATALYST DESIGN VARIABLES FOR VARIOUS CATALYST SHAPES
(Basis: 90% NO removal at NH3/NO ratio of 1:1,
350-400°C)
25
Catalyst size (ram)
Thickness
Opening
Gas velocity (m/sec)
Bed depth (m)
SV (1,000 hr~1)b
Pressure drop (mmHjO)
Honeycomb
(metallic)
1
4-8
2-6
1-2
5-8
40-80
Honeycomb ,
tube (ceramic)
2.3-5
6-20
5-10
1.5-5
4-8
40-160
Parallel
(Ceramic)
8-10
8-14
5-10
4-6
1.5-3
80-160
Plate
(Metallic)
1
5-10
4-8
2-5
2-4
60-120
o
Velocity at 350-400°C in open column (superficial velocity).
Gas volume (Nm3/hr)/catalyst bed volume (m3).
2-12
-------
The flue gas flow rate and control level determine the catalyst volume
(hence reactor size) necessary. Increases in either also increase the
reactor size. The N0x concentration is primarily a function of fuel type
used in the standard boilers. Higher concentrations require larger NH3
storage and vaporization equipment; reactor size is not affected. Boiler
load can affect several things including flue gas temperature, flow rate,
and NO concentration. It is necessary to maintain reaction temperatures
of 350 to 400°C. Temperature control equipment may be necessary to
accomodate large boiler load variations which may lower the flue gas
temperature. Where these variations are present, some equipment overdesign
may be warranted to insure a constant control level. These variables are
discussed in more detail in the section on moving bed SCR systems for coal-
fired boilers, Section 2.2.2.
Parallel flow SCR processes have been applied in Japan to several
residual oil-fired industrial boilers. Oil-fired utility boilers and other
sources with high participate concentrations are also being treated. Two
applications to coal-fired utility boilers are planned for 1980 (Table
2.2.3-3) although none exist at the present time. A coal-fired pilot unit
demonstration of one NO -only parallel flow design is currently underway in
the U.S. under EPA sponsorship and several have been conducted in Japan. The
EPA facility should be operational by mid-1979. Also, a parallel flow pilot
system will be applied to flue gas from a coal-fired boiler in a study
sponsored by the Electric Power Research Institute (EPRI). The unit is
expected to be operational by 1980. A list of vendors of parallel flow SCR
systems is presented in Table 2.2.3-4. The number of pilot unit demonstra-
tions indicates that application of parallel flow SCR processes to coal-fired
industrial boilers is feasible.
2.2.3.2 System Performance—
Performance da^.a based on pilot plant testing were not found in the
literature, however, data do exist for oil-fired applications. Since many
of the flue gas characteristics are similar for oil and coal-fired boilers,
2-13
-------
TABLE 2.2.3-3. PLANNED FGT INSTALLATIONS OF SCR COAL-FIRED UTILITY BOILERS26
Location
User
Process
Developer
Capacity Completion
Fuel (NmVhr)
Date
Takehara Electric Not yet announced Coal 800,000 July 1981
Power C.
Tomato Hokkaido Hitachi, Ltd.
Electric
Coal 88,000 October 1980
TABLE 2.2.3-4. PROCESS VENDORS OF PARALLEL FLOW SCR PROCESSES
28
Vendor
Hitachi Zosen
Hitachi, Ltd.
Japan Gasoline Corp.
Mitsui Engineering & Shipbuilding
Mitsubishi Heavy Industries
Ishikawaj ima-Harima Heavy Industries
Kobe Steel
Kawasaki Heavy Industries
Shell/UOP
Demonstrated
Yes /No
yes
yes
no
no
yes
yes
no
yes
by 1979
on Coal
Scale
pilot
pilot
—
—
pilot
pilot
—
pilot
pilot
2-14
-------
it is expected that the FGT performance will be roughly similar. Detailed
data on oil-fired applications are contained in Section 2.3.
There are some potential problems downstream of the SCR systems due to
the presence of the unreacted ammonia in the flue gas. Two things can
happen: 1) the NH3 can react with S02 or SO3 to form ammonium bisulfate or
ammonium bisulfate or 2) the NH3 can enter the downstream equipment unreacted
The bisulfate has been shown to cause air preheater pluggage and this is
the subject of ongoing research both at the EPA and the Electric Power
Research Institute (EPRI). Both the bisulfate and sulfate exist as a par-
ticulate, but may be difficult to collect if the particles are submicron in
size. Unreacted NH3 is not likely to present any operational problems. A
recent study has shown that if an ESP exists downstream, then most of the
NHs will exit with the ash. NH3 can actually improve the performance of
an FGD system.129
2.2.4 Absorption-Oxidation
2.2.4.1 System Description—
Absorption-oxidation processes remove NO from flue gas by absorbing
the NO or NO into a solution containing an oxidant which converts the NO
x x
to a nitrate salt. Two types of gas/liquid contactors can be used and
examples of each type are shown in Figure 2.2.4-1. Both perforated plate and
packed towers accomplish NO absorption by generating high gas/liquid inter-
facial areas. The choice of one type of contactor is a design decision made
to achieve a given removal for the least cost.
A generalized process flow diagram is shown in Figure 2.2.4-2. Flue
gas is taken from the boiler after the air preheater. Before the gas can
be sent to the NO absorber^ it must be S02-free since S02 consumes prohibi-
X
tive amounts of the costly liquid-phase oxidant. In most cases, the oxidant
is permanganate (MnOij), but Ca(C10)2 can also be used. Therefore, a conven-
tional FGD unit is required ahead of the NO absorber. A prescrubber to cool
2-15
-------
FLUE GAS OUT
Principo! —
interface
LIQUID OUT -*- Pi
:{]•*- LIQUID IN
— Coalesced
dispersed
-Perforated
plate
— Downspout
RUE GAS IN
FLUE GAS OUT
LIQUID IN-
FLUE GAS IN -
r-Interfoce
Pocking
LIQUID OUT
Perforated Plate Absorber
Packed Absorber
Figure 2.2.4-1. Gas/liquid contactor options for
Absorption-Oxidation Processes.2 9
2-16
-------
Flue
Gas
Prescrubber
and
S02 Scrubber
NOX
Absorber
To Reheat
and Stack
Holding
Tank
Oxidant
Make-up
Nitrate Treatment and
Oxidant Regeneration
Figure 2.2.4-2,
Process flow diagram for Absorption-
Oxidation Process.30
2-17
-------
the gas and remove both participates and Cl prior to FGD is also necessary.
After having passed through these two scrubbing sections, the flue gas enters
the distributing space at the bottom of the NOX absorber, below the packing
or plates. The gas passes upward through the column, countercurrent to the
flow of the liquid absorbent/oxidant (usually a KOH solution containing
KMnCM. NO is absorbed and then oxidized over the length of the column
according to the following reactions.3
NO tg) -> NO(aq) (2-6)
NO(aq) + KMnO^taq) + KN03(.aq) + Mn02(s) (2-7)
2N02(g) -> N204(g) (2-8)
N204(g) + N20i»(aq) (2-9)
4- 2K2Hn01+(aq) ->• 2KMnOit(aq) + 2KN02 (aq) (2-10)
Since most of the NOX from combustion processes occurs as NO,32
reactions 2-6 and 2-7 predominate. The clean gas passes out of the top
of the absorber to a heater for plume buoyancy and is sent to the stack.
The absorbing solution drops to a holding tank where makeup KOH and/or
KMnOit are added. This solution flows to a centrifuge to separate the
solid Mn02 which is then electrolytically oxidized to MnO^. The remaining
solution is either concentrated in an evaporator to form a weak KN03 solu-
tion or is electrochemically treated to produce a weak HN03 solution and a
mixed stream of KOH and KN03.
The fundamental design equation used for gas absorption column design
is
(2-11)
2-18
-------
where y = bulk NO concentration (mole fraction) of gas phase at any
given point in column
y-y* = overall driving force for absorption (y* being the NO con-
centration of a gas in equilibrium with a given liquid NO
concentration)
Y, = inlet NO concentration
b x
Y = outlet NO concentration
a x
K = overall gas-phase mass transfer coefficient, Ib-moles NO /
(ft2)(hr)(mole fraction)
a = area of gas-liquid interface per unit packed volume, ft2/ft3
G = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
Z = length of packed section of column, ft
In a column containing a given packing or plate configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow rate/
cross-sectional area of column) is called the flooding velocity. At this
point, the gas flow completely impedes the downward motion of the liquid
and blows the liquid out of the top of the column. The gas velocity, obvi-
ously, must be lower than the flooding ve1ocity. How much lower is a design
decision. Often, it is an economic tradeoff between power costs and equip-
ment costs. A low gas velocity will lower the pressure drop and, hence, the
power costs but the absorber will have a larger diameter and cost more. High
gas velocities have an opposite effect. Usually the optimum gas velocity is
about one-half the flooding velocity.3 "* The height of the column depends on
the desired level of removal and on the rate of mass transfer. The latter
is a major problem for these systems trying to achieve large NOV reductions
X
since NO is relatively insoluble in water. This can be seen in Table 2.2.4-1.
TABLE 2.2.4-1. NITROGEN OXIDES CHARACTERISTICS35
Boiling Point, Solubility in Cold Solubility in Hot
°C Water (0°C), cm3 Water (60°C), cm3
NO
NO 2
-151. '8
21.2
7.34/100 cc H20
soluble, decomposes
2.37/100 cc H20
2-19
-------
One can see that NO has a very limited solubility in water and, since most
NOX is present as NO, the rate of mass transfer (absorption) is going to be
relatively slow. This means that the absorber must be tall with a high
liquid flow rate. Table 2.2.4-2 presents the effects of boiler/flue gas
variables on the design of absorption-oxidation systems.
TABLE 2.2.4-2. SYSTEM DESIGN CONSIDERATIONS
Variable Design Effect
Presence of particulates Requires prescrubber
Presence of SOa Requires FGD pretreatment
Increased gas flow Requires larger column diameter; increased
liquid flow rate
Increased NOX concentration Requires larger column height; increased
oxidant concentration
Both flue gas flow rate and NOX concentration can be affected by boiler
operating conditions. Therefore a change in load on an industrial boiler
may alter these variables markedly. The absorber must be designed to accom-
modate any anticipated load changes. The column size and the liquid and
oxidant flows must be designed for each application after examining the
boiler operating history and establishing ranges of variation.
None of the sources consulted for this study could supply typical ranges
for operating variables such as liquid/gas ratio, reagent concentrations or
pressure drops and, as a result, none are presented here. Economic data were
not presented either. One source did estimate the removal for absorption-
oxidation processes to be 85 percent.36
Presently, absorption-oxidation processes are still in the pilot unit
stage of development. Table 2.2.4-3 presents a list of absorption-
oxidation process vendors and the status of development of their projects.
One can see from the table that no coal-fired flue gas tests have been
performed.
2-20
-------
TABLE 2.2.4-3. PROCESS VENDORS OF ABSORPTION-OXIDATION PROCESSES
37,38
Vendor Status of Development
Hodogaya No information available; stopped development
on process
Kobe Steel 1974: 1000 Nm /hr gas from iron-ore sintering
furnace; stopped development on process
MON (Mitsubishi Metal, MKK, 1974: 4000 Nm3/hr flue gas from oil-fired
Nikon Chemical) boiler
Nissan Engineering 1972: 4 pilot plants, 100-2000 Nm3/hr tail
gas from HNOs plant
2.2.4.2 System Performance—
No coal-fired tests have been made. No information has been published
on tests conducted with other fuels. The relative insolubility of NO in
water may present a major obstacle to achieving the stringent level of con-
trol (90 percent NO reduction) by absorption-oxidation processes. Another
X
primary drawback of absorption-oxidation systems is the production of nitrate
salts (see Equation 4-2), a secondary pollutant. These processes probably
could not be applied on a large scale as wastewater treatment systems
(chemical or biological) do not remove nitrogen compounds from the waste-
o q
water. Trying to recover the nitrates as nitric acid for industrial use
or potassium nitrate for fertilizer does not seem promising as the by-products
are of low quality. Also, the use of an expensive, liquid-phase oxidant
requires stainless steel and other corrosion resistant materials of construc-
tion. High sulfur coals require an FGD system prior to the NO absorber to
prevent excessive oxidant consumption by S02- The process steps of several
absorber columns in series (large fan requirements), oxidant regeneration
(electrolysis), and flue gas reheat (inline heater) are all energy intensive
and present technical and economic disadvantages.
2-21
-------
2.2.5 Selective Catalytic Reduction-N0x/S02 Removal
2.2.5.1 System Description—
From a NO removal standpoint, this process is very similar to those
X
discussed in Sections 2.2.2 and 2.2.3. The primary difference is the addi-
tional equipment necessary to collect and process the SOz- The main feature
of the process is the reactor and catalyst which remove both NOX and SO2.
This process was developed by Shell although the U.S. licensor, UOP, is
currently marketing and developing the process. The N0x/S0a version of the
process is commonly called the SFGT process which stands for the Shell Flue
Gas Treatment process.
A uniquely designed parallel flow type of reactor is used to avoid
problems with particulates. The reactor consists of a series of packages
containing catalyst material, arranged in a parallel configuration which
allows flue gas flow between the packages. Each package consists of catalyst
material placed between two layers of wire gauze. Figure 2.2.5-1 illustrates
the internals of the parallel passage reactor. The flue gas flows between
the catalyst packages and not directly through the catalyst material. This
prevents plugging of the catalyst with particulate matter in the flue gas.
For convenient fabrication and handling, catalyst packages of a standard
size are appropriately spaced and placed in a container to form a unit cell
or module. S02 removal efficiency and capacity are determined by the number
of unit cells placed in series in a cell stack. For a given level of S02
removal, a greater number of cells in the stack increases the capacity and
reduces the frequency of regeneration. The number of stacks is determined
largely by the flue gas rate and the flue gas velocity through a single stack
is generally not a design variable. For most design situations, 4 to 5 unit
cells in a stack are adequate to achieve high S02 removal, however, if a high
level of denitrification is required, more unit cells per stack may be neces-
sary .
2-22
-------
F1EGEN. GAS I t.
PURGE OFF-GAS
REGEN. OFF-GAS
TREATED
E GAS
FLUE GAS
PURGE STEAM
UOP 1S3-3
Figure 2.2.5-1. The SFGT parallel flow reactor.
1*0
The SFGT process is a dry process with two or more reactors operating
in a cyclic manner. The desulfurization aspect of the process is regenerable,
while NO removal is accomplished by catalytic reduction with ammonia. The
catalyst material is commonly called an acceptor since SO2 removal involves
adsorption or "acceptance" of S02. The desulfurization cycle consists of
the following steps:
1) oxidation of acceptor bed/acceptance of S025
2) purge reactor,
3) regeneration with reducing gas, and
4) purge reactor.
The products of the oxidation and acceptance reactions in step 1 above
catalyze the reaction of NO with ammonia to form nitrogen and water. NO
X X
removal is accomplished by metering ammonia into the untreated flue gas
upstream of the reactors. The catalytic reaction takes place across the
partially spent acceptor beds.
2-23
-------
Also associated with the SFGT process are facilities for generating
reducing gas and for the processing of S02 in regeneration off gases into
sulfur by-products. Figure 2.2.5-2 illustrates the process flow for a
typical SFGT system.
Boiler flue gas is withdrawn upstream of the air preheater and particu-
late removal device by the SFGT system fan and discharged to the reactor
inlets. The flue gas then flows through fixed bed reactors in open channels
alongside and in contact with the acceptor material. Ammonia is added to the
flue gas upstream of the SFGT system fan to insure complete mixing before the
flue gas enters the reactor.
Fresh acceptor material is elemental copper on an alumina support. This
is converted to the oxide form by flue gas oxygen shortly after initiation
of the acceptance cycle. S02 is removed by reaction with the copper oxide
and oxygen as the flue gas flows through the channels, converting the accep-
tor material to copper sulfate. Simultaneous with the desulfurization pro-
cess, the reduction of flue gas NO by ammonia is selectively catalyzed by
copper oxide and copper sulfate in the acceptor bed. As the flue gas leaves
the SFGT system reactors it is returned to the boiler flue gas duct down-
stream of SFGT fan suction.
Flue gas is fed to a reactor until an unacceptable amount of SOa begins
to pass through the reactor. This occurs when a large fraction of the accep-
tor has been converted to the sulfate form. Flue gas flow is then diverted
to another reactor and the spent reactor is isolated. Any flue gas remaining
in the spent reactor is purged with an inert gas such as steam, and the re-
generation cycle is initiated.
Regeneration is accomplished by passing a reducing gas through the bed
countercurrent to the direction of the flue gas flow. The reducing gas,
which is primarily hydrogen, reacts with the copper sulfate in the spent
reactor to convert it to elemental copper. An off gas of S02 and water is
2-24
-------
t-o
Ln
PARTICULATE REMOVAL
AND STACK
FLUE GAS
NH3
1
*TED
S
:GENE
:RATION
STEAM
NAPTHA
REFORMER
OFF
PRODUCT
(S,SOZ (t),OR
Figure 2.2.5-2. Flow diagram of the SFGT process.1*1
-------
produced by the reaction. After regeneration is complete, the reactor is
again purged with steam and is ready for another acceptance cycle. Regenera-
tion gas can be produced from a number of sources, but steam-naphtha reform-
tt 2
ing is proposed by UOP as being the most economical.
The regeneration off-gas treatment section consists of flow smoothing
equipment and SOa workup equipment. Typically, the regeneration off-gas is
cooled and most of the steam condensed, raising the SOa concentration from
10 percent to 80 percent by volume. The concentrated SOa is then compressed
into an intermediate holding vessel to provide a smooth flow rate to the
workup section. The workup section may be a modified Glaus unit which pro-
duces an elemental sulfur by-product, a f ractionation unit which produces
liquid SOa, or a sulfuric acid plant.
Each process step consists of different chemical reactions. The
is converted to the oxide form by the following reaction:
Cu + kO 2 ->• CuO (2-12)
This oxide readily reacts with flue gas SOa and oxygen, as described by:
CuO + %02 + S02 -»• CuS04 (2-13)
SO 3 in the flue gas is also removed by the following reaction:
CuO + S03 + CuSOtt (2-14)
The reaction scheme for reduction of NOX is described by the
following: X 2
4NO + 4NH3 + 02 £ 4N2 + 6H20 (2-1)
2N02 + 4NH3 + 02 t 3N2 + 6H20 (2-2)
2-26
-------
480
400
E
Ok
- 300
UJ
o
o
u
20O
IOO
AT REACTOR INLET:
FLOW • I37.OOO NmVh
SOt « 1260 ppmv
NO* i 293 ppmv
REACTOR BED LENGTH • 4 METER
NOX AT NHj /NO-QO
0 20 40 60 8O IOO 120
ACCEPTANCE TIME, MIN
Figure 2.2.5-3. SFGT reactor performance versus acceptance time.
4-if
2-27
-------
Excess ammonia which is not consumed in reactions 2-1 and 2-2 may be
catalytically oxidized to nitrogen and water by reaction with flue gas
oxygen, as described by:
4NH3 + 302 t 2N2 + 6H20 (2-15)
Maximum NOX removal efficiency is achieved at the point of S02 breakthrough,
where conversion of the acceptor material from the oxide to the sulfate form
is essentially complete. Figure 2.2.5-3 illustrates reactor outlet S02 and
NO concentrations during a typical SFGT acceptance cycle.
Copper sulfate is reduced to the elemental copper form by reducing gas
hydrogen as described by the following reaction:
CuSOi* + 2H2 + Cu + S02 + 2H20 (2-16)
Any acceptor material present in the reactor as the oxide will also be
reduced, according to the following reaction:
CuO + H2 + Cu + H20 (2-17)
The regeneration step occurs at the same temperature as the acceptance step,
400°C (750°F).
The general reactor design equation is the same as that described in
earlier sections for SCR processes. The primary variables are the gas rate,
reaction rate, and control level. Reaction rate data have not been released
for this process except that the NO reduction is first order.
X
The gas flow rate and control level will determine the reactor size.
Increases in either variable will increase the reactor volume. The effect
of control level can be seen in Figure 2.2.5-4. It is necessary for the
flue gas to enter the reactor at 400°C and therefore it must be taken from
2-28
-------
10
t
L
X
1
\
\
^
\
^
N
\
CM
\
\
^
i
\
\
\
\)
COND
400°
CuA
NH3
PEHF
• PERF
AFFE
OUT
\
\
\
•»
TIONS:
C
5 CuS04
/NO 1.1 ~
MALEXPEC
ORMANCE
ORMANCE
CTED BY
5IDE FACTO
V
\
\
1.S
RS
y
\
234
IED UNGTH. MCTCR1
Figure 2.2.5-4. Unconverted NOX as a function of catalyst bed length.1*5
an appropriate point in the boiler, most likely from between the economizer
and air preheater. Alternatively, a cooler gas can be heated to 400°C by an
inline heater.
The removal efficiency of NO for a given reactor size is determined
X
by the amount of NH3 injected as shown in Figure 2.2.5-5. Since the reac-
tion is first order in N() , control level is not affected by NO concentra-
x x
47
tion. The SOa control efficiency is primarily a function of the acceptance
time of the reactor (Figure 2.2.5-3). Typical ranges of operating variables
are shown in Table 2.2.5-1.
Since the SFGT system can handle full particulate loading (^10 gr/sft3)
it is not dependent on any pretreatment facilities. Also, the SFGT system
operation is independent of boiler operation. The system fan takes suction
from the flue gas duct between the economizer and air preheater and the reac-
tor discharge returns to the boiler flue gas duct just downstream of the
2-29
-------
100
90
80
70
* 60
| SO
J -
30
20
10
A
&
1
/
/
J
1
I
/
£?
• O O £
/ *
3
1.5 METER BENCH S
2000 SO}
4% Oj
TCCI CHSV
O3so eooo
O400 SOW
A 450 §000
£ALE UNIT
wl CM
NHj/NO MOLi RATIO
Figure 2.2.5-5. NOX reduction with NHs over commercial SFGT acceptor.
if 6
TABLE 2.2.5-1. DESIGN AND OPERATING VARIABLES FOR
SFGT SYSTEM48
Variable
Typical Range
Space Velocity
NH3:NOx Mole Ratio
Flue Gas Temperature
Pressure Drop
Maximum Particulate Loading
5,000 - 8,000 hr l*
1.0:1.0 to 1.2:1.0*
400°C
5-6 in. H20*
>23 g/Nm
*Actual value will depend on required removal level.
2-30
-------
suction point, with no valves between the two points. The system fan
provides a constant flow rate through the SFGT system. If the boiler flue
gas rate is greater than the fan rate, flue gas will bypass the system
through the open duct. If the boiler flue gas rate is lower than that of the
system fan, treated gas will recycle back to the system fan suction. Recycle
of treated gas to the reactor inlet with "open bypass" arrangement presents
no operating problems. This is due to the fact that both the level of
desulfurization and denitrification are independent of inlet concentrations,
and the system does not humidify the flue gas.
Tables 2.2.5-2 and 2.2.5-3 present test and commercial applications
of the SFGT process. The development history of the process can also be
seen in these tables.
In the U.S., from 1974 to 1976 a pilot scale unit at Tampa Electric
Company (TECO) was operated using flue gas from a coal-fired boiler.
Testing was for S02 removal only, NO control was not attempted during
this period. The process developer is currently modifying the TECO
pilot unit to accommodate 7 meters of bed height, up from the previous
maximum of 5 meters. This should permit bimultaneous removal of NO
and SOX to the 90 percent level. Also, provisions are being made for
injection of a CO/C02 gas mixture into the regeneration gas in order
to simulate medium-Btu gas from a coal gasifier.
The costs for an industrial size boiler have not been estimated. How-
ever, costs for a 500 MW utility boiler application are available and are
shown in Tables 2.2.5-4, 5, and 6. Also shown are the estimated energy and
raw material requirements.
2-31
-------
TABLE 2.2.5-2. SFGT PROCESS, PILOT AND DEMONSTRATION UNITS
Location/
Company
Shell Ref.
at Pernis
Rotterdam
Utility
Tampa Elec.
Big Bend
M
i
N>
JGC
Yokohama
Nippon
Steel
—
Fuel/
Designed By Application
Shell Residual
Fuel Oil-
Proc. Heater
Shell Coal-
Steam Boiler
UOP Coal-
Wet-Bottom
Utility Boiler
JGC* Fuel Oil
JGC Sintering
Furnace
JGC Coke Oven
Gas
Size, Type of
Nm3/hr Operation
600-1000 S0x-only
Heavy Fly Ash
Loading
1200-2000 S0x-only;
SOX-NOX
Simultaneous
250-700 N0x-only
2000 N0x-only
400 N0x-only
Dates Comments
1967-1972 SOX reduction -
approx. 90%
1971 Particulate mat-
ter - loadings to
20 Gr/Nm3
1974-1976 S0x - 90%;
1979- S0x-N0 - 90/90%
fly ash to
25 Gr/Nm3
1974- NOX reduction -
90-99%
1976-1978 NOV reduction -
A,
90-97%
1976-1977 NOX reduction -
90%; special low
temp . cat . evalua-
tion
*JGC Corporation, licensing agent in Japan.
-------
TABLE 2.2.5-3. SFGT PROCESS, COMMERCIAL UNITS
K5
CO
Co
Unit
SYS*
Yokkaichi
Kashima Oil
Co. Ltd.
Fuji Oil
Co. Ltd.
Nippon Steel
Corp .
Fuel/
Designed By Application
Shell Residual
Fuel Oil-
Ref. Boiler
JGC Fuel Oil-
Process Unit
Heater
JGC CO Boiler
JGC Sintering
Furnace
Size, Type of
Nm3/hr Operation Dates Comments
125,000 S0x-only; 1973-1975 SOX reduction - 90%;
NO -S0x 1975- Simultaneous - 90/50%
Simultaneous
50,000 N0x-only 1975- 95-98%
70,000 N0x-only 1976- 93-96%
150,000 N0x-only 1978- a-95% (low temp, cata-
lyst)
"Showa Yokkaichi Sekiyu
-------
TABLE 2.2.5-4. ECONOMICS OF SFGT SYSTEM
Incorporated Units:
Power Plant Size
Fuel
S-Content, Wt-%
Case 1
Case 2
Case 3
HHV
Heat Rate
Excess Air
Air Preheater Leakage
BASIS:
Steam-Naphtha Reformer
SFGD Reactor Section
Compressor/Gasholder Flow
Smooth Section
Modified Glaus Unit
500 MW
Coal
3.5
2.5
0.8
10,500 Btu/lb
9,000 Btu/kWh
20%
13%
Flue Gas Rate
SO2 Content, ppmv
Case 1
Case 2
Case 3
1,582,000 NmVh (983,000 SCFM)
2,580
1,850
590
Mid-1977, Gulf Coast Location
Load Factor
Capital Charges
Cost of:
Naphtha
Steam (40 psi, SAT.)
Electricity
Labor
Heat Credits
Sulfur
7,000 h/a
15%/a
$0.35/gal
$1.50/M Ib
$0.018/kWh
$10.00/hr
$2.50/MMBtu
$45.00/ton
2-34
-------
TABLE 2.2.5-5. ECONOMICS OF SFGT SYSTEM ESTIMATED
CHEMICALS AND UTILITY REQUIREMENTS
5 0
Case 1
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
Case 2
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
Case 3
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
SFGD
Section
kW 5,770
kmol/h 1,820
Gcal/h
Gcal/h
kg/h
kW 5 , 800
kmol/h 1,300
Gcal/h
Gcal/h
kg/h
kW 5,120
kmol/h 480
Gcal/h
Gcal/h
kg/h
Flow Mod.
Smooth Glaus
Section Section
850 115
-380* -740*
5250
570 82
-270* -530*
3760
180 30
-95* -170*
1200
Reformer
Section Total
480 7215
-600* 100
90.92 90.92
42.53
5250
300 6782
-415* 85
62.75 62.75
32.48
3760
110 5440
-140* 75
21.01 21.01
18.46
1200
*Produced
**40 psig, Saturated
***5.175 MMBtu/Bbl produces 11,500 SCF Hydrogen/Bbl
2-35
-------
TABLE 2.2.5-6.
ECONOMICS OF SFGT SYSTEM ESTIMATED
CAPITAL AND OPERATING COST51
EEC. (MM$)
SFGD Reactor Section
Compressor /Gasholder
Modified Glaus
Steam-Naphtha Reformer
Estimated Annual Revenue
Requirements (M$/a)
Capital Charges
Maintenance
Labor
Acceptor
Electricity
Steam
Naphtha
Heat Credits
Sulfur Credits
Capital Cost, Operating Cost,
Energy Requirement
Capital Cost, $/kW
Operating Cost, C/kWh
Energy Requirement, Btu/kWh*
Case 1
28.95
7.82
2.76
8.81
7251
967
123
1479
909
42
7174
-2977
-1570
97
0.38
525
Case 2
28.53
6.10 -
2.26
7.14
6604
881
123
1053
855
35
4951
-2273
-1126
88
0.32
371
Case 3
22.94
2.65
1.14
4.17
4634
618
123
411
685
31
1658
-1292
-359
62
0.19
124
^Defined as the sum of;
Electricity at
Steam at
Naphtha at
Heat Credits at
9000 Btu/kWh
40000 Btu/kmol
4 Btu/kcal
4 Btu/kcal
2-36
-------
2.2.5.2 System Performance—
NOX control by the SFGT process is shown graphically in Figure 2.2.5-5.
As can be seen, at a space velocity of 8000 hr 1, NO control of >80 percent
X
can be achieved. Figure 2.2.5-4 indicates that the process developers feel
the process to be capable of NOX control levels of >90 percent.
Several different test series were conducted using the TECO pilot plant
and the operating conditions for these tests are shown in Table 2.2.5-7.
The S02 removal efficiency for several of these runs is shown in Figure
2.2.5-6 plotted against the number of cycles, which can be converted to time.
No data of this type are available for NOX control using coal-fired flue gas,
however, these data should be available in about one year.
As mentioned earlier, the system is not impacted by changes in the
boiler gas rate or particulate concentrations. Changes in the NO concen-
tration due to boiler load changes can be compensated for by a conventional
control system used in conjunction with the NHs injection equipment. This
control system will be developed during the upcoming pilot tests at the TECO
pilot plant.
2.2.6 Adsorption
2.2.6.1 System Description.—
The adsorption process removes NO and SOa from flue gas by adsorbing
X
them onto a special activated char. Adsorbed NO is reduced to Na while SOa
is reduced and condensed to an elemental S by-product.
A process flow diagram is shown in Figure 2.2.6-1. Flue gas is taken
from the boiler air preheater and passed through a particulate removal device
to prevent blinding of the adsorption bed. The flue gas then enters the ad-
sorber, a vertical column with parallel louver beds containing the char in
2-37
-------
TABLE 2.2.5-7. SUMMARY OF BASE OPERATING CONDITIONS ON THE SFGT PILOT
PLANT AT TECO52
Run No.
Duration,
Cycles
Months
Cumulative Cycles
Flow Rate
Ace. Time
Reg. Time
Flue Gas
Eff. SOR
Eff. EOR
, SCFM
, Min.
, Min.
Source*
1
5
2488
2488
1090
20
20
1
2
2%
1520
4008
1090
20
20
1
92
82
3
1%
1292
5300
1090
20
20
2
95
95
4
2
1412
6712
1090/1420
20
20
1
95
80
5
5
4328
11040
1090
20
20
1
95
92
6
3
2210
13250
1090
20
20
3
93
93
100
o
2
.RUN NO. 3-
!jK-WtoSK^;a::u»«%3Va&:tf YW.CJESl A>" f-f. JV£. :TpWSK3»ti«»
I i i RUN NO. G
•_^WL^,tfJ-J-^TTl1*c^Wcj£KaS3JKEaM^:'EU^
400 600 800 1000 1200 1400 1600
CYuLES UOP1S3-H
Figure 2.2.5-6.
removal efficiency vs. cycles.
5 3
2-38
-------
K3
I
LO
FLUE
GAS
AIR
STACK
* ADSORBER -»i
,
'
:RA
i
' .
TC
Ft
>
<
,r
HEATER
CRUSHED
COAL
!
RESOX
REACTOR
CONDENSER
ASH
HEAT
SULFUR
,TO FLUE GAS
> ENTERING AIR
EXCHANGER, HEATER
AIR
Figure 2.2.6-1. Flow diagram of Foster Wheeler-Bergbau Forschung
Dry Adsorption Process. "*
-------
pellet form. NOX and S02 are adsorbed on the char which slowly moves
downward through the bed. The NOX adsorption mechanism is unknown but S02
undergoes the following reaction.
SQ2(g) + H20(g) + %02(g) + HzSOitCL) (2-18)
The reaction product is held in the pores of the char pellets. The flue
gas exits the adsorber and passes to the stack. The saturated char leaves
the bottom of the adsorber and is screened to remove any fly ash deposits.
It is then conveyed to a regenerator where it is mixed with hot sand (650°C)
and the following reactions take place. '
2H2S(K(1) + C(s) -> C02(g) + 2H20(g) + 2S02 (g) (2-19)
2NO(g) + C(s) + C02(g) + N2(g) (2-20)
This S02-rich gas product stream is sent to an off-gas treatment reactor
containing hot, crushed coal (650-820°C) and the following reactions take
place. 6
S02(g) + S(g) + Oz.(g) (2-21)
C(s) + 02 (g) + C02(g) (2-22)
The gas then passes to a condenser where the S vapor forms molten S. The
char/sand mixture from the regenerator is screened to separate the two solids.
The char is recycled to the adsorber via a spray cooler and the sand is re-
cycled to the regenerator after passing through a heater.
This process operates at 120-150°C, however, typical values for other
operating variables were not found. NO and S02 control levels were reported
X
to be 40-60 percent and 80-95 percent, respectively.57 The economics of the
process vary with the fuel sulfur level. For fuel sulfur levels of 0.9-4.3
2-40
-------
percent, the capital costs range from $40-90/kW and the operating costs range
from 1.0-2.3 mills/kWh. The costs were based on applying the process to a
utility boiler of >200 MW capacity.
Presently, the adsorption process is in the prototype unit stage of
development. The one reported process developer in the field, Foster Wheeler-
Bergbau Forschung has a 20 MW prototype unit and several small pilot plants
treating coal-fired flue gas.
2.2.6.2 System Performance—
Tests have shown the adsorption process to be primarily a S02 reduction
process as NC) removal efficiency averages 40-60 percent while SOa removal
X
had a range of 80-95 percent.
The primary drawback of this process, besides the low NC) removal level,
X
is its complexity: numerous process steps involving hot solids handling.
Solids flow can be difficult to control and high maintenance requirements
could be expected. The vendor has reported several mechanical problems
during testing which included control of adsorber-bed levels, poor char
distribution, char-sand separation, hot sand conveying, and char cooling
and feed. Some corrosion-resistant material is needed in the high tempera-
ture zones of the process. The ash waste stream from the off-gas treatment
reactor appears to be the sole secondary pollutant associated with the pro-
cess. The overall complexity and low N0y removal of the process present
definite technical disadvantages.
2.2.7 Electron Beam Radiation
2.2.7.1 System Description—
This dry process utilizes an electron beam to bombard the flue gas,
removing NO and S02 in the process. A block flow diagram for the process
is shown in Figure 2.2.7-1.
2-41
-------
Electron Beam
Accelerator
Flue
Gas
.Reactor
Fly Ash
Off-Gas
Solid
Residue
By-product
Treatment
I
Disposable or
Salable By-product
Figure 2.2.7-1. Process flow diagram for Ebara-JAERI
electron beam process.
60
Flue gas is taken from the boiler air preheater and passed through a
cold ESP to remove particulates. After a small amount of ammonia is added,
the gas enters a reactor where it is bombarded with an electron beam. (The
penetration of the gas stream by the beam will require a unique discharge
pattern or other special design considerations.) A powder containing both
ammonium nitrate and sulfate is generated by an unknown reaction mechanism.
The gas then exits the reactor, passes through a second ESP to remove the
solid by-product, and is sent to the stack. The by-product treatment system
is still being developed. Various methods investigated include thermal de-
composition in the presence of an inert gas, steam roasting with CaO, or
steam roasting with HaO. The by-product may eventually be useful as a fer-
tilizer.51
2-42
-------
The key subsystem of this process is the electron beam accelerator.
Control of this unit's power supply is based upon inlet composition, flow
rate, and temperature of the flue gas.
Some of the important variables and typical ranges are listed in
Table 2.2.7-1.
TABLE 2.2.7-1. SYSTEM VARIABLES62
Typical Value
Temperature
Reactor residence time 1-20 sec
Radiation rate 105-106 rad*/sec
Total radiation absorbed 1-3 Mrad*
*Rad is the radiation dose absorbed
1 rad = .01 J/Kg
The operating cost with NO removal only (low sulfur coals) is lower
due to lower radiation levels, but the capital cost would be just as high
as for simultaneous NO /SO removal. Capital costs are quite high for this
X X
process as the 2 ESP's and the accelerator are expensive. The cost for a
1000 Nm3/hr test unit are reported to be $1000/kW, however, the cost of a
full scale system is expected to be lower. Operating costs are not
available.
No coal-fired tests have been performed at this time. The Ebara
Manufacturing Company in conjunction with Japan Atomic Energy Research
Institute (JAERI) has operated a 1000 Nm3/hr pilot plant treating flue
gas from an oil-fired boiler. In 1976, a 3000 Nm3/hr pilot plant began
treating off-gas from an iron ore sintering furnace at Nippon Steel.
By-product treatment technology needs to be more fully developed before
this process can be applied commercially.
2-43
-------
In the U.S., the Department of Energy (DOE) is funding development of
an electron beam process offered by Research-Cottrell. Pilot unit tests
with flue gas are scheduled, however, the details of the program are not
yet available.
2.2.7.2 System Performance—
No coal-fired testing has been done.
A summary of the oil-fired pilot tests is shown in Figure 2.2.7-2.
100
60
of
i-i
«9 60
o
NO
2 3
Total beam (Mrad)
Figure 2.2.7-2. Oil-fired pilot plant test results.
One can see that N0x/S02 removal drops off drastically at a total radiation
dose below 1 Mrad while the maximum removal is obtained at about 3 Mrad.
The removal efficiencies decrease as the concentrations of NO and S02
• X
increase as can be seen in Figure 2.2.7-3.
2-44
-------
100
90
80
S 70
z
LJ
O
3 40
O
S 30
cc
20
10
0
I
I
I
I
2O
400
600 800
1000 1200
1400
1600
CONCENTRATION OF NOX OR S02 , PPM
Figure 2.2.7-3. Effect of pollutant concentration on removal efficiency.
65
2.2.8 Absorption-Reduction
2.2.8.1 jjystem Description—
Absorption-reduction processes simultaneously remove NO and SOa from
flue gas by absorbing them into a scrubbing solution. The processes are
based on the use of chelating compounds, such as ethylenediamine tetraacetic
acid (EDTA) complexed with iron, to "catalyze" the absorption of NCL. Most
X
process vendors prefer a perforated-plate type of gas-liquid contactor. The
advantages of a perforated-plate absorber over a packed bed absorber include
easier cleaning when solids are present, wider operating ranges, and more
economical handling of high liquid rates.66 An example of a perforated plate
absorber is shown in Figure 2.2.8-1. The most common design of a perforated
plate is one that employs liquid crossflow over the face of the plate with
the gas passing upward through the plate perforations. A schematic of the
2-45
-------
operation of a crossflow perforated plate is shown in Figure 2.2.8-2.
The liquid is prevented from flowing through the plates by the upward flow
of the gas. However, during periods of low gas flow (such as load changes
on industrial boilers) liquid can drain through the openings in the plates.
This reduces the liquid's time of contact with the gas on each plate and may
decrease the overall operating efficiency of the absorber. To prevent this
problem, there are two other types of dispersers utilized besides the basic
sieve-plate: the valve-plate and the bubble cap, depicted in Figure 2.2.8-3.
As the gas flow lowers, the valve or cap settles, sealing off the perforation
so liquid cannot drain through. This design feature allows the perforated
plate absorber to operate more efficiently at widely fluctuating gas rates.
RUE
-------
Plate n-\
Plate n
Figure 2.2.8-2.
Normal operation of sieve plate. Za, height of
station a above datum. Zcr, weir crest. Z^,
liquid-friction head. Zp, pressure head across
plate. Z^, net head in down pipe. Z^, weir
height.67
t t
Cos flow
Valve-plate dispenen.
•Valve closed
Valve open
Holes, punched
2 to 4 in. diom.
(a) (t>)
(a) Circular or bell cap. (fc) Tunnel cap.
Bubble cap dispersers
Figure 2.2.8-3. Other gas dispersers.
6 8
2-47
-------
While most all absorption-reduction processes utilize ferrous chelating
compounds to enhance NO absorption, the scrubbing solutions, the by-product
treatment and sorbent regeneration chemistry differ from process to process.
For this reason, one of the simpler absorption-reduction processes, that of
Kureha Chemical Industry Company, is examined here in detail.
A block flow diagram of the Kureha absorption-reduction process is
shown in Figure 2.2.8-4. Flue gas is taken from the boiler after the air
preheater. It passes through a prescrubber to adiabatically cool the gas
and remove both particulates and chlorides. The flue gas then enters the
distributing space at the bottom of the NOX/S02 absorber, below the plates
or packing. The gas flows upward, countercurrent to a sodium acetate
(CHsCOONa) scrubbing solution (^60°C) containing ferrous iron and EDTA and
a few seed crystals of gypsum (to prevent scaling). Most of the S02 is
rapidly absorbed at the bottom of the absorber according to the following
reactions.7
S02(g) -»• S02(aq) (2-23)
S02(aq) + 2CH3COONa(aq) + H20 -> Na2S03(aq) + 2CH3COOH(aq) (2-24)
The NOX (which consists mainly of NO) is relatively insoluble; therefore, it
is absorbed gradually over the length of the column. The ferrous chelating
compounds effect on NO absorption is described in Figure 2.2.8-5. The NOX
is absorbed and undergoes the following reactions.73
N0(g) -»• NO(aq) (2-6)
2N02(g) + N204(aq) (2-25)
N20n(g) -> N2(Maq) (2-9)
2-48
-------
Water
I
-P-
Gypsum
Figure 2.2.8-4. Process flow diagram of Kureha absorption-reduction process.69'70
-------
Cl
ro
o
CO
cc
o
CO
CO
<
o
I
0 0.01 0.02
EDTA-Fe(II), mole/1 Her
Figure 2.2.8-5. EDTA-Fe(II) concentration and NO absorption at 50°C.
72
2NO(aq) + 5Na2S03(aq) + 4CH3COOH(aq) + 2NH(S03Na)2 (aq)
+ 4CH3COONa(aq) + H20 (2-26)
2N2(V(aq) + 7Na2S03(aq) + 4CH3COOH(aq) -> 2NH(S03Na)2 (aq) + 3Na2SCh (aq)
+ 4CH3COONa(aq) + H20 (2-27)
Some of the acetic acid (CH3COOH) formed at the bottom of the absorber via
reaction (2-24) is vaporized. It must be captured and is done so by water
scrubbing at the very top of the absorber. From the top of the absorber
column the clean flue gas passes to a heater for plume buoyancy and is then
sent to the stack. The liquid effluent drops from the bottom of the absorber
to a gypsum, CaSOit*2H20, production reactor. Here, the solution is mixed with
with the purge stream from the acetic acid recovery section and a lime slurry
2-50
-------
stream. The lime, Ca(OH)2, treatment involves the following reactions.7"*
2CH3COOH(aq) + Ca'(OH) 2 (aq) -> (CH3COO)2Ca (aq) + 2H20 (2-28)
(CH3COO)2Ca(aq) + Na2S(H(aq) + 2H20 + CaS04 -2H20(s) 4- + 2CH3COONa(aq) (2-29)
The gypsum formed by reaction 2-29 is centrifuged. Most of the liquor
discharged is returned to the gypsum reactor and on to the absorber. The
remaining liquor is sent to a reactor where sulfuric acid (H2SOO is added
to hydrolyze the imidodisulfonate, NH(S03Na)2, by the following reaction.75
H+
NH(S03Na)2(aq) + 2H20 O NH^HSCK (aq) + Na2SOi,(aq) (2-30)
The effluent from this reactor is then recycled to the gypsum production
reactor. A small purge stream is taken from the gypsum reactor to another
reactor where the ammonium bisulfate (NHifHSOif) formed in the hydrolysis
reaction is treated with lime to yield gypsum and NH3 off-gas by the follow-
ing reaction.75
(aq) + Ca(OH)2(s) + CaSCK •2H20(s) 4- + NH3 (g)+ (2-31)
The gaseous ammonia is stripped from the solution by an air stream. If no
use for the ammonia can be found, the gas mixture is sent to a catalytic
reactor where ammonia is oxidized by the following reaction.
4NH3(g) + 302(g) 2N2(g) + 6H20(g) (2-32)
350°C
The product stream is then sent to the deacetating section of the absorber
column .
The fundamental design equation used for gas absorption column design
is32
2-51
-------
f
J V
(y-y*)
(2-11)
where y = bulk NOX concentration (mole fraction of gas phase at any
given point in column
y_y* = overall driving force for absorption (y* being the NOX concen-
tration of a gas in equilibrium with given liquid Npy
concentration)
Y, = inlet NC> concentration
D x
Y = outlet NOV concentration
a
x
K = overall gas-phase mass transfer coefficient, Ib-moles N0x/
(ft2)(hr)(mole fraction)
a = area of gas-liquid interface per unit packed volume, ft2/ft3
G = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
Z = length of packed section of column, ft
In a column containing a given plate or packing configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow rate/
cross-sectional area of column) is called the flooding velocity. At this
point, the gas flow completely impedes the downward motion of the liquid and
blows the liquid out of the top of the column. The gas velocity, obviously,
must be lower than the flooding velocity. How much lower is a design deci-
sion. Often it is an economic tradeoff between power costs and equipment
costs. A low gas velocity will lower the pressure drop and, hence, the
power costs but the absorber will have a larger diameter and cost more.
High gas velocities have an opposite effect. Usually the optimum gas
velocity is about one-half the flooding velocity.33 The height of the
column depends on the desired level of removal and on the rate of mass
transfer. The latter consideration is the reason why a chelating compound
is used in absorption-reduction processes to aid in NO absorption. Table
2.2.8-1 presents the effects of boiler/flue gas variables on the design of
2-52
-------
absorption-reduction systems. Both flue gas flow rate and NO concentration
can be affected by boiler operating conditions. Therefore a change in load
on an industrial boiler may alter these variables markedly. The absorber
must be designed to accommodate any anticipated load change. The column
size and the liquid flows must be designed for each application after
examining the boiler operating history and establishing ranges of variation.
TABLE 2.2.8-1. SYSTEM DESIGN CONSIDERATIONS
Variable Design Effect
Presence of particulates Requires prescrubber
Presence of SOz Requires S02:NO mole ratio of at least
3-5 (depending on process) for absorption-
reduction to be effective.
Increased gas flow Requires larger column diameter; increased
liquid flow rate
Increased NO concentration Requires larger column height; increased
catalyst concentration
The process vendors have not released much information on the operating
conditions of these processes. This is primarily due to the competitive
status of these similar processes at this early stage of development. Typi-
cal values for some of the process variables are shown in Table 2.2.8-2.
TABLE 2.2.8-2. TYPICAL VALUES FOR PROCESS VARIABLES
OF ABSORPTION-REDUCTION PROCESSES78
Variable Range
Liquid/Gas ratio, 1/Nm3 10-30
S0x/N0x mole ratio 2.5-3.C
Superficial Gas Velocity, m/s 1-3
2-53
-------
Cost estimates for this type of process cover a large range, presumably
due to the differences in sorbent regeneration technique. Capital costs
for utility applications are reported to range from $65-127/kW and operating
costs from 4.8-7.4 mills/kWh.79
Presently, absorption-reduction processes are in the pilot-unit stage
of development. Table 2.2.8-3 presents a list of absorption-reduction
process vendors and the status of development of their projects. One can
see from the table that only one coal-fired flue gas test has been performed.
TABLE 2.2.8-3. PROCESS VENDORS OF ABSORPTION-REDUCTION PROCESSES80
Vendor Status of Development
Asahi 1974: 600 Nm3/hr flue gas from residual oil-
fired boiler (1000 hours continuous).
Chisso 1975: 300 Nm3/hr flue gas from oil-fired boiler
(335 hours continuous)
Kureba 1976: 5000 Nm3/hr flue gas from heavy oil-fired
boiler (3000 hours continuous)
Mitsui Engineering and 1974: 150 Nm3/hr flue gas from oil-fired boiler
Shipbuilding
Pittsburgh Environmental 1976: 3000 Nm3/hr flue gas from coal-fired
boiler (52 hours continuous, absorption section
_ only)________ ___ _^
2.2.8.2 System Performance—
The one coal-fired test showed 60-70 percent NO and 90 percent SOz
X
reductions are possible. The longest continuous operation was for 52 hours
and the absorption section was the only part of the process tested. Pilot-
plant testing was discontinued after two months. Plans are being made for
further coal-fired pilot tests on the integrated system.
2-54
-------
Absorption-reduction processes are readily applicable only to high
sulfur coals as a S02:NO^ mole ratio in the flue gas of at least 3-5 is
X
required for maximum performance. This can easily be shown by observing
reactions 2-24 and 2-26 reprinted below.
S02(aq) + 2CH3COONa(aq) + H20 + Na2S03(aq) + 2CH3COOH(aq)
2NO(aq) + 5Na2S03(aq) + 4CH3COOH(aq) + 2NH(S03Na) 2 (aq) + Na2S(H(aq)
+ 4CH3COONa(aq) + H20
One can see that 1 mole of S02 absorbed in solution reacts to form 1 mole of
sodium sulfite (Na2S03). Then, 5 moles of sodium sulfite are required to
reduce 2 moles of NO. So, the minimum stoichiometric S02:NO mole ratio
5 x
required is y or 2.5. Also, some of the sodium sulfite is oxidized to
sodium sulfate by oxygen present in the flue gas according to:
Na2S03(aq) + hOz (aq) + Na2SOif(aq) (2-33)
and is not available for NO reduction. Low-sulfur coals would require S02
to be added to the flue gas for these processes to perform; therefore, they
should be considered applicable to high sulfur coals only.
Absorption-reduction processes require large absorbers with high liquid
rates due to relative insolubility of NO, even when the absorption catalyst
is used. Also, the regeneration of the absorption catalyst and the flue gas
reheat for plume buoyancy are energy intensive. Some corrosion-resistant
material is necessary due to the corrosive nature of the absorbing solution.
However, absorption-reduction appears to be the most promising of the "wet"
N0x/S02 removal processes. This is due primarily to its not utilizing oxi-
dants which require much corrosion-resistant material and, more importantly,
create serious secondary pollution problems. Also, the primary by-products
of absorption-reduction processes, gypsum, can be used as landfill material
2-55
-------
or in building materials. For all the above reasons, absorption-reduction
processes appear, at this preliminary stage, competitive with other wet
N0x/S0z removal processes.
2.2.9 Oxidation-Absorption-Reduction
2.2.9.1 System Description—
Oxidation-absorption-reduction processes simultaneously remove NOV and
X
from flue gas by oxidizing relatively insoluble NO to relatively soluble
and then absorbing both N02 and S02 into a scrubbing solution. The pro-
cesses are based on the use of gas-phase oxidants, either ozone (Oa) or
chlorine dioxide (ClOa), to selectively oxidize NO to NOz. Both perforated-
plate and packed bed absorption columns are utilized by various process
vendors.
Most of the oxidation-absorption-reduction processes are similar in
that they consist of five major sections:
prescrubbing
gas-phase oxidation
N02/S02 absorption
reduction of absorbed NO., and oxidation of S07
X
• wastewater treatment
The areas where processes differ are gas-phase oxidation - 03 or C102;
absorption solutions - limestone slurry (CaCOa), HzSOit, or NaOH; and the
amount and type of waste treatment required. Thermal decomposition, bio-
logical denitrification, or wastewater evaporation wastewater treatment
systems can je used. Because of these differences, only one of the oxidation-
absorption-reduction processes, that of Mitsubishi Heavy Industries, is
examined here in detail.
2-56
-------
A block flow diagram of the MHI oxidation-absorption-reduction process
is shown in Figure 2.2.9-1.
Gypsum
NKuOH
Figure 2.2.9-1. Process flow diagram for MHI oxidation-
absorption-reduction process.
2-57
-------
Flue gas is taken from the boiler after the air preheater and passed through
a prescrubber to cool the gas and remove particulates and chlorides. The
flue gas then enters a duct where it is injected with ozone (about 1 percent
by weight in air)82 such that the Os: NO ratio is 1:1. Ozone selectively
8 3
oxidizes NO by the following reatcion.
N0(g) + 03(g) + N02(g) + 02(g) (2-34)
After injection, the flue gas passes countercurrent to a lime/limestone
slurry in a grid-packed absorption column. A water-soluble catalyst is
added to the slurry to enhance N02 absorption (even though N02 is more
soluble than NO, it is still less soluble than S02). S02 is absorbed quickly
at the bottom of the column and undergoes the following reactions.15
S02(g) -»• S02(aq) (2-23)
S02(aq) + CaC03(s) + %H20 -* CaS03 •JSH20(s) + C02 (g) (2-35)
S02(aq) + CaS03(aq) + H20 + Ca(HS03 ) 2 (aq) (2-36)
N02 is absorbed gradually over the length of the column and reacts as
follows.15
2N02(g) + Ca(OH)2(s) + CaSOs -JsH20(s) + %H20 + Ca(N02)2(aq) + CaSO^ 2H20(s)
(2-37)
Once both the N02 and S02 are absorbed, the nitrite ion formed by reaction
2-37 is reduced by the bisulfate ion formed by reaction 2-36. 8if
Ca(N02)2(aq) + 3Ca(HS03 ) 2 (aq) + 2Ca[NOH(S03)2 ] (aq) + 2CaS03 -i;>H20(s) 4- + H20
(2-38)
2-58
-------
These hydroxylamine [NOH(S03)2] compounds are reduced further by the sulfite
ion85
Ca[NOH(S03)2](aq) + CaS03'%K20(s) + -^ H20 + Ca[NH(S03) 2 ] (aq) + CaSO^ -2H20(s)4-
(2-39)
Upon leaving the top of the absorber, the clean flue gas is reheated for
plume buoyancy and sent to the stack. The slurry solution drops to a holding
tank from which most of the solution is returned to the top of the absorber.
A small stream passes to a neutralization reactor where sulfuric acid is
fl c
added to convert the sulfite solid to soluble bisulfite and solid gypsum.'
(aq)' + H20 -»• CaSCK '2H20(s) 4- + Ca(HS03)2 (aq)
(2-40)
This stream passes to a thickener from which the bottoms are sent to a
centrifuge to separate the solid gypsum by-product from the liquor which is
returned to the absorber. The overflow from the thickener is primarily
recycled to the limestone slurry preparation tank. The remainder is sent
to a thermal decomposer where sulfuric acid is added to hydrolyze the N-S
compounds . l 8
H+
2Ca[NH(S03)2](aq> + 2H20 -»- Ca(NH2S03)2 (aq) + Ca(HSCK)2 (aq) (2-41)
TJ_1_
Ca(NH2S03)2(aq) + Ca(HS(K)2 (aq) + 6H20 ^ 2NHitHSO
-------
• decompose by increasing pH
decompose thermally
strip out with makeup HzO
The remaining gypsum slurry is pumped to the limestone slurry preparation
tank.
The fundamental design equation used for gas absorption column design
is32
(y-y*)
(2-11)
where y = bulk NO concentration (mole fraction) of gas phase at any
given point in column
y-y* = overall driving force for absorption (y* being the NO concen-
X
tration of a gas in equilibrium with a given liquid NOX con-
centration)
Y, = inlet NO concentration
b x
Y = outlet NO, concentration
a x
Kv = overall gas-phase mass transfer coefficient, Ib-moles N0_/
y *•
(ft2)(hr)(mole fraction)
a = area of gas-liquid interface per unit packed volume, ft2/ftc
y = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
Z = length of packed section of column, ft
In a column containing a given plate or packing configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow
rate/cross-sectional area of column) is called the flooding velocity. At
this point, the gas flow completely impedes the downward motion of the
2-60
-------
liquid and blows the liquid out of the top of the column. The gas velocity
obviously, must be lower than the flooding velocity. How much lower is a
design decision. Often, it is an economic tradeoff between power costs and
equipment costs. A low gas velocity will lower the pressure drop and, hence,
the power costs but the absorber will have a larger diameter and cost more.
High gas velocities have an opposite effect. Usually the optimum gas veloc-
ity is about one-half the flooding velocity.33 The height of the column
depends on the desired level of removal and on the rate of mass transfer.
The latter consideration is why oxidation-absorption-reduction processes
oxidize NO to more soluble NOz prior to the absorber and why some processes
add water soluble catalysts to the scrubbing solution to aid NOa absorption.
The oxidation step enables these processes to use shorter absorbers with
lower liquid rates than either the absorption-oxidation or absorption-reduc-
tion processes. Table 2.2.9-1 presents the effects of boiler/flue gas
variables on the design of oxidation-absorption-reduction systems. Both
flue gas flow rate and NO concentration can be affected by boiler opera-
ting conditions. Therefore a change in load on an industrial boiler may
alter these variables markedly. The absorber must be designed to accommodate
any anticipated load change. The column size and the liquid, oxidant, and
catalyst flows must be designed for each application after examining the
boiler operating history and establishing ranges of variation.
Typical ranges for several operating parameters for this type of
process are shown in Table 2.2.9-2. Reagent concentrations were not avail-
able. Economics for the various processes cover a wide range presumably
due to different techniques for oxidant generation and treatment of the
scrubbing solution. Costs are reported to range from $84-134/kW for utility
applications' capital expense and 6.7-9 mills/kWh for operating expense.
Presently, some of the oxidation-absorption-reduction processes have
reached the prototype stage of development. Table 2.2.9-3 presents a list
of oxidation-absorption-reduction process vendors and the status of develop-
ment of their projects. One can see from the table that no coal-fired flue
gas tests have been made as of yet.
2-61
-------
TABLE 2.2.9-1. SYSTEM DESIGN CONSIDERATIONS
Variable
Design Effect
Presence of participates
Presence of
Increased gas flow
Increased NO concentration
Requires prescrubber
Depends on individual process: if NOz is com-
pletely reduced to Na or NHa by SOJ (as does
MHI), then at least the stoichiometric S02:NOX
mole ratio of 3:1 is required [see equation
(9-6)]; if N02 is not reduced completely, then
a different ratio will be necessary
Requires larger column diameter; increased
liquid flow rate
Requires larger column height; increased gas-
phase oxidant flow rate; increased liquid-
phase catalyst concentration
TABLE 2.2.9-2.
TYPICAL RANGES OF OPERATING VARIABLES FOR
OXIDATION-ABSORPTION-REDUCTION PROCESSES8 9'9 °
Variable
Liquid/Gas Ratio, 1/Nm3
Oxidant/NO Mole Ratio 03 systems
ClOa systems
S02/NOX Mole Ratio
Superficial Gas Velocity, m/s
Pressure Drop, mmH20
Range
2-12
0.6-1.0
0.55
2.5-5.0
3-5
200-500
2-62
-------
TABLE 2.2.9-3. PROCESS VENDORS OF OXIDATION-ABSORPTION-
REDUCTION PROCESSES
92,93
Vendor
Status of Development
Chiyoda
Ishikawaj ima-Harima Heavy
Industries
Mitsubishi Heavy Industries
Osaka Soda
Shirogane
Sumitomo Metal-Fujikasui:
Calcium Process
Sumitomo Metal-Fuj ikasui:
Sodium Process
1975: 1000 Nm3/hr flue gas from heavy oil-
fired boiler
1975: 5000 Nm3/hr flue gas from oil-fired
boiler (3000 hours continuous)
1975: 2000 Nm3/hr flue gas from heavy oil-
fired boiler (700 hours continuous)
1976: 60,000 Nm3/hr flue gas from oil-fired
boiler
1974: 48,000 Nm3/hr flue gas from oil-fired
boiler
1976: 25,000 Nm3/hr flue gas from sintering
furnace
1973: 62,000 Nm3/hr flue gas from heavy oil-
fired boiler (5 others)
2.2.9.2 System Performance —
No coal-fired testing has been performed. Results of oil-fired tests
show up to 90 percent NO reduction and >95 percent
reduction.
The primary disadvantage of these processes is the utilization of
costly gas-phase oxidants which create secondary wastewater pollution prob-
lems. Both ozone and chlorine dioxide are highly unstable so they cannot be
stored and must be generated onsite. 03, the more expensive of the two, is
generated by a high energy corona discharge in air. This instantaneous pro-
cess requires significantly large amounts of electricity. CIOz is generated
by a slower chemical reaction (requires about 20 minutes to respond to a
change in demand) which could make it less responsive to boiler load changes.
The use of CIOz introduces an additional secondary pollutant, chlorides,
besides the nitrite salt problem. Significant amounts of corrosion-resistant
2-63
-------
material are required for oxidation-absorption-reduction processes,,
regardless of which oxidant is utilized. Some of the processes would not
be applicable to low sulfur coals as they require large amounts of S02 to
obtain N02(aq) or N02 reduction.
2.2.10 Oxidation-Absorption
2.2.10.1 System Description—
As a group, oxidation-absorption processes include those oxidation
processes which do not qualify for the oxidation-absorption-reduction cate-
gory. Basically, there are two types of oxidation-absorption processes.
One is a simplified version of the oxidation-absorption-reduction process
and uses an excess of ozone to selectively oxidize NO,, to N20s which is
X
absorbed into aqueous solution and concentrated to form a 60 percent nitric
acid (HNOs) by-product. There is no reduction of NO (N0a~) by the absorption
X
of SOz(as SOs) and no wastewater treatment facility. The other type of
oxidation-absorption process is based on equimolar NO-NOa absorption:
absorbing NaOs which is formed by the gas-phase reaction of NO and N02.
A flow diagram of the Kawasaki Heavy Industries oxidation-absorption
process is shown in Figure 2.2.10-1. Flue gas is taken from the boiler
after the air preheater. It passes countercurrent to a magnesium hydroxide
[Mg(OH)2] slurry in the first section of the absorber. There, SOa is absorbed
and undergoes the following reactions.95
S02(g) ->• SOa(aq) (2-23)
Mg(OH)2(s) + S02(aq) + 5H20 -+ MgS03 •6H20(s)4- (2-44)
The S02-free flue gas passes to the first denitrification section of the
absorber while the liquid effluent drops to a holding tank. A recycle N02
stream is added to the flue gas to bring the NO:N02 mole ratio to 1:1. The
2-64
-------
ro
Ln
^
AIR
HEATER
f
AIR
>
!!!• 1
02
L
Mtf
•.-"i—
S02
ABSORBER
SECTION
I\!0+N02
ABSORBER
SECTION
N02
ABSOR8EF
SECTION |
Mg[NC^2 tj MgSOa.MgSO,,
CLEAN
FLUE GAS
CRYSTALUZER 6a(OH)2
Figure 2.2.10-1. Flow diagram of Kawasaki Heavy Industries process.;
-------
resulting mixture then passes countercurrent to a Mg(OH)2 slurry. Equimolar
amounts of NO and N02 react and are absorbed in the following manner.96
N0(g) + N02(g) -»• N203(g) (2-45)
N203(g) -* N203(aq) (2-46)
Mg(OH)2(aq) + N203(aq) -> Mg(N02)2(aq) + H20 (2-47)
The flue gas passes out of the top of this absorption section while the
liquid effluent drops to the holding tank. Because the rate of reaction
2-45 decreases with NOX concentration (below 200 ppm it becomes negligible),
it is necessary to further reduce NO by injecting ozone to oxidize the
remaining NO to N02. The mixture then passes to the final denitrification
section of the absorber and is passed countercurrent to a Mg(OH)2 slurry.
q T
This section of the absorber is described by the following reactions.
2N02(g) + N2(Mg) (2-8)
NaCMg) + N204(aq) (2-9)
2N20.t(aq) + 2Mg(OH)2(s) -»• Mg(N03)2(aq) + Mg(N02)2(aq) + 2H20 (2-48)
The clean flue gas leaves the top of this absorber section, is passed to a
reheater for plume buoyancy and sent to the stack. Part of the liquid efflu-
ent from this section is recycled to the tops of the absorber sections while
the rest drops to the holding tank. The slurry solution is pumped to a
thickener which separates the soluble nitrite (N02) and nitrate (NOl) salts
from the solid magnesium sulfite. The overflow from the thickener passes to
a N02 decomposition reactor where sulfuric acid is added.98
3Mg(N02)2(aq) + 2H2SO.,(aq) + 2MgSOit (aq) + Mg(N03)2(aq) + 4NO(g) + + 2H20
(2-49)
2-66
-------
The NO off-gas passes through an oxidizer where it is oxidized by air to N02
and sent to the first denitrif ication section of the absorber. The effluent
from the decomposition reactor is mixed with the thickener bottoms and pumped
q n
to a second oxidizer.
MgS03'6H20(s) + ^02(g) -> MgSCH(aq) + 6H20 (2-50)
The magnesium sulfate formed in the oxidizer is treated with calcium nitrate
p c
in a gypsum production reactor.
Ca(N03)2(aq) + MgSCK (aq) + 2H20 -> CaSO.* -2H20(s) 4- + Mg(N03)2(aq)
(2-51)
The products of this reaction are sent to a centrifuge to remove the solid
gypsum by-product. The liquid from the centrifuge goes to another decomposi-
tion reactor where makeup lime slurry is added. lcc
Mg(N03)2(aq) + Ca(OH)2(s) -> Ca(N03)2(aq) + Mg(OH)2 (s) (2-52)
The magnesium hydroxide product is separated in a thickener and recycled to
the absorbers. The thickener overflow stream is split and part is recycled
to the gypsum production reactor and the rest is concentrated to form low-
grade liquid fertilizer by-product, Ca(N03)2.
Since the processes in this category are all very different, especially
with respect to chemistry, generalization of typical ranges of operating
variables is not meaningful and, therefore, not presented. No published
economics for these processes were found.
Presently, the equimolar absorption-type oxidation-absorption processes
are still in the pilot-unit stage of development. Table 2.2.10-1 presents
a list of all oxidation-absorption process vendors and their project's status
of development .
2-67
-------
TABLE 2.2.10-1. PROCESS VENDORS OF OXIDATION-ABSORPTION PROCESSES100
Vendor Status of Development
Kawasaki Heavy Industries 1975: 5000 Nm3/hr flue gas from coal-
fired boiler
Tokyo Electric-MHI (N0y only) 1974: 100,000 Nm3/hr flue gas from natural
gas-fired boiler
Ube Industries No information available
2.2.10.2 System Performance—
Only one coal-fired test has been performed. No information has been
published on any of the tests conducted.
The production of nitrate salts poses a potential secondary pollution
problem. The plan for reclaiming and concentrating the nitrates as
Ca(N03)2(aq) for liquid fertilizer is questionable as the by-product is of
low quality and may not be easily marketable in the U.S. Also, the gypsum
by-product would be contaminated with various nitrate and sulfite salts, and
therefore, would probably be useful only as landfill material. Much corro-
sion-resistant material is necessary due to the utilization of ozone and
circulating magnesium slurries. The three absorber sections, with their
respective operating conditions, and ozone generation present complex pro-
cess control problems. The process steps of several absorber sections in
series (large fan requirements), ozone generation (corona discharge), flue
gas reheat (inline heater), and by-product and wastewater treatment are all
energy intensive and present technical and economic disadvantages when com-
pared to other simpler FGT processes.
2-68
-------
2.3 CONTROLS FOR OIL-FIRED BOILERS
2.3.1 Selective Catalytic Reduction-Fixed Packed Bed Reactors
2.3.1.1 System Description—
Fixed packed bed systems are applicable only to flue gas streams
containing less than 20 mg/Nm of particulates. As such, they are applicable
to distillate oil-fired boilers (19 mg/Nm3) but not to residual oil-fired
boilers (330 mg/Nm3).
The primary feature of these systems is the reactor which contains the
catalyst. As the name implies, the granular catalyst is randomly packed in
a stationary bed. An example of a typical fixed bed reactor is shown in
Figure 2.3.1-1. The important features of the reactor are:
the catalyst
• the catalyst support
• the gas distributor
The catalyst can be either spherical or cylindrical in shape. Spherical
granules typically range in size from 4-10 mm in diameter. The composi-
tion varies from process to process and most formulations are proprietary.
All of the catalysts considered here for use in treating flue gas containing
S02 and SOa are resistant to poisoning by these compounds. Long term tests
of these catalysts in the presence of SO have shown very little or no decrease
X
in activity or selectivity. The catalyst is supported either by inert packing
(as shown in Figure 2.3.1-1) or by a perforated support plate (Figure 2.3.1-2).
The catalyst supports hold the catalyst fixed in place in order to pre-
vent both mobilization of the particles by the gas stream and catalyst rear-
rangement which would allow channelling of the flue gas. The gas distributor
can be a perforated plate or similar device which spreads the gas flow across
the entire cross-section of the catalyst bed.
2-69
-------
6* loyer I" bolls-
6'optionol odditionol layers-
of progressively smaller bolls
for improved distribution and
scale removal
Catalyst Bed
(1/8" i 1/8" pellets)
3* loyer 1/4" bolls
4" loyer Ml' bolls
5" loyer 3/4" bolls
3/4" balls
Reactor Outlet Screen
with Continuous Slotted
Openings
Catalyst Bed
/l/4"x 1/4" \
V pellets /
~3 layer 3/8 bolls
4" loyer 1/2 bolls
5" loyer 3/4" balls
3/4" balls
Catalyst Dump Flange
Figure 2.3.1-1. Example of typical fixed packed bed reactor.
i o i
SP
f""" ;!j:|:xx|:;:|:
VfVVti t Vt
f :-:-:•:-:.:;
n
•»v
fV
•/•«*.•.*.•
Vr'frr
*r**fe
"I"' SUPPORT BEAMS REQUIRED ONLY FOR ^i^T
I '.[ ! LARGER VESSELS OR HIGH LOADINGS ; ^|
11 ll__- _-.____ _'_"-!.
Figure 2.3.1-2. Example of catalyst support plate.
1 02
2-70
-------
A typical fixed bed SCR process layout is presented in Figure 2.3.1-3.
Several arrangements are possible, however, for application to new boilers
this arrangement is the most desirable.8
Flue Gas
Reactor
Air
Heater
Stack
NH3 Air
Figure 2.3.1-3. Process layout for fixed bed SCR process.
The principle of operation of these systems involves a gas phase
reaction between ammonia (NHs) and NO (NO and NOa). These reactions are
presented most accurately by
1 2
4NO + 4NH3 + 02 ^ 4K2 + 6H20
(2-1)
2N02 + 4NH3 + 02 £ 3N2 + 6H20
(2-2)
The first reaction predominates since flue gas NO is typically 90-95 percent
X
NO. As shown, the NO is reduced to molecular nitrogen (N2) which exits with
the flue gas stream.
The primary design equation used with these processes is the standard
equation for reactor design,
1 3
V
F
.x
dx
r
(2-3)
2-71
-------
where V is the catalyst volume
F is the mass (or molar) flow rate
x is the conversion of NO., to Na
A.
r is the reaction rate mass (or moles)
volume of catalyst x time
The reaction rate, r, for each NO reduction reaction can be represented by
r = k[NH3]a[NO]b[02]C (2-4)
where k is the reaction rate constant
[NHa], [NO], [Oz] are the reactant concentrations, and
a, b, c are empirically determined exponents
The catalyst volume can also be determined if the space velocity is known
for the catalyst and removal level of interest. The space velocity is
defined as the gas flow rate divided by the catalyst volume. The reaction
rate is different for each catalyst formulation and therefore, values for
k, a, b, and c must be determined for the particular catalyst to be used
before any design can be performed. The reaction rate constant is usually
described by the Arrhenius equation.
_ E_
1 A RT
k = Ae (2-5)
where A is the frequency factor
E is the activation energy
R is the universal gas constant, and
T is the temperature
2-72
-------
Values for k, a, b and c for two catalyst formulations are shown in Table
2.3.1-1. Values for other catalyst formulations will be different. The
most important design and operating variables are similar to those for
moving bed systems using granular catalysts. These are listed, along with
typical ranges, in Table 2.3.1-2.
Other variables that affect the process are
• flue gas flow rate
NO control level
X
• NO concentration
boiler load variation
The flue gas flow rate and control level determine the catalyst volume
(hence reactor size) necessary. Increases in either also increases the
reactor size. The NOX concentration is a function of fuel type used in
the standard boilers. Higher concentrations require larger NHs storage
and vaporization equipment; reactor size is not affected. Boiler load can
affect several things including flue gas temperature, flow rate and NO con-
X
centration. It is necessary to maintain reactions temperatures of 350 to
400°C and temperature control equipment may be necessary if the boiler
experiences large load variations. Where these variations are present,
some equipment overdesign may be warranted to insure a constant control
level. These variables are discussed in more detail in the section on moving
bed SCR systems for coal-fired boilers, Section 2.3.2. Costs of fixed packed
bed systems range from $16-49/kW (capital) and 1.2-1.8 mills/kWh (operating).
These costs are based on utility applications as well as a variety of pro-
cesses and operating conditions.
There are vendors of fixed packed bed SCR systems and all are Japanese.
Vendors are listed in Table 2.3.1-3 and the scale of development is also
noted. Fixed packed systems have been applied to industrial but not utility
boilers in Japan. Existing and planned installations are shown in Table
2.3.1-4. Currently, there are no installations in the U.S.
2-73
-------
TABLE 2.3.1-1. REACTION RATE DATA FOR TWO
CATALYST FORMULATIONS11
Catalyst: V205 on A1203
_ 9650
1?T
k = 2.05 x 103e
a = 0.30
b = 0.22
c = 0.05
Catalyst: Fe-Cr on
10,860
"RT
k = 3.25 x 103e
a = 0.45
b'= 0.10
c = 0.15
TABLE 2.3.1-2. DESIGN AND OPERATING VARIABLES FOR
FIXED PACKED BED SYSTEMSllf
Variable Typical Range
Gas Velocity, m/s 1-1.5
Bed Depth, m 0.2-0.6
Space Velocity, hiT1 6,000 - 10,000
Pressure Drop, mmH20 40 - 80
Temperature, °C 350 - 400
2-74
-------
TABLE 2.3.1-3. VENDORS OF SCR FIXED BED SYSTEMS FOR
OIL-FIRED APPLICATIONS21
Vendor Notes
Sumitomo Chemical Tested on commercial scale equipment
Hitachi Zosen Tested on commercial scale equipment
Hitachi, Ltd. Tested on commercial scale equipment
Mitsubishi Heavy Industries Tested on commercial scale equipment
Ishikawjima-Harima Heavy Industries Tested on commercial scale equipment
Mitsui Toatsu Chemical Has not been applied to boilers
Kawasaki Heavy Industries Tested on pilot scale equipment
Mitsubishi Kakoki Kaisha Tested on commercial scale equipment
2.3.1.2 System Performance—
Typical performance data for packed fixed bed SCR systems are shown in
Figures 2.3.1-4 and 2.3.1-5 and Tables 2.3.1-5 through 2.3.1-7. These data
indicate that NO removals up to 90 percent are achievable with these sys-
tems. This allows them to be considered for all control levels of interest
in this study.
There are some potential problems downstream of the SCR systems (fixed
packed bed, moving, and parallel flow) due to the presence of the unreacted
ammonia in the flue gas.. Two things can happen: 1) the NH3 can react with
S02 or S03 to form ammonium bisulfate or ammonium sulfate or 2) the NH3 can
enter the downstream equipment unreacted. The bisulfate has been shown to
cause air preheater pluggage and this is the subject of ongoing research both
at the EPA and the Electric Power Research Institute (EPRI). Both the bi-
sulfate and sulfate exist as a particulate, but may be difficult to collect
if the particles are submicron in size. Unreacted NH3 is not likely to pre-
sent any operational problems. A recent study has shown that if an ESP
exists downstream, then most of the NH3 will exit with the ash. NH3 can ac-
tually improve the performance of an FGD system.129
2-75
-------
TABLE 2.3.1-4. EXISTING FGT INSTALLATIONS OF SCR FIXED BED SYSTEMS OIL-FIRED INDUSTRIAL BOILERS
21
Location
(Japan)
User
Process Developer
Fuel
Capacity
(Nm3/hr)
Completion
Date
to
—i
Amagasaki
Amagasaki
Amagasaki
Sakai
Hokkaichi
Sodegaura
Sodegaura
Sorami
Sorami
Sorami
Sorami
Kawasaki
Kawasaki
Chita
Kansai Paint
Nisshin Steel
Nisshin Steel
Nisshin Steel
Shindaikyowa P.C.
Sumitomo Chemical
Sumitomo Chemical
Toho Gas
Toho Gas
Toho Gas
Toho Gas
Nippon Yakin
Toho Gas
Toho Gas
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi, Ltd.
Hitachi Zosen
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Sumitomo Chemical Eng.
Mitsubishi Kakoki
Mitsubishi Kakoki
Mitsubishi Kakoki
Distillate
Resid
Resid
Distillate
Resid
Resid
Resid
Distillate
Distillate
Distillate
Distillate
Resid
Distillate
Distillate
16,000
20,000
19,000
30,000
440,000
30,000
240,000
62,000
23,000
23,000
19,000
14,000
30,000
30,000
October 1977
August 1977
July 1977
December 1978
November 1975
July 1973
March 1976
October 1977
December 1977
June 1978
July 1978
July 1978
October 1977
October 1977
-------
100
90
SU 80
70
60
50
CATALYST - 8
CATALYST - C
SOj 600-800 ppa
TEMP. 350 C
NH3 / NOX 0.9
SV CAT. -A 15,000 HR'l
CAT. -C 5.000 HR'1
0.5 1.0 1.5 2.0
LAPSE OF TIME, YEARS
CATALYST - 8 : Based on T102
CATALYST - C : Based on d-A1203
Figure 2.3.1-4. Performance of experimental catalyst of
Sumitomo Chemical.l ° "*
100 r
20
10
1.4
Figure 2.3.1-5.
NHs/NOx Mole Ratio
Typical example of operation data (oil-fired boiler,
350-400°C, granular or honeycomb catalyst).105
2-77
-------
TABLE 2.3.1-5.
OPERATION PARAMETERS OF MAJOR PLANTS CONSTRUCTED
BY HITACHI ZOSEN106
Completed
Plant site
Gas source
Capacity
(Nm3/hr)
Load factor (%)
Pretreatment of gas
Reactor inlet
N0x (ppm)
S0x (ppm)
Dust (mg/Nm3)
02 (%)
Reactor type
Reaction temp.
NO/NH3 ratio
A
Catalyst No.
SV (hr'1)
NOV removal (%)
Pv.
Pressure drop by
SCR reactor (mmH20)
Catalyst life
Idemitsu
Kosan
Oct. 1975
Chiba
FCC-CO
Boiler and
furnace
350,000
50-100
Heating
230
50-80
20-50
2.3
Fixed bed
400
1.0
204
5,000
93
170
1 year
Shindaikyowa
Petrochemical
Nov. 1975
Yokkaichi
Oil-fired
Boiler
440,000
50-100
EP*, FGD,
Heating
150
80-130
30-100
3.2
Fixed bed
420
1.0
304
10,000
80^
160
1 year
Kawasaki
Steel
Nov. 1976
Chiba
Iron-ore
Sintering
machine
762,000
70-100
EP, FGD
WEpt,
Heating
200-300
5-20
3-10
11.2
Fixed bed
1.0
304
4,000
95
50
1 year
*Electrostatic precipitator
"""Wet electrostatic precipitator
Tlncluding leakage in heat exchanger
2-78
-------
TABLE 2.3.1-6. SCR PLANT BY MITSUI ENGINEERING &
SHIPBUILDING CO.
1 0 7
Mitsui Petro-
Chendcal Co.
Capacity (Nm3/hr)
Gas composition
NOX (ppm)
SOX (ppm)
Dust (mg/Nm3)
Catalyst and reactor
Catalyst carrier
Catalyst shape
SV (hr'1)
Temperature (°C)
NH3/NOV mole ratio
X
NC- removal (%)
X
Total pressure drop
Leak NH3 (ppm)
Operation start
Plant cost (106 yen)
Denitrification cost
(yen/kWhr)*
200,000
190
None
20-50
A1203
Granule
2,600
350
1.0
Above 90
Oct. 1975
*Including 7 years depreciation.
2-79
-------
TABLE 2.3.1-7. OPERATION DATA OF SCR PLANTS FOR DIRTY GAS
108
Gas for SCR (Nm3/hr)
Fuel
Load fluctuation
Stack height (m)
Inlet gas composition
02 (%)
SOX (ppm)
NOX (ppm)
Particulates after EP (mg/Nm )
FGD unit
SV (hr"1)
Temperature (°C)
NOV removal (%)
X
NHa/NO mole ratio
Leak ammonia (ppm)
Type of reactor
Pressure drop (mmHzO)
Reactor
Total system
Plant completed
Pilot
30,000
011(8=0.7%)
60-100%
70
6
400
200
5-20
None
5,000
320
Over 90
1.0
10-20
Fixed bed
July 1973
Commercial
240,000
011(8=0.7%)
60-100%
140
6
400
200
5-10
None
5,000
320
Over 90
1.0
10-20
Fixed bed
200
500
March 1976
2-80
-------
2.3.2 Selective Catalytic Reduction-Moving Bed Reactor
2.3.2.1 System Description—
The primary feature of this and other selective catalytic reduction
(SCR) processes is the reactor. The reactor contains the catalyst which
allows the reduction reaction to proceed at 350-400°C. In this case the
reactor is a moving-bed type in which a portion of the catalyst is either
continuously or intermittently removed from service in order to remove the
accumulated particulates. Some moving bed reactors are shown in Figure
2.3.2-1. The actual reactor arrangement is highly process specific, however,
the principle of operation is the same for all of the processes.
For moderate particulate loadings the bed is moved intermittently and
operated as a fixed on bed system most of the time. High particulate load-
ings require that the bed be moved continuously. Moving bed reactors are
reportedly capable of handling up to 2 g/Nm3 of particulates. However, this
is more a theoretical than a practical particulate load limit.9 If possible,
this would be sufficient to handle the residual oil-fired boilers (0.33
g/Nm3) .
Entrained particulates are generally removed from the catalyst bed by
vibrating the bed and screening the catalyst or some other mechanical means.
Particulate removal by the bed can be as high as 70 percent.
An example flow diagram for a moving bed SCR process is shown in Figure
2.3.2-2. Flue gas is taken from the boiler between the boiler and the air
preheater. An economizer bypass is utilized for temperature control. Ammo-
nia is injected and mixed with the flue gas stream just prior to the reactor.
The flue gas passes through the catalyst bed where N0x is reduced to N2 and
is then sent to the stack or further treatment facilities. The catalyst cir-
culates through the reactor and is screened to remove particulates. The par-
ticulates are blown to a small baghouse where they are collected.
2-81
-------
ELEMENT
CASING
Kurabo
Mitsubishi Heavy
Industries7
GAS
CATALYST
CATALYST
ELEVATOR
DUST
GAS
Sumitomo Heavy Industries
Figure 2.3.2-1. Moving bed reactors of three process vendors.
2-82
-------
AIR
PREHEATER
REACTOR^""
w
AIR
AMMONIA
GENERATOR
Figure 2.3.2-2. Process flow diagram for moving bed SCR process.
11
2-83
-------
The NO reduction reactions are represented most accurately by
x
12
4NO + 4NH3 + 02 J 4N2 + 6H20 (2-1)
2N02 + 4NH3 + 02 J 3N2 + 6H20 (2-2)
The first reaction predominates since flue gas N0x typically consists of
90-95 percent NO. As shown, the N0x is reduced to molecular nitrogen, N2,
which exits with the flue gas. 02 is also a reactant, but is in large
excess (>3 percent) in the flue gas.
The fundamental design equation used for catalytic reactor design is
13
}-
where V = catalyst volume
F = mass flow rate (or molar flow rate)
x = conversion of NO to N2
r = reaction rate, mass (or moles)
volume of catalyst x time
The reaction rate, r, for each of the NOX reduction reactions can be repre-
sented by
r = k [NH3]a [N0]b [02]C (2-4)
where k = reaction rate constant
[NHa], [NO], [02] = reactant concentrations, and
a, b, c = empirically determined exponents
The reaction rate is different for each catalyst formulation and, therefore,
values for k, a, b, and c must be determined for each particular catalyst.
The catalyst volume can also be determined if the space velocity is known
for the catalyst and removal level of interest. The space velocity is
defined as the gas flow rate divided by the catalyst volume.
2-84
-------
The most important design and operating variables are presented in
Table 2.3.2-1 and typical values for these are also shown. Although some
of the data used in developing this table are for utility applications, the
values should not be different for industrial applications. There are other
variables that are important, but must be determined for each individual
case. These are:
flue gas flow rate
NO control level
X
NO concentration
X
TABLE 2.3.2-1. DESIGN AND OPERATING VARIABLES FOR
MOVING BED SCR SYSTEMS1
,!<» ,1 5
Variable Typical Range
Space Velocity 6000 - 10,000 hr"1*
NH3:NO Mole Ratio 0.7 - 1.0*
X
Flue Gas Temperature 350 - 400°C*
Pressure Drop 40 - 80 mm Hg
Catalyst Diameter (ring) 4 - 8 mm
*Actual value will depend on required removal level.
The first two of these variables are the most important since they
determine the size of the reactor. Higher flow rates or removal levels
require larger reactors. Pressure drop for FGT systems does not usually
change for a particular process. To accommodate the higher flow rates,
the reactor cross-sectional area is increased to provide a constant linear
velocity. The NO concentration will affect the NHa and dilution steam
requirements, but will not affect reactor size. Both flow rate and NOX con-
centration can be affected by boiler operating conditions. Since industrial
boilers have fewer burners than utility boilers,2 one burner represents a
more significant fraction of the total boiler capacity. Therefore, a change
in load on an industrial boiler may change these variables substantially if
a burner is taken out of service
2-85
-------
The FGT system will have to be designed to accommodate load changes.
The necessary design accomodations will have to be determined for each
application after examining operating data and establishing ranges of varia-
tion. Most likely this will involve a slight overdesign of the reactor and
other equipment. The process control loops used with utility boiler applica-
tions should be capable of following load changes in industrial boilers.
Space velocity is usually defined as the volume of catalyst or reactor
required to treat a given flow rate of flue gas.16 The magnitude of the
space velocity is dependent entirely on the particular catalyst being con-
sidered. As can be seen in Table 2.3.2-1, the range for moving bed processes
is 6000-10,000 hr"1. These values are typically reported for 90 percent
removal. For lower control levels, the value will be proportionally greater.
Almost all SCR processes require temperatures in the 350-400°C range
in order to achieve good reaction rates. The temperature can vary with such
things as boiler load, excess air, and ambient air temperatures. To control
temperature two techniques are possible. The first involves bypassing a por-
tion of the hot flue gas around the economizer and mixing with the economizer
outlet gas so that the desired temperature is maintained.11 The other tech-
nique uses inline heaters to obtain the desired temperature.
1 7
The NH3:NOX mole ratio is also a function of the necessary removal
level and, to a lesser extent, space velocity.18 For the three levels of
removal considered in this study, 70, 80 and 90 percent, NH3:NO mole ratios
X
of 0.7, 0.8, and 0.9 are required, respectively. These data are for oil-
fired boilers.
The catalyst shape and size is determined by the process vendor and is
simply a design decision. Ring shapes (shaped like Raschig rings) are the
most resistant to particulate plugging and, for this reason, were selected
for this study.
2-86
-------
The most recently published cost estimates for SCR systems are those
of the Japanese Environment Agency which were published in Dr. Jumpei Ando's
most recent report on Japanese NO control technology. Values taken from
A
this study for two gas flow rates are shown below. The smaller gas flow
rate is typical of industrial sized units while the larger flow rate is
typical of utility installations.19
Gas Flow Rate Capital Cost Operating Cost
50,000 Nm3/hr $0.5 x 105 $0.2 x 105
1,200,000 NrnVhr $5.0 x 106 $3.5 x 106
The cost for the large unit was included for comparison with other costs
for large units that were reported for other process types in Section II.
The reactor and catalyst type (fixed packed bed, moving bed or parallel
flow) were not disclosed and, as a result, those costs are assumed to apply
to all N0y-only SCR systems.
Vendors of moving bed SCR systems are listed in Table 2.3.2-2 and the
relative levels of application are noted. Although there are seven vendors,
only four have applied their process to boilers. Of these, three have been
applied to commercial scale equipment. Table 2.3.2-3 lists the moving bed
systems that have been applied to oil-fired industrial boilers in Japan.
Moving bed systems have not been applied to utility boilers in that country.
Presently, there are no moving bed systems operating in the U.S. The Japan-
ese installations all treat gas from residual oil-fired boilers, implying
that the technology is not necessary for distillate oil-fired applications
which can use fixed packed beds.
2.3.2.2 System Performance—
The performance of several moving bed catalysts and plants is illustrated
in Figures 2.2.3-2 through 2.3.2-8. The data presented indicate that NO con-
X
trol greater than 90 percent is possible through the correct selection of
process design variables. Outlet NHa concentrations are also shown. These
are discussed in detail in Section VI. Table 2.3.2-4 shows several operating
values from a commercial installation.
2-87
-------
TABLE 2.3.2-2. VENDORS OF SCR MOVING BED SYSTEMS
o l
FOR OIL-FIRED APPLICATIONS
21
Vendor
Notes
Sumitomo Chemical & Mitsubishi Heavy Tested on commercial scale equipment
Industries
Hitachi, Ltd. Tested on commercial scale equipment
Ishikawajima-Harima Heavy Industries Tested on pilot scale equipment
Kurabo
Kobe Steel
Sumitomo Heavy Industries
Asahi Glass Company
Tested on commercial scale equipment
Has not been applied to boilers
Has not been applied to boilers
Has not been applied to boilers
TABLE 2.3.2-3. EXISTING FGT INSTALLATIONS OF SCR MOVING BED SYSTEMS
OIL-FIRED INDUSTRIAL BOILERS21
Location
(Japan)
User
Process
Developer
Capacity
Fuel (Nm3/hr)
Completion
Date
Kaizuka Chiyoda Kenzai Hitachi, Ltd.
Amagasaki Nippon Oils & Hitachi, Ltd.
Fats
Sodegaura Sumitomo
Chemical
Sodegaura Sumitomo
Chemical
Hirakatu I'urabo
Mitsubishi H.I.
Sumitomo Chemical/
Mitsubishi H.I.
Kurabo
Resid 15,000
Resid 20,000
Oct 1977
Apr 1978
Resid 300,000 Sept 1976
Resid 300,000 Oct 1976
Resid 30,000 Aug 1975
2-88
-------
RING TYPE
CATALYST
TOO
80
~ 60
40
20
60
40
20
0246
SV (x 1000, hr'1)
10
Figure 2.3.2-3.
SV Vs. NOX removal and NHs leak (ring type catalyst, 15 mm
diameter, 350°C NH3/NO 1.0, inlet NOX 250 ppm).109
2-89
-------
100
90
s 80
Test Plant at Shinnagoya Power Station
(In..case of intermittent moving bed)
LS crude oil S0x=50^70 ppm
N0x=50^130 ppm
Dust=20^60mg/N
NHi/NOx molt ratio=1.0
o
70
is
o
£ 60
JJ
•H
* 50
o
-U4J
Q) (Q
r-t 14
4J 4J
9 C
O 0)
u
w c
o o
jj o-~ in
o e JU
IB •> a
30
20
!Denitri fication
ef ficienc;'
Gas temperature
: j
j Reactor outlet NH3 jconcentrat
8000
7000
6000
5000
4000
3000
2000
1000
0
400
350
300
250
200
i
§
M
ion
2-
4J
19
•w
fl
u
Boiler load
Figure 2.3.2-4.
Relation between boiler load and denitrification
efficiency (one example).110
2-90
-------
Intermittent Moving Bed Reactor Test Data
1UU
c -~
0 *"
fO
*W C on
•p< 4) bU
*> U
IS
aw
GO
wu
o-^"'
G-^"
_c
_ Q_ ^
(•)• " 1 1
o-. o
SV= 8,000 hf1
Gas temperature 350 °C
Inlet NOx con-
centration 130 ppm
^*
^^^.G
^Q •~^^
j^
1 1
e
a
£•
o
^j
•^
4J
<0
J-l
4J
C
Q>
O
O
o
(0
•H
c
(0
4J
0)
9n aJ
t,y n^
3
0
M
10 S
1 w 4-*
O
nj
O
«
n
0.8 1.0
NHs/NOx mol ratio ( mol/mole)
1.2
Figure 2.3.2-5. NHs/NOx mole ratio vs. denitrification efficiency
and reactor outlet ammonia concentration.
2-91
-------
1UU
0^*
•H— on
10
0 >,
•H O
>u c
•H 4)
IJ-H
4J O
•H-H
2w60
• -
-------
TABLE 2.3.2-4. OPERATION DATA OF A COMMERCIAL SCR
PLANT FOR DIRTY GAS108
Gas for SCR (Nm3/hr) 300,000
Fuel Oil (S=0.7%)
Load fluctuation 60-100%
Stack height (m) 140
Inlet gas composition
02 (%) 6
SOX (ppm) 400
NOV (ppm) 200
P^
Particulates after EP (mg/Nm3) 10-20
FGD unit Scheduled
SV (hr"1) 5,000
Temperature (°C) 320
NOX removal (%) Over 90
NHs/NO mole ratio 1.0
Leak ammonia (ppm) 10-20
Type of reactor Moving bed
Plant completed Oct. 1976
2.3.3 Selective Catalytic Reduction-Parallel Flow Reactor
2.3.3.1 System Description—
The distinguishing aspect of this process is the catalyst shape which
is produced in a variety of shapes. The catalysts are produced in either a
honeycomb, pipe, or plate shape. Both metal and ceramic supports are employed
Several shapes are illustrated in Figure 2.3.3-1. The catalyst shapes allow
particulate laden flue gas to pass through the reactor with no inertial impac-
tion of the particles while the NO is transported to the catalyst surfaces
by basic diffusion. The catalysts can handle all of the particulate levels
emitted by the standard boilers.
2-93
-------
Honeycomb
(Ceramic)
(Grid Type)
oooc
Honeycomb
(Ceramic)
(Hexagonal)
Honeycomb
(Metal).
(Wave Type)
Plate (Ceramic)
Plate (Metal)
Tube (Ceramic)
Figure 2.3.3-1.
Parallel Passage
Shapes of parallel flow catalysts.22
2-94
-------
The reactors used are similar to standard fixed bed units and an
example is shown in Figure 2.3.3-2. The catalyst is usually prepared
in small modules and manually stacked within the reactor. The specific
arrangement will depend on the particular process under consideration.
CATALYST LAYER
Figure 2.3.3-2. Typical reactor used with parallel flow SCR process.
23
A typical flow diagram for a parallel flow SCR system is shown in
Figure 2.3.3-3. The arrangement is similar to the other SCR processes in
that hot flue gas leaving the boiler economizer is injected with NHs and
passed through a catalyst bed. Temperature control is important and can
be accomplished with either a fired heater or an economizer bypass. NHa
can be controlled using boiler operating condition inputs to conventional
control components.
2-95
-------
PARTICULATE REMOVAL,
TO FGD
AND/OR STACK
AIR
Figure 2.3.3-3. Flow diagram for parallel flow SCR process.'
Within the reactor, NO^ reacts with NHa to form N2 and H20 according
to the following reactions.
12
4NO + 4NH3 + Oa 2 4N2 + 6H20
(2-1)
2N02 + 4NH3 + 02 £ 3N2 + 6H20
(2-2)
Reaction (2-1) is the primary reaction since flue gas NO is typically 90-
95 percent NO. 02 is necessary for both reactions and is present in suffi-
cient quantities (>3 percent) in all of the flue gases from the standard
boilers.
The catalyst volume for a desired NO removal can be determined by the
fundamental design equation for a plug flow reactor.
1 3
V
F
f*
I dx
Jo r
(2-3)
The reaction rate, r, can be expressed as
r = k[NH3]a [N0]b [02]C
(2-4)
2-96
-------
The variables presented here have the same definitions as those presented in
equations 2-3 and 2-4 of Section 2.2.2. The catalyst volume can also be
determined if the space velocity is known for the catalyst and removal level
of interest. The space velocity is defined as the gas flow rate divided by
the catalyst volume.
The reaction rate is different for each catalyst formulation since
different catalysts will lower the activation energy by different amounts.
The activation energy affects the reaction rate constant, k, according to
the Arrhenius equation.
_ E_
k = Ae RT (2-5)
An important design variable with catalytic systems is the space
velocity which expresses the volume of catalyst required to treat one
volume per hour of flue gas. Space velocity varies with catalyst formula-
tion, catalyst shape, and control level. Typical values of space velocity
for various catalyst shapes are shown in Table 2.3.3-1. Also shown are
other catalyst design variables such as catalyst dimensions, gas velocities,
bed depth and pressure drop. Ranges of values are used since specific values
are different for each catalyst. The values shown pertain to 90 percent NO
removal and an NHs/NO,. mole ratio of 1:1.
X
Both NHa/NO ratio and space velocity will change with removal level.
The NH3/NOy mole ratio will range from 0.7-1.0 and the space velocity will
range approximately as shown in the table for control levels of 70 to 90
percent.l5
Variables associated with the boiler can also affect the performance
of these systems. These are
• flue gas flow rate
• NO concentration
boiler load variability
2-97
-------
TABLE 2.3.3-1. CATALYST DESIGN VARIABLES FOR VARIOUS CATALYST SHAPES
(Basis: 90% NOX removal at NH3/NOX ratio of 1:1,
350-400°C)
Catalyst size (mm)
Thickness
Opening
Gas velocity (m/sec)
Bed depth (m)
SV (1,000 hr"!)b
Pressure drop (mrnHjO)
Honeycomb
(metallic)
0.5-1
4-8
2-6
1-2
5-8
30-80
Honeycomb ,
tube (ceramic)
1.5-3
6-20
5-10
1.5-5
4-8
40-160
Parallel
(Ceramic)
8-10
8-14
5-10
4-6
1.5-3
80-160
Plate
(Metallic)
1
5-10
4-8
2-5
2-5
60-120
Velocity at 350-400°C in open column (superficial velocity)-
Gas volume (Nm /hr)/catalyst bed volume (m3).
The flue gas flow rate and control level determine the catalyst volume
(hence reactor size) necessary. Increases in either also increase the
reactor size. The NOV concentration is a function only of fuel type used
X
in the standard boilers. Higher concentrations require larger NHs storage
and vaporization equipment; reactor size is not affected. Boiler load can
affect several things including flue gas temperature, flow rate and NOV con-
X
centration. It is necessary to maintain reactions temperatures of 350 to
400°C and temperature control equipment may be necessary if the boiler
experiences large load variations. Where these variations are present,
some equipment overdesign may be warranted to insure a constant control
level. These variables are discussed in more detail in the section on moving
bed SCR systems for oil-fired boilers, Section 2.3.2.
Parallel flow SCR processes have been applied in Japan to several
residual oil-fired industrial boilers. Oil-fired utility boilers and other
sources with high particulate concentrations are also being treated. A list
of vendors of parallel flow SCR systems is presented in Table 2.3.3-2. Notes
2-98
-------
on the relative level of application are also shown. Four of the eight
vendors have applied their systems to oil-fired boilers indicating that
application of this technology to industrial boilers is technically feasible.
Parallel flow SCR systems have been applied to both industrial and utility
boilers. Specific applications are listed in Tables 2.3.3-3 and 2.3.3-4.
There have been no applications in the U.S. The tables indicate that the
parallel flow technology is designed primarily for residual oils and not
distillate oils.
TABLE 2.3.3-2.
VENDORS OF SCR PARALLEL FLOW SYSTEMS FOR
OIL-FIRED APPLICATIONS21
Vendor
Notes
Hitachi Zosen
Hitachi, Ltd.
JGC
Mitsui Engineering & Shipbuilding
Mitsubishi Heavy Industries
Ishikawajima-Harima Heavy Industries
Kobe Steel
Kawasaki Heavy Industries
Tested on pilot scale equipment
Tested on commercial scale equipment
Has not been tested on boilers
Tested on commercial scale equipment
Tested on commercial scale equipment
Tested on commercial scale equipment
Has not been tested on boilers
Tested on pilot scale equipment
TABLE 2.3.3-3.
EXISTING FGT INSTALLATIONS OF SCR PARALLEL FLOW
SYSTEMS OIL-FIRED INDUSTRIAL BOILERS21
Location
(Japan)
Sodegaura
Kawasaki
Chiba
User
Fuji Oil
Ajinomoto
Ukishima
Pet. Chem.
Process
Developer
Mitsubishi
H.I.
Ishikawaj ima
H.I.
Mitsui
Engineering
Fuel
Res id
Resid
Resid
Capacity
(Nm3/hr)
200,000
180,000
220,000
Completion
Date
January 1978
January 1978
April 1978
2-99
-------
TABLE 2.3.3-4. EXISTING FGT INSTALLATIONS OF SCR PARALLEL FLOW
SYSTEMS OIL-FIRED UTILITY BOILERS21
Location
(Japan)
Yokosuka
Chita
Kudamatsu
Niigata
User
Tokyo
Electric
Chubu
Electric
Chugoku
Electric
Tohoku
Electric
Process
Developer
Mitsubishi H.I.
Mitsubishi H.I.
Ishikawaj ima
H.I.
Ishikawaj ima
H.I.
Capacity
Fuel (Nm3/hr)
Resid 40,000
Resid 1,920,000
Resid 1,900,000
Resid 1,660,000
Completion
Date
March 1977
February 1980
July 1979
August 1981
2.3.3.2 System Performance—
The performance of several parallel flow catalysts is illustrated in
Figures 2.3.3-4 through 2.3.3-9. Table 2.3.3-5 shows several operating data
for a single parallel flow SCR installation. The data presented indicated
that NOX control levels of greater than 90 percent are obtainable through
selection of the appropriate process design variables. Other data are also
presented and these are discussed in subsequent sections.
2-100
-------
100
I
H
O
4J
IQ
>HC
•H 01
•MU
•rl-H
Q 01
V)
w
O
M
3
CO
in
S
a
1 cr
90 i
80 -
70 -
^"^a
50 -
40
30 •
20 -
10 -
0 -
Capacity
(Nm»/h)
2aooo
Fuel
Crude
oil and
hoavy
oil
SV value
and the like
SV-6.0001"1
360 1C
KJl»/HOx mol.
ratio 1.0
1.000 2,000 3,000 4,000 5,000 6,000 7,'OOd
Operation time (hrs)
8,000
Figure 2.3.3-4. Catalyst life test results.115
-------
o
to
^^
c
LU
I — f
CJ
1— t
U_
LU
O
.
0
Q
LU
Di
X
O
i.
ex
100 n
<
90-
80-
70-
60-
i
50-
)
40 -J
-» f
KXX>0ox/>tvo^ NOX
^">°^:H><>cK>xy<>CKX)^^
^p
K>-<*«l-«»K»-<><*<>-<>O-(M>K>K»-<»^^
A *f^ ^"^3
1 1 I 1
LU
03
1—
_l
CJ
o;
LU
Lu
CO
ni
-10^
o
^^
-202
o;
L0 ^
o
CM
-200 |
•150"-
n.
-100 |
LU
LO- i
• • LU I/)
0 1000 2000 3000 4000 ^ <2
TEST PERIOD (HR)
o
o
Figure 2.3.3-5. Durability of NOX removal catalyst for exhaust gas
of high sulfur oil burning boiler. 5
-------
100
90
80
70
60
50
40
30
20
10
Q.
Q.
re
-------
TOO
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
NH3/NOX MOLE RATIO
Figure 2.3.3-7. NHa/NOX mole ratio vs. NO removal
(plate catalyst; 350°C, LV 5.9 m/sec).117
2-104
-------
UJ
E
UJ
o
uj
tr
111
So*
P-"
z --5
U.S£
OC3
2 - O
811
o o
X O O
o « «>
98
96
94
92
100
8.0
6.0
4X5-
200
oT 160
O
cc
o
u 120
a:
tn
{2 80
-------
100
^»
<#>
c
o
•••i fin
•H
0)
c
o
•H
-------
TABLE 2.3.3-5.
SCR PLANT BY MITSUI ENGINEERING
AND SHIPBUILDING CO.20
Capacity (Nm /hr)
Gas composition
NOX (ppm)
SOX (ppm)
Dust (mg/Nm3)
Catalyst and reactor
Catalyst carrier
Catalyst shape
SV (hr'1)
Temperature (°C)
NHs/NC) mole ratio
X
NOX removal (%)
Total pressure drop
Leak NHs (ppm)
Operation start
Plant cost (106 yen)
220,000
150
300
100-150
TiO
PP
4,000
350-400
1.0
Above 90
180
Below 10
July 1977
260
2-107
-------
2.3.4 Ab sorpt ion-Oxidat ion
2.3.4.1 System Description—
Absorption-oxidation processes remove NOX from flue gas by absorbing
the NO or NOz into a solution containing an oxidant which converts the NO
X
to a nitrate salt. Two types of gas/liquid contactors can be used and exam-
ples of each type are shown in Figure 2.3.4-1. Both perforated plate and
packed towers accomplish NOX absorption by generating high gas/liquid inter-
facial areas. The choice of one type of contactor is a design decision made
to achieve a given removal for the least cost.
A generalized process flow diagram is shown in Figure 2.3.4-2. Flue
gas is taken from the boiler after the air preheater. Before the gas can
be sent to the NOx absorber, it must be S02-free since SOa consumes prohibi-
tive amounts of the costly liquid-phase oxidant. In most cases, the oxidant
is permanganate (MnOi*); however, Ca(C10)2 can also be used. Therefore, a
conventional FGD unit is required ahead of the NOX absorber. A prescrubber
to cool the gas and remove both particulates and Cl prior to FGD is also
necessary. After having passed through these two scrubbing sections, the
flue gas enters the distributing space at the bottom of the NOx absorber,
below the packing or plates. The gas passes upward through the column,
countercurrent to the flow of the liquid absorbent/oxidant (usually a KOH
solution containing KMnOi+). NOX is absorbed and then oxidized over the
length of the column according to the following reactions.31
N0(g) + NO(aq) (2-6)
NO(aq) + KMnO^aq) -> KN03 (aq) + Mn02(s) (2-7)
2N02(g) -> N20^(g) (2-8)
2-108
-------
FLUE GAS OUT
FLUE GAS OUT
Prindpol -
inJedoce
;
LIQUID IN
— Coolesced
dispersed
-Perforoted
plote
— Downspout
FLUE GAS IN
LIQUID IN
FLUE GAS IN
LIQUID OUT
Perforated Plate Absorber
Packed Absorber
Figure 2.3.4-1. Gas/liquid contactor options for
Absorption-Oxidation Processes.29
2-109
-------
Flue
Gas
Prescrubber
and
S02 Scrubber
•€
NOX
Absorber
To Reheat
and Stack
Holding
Tank
Oxidant
Make-up
Nitrate Treatment and
Oxidant Regeneration
Figure 2.3.4-2. Process flow diagram for absorption-
oxidation process.
30
2-110
-------
N20.,(g) -f N20i>(aq) (2-9)
N20i»(aq) + 2K2MnOit(aq) -»- 2KMn0.t (aq) + 2KN02(aq) (2-10)
Since most of the NO from combustion processes occurs as NO,32
reactions 2-6 and 2-7 predominate. The clean gas passes out of the top
of the absorber to a heater for plume buoyancy and is sent to the stack.
The absorbing solution drops to a holding tank where makeup KOH and/or
KMnCK are added. This solution flows to a centrifuge to separate the
solid Mn02 which is then electrolytically oxidized to MnCK . The remaining
solution is either concentrated in an evaporator to form a weak KNOs solu-
tion or is electrochemically treated to produce a weak HNOs solution and a
mixed stream of KOH and KNOs-
The fundamental design equation used for gas absorption column design
is
where y = bulk NO concentration (mole fraction) of gas phase at any
given point in column
y-y* = overall driving force for absorption (y* being the NO con-
centration of a gas in equilibrium with a given liquid NO
concentration)
Y, = inlet NO concentration
b x
Y = outlet NO concentration
a *
K = overall gas-phase mass transfer coefficient, Ib-moles NOX/
(ft2)(hr)(mole fraction)
2-111
-------
a = area of gas-liquid interface per unit packed volume, ft2/ft'
gas/ v.ii"
length of packed section of column, ft
G = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
In a column containing a given packing or plate configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow rate/
cross-sectional area of column) is called the flooding velocity. At this
point, the gas flow completely impedes the downward motion of the liquid
and blows the liquid out of the top of the column. The gas velocity, obvi-
ously, must be lower than the flooding velocity. How much lower is a design
decision. Often, it is an economic tradeoff between power costs and equip-
ment costs. A low gas velocity will lower the pressure drop and, hence, the
power costs but the absorber will have a larger diameter and cost more. High
gas velocities have an opposite effect. Usually the optimum gas velocity is
about one-half the flooding velocity.34 The height of the column depends on
the desired level of removal and on the rate of mass transfer. The latter
is a major problem for these systems trying to achieve large NO reductions
since NO is relatively insoluble in water. This can be seen in Table 2.3.4-1.
TABLE 2.3.4-1. NITROGEN OXIDES CHARACTERISTICS35
Boiling Point, Solubility in Cold Solubility in Hot
°C Water (0°C), cm3 Water (60°C), cm3
NO
NO 2
-151.8
21.2
7.34/100 cc H20
soluble, decomposes
2.37/100 cc H20
One can see that NO has a very limited solubility in water and, since most
NOX is present as NO, the rate of mass transfer (absorption) is going to be
relatively slow. This means that the absorber must be tall with a high
2-112
-------
liquid flow rate. Table 2.3.4-2 presents the effects of boiler/flue gas
variables on the design of absorption-oxidation systems.
TABLE 2.3.4-2. SYSTEM DESIGN CONSIDERATIONS
Variable Design Effect
Presence of participates Requires prescrubber
Presence of SOa Requires FGD pretreatment
Increased gas flow Requires larger column diameter; increased
liquid flow rate
Increased NOX concentration Requires larger column height; increased
oxidant concentration
Both flue gas flow rate and NOX concentration can be affected by boiler
operating conditions. Therefore a change in load on an industrial boiler
may alter these variables markedly. The absorber must be designed to accom-
modate any anticipated load changes. The column size and the liquid and
oxidant flows must be designed for each application after examining the
boiler operating history and establishing ranges of variation.
None of the sources consulted for this study could supply typical ranges
for operating variables such as liquid/gas ratio, reagent concentrations or
pressure drops and, as a result, none are presented here. Economic data were
not presented either. One source did estimate the removal for absorption-
oxidation processes to be 85 percent.
Presently, absorption-oxidation processes are still in the pilot unit
stage of development. Table 2.3.4-3 presents a list of absorption-
oxidation process vendors and the status of development of their projects.
2-113
-------
TABLE 2.3.4-3. PROCESS VENDORS OF ABSORPTION-OXIDATION PROCESSES37'38
Vendor Status of Development
Hodogaya No information available; stopped development
on process
Kobe Steel 1974: 1000 Nm3/hr gas from iron-ore sintering
furnace; stopped development on process
MON (Mitsubishi Metal, MKK, 1974: 4000 Nm3/nr flue gas from oil-fired
Nikon Chemical) boiler
Nissan Engineering 1972: 4 pilot plants, 100-2000 Nm3/hr tail
gas from HNOa plant
Only one of the process vendors has piloted this process on flue gas from an
oil-fired boiler and there have been no commercial applications either in
Japan or the U.S.
2.3.4.2 System Performance—
No information has been published on tests conducted with flue gas from
oil-fired boilers. The relative insolubility of NO in water may present a
major obstacle to achieving the stringent level of control (90 percent NO
reduction) by absorption-oxidation processes. Another primary drawback of
absorption-oxidation systems is the production of nitrate salts (see Equation
2-7), a secondary pollutant. These processes probably could not be applied
on a large scale as wastewater treatment systems (chemical or biological) do
not remove nitrogen compounds from the wastewater.39 Trying to recover the
nitrates as nitric acid for industrial use or potassium nitrate for fertilizer
does not seem promising as the by-products are of low quality. Also, the use
of an expensive, liquid-phase oxidant requires stainless steel and other cor-
rosion resistant materials of construction. High sulfur coals require an FGD
system prior to the NO absorber to prevent excessive oxidant consumption by
X
S02. The process steps of several absorber columns in series (large fan re-
quirements), oxidant regeneration (electrolysis), and flue gas reheat (inline
heater) are all energy intensive and present technical and economic disadvan-
tages.
2-114
-------
2.3.5 Selective Catalytic Reduction-N0x/S02 Removal
2.3.5.1 System Description—
From a NOX removal standpoint, this process is very similar to those
discussed in Sections 2.3.1 through 2.3.3. The primary difference is the
additional equipment necessary to collect and process the SOz. The main
feature of the process is the reactor and catalyst which remove both NO and
SOz • This process was developed by Shell although the U.S. licensor, UOP,
is currently marketing and developing the process. The NO /SOz version of
the process is commonly called the SFGT process which stands for the Shell
Flue Gas Treatment Process.
A uniquely designed parallel flow type of reactor is used to avoid
problems with particulates. This design is necessary only with flue gas
from residual oil-fired boilers. The reactor consists of a series of pack-
ages containing catalyst material, arranged in a parallel configuration which
allows flue gas flow between the packages. Each package consists of catalyst
material placed between two layers of wire gauze. Figure 2.3.5-1 illustrates
the internals of the parallel passage reactor. The flue gas flows between
the catalyst packages and not directly through the catalyst material. This
prevents plugging of the catalyst with particulate matter in the flue gas.
For convenient fabrication and handling, catalyst packages of a standard
size are appropriately spaced and placed in a container to form a unit cell
or module. S02 removal efficiency and capacity are determined by the number
of unit cells placed in series in a cell stack. For a given level of S02
removal, a greater number of cells in the stack increases the capacity and
reduces the frequency of regeneration. The number of stacks is determined
largely by the flue gas rate and the flue gas velocity through a single stack
is generally not a design variable. For most design situations, 4 to 5 unit
cells in a stack are adequate to achieve high SO2 removal, however, if a high
level of denitrification is required, more unit cells per stack may be neces-
sary.
2-115
-------
REGE1M. GAS 11
tl PURGE OFF-GAS
TREATED
FLUE GAS
FLUE GAS
REGEN. OFF-GAS IA
tj PURGE STEAM
UOP1U4
Figure 2.3.5-1. The SFGT parallel flow reactor.
1*0
The SFGT process is a dry process with two or more reactors operating
in a cyclic manner. The desulfurization aspect of the process is regenerable,
while NO removal is accomplished by catalytic reduction with ammonia. The
catalyst material is commonly called an acceptor since S02 removal involves
adsorption or "acceptance" of S02. The desulfurization cycle consists of
the following steps:
1) oxidation of accaptor bed/acceptance of S02>
2) purge reactor,
3) regeneration with reducing gas, and
4) purge reactor.
The products of the oxidation and acceptance reactions in step 1 above
catalyze the reaction of NO with ammonia to form nitrogen and water. NO
X X
removal is accomplished by metering ammonia into the untreated flue gas
upstream of the reactors. The catalytic reaction takes place across the
partially spent acceptor beds.
2-116
-------
Also associated with the SFGT process are facilities for generating
reducing gas and for the processing of SO2 in regeneration off gases into
sulfur by-products. Figure 2.3.5-2 illustrates the process flow for a
typical SFGT system.
Boiler flue gas is withdrawn upstream of the air preheater and particu-
late removal device by the SFGT system fan and discharged to the reactor
inlets. The flue gas then flows through fixed bed reactors in open channels
alongside and in contact with the acceptor material. Ammonia is added to the
flue gas upstream of the SFGT system fan to insure complete mixing before the
flue gas enters the reactor.
Fresh acceptor material is elemental copper on an alumina support. This
is converted to the oxide form by flue gas oxygen shortly after initiation
of the acceptance cycle. SOz is removed by reaction with the copper oxide
and oxygen as the flue gas flows through the channels, converting the accep-
tor material to copper sulfate. Simultaneous with the desulfurization pro-
cess, the reduction of flue gas NOX by ammonia is selectively catalyzed by
copper oxide and copper sulfate in the acceptor bed. As the flue gas leaves
the SFGT system reactors it is returned to the boiler flue gas duct down-
stream of SFGT fan suction.
Flue gas is fed to a reactor until an unacceptable amount of S02 begins
to pass through the reactor. This occurs when a large fraction of the accep-
tor has been converted to the sulfate form. Flue gas flow is then diverted
to another reactor and the spent reactor is isolated. Any flue gas remaining
in the spent reactor is purged with an inert gas such as steam, and the re-
generation cycle is initiated.
Regeneration is accomplished by passing a reducing gas through the bed
countercurrent to the direction of the flue gas flow. The reducing gas,
which is primarily hydrogen, reacts with the copper sulfate in the spent
reactor to convert it to elemental copper. An off gas of 862 and water is
2-117
-------
I
M
H
OO
GAS
PflRTICULATE REMOVAL
AND STACK
NH
OFF
GAS
PRODUCT
(S,S02 U),ORH2SO«>
Figure 2.3.5-2. Flow diagram of the SFGT process.
11
-------
produced by the reaction. After regeneration is complete, the reactor is
again purged with steam and is ready for another acceptance cycle. Regenera
tion gas can be produced from a number of sources, but steam-naphtha reform-
ing is proposed by UOP as being the most economical.42
The regeneration off-gas treatment section consists of flow smoothing
equipment and SOa workup equipment. Typically, the regeneration off-gas is
cooled and most of the steam condensed, raising the SOa concentration from
10 percent to 80 percent by volume. The concentrated SOa is then compressed
into an intermediate holding vessel to provide a smooth flow rate to the
workup section. The workup section may be a modified Glaus unit which pro-
duces an elemental sulfur by-product, a fractionation unit which produces
liquid SOa, or a sulfuric acid plant.
Each process step consists of different chemical reactions. The
acceptor material is converted to the oxide form by the following reaction:
Cu + J^Oa ->• CuO (2-12)
This oxide readily reacts with flue gas SOa and oxygen, as described by:
CuO + hQz + S02 -»• CuSOit (2-13)
SOs in the flue gas is also removed by the following reaction:
CuO + 80s -*1 CuSOn (2-14)
The reaction scheme for reduction of NOX is described by the following:12
4NO + 4NH3 + 02 * 4N2 + 6H20 (2-1)
2NOa + 4NH3 + 02 £ 3N2 + 6H20 (2-2)
2-119
-------
Excess ammonia which is not consumed in reactions 2-1 and 2-2 may be cataly-
tically oxidized to nitrogen and water by reaction with flue gas oxygen, as
described by:
4NH3 + 302 ->- 2N2 + 6H20 (2-15)
Maximum NOX removal efficiency is achieved at the point of S02 breakthrough,
where conversion of the acceptor material from the oxide to the sulfate form
is essentially complete. Figure 2.3.5-3 illustrates reactor outlet S02 and
NO concentrations during a typical SFGT acceptance cycle.
A different set of reactions is involved during regeneration of the
catalyst bed.
Copper sulfate is reduced to the elemental copper form by reducing gas
hydrogen as described by the following reaction:
CuSO.* + 2H2 •*• Cu + S02 + 2H20 (2-16)
Any acceptor material present in the reactor as the oxide will also be
reduced, according to the following reaction:
CuO + H2 -*• Cu + H20 (2-17)
The regeneration step occurs at the same temperature as the acceptance step,
400°C (750°F).
The general reactor design equation is the same as that described in
earlier sections for SCR processes. The primary variables are the gas rate,
reaction rate, and control level. Reaction rate data have not been released
for this process except that the N0x reduction is first order.
2-120
-------
450
4001
E
Q.
O.
z
g
UJ
O
z
O
O
3
O
-------
The gas flow rate and control level will determine the reactor size.
Increases in either variable will increase the reactor volume. The effect
of control level can be seen in Figure 2.3.5-4. It is necessary for the
flue gas to enter the reactor at 400°C and therefore it must be taken from
an appropriate point in the boiler, most likely from between the economizer
and air preheater. Alternatively, a cooler gas can be heated to 400°C by an
inline heater.
The removal efficiency of NO for a given reactor size is determined
by the amount of NH3 injected as shown in Figure 2.3.5-5. Since the reac-
tion is first order in NOX, control level is not affected by NO concentra-
tion. '+7 The S02 control efficiency is primarily a function of the acceptance
time of the reactor (Figure 2.3.5-3). Typical ranges of operating variables
are shown in Table 2.3.5-1.
Since the SFGT system can handle full particulate loading (£10 gr/sft3)
it is not dependent on any pretreatment facilities. Also, the SFGT system
operation is independent of boiler operation. The system fan takes suction
from the flue gas duct between the economizer and air preheater and the reac-
tor discharge returns to the boiler flue gas duct just downstream of the
suction point, with no valves between the two points. The system fan pro-
vides a constant flow rate through the SFGT system. If the boiler flue gas
rate is greater than the fan rate, flue gas will bypass the system through
the open duct. If the boiler flue gas rate is lower than that of the system
fan, treated gas will recycle back to the system fan suction. Recycle of
treated gas to the reactor inlet with "open bypass" arrangement presents no
operating problems. This is due to the fact that both the level of desulfuri-
zation and denitrification are independent of inlet concentrations, and the
system does not humidify the flue gas.
2-122
-------
V
\
^
v
^
\;
^
hN
\
\
^
i
\
\
\
\
CONO
, 400°
CuA
NX,
«RF
BPERF
AFFC
i
'\
\
\
X
ITIONS:
C
SCuSO4
NO 1.1 ~
MAL tXPEC
ORMANCE
ORMANCI
CTEOBY
SIDE FACTC
'»
\
\
1.1
TfO
*.
\
1 4
UO UN01R UttttM
Figure 2.3.5-4. Unconverted NO as a function of catalyst bed length.
1UU
90
80
70
60
SO
30
20
10
£
A
1
¥
/
/
!
cf
• o o ,,
^r ft
/
O
15 METER BENCH &
2000 SOj
4% O}
TTCI CHSV
OJS4 (000
O 400 tooe
A 450 (000
CA1EUNIT
ol Cu
14) 10
MH)/MO MOlf RATIO
Figure 2.3.5-5.
NOX reduction with NHa over commercial SFGT acceptor.
2-123
-------
TABLE 2.3.5-1. DESIGN AND OPERATING VARIABLES FOR SFGT SYSTEM1*8
Variable Typical Range
Space Velocity 5,000 - 8,000 hr~l*
NH3:NOX Mole Ratio 1.0:1.0 to 1.2:1.0*
Flue Gas Temperature 400°C
Pressure Drop 5-6 in.
Maximum Particulate Loading >23 g/Nm
*Actual value will depend on required removal level.
-------
Tables 2.3.5-2 and 2.3.5-3 present the test and commercial applications
of r.he SFGT process. The development history of the process can also be
seen in these tables. In the U.S., from 1974 to 1976 a pilot scale unit
at Tampa Electric Company (TECO) was operated using flue gas from a coal-
fired boiler. Testing was for 862 removal only, NO control was not at-
tempted during this period. The process developer is currently modifying
the TECO pilot unit to accommodate 7 meters of bed height, up from the
previous maximum of 5 meters. This should permit simultaneous removal
of NO^ and SO^ to the 90 percent level. Also, provisions are being made
X X
for injection of a CO/C02 gas mixture into the regeneration gas in order
to simulate medium-Btu gas from a coal gasifier.
The costs for an industrial size boiler have not been estimated. The
only detailed cost estimates currently available are for coal-fired utility
boilers. These are shown in Tables 2.3.5-4 through 2.3.5-6. Also shown are
the estimated energy and raw material requirements.
2-125
-------
TABLE 2.3.5-2. SFGT PROCESS, PILOT AND DEMONSTRATION UNITS
K>
Location/
Company
Shell Ref.
at Pernis
Rotterdam
Utility
Tampa Elec.
Big Bend
JGC
Yokohama
Nippon
Steel
Fuel/
Designed By Application
Shell Residual
Fuel Oil-
Proc. Heater
Shell Coal-
Steam Boiler
UOP Coal-
Wet-Bottom
Utility Boiler
JGC* Fuel Oil
JGC Sintering
Furnace
JGC Coke Oven
Gas
Size, Type of
Nm /hr Operation
600-1000 S0x-only
Heavy Fly Ash
Loading
1200-2000 S0x-only;
SOX-NOX
Simultaneous
250-700 NO -only
2000 N0x-only
400 N0x-only
Dates Comments
1967-1972 S0x reduction -
1 approx. 90%
1971 Particulate mat-
ter - loadings to
20 Gr/Nra3
1974-1976 S0x - 90%;
1979- S0x-N0 - 90/90%
fly ash to
25 Gr/Nm3
1974- NOX reduction -
90-99%
1976-1978 NOX reduction -
90-97%
1976-1977 NOX reduction -
90%; special low
temp. cat. evalua-
tion
*JGC Corporation, licensing agent in Japan.
-------
TABLE 2.3.5-3. SFGT PROCESS, COMMERCIAL UNITS
Unit
SYS*
Yokkaichi
Kashima Oil
Co. Ltd.
Fuji Oil
Co. Ltd.
i
£| Nippon Steel
Corp.
Fuel/
Designed By Application
Shell Residual
Fuel Oil-
Ref. Boiler
JGC Fuel Oil-
Process Unit
Heater
JGC CO Boiler
JGC Sintering
Furnace
Size, Type of
Nm3/hr Operation
125,000 S0x-only;
NOX-SOX
Simultaneous
50,000 N0x-only
70,000 NO -only
X
150,000 NO -only
X
Dates Comments
1973-1975 SOX reduction
1975- Simultaneous -
1975- 95-98%
1976- 93-96%
1978- ^95% (low temp
lyst)
- 90%;
90/50%
. cata-
*Showa Yokkaichi Sekiyu
-------
TABLE 2.3.5-4. ECONOMICS OF SFGT SYSTEM
>f 9
Incorporated Units:
Power Plant Size
Fuel
S-Content, Wt-%
Case 1
Case 2
Case 3
HHV
Heat Rate
Excess Air
Air Preheater Leakage
BASIS:
Steam-Naphtha Reformer
SFGD Reactor Section
Compressor/Gasholder Flow
Smooth Section
Modified Claus Unit
500 MW
Coal
3.5
2.5
0.8
10,500 Btu/lb
9,000 Btu/kWh
20%
13%
Flue Gas Rate
SO2 Content, ppmv
Case 1
Case 2
Case 3
1,582,000 Nm3/h (983,000 SCFM)
2,580
1,850
590
Mid-1977, Gulf Coast Location
Load Factor
Capital Charges
Cost of:
Naphtha
Steam (40 psi, SAT.)
Electricity
Labor
Heat Credits
Sulfur
7,000 h/a
15%/a
$0.35/gal
$1.50/M Ib
$0.018/kWh
$10.00/hr
$2.50/MMBtu
$45.00/ton
2-128
-------
TABLE 2.3.5-5. ECONOMICS OF SFGT SYSTEM ESTIMATED
CHEMICALS AND UTILITY REQUIREMENTS
50
Case 1
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
Case 2
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
Case 3
Electricity
Steam**
Naphtha***
Heat Credits
S° Produced
SFGD
Section
kW 5,770
kmol/h 1,820
Gcal/h
Gcal/h
kg/h
kW 5,800
kmol/h 1,300
Gcal/h
Gcal/h
kg/h
kW 5,120
kmol/h 480
Gcal/h
Gcal/h
kg/h
Flow Mod .
Smooth Glaus
Section Section
850 115
-380* -740*
5250
570 82
-270* -530*
3760
180 30
-95* -170*
1200
Reformer
Section
480
-600*
90.92
300
-415*
62.75
110
-140*
21.01
Total
7215
100
90.92
42.53
5250
6782
85
62.75
32.48
3760
5440
75
21.01
18.46
1200
*Produced
**40 psig, Saturated
***5.175 MMBtu/Bbl produces 11,500 SCF Hydrogen/Bbl
2-129
-------
TABLE 2.3.5-6. ECONOMICS OF SFGE SYSTEM ESTIMATED
CAPITAL -AND OPERATING COST
51
EEC. (MM$)
SFGD Reactor Section
Compressor/Gasholder
Modified Glaus
Steam-Naphtha Reformer
Estimated Annual Revenue
Requirements (M$/a)
Capital Charges
Maintenance
Labor
Acceptor
Electricity
Steam
Naphtha
Heat Credits
Sulfur Credits
Capital Cost, Operating Cost,
Energy Requirement
Capital Cost, $/kW
Operating Cost, c/kWh
Energy Requirement, Btu/kWh*
Case 1
28.95
7.82
2.76
8.81
7251
967
123
1479
909
42
7174
-2977
-1570
97
0.38
525
Case 2
28.53
6.10
2.26
7.14
6604
881
123
1053
855
35
4951
-2273
-1126
88
0.32
371
Case 3
22.94
2.65
1.14
4.17
4634
618
123
411
685
31
1658
-1292
-359
62
0.19
124
*Defined as the sum of:
Electricity at
Steam at
Naphtha at
Heat Credits at
9000 Btu/kWh
40000 Btu/kmol
4 Btu/kcal
4 Btu/kcal
2-130
-------
2.3.5.2 System Performance—
NOX control by the SFGT process is shown graphically in Figure 2.3.5-5.
and in Figure 2.3.5-3 presented earlier. As can be seen, at a space velocity
of 8000 hr"1, NOX control of >80 percent can be achieved. Figure 2.3.5-4
indicates that the process developers feel the process to be capable of NO
control levels of >90 percent. Ando indicates that NO and SO removals of
70 percent and 90 percent, respectively are achievable at an NHa/NO mole ratio
of 0.99.120 He also indicates that higher NOX control may be possible. But
unless some process modifications are made, S02 control will decrease and
emissions will increase.
As mentioned earlier, the system is not impacted by changes in the
boiler gas rate or particulate-concentrations. Changes in the NO concen-
tration due to boiler load changes and be compensated for by conventional
control system used in conjunction with the NHs injection equipment. This
control system will be developed during the upcoming pilot tests at the TECO
pilot plant.
2.3.6 Adsorption
2.3.6.1 System Description—
The adsorption process removes NOV and SOz from flue gas by adsorbing
X
them onto a special activated char. Adsorbed NO is reduced to Na while SOz
X
is reduced and condensed to an elemental sulfur by-product.
A process flow diagram is shown in Figure 2.3.6-1. Flue gas is taken
from the boiler air preheater and passed through a particulate removal device
to prevent blinding of the adsorption bed. The flue gas then enters the ad-
sorber, a vertical column with parallel louver beds containing the char in
pellet form. NO and S02 are adsorbed on the char which slowly moves down-
ward through the bed. The NO adsorption mechanism is unknown but SOa under-
goes the following reaction.55
2-131
-------
to
I
OJ
N3
J '
FLUE
GAS
AIR
STACK
—* AOSOR3ER -ii
REGENERATOR
CRUSHED
COAL
CONDENSER
ASH
SULFUR
TO FLUE GAS
ENTERING AIR
HEATER
AIR
Figure 2.3.6-1. Flow diagram of Foster Wheeler-Bergbau Forschung
Dry Adsorption Process.51*
-------
S02(g) + H20(g) + hOz(g) ->• H2S04(1) (2-18)
The reaction product is held in the pores of the char pellets. The
flue gas exits the adsorber and passes to the stack. The saturated char
leaves the bottom of the adsorber and is screened to remove any fly ash
deposits. It is then conveyed to a regenerator where it is mixed with hot
sand (650°C) and the following reactions take place.55'56
2H2SCM1) + C(s) -> C02(g) + 2H20(g) + 2S02 (g) (2-19)
2NO(g) + C(s) -»• C02(g) + N2(g) (2-20)
This S02-rich gas product stream is sent to an off-gas treatment reactor
containing hot, crushed coal (650-820°C) and the following reactions take
place.56
S02(g) -»• S(g) + 02(g) (2-21)
C(s) + 02(g) -»• C02(g) (2-22)
The gas then passes to a condenser where the S vapor forms molten S. The
char/sand mixture from the regenerator is screened to separate the two solids.
The char is recycled to the adsorber via a spray cooler and the sand is re-
cycled to the regenerator after passing through a heater.
This process operates at 120-150°C, however, typical values for other
operating variables were not found. NOX and SOz control levels were reported
to be 40-60 percent and 80-95 percent, respectively.57 The economics of the
process vary with the fuel sulfur level. For coal-fired boilers with fuel
sulfur levels of 0.9-4.3 percent, the capital costs range from $40-90/kW and
the operating costs range from 1.0-2.3 mills/kWh.58 The costs were based on
applying the process to a utility boiler of >200 MW capacity. Costs for oil-
fired applications were not found.
2-133
-------
Presently, the adsorption process is in the prototype unit stage of
development. The one reported process developer in the field, Foster Wheeler-
Bergbau Forschung has a 20 MW prototype unit and several small pilot plants
treating coal-fired flue gas. The process should also be applicable to oil-
fired boilers.
2.3.6.2 System Performance—
Tests have shown the adsorption process to be primarily a SOa reduction
process as NO removal efficiency averages 40-60 percent while SOa removal
r Q
had a range of 80-95 percent.
The primary drawback of this process, besides the low NO removal level,
X
is its complexity: numerous process steps involving hot solids handling.
Solids flow can be difficult to control and high maintenance requirements
could be expected. The vendor has reported several mechanical problems
during testing which included control of adsorber-bed levels, poor char
distribution, char-sand separation, hot sand conveying, and char cooling
and feed. Some corrosion-resistant material is needed in the high tempera-
ture zones of the process. The ash waste stream from the off-gas treatment
reactor appears to be the sole secondary pollutant associated with the pro-
cess. The overall complexity and low NOX removal of the process present
definite technical disadvantages.
2.3.7 Electron Beam Radiation
2.3.7.1 System Description—
This dry process utilizes an electron beam to bombard the flue gas,
removing NOX and S02 in the process. A block flow diagram for the process
is shown in Figure 2.3.7-1.
2-134
-------
Electron Beam
Accelerator
Flue
Gas
V Y
Reactor
Fly Ash
Off-Gas
Solid
Residue
By-product
Treatment
Figure 2.3.7-1.
Disposable or
Salable By-product
Process flow diagram for Ebara-JAERI
electron beam process.60
Flue gas is taken from the boiler air preheater and passed through a
cold ESP to remove particulates. After a small amount of ammonia is added,
the gas enters a reactor where it is bombarded with an electron beam. (The
penetration of the gas stream by the beam will require a unique discharge
pattern or other special design considerations.) A powder containing both
ammonium nitrate and sulfate is generated by an unknown reaction mechanism.
The gas then exits the reactor, passes through a second ESP to remove the
solid by-product, and is sent to the stack. The by-product treatment system
is still being developed. Various methods investigated include thermal de-
composition in the presence of an inert gas, steam roasting with CaO, or
steam roasting with HaO. The by-product may eventually be useful as a fer-
tilizer.61
2-135
-------
The key subsystem of this process is the electron beam accelerator.
Control of this unit's power supply is based upon inlet composition, flow
rate, and temperature of the flue gas.
Some of the important variables and typical ranges are listed in
Table 2.3.7-1.
TABLE 2.3.7-1. SYSTEM VARIABLES62
Typical Value
Temperature
Reactor residence time 1-20 sec
Radiation rate 10s-106 rad*/sec
Total radiation absorbed 1-3 Mrad*
*Rad is the radiation dose absorbed
1 rad = .01 J/Kg
The operating cost with NO removal only (low sulfur oils) is lower
due to lower radiation levels, but the capital cost would be just as high
as for simultaneous NOX/SOX removal. Capital costs are quite high for this
process as the 2 ESP's and the accelerator are expensive. The costs for a
1000 Nm3/hr test unit are reported to be $1000/kW; however, the costs of a
full scale system are expected to be lower. Operating costs are not
available.
The Ebara Manufacturing Company in conjunction with Japan Atomic Energy
Research Institute (JAERI) has operated a 1000 Nm3/hr pilot plant treating
flue gas from an oil-fired boiler. In 1976, a 3000 Nm3/hr pilot plant began
treating ofi-gas from an iron ore sintering furnace at Nippon Steel. By-
product treatment technology needs to be more fully developed before this
process can be applied commercially.
2-136
-------
In the U.S., the Department of Energy (DOE) is funding development of
an electron beam process offered by Research-Cottrell. Pilot unit tests
with flue gas are scheduled, however, the details of the program are not yet
available.
2.3.7.2 System Performance—
A summary of the oil-fired pilot tests is shown in Figure 2.3.7-2.
100
80
o
-ri
u
td
"Is 60
o
2 3
Total beam (Mrad)
Figure 2.3.7-2. Oil-fired pilot plant test results.
61*
One can see that NOX/SOX removal drops off drastically at a total radiation
dose below 1 Mrad while the maximum removal is obtained at about 3 Mrad.
The removal efficiencies decrease as the concentrations of NOX and SOX
increase as can be seen in Figure 2.3.7-3.
2-137
-------
OJ
00
2O 400 600 800 1000 1200
CONCENTRATION OF NO, Oft SOZ , PPM
1400 1600
Figure 2.3.7-3. Effect of pollutant concentration on removal efficiency.
65
-------
2.3.8 Absorption-Reduction
2.3.8.1 System Description—
Absorption-reduction processes simultaneously remove NO and SOa from
flue gas by absorbing them into a scrubbing solution. The processes are
based on the use of chelating compounds, such as ethylenediamine tetraacetic
acid (EDTA) complexed with iron, to "catalyze" the absorption of NOX- Most
process vendors prefer a perforated-plate type of gas-liquid contactor. The
advantages of a perforated-plate absorber over a packed bed absorber include
easier cleaning when solids are present, wider operating ranges, and more
economical handling of high liquid rates.66 An example of a perforated plate
absorber is shown in Figure 2.3.8-1. The most common design of a perforated
plate is one that employs liquid crossflow over the face of the plate with
the gas passing upward through the plate perforations. A schematic of the
operation of a crossflow perforated plate is shown in Figure 2.3.8-2. The
liquid is prevented from flowing through the plates by the upward flow of
the gas. However, during periods of low gas flow (such as load changes on
industrial boilers) liquid can drain through the openings in the plates.
This reduces the liquid's time of contact with the gas on each plate and may
decrease the overall operating efficiency of the absorber. To prevent this
problem, there are two other types of dispersers utilized besides the basic
sieve-plate: the valve-plate and the bubble cap, depicted in Figure 2.3.8-3.
As the gas flow lowers, the valve or cap settles, sealing off the perforation
so liquid cannot drain through. This design feature allows the perforated
plate absorber to operate more efficiently at widely fluctuating gas rates.
While most all absorption-reduction processes utilize ferrous chelating
compounds to enhance NO absorption, the scrubbing solutions, the by-product
treatment and sorbent regeneration chemistry differ from process to process.
For this reason, one of the simpler absorption-reduction processes, that of
Kureha Chemical Industry Company, is examined here in detail.
2-139
-------
Principol
interface
Coalesced
dispersed
Perforated
plate
Downspout
RUE ws
^igure 2.3.8-1. Perforated plate absorber option for
Absorption-Reduction Processes.z8
2-140
-------
Plate n-\
Plate n
Figure 2.3.8-2.
Normal operation of sieve plate. Za, height of
station a above datum. ZGr, weir crest. Z^,
liquid-friction head. Zp, pressure head across
plate. Z^, net head in down pipe. Zy, weir
height.67
GOJ flow
Valve-plate diipenen.
•Valve closed
Valve open
Holes, punched
Z to 4 in. dcom.
(a) Circular or bell cap. (fc) Tunnel cap.
Bubble cap dispersers
Figure 2.3.8-3. Other gas dispersers.
6 8
2-141
-------
A block flow diagram of the Kureha absorption-reduction process is
shown in Figure 2.3.8-4. Flue gas is taken from the boiler after the air
preheater. It passes through a prescrubber to adiabatically cool the gas
and remove both particulates and chlorides. The flue gas then enters the
distributing space at the bottom of the NOX/SC>2 absorber, below the plates
or packing. The gas flows upward, countercurrent to a sodium acetate
(CHsCOONa) scrubbing solution (^60°C) containing ferrous iron and EDTA and
a few seed crystals of gypsum (to prevent scaling). Most of the S02 is
rapidly absorbed at the bottom of the absorber according to the following
reactions.
S02(g) + S02(aq) (2-23)
S02(aq) + 2CH3COONa(aq) + H20 + Na2S03(aq) + 2CH3COOH(aq) (2-24)
The NOX (which consists mainly of NO) is relatively insoluble; therefore, it
is absorbed gradually over the length of the column. The ferrous chelating
compounds effect on NO absorption is described in Figure 2.3.8-5. The NOX
is absorbed and undergoes the following reactions.73
N0(g) •* NO(aq) (2-6)
2N02(g) •* N20n(aq) (2-25)
N20<4(g) -> N204(aq) (2-9)
2NO(aq) + 5Na2S03(aq) + 4CH3COOH(aq) -»• 2NH(S03Na)2 (aq) + Na/S0n(aq)
+ 4CH3COONa(aq) + H20 (2-26)
2N2Ou
-------
Water
K3
I
Gypsum
Figure 2.3.8-4. Process flow diagram of Kureha absorption-reduction process.69'70
-------
a
ui
CO
CO
<
0 0.01 0.02
EDTA-Fe(II), mole/liter
Figure 2.3.8-5. EDTA-Fe(II) concentration and NO absorption at 50°C.
72
Some of the acetic acid (CH3COOH) formed at the bottom of the absorber via
reaction (2-24) is vaporized. It must be captured and is done so by water
scrubbing at the very top of the absorber. From the top of the absorber
column the clean flue gas passes to a heater for plume buoyancy and is then
sent to the stack. The liquid effluent drops from the bottom of the absorber
to a gypsum, CaSOi+"2H20, production reactor. Here, the solution is mixed with
with the purge stream from the acetic acid recovery section and a lime slurry
stream. The lime, Ca(OH)2, treatment involves the following reactions.7"*
2CH3COOH(aq) + Ca(OH)2(aq) + (CH3COO)2Ca(aq) + 2H20
(2-28)
(CH3COO)2Ca(aq) + Na2SQlt(aq) + 2H20 + CaSCK • 2H20(s) 4- + 2CH3COONa(aq) (2-29)
2-144
-------
The gypsum formed by reaction 2-29 is centrifuged. Most of the liquor
discharged is returned to the gypsum reactor and on to the absorber. The
remaining liquor is sent to a reactor where sulfuric acid (HaSOO is added
to hydrolyze the imidodisulfonate, NH(S03Na)2, by the following reaction.75
H«-
NH(S03Na)2(aq) + 2H20 0, NH^HSO* (aq) + Na2SO^(aq) (2-30)
The effluent from this reactor is then recycled to the gypsum production
reactor. A small purge stream is taken from the gypsum reactor to another
reactor where the ammonium bisulfate (NHitHSCH) formed in the hydrolysis
reaction is treated with lime to yield gypsum and NH3 off-gas by the follow-
ing reaction.76
NHijHSOit (aq) + Ca(OH)2(s) -»• CaS(K -2H20(s) -I- + NH3 (g)+ (2-31)
The gaseous ammonia is stripped from the solution by an air stream. If no
use for the ammonia can be found, the gas mixture is sent to a catalytic
reactor where ammonia is oxidized by the following reaction.
4NH3(g) + 302(g) 2N2(g) + 6H20(g) (2-32)
350°C
The product stream is then sent to the deacetating section of the absorber
column.
is
The fundamental design equation used for gas absorption column design
32
fib
I dy
I (y-y*)
•'Y
a
2-145
-------
where y = bulk NOX concentration (mole fraction of gas phase at any
given point in column
y-y* = overall driving force for absorption (y* being the NOX concen-
tration of a gas in equilibrium with given liquid NOX
concentration)
Y, = inlet NCI concentration
b x
Y = outlet NOV concentration
a x
K = overall gas-phase mass transfer coefficient, Ib-moles N0x/
(ft2)(hr)(mole fraction)
n n
a = area of gas-liquid interface per unit packed volume, ft /ft
G = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
Z = length of packed section of column, ft
In a column containing a given plate or packing configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow rate/
cross-sectional area of column) is called the flooding velocity. At this
point, the gas flow completely impedes the downward motion of the liquid and
blows the liquid out of the top of the column. The gas velocity, obviously,
must be lower than the flooding velocity. How much lower is a design deci-
sion. Often it is an economic tradeoff between power costs and equipment
costs. A low gas velocity will lower the pressure drop and, hence, the
power costs but the absorber will have a larger diameter and cost more.
High gas velocities have an opposite effect. Usually the optimum gas
velocity is about one-half the flooding velocity.33 The height of the
column depends on the desired level of removal and on the rate of mass
transfer. The latter consideration is the reason why a chelating compound
is used in absorption-reduction processes to aid in NO absorption. Table
2.3.8-1 presents the effects of boiler/flue gas variables on the design of
absorption-reduction systems. Both flue gas flow rate and NO concentration
X
can be affected by boiler operating conditions. Therefore a change in load
on an industrial boiler may alter these variables markedly. The absorber
2-146
-------
mast be designed to accommodate any anticipated load change. The column
size and the liquid flows must be designed for each application after exam-
ining the boiler operating history and establishing ranges of variation.
TABLE 2.3.8-1. SYSTEM DESIGN CONSIDERATIONS
Variable
Design Effect
Presence of particulates
Presence of
Increased gas flow
Increased NO concentration
Requires prescrubber
Requires SOa :NOx mole ratio of at least
3-59 (depending on process) for absorption-
reduction to be effective.
Requires larger column diameter; increased
liquid flow rate
Requires larger column height; increased
catalyst concentration
The process vendors have not released much information on the operating
conditions of these processes. This is primarily due to the competitive
status of these similar processes at this early stage of development. Typi-
cal values for some of the process variables are shown in Table 2.3.8-2.
TABLE 2.3.8-2. TYPICAL VALUES FOR PROCESS VARIABLES
OF ABSORPTION-REDUCTION PROCESSES
78
Variable
Range
Liquid/Gas ratio, 1/Nm3
SO /NOV mole ratio
X X
Superficial Gas Velocity, m/s
10-30
2.5-3.0
1-3
2-147
-------
Cost estimates for this type of process cover a large range, presumably due
to the differences in sorbent regeneration technique. Capital costs for an
oil-fired system were not found, however, for coal-fired utility applications,
capital costs ranged from $65-127/kW. Operating costs for an oil-fired appli-
cation are shown below for two flue gas flow rates.
Gas Flow Rate, Nm3/hr
150,000
500,000
Operating Costs, mills/kWh
9.1
8.3
These costs are based on ¥200/$ and do not include flue gas reheat.
Presently, absorption-reduction processes are in the pilot-unit stage of
development. Table 2.3.8-3 presents a list of absorption-reduction process
vendors and the status of development of their projects. One can see from
the table that several oil-fired flue gas tests have been performed.
TABLE 2.3.8-3. PROCESS VENDORS OF ABSORPTION-REDUCTION PROCESSES
80
Vendor
Status of Development
Asahi
Chisso
Kureba
Mitsui Engineering and
Shipbuilding
Pittsburgh Environmental
1974: 600 Nm3/hr flue gas from residual oil-
fired boiler (1000 hours continuous).
1975: 300 Nm3/hr flue gas from oil-fired boiler
(335 hours continuous)
1976: 5000 Nm3/hr flue gas from heavy oil-fired
boiler (3000 hours continuous)
1974: 150 Nm3/hr flue gas from oil-fired boiler
1976: 3000 Nm3/hr flue gas from coal-fired
boiler (52 hours continuous, absorption section
only)
2-148
-------
2.3.8.2 System Performance—
Four of the vendors listed in Table 2.3.8-3 report NOX removals of at
least 80 percent with oil-fired boiler flue gas. The Pittsburgh Environmen-
tal system, however, only achieves 60 percent with coal. S02 control for all
of the systems is 90+ percent. No plots of system performance could be found
for any of the oil-fired systems.
Absorption-reduction processes are readily applicable only to high
sulfur oils as a S02:NOX mole ratio in the flue gas of at least 3-5 is
required for maximum performance. This can easily be shown by observing
reactions 2-24 and 2-26 reprinted below.
S02(aq) + 2CH3COONa(aq) + H20 + Na2S03(aq) + 2CH3COOH(aq)
2NO(aq) + 5Na2S03(aq) + 4CH3COOH(aq) + 2NH(S03Na) 2 (aq) + Na2SO.,(aq)
+ 4CH3COONa(aq) + H20
One can see that 1 mole of S02 absorbed in solution reacts to form 1 mole of
sodium sulfite (Na2S03). Then, 5 moles of sodium sulfite are required to
reduce 2 moles of NO. So, the minimum stoichiometric S02:NOX mole ratio
required is y or 2.5. Also, some of the sodium sulfite is oxidized to
sodium sulfate by oxygen present in the flue gas according to:
Na2S03(aq) + hQz (aq) -> Na2SOlt (aq) (2-33)
and is not available for NO reduction. Low-sulfur oils would require S02
X
to be added to the flue gas for these processes to perform; therefore, they
should be considered applicable to high sulfur oils only.
Absorption-reduction processes require large absorbers with high liquid
rates due to relative insolubility of NO, even when the absorption catalyst
is used. Also, the regeneration of the absorption catalyst and the flue gas
reheat for plume buoyancy are energy intensive. Some corrosion-resistant
2-149
-------
material is necessary due to the corrosive nature of the absorbing solution.
However, absorption-reduction appears to be the most promising of the "wet"
NO /SOz removal processes. This is due primarily to its not utilizing oxi-
dants which require much corrosion-resistant material and, more importantly,
create serious secondary pollution problems. Also, the primary by-product
of absorption-reduction processes, gypsum, can be used as landfill material
or in building materials. For all the above reasons, absorption-reduction
processes appear, at this preliminary stage, to be competitive with other
wet NOX/SOX removal processes.
2.3.9 Oxidation-Absorption-Reduction
2.3.9.1 System Description—
Oxidation-absorption-reduction processes simultaneously remove NO and
SOa from flue gas by oxidizing relatively insoluble NO to relatively soluble
NOa and then absorbing both N02 and S02 into a scrubbing solution. The pro-
cesses are based on the use of gas-phase oxidants, either ozone (Os) or
chlorine dioxide (ClOz), to selectively oxidize NO to NOa. Both perforated-
plate and packed bed absorption columns are utilized by various process
vendors.
Most of the oxidation-absorption-reduction processes are similar in
that they consist of five major sections:
prescrubbing
gas-phase oxidation
NOX/S02 absorption
reduction of absorbed NOX and oxidation of SOl
• wastewater treatment
The areas where processes differ are gas-phase oxidation - 03 or Cl02 ;
absorption solutions - limestone slurry (CaCOs), H2SOit, or NaOH; and
the amount and type of waste treatment required. Thermal decomposition,
•2-150
-------
biological denitrification, or wastewater evaporation wastewater treatment
systems can be used. Because of these differences, only one of the oxidation-
absorption-reduction processes, that of Mitsubishi Heavy Industries, is exam-
ined here in detail.
A block flow diagram of the MHI oxidation-absorption-reduction process
is shown in Figure 2.3.9-1.
Gypsum
Figure 2.3.9-1. Process flow diagram for MHI oxidation-
absorption-reduction process.8
2-151
-------
Flue gas is taken from the boiler after the air preheater and passed through
a prescrubber to cool the gas and remove particulates and chlorides. The
flue gas then enters a duct where it is injected with ozone (about 1 percent
by weight in air)82 such that the Os:NO ratio is 1:1. Ozone selectively
oxidizes NO by the following reatcion.83
N0(g) + 03(g) + N02(g) + 02(g) (2-34)
After injection, the flue gas passes countercurrent to a lime/limestone
slurry in a grid-packed absorption column. A water-soluble catalyst is
added to the slurry to enhance N02 absorption (even though N02 is more
soluble than NO, it is still less soluble than S02). S02 is absorbed quickly
at the bottom of the column and undergoes the following reactions.15
S02(g) -»• S02(aq) (2-23)
S02(aq) + CaC03(s) + %K20 ->• CaS03 •J5H20(s) + C02 (g) (2-35)
S02(aq) + CaS03(aq) + H20 + Ca(HS03) 2 (aq) ' (2-36)
N02 is absorbed gradually over the length of the column and reacts as
follows.16
2ND?, (g) + Ca(OH)2(s) + CaS03 -isH20(s) + ^H20 + Ca(N02)2(aq) + CaSOi* 2H20(s)
(2-37)
Once both the N02 and S02 are absorbed, the nitrite ion formed by reaction
2-37 is reduced by the bisulfate ion formed by reaction 2-36. 81*
Ca(N02)2(a.T) + 3Ca(HS03 )2 (aq) + 2Ca[NOH(S03 )2 ] (aq) + 2CaS03 ^H20(s) I + H20
(2-38)
2-152
-------
These hydroxylamine [NOH(S03)2] compounds are reduced further by the sulfite
ion85
Ca[NOH(S03)2](aq) + CaS03«isH20(s) + y H20 -> Ca[NH(S03) 2 ] (aq) + CaSO^ •2H20(s)4-
(2-39)
Upon leaving the top of the absorber, the clean flue gas is reheated for
plume buoyancy and sent to the stack. The slurry solution drops to a holding
tank from which most of the solution is returned to the top of the absorber.
A small stream passes to a neutralization reactor where sulfuric acid is
R fi
added to convert the sulfite solid to soluble bisulfite and solid gypsum.'
2CaS03'%H20(s) + H2SOi»(aq) + H20 + CaS(H'2H20(s) 4- + Ca(HS03 )2 (aq)
(2-40)
This stream passes to a thickener from which the bottoms are sent to a
centrifuge to separate the solid gypsum by-product from the liquor which is
returned to the absorber. The overflow from the thickener is primarily
recycled to the limestone slurry preparation tank. The remainder is sent
to a thermal decomposer where sulfuric acid is added to hydrolyze the N-S
compounds. 8
trf
2Ca[NH(S03)2](aq) + 2H20 -> Ca(NH2S03)2 (aq) + Ca(HSCK)2 (aq) (2-41)
Ca(NH2S03)2(aq) + Ca(HSOO2 (aq) + 6H20 " 2NH.tHSOif (aq) + 2CaS0lt•2H20(s)4'
(2-42)
The ammonium bisulfate solution is pumped to another neutralization reactor
where lime is added.87
(aq) + Ca(OH)2 + H20 + CaSCK -2H20(s) 4- + NIUOH(aq) (2-43)
MHI has three possible methods of removing this ammonium hydroxide:
2-153
-------
decompose by increasing pH
• decompose thermally
• strip out with makeup HaO
The remaining gypsum slurry is pumped to the limestone slurry preparation
tank.
The fundamental design equation used for gas absorption column design
is32
r
dy
(y-y*)
(2-11)
where y = bulk NO concentration (mole fraction) of gas phase at any
X
given point in column
y-y* = overall driving force for absorption (y* being the NO concen-
tration of a gas in equilibrium with a given liquid NOX con-'
centration)
Y, = inlet NOV concentration
b x
Y = outlet NO concentration
a x
Ky = overall gas-phase mass transfer coefficient, Ib-moles NOX/
(ft2)(hr)(mole fraction)
a = area of gas-liquid interface per unit packed volume, ft2/ftc
Gy = molal gas mass velocity, Ib-moles flue gas/(ft2)(hr)
Z = length of packed section of column, ft
In a column containing a given plate or packing configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow
rate/cross-sectional area of column) is called the flooding velocity. At
this point, the gas flow completely impedes the downward motion of the
2-154
-------
liquid and blows the liquid out of the top of the column. The gas velocity
obviously, must be lower than the flooding velocity. How much lower is a
design decision. Often, it is an economic tradeoff between power costs and
equipment costs. A low gas velocity will lower the pressure drop and, hence,
the power costs but the absorber will have a larger diameter and cost more.
High gas velocities have an opposite effect. Usually the optimum gas veloc-
ity is about one-half the flooding velocity.33 The height of the column
depends on the desired level of removal and on the rate of mass transfer.
The latter consideration is why oxidation-absorption-reduction processes
oxidize NO to more soluble NOa prior to the absorber and why some processes
add water soluble catalysts to the scrubbing solution to aid N02 absorption.
The oxidation step enables these processes to use shorter absorbers with
lower liquid rates than either the absorption-oxidation or absorption-reduc-
tion processes. Table 2.3.9-1 presents the effects of boiler/flue gas
variables on the design of oxidation-absorption-reduction systems. Both
flue gas flow rate and NO concentration can be affected by boiler opera-
ting conditions. Therefore a change in load on an industrial boiler may
alter these variables markedly. The absorber must be designed to accommodate
any anticipated load change. The column size and the liquid, oxidant, and
catalyst flows must be designed for each application after examining the
boiler operating history and establishing ringes of variation.
Typical ranges for several operating parameters for this type of process
are shown in Table 2.3.9-2. Reagent concentrations were not available. Eco-
nomics for the various processes cover a wide range presumably due to differ-
ent techniques for oxidant generation and treatment of the scrubbing solution,
Costs are reported to range from $84-134/kW for utility applications' capital
expense and 6.7-9 mills/kWh for operating expense.91
Presently, some of the oxidation-absorption-reduction processes have
reached the prototype stage of development. Table 2.3.9-3 presents a list
of oxidation-absorption-reduction process vendors and the status of develop-
ment of their projects. The applications of this process have been predomi-
nately to oil-fired boilers. Some of the applications treat flue gas flow
2-155
-------
rates similar to those for the standard boilers of this study. Application
to industrial boilers, therefore, is technically feasible.
TABLE 2.3.9-1. SYSTEM DESIGN CONSIDERATIONS
Variable
Design Effect
Presence of particul-ates
Presence of S02
Increased gas flow
Increased NO concentration
Requires prescrubber
Depends on individual process: if NOa is com-
pletely reduced to Nz or NHs by S07 (as does
MHI), then at least the stoichiometric SOz:NO
mole ratio of 3:1 is required [see equation
(9-6)]; if NOz is not- reduced completely, then
a different ratio will be necessary
Requires larger column diameter; increased
liquid flow rate
Requires larger column height; increased gas-
phase oxidant flow rate; increased liquid-
phase catalyst concentration
TABLE 2.3.9-2.
TYPICAL RANGES OF OPERATING VARIABLES FOR
OXIDATION-ABSORPTION-REDUCTION PROCESSES89'9 °
Variable
Range
Liquid/Gas Ratio, 1/Nm3
Oxidant/NO Mole Ratio 0$ systems
CK>2 systems
S02/NOX Mole Ratio
Superficial Gas Velocity, m/s
Pressure Drop,
2-12
0.6-1.0
0.55
2.5-5.0
3-5
200-500
2-156
-------
TABLE 2.3.9-3. PROCESS VENDORS OF OXIDATION ABSORPTION-
REDUCTION PROCESSES
92,93
Vendor
Status of Development
Chiyoda
Ishikawaj ima-Harima Heavy
Industries
Mitsubishi Heavy Industries
Osaka Soda
Shirogane
Sumitomo Metal-Fuj ikasui;
Calcium Process
Sumitomo Metal-Fujikasui:
Sodium Process
1975: 1000 Nm3/hr flue gas from heavy oil-
fired boiler
1975: 5000 Nm3/hr flue gas from oil-fired
boiler (3000 hours continuous)
1975: 2000 Nm3/hr flue gas from heavy oil-
fired boiler (700 hours continuous)
1976: 60,000 Nm3/hr flue gas from oil-fired
boiler
1974: 48,000 Nm3/hr flue gas from oil-fired
boiler
1976: 25,000 Nm3/hr flue gas from sintering
furnace
1973: 62,000 Nm3/hr flue gas from heavy oil-
fired boiler (5 others)
2.3.9.2 System Performance —
Results of oil-fired tests show up to 90 percent M) reduction and >95
percent
reduction. Figures 2.3.9-2 and 2.3.9-3 illustrate NO removals
as a function of pH and additive concentrations.
The primary disadvantage of these processes is the utilization of
costly gas-phase oxidants which create secondary wastewater pollution prob-
lems. Both ozone and chlorine dioxide are highly unstable so they cannot be
stored and must be generated onsite. 03, the more expensive of the two, is
generated by a high energy corona discharge in air. This instantaneous pro-
cess requires significantly large amounts of electricity. C102 is generated
by a slower chemical reaction (requires about 20 minutes to respond to a
change in demand) which could make it less responsive to boiler load changes,
The use of C102 introduces an additional secondary pollutant, chlorides,
2-157
-------
besides the nitrite salt problem. Significant amounts of corrosion-
resistant material are required for oxidation-absorption-reduction pro-
cesses, regardless of which oxidant is utilized. Some of the processes
would not be applicable to low sulfur oils as they require large amounts
of S02 to obtain N02(aq) or N02 reduction.
CONDITIONS
CONCENTRATION OF CaS03: 5 wt%
pH OF LIQUOR : 5.5
N02/(NO + N02) : 0.95
100
0.5
1.0
1.5
NaCl 0.5 mol/£
NaCl 0.17 mol/H
No NaCl, CuCl2 only
2.0 2.5xlO~2
CONCENTRATION OF CuCl2 (moles/&)
Figure 2.3.9-2. Effect of CaCl2 and NaCl concentration
on NO removal efficiency.
22
2-158
-------
CONDITIONS
CONCENTRATION OF CaSOs
CONCENTRATION OF CuCl2
CONCENTRATION OF NaCl
N02/(NO + N02)
5 wt%
0.01 mole/£
1 wt%
0.95
100
80
o
H
O
H
p-l
Pn
w
<
>
i
s
O
60
. 20
\
SOx (CaS03^CuCl2+NaCl)
J- 0
o
SOx(CaS03 only)
NOx (CaS03+CuCl2+NaCl)
NOx (CaSO3 only)
567
pH OF LIQUOR
Figure 2.3.9-3. Effect of pH on SO and NO removal efficiency.
1 23
2-159
-------
2.3.10 Oxidation-Absorption
2.3.10.1 System Description—
As a group, oxidation-absorption processes include those oxidation
processes which do not qualify for the oxidation-absorption-reduction cate-
gory. Basically, there are two types of oxidation-absorption processes.
One is a simplified version of the oxidation-absorption-reduction process
and uses an excess of ozone to selectively oxidize NO to N20s which is
absorbed into aqueous solution and concentrated to form a 60 percent nitric
acid (HNOs) by-product. There is no reduction of NOX(N02) by the absorption
of S02(as SOa) and no wastewater treatment facility.- The other type of
oxidation-absorption process is based on equimolar N0-N02 absorption:
absorbing N20s which is formed by the gas-phase reaction of NO and N02.
A flow diagram of the Kawasaki Heavy Industries oxidation-absorption
process is shown in Figure 2.3.10-1. Flue gas is taken from the boiler
after the air preheater. It passes countercurrent to a magnesium hydroxide
[Mg(OH)2] slurry in the first section of the absorber. There, S02 is absorbed
and undergoes the following reactions.95
S02(g) -> S02(aq) (2-23)
Mg(OH)2(s) + S02(aq) + 5H20 -»• MgS03-6H20(s)i (2-44)
The S02-free flue gas passes to the first denitrification section of the
absorber while the liquid effluent drops to a holding tank. A recycle N02
stream is added to the flue gas to bring the NO:N02 mole ratio to 1:1. . The
resulting mixture then passes countercurrent to a Mg(OH)2 slurry. Equimolar
amounts of 10 and N02 react and are reabsorbed in the following manner.96
N0(g) + N02(g) •* N203(g) (2-45)
N203(g) -> N203(aq) (2-46)
2-160
-------
i
M
ON
f
;• 1 1 ;r
i J
SO2
ABSORBER
SECTION
\ I
MO * NOg
ABSORBER
SECTION
* VJ
N02
ABSORBE
SECTION
' I !l
jn 4 -I
R —
02
CLEAN
FLUE GAS
REACTD
CRYSTALLIZER
REACTOR
CRYSTALUZER Co(OH)2
Figure 2.3.10-1. Flow diagram of Kawasaki Heavy Industries process.'
-------
Mg(OH)2(aq) + N203(aq) -»• Mg(N02)2(aq) + H20 (2-47)
The flue gas passes out of the top of this absorption section while the
liquid effluent drops to the holding tank. Because the rate of reaction
2-45 decreases with NOX concentration (below 200 ppm it becomes negligible),
it is necessary to further reduce NOX by injecting ozone to oxidize the
remaining NO to N02 . The mixture then passes to the final denitrif ication
section of the absorber and is passed countercurrent to a Mg(OH)2 slurry.
This section of the absorber is described by the following reactions.97
2N02(g) + NaOitCg) (2-8)
N20<*(g) -»• N2OUaq) (2-9)
2N20lt(aq) + 2Mg(OH)2(s) + Mg(N03)2(aq) + Mg(N02)2(aq) + 2H20 (2-48)
The clean flue gas leaves the top of this absorber section, is passed to a
reheater for plume buoyancy and sent to the stack. Part of the liquid efflu-
ent from this section is recycled to the tops of the absorber sections while
the rest drops to the holding tank. The slurry solution is pumped to a
thickener which separates the soluble nitrite (NO^) and nitrate (N03) salts
from the solid magnesium sulfite. The overflow from the thickener passes to
a N02 decomposition reactor where sulfuric acid is added.98
3Mg(N02)2(aq) + 2H2SOit(aq) -> 2MgSOi» (aq) + Mg(N03)2(aq) + 4NO(g) t + 2H20
(2-49)
The NO off-gas passes through an oxidizer where it is oxidized by air to N02
and sent to the first denitrif ication section of the absorber. The effluent
from the decomposition reactor is mixed with the thickener bottoms and pumped
to a second oxidizer.99
MgS03'6H20(s) + hz02(g) -»• MgSO^aq) + 6H20 (2-50)
2-162
-------
The magnesium sulfate formed in the oxidizer is treated with calcium nitrate
f\ c
in a gypsum production reactor.
Ca(N03)2(aq) + MgS04 (aq) + 2H20 + CaSO^ '2H20(s) 4- + Mg(N03)2(aq)
(2-51)
The products of this reaction are sent to a centrifuge to remove the solid
gypsum by-product. The liquid from the centrifuge goes to another decomposi-
tion reactor where makeup lime slurry is added.
Mg(N03)2(aq) + Ca(OH)2(s) •> Ca(N03)2(aq) + Mg(OH)2(s) (2-52)
The magnesium hydroxide product is separated in a thickener and recycled to
the absorbers. The thickener overflow stream is split and part is recycled
to the gypsum production reactor and the rest is concentrated to form low-
grade liquid fertilizer by-product, Ca(N03)2.
Since the processes in this category are all very different, especially
with respect to chemistry, generalization of typical ranges of operating
variables is not meaningful and, therefore, not presented. No published
economics for these processes were found.
Presently, the equimolar absorption-type oxidation-absorption processes
are still in the pilot-unit stage of development. Table 2.3.10-1 presents a
list of all oxidation-absorption process vendors and their project's status
of development. These processes have not yet been applied to oil-fired
boilers.
2-163
-------
TABLE 2.3.10-1. PROCESS VENDORS OF OXIDATION-ABSORPTION PROCESSES100
Vendor Status of Development
Kawasaki Heavy Industries 1975: 5000 Nm3/hr flue gas from coal-
fired boiler
Tokyo Electric-MHI (NOX only) 1974: 100,000 Nm3/hr flue gas from natural
gas-fired boiler
Ube Industries No information available
2.3.10.2 System Performance—
No oil-fired tests have been performed. Very little information has
been published on any of the tests conducted.
The production of nitrate salts poses a potential secondary pollution
problem. The plan for reclaiming and concentrating the nitrates as
Ca(N03)2(aq) for liquid fertilizer is questionable as the by-product is of
low quality and may not be easily marketable in the U.S. Also, the gypsum
by-product would be contaminated with various nitrate and sulfite salts, and
therefore, would probably be useful only as landfill material. Much corro-
sion-resistant material is necessary due to the utilization of ozone and
circulating magnesium slurries. The three absorber sections, with their
respective operating conditions, and ozone generation present complex pro-
cess control problems. The process steps of several absorber sections in
series (large fan requirements), ozone generation (corona discharge), flue
gas reheat (inline heater), and by-product and wastewater treatment are all
energy intensive and present technical and economic disadvantages when com-
pared to other simpler FGT processes.
2-164
-------
2.4 CONTROLS FOR NATURAL GAS-FIRED BOILERS
In the previous two sections which discuss controls for coal and oil-
fired boilers, up to ten different process types are presented. Many of
these process types are not considered here for application to natural gas-
fired boilers for two reasons. First, natural gas-fired boilers have no S02
emission problems, and therefore, the simultaneous systems are not considered.
Second, these boilers have no particulate emissions, and as a result, the sys-
tems designed specifically for high particulate applications are not consid-
ered. This leaves two systems to be considered for application to natural
gas-fired boilers and these are discussed in the following sections.
2.4.1 Selective Catalytic Reduction-Fixed Packed Bed Reactor
2.4.1.1 System Description—
Fixed packed bed systems are applicable only to flue gas streams con-
taining less than 20 mg/Nm of particulates. As such, they are applicable
to natural gas-fired boilers.
The primary feature of these systems is the reactor which contains the
catalyst. As the name implies, the granular catalyst is randomly packed in
a stationary bed. An example of a typical fixed bed reactor is shown in
Figure 2.4.1-1. The important features of the reactor are:
the catalyst
the catalyst support
the gas distributor
The catalyst can be either spherical or cylindrical in shape. Spherical
granules typically range in size from 4-10 mm in diameter. The composi-
tion varies from process to process and most formulations are proprietary.
The catalyst is supported either by inert packing (as shown in Figure 2.4.1-1)
or by a perforated support plate (Figure 2.4.1-2). The catalyst supports
2-165
-------
6* loytr l" bolls-
{'optionol additional layers-
of progressively smaller bolls
for improved distribution and
tcalt removal
Catalyst Bed
(1/8" x 1/8'pellets)
3" layer 1/4* balls ~f
4" layer 1/2* bolls
5" layer 3/4' bolls
3/4" balls
Reactor Outlet Screen
•ith Continuous Slotted
-Openings
Catalyst Bed
M/4"x 1/4" \
V pellets /
HO" layer 3/8" bolls
4" layer 1/2" bolls
5" layer 3/4" bolls
3/4" balls
Catalyst Dump Flange
Figure 2.4.1-1. Example of typical fixed packed bed reactor,
i o i
Tir'
SUPPORT SEAMS REQUIRED ONLY FOR
LARGER VESSELS OR HIGH LOADINGS
Figure 2.4.1-2. Example of catalyst support plate.1
02
2-166
-------
hold the catalyst fixed in place in order to prevent both mobilization of
the particles by the gas stream and catalyst rearrangement which would allow
channelling of the flue gas. The gas distributor can be a perforated plate
or similar device which spreads the gas flow across the entire cross-section
of the catalyst bed.
A typical fixed bed SCR process layout is presented in Figure 2.4.1-3.
Several arrangements are possible, however, for application to new boilers
this arrangement is the most desirable.
Boiler
Flue Gas
NH3
Reactor
Air
Heater
Stack
Air
Figure 2.4.1-3. Process layout for fixed bed SCR process.
The principle of operation of these systems involves a gas phase
reaction between ammonia (NH3) and NO (NO and N02). These reactions are
X
presented most accurately by
1 2
4NH3 + 4NO + 02 * 4N2 + 6H20
(2-53)
4NH3 + 2N02 + 02 2 3N2 + 6H20
(2-54)
The first reaction predominates since flue gas NO is typically 90-95 percent
NO. As shown, the NO is reduced to molecular nitrogen (N2) which exits with
the flue gas stream.
2-167
-------
The primary design equation used with these processes is the standard
equation for reactor design, 3 represented by
x
dx
V f d:
F-] 7
•f (->
(2-3)
where V is the catalyst volume
F is the mass (or molar) flow rate
x is the conversion of NOX to Na
r is the reaction rate mass (or moles)
volume of catalyst x time
The reaction rate, r, for each NO reduction reaction can be represented by
r = k[NH3]a[NO]b[02]C (2-4)
where k is the reaction rate constant
[NHa], [NO], [02] are the reactant concentrations
a, b, c, are empirically determined exponents
The reaction rate is different for each catalyst formulation and, therefore,
values for k, a, b, and c must be determined for the particular catalyst to
be used before any design can be performed. The reaction rate constant is
usually described by the Arrhenius equation
_ E_
RT
k = Ae (2-5)
where A is the frequency factor
E is the activation energy
R is the universal gas constant, and
T is the temperature
2-168
-------
Values for k, a, b and c for two catalyst formulations are shown in Table
2.4.1-1. Values for other catalyst formulations will be different.
TABLE 2.4.1-1. REACTION RATE DATA FOR TWO
CATALYST FORMULATIONS
11
Catalyst: V205 on A1203
k = 2.05 x 103e
a = 0.30
b = 0.22
c = 0.05
9650
RT
Catalyst: Fe-Cr on Al20s
k = 3.25 x 103e
a = 0.45
b = 0.10
c = 0.15
10.860
RT
The most important design and operating variables are similar to those for
moving bed systems using granular catalysts. These are listed, along with
typical ranges, in Table 2.4.1-2.
TABLE 2.4.1-2.
DESIGN AND OPERATING VARIABLES FOR
FIXED PACKED BED SYSTEMS1*
Typical Range
Variable
Gas Velocity, m/s
Bed Depth, m
Space Velocity, hr l
Pressure Drop, mmHaO
Temperature, °C
(For Oil)
0.5 - 1.5
0.2 - 0.6
6,000 - 10,000
40 - 80
350 - 400
(For Gas)
0.5 - 1.5
0.2 - 0.4
8,000 - 15,000
40 - 70
300 - 400
2-169
-------
Other variables that affect the process are:
flue gas flow rate
NOX control level
NO concentration
boiler load variation
The flue gas flow rate and NOX control level determine the catalyst volume
required (hence reactor size). Increases in either parameter also increase
the reactor size. The NOX concentration is primarily a function of fuel type
used in the standard boilers. Higher concentrations require larger NHa
storage and vaporization equipment; reactor size is not affected. Boiler
load can affect several things including flue gas temperature, flow rate and
NOx concentration. It is usually necessary to maintain reaction temperatures
of 350 to 400°C. Temperature control equipment may be necessary to accomo-
date large boiler load variations which cause lower flue gas temperatures.
Where these variations are present, some equipment overdesign may be war-
ranted to insure a constant control level. These variables are discussed in
more detail in the section on moving bed SCR systems for coal-fired boilers,
Section 2.2.2. Costs of fixed packed bed systems range from $16-49/kW
(capital) and 1.2-1.8 mills/kWh (operating). These costs are based on util-
ity applications as well as a variety of process and operating conditions.
There are vendors of fixed packed bed SCR systems and all are Japanese.
Vendors are listed in Table 2.4.1-3 and the scale of development is also
noted. Fixed packed systems have been applied to industrial and utility
boilers in Japan. Existing installations are shown in Tables 2.4.1-4 and
2.4.1-5. Currently, there are no installations in the U.S.
2.4.1.2 System Performance—
Typical performance data for fixed packed bed SCR systems are shown in
Figures 2.4.1-4 through 2.4.1-8. These data indicate that NO removals 90
X
percent and higher are achievable with these systems. This allows them to
be considered for all control levels of interest in this study.
2-170
-------
TABLE 2.4.1-3.
VENDORS OF SCR FIXED BED SYSTEMS
FOR GAS-FIRED APPLICATIONS21
Vendor
Notes
Sumitomo
Hitachi Zosen
Hitachi, Ltd.
Mitsubishi Heavy Industries
Tested on commercial scale equipment
Tested on commercial scale equipment
Tested on commercial scale equipment
Tested on commercial scale equipment
Ishikawajima-Harima Heavy Industries Tested on commercial scale equipment
Mitsui Toatsu Chemical Has not been to boilers
Kawasaki Heavy Industries Tested on pilot scale equipment
Mitsubishi Kakoki Kaisha Tested on commercial scale equipment
TABLE 2.4.1-4. EXISTING FGT INSTALLATIONS OF SCR FIXED BED SYSTEMS
GAS-FIRED INDUSTRIAL BOILERS
21
Location
Takaishi
Process
User Developer
Osaka Gas Mitsubishi H.I.
Capacity
Fuel (Nm3/hr)
LNG 30,000
Completion
Date
December 1976
TABLE 2.4.1-5. EXISTING FGT INSTALLATIONS OF SCR FIXED BED SYSTEMS
GAS-FIRED UTILITY BOILERS
21
Location
User
Process
Developer
Capacity Completion
Fuel (Nm /hr)
Date
Kokura Kyushu Mitsubishi H.I. LNG 3,380,000* October 1978
Electric
Chita Chubu Hitachi, Ltd.
Electric
LNG 4,000,000* April 1978
*Flow rate is combined value from two boilers.
2-171
-------
•=C
>-
o
LNG BOILER
1000
Low-s 011 Boiler
H1gh-s 011 Boiler
2000
3000
4000
5000
OPERATION PERIOD (Hours)
o-o-o-o
6000
Circled figures show times when SV and MH3/NO mole ratio were changed.
1. SV 10,000 20,000 hr'1 2. SV 10,000 15,000 hr'1
3. SV 15,000 20,000 hr"1 4. SV 6,200 4,500 hr'1
5. SV 4,500 6,200 and the mole ratio 0.95 0.83
7000
Figure 2.4.1-4. Test results at gas- and oil-fired boilers.
124
-------
I
M
U)
IOO
x
O
90
O
z
UJ
5 80
u.
UJ
_J
-------
0)
o:
100
90
80
70
60
50
SV = 5,000
180 200 220 240 260 280
Temperature (°C)
Figure 2.4.1-6. Performance of catalyst MTC-102
(flue gas by LPG burning).
126
300
100 \-
5 80
o
60
(at 240°C)
_L
5,000
10,000
SY (hr'1)
Figure 2.4.1-7. SV and NO removal (MTC-102)
(flue gas by LPG burning).
127
_L
15,000
2-174
-------
ho
I
—J
Ln
100
* 80
>
o
z
UJ
2 60
UJ
§
o
2
ui
o:
X
O
z
40
20
300°C,
C-l CATALYST
SV=ZO,000 hr'1
'350°C
I
0.5 1.0 1.5
MOL NH3PER MOL NOX INLET
120
a.
lOOQ-
80 S
CD
60 t
x
40 z
20^
z
Figure 2.4.1-^
Relationship among inlet NH3:NOX mol ratio, NOX removal efficiency, and exiting
NHs concentration using the Sumitomo Chemical C-l Catalyst.128
-------
2.4.2 Absorption-Oxidation
2.4.2.1 System Description —
Absorption-oxidation processes remove NOX from flue gas by absorbing
the NO or NO into a solution containing an oxidant which converts the NO
X X
to a nitrate salt. Two types of gas/liquid contactors can be used and exam-
ples of each type are shown in Figure 2.4.2-1. Both perforated plate and
packed towers accomplish N0y absorption by generating high gas/liquid inter-
facial areas. The choice of one type of contactor is a design decision made
to achieve a given removal for the least cost .
A generalized process flow diagram is shown in Figure 2.4.2-2. Flue
gas is taken from the boiler after the air preheater. Before the gas can
be sent to the NO absorber, it must be S02-free since SOz consumes prohibi-
tive amounts of the costly liquid-phase oxidant. This is not a problem with
natural gas f ired-boilers since they have no S02 emissions. In most cases,
the oxidant is permanganate (MnOi*). The flue gas enters the distributing
space at the bottom of the NOX absorber, below the packing or plates. The
gas passes upward through the column, countercurrent to the flow of the
liquid absorbent/oxidant (usually a KOH solution containing KMnOi*). NO
X
is absorbed and then oxidized over the length of the column according to
the following reactions.31
N0(g) + NO(aq) (2-6)
NO(aq) + KMnO!t(aq) -> KN03 (aq) + Mn02 (sH (2-7)
2N02(g) + NaO^g) (2-8)
NaO^g) -> NaCMaq) (2-9)
+ 2K2Mn0lt(aq) -> ZKMnO^aq) + 2KN02 (aq) (2-10)
2-176
-------
FLUE GAS OUT
Principal -
interface
LIQUID OUT1'
LIQUID IN
'— Coalesced
dispersed
-Perforated
plate
— Downspout
FLUE GAS i\
FLUE GAS OUT
LIQUID IN
Perforated Plate Absorber
Packed Absorber
Figure 2.4.2-1. Gas/liquid contactor options for
Absorption-Oxidation Processes.
2-177
-------
Flue
Gas
Absorber
To Reheat
and Stack
Holding
Tank
Oxidant
Make-up
Nitrate Treatment and
Oxidant Regeneration
Figure 2.4.2-2. Process flow diagram for absorption-
oxidation process. °
Since most of the NO from combustion processes occurs as NO,
reactions 2-6 and 2-7 predominate. The clean gas passes out of the top
of the absorber to a heater for plume buoyancy and is sent to the stack.
The absorbing solution drops to a holding tank where makeup KOH and/or
KMnOii are added. This solution flows to a centrifuge to separate the
solid MnOa which is then electrolytically oxidized to MnOit. The remaining
solution is either concentrated in an evaporator to form a weak KNOs solu-
tion or is electrochemically treated to produce a weak HNOs solution and a
mixed stream of KOH and
2-178
-------
The fundamental design equation used for gas absorption column design is
(2-11)
where y = bulk NO concentration (mole fraction) of gas phase at any
X
given point in column
y-y* = overall driving force for absorption (y* being the NO con-
X
centration of a gas in equilibrium with a given liquid NO
concentration)
Y, = inlet NO concentration
b x
Y = outlet NO concentration
a *
K = overall gas-phase mass transfer coefficient, Ib-moles NOX/
(ft2)(hr)(mole fraction)
a = area of gas-liquid interface per unit packed volume, ft2/ft3
G = molal gas mass velocity, Ib-moles flue gas/(ft )(hr)
Z = length of packed section of column, ft
In a column containing a given packing or plate configuration and being
irrigated with a certain liquid flow, there is an upper limit to the gas
flow rate. This limit's superficial gas velocity (volumetric gas flow rate/
cross-sectional area of column) is called the flooding velocity. At this
point, the gas flow completely impedes the downward motion of the liquid
and blows the liquid out of the top of the column. The gas velocity, obvi-
ously, must be lower than the flooding velocity. How much lower is a design
decision. Often, it is an economic tradeoff between power costs and equip-
ment costs. A low gas velocity will lower the pressure drop and, hence, the
power costs but the absorber will have a larger diameter and cost more. High
gas velocities have an opposite effect. Usually the optimum gas velocity is
about one-half the flooding velocity.31* The height of the column depends on
2-179
-------
the desired level of removal and on the rate of mass transfer. The latter
is a major problem for these systems trying to achieve large NOX reductions
since NO is relatively insoluble in water. This can be seen in Table 2.4.2-1.
TABLE 2.4.2-1. NITROGEN OXIDES CHARACTERISTICS
35
Boiling Point,
°C
Solubility in Cold
Water (0°C), cm3
Solubility in Hot
Water.. (60°C), cm3
NO
NO 2
-151.8
21.2
7.34/100 cc H20
soluble, decomposes
2.37/100 cc H20
One can see that NO has a very limited solubility in water and, since most
N0y is present as NO, the rate of mass transfer (absorption) is going to be
relatively slow. This means that the absorber must be tall with a high
liquid flow rate. Table 2.4.2-2 presents the effects of boiler/flue gas
variables on the design of absorption-oxidation systems.
TABLE 2.4.2-2. SYSTEM DESIGN CONSIDERATIONS
Variable
Design Effect
Presence of participates
Presence of SOa
Increased gas flow
Increased NO,, concentration
Requires prescrubber
Requires FGD pretreatment
Requires larger column diameter; increased
liquid flow rate
Requires larger column height; increased
oxidant concentration
Both flue gas flow rate and NOX concentration can be affected by boiler
operating conditions. Therefore a change in load on an industrial boiler
may alter these variables markedly. The absorber must be designed to accom-
modate any anticipated load changes. The column size and the liquid and
2-180
-------
oxidant flows must be designed for each application after examining the
boiler operating history and establishing ranges of variation.
None of the sources consulted for this study could supply typical ranges
for operating variables such as liquid/gas ratio, reagent concentrations or
pressure drops and, as a result, none are presented here. Economic data were
not presented either. One source did estimate the removal for absorption-
oxidation processes to be 85 percent.36
Presently, absorption-oxidation processes are still in the pilot unit
stage of development. Table 2.4.2-3 presents a list of absorption-
oxidation process vendors and the status of development of their projects.
One can see from the table that no gas-fired flue gas tests have been
performed.
TABLE 2.4.2-3. PROCESS VENDORS OF ABSORPTION-OXIDATION PROCESSES37'38
Vendor Status of Development
Hodogaya No information available; stopped development
on process
Kobe Steel 1974: 1000 Nm3/hr gas from iron-ore sintering
furnace; stopped development on process
MON (Mitsubishi Metal, MKK, 1974: 4000 Nm3/hr flue gas from oil-fired
Nikon Chemical) boiler
Nissan Engineering 1972: 4 pilot plants, 100-2000 Nm3/hr tail
gas from HNOs plant
2.4.2.2 System Performance—
No gas-fired tests have been made. No information has been published
on tests conducted with other fuels. The relative insolubility of NO in
water may present a nu.jor obstacle to achieving the stringent level of con-
trol (90 percent NO reduction) by absorption-oxidation processes. Another
primary drawback of absorption-oxidation systems is the production of nitrate
2-131
-------
salts (see Equation 2-7), a secondary pollutant. These processes probably
could not be applied on a large scale as wastewater treatment systems
(chemical or biological) do not remove nitrogen compounds from the waste-
water.39 Trying to recover the nitrates as nitric acid for industrial use
or potassium nitrate for fertilizer does not seem promising as the by-products
are of low quality. Also, the use of an expensive, liquid-phase oxidant
requires stainless steel and other corrosion resistant materials of construc-
tion. The process steps of oxidant regeneration (electrolysis) and flue gas
reheat (inline heater) are all energy intensive and present technical and
economic disadvantages.
2-182
-------
REFERENCES
1. Babcock & Wilcox. Steam, It's Generation and Use. 1978. pp. 25-1 -
25-10.
2. Blue, George. Hartford Steam Boiler & Insurance Company. Private
Conversation with Gary Jones. September 15, 1978.
3. United States Environmental Protection Agency. Task 2 Summary Report.
"Preliminary Summary of the Industrial Boiler Population." PEDCo
Environmental, Inc. Cincinnati, Ohio. June 29, 1978.
4. Bartok, W. Systems Study -of Nitrogen Oxide Control Methods for Station-
ary Sources. November 20, 1969. p. 4-10.
5. Ando, Jumpei. NOX Abatement from Stationary Sources in Japan. EPA
Report in Preparation. October 1978. p. 2-23.
6. Ando, Jumpei. NOX Abatement from Stationary Sources in Japan. EPA-600/
7-77-l03b. September 1977. p. 78.
7- Ibid., p. 59.
8. Ando, Jumpei. October 1978^ op cit.3 p. 3-36.
9. Faucett, H.L., et at. Technical Assessment of NOX Removal Processes
for Utility Application. EPA-600/7-77-127. November 1977. p. 243.
10. Ando, Jumpei. Octo'ber 1978, op oit.3 3-27.
11. Ando, Jumpei. September 1977, op cit.,, p. 77.
12. Matsuda, S., et at. Selective Reduction of Nitrogen Oxides in Com-
bustion Flue Gases. Journal of the Air Pollution Control Association.
April 1978. p. 350-353.
13. Smith, J.M. Chemical Engineering Kinetics. 2nd Edition. McGraw-Hill.
1970. p. 112.
14. Ando, Jumpei. October 1978, op cit.3 p. 3-31.
15. Faucett, H.L., op oit.3 General Reference, All Described Processes Were
Surveyed.
2-183
-------
16. Levenspiel, 0. Chemical Reaction Engineering. 2nd Edition. John
Wiley & Sons. 1972. p. 100.
17- Faucett, H. L., op cit., p. 249.
18. Ando, Jumpei. October 1978, op cit.3 p. 3-7.
19. Ibid. , p. 3-70 - 3-79.
20. Ibid., p. 3-4.
21. Ibid., p. 3-4, 3-5.
22. Ibid., p. 3-28.
23. Ibid., p. 3-34.
24. Faucett, H. L., op cit., p. 217.
25. Ando, Jumpei. October 1978, op cit. , p. 3-31.
26. Ibid. , p. 1-35.
27. "NOX Control Review." Vol. 3, No. 4. Fall 1978. p. 3.
28, Ando, Jumpei. October 1978, op cit., pp. 4-1 - 4-133.
29. Treybal, R. E. Mass-Transfer Operations. Second Edition. McGraw-Hill,
1968. pp. 419, 425,
30. Faucett, H. L., op cit.., p. 108.
31. Ibid., p. 132.
32. Ibid., p. 7-
33. McCabe, W. L. and Smith, J. C. Unit Operations of Chemical Engineering.
Second Edition. McGraw-Hill. 1967. p. 664.
34. Ibid., p. 645.
35. Weast, R. C. Handbook of Chemistry and Physics. 52nd Edition. The
Chemical Rubber Company. 1971. p. B-115.
36. Faucett, H. L. , op ait.., p. 20.
37. Faucett, H. L. , op cit.., pp. XV, 80, 109, 134.
38. Ando, Jumpei. October 1978, op ait.., p. 7-51.
2-184
-------
39. Ibid. , p. 7-4.
40. Arneson, A. D., et al. "The Shell FGD Process-Pilot Plant Experience
at Tampa Electric." Paper Presented at Fourth Symposium on Flue Gas
Desulfurization. Hollywood, Florida. November 8-11, 1977. p. 3.
41. Faucett, H. L., op oit. 3 p. 351.
42. Arneson, A. D. , op cit. , pp. 15-16.
43. Faucett, H. L. , op cit.3 p. 361.
44. Nooy, F. M. and Pohlenz, J. B. "S02 Stack Gas Scrubbing Technology."
p. 354.
45. Ibid., p. 355.
46. Ibid., p. 354.
47. Ibid., p. 453.
48. Faucett, H. L. , op oit.3 p. 352.
49. Arneson, A. D. , op oit.3 p. 16.
50. Ibid., p. 16.
51. Ibid., p. 20.
52. Ibid., p. 9.
53. Ibid., p. 11.
54. Faucett, H. L., op cit., 201.
55. Ibid., p. 200.
56. Ibid., p. 202.
57. Ibid., p. 159.
58. Ibid., p. 204.
59. Ibid., p. 203.
60. Ando, Jumpei. October 1978, op ait.3 p. 6-32.
61. Ibid. , p. 6-31.
62. Faucett, H. L., op cit., p. 163.
2^-185
-------
63. Ibid. , p. 168.
64. Ando, Jumpei. October 1978, op oit.3 p. 6-33.
65. Faucett, H. L., op oit., p. 168.
66. Perry, R. H. Chemical Engineers Handbook. 5th Edition. McGraw-Hill
1973. p. 18-20.
67. McCabe, W. L., op cit., p. 588.
68. Perry, R. H., op oit., p. 18-45.
69. Ando, Jumpei. October 1978, op ait.., p. 7-39.
70. Faucett, H. L. , op cit., p. 85.
71. Ibid. , p. 84.
72. Ando, Jumpei. October 1978, op oit., p 7-7.
73. Faucett, H. L., op oit. , p. 86.
74. Ibid., p. 87.
75. Ando, Jumpei. October 1978, op eit. , p. 7-38.
76. Ibid., p. 7-40.
77. Ibid., p. 7-39.
78. Faucett, H. L., op oit., pp. 34-35, 42-44, 89-90, 141-42.
79. Ibid., p. 27.
80. Ibid., pp. 32, 40, 87, 104, 139.
81. Ibid., p. 94.
82. Ibid.t p. 389.
83. Ibid., p. 93.
84. Ibid., p. 95.
85. Ando, Jumpei. October 1978, op oit., p. 7-12.
86. Faucett. H. L., op oit., p. 96.
87. Ando, Jumpei. October 1978, op oit.., p. 7-14.
2-186
-------
88. Faucett, H. L., op ait. , pp. 63, 99.
89. Ibid., pp. 42-44, 63-66, 98-100, 117-121, 125-130.
90. Audo, Jumpei. October 1978, op sit., p. 7-25.
91. Faucett, H. L. , op oit. , p. 25.
92. Ibid., pp. 52, 61, 97, 116, 126.
93. Ando, Jumpei. October 1978, op oit., p. 110.
94. Faucett, H. L., op cit. , p. 70.
95. Ibid., p. 69.
96. Ibid., p. 71.
97. Ibid., p. 71.
98. Ibid., p. 72.
99. Ibid., p. 72.
100. Ibid., pp. 73, 146, 149.
101. Rase, Howard F. Chemical Reactor Design for Process Plants. Volume 1
Wiley-Interscience. 1977. p. 515.
102. Ibid., p. 514.
103. Ando, Jumpei. October 1978, op cit., p. 3-30.
104. Ibid., p. 4-10.
105. Ibid., p. 3-7.
106. Ibid., p. 4-21.
107. Ibid., p. 4-71.
108. Ibid., p. 4-5.
109. Ibid., p. 4-37.
110. Ibid., p. 4-94.
111. Ibid. , p. 4-93.
112. Ibid., p. 4-92.
2-187
-------
113. Ibid., p. 4-126.
114. Ibid., p. 4-21.
115. Ibid., P. 4-96.
116. Ibid., p. 3-45.
117. Ibid., p. 4-41.
118. Faucett, H. L., op cit. , p. 224
119. Ando, Jumpei. October 1978, op cit., p. 4-95.
120. Ibid., p. 6-29.
121. Ibid., p. 7-31.
122. Ibid., p. 7-19.
123. Ibid., p. 7-20.
124. Ibid., p. 4-43.
125. Faucett, H. L., op cit., p. 214.
126. Ando, Jumpei. October 1978, op cit., p. 4-121.
127. Ibid., p. 121.
128. Faucett, H. L., op cit., p. 298.
129. Noblett, J.G., et al• "Impact of NOX Selective Catalytic Deduction
Processes on Flue Gas Desulfurization Processes," Draft Final Report
EPA Contract No. 68-02-2608, Task 70, Radian Corporation, September
1979.
2-188
-------
SECTION 3
CANDIDATES FOR BEST SYSTEMS OF EMISSION REDUCTION
The ten systems discussed in Section 2 are not applicable to all combi-
nations of boiler types and fuels of interest in this study. However, several
of these systems may be applicable to a specific boiler/fuel combina-
tion (i.e., capable of removing sufficient NO to meet proposed emission
regulations). In this section, NOX control techniques which are applicable
to the various boilers and fuels considered in this study are selected. The
section is organized to compare N0x-only and simultaneous NOX/SOX reduction
systems separately. The result is a set of candidate control techniques
that will be evaluated in detail in subsequent sections to determine the
"best" system for NOX control by FGT.
3.1 CRITERIA FOR SELECTION
Two sets of evaluation criteria are used to determine the set of candi-
date systems. One is the level of NO control desired which determines the
set of systems available for further evaluation. The other is a set of
evaluation criteria that will allow comparison of the systems capable of
meeting a particular level of control.
3.1.1 Factors Considered in Selection of Best Systems
A consistent set of rating criteria was used to evaluate and compare
each of the FGT systems described in section 2 that are capable of achieving
the proposed NO removal levels. These criteria and the weighting factors
X
are shown in Table 3.1.1-1. As can be seen, the criteria receiving most
emphasis are status of development, economics, performance, and reliability.
3-1
-------
TABLE 3.1.1-1. RATING CRITERIA AND WEIGHTING FACTORS
Evaluation Category Total Points
Performance 14
Operational/Maintenance Impacts on Performance 7
Preliminary Environmental Impacts 9
Preliminary Economic Impacts 15
Preliminary Energy/Material Impacts 10
Boiler Operation and Safety 4
Reliability 14
Status of Development 16
Adaptability to Existing Sources 6
Compatability with Other Control Systems 5
100
Emphasis is placed on the most developed FGT systems since they repre-
sent the most likely controls to be applied if a high degree of NO control
/ X
is required on industrial boilers. An FGT system must achieve the necessary
NO reduction and do so as economically as possible, hence the heavy emphasis
on performance and economics. These are important considerations for any
application. Reliability is heavily weighted because it is common for an
industrial boiler to supply one or several continuous manufacturing processes.
A high reliability is required to avoid frequent boiler shutdowns with sub-
sequent loss of revenues due to dependency of the manufacturing process on
the boiler.
It should be pointed out that only large differences in point values are
significant while small differences are not. For example, ratings which dif-
fer by a factor of two are significant. However, two ratings 10 points apart
do not necessarily indicate the superiority of one process. A more detailed
breakdown of the evaluation criteria and the point values assigned is present-
ed in Table 3.1.1-2. The basis for the detailed breakdown is discussed below.
The analysis of each system using these criteria is discussed in Section 3.2.
3-2
-------
TABLE 3.1.1-2. SPECIFIC POINT VALUES ASSOCIATED WITH SELECTION FACTORS
Item
1. Performance
a. Desired control level (stringent,
intermediate, or moderate) as percent
of system's maximum design capability
/ Desired Control Level \
\Maximum Design Control Level/ X 10°
b. Particulate handling capability
c. Load following ability
2. Operation and Maintenance impacts on
Performance
a. Moving parts
b. Solids handling
c. Process separability
d. Flue gas composition sensitivity
e. Prescrubbing necessary
f. Process stability
3. Preliminary Environmental Impacts
a. Secondary pollutants - Air
- Liquid
Quality
<70
70 - 80
80 - 90
90 -100
>100
Great
Some
None
Good
Fair
Poor
Few
Many
No
Yes
Once-through
Regenerable
No
Yes
No
Yes
Simple process &
insensitive control needs
Complex process or
sensitive control needs
Complex process &
sensitive control needs
None
Potential
Some
Major
None
Some
Major
Points
8
6
4
2
No Go
4
1
0
2
1
0
1
0
1
0
1
0
1
0
1
0
2
1
0
3
2
1
0
3
1
0
3-3
-------
TABLE 3.1.1-2. (Continued)
3.
4.
5.
6.
7.
Item
a. Secondary pollutants (Cont'd)
- Solid
Preliminary Economic Impacts
a. Capital investment
b. Operating costs
c. Marketable by-product
Preliminary Energy/Material Impacts
a. Electrical demand
b. Auxiliary fuel use
c. Energy intensive regeneration or
by-product treatment
d. Raw material demand
Boiler Operation and/or Safety
Boiler impacts or safety hazards
Reliability
a. Plugging and scaling
Quality
None
Some
Major
<50% mean
50% mean
75% mean
Mean
125% mean
150% mean
>150% mean
Potential
None
<1% output
1-2%
2-3%
3-4%
4-5%
>5%
No
Yes
None
Some
Heavy
Light
Moderate
Heavy
None
Potential
Yes
None
Some
Much
Points
3
1
0
7
6
5
4
3
2
1
1
0
5
4
3
2
1
0
1
0
2
1
0
2
1
0
4
2
0
5
2
0
3-4
-------
TABLE 3.1.1-2. (Continued)
Item
7. b. Simplicity - Number process steps
c. Material of construction
8. Development Status
a. Scale demonstrated
b. Length of operation
c. Uncertainties in technology
9. Adaptability to Existing Sources
a. Retrofit
b. Land required
10. Compatability with Other Control Systems
a. FGD
b. ESP, other
Quality
<3
3
4
5
6
7
>7
Carbon steel
Some corrosion resistant
material
Much corrosion resistant
material
Commercial
Prototype
Pilot
Bench
Conceptual
>5000 hours
3000 - 5000
1000 - 3000
<1000
No
Yes
Easy
Difficult
Small
Large
Yes
No
Yes
No
Points
6
5
4
3
2
1
0
3
1
0
10
8
5
2
0
3
2
1
0
3
0
3
0
3
0
3
0
2
0
3-5
-------
3.1.1.1 Performance—
A primary concern in the selection of an N0x flue gas treatment system
is the system's performance. The first aspect to consider here is the N0x
removal capability. This study is organized by different levels of N0x
control (stringent, intermediate, moderate). The processes' maximum removal
capability is compared to these various control levels to show the ease with
which the system can meet the removal requirement. Another measurement of
a system's performance is its load following capability—how well the system
responds to a sudden change in boiler load. Generally, large, complex
systems do not respond to load changes as quickly as small, simple systems.
Slow response is a disadvantage since it may result in increased emissions
during load changes.
3.1.1.2 Operational and Maintenance Impacts—
This category is important for several reasons. A system with diffi-
cult operational steps or high maintenance requirements is not as desirable
since it will require more manpower and increase operating costs. Reliabil-
ity may also be adversely affected. For most FGT systems, this type of data
is not available. In this study these impacts are inferred by examining
each system and applying engineering judgment. The more mechanically complex
a system is, the more likely it is to have operation and maintenance problems.
3.1.1.3 Preliminary Environmental Impacts—
This category, along with the economic and energy categories, relies on
published information for data. Detailed analyses of the candidate systems
in these areas will be conducted in a subsequent section. The data presented
in this section are used for comparison purposes only. Obviously it is
undesirable for an FGT system to remove NO at the expense of emitting a
X
secondary pollutant. For this reason secondary pollutants (air, liquid, and
solid) emitted by the process, or potentially so, are identified. Systems
with no secondary pollutants receive the highest ratings.
3-6
-------
3.1.1.4 Preliminary Economic Impacts—
With an industrial boiler it is probable that application of FGT will
affect the price of products from a new or modified facility and thereby
affect the salability of these products. For this reason, the lowest cost
system that will adequately control NOX is desirable. The areas considered
are capital investment ($/kW) , operating costs (mills/kWh), and credits for
marketable by-products. Cost data that are available are primarily for
utility installations. While there is some economy of scale in the invest-
ment cost due to the large size of the facilities, the values are adequate
for preliminary cost comparisons. Sample economy of scale calculations show-
ing how the preliminary economic figures were generated are contained in
Appendix II.
3.1.1.5 Preliminary Energy/Material Impacts—
It is desired to minimize energy and raw material consumption by an FGT
process since this also minimizes operating costs. In addition, dependence
on outside factors such as raw material supplies is reduced. The main sys-
tem parameters considered are the electrical demand of the system, use of
auxiliary fuels and energy, and intensive regeneration or by-product treat-
ment processes. Also, heavy raw material demands are noted. Again, utility
data are used for comparative purposes since very little industrial boiler
data are available.
3.1.1.6 Boiler Operation and/or Safety—
It is desirable to minimize impacts of the FGT system on the boiler.
The main areas of potential impacts are air heater fouling, duct scaling and
stack corrosion. These impacts as well as safety aspects of the process are
determined by inspection of the process equipment and chemistry.
3.1.1.7 Reliability--
Reliability data are not generally available for all of the process
types considered. Many have not been applied on commercial scale equipment.
Some reliability data are available for SCR systems, but data from other
3-7
-------
systems are necessary before the reliability of SCR systems can be compared
on a relative basis. For most systems it can be said that simplicity is
concommitant with reliability and this concept is used in the evaluation.
3.1.1.8 Development Status—
A crucial consideration in the selection of the best NOX control tech-
niques by flue gas treating is the status of development of the processes.
Presently, there are but a few commercial-size NOX FGT units in operation on
industrial boilers—all in Japan. Because most of the flue gas treatment
development work has been conducted fairly recently, it is vital that those
systems which have been demonstrated most fully be given primary considera-
tion for implementation to industrial boilers. For this study, availability
by the year 1981 was estimated using the current status of development and
reported on-going development. The size of the unit, length of operation,
and any uncertainties in technology were all taken into account.
3.1.1.9 Adaptability to Existing Sources—
Since applying FGT to modified existing sources is generally more
difficult than with new sources, the ease of retrofit was examined. Struc-
tural and equipment modifications necessary for retrofit are considered since
existing boilers are not constructed to accommodate FGT systems. Land
requirements of the FGT system are also considered, since existing industrial
boilers are not necessarily located near large land areas. Quite frequently,
they are located in the center of a plant and surrounded by equipment. Small
systems requiring little boiler modification are desired.
3.1.1.10 Compatibility with Other Control Systems—
This category is related to retrofit and new installation. Where addi-
tional cont ol equipment is existing or planned for installation, an FGT
system which ioes not affect and is not affected by other control systems is
desirable. This aspect of the processes is determined by inspection of the
chemistry and equipment of the FGT system as well as other pollutant control
systems.
3-8
-------
3.1.2 Selection of Control Levels—Moderate, Stringent, and Intermediate
The control levels selected are applied to the following boilers:
Fuel
Gas
Gas
011-
dist.
011-
dist.
Oil-
res id.
011-
resid.
Coal
Coal
Coal
Load
Type (MWf-)
Firetube
Watertube
Firetube
Watertube
Watertube
Watertube
Underfeed
Stoker
Chaingrate
Spreader
Stoker
4.4
44
4.4
44
44
8.8
8.8
22
44
Uncontrolled NOx Emissions
(Ib/hr) (lb/106 Btu)
2.63
26.26
2.38
23.76
60.00
12.00
High S Low S Low S High S
Eastern Eastern Western E
19.05 16.35 23.40 0.64
47.70 40.80 58.65 0.64
95.40 81.45 117.15 0.64
0.18
0.18
0.16
0.16
0.40
0.40
Low S
E
0.55
0.54
0.54
Low S
W
0.78
0.78
0.78
Coal Pulverized 58.6
Coal
152.46 130.50 187.56 0.76 0.65 0.94
These NOX emission levels are all lower than the following average State
Implementation Plan (SIP) requirements except for one oil-fired boiler, one
coal-fired boiler burning high sulfur eastern coal, and all coal-fired boilers
burning low sulfur western coal.
3-9
-------
Coal 0.7
Oil 0.3
Gas 0.2
The moderate level of control is defined as representing that level
which is achievable applying techniques in current practice within industry.
This is the least stringent emission reduction achievable applying accepted
engineering practice. For FGT systems, this represents an NOX removal of
approximately 70 percent. When considering NOx FGT, it is not reasonable to
consider a removal level less than 70% since such levels can probably be
achieved by combustion modification techniques at lower costs. Allowable NOx
emissions at this control level are shown below:
Fuel Emission Level ^
VlOb Btu
Coal 0.24
Oil 0.09
Gas 0.06
Most of the control techniques are capable of controlling the standard
boilers with the highest NO emissions at this level.
x
The stringent level of control is defined as a technology-forcing level
and represents the most rigorous control which might be considered. This
represents an N0x removal of 90 percent. Allowable emissions at this control
level are shown below:
Fuel Emission Level
\ J-U iitu
Coal 0.08
Oil 0.03
Gas 0 . 02
3-10
-------
These systems are operating at their upper limit of practical NO removal
capability to achieve this level of control and are definitely technology-
forcing.
Intermediate level of control is defined as a level between moderate and
stringent and probably representing a technological or cost breakpoint. At
this point in time, it is difficult to say if those logical breakpoints exist
and, if so, where they are. Therefore, the intermediate level was chosen
between moderate and stringent levels. The intermediate levels of control
considered here represent about 80 percent NO removal. Allowable emissions
X
at this level are shown below:
Ib NOV
Fuel Emission Level
10b Btu
Coal 0.16
Oil 0.06
Gas 0.04
The best FGT systems should be able to achieve steady-state control at this
level. This control level provides an alternative choice between the least-
stringent and technology-forcing options.
The allowable emission rates for each of the control levels are
summarized in Table 3.1.2-1.
TABLE 3.1.2-1. CONTROLLED EMISSION LEVELS IN THIS STUDY (lb/106 Btu)
Coal
Oil
Gas
Moderate
0.24
0.09
0.06
Intermediate
0.16
0.06
0.04
Stringent
0.08
0.03
0.02
3-11
-------
3.2 BEST CONTROL SYSTEMS FOR COAL-FIRED BOILERS
A three phase selection process was used to determine the best NOX
control systems. The first phase involves comparing the maximum removal
level obtainable by each process with the level of control desired—moderate,
intermediate or stringent. Those process types which cannot achieve this
level are eliminated from further consideration. The remaining process types
are then evaluated using the criteria established in Section 3.1.1. The
result is a set of process types that are most desirable for a particular
consideration of special characteristics of the process types in the set in
order to determine the best system candidates. For example, all SCR
processes may rate high for application to gas-fired boilers. However, the
SCR fixed packed bed process may be more applicable than the moving bed or
parallel flow SCR processes since ability to tolerate particulates is not re-
quired for gas-fired boiler applications.
For use in the application of the selection factors, tables are compiled
which list the process features pertinent to each selection factor. The data
in these tables was derived from information presented in Section 2. For
coal-fired boilers, this information is presented in Tables 3.2-1 and 3.2-2.
3.2.1 Moderate Reduction Controls
The first phase evaluation eliminated the adsorption process from con-
sideration since it cannot achieve 70% NO reduction at high NO concentra-
tions (400-600 ppm). Application of the selection factors resulted in numeri-
cal ratings for the remaining processes as shown in Table 3.2.1-1. As can be
seen, the four SCR processes were superior. The fixed packed bed technique
was eliminate i since it would rapidly plug due to the high particulate levels
encountered wir.h coal-fired applications. Therefore, the candidate systems
for moderate control of coal-fired boilers are SCR parallel flow and SCR mov-
ing bed for NO -only removal and SCR parallel flow for simultaneous NO /SOV
X A
removal.
3-12
-------
TABLE 3.2-1.
COMPARISON INFORMATION OF NO -ONLY SYSTEMS FOR COAL-FIRED BOILERS
SCR Fixed
Packed Bed
SCR Moving Bed
SCR Parallel
Flow
OJ
I
Performance
Operational and
maintenance impacts
Preliminary
environmental
impacts
Capable of at taining
>90% NOX control level;
cannot be used with high
particulate levels; good
load following
capability.
Capable of attaining
>90% NOx control level;
can be used with some
particulates (up to
2.0 g/Nm3); adequate
load following
capability.
Capable of attaining
>90% NOX control level;
can be used wJ th full
particulate loading (up
to 20 g/Nm3); good load
following capability.
Few moving parts;
gas phase chemistry;
simple process - good
controllability; need
high removal of
particulates - ESP;
large pressure drop.
Moving parts, solids
handling - increased
maintenance; gas phase
chemistry; fairly
simple - controllable;
need particulate
removal; low pressure
drop.
Few moving parts;
gas phase chemistry;
simple process - good
controllabiltiy;
moderate pressure
drop; no particulate
removal needed.
Potential for some NHa
and NIK HSOi4 emissions.
Potential for some NHa
and NHi|HSOi4 emissions.
Potential for some NH3
and NH^HSO^ emissions.
Prelim Lnary
economic
Lmpacl.R
20 MW estimates:
Capital: $130/kWi
Operating:
2.1 raills/kWh1'2
Cost is higher than
other SCR's due to ESP.
20 MW est iraates:
Capital: $92/kW3
Operating: 2.0 mills/kWh
20 MW estimates:
Capital: $Wkw"
Operating: 1.5 mills/kWh
Preliminary
energy anr'
material impacts
El'ectrical usage:
1.2% of total output;
large NHs demand (1:1
NH3:NOX mole ratio);
may require auxiliary
heater.
Electrical usage:
unknown - should be
<1%; large NH3 demand
(1:1 NH3:NOx mole
ratio); may require
auxiliary heater;
greater catalyst
attrition due to
moving bed.
Electrical usage:
0.2% of total output;
large NHa demand (1:1
NH3:NOx mole ratio);
may require auxiliary
heater.
Absorption-
Oxidalion
No removal data are
available - should be
able to achieve
moderate control level;
can be used with full
particulate loading;
fair load following
capability.
Complex process with
sensitive control needs;
sensitive to flue gas
sulfur content -
separate SOx scrubber
before NOX absorber;
prescrubber needed to
remove particulates
and Cl~"; very large
pressure drops.
salts in wastewater.
20 MW estimates: none
available, but since
process contains extra
scrubber train,
Capital: $500/kW
Operating: 8 mills/kHh
Electrical usage:
unknown, estimate -3%;
uses large amounts of
gas-phase oxidant and
by-product treatment
materials.
-------
TABLE 3.2-1. (Continued)
Boiler operation
and/or safety
SCR Fixed No safety hazards.
Packed Bed
SCR Moving Bed No safety hazards.
Reliability
Catalyst easily plugged;
possible NHuHSOi* scaling;
simple - few process
steps; little corrosion
resistant material.
Little catalyst plugging;
possible NHitHSOi) scaling;
fairly simple - few
process steps; little
Status of
development
Has only been tested on
bench-scale (8 oil- and
numerous gas- fired
commercial operations).
Has only been tested on
bench-scale (5 oil- and
3 coke oven gas-fired
commercial operations) .
Adaptability to
existing sources
Some difficulty; few
pieces of process
equipment; little land
needed.
Some difficulty; few
pieces of process
equipment ; little land
needed .
Compatibility
with other
control systems
Excessive reheat
required if after FGD;
needs ESP.
Excessive reheat
required if after FGD;
needs particulate
removal.
corrosion resistant
material.
SCR Parallel
Flow
No safety hazards.
I
M
•P-
Little catalyst plugging
(must be packed well);
possible NH^HSOi, scaling;
simple - few process
steps; little corrosion
resistant material.
Has only been tested on
bench-scale; pilot
plants due to start up
in 1979 (some oil-fired
operations); commercial
operation by 1981.
Some difficulty;
catalyst can be placed
in duct between
economizer and preheater
without a separate
reactor; few pieces of
equipment; little land
needed.
Completely compatible
with FGD.
Absorption-
Oxidation
Oxidant handling could
be hazardous.
Numerous process steps
and corrosion resistant
material.
Has not been tested on
coal-fired flue gas
(a few pilot plants
treating oil-fired and
furnace gases).
Much land needed for
numerous pieces of
process equipment and
wastewater treatment.
Existing FGD would be
helpful as process
cannot tolerate sulfur.
-------
TABLE 3.2-2.
COMPARISON INFORMATION OF SIMULTANEOUS 1TO/SC) SYSTEMS FOR COAL-FIRED BOILERS
_s\ X
Performance
Operational and
maintenance impacts
Preliminary
environmental
impacts
Preliminary
economic
impacts
Preliminary
energy and material
impacts
SRC Parallel
Flow
Adsorption
I
H
Ln
Electron Beam
Radiation
Ahsorp tion-
Reduc tIon
Oxldation-
Absorption-
Reduction
Oxidation-
Absorption
Capable of attaining 90%
control of both NOx and
SOx; can be applied to
gases with high particu-
late loadings; process
can follow boiler load
easily through use of
gas bypass arrangement.
Capable of attaining 60%
NOx control level;
cannot be used with high
particulate levels; poor
load following capability;
primarily SO* removal.
Capable of attaining 80%
NOX control level;
cannot be used with
particulates; fair load
following capability;
also removes SOX.
Capable of attaining 85%
NOx control level; can
be used with full
particulate loading;
good load following
capability; removes SOx.
Capable of attaining 90%
NOX control level; can
be used with full
particulate loading;
poor load following
capability - oxidant
generation lagtime;
removes SOX.
Capable of attaining 90%
NOx control level; can
be used with full
particulate loading;
poor load following
capability; removes SO^.
Process has several
sections but all
except NOx/SOx
reactor are based on
well established
technology; average
maintenance require-
ments .
Many moving parts,
hot solids handling;
complex process; need
ESP for particulate
removal; major mainte-
nance requirements;
high pressure drop.
Simple process but
complex control;
sensitive to flue gas
composition (at least
1% 02 and H70>NOxO.
Complex process with
very sensitive control
needs; sensitive to
flue gas composition
(low 02 and SOx:NOx
ratio >2.5); need
prescrubber to remove
particulates and Cl~.
Complex process with
very sensitive control
needs; prescrubber
needed to remove
particulates and Cl~;
large pressure drop.
Complex process with
very sensitive control
needs; prescrubber
needed; large pressure
drop.
Potential Nils emissions.
Ash disposal.
H2S04 mist and a powder
containing ammonium
nitrates and sulfates
are generated.
Possibility of plume
from absorbent (sulfate
or NH3).
N03 or N-S salts or
NHs" based compounds in
wastewater.
N03 salts in wastewater.
20 MW estimates:
Capital: $475/kW
Operating: 5 mills/kWh
20 MW estimates:
Capital: $215 kW5
Operating: 2.3 mills/kWh
20 MW estimates:
Capital: $202 kW6
Operating: unknown
Electricity is only
major.
20 MW estimates:
Capital: $413/kW7
Operating: 7.4 mills/kWli
Gypsum by-product
(landfill).
Economic estimates:
unknown for coal-fired
plant; gypsum by-product
(landfill).
Economic estimates:
unknown; gypsum
by-product and liquid
fertilizer, or HNO,.
Electrical usage:
1.5% of total output;
also consumes NHj,
naphtha, and steam.
Electrical usage:
unknown - should be
?2%; activated char
usage high due to
attrition.
Electrical usage:
3.3% of total output
(excluding ESP);
treatment of by—product
is unknown.
Electrical usage:
1.8% of total output;
large amounts of
chelating compound,
absorbent, and
regeneration chemicals
are used.
Electrical usage:
9.0% of total output;
uses large amounts of
gas-phose oxidant and
by-product treatment
materials.
Electrical usage:
unknown (will be =10%
of total output): uses
large amounts of gas-
phase oxidant and
by-product treatment
materials.
-------
TABLE 3.2-2. (Continued)
SCR Parallel
Flow
Boiler operation
and/or safety
H? usage may present
safety hazard.
Reliability
Process steps well
established; should be
reliable.
Status of
development
SOa system has been
tested on coal-fired
flue gas; NOX/SOX
Adaptability to
existing sources
Will need land for
equipment .
Compatibility
with other
control systems
Compatible with
particulate systems.
Adsorption
Electron Beam
Radiation
Absorption-
Reduction
Oxidation-
Absorption-
Reduction
Oxidation-
Absorption
Possible safety hazard
due to poor char
distribution in beds.
Radiation safety
hazards are unknown as
are those of byproduct.
No safety hazards.
Gas-phase oxidant
presents serious
safety hazard.
Gas-phase oxidant
presents serious
safety hazard.
Char plugged by
part iculates; numerous
process steps; some
corrosion resistant
material in high
temperature zones.
Few process steps;
stainless steel
reactor.
Many process steps;
much glass- and
elastomer-lined
equipment.
Numerous process steps
and corrosion resistant
material; oxidant
generation system
subject to periodic
failure.
Numerous process steps
and corrosion resistant
material; oxidant
generation system
subject to periodic
failure.
operation with coal-
fired flue gas to begin
late 1979; pilot unit
tests; S0£ work up and
H2 generation not
tested, but are
established technology.
One prototype unit
treat ing coal-fired
flue p,as.
Has not been tested on
coal-fired flue gas
(one pilot plant treat-
ing gas from sintering
machine); uncertain
by-product treatment
method.
Has not been tested on
coa]-fired flue gas
(several pilot plants
treating oil-fired flue
gas); NOx absorption
chemistry uncertain.
Has not been tested on
coal-fired flue gas (6
prototype units treating
oil-fired flue gas in
operation).
One pilot plant treating
flue gas from coal-fired
boiler.
Need land for pieces
of process equipment.
Need land for pieces
of process equipment..
Much land needed for
numerous pieces of -
process equipment.
Much land needed for
numerous pieces of
process equipment,
oxidant generation, and
wastewater treatment.
Much land needed for
numerous pieces of
process equipment,
oxidant generation, and
wastewater treatment.
Suitable for placement
after ESP; not useful
with FGD system as NOX
removal is secondary.
Needs ESP; with or
without existing FGD
but capital cost will
be the same.
Cannot be used in
conjunction with FGD.
Cannot be used in
conjunction with FGD.'
Compatible.
-------
TABLE 3.2.1-1, CANDIDATE SYSTEMS SELECTION: COAL-FIRED BOILERS - MODERATE CONTROL
U)
Control technique
N0x-0nly
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidat ion
Simultaneous NOX/SOX
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption-Reduction
Oxidation- Absorption-Reduction
Oxidation- Absorption
Total point Candidate
rating system
69 no
70 no
83 yes
43 no
72 yes
NA no
41 no
52 no
51 no
51 no
Comments
Adversely affected
by particulates
Adversely affected
by particulates
Low rating
Low rating
Low rating
Low rating
Low rating
NA - Not applicable (see Appendix)
-------
A detailed listing of how each process was evaluated on each selection
factor is contained in Tables A3.1 and A3.2 in the Appendix.
3.2.2 Stringent Reduction Controls
In a similar manner, candidate systems for stringent control were
selected. The results appear in Table 3.2.2-1. A detailed listing of the
selection factors and point values for each system is contained in Tables
A3.3 and A3.4. The candidate systems selected are SCR parallel flow and SCR
moving bed for N0x~only removal and SCR parallel flow for simultaneous NO /
SO removal.
3.2.3 Intermediate Reduction Controls
The selection results for this level are presented in Table 3.2.3-1.
Detailed application of the selection factors is presented in Tables A3.5
and A3.6. The candidate systems selected are SCR parallel flow and SCR
moving bed for N0x-only removal and SCR parallel flow for simultaneous
NO /SO removal.
X X
3.3 BEST CONTROL SYSTEMS FOR OIL-FIRED BOILERS
The control systems for oil-fired boilers were evaluated using the same
method described in the previous section on coal-fired boilers. Tables
3.3-1 and 3.3-2 present a side-by-side comparison of all potential
systems with data categorized with respect to the selection factors. The
information in this table is summarized from Section 2. The table is
similar in many respects to the equivalent table for coal. This is due to
the fact that, since FGT systems are applied after the boiler, they are
relatively insensitive to the types of fuel burned. Two notable exceptions
are particulate and sulfur emissions which are a function of the fuel type.
Process characteristics that change with fuel type are noted in the table.
3-18
-------
TABLE 3.2.2-1. CANDIDATE SYSTEMS SELECTION: COAL-FIRED BOILERS - STRINGENT CONTROL
Control technique
Total point
rating
Candidate
system
Comments
NOx-Only
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidation
62
60
73
NA
no
no
yes
no
Adversely affected
by particulates
Adversely affected
by particulates
Simultaneous NOX/SOX
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption-Reduction
Oxidation- Absorption-Reduction
Oxidation- Absorption
68
NA
NA
NA
48
49
yes
no
no
no
no
no
Low rating
Low rating
NA - Not applicable (see Appendix)
-------
TABLE 3.2.3-1. CANDIDATE SYSTEMS SELECTION: COAL-FIRED BOILERS - INTERMEDIATE CONTROL
to
o
Control technique
NOx-Only
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidation
Simultaneous NOX/SOX
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption-Reduction
Oxidat ion-Absorption-Reduction
Oxidation- Absorption
Total point Candidate
rating system
67 no
69 no
81 yes
43 no
70 yes
NA no
41 no
50 no
49 no
46 no
Comments
Adversely affected
by particulates
Adversely affected
by particulates
Low rating
Low rating
Low rating
Low rating
Low rating
NA - Not applicable (see Appendix),
-------
TABLE 3.3-1.
COMPARISON INFORMATION OF NO -ONLYSYSTEMS FOR OIL-FIRED BOILERS
Performance
Operational and
maintenance ijnpacts
Preliminary
environmental
impacts
Preliminary
economic
impacts
Preliminary
energy and
material impacts
SCR Fixed
Packed Bed
SCR Moving Bed
SCR Parallel
Flow
Absorption-
Oxidation
Capable of achieving
>90% NOX reduction;
cannot be used with
high particulate levels;
good load following
capability.
Capable of achieving
>90% NOX reduction;
can tolerate particulate
level of moat oils (<1
g/Nra3); adequate load
following capability.
Capable of achieving
>9Q% NOx reduction;
can tolerate full
particulate loading (up
to 20 g/Nm'); good load
following capability.
No removal data are
available; can tolerate
particulates; fair load
following capability;
removes SOX.
Few moving parts;
gas phase chemistry;
simple process - good
controllability; need
high removal of
particulates - ESP;
large pressure drop.
Some moving parts,
solids handling -
increased maintenance;
gas phase chemistry;
simple - controllable;
low pressure drop.
Few moving parts;
gas phase chemistry;
simple process - good
controllability; no
particulate removal
needed; moderate
pressure drop.
Complex process with
sensitive control
needs; sensitive to
flue gas sulfur
content - separate SOX
scrubber before NOX
absorber; prescrubber
needed; very large Ap.
Potential for some NHj
and NH^HSOi, emissions.
Potential for some NHs
and NHnHSOi, emissions.
Potential for some NHa
and NHuHSCH emissions.
N03~ salts in wastewater.
20 HU estimates:
Capital: $70/kW8'9
Operating:
1.9 mills/kWh18.11
20 MW estimates:
Capital: $70/kw''''2
Operating: 1.8 mills/kWh
20 MW estimates:
Capital: $39/kw'3
Operating: unknown.
Economic estimates:
unknown.
Electrical usage:
unknown for oil-fired
plant; moderate NH3
demand (1:1 NH3:NOX
mole ratio); may
require auxiliary
heater.
Electrical usage:
unknown for oil-fired
plant; moderate NHs
demand (1:1 NH3:NOX
mole ratio); may
require auxiliary
heater; greater
catalyst attrition
due to moving bed.
Electrical usage:
unknown for oil-fired
plant; moderate NH3
demand Cl:l NH3:NOX
mole ratio); may
require auxiliary
heater.
Electrical usage:
unknown; uses large
amounts of liquid
phase oxidant and
regeneration
materials.
-------
TABLE 3.3-1. (Continued)
Boiler operation
and/or safety
SCR Fixed No safety hazards.
Packed Bed
Status of
Reliability development
Catalyst easily plugged; 8 commercial operations
possible NH*HSOi, scaling; in Japan.
simple - few process
steps; little corrosion
resistant material.
Adaptability to
existing sources
Some difficulty; few
pieces of process
equipment; little land
required.
Compatibility
with other
control systems
Excessive reheat
required if after FGD;
needs ESP.
SCR Moving Bed
SCR Parallel
Flow
Absorption-
Oxidation
No safety hazards.
No safety hazards.
Oxidant handling can
be dangerous.
Some catalyst plugging;
possible NHuHSOt, scaling;
fairly simple - few
process steps; little
corrosion resistant
material.
Little catalyst plugging;
possible NHiiHSOi scaling;
simple - few process
steps; little corrosion
resistant material.
Numerous process steps
and corrosion resistant
material.
6 commercial oil-fired
operations in Japan.
Numerous commercial
oil-fired operations
in Japan.
2 pilot plants treating
oil-fired flue gas.
Some difficulty; few
pieces of process
equipment; little land
required.
Some difficulty;
catalyst can be placed
in duct between
economizer and preheater
without a separate
reactor; few pieces of
equipment; little land
needed.
Much land needed for
numerous pieces of
process equipment and
wastewater treatment.
Excessive reheat
required if after FGD.
Completely compatible
with FGD.
Existing FGD would be
helpful as process
cannot tolerate sulfur.
-------
TABLE 3.3-2.
COMPARISON INFORMATION OF SIMULTANEOUS NOX/SOX SYSTEMS FOR OIL-FIRED BOILERS
Performance
Operational and
maintenance Impacts
Preliminary
env ironmen tal
Impacts
Preliminary
economic
impacts
Preliminary
energy and
mat er ial impact s
SCR Parallel
Flow
Adsorption
Electron Beam
Radiation
Absorption-
Reduction
Oxidation-
Absorption-
Reduction
Oxidation-
Absorption
Capable of attaining 90%
control of both NO* and
S02; can be used with
full particulate loading;
good load following
capability.
Capable of attaining 60%
NOx reduction; cannot be
used with high
particulate levels; poor
load following
capability; primarily
SOX removal.
Capable of attaining 80%
NOX removal; cannot be
used with particulates;
fair load following
capability; also removes
SO*.
Capable of attaining 85%
removal; can tolerate
particulates; good load
following capability;
cannot be used on
distillate oil; also
removes SOX .
Capable of attaining 90%
NOX reduction; can
tolerate particulates;
poor load following
capability - oxldant
generation lagtime;
cannot be used on distil-
late oils; removes SOX.
Capable of attaining 90%
NOX reduction; can
tolerate particulates;
poor load following
capability; removes SOX.
Most of the process
steps are based on
well established
technology; average
maintenance require-
ments.
Many moving parts,
hot solids handling -
high maintenance;
complex process; may
need particulate
removal on residual
oils; large pressure
drop.
Simple process with
complex control;
sensitive to flue gas
composition (at least
1% Oj and HzO>NOx);
may need particulate
removal on residual
oils.
Complex process with
very sensitive control
needs; sensitive to
flue gas composition
(low Oa and SOxiNOy
ratio >2.5); need
prescrubber; large AP.
Complex process with
very sensitive control
needs; prescrubber
needed; large pressure
drop.
Complex process with
very sensitive control
needs; prescrubber
needed; large pressure
drop.
Potential NHa emissions.
Ash disposal.
H2 SO, mist and a
powder containing
ammonium-nitrates and
sulfates are generated.
Possibility of plume
(sulflte or NHj) from
absorbent.
NOS~ or N-S salts or
NH-base compounds In
wastewater.
N0j~ salts in wastewater.
Economic estimates:
unknown for oil-fired
plant; assumed to be
similar to those for
coal.
20 MW estimates for coal:
Capital: $475/kW
Operating: 5 mills/kWh
Ecomonlc estimates:
unknown for oil-fired
plant; elemental S by-
product.
Economic estimates:
unknown for oil-fired
plant; electricity is
primary operating
expense.
20 MW estimates:
Capital: $187/kW11<>15
Operating: 5.4 mills/kWh
Gypsum by-product
(landfill).
20 MW estimates:
Capital: $231/kW16'17
Operating: 6.4 mills/kWh
Gypsum by-product
(landfill).
Economic estimates:
unknown; gypsum by-product
and liquid fertilizer or
HN03.
Electrical usage:
unknown for oil-fired
plant; assumed to be
similar to those for
coal, i.e. 1.5% of
boiler output as
electricity; also
uses steam, naphtha
and NHs .
Electrical usage:
unknown for oil-fired
plant; large
activated char demand
due to attrition.
Electrical usage:
unknown for oil-fired
plant; treatment of
by-product is unknown.
Electrical usage:
1.8% of total output;
extremely large
amounts of chelatlng
compound; absorbent
and regeneration
chemicals are used.
Electrical usage:
5-10% of total
output; large amount
of gas phase oxldant
and by-product treat-
ment materials.
Electrical usage:
unknown (will be 5-
10% of total output);
uses large amounts
of gas phase oxidant
and by-product treat-
ment materials.
-------
TABLE 3.3-2. (Continued)
SCR Parallel
Flow
Boiler operation
and/or safety
H2 is potential
safety hazard .
Reliability
Process steps well
established. Should
be reliable.
Status of
development
Both S0x and N0x removal
systems have been tested
on oil. S02 workup and
H2 generation steps not
tested but are established
technology .
Adaptability to
existing sources
Will need land for
equipment
Compatibility
with other
control systems
Compatible with par-
ticulate control
systems
Adsorption
Electron Beam
Radiation
Absorption-
Reduction
Oxidation-
Absorption-
Reduction
Oxidation-
Absorption
Possible safety hazard
due to poor char distri-
bution in beds.
Radiation safety hazards
are unknown as are those
of by-product.
No safety hazards.
Gas phase oxidant
presents serious safety
hazard.
Gas phase oxidant
presents serious
safety hazards.
Particulate plugging;
numerous process steps;
some corrosion resistant
material in high temp-
erature areas.
Few process steps;
stainless steel reactor.
Many process steps;
much glass - and
elastomer-lined
equipment.
Numerous process steps
and corrosion resistant
material; oxidant gen-
eration system subject
to periodic failure.
Numerous process steps
and corrosion resistant
material; oxidant gener-
ation system subject to
periodic failure.
No tests on oil-fired gas. Need land for pieces of
process equipment
One oil-fired pilot plant;
by-product treating
method is uncertain.
3 pilot plants treating
oil-fired flue gas;
NOX absorption mechanism
uncertain.
6 prototype units treat-
ing oil-fired flue gas
in operation.
One bench-scale test on
oil-fired flue gas.
Need land for pieces of
process equipment.
Much land needed for
numerous pieces of
process equipment.
Much land needed for
numerous pieces of
process equipment,
oxidant generation and
wastewater treatment.
Much land needed for
numerous pieces of process
equipment, oxidant gener-
ation, and wastewater
treatment.
Not useful with FGD
systems NOX removal
is secondary.
Operate with or without
FGD but capital cost
is same.
Cannot be used in
conjunction with FGD.
Cannot be used in
conjunction with FGD.
Compatible
-------
3.3.1 Moderate Reduction Controls
One system, adsorption, was eliminated because it was not capable of
achieving sufficient emission reduction. The remaining systems were rated
using the selection factors and the results are presented in Table 3.3.1-1.
A detailed breakdown of this evaluation is contained in Tables A3.7 and A3.8.
SCR fixed packed bed was selected as the NO -only candidate system for
distillate-oil-fired boilers since these have low particulate emissions.
For resid-fired boilers, which have higher particulate emissions, the NO -
only candidate systems are SCR parallel flow and SCR moving bed and the
simultaneous NO /SO candidate system is SCR parallel flow.
XX
3.3.2 Stringent Reduction Controls
The results of system evaluations for stringent control levels are shown
in Table 3.3.2-1. The detailed evaluation breakdown is contained in Tables
A3-9 and A3-10. The candidate systems are the same as for moderate control.
3.3.3 Intermediate Reduction Controls
The results of system evaluations for intermediate control levels are
shown in Table 3.3.3-1. A detailed breakdown of the selection factor ratings
is presented in Tables A3.11 and A3.12. The candidate systems are the same
as for the other two levels: NO -only, SCR fixed packed bed for distillate
X
oil plus SCR parallel flow and SCR moving bed for resid oil; simultaneous
NO /SO _ SCR parallel flow.
X A
3.4 BEST CONTROL SYSTEMS FOR GAS-FIRED BOILERS
Table 3.4-1 compares all of the FGT systems as applied to gas-fired
boilers for each of tha selection factors. This table was used to arrive
at the point values shown on the candidate selection tables.
3-25
-------
TABLE 3.3.1-1. CANDIDATE SYSTEMS SELECTION: OIL-FIRED BOILERS - MODERATE CONTROL
Control technique
NOx-Only
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidation
Simultaneous NOx/SOx
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption-Reduction
Oxidation- Absorption- Reduction
Oxidation- Absorption
Total point
rating
81
88
90
53
75
NA
47
58
59
52
Candidate
system
yes
yes
yes
no
yes
no
no
no
no
no
Comments
Distillate oil-fired boilers
Residual oil-fired
Residual oil-fired
Low rating
Residual oil-fired
Low rating
Low rating
Low rating
Low rating
boilers
boilers
boilers
NA - Not applicable (see Appendix).
-------
TABLE 3.3.2-1. CANDIDATE SYSTEMS SELECTION: OIL-FIRED BOILERS - STRINGENT CONTROLS
Control technique
N0x-0nly
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidation
Simultaneous N0x/S0x
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption- Reduction
Oxidation-Absorption-Reduction
Oxidation- Absorption
Total point
rating
74
81
83
NA
71
NA
NA
NA
54
50
Candidate
system
yes
yes
yes
no
yes
no
no
no
no
no
Comments
Distillate oil-fired boilers
Residual oil-fired boilers
Residual oil-fired boilers
Low rating
Low rating
NA - Not applicable (see Appendix),
-------
TABLE 3.3.3-1. CANDIDATE SYSTEMS SELECTION: OIL-FIRED BOILERS - INTERMEDIATE CONTROL
M
OO
Control technique
N0x-0nly
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption-Oxidation
Simultaneous NOX/SOX
SCR Parallel Flow
Adsorption
Electron Beam Radiation
Absorption- Reduction
Oxidation- Absorption-Reduction
Oxidation-Absorption
Total point
rating
79
86
88
54
73
NA
47
58
55
48
Candidate
system
yes
yes
yes
no
yes
no
no
no
no
no
Comments
Distillate oil-fired boilers
Residual oil-fired
Residual oil-fired
Low rating
Residual oil-fired
Low rating
Low rating
Low rating
Low rating
boilers
boilers
boilers
NA - Not applicable (see Appendix).
-------
TABLE 3.4-1.
COMPARISON INFORMATION OF NOx~ONLY SYSTEMS FOR GAS-FIRED BOILERS
Performance
Operational and
maintenance impacts
Preliminary
environmenta1
impacts
Preliminary
economic
impacts
Preliminary
energy and
material impacts
SCR Fixed Capable of achieving
Packed Bed >90% NOX removal; good
load following
capability.
Few moving parts;
gas phase chemistry;
simple process - good
controllability; large
pressure drop.
Potential for some NHa
and Ntii,HSOi, emissions.
20 MW estimates for
clean gas:
Capital: $27/kVT
Operating: 1.2 mills/kWh
Electrical usage:
unknown for gas-fired
flue gas; light NH3
demand (1:1 NH3:NOX
mole ratio).
SCR Moving Bed
Capable of achieving
>90% NOX removal;
adequate load following
capability.
Some moving parts,
solids handling -
increased maintenance;
gas phase chemistry;
simple - controllable;
low pressure drop.
Potential for some NHs
and NHi,HSOi| emissions.
Economic estimates:
unknown for gas-fired
plant.
Electrical usage:
unknown for gas-fired
flue gas; light NH3
demand (1:1 NH3:NOX
mole ratio).
CO
SCR Parallel
Flow
Absorption-
Oxidation
Capable of achieving
>90% NOX removal; good
load following
capability.
No removal data are
available; fair load
following capability.
Few moving parts;
gas phase chemistry;
simple process - good
controllability;
moderate pressure drop.
Complex process with
sensitive control
needs; very large
pressure drop.
Potential for some NH3
and NH^HSOt, emissions.
NOa salts in waste-
waters.
Economic estimates:
unknown for gas-fired
plant,
Economic estimates:
unknown.
Electrical usage:
unknown for gas-fired
flue gas; light NIU
demand (1:1 NH3:NO*
mole ratio).
Electrical usage:
unknown; uses large
amount s o f 1iqu id
phase oxidant and
regeneration
materials.
-------
TABLE 3.4-1. (Continued)
u>
o
Boiler operation
and/or safety
SCR Fixed No safety hazards.
Packed Bed
SCR Moving No safety hazards.
Bed
SCR Parallel No safety hazards.
Flow
Absorption - Oxidant handling could
Oxidation be dangerous.
Reliability
Possible NH^HSOi, seal ing;
simple - few process
steps; little corrosion
resistant material.
Possible NH^HSO^ scaling;
simple - few process
steps; little corrosion
resistant material.
Possible NIUHSO., scaling;
simple - few process
steps; little corrosion
resistant material.
Numerous process steps
and corrosion resistant
material .
Status of
development
Numerous commercial
operations in Japan.
Three commercial coke
oven gas operations in
Japan .
No commercial operations
(many oil-fired; not
necessary for gas-fired -
no particulates)
Pilot plants treating
off gases from HN03 and
steel plants.
Adaptability to
existing sources
Some retrofit difficulty;
few pieces of process
equipment - little land
required .
Some retrofit difficulty;
few pieces of process
equipment - little land
required .
Some retrofit difficulty;
few pieces of process
equipment - little land
required; if space exists,
catalyst can fit in duct.
Much land needed for
numerous pieces of process
equipment and wastewater
treatment.
Compatibility
with other
control systems
Compatible
Compatible
Compatible
Compatible
-------
3.4.1 Moderate Reduction Controls
The first cut in FGT systems applied to gas-fired boilers eliminated
one process due to insufficient emission reduction and five processes due to
their removal of SOX which is not present in gas-fired flue gas. This can
be seen in Table 3.4.1-1 which presents the results of the candidate selec-
tion. SCR fixed packed bed was chosen as the candidate system. SCR parallel
flow and SCR moving bed were eliminated since their specialized ash handling
characteristics are not required for this application. A detailed selection
factor rating breakdown is contained in Table A3.13.
TABLE 3.4.1-1. CANDIDATE SYSTEMS SELECTION:
GAS-FIRED BO-ILERS - MODERATE CONTROL
Total point
Control technique rating
N0y-0nly
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption - Oxidation
93
91
93
58
Candidate
system
Yes
No C
No \
No
Comments
These specialized catalyst
arrangements are not neces-
sary for gas-fired sources.
Low rating
3.4.2 Stringent Reduction Controls
The results of system evaluations for stringent control levels are shown
in Table 3.4.2-1. The detailed evaluation breakdown is contained in Table
A3.14. The candidate system is SCR fixed packed bed.
3-31
-------
TABLE 3.4.2-1. CANDIDATE SYSTEMS SELECTION:
GAS-FIRED BOILERS - STRINGENT CONTROL
Total point Candidate
Control technique rating system
Comments
NO x-0nly
SCR Fixed Packed Bed
SCR Moving Bed
SCR Parallel Flow
Absorption - Oxidation
83
81
83
NA
Yes
No
No
These specialized catalyst
arrangements are not neces-
sary for gas-fired sources.
No
3.4.3 Intermediate Reduction Controls
The results of system evaluations for intermediate control levels are
shown in Table 3.4.3-1. The detailed evaluation is presented in Table A3.15.
The candidate system is SCR fixed packed bed.
TABLE 3.4.3-1. CANDIDATE SYSTEMS SELECTION:
GAS-FIRED BOILERS - INTERMEDIATE CONTROL
Control technique
Total point Candidate
rating system
Comments
N0y-0nly
SCR Fixed Packed Bed 87
SCR Moving Bed 85
SCR Parallel Flow 87
Absorption - Oxidation 58
Yes
No
No
No
These specialized catalyst
arrangements are not neces-
sary for gas-fired sources.
Low rating
3-32
-------
3.5 SUMMARY
A candidate or set of candidates has now been chosen for each of the
standard boilers under consideration. These are shown in Table 3.5-1. These
systems will be analyzed in detail in the subsequent sections in order to
determine the best overall system for NOX reduction by FGT on industrial
boilers. The major performance characteristics for the candidate processes
are presented in Table 3.5-2.
TABLE 3.5-1. SUMMARY OF CANDIDATE SYSTEMS: ALL LEVELS OF CONTROL
Fuel Candidate Systems
Coal SCR Parallel Flow
Residual Oil SCR Parallel Flow
SCR Moving Bed
Distillate Oil SCR Fixed Packed Bed
Natural Gas SCR Fixed Packed Bed
3-33
-------
TABLE 3.5-2. MAJOR PERFORMANCE CHARACTERISTICS OF CANDIDATE SYSTEMS
SCR-fixed
packed
bed
Collection
Efficiency
>90%
N0x
reduction
Environmental
Impacts
Possible NH3
and NH^HSO^
emissions
Energy
Impacts
Coal-need
ESP elec.=
1.2% of
total power
input
Reliability
Simple, few
process steps;
catalyst
easily plugged
Commercial Availability
Coal-bench scale; 8 oil
and numerous gas-fired
commercial operations
u>
.p-
SCR-parallel >90%
flow N0x
reduction
SCR-moving >90%
bed N0x
reduction
Possible NH3
and Nl^HSO^
emissions
Possible NH3
and NH^HSO^
emissions
Coal-no ESP
elec.=0.2%
of total
power output
Coal-some
particulate
removal
needed
elec.
-------
REFERENCES
1. Faucett, H.L., et al. Technical Assessment of NO Removal Processes
for Utility Applications. EPA-600/7-77/127. November 1977. pp.
210, 268, 301.
2. Rosenberg, H.S., et al. State-of-the-Art Review of Stack Gas Treatment
Techniques for N0x Control. EPRI-Batelle. April 1976. p. A-47.
3. Faucett, H.L., op.ait. , pp. 233, 319.
4. Ibid. , p. 218.
5. Ibid. , p. 204.
6. Ibid. , p. 166.
7. Ibid. , p. 33
8. Ibid. , p. 347.
9. Rosenberg, H.S., op.cit., p. A-14.
10. Faucett, H.L., op. ait. , pp. 280, 327.
11. Rosenberg, H.S., op.cit.,p. A-23.
12. Faucett, H.L., op. sit.,p. 242.
13. Faucett, H.L., op.cit., p. 223.
14. Ibid. , p. 41
15. Rosenberg, H.S., op.cit., p. A-139.
16. Ibid., pp. A-106, A-116.
17. Faucett, H.L., op.cit., pp. 48, 117, 127.
3-35
-------
SECTION 4
COST ANALYSIS OF CANDIDATES
FOR BEST EMISSION CONTROL SYSTEMS
4.1 NOx-ONLY SYSTEMS
4.1.1 Introduction
This section considers the costs involved with applying the "best" NOX
FGT systems selected in Section 3 to the standard boilers. The costs pre-
sented are based on several factors. First, typical process layouts were
determined to establish the equipment requirements. Material balances are
established for each case and the equipment sized. Process layouts and
material balances for all nineteen cases considered in detail are presented
in Appendices 3, 4 and 5 for coal, oil, and gas sources, respectively.
Purchased equipment lists for each process considered are shown in Table
4.1.1-1. The equipment is selected and sized by using standard engineering
techniques. Example calculations for equipment size and energy usage are
presented in the Appendix 8. Energy usage for all systems consists only of
electricity and steam. Other costs were based on cost factors supplied by
references 1 and 2 as well as other sources.
All of the equipment listed in Table 4.1.1-1 will require some mainte-
nance. The items requiring the most maintenance are the pump, fan motor
drive, vaporizer, screen, catalyst elevator, baghouse/blower, and all
associated process control elements. The catalyst has a lifetime of about
one year and its regeneration is presently uncertain. Therefore, in this
analysis, it is replaced annually and represents both a capital and operating
cost.
4-1
-------
TABLE 4.1.1-1. PURCHASED EQUIPMENT FOR NO FGT SYSTEMS
X
Parallel Flow SCR Moving Bed SCR Fixed Packed Bed SCR
Reactor Reactor Reactor
Catalyst Catalyst Catalyst
Fan Motor Drive Catalyst Screen Fan Motor Drive
NH3 Storage Tank Catalyst Elevator NH3 Storage Tank
NH3 Transfer Pump Baghouse/Blower NH3 Transfer Pump
NH3 Vaporizer Fan Motor Drive NH3 Vaporizer
NH3 Storage Tank
NH3 Transfer Pump
NH3 Vaporizer
The cost bases can be separated into several areas. Costs of materials
associated with all of the processes evaluated are presented in Table
4.1.1-2. Sources of the costs are also shown. Several costs were determined
by multiplying a factor times another cost. This is common with this type
of economic analysis and the cost factors used are shown in Table 4.1.1-3.
Direct costs were determined on a full year basis and then multiplied by the
boiler load factor to determine the annual direct costs. Load factors for
the standard boilers are shown in Table 4.1.1-4. The capital recovery factor
was calculated from the formula:
For i = 0.10 (interest) and n = 15 (years) the capital recovery factor is
0.13147.
The costs of each equipment item was determined using a variety of cost
references shown in Table 4.1.1-5. Installation costs were provided in the
references. As with the annual costs, some of the capital costs were deter-
mined by multiplying a factor times another cost. A list of the factors
used in the capital cost estimates is contained in Table 4.1.6.
4-2
-------
TABLE 4.1.1-2. ANNUAL COST PARAMETERS USED IN COST ANALYSIS
Item
Cost Used
Reference
Direct Labor, $/manhour
Maintenance Labor, $/manhour
Electricity, mills/kWh
Ammonia, $/ton delivered
Steam, $/1000 Ib
Catalyst, $/ft3
Parallel Flow
Moving Bed
Fixed Packed Bed
12.
14.
25.
130
3.
212
282
282
02
63
8
50
1
1
1
1
2
3
4
4
TABLE 4.1.1-3. ANNUAL COST FACTORS
Item
Amount
Reference
Maintenance Materials
Payroll Overhead
Plant Overhead
General and Administrative
Expenses (G&A),
Taxes & Insurance
Capital Recovery Factor
(10% interest rate)
3% of turnkey costs
30% of direct labor
26% of labor, parts & maintenance
4% of total turnkey costs
13.147% of total turnkey costs
5
1
1
1
1
TABLE 4.1.1-4. LOAD FACTORS FOR THE STANDARD BOILERS
Fuel
Load Factor
Coal
Residual Oil
Distillate Oil and Natural Gas
0.60
0.55
0.45
4-3
-------
TABLE 4.1.1-5. SOURCES OF COSTS FOR SPECIFIC EQUIPMENT ITEMS
Equipment Item
Reference
Reactor
Catalyst
Fan Motor Drive (Incremental)
NH3 Storage Tank
NH3 Transfer Pump
NH3 Vaporizer
Vibrating Screen
Catalyst Elevator
Baghouse
6
3,4
7,8
9
10
11
12
13,8
14
TABLE 4.1.1-6. CAPITAL COST FACTORS
Item
Amount
Reference
Engineering
Construction and Field Expense
Contractor Fee
Start-up
Performance Tests
Contingency
Working Capital
10% of installed cost of largest
NO removal system considered
(pulverized coal boiler; stringent
level of control)
10% of installed cost
10% of installed cost
2% of installed cost
$2000
Coal: 20% of total direct and
indirect costs
Oil and Gas: 15% of total direct
and indirect costs
25% of total direct operating
costs
1
1
1
1
1
4-4
-------
Capital costs were escalated to June 1978 costs using standard cost
indices. For example, costs in Guthrie6 are based on June 1970 costs.
Cost indices for this year and June 1978 for various types of equipment
are shown in Table 4.1.1-7. To obtain the mid 1978 costs the costs given
in Guthrie are multiplied by the escalation index.
TABLE 4.1.1-7. CHEMICAL ENGINEERING COST INDICES15
Item
Fabricated Equipment
Process Machinery
Pipes, Valves & Fittings
Process Instruments
Pumps & Compressors
Electrical Equipment
Miscellaneous
Construction Labor
June 1970
Index
124.0
122.7
133.0
132.0
124.1
98.9
118.5
134.8
June 1978
Index
237.4
226.6
268.4
214.8
258.2
167.9
250.1
184.3
Escalation
Index (E.I.)
(1978/1970)
1.91
1.85
2.02
1.63
2.08
1.70
2.11
1.37
The labor requirements were determined from the basis for an economic
analysis performed by a process vendor which indicated a requirement of one
person/shift/day per reactor. 16 Equipment life was estimated at 15 years
based on the average lifetime of chemical processing equipment.17 Capital
costs were annualized over a 15 year period to give constant annual costs
for the life of the boiler.
The capital and operating costs were collected and presented in a con-
sistent set of table^ and an annualized cost was calculated. These compre-
hensive tables are contained in separate appendices and discussed in the
subsequent subsections.
4-5
-------
Costs for modified or reconstructed facilities will most likely be
slightly higher than those for new facilities. This is due to the fact
the major cost items—i-.e. the fan motor drive, reactor plus catalyst, and
NHs storage tanks—are the same for both applications. There may be some
increased costs where additional ductwork, boiler modification or flue gas
heating is necessary and these factors are highly site specific. The cost
of a retrofit will have to be determined for each application since it is
dependent on site specific factors.
The cases considered include only one type of coal, low sulfur western.
Other coal types are not considered since process costs do not vary signifi-
cantly with coal type. Two of the most significant cost items for FGT sys-
tems are the reactor plus catalyst and the fan motor drive. These equipment
items are sized and costed based on flue gas flow rate which does not vary
significantly with coal type. Since including all three coal types would
not provide additional information, only low sulfur western was considered.
Since all catalysts considered in this study are assumed to be resistant to
SOX poisoning, low sulfur coal was chosen since it had the highest NOX
emissions and flue gas flow rate. Therefore, use of this coal provides a
"worst case" analysis.
SIP control levels are not considered since in many cases no control
is required. On cases that require some control the level can be easily
attained through use of combustion modifications. In no instance is FGT
required to meet the average SIP levels, except possibly in California. Los
Angeles hourly maximum N0x concentration occassionally exceeds the state
standard by a factor of 3.l8 Because of the topographic characteristics of
the area and its high concentration of mobile sources (automobiles which
also have mileage requirements to meet), strict legislation for stationary
sources has been proposed that would require NO FGT on boilers.
4-6
-------
4.1.2 Control Costs for Coal-Fired Boilers
Equipment costs are determined from equipment sizing calculations which
are in turn determined from material balances. Material balances for coal-
fired boilers are contained in Appendix 3. These and the factors discussed
in Section 4.1.1 were used to compute the various cost values. The cost com-
ponents are broken 'down into individual capital and operating costs in
Appendices 6 and 7, respectively.
The annualized costs for each of the standard boilers are summarized
in Tables 4.1.2-1 through 4.1.2-4. The costs are also presented as a percent
of cost of the uncontrolled boiler. These data are also plotted in Figures
4.1.2-1 through 4.1.2-4 to show the sensitivity of the process costs to
control level. The slight nonlinearities are a result of the cost of
catalyst which increases while several equipment costs and labor costs are
constant for all control levels.
The cost effectiveness of the various applications can be assessed
by dividing the annual cost by the annual NOx removal. The results of
this calculation are presented in Table 4.1.2-5. As can be seen, the
effectiveness of the parallel flow system on the largest boiler indicates
an optimum at 70% while the smallest boiler exhibits an optimum at 90%.
There are three primary cost components that determine these results:
equipment costs per unit size, catalyst costs, and labor costs. On the
smaller boilers the equipment required is obviously smaller and its costs
per unit size is greater due to the lack of economy of scale. This is shown
directly by the improving cost effectiveness with boiler size (less $
required per kg NOX removed). Now, on the smaller boilers, the catalyst
costs are not as dominant as the labor costs. This is due to less catalyst
required by the smaller boilers, yet the operating and maintenance labor
requirements for the NOX systems on smaller boilers are comparable to those
of larger boilers (at least within the size range of the standard boilers).
What this means is that on a small NOX system where maintenance and operation
4-7
-------
I
CO
TABLE 4.1.2-1. COSTS OF NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
X
System
Standard boilers
Heat input
MW (MBtu/hr)
8.8 (30)
Low
Sulfur
Western
Coal
Type
Package
Watertube
Underfeed
Stoker
Annual costs
Type and Control
level efficiency
of controlf (%) $/J/S ($/MBtu/hr)
PF SCR 80 0.0134 (3920)
Intermediate
Impacts*
% increase
in costs over
uncontrolled
boiler
10.7
*Based only on Annual Costs
tPF = parallel Flow SCR
-------
TABLE 4.1.2-2. COSTS OF NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
-p-
I
Heat input
MW (MBtu/hr)
22 (75)
Low
Sulfur
Western
Coal
System
Type
Package
Watertube
Chaingrate
Annual costs
Type and Control
level efficiency
of control^" (%)
PF SCR 90
Stringent
PF SCR 80
Intermediate
PF SCR 70
Moderate
$/J/S
0.00882
0.00769
0.00687
($/MBtu/hr)
(2620)
(2270)
(2030)
Impacts*
% increase
in costs over
uncontrolled
boiler
9.1
7.9
7.1
•''Based only on Annual Costs.
tPF = Parallel Flow SCR
-------
TABLE 4.1.2-3. COSTS OF NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Heat input
System
MW (MBtu/hr) Type
44 (150)
Low
Sulfur
Western
Coal
*Based only on
tpF = Parallel
i
o
Heat input
Field-
Erected
Watertube
Spreader
Stoker
Annual Costs
Flow SCR
TABLE 4.1.2-4.
System
MW (MBtu/hr) Type
58.6 (200)
Low
Sulfur
Western
Coal
Field-
Erected
Watertube
Pulverized
Coal
Type and
level
of control"'"
PF SCR
Intermediate
COSTS OF NOX
Type and
level
of control^
PF SCR
Stringent
PF SCR
Moderate
Impacts*
Annual costs % increase
Control in costs over
efficiency uncontrolled
(%) $/J/S ($/MBtu/hr) boiler
80 0.00567 (1680) 7.2
FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Impacts*
Annual costs % increase
Control in costs over
efficiency uncontrolled
(%) $/J/S ($/MBtu/hr) boiler
90 0.00599 (1760) 7.9
70 0.00433 (1270) 5.7
*Based only on Annual Costs
TPF
Parallel Flow SCR
-------
175
150
125
Annual Cost
($1000)
100
75
Parallel Flow SCR
70
80
Percent NOV Control
90
16
15
14
13
12
Percent
Increase
11 Over
Uncontrolled
Boiler
10
Figure 4.1.2-1. Annual cost of NOX control systems applied to
underfeed stoker standard boiler.
4-11
-------
250
200
150
Annual Cost
($1000)
100
50
70 80
Percent NOX Control
90
11
10
9
8
7 Percent
Increase
Over
6 Uncontrolled
Boiler
Figure 4.1.2-2. Annual cost of NOX control systems applied to
chaingrate standard boiler.
4-12
-------
400
350
300
250
Annual Cost
($1000) 200
150
100
50
11
10
Percent
6 Increase
Over
Uncontrolled
5 Boner
70 80
Percent NCL Control
90
Figure 4.1.2-3. Annual cost of NOX control systems applied to
spreader stoker standard boiler.
4-13
-------
500
450
400
350
300
Annual Cost
($1000) 250
200
200
150
100
11
10
Percent
Increase
6 Over
Uncontrolled
Boiler
5
70 80
Percent NOX Control
90
Figure 4.1.2-4. Annual cost of NOX control systems applied to
pulverized coal standard boiler.
4-14
-------
TABLE 4.1.2-5. COST EFFECTIVENESS OF NO . FGT ($/kg NO removed)
X X
Boiler type
Underfeed
Chaingrate
Spreader Stoker
Pulverized Coal
Control type
PF
PF
PF
PF
90
2.57
1.56
1.16
0.874
Percent NOX control
80
2.64
1.53
1.13
0.836
70
2.75
1.56
1.14
0.813
personnel are going to be needed regardless, the NOX system might as well be
a little larger to remove additional NOX. This trend is just the opposite
for larger boilers where catalyst costs become dominant. It requires larger
amounts of expensive catalyst to remove the additional NOX and thus increases
the cost substantially and decreases the cost effectiveness of the system at
higher removal levels. The data presented in Table 4.1.2-5 is plotted in
Figure 4.1.2-5.
The cost of applying NO FGT to modified or reconstructed facilities
X
is likely to be higher than the cost for applications to new facilities.
All of the equipment for new installations will be necessary for retrofit
installations, however, additional equipment may also be necessary. Specific
costs for retrofit applications were not calculated here, but can be esti-
mated. In a study for the Japanese Environment Agency, five process vendors
prepared economic analyses for three applications:19
4-15
-------
3.0
i
M
X
£ 2.0
O
u
1.0
70
80
Percent NO Control
Underfeed Stoker
Chaingrate
Spreader Stoker
Pulverized Coal
90
Figure 4.1.2-5.
Cost effectiveness of parallel flow SCR NOX control systems
applied to the coal-fired standard boilers.
4-16
-------
1) new boiler,
2) retrofit for gas taken upstream of the air preheater requiring
additional ductwork and a fan, and
3) retrofit for gas taken downstream of the ESP including gas/gas
heat exchanger, heater, fan.
The relative costs for each system treating 40,000 Nm3/hr of flue gas is
shown in Table 4.1.2-6.
TABLE 4.1.2-6. RELATIVE COSTS OF RETROFIT SCR SYSTEMS
System Relative annualized cost
1 1.00
2 1.23
3 2.20
These results indicate that SCR applications to modified or recon-
structed facilities can cost from 25 to 120 percent more than applications
to new boilers.
A.1.3 Costs To Control Oil-Fired Boilers
The cost calculations presented in this section are based on material
balances performed for each case considered. The material balances are
presented in Appendix 4. These are used to size the equipment which is
subsequently costed. The costing techniques are described in principle in
Section 4.1.1.
4-17
-------
The cost components are broken down into individual capital and
operating costs in Appendices 6 and 7. The annualized costs are summarized
in Tables 4.1.3-1 and 4.1.3-2. These tables show the annual cost as a
percentage of the uncontrolled boiler cost. These values are plotted as a
function of control level in Figures 4.1.3-1 and' 4.1.3-4.
The parallel flow system shows slightly more sensitivity to control
level. This is most likely due to the catalyst which is the most significant
cost component. The parallel flow catalyst is about as expensive as the
moving bed catalyst per cubic meter, but has a lower space velocity. This
causes the parallel flow systems to have a higher catalyst cost component.
The nonlinearity is due to this fact combined with the fact that the cost/
unit of equipment increases as size ('i.e., control level) decreases.
The cost effectiveness is also determined in Table 4.1.3-3 where the
cost per kg of NO removed is presented. The cost for the distillate oil-
X
fired boiler is very high due primarily to poor economy of scale since the
boiler is small. These costs are plotted in Figures 4.1.3-5 and 4.1.3-6.
The cost differences between the two systems applied to the residual
oil-fired boilers are not significant within the accuracy of this cost
estimate (+50 percent). The table indicates that the cost effectiveness of
the moving bed system increases as removal level increases. This seems to
be due to the effect of a greater economy of scale with the larger systems.
The reactor is smaller than the parallel flow so the catalyst cost is not as
dominant a cost component as the labor cost component. There are several dif-
ferent types of parallel flow type reactors. Some of them consume more energy
and cost more than the moving bed reactors, as described here. However, reac-
tors using thin-wall honeycomb or plate catalysts developed recently in
Japan are reported to require less energy and cost less than moving bed reac-
tors, and have been used for virtually all of the new SCR plants for dirty or
semi-dirty gases.
4-18
-------
TABLE 4.1.3-1. COSTS OF NOX FGT CONTROL TECHNIQUES FOR OIL-FIRED BOILERS
-p-
I
System
Standard boilers
Heat input
MWt (MBtu/hr)
4.4 (15)
Distillate
Oil
44 (150)
Type and Control
Annual Costs
level ^ efficiency
Type
Package
Firetube
Scotch
Package
Water tube-
of control (%)
FPB SCR 90
Stringent
FPB SCR 70
Moderate
FPB SCR 90
Stringent
FPB SCR 70
Moderate
$/J/S
0.0154
0.0145
0.0040
0.0031
($ /MBtu/hr)
(4500)
(4240)
(1170)
(915)
Impacts*
% increase
in costs over
uncontrolled
boiler
12.1
11.4
7.5
5.6
*Based only on Annual Costs
tFPB = Fixed Packed Bed
-------
TABLE 4.1.3-2. COSTS OF NOX FGT CONTROL TECHNIQUES FOR OIL-FIRED BOILERS
System
Standard boilers Type and
Heat input level
MW (MBtu/hr) Type of control^"
8.8 30 Package PF SCR
Watertube Stringent
Residual
Fuel Oil FF SCR
Moderate
MB SCR
Stringent
j>.
I MB SCR
o Moderate
44 (150) Package PF SCR
Watertube Stringent
Residual
Fuel Oil PF SCR
Moderate
MB SCR
Stringent
MB SCR
Moderate
Control
efficiency
90
70
90
70
90
70
90
70
Annual
$/J/S
0.0123
0.0110
0.0148
0.0137
0.00502
0.00408
0.00457
0.00377
costs
($ /MBtu/hr)
(3600)
(3200)
(4330)
(4010)
(1490)
(1210)
(1360)
(1120)
Impacts*
% increase
in costs over
uncontrolled
boiler
14
12
16
15
7.0
5.7
6.4
5.3
Based only on Annual Costs
tPF = Parallel Flow
MB = Moving Bed
-------
Annual Cost
($1000)
100
90
80
70
60
50
Fixed Packed Bed SCR
18
17
16
15
14
Percent
Increase
13 Over
Uncontrolled
Boiler
12
11
10
70
80
90
Percent NOV Control
Figure 4.1.3-1. Annual cost of NOX control system applied
to 4.4 MW distillate oil-fired standard
boiler.
4-21
-------
300
250
200
Annual Cost
($1000)
150
100
50
Fixed Packed Bed SCR
137,000
70 80
Percent NCy Control
176,000
90
11
10
Percent
Increase
6 Over
Uncontrolled
Boiler
5
Figure 4.1.3-2. Annual cost of NOX control system applied to
44 MW distillate oil-fired standard boiler.
4-22
-------
150
100
Annual Cost
($1000)
50
Moving Bed SCR
Parallel Flow SCR
70 80
Percent NO* Control
90
22
20
18
16
14
Percent
Increase
12 Over
Uncontrolled
Boiler
10
Figure 4.1.3-3. Annual cost of NOX control systems applied to
8.8 MW residual oil-fired standard boiler.
4-23
-------
350
300
250
200
Annual Cost
($1000)
150
100
50
Parallel FlowSCR.
__ •
Moving Bed SCR
11
10
9
Percent
Increase
6 Over
Uncontrolled
Boiler
5
70
80
90
Percent NOX Control
Figure 4.1.3-4. Annual cost of NOX control systems applied to
44 MW residual oil-fired standard boiler.
4-24
-------
TABLE 4.1.3-3. COST EFFECTIVENESS OF NOX FGT
I
K5
Fuel
Distillate Oil
Distillate Oil
Residual Oil
Residual Oil
Boiler size,
MWt
4.4
44
8.8
44
Control type*
FPB
FPB
PF
MB
PF
MB
Percent NOx control ($/kg NOx removed)
90
17.6
3.8
5.7
6.9
1.89
1.72
80
19.0
3.8
6.0
7.3
1.85
1.75
70
21.4
3,8
6.6
8.2
1.97
1.84
* FPB = Fixed Packed Bed SCR
PF = Parallel Flow SCR
MB = Moving Bed SCR
-------
I
o
ox
22
20
18
16
14
12
10
4.4 MWt Boiler
Fixed Packed Bed SCR
44 MWt Boiler
Fixed Packed Bed SCR
70
80
90
Percent NOX Control
Figure 4.1.3-5.
Cost effectiveness of FGT systems applied
to distillate oil-fired boilers.
4-26
-------
10
ox
z
8.8 MWt Boiler
Moving Bed SCR
8.8 MWt Boiler
Parallel Flow SCR
44 MWt Boiler
Parallel Flow SCR
44 MWt Boiler
Moving Bed SCR
70
80
90
Percent
Control
Figure 4.1.3-6.
Cost effectiveness of FGT systems applied
to residual oil-fired boilers.
4-27
-------
The overall conclusion is the systems applied to the residual oil-fired
boilers to not appear to be cost sensitive with respect to control level.
The small distillate oil-fired boiler appears to be very sensitive since
higher control can be achieved with only slightly higher annualized costs.
The cost of applying NOX FGT to modified or reconstructed facilities is
likely to be higher than the cost for applications to new facilities. All
of the equipment for new installations will be necessary for retrofit
installations, however, additional equipment may also be necessary. Specific
costs for retrofit applications were not calculated here, but can be esti-
mated. In a study for the Japanese Environment Agency, five process vendors
prepared economic analyses for three applications:19
1) new boiler,
2) retrofit for gas taken upstream of the air preheater requiring
additional ductwork and a fan, and
3) retrofit for gas taken downstream of the ESP including gas/gas
heat exchanger, heater, fan.
The relative costs for each system treating 40,000 Nm /hr of flue gas are
shown in Table 4.1.3-4.
TABLE 4.1.3-4. RELATIVE COSTS OF RETROFIT SCR SYSTEMS
System Relative annualized cost
1 1.00
2 1.23
3 2.20
These results indicate that SCR applications to modified or recon-
structed facilities can cost from 25 to 120 percent more than applications
to new boilers.
4-28
-------
4.1.4 Control Costs for Natural Gas-Fired Boilers
This section presents cost calculations for a FGT system applied to
the natural gas-fired standard boiler. The calculations are based on
material balances contained in Appendix 5. The costing techniques have been
described in Section 4.1.2.
The cost components are broken down into individual capital and operat-
ing costs in Appendices 6 and 7. Both total annualized costs and costs as a
percentage of the uncontrolled boiler cost are shown in Table 4.1.4-1. The
data presented are also plotted in Figures 4.1.4-1 and 4.1.4-2 to show the
sensitivity to control level. The costs are fairly linear with control level
indicating only a slight sensitivity.
The cost effectiveness of NOX control on natural gas-fired boilers is
determined in Table 4.1.4-2 and these values are plotted in Figure 4.1.4-3.
The smaller systems are not as cost effective as the larger systems since the
cost per unit size of equipment is less for larger systems. The cost effec-
tiveness of the small system is more sensitive to control level due to the
influence of labor costs, which are constant for all control levels.
The cost of applying NOX FGT to modified or reconstructed facilities
is likely to be higher than the cost for applications to new facilities.
All of the equipment for new installations will be necessary for retrofit
installations, however, additional equipment may also be necessary.
Specific costs for retrofit applications were not calculated here, but
can be estimated. In a study for the Japanese Environment Agency, five
1 9
process vendors prepared economic analyses for three spplications:
1) new boiler,
2) retrofit for gas taken upstream of the air preheater requiring
additional ductwork and a fan, and
4-29
-------
TABLE 4.1.4-1. COSTS OF N0x FGT CONTROL TECHNIQUES FOR NATURAL GAS-FIRED BOILERS
System
Standard boilers
Heat input
MWfc (MBtu/hr)
4.4 (15)
Natural
Gas
j>
L
to
O
44 (150)
Natural
Gas
Type
Package
Firetube
Scotch
Package
Water tube
Type and
level
of control^"
FPB SCR
Stringent
FPB SCR
Moderate
FPP SCR
Stringent
FPB SCR
Moderate
Control
efficiency
(%)
90
70
90
70
Annual costs
$/J/S
0.0154
0.0146
0.0040
0.0029
($/MBtu/hr)
(4510)
(4290)
(1160)
(863)
Impacts*
% increase
in costs over
uncontrolled
boiler
13.6
13.0
7.5
5.6
*Based only on Actual Costs
tFPB = Fixed Packed Bed
-------
Annual Cost
($1000)
90
80
70
60
50
40
Fixed Packed
Bed SCR
18
17
16
15
14
13
Percent
Increase
Over
Uncontrolled
Boiler
12
11
10
70
80
90
Percent NOX Control
Figure 4.1.4-1. Annual cost of NOX control system applied to
4.4 MW natural gas-fired standard boiler.
4-31
-------
250
200
150
Annual Cost
($1000)
100
50
175,000
129,000
70
Fixed Packed Bed SCR
80 90
Percent NOX Control
11
10
Percent
Increase
6 Over
Uncontrolled
Boiler
5
Figure 4.1.4-2. Annual cost of NOx control system applied to
44 MW natural gas-fired standard boiler.
4-32
-------
TABLE 4.1.4-2. COST EFFECTIVENESS OF NOX FGT
$/kg NOX removed
Fuel
Boiler size,
t
Percent NOX control
Control type
90
80
70
P-
i
LO
to
Natural Gas
Natural Gas
4.4
44
Fixed Packed Bed
Fixed Packed Bed
16.0
3.4
17.5
3.3
19.7
3.2
-------
o
z
20
18
16
14
12
10
4.4 MWt Boiler
'Fixed Packed Bed SCR
44 MWt Boiler
"Fixed Packed Bed
70
80
90
Percent NOy Control
Figure 4.1.4-3.
Cost effectiveness of FGT systems applied
to natural gas-fired boilers.
4-34
-------
3) retrofit for gas taken downstream of the air preheater including
gas/gas heat exchanger, heater, fan.
The relative costs for each system treating 40,000 Nm3/hr of flue gas is
shown in Table 4.1.4-3.
TABLE 4.1.4-3. RELATIVE COSTS OF RETROFIT SCR SYSTEMS
System Relative annualized cost
1 1.00
2 1.23
3 2.20
These results indicate that SCR applications to modified or reconstructed
facilities can cost from 25 to 120 percent more than applications to new
boilers.
4.1.5 Summary
In all cases the catalyst cost is a significant capital cost. Other
significant capital cost components are labor, fan motor drive, and NHa
storage tanks. The most significant operating cost component in all cases
was labor. The smaller systems are more significantly affected by this fact
than are the larger systems. As a result, the costs for small systems are
high, not only because they lack economy of scale, but due to labor con-
siderations as well. As a result the size of the unit has a greater effect
on costs than does control level.
4-35
-------
This dramatic effect is most readily observed in the cost effectiveness
numbers. The systems exhibited an order of magnitude larger cost/kg of NOX
when applied to the smallest systems.
4.2 NOx/SOy SYSTEM
4.2.1 Introduction
This section considers the costs of applying the NOy/SOy system selected
in Section 3 to two coal-fired and one oil-fired boilers. The costing tech-
niques are the same as used with the N0x-only processes and will not be
repeated here. The equipment items are more numerous due to the higher
number of process operations associated with the process. These items are
listed in Table 4.2.1-1.
TABLE 4.2.1-1. PURCHASED EQUIPMENT FOR NOX FGT SYSTEMS
NOX/SOX Parallel Passage
Reactors (2)
Catalyst
Fan Motor Drive
NH3 Storage Tank
NH3 Transfer Pump
NHs Vaporizer
Naphtha Reformer
H2SOit Plant
Compressor/Gasholder
With the coal-fired boilers, both high and low sulfur coals were
analyzed. However, only one set of control levels are considered (80 percent
NOX, 85 percent SOX) and therefore, it is not possible to present costs as
a function of control level as is done in the N0x-only section. Instead, the
costs are plotted against flue gas flow rate to show the effect of unit size
on cost. The results for the residual oil-fired boiler are presented in
4-36
-------
tabular form, but not plotted since only one boiler and control level are
considered.
4.2.2 Control Costs for Coal-Fired Boilers
The equipment listed in the previous table is sized based on material
balances performed for each case. These balances are presented in Appendix
3. Detailed breakdowns of both capital and operating costs are presented in
Appendices 6 and 7, respectively.
The annualized costs for the standard boilers considered are presented
in Table 4.2.2-1 and plotted in Figure 4.2.2-1. The costs are significantly
higher than those for the N0x-only processes because the additional require-
ment of SOa removal necessitates the use of small processing units for H2
production and SOz workup. In a real world situation where several indus-
trial boilers operate at a single location, it will be possible to reduce
costs by having large, central units for Ha production and SOa workup. This
option is not considered here since the cost impact is a function of the
total number of boilers serviced by the central facilities and this is
entirely site specific.
The cost to retrofit such a process can be calculated from the data
presented in Section 4.1.2. Depending on the modifications required by the
retrofit, the additional cost will be increased by an amount equivalent to
25 to 120 percent of the cost of an average N0x-only system. The cost of
special equipment necessary for SOa processing is not affected by a retrofit
application.
4.2.3 Control Costs .for the Oil-Fired Boiler
The equipment items necessary to-treat flue gas from the residual oil-
fired boiler are the same as for the coal-fired boilers. The annualized cost
of the dry NOX/SOX process applied to the residual oil-fired standard boiler
4-37
-------
TABLE 4.2.2-1. COSTS OF NOX/SOX FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
.p-
1
CO
CO
System
Control
Heat input efficiency
MW (MBtu/hr) Type (%) Coal"^
8.8 (30) Package 90% NOX HSE
Watertube 85% SOX
Underfeed LSW
Stoker
58.6 (200) Field 90% NOx HSE
Erected 85% SOX
Watertube LSW
Pulverized
Coal
Annual Costs
$/J/S
0.0811
0.0527
0.0309
0.0153
($/MBtu/hr)
(23,690)
(15,400)
(9,025)
(4,468)
Impacts*
% increase
in costs over
uncontrolled
boiler
75
49
42
21
*Based only on Annual Costs
tHSE = High sulfur eastern coal (3.5% S)
LSW = Low sulfur western coal (0.6% S)
-------
2000 _j
-P-
I
1500 _
1000 -
500 _
High Sulfur Coal
Low Sulfur Coal
1000 2000
Flue Gas Flow (m3/min)
3000
4.2.2-1. Annual cost of parallel flow SCR NOX/SOX FGT for coal-fired boilers
-------
is shown in Table 4.2.3-1. Detailed capital and annual costs are presented
in detail for this case in the appendices. Retrofit considerations for this
case are similar to those for the coal cases. The increased cost of a
retrofit will be increased by the same dollar amount as the cost increase
incurred when a similarly sized N0x-only unit is retrofitted.
4-40
-------
TABLE 4.2.3-1. COSTS OF THE DRY N0x/S0x CONTROL TECHNIQUE FOR THE RESIDUAL OIL-FIRED BOILER
Impact*
% increase
Standard boiler Annual costs in costs over
Heat input uncontrolled
MW (MBtu/hr) Type Control Level $/J/S ($/MBtu/hr) boiler
44 (150) Package 90% NOX 0.0249 (7280) 43
Watertube 85% SOX
*Based only on Annual Costs
-------
REFERENCES
1. PEDCo, Task 7, Section 3 Report. "Emission Control System Economics".
29 June 1978.
2. PEDCo, Task 7, Section 3 Draft Report, "Emission Control System
Economics". 29 August 1978.
3. Winkler, P., Chemico Air Pollution Control Corp. Telephone conversation
with Gary Jones. 25 October 1978.
4. Maxwell, J.D., TVA. Telephone conversation with Gary Jones. 15
January 1979.
5. Ando, J. "NOX Abatement from Stationary Sources in Japan". EPA draft
report in preparation. October 1978. p. 4-26.
6. Guthrie, K.M. Process Plant Estimating and Control. Craftsman, 1974.
pp. 150-154.
7. Ibid. , p. 174.
8. Woods, D.R. Financial Decision Making in the Process Industry.
Prentice-Hall. 1975. p. 301.
9. Guthrie, op'.oit. , pp. 349-350.
10. Ibid. , pp. 159, 163.
11. Ibid. , pp. 144-145.
12. Peters, M.S. and K.D. Timmerhaus. Plant Design and Economics for
Chemical Engineers. McGraw-Hill. Second Edition. 1968. p. 505.
13. Woods, op.ait., p. 296.
14. Maxwell, J.D., et al. "Preliminary Economic Analysis of NOx Removal
Processes for Utility Application". EPA draft report. November 1978.
pp. 201-202.
15. Chemical Engineering. "Economics Indicators". McGraw-Hill. July
1978. p. 7.
4-42
-------
16. Ando, 1978, op.oit., p. 4-26.
17. Peters, op.eit., p. 216.
18. SCE, EPRI, et al. "An Assessment of NOX Control Technology for Oil-
and Gas-Fired Utility Boilers". October 1978. p. 5.
19. Ando, 1978, op.cit. , pp. 3-67-3-82.
20. Ando, J., Review Comments on Draft Report, July 12, 1979.
4-43
-------
SECTION 5
ENERGY IMPACT
5.1 NOX-ONLY SYSTEMS
5.1.1 Introduction
The three types of control systems selected in Section III for further
comparison are analyzed with respect to energy requirements. All of the
control systems are basically similar in principle and differ mainly in the
design parameters. There are also a few differences in equipment require-
ments. Energy consumption steps considered in this analysis are listed in
Table 5.1.1-1 for each of the control systems considered. It was assumed
that flue gas could be taken from the boiler between the economizer and air
heater at a temperature of 375°C. This eliminates the need for flue gas
heating and heat exchange equipment. Since the hot flue gas is returned to
the boiler upstream of the air heater, there is no loss in boiler efficiency.
Energy consumption was calculated using the design information and
standard engineering principles. Design information from a variety of process
developers was compared and used to generate a range of values or specific
values. A range of values was determined for design parameters which changed
with control level. Specific values for analysis were chosen from this
range based on the level of control being considered, e.g. for 70% control a
value at the lower end of the range was used. Design data used in this analy-
sis is presented in Table 5.1.1-2.
5-1
-------
TABLE 5.1.1-1. AREAS OF ENERGY CONSUMPTION IN NOX FGT SYSTEMS
1 ,2
NO FGT system
Energy consumption step
(equipment)
Type of energy
consumed
Parallel Flow SCR
Moving Bed SCR
Fixed Packed
Bed SCR
Reactor Draft Loss (Fan) Electrical
Liquid NHs Transfer (Pump) Electrical
NHs Vaporization (Vaporizer) Steam
NHs Dilution Steam
Reactor Draft Loss (Fan) Electrical
Liquid NH3 Transfer (Pump) Electrical
Catalyst Screening & Transfer (Elevator) Electrical
Baghouse Draft Loss (Blower) Electrical
NHs Vaporization (Vaporizer) Steam
NHs Dilution Steam
Reactor Draft (Fan) Electrical
Liquid NHs Transfer (Pump) Electrical
NHs Vaporization (Vaporizer) Steam
NHs Dilution Steam
Soot Blowing-Distillate Oil Boiler Only Steam
TABLE 5.1.1-2. RANGE OF DESIGN PARAMETERS USED FOR
ENERGY IMPACT CALCULATIONS
1,2,3
Parameter
Space velocity
NH3:NO mole ratio
X
Dilution ratio (moles steam/mole
Dilution steam pressure
Flue gas temperature
Pressure drop
Catalyst type
Range
Parallel flow
3000-5000
0.7-1.0
NH3) 5:1
30 psig
375°C
80-160 mmH20
Square
honeycomb
or specific
Moving bed
6000-10000
0.7-1.0
5:1
30 psig
375°C
40-80 mmH20
Ring
values used
Fixed packed bed
6000-10000
0.7-1.0
5:1
30 psig
375°C
VL25 mmH20
Spherical
pellet
Void fraction of packed
catalyst particles
0.67-0.7
0.52
0.26
5-2
-------
Steam was chosen as the NHs dilution gas because of its ease of
application and safety considerations. Air, at 20:1 air:NHs mole ratio, can
also b.e used as an NHs diluent. Its use requires a compressor or blower and
a motor which are high maintenance items. Also, at dilution ratios less than
20:1 there is an explosion hazard. The optimum choice would ordinarily be
made by comparing the operating costs of steam use versus the capital charges
of the air handling equipment plus the operating costs of electricity. This
optimization is beyond the scope of this study and is site specific.
The analyses conducted in this study assumed that the boilers were
operated constantly at full load and, therefore, had constant flue gas temper-
atures. However, it is possible that the boiler may experience large and
frequent load swings which result in a variable flue gas temperature. FGT
systems in this service will require flue gas heating in order to maintain
sufficiently high temperatures. Temperature control can be accomplished by
either a heater or a slipstream around the economizer. The heater will
effectively decouple the FGT system from the boiler and does not require flow
control of a flue gas slipstream. The economizer bypass will not derate the
boiler since it will only be required during low load situations. Energy
usage calculations were not made for either of these approaches since the
amount of heating necessary is likely to be different for each boiler applica-
tion.
5.1.2 Energy Impact of Controls for Coal-Fired Boilers
This subsection presents the results of calculations on the energy
requirements of the candidate control systems applied to the standard boilers.
One simplification was made in order to reduce the number of cases necessary
for consideration and that is that only one coal was analyzed for each boiler.
The justification for this simplification is presented below.
The result of the energy impact analyses indicate that the most signi-
ficant energy consumption occurs in the fan required to overcome the reactor
pressure drop and NH3 dilution by steam. Coal sulfur content does not
5-3
-------
significantly affect the fan requirements which are a function of flue gas
flow rate and control level. W.3 dilution steam is affected; however, energy
consumption of this step is approximately a third of the fan requirement.
This is illustrated in an example calculation in Table 5.1.2-1. As can be
seen the sulfur content of the coal does not significantly affect the total
energy requirements especially when compared to the effect of control level.
The low sulfur coal was used for the analyses since the NO emissions
were somewhat higher and, therefore, energy usage for the other coals will
not exceed those presented here.
Also, SIP control levels were not considered since in cases where
control is required, it can be achieved through use of combustion modifica-
tions. The typical SIP control levels are shown in Table 5.1.2-2.
TABLE 5.1.2-2. SIP CONTROL LEVELS4
3
0
0
.5%
.9%
.6%
rue
s
s
s
;_l_
Coal
Coal
Coal
NO emissions
X
0.
0.
0.
lb -ir IP
' iob Btu "ir le
64
55
78
0
0
0
lb
10b Btu
.7
.7
.7
Required
control
efficiency
0
0
10%
Material balances were performed for each of the 7 cases considered for
the coal-fired standard boilers. The results of these calculations appear
in Appendices 3, 4, and 5. These results were used to calculate energy
requirements of the control systems and an example calculation is presented
in Appendix 8.
The results of the energy requirement calculations are presented in
Tables 5.1.2-3 through 5.1.2-6. Each table represents one standard boiler
and all control types and levels are included. It should be noted that the
megawatt values shown for electrical usage are thermal megawatts and not
5-4
-------
Ln
I
Ui
TABLE 5.1.2-1. RELATIVE SIGNIFICANCE OF PARAMETERS CONSIDERED IN ENERGY ANALYSIS
Example: Pulverized Coal Boiler, 90% Control, Parallel Flow SCR
a) Effect of sulfur content
Energy usage (MW thermal)
0.6% S coal
(187.56 Ib N0x/hr)
0.9% S coal
(130.50 Ib NO/hr)
3.5% S coal
(152.46 Ib N0x/hr)
Energy Consumer
b)
Fan 0.91
Liquid NH Pump 0.00373
NH Vaporizer 0.0383
NH Dilution Steam 0.325
Total 1.275
10%
7%
Effect of removal level 90% removal
Total Energy Consumed 1.28
(MW thermal)
36%
n-if f AT- 01
0.88 0.88
0.00373 0.00373
0.0275 0.0325
0.234 0.275
1.145 1.191
70% removal
0.821
->r« o - -
-------
TABLE 5.1.2-3. ENERGY CONSUMPTION FOR NOX FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
System
Standard boiler
Heat input Type Type and level
MW (MBtu/hr) of control
SCR Parallel Flow
58.6 (200) Field Erected, Moderate
Watertube,
Pulverized Coal
SCR Parallel Flow
Stringent
TABLE 5.1.2-4. ENERGY CONSUMPTION FOR NOX
System
Standard boiler Type and level
Heat input Type of control
MW (MBtu/hr)
44 (150) Field Erected, SCR Parallel Flow
Watertube, T ..
' , Intermediate
Spreader Stoker
Control
efficiency Energy
% types
Electrical
70 Steam
Electrical
90 Steam
Energy consumption
Energy consumed % increase
by control device in energy use over
MWt (MBtu/hr) uncontrolled
0.161 (0.549)
0.0797 (0.272) 0.41
0.268 (0.912)
0.108 (0.364) 0.64
boiler
FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Control
efficiency Energy
% types
Electrical
80 Steam
Energy consumption
Energy consumed % increase
by control device in energy use over
MWt (MBtu/hr) uncontrolled
0.126 (0.428)
0.0568 (0.194) 0.41
boiler
-------
TABLE 5.1.2-5. ENERGY CONSUMPTION FOR NOx FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Ln
I
System
Standard Boiler Type & Level
Heat Input Type of
MW (MBtu/hr) Control
SCR Parallel Flow
Moderate
Package
22 (75) Watertube SCR P"allel Flow
Chalngrate Intermediate
SCR Parallel Flow
Stringent
TABLE 5.1.2-6. ENERGY CONSUMPTION FOR NO
System
Standard boiler
Heat input TyPe Type and level
MW (MBtu/hr) of control
8.8 (30) Package SCR Parallel Flow
Watertube T ,
.. , , , Intermediate
Underfeed
Stoker
Control Energy
Efficiency Types
7,
Electrical
70 Steam
Electrical
80 Steam
Electrical
90 Steam
Energy Consumption
Energy Consumed 7, Increase
MWt (MBtu/hr) Uncontrolled Boiler
0.0408 (0.139)
0.0253 (0.0862) 0.30
0.0505 (0.172)
0.0289 (0.0988) 0.36
0.0669 (0.228)
0.0337 (0.115) 0.46
x FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Control
efficiency Energy
% types
Electrical
80 Steam.
Energy consumption
Energy consumed % increase
by control device in energy use over
MWt (MBtu/hr) uncontrolled boiler
0.0185 (0.0630)
0.0113 (0.0387) 0.34
-------
electrical megawatts. The data appearing in these tables is summarized in
Table 5.1.2-7. This data is plotted in Figures 5.1.2-1 through 5.1.2-4.
Each figure represents one standard boiler and shows the effect of removal
level on energy usage for both types of FGT candidate systems: parallel
flow SCR and moving bed SCR. Figure 5.1.2-5 presents energy usage for all
boilers and FGT systems as a percent of the boiler heat input.
In general, energy usage seems to increase with control level in a non-
linear manner. This is also true with regard to energy usage as a percent
of boiler input. Also, it appears that more energy is used per mole of NOX
controlled with the larger boilers. This increase with boiler size is not a
physical phenomena of SCR systems but rather an idiosyncrasy of the reactor
design scheme. In keeping reactor geometry consistent from boiler to boiler,
the linear gas velocity (m/s) decreased slightly with boiler size resulting
in a corresponding slight decrease in pressure drop. However, the pressure
drops of all the standard boilers are in the range of commercial operations
and the differences in energy usage as a percentage of boiler heat input
of the standard boilers are not large.
The effect of this energy usage on economics is examined in Section IV.
Very little work has been done with SCR systems to affect reductions in
energy consumption. Problem solving efforts thus far have been directed
toward improving the reliability of operating units and applying the control
techniques to coal-fired flue gas. It is likely that only after the pro-
cesses have been applied and demonstrated on coal-fired units will the over-
all energy consumption be examined in detail. It should be noted that the
SCR processes are the least energy intensive of all of the FGT control systems
mentioned ir Section II.1'5
There are two areas in which there is a potential for energy savings.
These are control of excess air and NHs dilution. By using only as much
excess air as necessary, the energy required for pressure drop will be
reduced. This has a twofold effect. Not only is the flow through the
-------
TABLE 5.1.2-7. SUMMARY OF ENERGY REQUIREMENTS FOR COAL-FIRED INDUSTRIAL BOILERS
Pulverized coal
Ui
1
VD
Parallel Flow SCR
90% Removal
C0% Removal
70% Removal
Total
thermal kW
376
241
% of boiler
heat input
0.64
0.41
Spreader stoker
Total
thermal kW
183
% of boiler
heat input
0.41
Chaingrate
Total
thermal kW
101
80
66
% of boiler
heat input
0.46
0.36
0.30
Underfeed stoker
Total
thermal kW
30
% of boiler
heat input
0.34
-------
500 —i
400 —
0)
f.
OJ
bO
cd
CO
oo
n
0)
300 —
200 —
100 —
Parallel
Flow SCR
50
\ I I
60 70 80
Percent NOX Removal
I
90
100
Figure 5.1.2-1. Energy usage of NO control systems for
pulverized coal standard boiler.
5-10
-------
500 -
400-^
300 -
0)
60
cd
en
13
60
200 H
100 -
•Parallel
Flow SCR
50
] I I
60 70 80
Percent NOX Control
I
90
100
Figure 5.1.2-2,
Energy usage of NOX control systems for
spreader stoker standard boiler.
5-11
-------
500 -|
400
300 H
CJ
t>o
to
CO
60
200 H
100 H
Parallel Flow S(
50
I
60
I
70
80
I
90
100
Percent NOX Control
Figure 5.1.2-3. Energy usage of NOX control systems for
chaingrate standard boiler.
5-12
-------
a)
4-1
0)
60
n)
W
(-1
01
c
500 -i
400 ~
300 ~
200 -
100 —
Parallel Flow SCR
50
1
60
70
80
90
I
100
Percent NOV Control
Figure 5.1.2-4. Energy usage of NOX control systems
for underfeed stoker standard boiler.
5-13
-------
3
O,
c
cfl
0)
S3
•H
O
PQ
4-1
Pi
(!)
O
M
cu
0.6-
0.5-
0.4-
0.3-
Pulverized Coal
Spreader Stoker
Chalngrate
Underfeed Stoker
Parallel
Flow
SCR
-------
reactor reduced, but the required reactor volume itself is reduced by lower
flue gas flow rates. It is likely that a boiler equipped with combustion
modifications will utilize low excess air for NO control. Energy consump-
tion by NHs dilution might be reduced by using air instead of steam at a
specific site. Use of air is less safe since some air:NHs mixtures can be
explosive.
The energy impact of FGT controls applied to modified or reconstructed
facilities (retrofit application) will be the same or greater than that for
new facilities. If flue gas can be taken from the economizer of the existing
boiler at 350-400°C and returned upstream of any existing heat exchange
equipment, then there will be no additional energy impact.
If the flue gas is only available at a lower temperature (<350-400°C)
then a heater will be required. The energy impact of the heater will depend
on the temperature of the flue gas. If the temperature is that of the out-
let gas of the standard boilers (approximately 180°C), calculations indicate
that energy requirement would be more than tripled even if heat exchange
equipment is used to recover 85% of the energy supplied by the heater. The
heater will probably be oil-fired for ease of control.
These results indicate that, on retrofit applications, there is a
considerable energy incentive to obtain the flue gas at the necessary
reaction temperature in order to avoid flue gas heating. Other energy
impacts would be the same as those for new facilities.
5.1.3 Energy Impact of Controls for Oil-Fired Boilers
In this subsection, the results of energy impact calculations for the
.candidate FGT systems as applied to the standard oil-fired boilers are pre-
sented. The combinations considered are
5-15
-------
Boiler Size, MWt Fuel ' FGT System
8.8, 44 Residual Oil Parallel Flow SCR
8.8, 44 Residual Oil Moving Bed SCR
4.4, 44 Distillate Oil Fixed Packed Bed SCR
Also, two levels of control are considered for each combination.
The first step in performing this energy impact analysis was to calcu-
late general material balances. The result of these balances were used to
determine energy requirements for each process step. Energy consuming steps
and the types of energy used were presented earlier in Table 5.1.1-1. All
calculations are similar to the example case presented in Appendix 8.
The results of these calculations are presented in Tables 5.1.3-1 and
5.1.3-2. The data in Table 5.1.3-1 represents energy consumption for
residual oil-fired boilers. Two candidate systems and two levels of control
are considered. Table 5.1.3-2 shows energy consumption for application of a
fixed packed bed SCR process to the standard boiler firing distillate oil.
All energy values presented are on a thermal basis. Actual electrical
usages have been converted to a heat input basis by assuming 10,000 Btu/hr
per kW.
The data appearing in Tables 5.1.3-1 and 5.1.3-2 is summarized in Table
5.1.3-3 and is plotted in Figures 5.1.3-1 through 5.1.3-4. The first two
figures show thermal energy usage as a function of NOx control for all fuels
and control systems. The next two figures illustrate energy usage as a
percent of boiler heat input for all cases.
Energy usage increases in a nonlinear manner with control level. The
energy usage as a percent of boiler input is also nonlinear. On this basis
the fixed packed bed SCR appears to be the most energy intensive and the
moving bed SCR the least. It is difficult to draw any definite conclusions
5-16
-------
TABLE 5.1.3-1. ENERGY CONSUMPTION FOR NOX FGT CONTROL TECHNIQUES FOR RESIDUAL OIL-FIRED BOILERS
I
M
-^J
System
Standard boiler
Heat input Type Type and level
MWt (MBtu/hr) of control
8.8 (30) Package SCR Parallel Flow
Watertube Moderate
SCR Parallel Flow
Stringent
SCR Moving Bed
Moderate
SCR Moving Bed
Stringent
44 (150) Package SCR Parallel Flow
Watertube Moderate
SCR Parallel Flow
Stringent
SCR Moving Bed
Moderate
SCR Moving Bed
Stringent
Energy consumption
Control
efficiency Energy
% types
70 Electrical
Steam
90 Electrical
Steam
70 Electrical
Steam
90 Electrical
Steam
70 Electrical
Steam
90 Electrical
Steam
70 Electrical
Steam
90 Electrical
Steam
Energy consumed % increase
by control device in energy use over
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
MWt
.011
.0071
.018
.0095
.0094
.0071
.014
.0095
.0813
.0253
.134
.0337
.0570
.0253
.0918
.0337
(MBtu/hr) uncontrolled boiler
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
(0.
0367)
0242)
0597)
0323)
0322)
0242)
0462)
0323)
277)
0864)
458)
115)
195)
0864)
314)
115)
0.20
0.31
0.19
0.26
0.24
0.38
0.19
0.29
-------
TABLE 5.1.3-2. ENERGY CONSUMPTION FOR NOX FGT CONTROL TECHNIQUES FOR DISTILLATE OIL-FIRED BOILERS
System
Standard boiler
Heat input Type Type and level
MWt (MBtn/yr) of control
4.4 (15) Package SCR Fixed Packed Bed
Firctube Moderate
Scotch
SCR Fixed Packed Bed
Stringent
4/i (150) Package SCR Fixed Packed Bed
Watertube Moderate
SCR Fixed Packed Bed
Stringent
1
Energy consumption
Control
efficiency Energy
7, types
70 Electrical
Steam
90 Electrical
Steam
70 Electrical
Steam
90 Electrical
Steam
Energy consumed % increase'
by control device in energy use over
MWt
0.00994
0.00697
0.0158
0.00888
n.200
0.0734
0.12J
0.0706
(MBtu/yr) uncontrolled hoJJer
(0.0339) 0.38
(0.0238)
(0.0539) 0.56
(0.0302)
(0.682) 0.62
(0.251)
(0.414) 0.44
(0.241)
oo
-------
Ul
I
TABLE 5.1.3-3. SUMMARY OF ENERGY REQUIREMENTS FOR OIL-FIRED INDUSTRIAL BOILERS
Parallel Flow SCR
907. Removal
70% Removal
Moving Bed SCR
90% Removal
70% Removal
8.8 MWt
Total
thermal
kW
27
18
23
17
Residual oil
% of
boiler heat
input
0.31
0.20
0.26
0.19
44 MWt
Total
thermal
kW
168
107
126
82
Residual oil
% of
boiler heat
input
0.38
0.24
0.29
0.19
4.4 MWt Distillate oil 44 MWt Distillate oil
Total 7. of. Total 7, of
thermal boiler heat thermal boijer heat
kW . input kW input
A *
* A
A A
* A
Fixed Packed Bed SCR
90% Removal
70% Removal
25
17
0.56
0.38
274
192
0.62
0.44
*Not considered as a candidate system.
-------
200
150
3
0)
60
60
M
0>
C
W
100
50
Parallel Flow SCR
44 MWt Boiler
Moving Bed SCR
Parallel Flow SCR
8.8 MWt Boiler
60 70 80
Percent NOX Control
90
100
Figure 5.1.3-1. Energy usage of NOX control systems for
residual oil-fired standard boilers.
5-20
-------
400
300
0)
60
60
t-t
OJ
C
W
200
100
44 MW Boiler
Fixed Packed Bed SCR
Fixed Packed Bed SCR
.4.4 MW Boiler
60 70 80
Percent NOX Control
90
100
Figure 5.1.3-2. Energy usage of NOX control systems
for distillate oil-fired boilers.
5-21
-------
0.6
0.5
ex
G
M
4-1
rt
0)
ffi
1-1
0)
H
•H
o
m
c
a)
o
>-i
01
PM
CO
Ctf
bO
0)
C
0.4
0.3
0.2
0.1
Parallel Flow SCR
Moving Bed SCR
44 MWt Boiler
8.8 MWt Boiler
60
70
80
90
100
Percent NOV Control
Figure 5.1.3-3.
Energy usage of NOX control systems applied to residual
oil-fired boilers as percent of boiler heat input.
5-22
-------
w
ft
C
H
cfl
0)
o
M
tn
rt
QJ
60
ni
ra
M
M
a)
0.6
0.5
0.4
0.3
0.2
0.1
Fixed Packed Bed SCR
44 MW Boiler
Fixed Packed Bed SCR
4.4 MW<- Boiler
60 70 80
Percent NOX Control
90
100
Figure 5.1.3-4.
Energy usage of NOX control systems applied to distillate
oil-fired boilers as percent of boiler heat input.
5-23
-------
when comparing the two fuels since the size of the standard boilers is dif-
ferent by an order of magnitude. It can be said then, for the residual oil
case, the moving bed systems are less energy intensive than parallel flow
systems due to the moving beds1 lower pressure drop across the length of the
reactor.
It is not clear as to whether or not these systems have been optimized
with respect to energy usage. The technology is relatively new and problem
solving efforts are probably directed toward improving reliability and
operability. It does seem possible that there is an optimum catalyst size
and reactor volume that would minimize the pressure drop. Another potential
method of lowering the pressure drop is to minimize the excess air. This
reduces both the required reactor volume and the AP. It is likely that a
boiler equipped with low NO burners will utilize low excess air for NO
X X
control.
NHs dilution by air instead of steam might possibly use less energy.
There is, however, a safety aspect to consider since some air/NHs mixtures
(<20:1) are explosive.
The energy impact of FGT controls applied to modified or reconstructed
facilities (retrofit application) will be the same or greater than that for
new facilities. If flue gas can be taken from the economizer of the existing
boiler at 350-400°C and returned upstream of any existing heat exchange
equipment, then there will be no additional energy impact.
If the flue gas is only available at a lower temperature (<350-400°C)
then a heater will be required. The energy impact of the heater will depend
on the temperature of the flue gas. If the temperature is that of the outlet
gas of the standard boilers (approximately 180°C) calculations indicate that
energy requirement would be more than tripled even if heat exchange equipment
is used to recover 85% of the energy supplied by the heater. The heater will
probably be oil-fired for ease of control.
5-24
-------
These results indicate that, on retrofit applications, there is a con-
siderable energy incentive to obtain the flue gas at the necessary reaction
temperature in order to avoid flue gas heating. Other energy impacts would
be the same as those for new facilities.
5.1.4 Energy Impact of Controls for Natural Gas-Fired Boilers
This subsection presents the results of energy and material balance for
natural gas-fired industrial boilers. For new facilities two standard boilers
and one M) FGT system is considered. Results are presented for two levels
A
of control.
The data presented is the result of several calculations. First,
material balances were performed and the necessary equipment sized. Then,
knowing the equipment size and flow rates, it was possible to calculate
energy usage for each process step.
The results of these calculations are presented in Table 5.1.4-1. Both
thermal energy requirements and energy requirements as a percentage of
boiler heat input are shown. The candidate system for natural gas-fired
boilers is fixed packed bed SCR. For the calculations, it is assumed that
flue gas is available from the boiler economizer at 375°C and can be returned
upstream of the air heater. Therefore, no energy is necessary for flue gas
heating.
The data appearing in Table 5.1.4-1 is summarized in Table 5.1.4-2 and
plotted in Figures 5.1.4-1 and 5.1.4-2. Figure 5.1.4-1 presents total energy
usage and Figure 5.1.4-2 shows the energy usage as a percent of boiler heat input.
There are some areas of potential energy usage reduction. The catalyst
particle size and reactor volume may be optimized to minimize reactor pressure
drop. Reduction of excess air may also reduce the pressure drop and this
may be standard practice on boilers with low NC) burners. It may be more
X
5-25
-------
TABLE 5.1.4-1. ENERGY CONSUMPTION FOR NOX FGT CONTROL TECHNIQUES FOR NATURAL GAS-FIRED BOILERS
System
Standard boiler
Heat input Type and level
MWt (MBtu/hr) of control
4.4 (15) Package SCR Fixed Packed Bed
Flretube Moderate
Scotch
SCR Fixed Packed Bed
Stringent
44 (150) Package SCR Fixed Packed Bed
Watertube Moderate
SCR Fixed Packed Bed
Stringent
Control
efficiency Energy
7, types
70 Electrical
Steam
90 Electrical
Steam
70 Electrical
Steam
90 Electrical
Steam
Energy consumed % increase
by control device In energy use over
MWt
0.0108
0.00106
0.0173
0.00133
0.123
0.0110
0.203
0.0142
(MBtu/yr) uncontrolled boiler
(0.0369) 0.27
(0.00363)
(0.0590) 0.42
(0.00455)
(0.421) 0.30
(0.0345)
(0.692) 0.49
(0.0483)
Ul
I
TABLE 5.1.4-2,
SUMMARY OF ENERGY REQUIREMENTS FOR NATURAL
GAS-FIRED BOILERS
Total thermal kW
Natural gas
of boiler heat input
4.4 MWt Boiler
Fixed Packed Bed SCR
90% Removal
70% Removal
19
12
0.42
0.27
44 MWt Boiler
Fixed Packed Bed SCR
90% Removal
70% Removal
217
134
0.49
0.30
-------
60
n)
ra
6D
»-l
0)
c
500-
400-
300-
200-
100-
50
Fixed Packed Bed SCR
44 MW Boiler
4.4 MWt Boiler
Fixed Packed
Bed SCR
60
70
80
I
90
100
Percent NO,, Control
Figure 5.1.4-1. Energy usage of NOx control systems for
natural gas-fired standard boiler.
5-27
-------
0.5 -1
.u
ex
id
QJ
S-i
0)
rH
•H
O
cl
0)
u
M
OJ
PH
CO
cd
0)
oo
M
0)
0.4-
0.3-
0.2-
0.1-
44 MW
Fixed
4.4 MWt Boiler
Pixed Packed Bed SCR
50
60
I
70
I
80
I
90
100
Percent NOX Control
Figure 5.1.4-2. Energy usage of NOX control systems as
percent of boiler heat input.
5-28
-------
energy efficient to use air instead of steam for NH3 dilution, however, there
is an explosion hazard with some air:NH3 mixtures (<20:1).
The energy impact of FGT controls applied to modified or reconstructed
facilities (retrofit application) will be the same or greater than that for
new facilities. If flue gas can be taken from the economizer of the exist-
ing boiler at 350-400°C and returned upstream of any existing heat exchange
equipment, then there will be no additional energy impact,
If the flue gas is only available at a lower temperature (<350-400°C)
then a heater will be required. The energy impact of the heater will depend
on the temperature of the flue gas. If the temperature is that of the outlet
gas of the standard boilers (approximately 180°C) calculations indicate that
energy requirement would be more than tripled even if heat exchange equipment
is used to recover 85% of the energy supplied by the heater. The heater
will probably be oil-fired for ease of control.
These results indicate that, on retrofit applications, there is a
considerable energy incentive to obtain the flue gas at the necessary reac-
tion temperature in order to avoid flue gas heating. Other energy impacts
would be the same as those for new facilities.
5.2 NOx/SO SYSTEMS
5.2.1 Introduction
This section considers the energy impacts associated with applying the
UOP NOX/SOX FGT system to three industrial boilers. The combinations
analyzed are presented in Table 5.2.1-1.
The N0x/S0x system has several more energy inputs than the NO only
systems; however, much of this energy is recovered by the air preheater
resulting in an energy credit. The areas of energy utilization are shown
in Table 5.2.1-2.
5-29
-------
TABLE 5.2.1-1. NQx/SOx FGT/BOILER COMBINATIONS ANALYZED FOR ENERGY IMPACT
NOX/SOX
System
Boiler
Fuel*
Control Level
NOX
S0>
UOP
UOP
UOP
Pulverized Coal
Underfeed Stoker
Oil-Fired
LSW
HSE
LSW
HSE
Residual
Oil
80
80
80
85
85
85
* LSW = Low sulfur western coal (0.6% S)
HSE = High sulfur eastern coal (3.5% S)
TABLE 5.2.1-2. AREAS OF ENERGY UTILIZATION IN THE NOX/SOX FGT SYSTEM
Process Step
Type of Energy Consumed
Reactor Draft Loss (Fan)
Liquid NH3 Transfer (Pump)
NHa Vaporization (Vaporizer)
NH3 Dilution
Naphtha Reformer
Compressor/Gasholder
H2SOit Plant
Electrical
Electrical
Steam
Steam
Electrical, Steam, Fuel
Electrical
Electrical, Steam
5-30
-------
For each case, a heat and material balance is performed and these are
contained in Appendices 4 and 5. These are used to size the equipment and
determine the energy requirements. These requirements are listed in tabular
form in each section and summarized. Since only one removal level is con-
sidered, the energy usage is not plotted against removal level as in the
N0x-only section.
5.2.2 Energy Impact of N0x/S0x Controls for Coal-Fired Boilers
Energy usage by these NOX/SOX applications is fairly evenly divided
among three energy types: electrical, steam and fuel. These data are
presented in Tables 5.2.2-1 and 5.2.2-2. Also shown in the tables are the
heat credits for energy recovered by the air preheater.
The net energy usage by the NOX/SOX system is higher than that of the
N0><-only systems. When put on the basis of percent increase in energy over
that of the uncontrolled boiler, the energy usage appears to be a function
of the coal sulfur content, but not unit size. Removal level will also
impact the energy usage; however, the magnitude of this impact is not known.
Energy usage is summarized in Table 5.2.2-3 and plotted in Figure 5.2.2-1.
TABLE 5.2.2-3.
SUMMARY OF ENERGY USAGE OF NOX/SOX SYSTEMS
APPLIED TO COAL-FIRED BOILERS
Fuel
Pulverized coal
Underfeed stoker
Thermal
kW
% of boiler
heat input
Thermal
kW
% of boiler
heat input
Low Sulfur
Western Coal
High Sulfur
Eastern Coal
1,240
11,200
2.1
7.7
200
680
2.3
7.7
5-31
-------
I
OJ
ho
TABLE 5.2.2-1. ENERGY CONSUMPTION FOR NOX/SOX FGT CONTROL TECHNIQUES FOR COAL FIRED BOILERS
Standard
Heat Input
MWt (MBtu/hr)
58.6 200
Systen
Boiler
Type Coal Type
Field Erected High Sulfur
Watertube Eastern Coal
Pulverized Coal
Type and Level
of Control
SCR Parallel Flow
Intermediate
Control Efficiency Energy Types
(Z NOx/SOx)
80/85 Electrical
Steam
Fuel
Heat Credit
Enemy Consumption
Energy Consumed
By Control Device
MWt (MBtu/hr)
9.45 (32.25)
2.79 ( 9.52)
5.18 (17.69)
-(6.24) -(21.3 )
Z Increase in
Energy Use Over
Uncontrolled Boiler
7.7
58.6
200
Low Sulfur
Western Coal
SCR'Parallel Flow
Intermediate
80/85
Electrical
Steam
Fuel
0.941 ( 3.21)
0.703 ( 2.40)
1.09 ( 3.72)
2.1
Heat Credit -(1.49) -( 5.1)
-------
TABLE 5.2.2-2. ENERGY CONSUMPTION FOR NOX/SOX FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Ln
I
UJ
System
Standard Boiler Control Efficiency Energy Types
Heat Input Type Coal Type Type and Level (% NOX/SOX)
MWt (MBtu/hr) of Control
8.8 30 Package High Sulfur SCR Parallel Flow 80/85 Electrical
Watertube Eastern Coal Intermediate
Underfeed Steam
Fuel
Heat Credit
8.8 30 Low Sulfur SCR Parallel Flow 80/85 Electrical
Western Coal Intermediate
Steam
Fuel
Heat Credit
Energy Consumption
Energy Consumed
By Control Device
MWt
0.428
0.416
0.774
-(0.935)
0.151
0.102
0.163
-(0.217)
(MBtu/hr)
(1.46 ) "
(1-42 )
(2.64 )
-(3.19 )
(0.52)
(0.348)
(0.558)
-(0.74 )
% Increase in
Energy Use Over
Uncontrolled Boiler
7.7
2.3
-------
Ln
I
CO
cd
g
0)
o
•H
C
O
o
60
M
a)
c
w
1
High Sulfur Eastern Coal
Low Sulfur Western Coal
1000
2000
Flue Gas Rate (m3/min)
3000
Figure 5.2.2-1. Energy consumption of parallel flow SCR N0x/S0x FGT systems
for coal-fired boilers.
-------
The energy impact of FGT controls applied to modified or reconstructed
facilities (retrofit application) will be the same or greater than that for
new facilities. If flue gas can be taken from the economizer of the existing
boiler at 350-400°C and returned upstream of any existing heat exchange
equipment, then there will be no additional energy impact.
If the flue gas is only available at a lower temperature ( 350-400°C)
then a heater will be required. The energy impact of the heater will depend
on the temperature of the flue gas. The energy used in heating the gas is
not completely lost since the air preheater can recover about 85% of the
energy supplied by the gas heater. The heater will probably be oil-fired
for ease of control.
These results indicate that, on retrofit applications, there is an
energy incentive to obtain the flue gas at the necessary reaction temperature
in order to avoid flue gas heating. Other energy impacts would be the same
as those for new facilities.
5.2.3 Energy Impact of NOX/SOX Controls for Oil-Fired Boilers
Three types of energy are utilized by these systems—electrical, steam
and fuel. The amounts of each type are shown in Table 5.2.3-1. Energy
consumption of each energy type is of the same order of magnitude. Also
shown is the heat credit that is obtained by partially recovering heat from
the energy inputs with the air preheater.
Since only one case is considered, the variables that affect the energy
impact cannot be quantified. It.can be stated qualitatively, however, that
the primary variables that affect energy usage are oil sulfur content and
control level (both NOX and SOX). The effect of fuel sulfur content was
examined in the section on coal-fired applications.
5-35
-------
TABLE 5.2.3-1. ENERGY CONSUMPTION FOR NOX/SOX FGT CONTROL TECHNIQUES FOR OIL FIRED BOILERS
Syetea Energy Consumption
Standard Boiler Control Efficiency Energy Types Energy Consumed
Heat Input Type Oil Type Type and Level (X HOx/SOx) By Control Device
MWt (MBtu/hr) of Control MWt (MBtu/hr)
44 150 Package Residual SCR Parallel Flow 80/85 Electrical 1.26 (4.29) '
Watertube Intermediate
Steam 1.16 (3.96).
Fuel 2.16 (7.37)
Heat Credit -(2.63) -(8.98) _
7. Increase in
Energy Use Over
Uncontrolled Boiler
4.4
-------
The energy impact of FGT controls applied to modified or reconstructed
facilities (retrofit application) will be the same or greater than that for
new facilities. If flue gas can be taken from the economizer of the existing
boiler at 350-400°C and returned upstream of any existing heat exchange
equipment, then there will be no additional energy impact.
If the flue gas is only available at a lower temperature (<350-400°C)
then a heater will be required. The energy impact of the heater will depend
on the temperature of the flue gas. The energy used in heating the gas is
not completely lost since the air preheater can recover about 85% of the
energy supplied by the heater. The heater will probably be oil-fired for
ease of control.
These results indicate that, on retrofit applications, there is an
energy incentive to obtain the flue gas at the necessary reaction temperature
in order to avoid flue gas heating. Other energy impacts would be the same
as those for new facilities.
5.3 SUMMARY
All of the NOy-only systems and cases required <1% of the total heat
input to the boiler. By far, the item contributing the most to energy con-
sumption was the flue gas fan which supplied the draft loss caused by the
catalyst bed.
The parallel flow systems appear to use more energy than the moving bed
systems; however, both are of the same order of magnitude. Within the
accuracy of the calculations, the systems examined should be considered to
have approximately similar energy impacts.
The NOX/SOX syrtems require 2-8% of the total heat input to the boiler.
This is primarily due to the requirement for S02 workup. Although this
requirement is higher than that for N0x-only processes, it may be less than
that for the combination of N0x-only and FGD.
5-37
-------
REFERENCES
1. Ando, Jumpei. "NO Abatement for Stationary Sources in Japan."
EPA report currently in preparation, April 1978.
2. Marcos, Chemico Air Pollution Control Corporation. Telephone
Conversation. 29 September 1978.
3. Perry, Robert H. Chemical Engineers Handbook. 5th Edition. 1973.
McGraw-Hill, pp. 5-52, 53.
4. Broz, Larry. Acurex Memo: "ITAR Average SIP Requirements." August
29, 1978.
5. Faucett, H.L., et al. Technical Assessment of NO Removal Processes
for Utility Application. EPA-600/7-77-127. November 1977.
5-38
-------
SECTION 6
ENVIRONMENTAL IMPACT OF CANDIDATES FOR
BEST EMISSION CONTROL SYSTEMS
6.1 INTRODUCTION
The three best candidate control techniques identified in Section III
are as follows:
SCR - Fixed Packed Bed
SCR - Parallel Flow
SCR - Moving Bed
These techniques have similar environmental concerns as they all utilize
to reduce NO to N2. All are capable of achieving the levels of control con-
sidered in this study, although the applicability of a particular system is
fuel dependent. While NO reduction is the primary beneficial environmental
impact of these systems, particulate removal is a secondary beneficial impact
of the moving bed systems. The moving bed system will reduce the level of
particulates in the flue gas by 70-80%.1>2 The particulates are embedded on
the catalyst (rings or granules) as the catalyst moves downward through the
reactor. The dirty catalyst is removed to a vibrating screen which separates
the dust and the clean catalyst is then recycled to the top of the reactor.
Pilot plant tests on the moving bed reactor have shown it capable of handling
<1 g/Nm3 of particulates.3 The uncontrolled particulate levels in the flue
gas from the pulverized coal (5-9 g/Nm3), spreader stoker (3.5-6.3 g/nM3),
chaingrate and underfeed (both 1.4-2.4 g/Nm3) standard boilers are all greater
than this figure. As a result, moving bed systems are not considered for
application to the coal-fired standard boilers. The fixed packed bed system
6-1
-------
cannot tolerate participates so it is applied only to natural gas- and
distillate oil-fired flue gas which have low particulate loadings (13 and
19 mg/Nm3, respectively, for the standard boilers). Conversely, the paral-
lel flow system can tolerate full particulate loadings (up to 20 g/nM3)1*
as the open passageways present unobstructed paths for particulates to pass
through with little impaction on the catalyst surface.
There are some potential adverse environmental impacts of the selective
catalytic reduction processes. First, the use of NH3 as the gaseous reducing
agent introduces the possibility of ammonia emissions. Commercial operations
of the three reactor types on industrial and utility boilers have demon-
strated emissions of <10 ppm NH3 at the NH3:NO mole ratio required for
stringent level of control. These levels are shown graphically in Figures
6.1-1 through 6.1-3 on the following pages. (These plots are composites of
the available commercial data.) The only data available on NH3 emissions are
from Japanese process vendors and these data indicate NH3 emissions to be <10
ppm. This number may be optimistic, especially considering that currently
there is no continuous monitoring technique for measuring NH3 in the presence
of SO . The data, therefore, represent spot measurements and not continuous
X
data. It seems reasonable to assume that 10 ppm represents a minimum level
of NH3 emissions.
A potential environmental problem in commercial SCR operations is the
formation of ammonium bisulfate, NH^HSO^, or ammonium sulfate, (NH^SC^.
The presence of NH3 , S03, and H20 in the hot flue gas leads to the formation
of liquid NH4HS04 upon cooling to approximately 180-220°C by the following
reaction.
NH3(g) + S03(g) + H2OCg) ~t NH^HSC^a) C6-D
This can create a plugging and corrosion problem in heat exchange equipment,
particularly for boilers burning medium- or high-sulfur fuels. Further
cooling to about 190°C precipitates the formation of solid ammonium sulfate
[ (NHi, )2SOif ] by the following reaction.
6-2
-------
30
NH3(ppm)
Level of Control NHa:NO^ NH3 Emissions
Moderate
Intermediate
Stringent
.7
.8
.9
1 ppm
1 ppm
1 ppm
.6
.7 .8- .9
NH3 :N<5 Mole Ratio
1.0
1.1
1.2
Figure 6.1-1. NH3 Emissions - Fixed Packed Bed Reactor.5'6'7
6-3
-------
NH3(ppm)
Moderate
Intermediate
Stringent
1.0
1.1
1.2
NH3:NC> Mole Ratio
Figure 6.1-2. NH3 Emissions - Parallel Flow Reactor.8'9'10'11'12
6-4
-------
NH3(ppm)
Moderate
Intermediate
Stringent
10 _
.5
.7
.8 .9
NH3:NC> Mole Ratio
1.0
1.1
1.2
Figure 6.1-3. NH3 Emissions - Moving Bed Reactor.1:
6-5
-------
00 + NH3(g) t (NHO2S04(s-) (6-2)
The impact of the solid sulfate and liquid bisulfate on downstream particu-
late collection equipment and FGD systems is unknown at present and is
currently being investigated by the EPA and others. It is speculated that
minor, if any, amounts of these sulfates will be emitted to the atmosphere in
situations where particulate control equipment exists downstream of the NO
X
control system.
The final environmental concern of the SCR processes is disposed of
spent catalyst. Catalysts such as titanium dioxide (Ti02) and vanadium
pentoxide (V205) are probably recycled due to their high cost. To date,
virtually no catalyst has been used commercially yet for over 10,000 hours,
and, as a result, there has been no commercial experience on the treatment of
spent catalyst. Reprocessing or disposal of spent catalyst will most likely
o *?
be carried out by the catalyst vendor. This question is not currently
addressed in literature. Another potential problem related to catalysts is
that of catalyst erosion, especially with the moving bed systems. Catalyst
erosion may generate small particulates which may present a stack fume pro-
blem if particulate control devices are not present or not effective at re-
moving the catalyst particles. No problems of this nature have been reported
at this time.
6.2 ENVIRONMENTAL IMPACTS OF CONTROLS FOR COAL-FIRED BOILERS
6.2.1 Air Pollution
The emission rates for primary and secondary pollutants are presented in
Tables 6.2.1-1 through 6.2.1-12 on the following pages. There are three
tables for tach of the 4 coal-fired standard boilers. Each table is broken
down according to coal type (high sulfur eastern and low sulfur western) and
control level (uncontrolled, moderate, intermediate, and stringent). For each
entry the impact on the primary pollutant, NOX, is shown. Then, the adverse
impact of the secondary pollutant, NHs, is given for each case.
-------
TABLE 6.2.1-1.
AIR POLLUTION IMPACTS FROM BEST NOX FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
1 > 2 » 8"17
Control level Type of control
Uncontrolled -
Moderate-70% SCR-Parallel Flow
Stringent-90% SCR-Parallel Flow
TABLE 6.2.1-2. AIR POLLUTION IMPACT
Control level Type of control
Uncontrolled -
Moderate-70% SCR-Parallel Flow
Stringent-90% SCR-Parallel Flow
Standard Boiler: Pulverized Coal
Heat Rate: 200 MBtu/hr
Coal: High Sulfur Eastern
NOV Participates
g/s ng/J g/s ng/J
(Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
19.2 328.0 181.2 3090
(152.46) (.762) (1436.5) (7.18)
5.77 98.2 Negligible Effect
(45.7) (.229)
1.92 32.8 Negligible Effect
(15.2) (.0762)
S FROM BEST N0x FGT CONTROL TECHNIQUES
Standard Boiler: Pulverized Coal
Heat Rate: 200 MBtu/hr
Coal: Low Sulfur Western
NOX Participates
g/s ng/J g/s ng/J
(Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
23.7 403.0 113.5 1936
(187.56) (.938) (900.3) (4.50)
7.10 121.0 Negligible Effect
(56.3) (.282)
2.37 40.3 Negligible Effect
(18.8) (.0938)
NH3
g/s ng/J
(Ib/hr) (Ib/MBtu) Bisulfate
00 0
.0154 .261 See Text
(.122) (.000608)
.0767 1.31 See Text
(.608) (.00304)
FOR COAL-FIRED BOILERS1'2'8"17
NH3
8/s ng/J
(Ib/hr) (Ib/MBtu) Bisulfate
00 0
.0159 .272 See Text
(.126) (.000632)
.0797 1.36 See Text
(.632) (.00316)
-------
TABLE 6.2.1-3. AIR POLLUTION IMPACTS FROM BEST NO FGT CONTROL TECHNIQUES FOR COAL^FIRED BOILERS
1> 2 » 8-17
Standard Boiler: Spreader Stoker
Heat Rate: 150 MBtu/hr
Coal: High Sulfur Eastern
Uncontrolled
Internedlate-8
NOX
Particulates
NH3
Control
level
Type
of
control
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
Bisulf ate
SCR-Parallel Flow
12.0
(95.4)
2.41
(19.1)
273.0
(.636)
54.7
(.127)
111.0
(876.4)
2512
(5.84)
Negligible Effect
.0266
(.211)
.604
(.00140)
See Text
I
CO
TABLE 6.2.1-4. AIR POLLUTION IMPACTS FROM BEST NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
1,2,3-17
Control level
Uncontrolled
Intermedlate-80%
g/s
Type of control (Ib/hr)
14.8
(117.15)
SCR-Parallel Flow 2.95
(23.4)
Standard
Boiler:
Heat Rate:
NO*
ng/J
(Ib/MBtu)
336.0
(.781)
67.2
(.156)
Coal:
Spreader Stoker
150 MBtu/hr
Low Sulfur
Particulates
g/s
(Ib/hr)
69.1
(548.3)
ng/J
(Ib/MBtu)
1572
(3.66)
Negligible Effect
Western
NH3
g/s
(Ib/hr)
0
.0273
(.217)
ng/J
(Ib/MBtu)
0
.622
(.00145)
Blsulfate
0
See Text
-------
TABLE 6.2.1-5. AIR POLLUTION IMPACTS FROM BEST NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
1 > 2 > 8—17
Standard Boiler: Chaingrate
Heat Rate: 75 MBtu/hr
Coal: High Sulfur Eastern
Uncontrolled
Moderate-70%
Intermediate-80%
Strlngent-9
NOX
Particulates
KH3
Control
level
Type
of
control
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
Bisulfate
SCR-Parallel Flow
SCR-Parallel Flow
SCR-Parallel Flow
6.02
(47.7)
1.80
(14.3)
1.20
(9.54)
.602
(4.77)
273.0
(.636)
82.0
(.191)
54.7
(.127)
27.3
(.0636)
21.2
(168.5)
966.0
(2.25)
Negligible Effect
Negligible Effect
Negligible Effect
.00662 .301 See Text
(.0525) (.000700)
.0132 .602 See Text
(.105) (.00140)
.0331 1.50 See Text
(.262) (.00350)
TABLE 6.2.1-6. AIR POLLUTION IMPACTS FROM BEST NO FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
1,2,&~17
Standard Boiler:
Heat Rate:
Coal:
Chaingrate
75 MBtu/hr
Low Sulfur Western
Uncontrolled
Moderate-70%
Intermediate-fi
Stringent-90%
NOX
Particulates
NH3
Control
level
Type
of
control
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
Bisulfate
SCR-Parallel Flow
SCR-Parallel Flow
SCR-Parallel Flow
7.40
(58.65)
2.22
(17.6)
1.48
(11.7)
.740
(5.87)
336.0
(.782)
101.0
(.235)
67.2
(.156)
33.6
(.0782)
13.3
(105.6)
605.0
(1.41)
Negligible Effect
Negligible Effect
Negligible Effect
.00683
(.0542)
.0137
(.108)
.0342
(-271)
.311 See Text
(.000723)
.621
(.00145)
See Text
1.55 See Text
(.00361)
-------
TABLE 6.2.1-7. AIR POLLUTION IMPACTS FROM BEST NOV FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
19 2, S~17
Standard Boiler: Underfeed Stoker
Heat Rate: 30 MBtu/hr
Coal: High Sulfur Eastern
Uncontrolled
Intermedlate-80%
NO*
Participates
NH3
Control
level
Type
of
control
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
Bisulfate
SCR-Parallel Flow
2.40
(19.05)
.480
(3.81)
273.0
(.635)
54.6
(.127)
8.49
(67.31)
965.0
(2.24)
Negligible Effect
.00529
(.0419)
.601
(.00140)
See Text
I
I—1
o
TABLE 6.2.1-8.
AIR POLLUTION IMPACTS FROM BEST N0x FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
1> 2> 8~17
Control level Type of control
Uncontrolled -
Intermediate-80% SCR-Parallel Flow
Standard Boiler: Underfeed Stoker
Heat Rate: 30 MBtu/hr
Coal: Low Sulfur Western
NOX Particulates
g/e ng/J g/s ng/J
(Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
2.95 335.0 5.31 604.0
(23.40) (.780) (42.12) (1.40)
.590 67.1 Negligible Effect
(4.68) (.156)
g/s
(Ib/hr)
0
.00544
(.0431)
NH3
ng/J
(Ib/MBtu)
0
.618
(.00144)
Bisulfate
0
See Text
-------
TABLE 6.2.1-9.
AIR POLLUTION IMPACTS FROM BEST NOx/SOx FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
I
H
H-1
Standard Boiler:
Heat Rate:
Coal:
Pulverized Coal
200 MBtu/hr
High Sulfur Eastern
NOV SO, Pa.rttculates
B/s ng/J g/s ng/J g/s ng/J
Control level Type of control (Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
Uncontrolled - 19.2 328.0 142.0 2423.0 181.0 3090.0
(152.5) (0.762) (1127.0) (5.64) (1437.0) (7.18)
Intermediate SRC-Parallel Flow 3.85 65.6 21.3 363.0 Negligible Effect
(80% N0y) (30.5) (0.153) (169.0) (0.865)
(85% S02)
TABLE 6.2.1-10. AIR POLLUTION
IMPACTS FROM BEST NOX/SOX FGT CONTROL TECHNIQUES
Standard Boiler: Pulverized Coal
Heat Rate: 200 MBtu/hr
Coal: LOW Sulfur Western
NOV SOj Ps(rttculatea
g/s ng/J g/s ng/J g/s ng/J
Control level Type of control (Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
Uncontrolled - 23.7 403.0 30.0 511.0 113.5 1936.0
(187.6) (0.938) (237.6) (1.19) (900.3) (4.50)
Intermediate SRC-Parallel Flow 4.73 80.6 4.49 7.65 Negligible Effect
(80% NO ) (37.5) (0.188) (35.6) (0.178)
(85% S02)
NH3
g/s ng/J
(Ib/hr) (Ib/MBtu)
0 0
0.307 5.22
(2.43) (0.0122)
FOR COAL-FIRED
NH^
g/s ng/J
(Ib/hr) (Ib/MBtu)
0 0
0.318 5.42
(2.52) (0.0126)
Bisulfate
0
See Text
BOILERS
Bisulfate
0
See Text
-------
TABLE 6.2.1-11.
AIR POLLUTION IMPACTS FROM BEST N0x/S0x FGT CONTROL TECHNIQUES FOR COAL-FIRED BOILERS
Standard Boiler:
Heat Rate:
Coal:
Underfeed Stoker
30 MBtu/hr
High Sulfur Eastern
Control level
Type of control
NOV
SO;
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
Pa.rtJ.culatea
S/B
(Ib/hr)
ng/J
(Ib/MBtu)
g/B
(Ib/hr)
ng/J
(Ib/MBtu)
g/s.
(Ib/hr)
ng/J
(Ib/MBtu)
Bisulfate
Uncontrolled
Intermediate
(80% NOX)
(85% S02)
SRC-Parallel Flow
2.40
(19.05)
0.481
(3.81)
273.0
(0.635)
54.6
(0.127)
21.3
(168.9)
3.20
(25.3)
2421.0
(5.63)
363.0
(0.845)
8.49
(67.31)
965.0
(2.24)
Negligible Effect
0.0529
(0.419)
6.01 See Text
(0.0140)
TABLE 6.2.1-12.
X X
Standard Boiler: Underfeed Stoker
Heat Rate: 30 MBtu/hr
Coal: Low Sulfur Western
Control level Type of control
Uncontrolled
Intermediate SRC-Parallel Flow
(80% NO )
(85% S02)
g/s
(Ib/hr)
2.95
(23.40)
0.590
(4.58)
NOV SO,
ng/J g/s ng/J
(Ib/MBtu) (Ib/hr) (Ib/MBtu)
Fa,rtl,culates
g/s
(Ib/hr)
ng/J
(Ib/MBtu)
335.0 4.49 510.0 5.31 604.0
(0.780) (35.6) (3.19) (42.12) (1.40)
67.1 0.674 76.8 Negligible Effect
(0.156) (5.34) (0.179)
NH3
g/s ng/J
(Ib/hr) (Ib/MBtu)
0 0
0.0544 6.18
(0.431) (0.0144)
Bisulfate
0
See Text
-------
Table 6.2.1-13 shows the uncontrolled NO emissions for each standard
boiler and the SIP level for each fuel. One can see that a majority of the
uncontrolled emissions are less than SIP allowable levels. For coal, the
worst case is the pulverized coal boiler burning low sulfur western coal. To
meet the SIP control level the degree of removal required is as follows.
QQS _ 7
% reduction = 938 * 100% = 25%
This level of control is easily obtained via combustion modifications,18
therefore, this study does not address the SIP control level.
Also, in Tables 6.2.1-1 through 6.2.1-12 the secondary pollutant bisul-
fate is not quantified, for several reasons. First, kinetic and thermody-
namic data for the reaction
NH3(g) + S03(g) + H20(g)
have not yet been evaluated. Therefore, the extent of reaction cannot
be determined for the residence time of the flue gas in the duct. Second,
bisulfate emissions are not constant since ^hey are at a high level during
soot blowing and at a lower level during other periods. Finally, it is not
known how much of the bisulfate is caught by downstream particulate removal
equipment (assuming that, the equipment is present) and how much is emitted.
A hypothetical calculation can be made for the case of the pulverized coal
standard boiler burning high sulfur eastern coal (Table 6.2.1-1) equipped
with an SCR parallel flow control system operating at the stringent level
of control. The amount of bisulfate formed is as follows.
6-13
-------
TABLE 6.2.1-13. NO EMISSION LEVELS AND SIP CONTROL LEVELS
(All values in lb/106 Btu)
Uncontrolled standard boilers
SIP Natural Distillate Residual
control gas oil oil
Gas .2 .175
Oil .3 .159 .400
Coal .7
Coal
type
High S
Eastern
Low S
Eastern
Low S
Western
Underfeed
stoker
.635
.545
.780
Chaingrate Spreader
stoker
.636 .636
.544 .543
.782 .781
Pulverized
coal
.762
.653
.938
-------
Assuming all NH3 ->
wt. bisulfate =
L9 tons
year
There are beneficial aspects of this reaction. The bisulfate formation ties
up SO3 which is more hazardous than S02 and is difficult to catch with FGD. l9
If the bisulfate can be collected adequately and disposed of safely, an
overall improvement could be achieved.
The fate of bisulfate formed in this manner has not been completely
resolved and is currently an aspect of NO control receiving much attention.
In Japan there have been problems at some installations of precipitation
of the bisulfate or sulfate on elements of regenerative air preheaters and
tubes of tubular air preheaters. This deposit can be removed by
periodically water-washing the air preheater.20 Bisulfate or sulfate parti-
culates that pass through the air preheater may be collected by downstream
particulate control or FGD equipment if such equipment exists. The collec-
tion efficiency of particulate control equipment cannot be determined quan-
titatively without knowing the particle size and resistivity (for ESP's)
or the K-factor and concentration (for baghouses). These data are not
currently known for these compounds, however, it is reasonable to expect
that some fraction of the particles will be collected.21 Similarly, the
collection efficiency for an FGD unit has not been examined. Where neither
particulate control nor FGD equipment exist, there may be stack emissions
of sulfates. An FGD system may also absorb NH3 emitted by an SCR system,
however, the removal cannot be determined from the information currently
available.
3-15
-------
To reduce the adverse environmental effects, improved combustion control
utilizing less 62 minimizes the formation of NO It also minimizes forma-
tion of SO3 which is necessary for ammonium bisulfate formation. Careful
operation of the FGT system should keep the NHs injection ratio as low as
possible to minimize NH3 emissions and bisulfate formation. Also, careful
operation of downstream heat exchange equipment to keep the flue gas above
the acid (SO3) dew point is required. The use of corrosion-resistant
material in any heat exchanger is advisable where NH^HSOi,. deposits are
probable.22 A multitube type heat exchanger with the tubes placed vertically
is a possible configuration to prevent bisulfate deposits from causing prob-
lems. Any bisulfate liquid formed in the tube will drip downward as long
as the temperature of the tube is kept above the melting point of bisulfate.23
It will be necessary to design the exchanger out of corrosion resistant
materials.
6.2.2 Water Pollution
There are no water streams that are associated with NO -only SCR systems,
however, there is one potential source of water pollution. In some Japanese
installations NH^HSOit deposits are removed from the air preheater by water
washing. The blowdown from this operation will contain both ammonium and sul-
fate ions which, if not treated, present a water pollution source. Since the
amounts of NH^HSOij and water are not known it is possible to estimate the con-
centration or flow additional of this potential source.
6.2.3 Solid Waste
The only potential solid waste associated with this system is spent
catalyst. Presently, the life of SCR catalysts is 1-2 years and the topic
of recycling is not addressed in the literature. Since, to date, virtually
no commercial SCR units have operated long enough to require catalyst replace-
ment, there is no commercial experience on the treatment of spent catalyst.27
With the high cost of some of the catalysts, recycling seems to be warranted.
Recycling is feasible where the spent catalyst support is composed of valuable
6-16
-------
materials such as titanium, stainless steel, or possibly a ceramic material.
Alumina catalyst supports probably do not warrant recycling unless required
for environmental reasons. These catalysts can probably be disposed of in the
same manner as other industrial cataylsts.
The amount of catalyst that must be recycled or disposed of is one
reactor volume since replacement involves total catalyst replacement. The
actual frequency of catalyst replacement is unique to each specific process,
however, catalyst lifetimes are typically one or more years.
Few process vendors have published their catalyst formulations since
the field is very competitive at the present time. Base metal oxides are
typically used.24 The environmental impact of catalyst disposal will depend
on what materials and compounds are involved.
6.2.4 Other Environmental Impacts
The only other environmental impact is a secondary impact. NH3 is
commonly made from natural gas and its consumption is considered a secondary
environmental impact. This impact is quantified in a report on the impact of
NO regulations on the NHs industry. The report was prepared by TVA under
contract to EPA-IERL. Other adverse environmental impacts (noise, thermal
pollution, electrical discharges, etc.) are not present with SCR systems.
6.2.5 Environmental Impact on Modified and Reconstructed Facilities
The environmental impacts of a new facility and a retrofitted facility
should be similar. There is not enough difference between new and retrofit
systems to indicate that environmental impacts would be significantly dif-
ferent with retrofitted systems.
6-17
-------
6.3 ENVIRONMENTAL IMPACTS OF CONTROLS FOR OIL-FIRED BOILERS
6.3.1 Air Pollution
Emission rates for primary and secondary pollutants are presented in
Tables 6.3.1-1 and 6.3.1-2 on the next pages. The tables are organized by
fuel type (residual and distillate oil); control level (uncontrolled, moderate,
and stringent); and type of control (residual - SCR moving bed and SCR
parallel flow; distillate - SCR fixed packed bed). The impact on the primary
pollutant, NO , is given for each case. Also, the moving bed's impact on
A
particulates of residual oil is shown. The adverse impact of the secondary
pollutant, NH3, is given for each entry.
Table 6.2.1-13 shows that only the residual oil-fired flue gas has uncon-
trolled NO emissions greater than the SIP control level. To achieve the SIP
X
level of control the removal required is as follows.
% reduction = °'o~°'3 x 100%
= 25%
This control level is readily achieved by combustion modifications; hence,
FGT to achieve the SIP control level is not examined for oil-fired boilers.
In Table 6.3.1-1 the secondary pollutant bisulfate is not quantified.
This is due to a lack of developed kinetic and thermodynamic data to predict
the extent of reaction. Also, removal levels are not constant since the
degree of downstream particulate removal is uncertain. However, one can see
that the bisulfate problem is worse for residual oil than for distillate oil
because ther ^ is more S03 available for reaction. Bisulfate is formed by a
one-to-one reaction between NH3 , SOa and HzO.
NH3(g) + S03(g) + H20(g)
6-18
-------
TABLE 6.3.1-1.
Standard boiler
Heat rate
(MBtu/hr) Type Control level
150 Residual Uncontrolled
Moderate - 70%
Stringent - 90%
Moderate - 70%
Stringent - 90%
30 Residual Uncontrolled
Moderate - 70%
Stringent - 90%
Moderate - 70%
Stringent - 90%
75 Distillate Uncontrolled
Moderate - 70%
Stringent - 90%
15 Distillate Uncontrolled
Moderate - 70%
Stringent - 90%
AIR POLLUTION IMPACTS
FOR OIL-FIRED BOILERS1
Type of control
_
SCR Moving Bed
SCR Moving Bed
SCR Parallel Flow
SCR Parallel Flow
_
SCR Moving Bed
SCR Moving Bed
SCR Parallel Flow
SCR Parallel Flow
_
SCR Fixed
Packed Bed
SCR Fixed
Packed Bed
_
SCR Fixed
Packed Bed
SCR Fixed
Packed Bed
NO
g/B
(Ib/hr)
7.57
(50.0)
2.27
(18.0)
.757
(6.00)
2.27
(18.0)
.757
(6.001
2.02
(16.0)
0.606
(4.80)
0.202
(1.60)
0.606
(4.80)
0.202
(1.60)
2.99
(23.76)
0.898
(7.13)
0.299
(2.38)
.300
(2.38)
.0900
(.71/0
.0300
(.238)
FROM BEST NO FGT CONTROL TECHNIQUES
, 2, 5-17 X
ng/J
(Ib/MBtu)
172.0
(.400)
51.6
(.120)
17.2
(.0400)
51.6
(.120)
17.2
(.0400)
229.0
(0.533)
68.7
(0.160)
22.9
(0.0533)
68.7
(0.160)
22.9
(0.0533)
68.0
(0.158)
20.4
(0.047)
6.80
(0.0158)
68.2
(.159)
20.5
(.0476)
6.82
(.0159)
Particulates
g/s ng/J
(Ib/hr) (Ib/MBtu)
4.16 94.6
(33.0) (.220)
1.25 28.4
(9.90) (.0660)
1.25 28.4
(9.90) (.0660)
Negligible Effect
Negligible Effect
.580 65.9
(4.60) (.153)
.174 20.1
(1.38) (.0459)
.174 20.1
(1.38) (.0459)
Negligible Effect
Negligible Effect
1.02 46.4
(8.10) (.108)
Negligible Effect
Negligible Effect
Negligible Amount
NH3
g/s
(Ib/hr)
0
.00957
(.0759)
.0766
(.607)
.00957
(.0759)
.0479
(.380)
0
.00201
(.0160)
.0161
(.128)
.00201
(.0160)
.0101
(.0798)
0
.00502
(.0398)
.00502
(.0398)
0
.00109
(.00864)
.00109
(.00864)
ng/J
(Ib/MBtu)
0
.218
(.000506)
1.74
(.00405)
.218
(.000506)
1.09
(.00253)
0
.229
(.000533)
1.83
(.00426)
.229
(.000533)
1.14
(.00266)
0
.228
(.000531)
.228
(.000531)
0
.248
(.000576)
.248
(.000576)
Bisulfate
0
See Text
See Text
See Text
See Text
0
See Text
See Text
See Text
See Text
0
Less than
Residual
Less than
Residual
n
Less than
Residual
Less than
Residual
-------
TABLE 6.3.1-2.
AIR POLLUTION IMPACTS FROM BEST NO /S0v FGT CONTROL TECHNIQUES FOR OIL-FIRED BOILERS
X X
Boiler Type: Watertube
Heat Rate: 150 MBtu/hr
Oil: Residual
NOV S02
g/s ng/J g/s ng/J
Control level Type of control (Ib/hr) (Ib/MBtu) (Ib/hr) (Ib/MBtu)
Uncontrolled - 7.57 172.0 59.4 1350.0
(60.0) (0.400) (471.0) (3.14)
Intermediate SRC-Parallel Flow 1.51 34.4 8.91 203.0
(80% NOX) (12.0) (0.0800) (70.7) (0.471)
(85% SO?)
Particulates
g/s ng/J
(Ib/hr) (Ib/MDtu)
4.16 94.6
(33.0) (0.220)
Negligible Effect
NHj
g/s np,/J
(Ib/hr) (Ib/MJHu)
0 0
0.191 4.36
(1.52) (0.0101)
Bisulfate
0
See Text
I
M
O
-------
The flue gas S03 concentration can be calculated as follows:
Residual Oil
Flue gas
S02
Fuel S
= 46,700
111
min V400 + 460 °R/ \359 scf
32+ 460 °R\ /lb-mole\ / 60 min
hr
4465 Ib-moles/hr
Ibs
7.359 Ib-moles/hr
3.0%
Residual
2 3
Sulfur In heavy oil (%)
Figure 6.3.1-1. Formation ratio of SO3 •
25
from Figure 6.3.1-1 % S03 =2.3%
6-21
-------
The SO3 concentration can be determined by calculation to be
[S03] = 39 ppm
Distillate Oil
Flue gas
S02
Fuel S
from Figure 6.3.1-1 % S03
= 5000
ftd \ / 32 + 460 °R \ /Ib-mole
min / 1350 + 460 °R / \359 scf
507.6 Ib-moles/hr
Ibs /Ib-mole
= 7.67
/ Ib-mc
\64.0
hr \64.0 Ib
.1198 Ib-moles/hr
0.5%
4.3%
60 min\
hr /
The SO3 concentration can be determined by calculation to be
[SO3] =11 ppm
One can see in Figure 6.3.1-2 below that the residual oil-fired flue gas
will form bisulfate at a higher temperature (earlier in the exchanger).
Also, if the NHs concentration does not become limiting, the greater SOs
concentration will drive the equilibrium of the reaction further to the
right, creating more bisulfate and eventually sulfate, (NHif)2SOi(.
6-22
-------
1000 -
I
a.
100 -
10
100
SO3, ppm
1000
Figure 6.3.1-2. Temperatures below which
forms.
26
Removing the SO3 as bisulfate using particulate control equipment may be
a more effective method of removing SO 3 from the environment than FGD. This
would be a beneficial impact of bisulfate formation. A downstream FGD sys-
tem could potentially absorb the small NHs emissions and, therefore, NHs
emissions may be negligible if FGD is used on conjunction with FGT. The
level of removal that can be achieved by an FGD scrubber has not yet been
examined. Also, the effect of absorbed NHs on the FGD chemistry has not
been resolved, although this question is being studied by the EPA. This is
due primarily to the fact that there is only one installation where FGD is
applied downstream of an SCR unit and data from this Japanese installation
has not been published in the U.S. Several things can be done to reduce the
adverse environmental impacts. Combustion control with less 02 minimizes
formation of NO and SO3. This would be the case for a boiler equipped with
X
low NO., burners. A rrlnimum NH injection ratio is needed for low NHs
X
emissions and bisulfate formation. Heat exchanger temperatures must be kept
above bisulfate formation and acid condensation points. Use of corrosion-
resistant material is warranted where bisulfate deposits are probable. Ver-
ticle tube heat exchangers are preferable since they are less prone to plugging,
6-23
-------
6.3.2 Water Pollution
There are no water streams that are associated with N0x~only SCR systems,
however, there is one potential source of water pollution. In some Japanese
installations NH^SC^ deposits are removed from the air preheater by water
washing. The blowdown from this operation will contain both ammonium and
sulfate ions which, if not treated, present a water pollution source. Since
the amounts of NH^HSOij and water are not known, it is impossible to estimate
the concentration or flow rate of this potential source.
6.3.3 Solid Waste
The only potential solid waste associated with this system is spent
catalyst. Presently, the life of SCR catalysts is 1-2 years and the topic
of recycling is not addressed in the literature. Since, to date, virtually no
commercial SCR units have operated long enough to require catalyst replacement,
there is no commercial experience on the treatment of spent catalyst.27 With
the high cost of some of the catalysts, recycling seems to be warranted. Re-
cycling is feasible where the spent catalyst support is composed of valuable
materials such as titanium, stainless steel or possibly a ceramic material.
Alumina catalyst supports probably do not warrant recycling unless required
for environmental reasons. These catalysts can probably be disposed of in the
same manner as other industrial catalysts.
The amount of catalyst that must be recycled or disposed of is one
reactor volume since replacement involves total catalyst replacement. The
actual frequency of catalyst replacement is unique to each specific process,
however, catalyst lifetimes are typically one or more years.
Few process vendors have published their catalyst formulations since
the field is very competitive at the present time. Base metal oxides are
typically used.21* The environmental impact of catalyst disposal will depend
on what materials and compounds are involved.
6-24
-------
6.3.4 Other Environmental Impacts
The only other environmental Impact is a secondary impact. NH3 is
commonly made from natural gas and its consumption is considered a secondary
environmental impact. This impact will be quantified in a forthcoming report
on the impact of NO regulations on the NH3 industry. The report is being
X
prepared by TVA under contract to EPA-IERL.
Other adverse environmental impacts (noise, thermal pollution, electri-
cal discharges, etc.) are not present with SCR systems.
6.3.5 Environmental Impacts on Modified and Reconstructed Facilities
The environmental impacts of new and retrofitted facilities should be
similar. There is not enough difference between new and retrofit systems
to indicate that environmental impacts would be significantly different
with retrofitted systems.
6.4 ENVIRONMENTAL IMPACTS OF CONTROLS FOR GAS-FIRED BOILERS
6.4.1 Air Pollution
Emission rates for primary and secondary pollutants are listed in Table
6.4.1-1. The table is organized according to control level (uncontrolled,
moderate, and stringent). The impact on the primary pollutant, NO is given
for each case, as is that of the secondary pollutant, NH3. There is an
insignificant amount of particulates in the flue gas and, therefore, these
are not considered to be a pollutant. There is also no problem with bisul-
fate formation since the fuel has only a trace of sulfur.
Table 6.2.1-13 shows the uncontrolled N0x emission for the natural gas-
fired standard boiler to be less than the SIP control level.
6-25
-------
TABLE 6.4.1-1. AIR POLLUTION IMPACTS FROM BEST NOX FGT CONTROL TECHNIQUES FOR GAS-FIRED BOILERS
5> 6 > 7
Standard boiler
Heat rate
(MBtu/hr) Type Control level
15 Firetube Uncontrolled
Moderate-70%
Stringent-90%
150 Watertube Uncontrolled
^ Moderate-70%
ON
Stringent-90%
g/s
Type of control (Ib/hr)
0.332
(2.63)
SCR Fixed Packed Bed 0.0995
(.789)
SCR Fixed Packed Bed .0332
(.263)
3.31
(26.26)
SCR Fixed Packed Bed 0.993
(7.88)
SCR Fixed Packed Bed 0.331
(2.63)
NOX
ng/J
(Ib/MBtu)
75.4
(.175)
22.6
(.0526)
7.54
(.0175)
75.3
(.175)
22.8
.0525)
7.53
.(.0175)
NH3
g/s
(Ib/hr)
0
0.00113
(.00898)
0.00113
(.00898)
0
0.00511
(.0405)
0.00511
(.0405)
ng/J
(Ib/MBtu)
0
0.257
(.000598)
0.257
(.000598)
0
0.232
(.000540)
0.232
(.000540)
-------
The only environmental impacts are NCI and NH3 emissions. The uncon-
X
trolled NO emissions, 0.332 -°-, are the lowest for all standard boilers
x s
except distillate oil. Moderate - stringent controls reduce this figure to
0.0995-0.0332 •**. NH3 emissions for all control levels are 1 ppm (Figure
S
6.1-1). This corresponds to a mass rate of 0.00113 —.
S
To reduce the adverse environmental impacts, combustion controls utiliz-
ing less 02 minimizes NO formation could be implemented. NHs emissions are
presently quite low. Care needs to be taken to see that an excessive
injection ratio is not used thus increasing the low emission level.
6.4.2 Water Pollution
There are no water streams in SCR fixed packed bed systems.
6.4.3 Solid Waste
The only potential solid waste associated with this system is spent
catalyst. Presently, the life of SCR catalysts is 1-2 years and the topic
of recycling is not addressed in the literature. Since, to date, virtually
no commercial SCR units have operated long enough to require catalyst replace-
o *j
ment, there is no commercial experience on the treatment of spent catalyst.
With the high cost of some of the catalysts, recycling seems to be warranted.
Recycling is feasible where the spent catalyst support is composed of valuable
materials such as titanium, stainless steel, or possibly a ceramic material.
Alumina catalyst supports probably do not warrant recycling unless required
for environmental reasons. These catalysts can probably be disposed of in
the same manner as other industrial catalysts.
The amount of catalyst that must be recycled or disposed of is one
reactor volume since replacement involves total catalyst replacement. The
actual frequency of catalyst replacement is unique to each specific process,
however, catalyst lifetimes are typically one or more years.
6-27
-------
Few process vendors have published their catalyst formulations since
the field is very competitive at the present time. Base metal oxides are
typically used.21* The environmental impact of catalyst disposal will depend
on what materials and compounds are involved.
6.4.4 Other Environmental Impacts
The only other environmental impact is a secondary impact. NHa is
commonly made from natural gas and its consumption is considered a secondary
environmental impact. This impact will be quantified in a forthcoming
report on the impact of NO regulations on the NHs industry. The report is
X
being prepared by TVA under contract to EPA-IERL.
Other adverse environmental impacts (noise, thermal pollution, electri-
cal discharge, etc.) are not present with SCR systems.
6.4.5 Environmental Impacts on Modified and Reconstructed Facilities
The environmental impacts of new and retrofitted systems should be
similar. Retrofitted systems are not so different as to create a greater
adverse environmental impact for these systems.
6-28
-------
REFERENCES
1. Ando, Jumped.. "NO Abatement for Stationary Sources in Japan." EPA
Report Currently in Preparation, October 1978. p. 3-27.
2. Faucett, H.L., et al. Technical Assessment of NO Removal Processes
for Utility Application. EPA-600/7-77-127. November 1977. p. 240.
3. Ibid. , p. 239, 240.
4. Ando, J., op.Git. , p. 3-33.
5. Ibid. , p. 4-4.
6. Faucett, H.L., op.Git., p. 214.
7. Ibid. , p. 259.
8. Ando, J., op. Git. , p. 3-7-
9. Ibid. , p. 4-41.
10. Ibid. , p. 4-95.
11. Wong-Woo, Harmon. "Observation of FGD and Denitrification Systems
in Japan." State of California Air Resources Board-SS-78-004. March
7, 1978. Appendix IV. p. 30.
12. Ibid. , p. 32.
13. Ando, J., op. cit. , p. 4-37.
14. Ibid. , p. 4-38.
15. Ibid. , p. 4-93.
16. Ibid. , p. 4-104.
17. Ibid., p. 4-126.
18. Faucett, H.L., op.cit. , p. 5.
19. Ando, J., op.cit. , p. 3-47.
6-29
-------
20. Ando, J., op.Git., p. 3-55-363.
21. Telephone conversation, Gary Jones of Radian with Jim Turner of
EPA-IERL. 10 January 1979.
22. Ando, Jumpei. "NO Abatement for Stationary Sources in Japan."
EPA-600/7-77-103b. September 1977- p. 47.
23. Ando, J., op.Git., October 1978. pp. 3-61, 3-62.
24. Ibid. , p. 3-2.
25. Ibid. , p. 3-49.
26. Faucett, H.L., op.ait. , p. 302.
27. Ando, J., Letter of April 4, 1979 to J. David Mobley.
6-30
-------
SECTION 7
EMISSION SOURCE TEST DATA
7.1 INTRODUCTION
Test data from operating units are necessary to demonstrate that the
control technology will perform as claimed. For this purpose the most
meaningful test data are those that represent 24 hour averages over 30 days
of continuous operation. At the present time, very little of this type of
information is published. However, some continuous data have been presented
at recent seminars and obtained from the process operators.
The EPA approved test methods are the same for all fuel types and are
discussed here to avoid unnecessary repetition. There are two methods for
measuring NOX (expressed as NOz) in gas streams. First, the EPA Reference
Method 7 is for the determination of nitrogen oxides emissions from station-
ary sources. Presently, this method is the only one approved by the EPA for
measuring NOX levels in flue gas from industrial boilers for emission source
test data.
Method 7 is based on grab sampling for wet chemical analysis and is used
for spot-checking SCR systems' performance and in calibrating an instrument
analyzer. Continuous monitoring by Method 7 for process control purposes is
impractical as the method requires a collected sample to set a minimum of 16
hours. However, continuous data can be developed using Method 7 by taking
samples at several intervals during a 24 hour period and computing a 24 hour
average. Daily values computed in this manner can represent continuous data
when computed for a period of 30 days or more.
7-1
-------
The second method, EPA Reference Method 20, is for the determination of
nitrogen oxide emissions from stationary gas turbines. While this method is
not approved for industrial boilers, it is applicable to continuous monitoring
due to its utilization of an instrumental analyzer based on chemiluminescence.
This instrument provides a sound basis for process control, which is most
important when hourly ambient NOX standards are in effect.
The methodology and test procedures for each method are described in the
following paragraphs. With Method 7, grab samples are collected in an
evacuated flask containing a dilute sulfuric acid-hydrogen peroxide
aOa) absorbing solution. The nitrogen oxides, except nitrous oxide
, are measured colorimetrically using the phenoldisulfonic acid (PDS)
procedure. The apparatus for this system is shown in Figure 7.1-1.
A 25 ml aliquot of absorbing solution is added to the flask. The flask
is stoppered and then evacuated by use of the pump. After checking for leaks,
the probe and vacuum tube are purged using the squeeze bulb. Then the flask
valve is turned to the "sample" position allowing the gas to enter the flask.
When the pressures in the flask and sample line (i.e., duct) are equalized
the flask is isolated, disconnected from the sampling train, and shaken for
5 minutes. The sample flask is allowed to set for a minimum of 16 hours.
After transferring the sample and then washing out the sample flask into a
volumetric flask, 25 ml aliquot is pipetted into a porcelain evaporating
dish. This aliquot is evaporated to dryness on a steam bath and allowed to
cool. Two ml of phenoldisulforic acid solution is added to the dried residue
and the residue is ground to a powder with a polyethylene policeman. After
adding deionized, distillated water and concentrated sulfuric acid, concen-
trated ammonium hydroxide is added dropwise until the pH is 10. The contents
of the flask are mixed thoroughly and the absorbance of a sample is measured
by a spectrophotometer. The total mass of NOX per sample is expressed by
the following equation:
7-2
-------
PROSE
FLASK VALVE.
FILTER
GROUND-GLASS SOCKET.
5 NO. 1Z/S
3-WAY STOPCOCKr
T-BORE. J PYREJ.
2-mm BORE. 8-?nm QO
FLASK
FLASK SHIELD \
SQUEEZE BULB
'MP VALVE
PUMP
THERMOMETER
GROUND-GLASS CONE.
STANDABD TAPER, GROUND-GLASS
I SLEEVE NO. 24/40 SOCKET. § NO. 12/5
F-YREX
•FOAM ENCASEMENT
BOILING FLASK -
2-LITER. ROUMO-SOTTOM. SHORT NECK.
WITH J SLEEVE NO. 24/40
Figure 7.1-1. Sampling train, flask valve, and flask.1
7-3
-------
m = 2 K A
c
where m = mass of NC> as N02 in gas sample, Ug
X
K = spectrophotometer calibration factor
A = absorbance of sample
The sample volume, dry basis, corrected to standard conditions is found by
the equation
P P.
V = 0.3858 -^- fv -25ml) (^ - ^
s mmng \ i y y i _ i.
where V = sample volume, ml
S
Vf = volume of flask and valve, ml
P = final absolute flask pressure, mmHg
T_ = final absolute flask temperature, °K
P. = initial absolute flask pressure, mmHg
T. = initial absolute flask temperature, °K
Finally, the NO concentration in the gas sample is determined by
X
r - in3 mg/m m
C ' 10 Ug/ml
Method 20, for determining nitrogen oxides emissions from stationary gas
turbines, utilizes an instrumental analyzer to which a continuous gas sample
from the exhaust stream is conveyed. The apparatus for this system is shown
in Figure 7.1-2. Particulate matter and water vapor are the primary inter-
ferring species for most instrumental analyzers, but these are removed by
the filter and condenser, respectively, present in the sampling train. In
application to SCR systems on boilers, the presence of NHa may interfere with
the instruments performance. This problem can be circumvented via the use
7-4
-------
of an ammonia decomposition catalyst before the probe measuring the reactor
outlet NO concentration.
\
\
STACK WAU
FILTER
ALT.
' FILTER '
1\ ""»
\
PHOBE
CAIIBSATI
V CAS
\
y
y
HEATSO
SAMPLE
lltiE
' 1 MOIS
TB
1
_
URaCE.'l OXIDES
AKALYZEB
^r1
>•--.
OXYCIM
AtiALYZSa
EXCiSS
SAV,?LSTa V
Figure 7.1-2. Measurement system design for stationary gas turbine tests.2
7.2 EMISSION SOURCE TEST DATA FOR COAL-FIRED BOILERS
At this writing, there have been only a few pilot plant tests performed
in Japan utilizing SCR systems to treat coal-fired flue gas. Little data
from these tests have been released. Most all SCR work has been done recently
in Japan (the oldest SCR system, on an oil-fired boiler, has been operating
for 5 years). Because of Japan's lack of large coal reserves, there are few
coal-fired utility or industrial boilers in the country. However, more coal-
fired boilers are planned for the near future which will utilize imported
coal. This is a result of the scarcity and high cost of cleaner fuels.
There are two coal-fired utility boilers equipped with SCR NO removal
systems due to start-up in 1980.3 Hokkaido Electric Company plans a 90 MW
coal-fired boiler with an SCR unit to be started up at Tomato in March 1980.
The Electric Power Development Corporation has a 250 MW coal-fired SCR unit
due for completion at Takehara in November, 1980. Also, in the United States,
7-5
-------
there are 2 SCR pilot plants presently under construction at coal-fired
utilities located in Tampa, Florida and Albany, Georgia. These are sched-
uled for completion in 1979. Once these units are in operation it will be
possible to obtain more test data.
The available coal-fired source test data is summarized below. Figure
7.2-1 shows the performance of a cylindrical catalyst treating coal-fired
flue gas after particulate removal. Figures 7.2-2 and 7.2-3 show the per-
formance of a parallel flow and a moving bed reactor, respectively, treating
coal-fired flue gas after an ESP. Figure 7.2-4 is for a parallel flow
system. One can see from the plots that the SCR systems are capable of
achieving the stringent level of control.
= 1
8 I
i fr
s "
«
oil
30
20
100
90
80
o o n n n
_L
1000
2000
Time (h)
3000
4000
Figure 7.2-1.
Change of NOX removal efficiency and pressure drop
(Kawasaki Heavy Industries process, Takehara power
station, Hiroshima, Japan) . ^
7-6
-------
Catalyst Plate Type
Gas Temperature 350 °C
(t!H3)/UIOx) 0.83
(Pretreated By Hot EP)
QJ C7I
-(-> O G
300
200
100
0
12
Hot EP Trip
O)
i.
t £
100
50
0
100
90
80
70
1000
2000
Hours (h)
3000
4000
Figure 7.2-2.
Pilot plant test of a parallel flow reactor treating a flue gas
from a coal-fired utility boiler (Hitachi, Ltd. process,
unknown location, Japan).5
-------
Catalyst
Gas Temperature
Dust Concentration
(Pretreated By Hot EP)
Pellet Type
350 oc
20~40 mg/Mm3
to O
-------
6
z
100 r-
PCXX^lXvO.
'2 90
80
Reaction Temp. : 350°c
SV : S.OOOh'1
T_
0
100
200
300
400
500
Time (d)
Figure 7.2-4.
Durability test of NOX removal catalyst
(Kawasaki Heavy Ind. process, Takehara
power station, Hiroshima, Japan).
-------
7.3 EMISSION SOURCE TEST DATA FOR OIL-FIRED BOILERS
While there are a number of commercial SCR systems presently treating
oil-fired flue gas in Japan, the data on these units are limited mostly to
a single reported removal level. Catalyst life tests on heavy oil-fired
pilot unit equipment do provide an indication as to how commercial units
will perform. These continuous test results are shown in Figures 7.3-1
through 7.3-3. Most of the data available are presented as summaries of
pilot test results and are usually expressed in plots of NOX removal (%)
vs. NHs:NO mole ratio or reactor space velocity (hr ). In a few cases,
tables of operating parameters of commercial SCR plants are given. These
results are given in Tables 7.3-1 through 7.3-3 on the following pages.
Most data give only point values of removal and not a set of continuous
data. In addition, the test method and boiler operating conditions are not
given. Included in these figures are data recently obtained from commercial
Japanese installations on industrial boilers and, as such, they represent the
most complete set of continuous data currently available.
Summaries of the oil-fired industrial (larger than 3 MW) and utility
SCR plants in Japan are shown in Tables 7.3-4 and 7.3-5. These tables are
presented since they represent locations where operating data on SCR units
can be obtained. The data can be obtained by either contacting the boiler
owners and requesting available data or by arranging for independent on-site
testing.
The data shown in Figures 7.3-4 and 7.3-5 represent the most recently
available continuous daily average data available from SCR systems applied to
industrial boilers. EPA Method 20 was used to obtain the data in both of the
figures. The average removal level represents the level of control necessary
to meet the local Japanese emission regulations. Continuous daily averages
are not available from most SCR units; however, Table 7.3-6 shows maximum and
minimum NOX removal values for several industrial units.
7-10
-------
100
C
o—
•HOP
o >,
•H 0
•H4I
M-H
•M O
-H-H
Q 4)
to
v)
O
3
in
V)
v
M
a
80 -
70 -
50 -
40
30 -
20 -
10 -
0
Capuclt)
(Nm»/h)
2ttOOO
Fuel
Crude
oil and
heavy
oil
SV value
and the like
RV=6.000kH
3Go r
Hlli/HOx mol.
ratio 1.0
1,000 2.000 3,000 4,000 5,000 6,000 7,"000'
Operation time (hrs)
8,000
Figure 7.3-1.
Catalyst life test results (IHI process,
Taketoyo power station, Japan).7
-------
Catalyst Plate Type
Gas Temperature 350 °C
|(NH3)/(NOx)
0.95
1.0
V
0.95
47 Days
Shut-Down For Boiler Outages
CL
o
100
50
0
Reactor A
Reactor B
25 % Gas Flow Increase
c
-------
X
o
100
90
80 l
Low-s Oil Boiler
Hig'n-s Oil Boiler
0-D-O-CI
OPERATION PERIOD (Hours)
Circled figures show times when SV and NH3/NO mole ratio were changed,
1. SV 10,000 - 15,000 hr"1 3. SV 6,200 - 4,500 hr"1
2. SV 15,000 - 20,000 hr"1 4. SV 4,500-6,200 and the mole
ratio 0.95 - 0.83.
Figure 7.3-3. Test results of oil-fired boilers (Hitachi,
Ltd. process, unknown location, Japan).15
-------
TABLE 7.3-1. OPERATION PARAMETERS OF MAJOR PLANTS CONSTRUCTED BY
HITACHI ZOSEN
10
Completed
Plant site
Gas source
Capacity
(NmVhr)
Load factor (%)
Pretreatment of gas
Reactor inlet
NO (ppm)
A
S0x (ppm)
Dust (mg/Nm3)
02 (%)
Reactor type
Reaction temp.
NOX/NH3 ratio
Catalyst No.
SV (hr'1)
NOX removal (%)
Pressure drop by
SCR reactor (mmH20)
Catalyst life
Idemitsu
Kosan
Oct. 1975
Chiba
FCC-CO
Boiler and
furnace
350,000
50-100
Heating
230
50-80
20-50
2.3
Fixed bed
400
1.0
204
5,000
93
170
1 year
Shindaikyowa
Petrochemical
Nov. 1975
Yokkaichi
Oil-fired
Boiler
440,000
50-100
EP*, FGD,
Heating
150
80-130
30-100
3.2
Fixed bed
420
1.0
304
10,000
80^
160
1 year
Kawasaki
Steel
Nov. 1976
Chiba
Iron-ore
Sintering
machine
762,000
70-100
EP, FGD
WEPT,
Heating
200-300
5-20
3-10
11.2
Fixed bed
1.0
304
4,000
95
50
1 year
*Electrostatic precipitator
'Wet electrostatic precipitator
flncluding leakage in heat exchanger
7-14
-------
TABLE 7.3-2. SCR PLANTS BY MITSUI ENGINEERING & SHIPBUILDING CO.11
Mitsui Petro-
chemical Co.
Ukishima Petro-
chemical Co.
Capacity (Nm3/hr)
Gas composition
NOX (ppm)
SOX (ppm)
Dust (mg/Nm3)
Catalyst and reactor
Catalyst carrier
Catalyst shape
SV (hr'1)
Temperature (°C)
NH3/NOX mole ratio
N0x removal (%)
Total pressure drop (mmH20)
Leak NHa (ppm)
Operation start
Plant cost (10s yen)
Denitrification cost
(yen/kWhr) *
200,000
190
None
20-50
A1203
Granule
2,600
350
1.0
Above 90
Oct. 1975
220,000
150
300
100-150
Ti02
Tube
4,000
350-400
1.0
Above 90
180
Below 10
July 1977
260
*Including 7 years depreciation.
7-15
-------
TABLE 7.3-3. OPERATION DATA OF SCR PLANTS FOR DIRTY GAS12
Gas for SCR (NM3/hr)
Fuel
Load fluctuation
Stack height (m)
Inlet gas composition
02 (%)
SOX (ppm)
NOX (ppm)
Particulates after EP (mg/Nm3)
FGD unit
SV (hr'1)
Temperature (°C)
N0x removal (%)
NH3/NO mole ratio
Leak ammonia (ppm)
Type of reactor
Pressure drop (mmHzO)
Reactor
Total system
Plant completed
Pilot
30,000
Oil(S=0.7%)
60-100%
70
6
400
200
5-20
None
5,000
320
Over 90
1.0
10-20
Fixed bed
July 1973
No.
240,000
Oil(S=0.
60-100%
140
6
400
200
5-10
Scheduled
5,000
320
Over 90
1.0
10-20
Fixed bed
200
500
Mar. 1976
Commercial
1 No. 2
300,000
7%) (011(8-0.7%)
60-100%
140
6
400
200
10-20
Scheduled
5,000
320
Over 90
1.0
10-20
Moving bed
Oct. 1976
7-16
-------
TABLE 7.3-4. OIL-FIRED INDUSTRIAL SCR PLANTS13
Company
Sumitano Chemical
Kurabo
Mitsui Petrochemical
Idemitsu Kosan
Shindaikyowa P . C .
Sumitomo Chemical
Fuji Oil
Sumitomo Chemical
Sumitomo Chemical
Nisshin Steel
Nisshin Steel
Chiyoda Kenzai
Fuji Oil
Ajinomoto
Nippon Oils & Fats
Nippon Yakin
Site
Sodegaura
Hirakata
Chiba
Chiba
Yokkaichi
Sodegaura
Chiba
Sodegaura
Sodegaura
Amagasaki
Amagasaki
Kaizuka
Sodegaura
Kawasaki
Amagasaki
Kawasaki
Capacity
(Nm3/hr)
30,000
30,000
200,000
350,000
440,000
240,000
70,000
300,000
300,000
19,000
20,000
15,000
200,000
180,000
20,000
14,000
Reactor
type*
FPB
MB
FPB
FPB
FPB '
FPB '
PF
MB
MB
FPB
FPB
MB
PF
PF
MB
FPB
Completion
date
July 1973
August 1975
October 1975
October 1975
November 1975
March 1976
July 1976
September 1976
October 1976
July 1977
August 1977
October 1977
January 1978
January 1978
April 1978
July 1978
*FPB = Fixed Packed Bed
MB = Moving Bed
PF = Parallel Flow
7-17
-------
TABLE 7.3-5. OIL-FIRED UTILITY SCR PLANTS
Power company
Kansai Electric
Company C
Kansai Electric
Company A
Kansai Electric
Company D
Company G
Chugoku Electric
Chubu Electric
Tohoku Electric
Site
Kainan
-
Amagasaki
-
Osaka
-
-
Kudamatsu
Chita
Niigata
Capacity
(Nm3/hr)
300,000
1,010,000
410,000
490,000
490,000
1,000,000
1,900,000
1,920,000
1,660,000
Reactor
type*
FPB
PF
-
PF
PF
PF
PF
PF
PF
PF
Completion
date
June 1977
February 1978
June 1978
June 1978
July 1978
July 1978
April 1979
July 1979
February 1980
August 1981
*FPB = Fixed Packed Bed
PF = Parallel Flow
7-18
-------
100 n
90H
o
§
x
o
C
0)
O
M
01
P-i
80H
70H
60.
10
15
Days
I
20
25
30
Figure 7.3-4.
NOX removal for the month of May 1977 (Hitachi Zosen fixed bed process,
Shindaikyowa Petrochemical, Yokkaichi, Japan, Chemiluminescence Method).20
-------
i
(S5
O
§
c§
53
0
0)
O
80
70
60-
50-
40
-------
TABLE 7.3-6. NOX REMOVAL LEVELS AT SEVERAL JAPANESE INDUSTRIAL BOILERS
WITH NOX CONTROL BY SCR
20
Plant owner
Nippon Yakin
Shindaikyowa
Fuji Oil
Fuji Oil
7" Kurabo
NJ
1-1 Nippon Oil & Fats
Kansai Paint
Nisshin Steel
Plant site
Kawasaki
Yokkaichi
Sodegaura
Sodegaura
Hirakata
Amagasaki
Amagasaki
Amagasaki
Capacity*
(MW)
5
135
23
53
10
1
5.3
6.5
Process
Fixed bed + sodium scrubbing
Sodium scrubbing + fixed bed
Fixed (parallel passage) bed
Fixed (honeycomb) bed
Moving bed (continuous)
Moving bed (intermittent)
Fixed bed
Moving bed
Percent NO
removal range
86-98
53-62
93-97
60-80
90-94
94-97
90-92
94-96
^Assumed to be MWP equivalent.
-------
7.4 EMISSION SOURCE TEST DATA FOR GAS-FIRED BOILERS
Although gas-fired boilers, both industrial and utility, are numerous
in Japan, few have been equipped with SCR units so far. This is due to the
fact that combustion modifications on the boilers have been installed because
of their lower cost and the lack of fuel-bound nitrogen to contend with. The
data available on gas-fired SCR systems in Japan are presented in Figures
7.4-1 through 7.4-5 on the following pages. Figure 7.4-1 is a plot of a
long-term performance test while the rest are summaries of pilot tests.
Summaries of the gas-fired industrial and utility SCR plants in Japan
are shown in the table below.
TABLE 7.4-1. GAS-FIRED SCR PLANTS15
Company
Osaka Gas
Chubu Electric
Kyushu Electric
Chubu Electric
Hyushu Electric
Site
Takaishi
Chita
Kokura
Chita
Kokura
Capacity Reactor
(Nm3/hr) type*
15,000x2
1,910,000
1,610,000
1,910,000
1,610,000
FPB
FPB
FPB
FPB
FPB
Completion
date
December 1976
April 1978
July 1978
September 1978
December 1978
*FPB = Fixed Packed Bed
As with oil-fired installations, the type of test data desired may
exist, but has not yet been published. To obtain this data it will be
necessary to contact the boiler owners to possibly get available data or
conduct on-site testing.
7-22
-------
INJ
U>
LHG BOILER
OPERATION PERIOD (Hours)
Circled figure shows time when SV and NH3/NO mole ratio were changed.
1. SV 10,000 - 20,000 hr
-i
Figure 7.4-1. Test results of gas-fired boilers (Hitachi, Ltd.
process, unknown location, Japan).
1 6
-------
N3
100
90
o
z
IU
o 80
o
UJ
a:
X
O
70
60
50
0.2
REACTION TEMPERATURE* 350°C
SV* 10,000 HR"8
I
I
I
0.4
0.6 0.8 1.0 1.2
MOL RATIO OF NH3:NOX
1.4
1.6
Figure 7.4-2. Characteristic curve of the effect of mole ratio of NHs:NOx on
removal efficiency for Hitachi, Ltd. process.
1 7
-------
(0
>
o
cu
o;
100
80
60
320
SV = 5,000
340
360
380
Temperature (*C)
Figure 7.4-3.
Performance of catalyst MTC-102 (Mitsui Toatsu
process, unknown location, Japan).18
o
TOO
80
60
(350°C)
i
5,000
7,500
SV (hr"1)
10,000
Figure 7.4-4. SV and NOX removal (MTC-102) (Mitsui Toatsu
process, unknown location, Japan).
7-25
-------
N3
ON
100
80
o
z
Ul
o
u.
S> 60
UJ
§ 40
O
UJ
o:
C-l CATALYST
SV =20,000 hr
350°C
0.5 1.0 1.5
MOL NH3PER MOL NOX INLET
120^
a.
lOOo-
80 S
CD
60 t
40
Figure 7.4-5. Relationship among inlet NHs:NOX mol ratio, NOX removal efficiency, and exiting
concentration using the Sumitomo Chemical C-l Catalyst.
-------
REFERENCES
1. Environmental Reporter, Appendix A, Oct. 21, 1977, p. 92.
2. Federal Register, Volume 42, Number 191, Oct. 3, 1977, p. 53790.
3. Ando, J. NO Abatement from Stationary Sources in Japan. EPA
draft report in preparation. October 1978, pp. 1-35.
4. Niwa, Senji. Characteristic of Cylindrical DeNOx Catalyst for
a Coal-Fired Boiler. Paper presented at EPRI NO Control Technology
Seminar II. Denver, Colorado. November 8-9, 1978. p. 21-8.
5. Kuroda, H., Nakajima, F. Experiences of NO Removal in Pilot Plants
and Utility Boilers. Paper presented at EPRI NO Control Technology
Seminar II. Denver, Colorado. November 8-9, 19^8. p. 20-13.
6. Ibid., p. 20-10.
7. Ando, J., op. oi,t.3 p. 4-96.
8. Niwa, S., op. cit.3 p. 21-5.
9. Kuroda, H., et al.3 op. oit., p. 2C-12.
10. Ando, J., op. cit., p. 4-21.
11. Ibid., p. 4-71.
12. Ibid., p. 4-5.
13. Ibid., p. 3-4, 3-5.
14. Ibid., p. 1-35
15. Ibid., p. 3-4, 3-5.
16. Ibid., p. 4-43.
17. Faucett, H. L., et al. Technical Assessment of NO Removal Processes
for Utility Applications. EPA 600/7-77-127. November 1977. p. 214.
18. Ando, J., op oit., p. 4-121.
19. Faucett, H. L., op. ait., p. 298.
7-27
-------
APPENDIX 1
DETAILED SYSTEM EVALUATIONS
Al-1
-------
TABLE Al.l. POINT VALUE NO -ONLY PROCESS RATINGS: COAL-FIRED BOILERS - MODERATE CONTROL
Performance
SCR Fixed
Packed Bed 8
SCR Moving
Bed 8
SCR Parallel
Flow 12
Absorption -
Oxidation 7
TABLE
Operation
and
ma in t enanc e
6
4
7
2
Environ-
mental
impact
7
7
7
6
Economic
impact
11
14
14
2
Energy/
material
impact
7
7
8
3
A1.2. POINT VALUE SIMULTANEOUS
COAL-FIRED
i
M
Performance
SCR Parallel
Flow 12
Adsorption *
Electron Beam
Radiation 4
Absorption -
Reduction 10
Oxidation -
Absorption/Reduction 10
Oxidation -
Absorption 8
Operation
and
maintenance
3
5
2
3
4
Environ-
mental
impact
9
3
7
6
6
Boiler
operation
and safety
3
3
3
2
NOX/SOX
Relia-
bility
9
9
14
7
Status of
development
10
10
10
6
Adapt-
ability
3
3
3
3
Compati-
bility
5
5
5
5
Total
69
70
83
43
PROCESS RATINGS:
BOILERS - MODERATE CONTROL
Economic
impact
11
6
10
6
5
Energy/
material
impact
5
4
4
1
1
Boiler
operation
and safety
3
2
4
2
2
Relia-
bility
12
6
7
8
8
Status of
development
13
3
3
10
11
Adapt-
ability
2
3
3
3
3
Compati-
bility
2
5
2
2
5
Total
72
41
52
51
51
*Not Applicable - Does not meet removal requirements.
-------
TABLE A1.3. POINT VALUE NOx~ONLY PROCESS RATINGS: COAL-FIRED BOILERS - STRINGENT CONTROL
Operation Environ- Energy/
SCR Fixed
Packed Bed
SCR Moving
Bed
. SCR Parallel
Flow
Absorption -
Oxidation
and mental
Performance maintenance impact
4 67
4 47
8 77
A
Economic material
impact Impact
8 7
8 8
8 8
Boiler
operation Relia- Status of Adapt- Compati-
and safety bility development ability bility Total
3 9 10 3 5 62
3 9 10 3 5 60
3 14 10 3 5 73
*Not Applicable - Does not meet removal requirements
fc
OJ
SCR Parallel
Flow
Adsorption
Electron Beam
Rad iat ion
Absorption -
Reduction
Oxidation -
Absorption/Reduction
Oxidation -
Absorption
TABLE A1.4. POINT VALUE
COAL -FIRED
Operation Environ-
and mental
Performance maintenance impaCL.
8 39
A
A
A
6 36
6 46
SIMULTANEOUS NOX/SOX PROCESS RATINGS:
BOILERS - STRINGENT CONTROL
Energy/
Economic material
impact Impact
11 5
6 1
5 1
Boiler
operation Relia- Status of Adapt- Compati-
and safety bility development ability bility Total
3 12 13 2 2 68
2 8 10 3 2 48
2 8 11 3 5 49
*Not Applicable - Does not meet removal requirements.
-------
TABLE A 1.5. POINT VALUE NO -ONLY PROCESS RATINGS: COAL-FIRED BOILERS - INTERMEDIATE CONTROL
X
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
Operation
and
Performance maintenance
6 6
6 4
10 7
7 2
Env Iron-
men tal
impact
7
7
7
6
TABLE A 1.6. POINT VALUE
COAL-FIRED
SCR Parallel
Flow
Adsorption
Electron Beam
Radiation
Absorption -
Reduction
Oxidation -
Absorption/Reduction
Oxidation -
Absorption
Operation
and
Performance maintenance
10 3
A
4 5
8 2
8 3
6 4
Environ-
mental
impact
9
3
7
6
6
Economic
impact
11
14
14
2
Energy/
material
impact
7
8
8
3
SIMULTANEOUS
BOILERS
Economic
impact
11
6
10
6
5
Boiler
operation
and safety
3
3
3
2
Relia-
bility
9
9
14
7
NOx/SOx PROCESS
- INTERMEDIATE
Energy/
material
impact
5
4
4
1
1
Boiler
operation
and safety
3
2
4
2
2
Status of
development
10
10
10
6
RATINGS :
Adapt-
ability
3
3
3
3
Compati-
bility Total
5 67
5 69
5 81
5 43
CONTROL
Relia-
bility
12
6
7
8
8
Status of
development
13
3
3
10
11
Adapt-
ability
2
3
3
3
3
Compati-
bility Total
2 70
5 41
2 50
2 49
5 46
*Not Applicable - Does not meet removal requirements.
-------
TABLE Al. 7. POINT VALUE NO -ONLY PROCESS RATINGS: OIL-FIRED BOILERS - MODERATE CONTROL
Performance
SCR Fixed
Packed Bed 8
SCR Moving
Bed 12
SCR Parallel
Flow 12
Absorption -
Oxidation 7
TABLE
Operation
and
maintenance
6
5
7
3
Environ-
mental
impact
7
7
7
6
Economic
Impact
14
14
14
2
Energy/
material
impact
8
9
9
4
A 1.8. POINT VALUE SIMULTANEOUS
OIL-FIRED
i
Ln
Performance
SCR Parallel
Flow 12
Adsorption *
Electron Beam
Radiation 4
Absorption -
Reduction 8
Oxidation -
Absorption/Reduction 10
Oxidation -
Absorption 8
Operation
and
maintenance
4
5
3
4
5
Environ-
mental
impact
9
3
7
6
6
BOILERS
Economic
impact
11
6
9
6
5
Boiler
operation
and safety
3
3
3
2
NOX/SOX
Relia-
bility
11
14
14
8
Status of Adapt-
development ability
16 3
16 3
16 3
13 3
Compati-
bility Total
5 81
5 88
5 90
5 53
PROCESS RATINGS:
- MODERATE CONTROL
Energy/
material
impact
7
4
5
2
2
Boiler
operation
and safety
3
2
4
2
2
Relia-
bility
12
8
7
8
9
Status of Adapt-
development ability
13 2
7 3
10 3
16 3
7 3
Compati-
bility Total
2 75
5 47
2 58
2 59
5 52
*Not Applicable - Does not meet removal requirements.
-------
TABLE A1.9. POINT VALUE NO -ONLY PROCESS RATINGS: OIL-FIRED BOILERS - STRINGENT CONTROL
X
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
Operation Environ-
and mental
Performance maintenance impact
4 67
8 57
8 77
*
Energy/
Economic material
impact impact
8 8
8 9
8 9
Boiler
operation Relia- Status of Adapt-
and safety bility development ability
3 11 16 3
3 14 16 3
3 14 16 3
Compati-
bility Total
5 74
5 81
5 83
*Not Applicable - Does not meet removal requirements
TABLE A1.10. POINT VALUE SIMULTANEOUS
SCR Parallel
Flow
Adsorption
Electron Beam
Radiation
Absorption -
Reduction
Oxidation -
Absorption/Reduction
Oxidation -
Absorption
OIL-FIRED
Operation Environ-
and mental
Performance maintenance impact
8 49
*
*
*
6 46
6 56
N0x/S0x PROCESS RATINGS:
BOILERS - STRINGENT CONTROL
Energy/
Economic material
impact impact
11 7
6 2
5 2
Boiler
operation Relia- Status of Adapt-
and safety bility development ability
3 12 13 2
2 8 16 3
2973
Compati-
bility Total
2 71
2 54
5 50
*Not Applicable - Does not meet removal requirements
-------
TABLE Al.ll. POINT VALUE NO -ONLY PROCESS RATINGS: OIL-FIRED BOILERS - INTERMEDIATE CONTROL
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
Operation
and
Performance maintenance
6 6
10 5
10 7
7 3
Environ-
mental
impact
7
7
7
6
Economic
impact
14
14
14
3
Energy/
material
impact
8
9
9
4
TABLE A1.12. POINT VALUE SIMULTANEOUS
OIL-FIRED
SCR Parallel
Flow
Adsorption
Electron Beam
Radiation
Absorption -
Reduction
Oxidation -
Absor p tion/Reduct ion
Oxidation -
Absorption
Operation
and
Performance maintenance
10 4
*
4 5
8 3
8 4
6 5
Environ-
mental
impact
9
3
7
6
6
BOILERS
Economic
impact
11
6
9
4
3
Boiler
operation
and safety
3
3
3
2
NOx/SOx
- INTERMEDIATE
Energy/
material
impact
7
4
5
2
2
Boiler
operation
and safety
3
2
4
2
2
Relia-
bility
11
14
14
8
PROCESS
CONTROL
Relia-
bility
12
8
7
8
9
Status of
development
16
16
16
13
RATINGS :
Status of
development
13
7
10
16
7
Adapt-
ability
3
3
3
3
Adapt-
ability
2
3
3
3
3
Compati-
bility Total
5 79
5 86
5 88
5 54
Compati-
bility Total
2 73
5 47
2 58
2 55
5 48
*Not Applicable - Does not meet removal requirements.
-------
TABLE A1.13. POINT VALUE NOX~ONLY PROCESS RATINGS: GAS-FIRED BOILERS - MODERATE CONTROL
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
E> TABLE A 1.14
H
CO
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
Operation Environ-
and mental
Performance maintenance impact
12 78
12 58
12 78
8 47
Economic
impact
14
14
14
2
Energy/
material
impact
10
10
10
6
. POINT VALUE NOX-ONLY PROCESS RATINGS:
Operation Environ-
and mental
Performance maintenance impact
8 78
8 58
8 78
*
Economic
impact
8
8
8
Energy/
material
impact
10
10
10
Boiler
operation
and safety
4
4
4
2
Relia-
bility
14
14
14
8
Status of Adapt-
development ability
16 3
16 3
16 3
9 3
GAS-FIRED BOILERS - STRINGENT
Boiler
operation
and safety
4
4
4
Relia-
bility
14
U
14
Status of Adapt-
development ability
16 3
16 3
16 3
Compati-
bility
5
5
5
5
CONTROL
Compati-
bility
5
5
5
Total
93
91
93
58
Total
83
81
83
*Not Applicable - Does not meet removal requirements.
-------
TABLE A1.15. POINT VALUE NO^-ONLY PROCESS RATINGS: GAS-FIRED BOILERS - INTERMEDIATE CONTROL
SCR Fixed
Packed Bed
SCR Moving
Bed
SCR Parallel
Flow
Absorption -
Oxidation
Operation
and
Performance maintenance
10 7
10 5
10 7
8 4
Environ-
mental
impact
8
8
8
7
Economic
impact
14
14
14
2
Energy/
material
impact
10
10
10
6
Boiler
operation
and safety
4
4
4
2
Relia-
bility
14
14
14
8
Status of
development
16
16
16
9
Adapt-
ability
3
3
3
3
Compati-
bility
5
5
5
5
Total
87
85
87
58
-------
APPENDIX 2
EXAMPLE OF TECHNIQUE FOR ECONOMIC SCALING
A2-1
-------
SAMPLE CALCULATION
Most of the available economic data are for utility boilers. Various
base capacities are utilized in the process developers' economic calcula-
tions. For the preliminary economic screening of alternative processes a
base capacity of 20 MW was selected to represent industrial boilers. The
capital cost figures were adjusted to a 20 MW cost by using the six-tenths
rule.1
I = IB *- (Reference 1)
where I = estimated 20 MW investment
I_ = known base investment
D
Q = 20 MW
Q_ = known base capacity
A sample calculation for the SCR-parallel passage (coal) process is shown
below.
Size = 250 MW
Investment = $4,000,000
1= ($4,000,000) '6
= ($4,000,000) (.220)
= $879,000
Capital Cost ($/kW) =
= $44/kW
A2-2
-------
The results of these calculations for all systems considered are contained
in Tables A2.1 and A2.2.
Reference: 1. Rudd, D. F., C. C. Watson, Strategy of Process Engineering,
1968, p.121.
Note: The six-tenths factor was used only during this preliminary phase to
put reported costs for the various processes on a consistent basis.
The six-tenths rule was not used in the development of process
economics for this report.
A2-3
-------
TABLE A2.1. ECONOMICS - CAPITAL COST, $/kW
SCR fixed packed bed
SCR parallel flow
SCR moving bed
SCR parallel flow, NO /SO,,
XX
Adsorption
Electron Beam Radiation
Absorption - Reduction
Absorption - Oxidation
Oxidation - Absorption - Reduction
Oxidation - Absorption
*Includes particulate removal
NA = Not Available
TABLE A2.2. ECONOMICS
SCR fixed packed bed
SCR parallel flow
SCR moving bed
SCR parallel flow, NO /SO
X X
Adsorption
Electron Beam Radiation
Absorption - Reduction
Absorption - Oxidation
Oxidation - Absorption - Reduction
Oxidation - Absorption
Coal
130*
44
92*
475
215
302
413
NA
NA
NA
- OPERATING COST,
Coal
2.1
1.5
2.0
5.0
2.3
NA
7.4
NA
NA
NA
Oil
70
39
70
NA
NA
NA
187
NA
231
NA
MILLS /kWh
Oil
1.9
NA
1.8
NA
NA
NA
5.4
NA
6.4
NA
Gas
27
NA
NA
NA
NA
NA
NA
NA
NA
254
Gas
1.2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA = Not Available
A2-4
-------
APPENDIX 3
MATERIAL BALANCES FOR COAL-FIRED BOILERS
A3-1
-------
MATERIAL BALANCE
Underfeed Stoker
Parallel Flow SCR
Intermediate Control
From
Economizer
^N
i
Steam
*
OJ
NJ
To
Preheater
\/
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
HH,
648
98 , 600
431
67.9
49.9
37.2
0.23
0.25
-
<2>
648
98,040
431
67.9
51.2
37.2
0.05
0.25
•c.Ol
283
615,000
_
_
-
-
-
-
0.19
<£>
289
752,000
_
_
-
-
-
-
0.19
<£>
429
552,000
_
_
0.10
-
-
-
-
^
408
310,000
_
_
0.94
-
-
-
-
-------
MATERIAL BALANCE
Chaingrate Stoker
Parallel Flow SCR
Stringent Control
From
Economizer
XX
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
sox
NH3
<1>
648
98,600
1081
170
125
93.2
0.58
0.63
-
<*>
648
97,770
1082
170
129
93.1
0.06
0.63
0.01
<3>
283
615,000
_
_
-
-
-
-
0.55
<$>
289
752,000
_
_
-
-
-
-
0.55
429
552,000
_
_
0.31
-
_
-
-
<6>
408
310,000
_
_
2.8
_
_
-
-
-------
MATERIAL BALANCE
Chaingrate Stoker
Parallel Flow SCR
Intermediate Control
From
Economizer
I
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
H0x
S°x
NH3
<^
648
98,600
1081
170
125
93.2
0.58
0.63
-
<2>
648
1082
170
128
93.1
0.12
0.63
<.01
<^>
283
615,000
-
-
-
-
-
-
0.48
$>
289
752,000
-
-
-
-
-
-
0~.48
<£>
429
552,000
-
-
0.26
-
-
-
-
<£>
408
310,000
_
_
2.4
-
-
-
-
-------
MATERIAL BALANCE
Chaingrate Stoker
Parallel Flow SCR
Moderate Control
From
Economizer
^N
Steam
Co
Ln
To +
Preheater
^2^-
XX
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
HjO
02
N0x
sox
NH3
648
98,600
1081
170
125
93.2
0.58
0.63
-
<2>
648
1081
170
128
93.1
0.17
0.63
<.01
<3>
283
615,000
-
-
-
-
-
-
0.41
4>
289
752,000
-
-
-
-
-
-
0.41
<£>
429
552,000
-
-
0.23
-
-
-
-
<£>
408
310,000
-
-
2.1
-
-
-
-
-------
MATERIAL BALANCE
Spreader Stoker
Parallel Flow SCR
Intermediate Control
From
Economizer
I
Steam
To
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
»°x
S0x
m,
648
98,600
2160
339
250
186
1.15
1.26
-
<2>
648
97,810
2161
339
256
186
0.23
1.26
0.01
<£>
283
615,000
-.
-
-
-
-
-
0.94
289
752,000
-
-
-
-
-
-
0.94
429
552,000
-
-
0.52
-
-
-
-
<£>
408
310,000
-
-
4.7
-
-
-
-
-------
MATERIAL BALANCE
From
Economizer
To +
Preheater
Reactor
Pulverized Coal
Parallel Flow SCR
Stringent Control
Steam
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
C02
H20
NO
X
S0x
NHj
<£>
648
98,600
2495
453
325
149
1.86
1.69
-
<£>
648
97,150
2497
453
336
149
0.19
1.69
0.03
<3>
283
615,000
-
-
-
-
-
-
1.76
<3>
289
752,000
_
-
-
-
-
-
1.76
<£>
429
552,000
_
-
0.98
-
-
-
-
<£>
408
310,000
_
_
8.8
-
-
-
-
-------
MATERIAL BALANCE
Pulverized Coal
Parallel Flow SCR
Moderate Control
From
Economizer
Steam
LO
00
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
Nil 3
648
98,600
2495
453
325
149
1.86
1.69
-
<2>
648
97,730
2496
453
334
149
0.56
1.69
<.01
<3>
283
615,000
-
-
-
-
-
-
1.32
<$>
288
752,000
-
-
-
-
-
-
1.32
<£>
429
552,000
-
-
0.73
-
-
-
-
<6>
408
310,000
-
-
6.6
-
-
-
-
-------
MATERIAL BALANCE
Underfeed Stoker (High Sulfur Eastern)
Parallel Flow SCR - NOX/SOX
Intermediate Control
From
Economizer
165,000
Naphtha
-------
MATERIAL BALANCE
OJ
I
Underfeed Stoker (Low Sulfur Western)
Parallel Flow SCR - NOX/SOX
Intermediate Control
T, "K
P, Pa
kE-mole/hr
H20
NOX
SDX
NH,
H;
Naphtha
s\
xx
648
98,600
67.9
49.9
0.23
0.25
-
-
-
xx.
\/
648
96,150
51.6
0.05
0.04
0.01
-
-
/x
\/
283
615,000
-
-
-
0.28
-
-
XX
XX
289
752,000
-
-
-
0.28
-
-
XX
XX
429
552,000
0.16
-
-
-
-
-
XX
\/
408
310,000
1.38
-
-
-
-
-
XX
\/
789
665,000
1.51
-
-
-
-
-
/\
V/
666
860,000
-
-
-
-
-
0.07
s\
\/
704
274,000
0.53
-
-
-
1.50
-
s^<
\/
700
177,000
2.03
-
0.21
-
-
-
''^
XX
411
345,000
0.63
-
-
-
-
-
y V
\/
450
170,000
2.03
-
0.21
-
-
-
XV
300
276,000
1.59
-
-
-
-
-
S \
346
276,000
1.59
-
<0.0001
-
-
-
y\
389
165,000
2.03
-
0,21
-
-
-
S \
389
-
-
0.21
-
-
-
-------
MATERIAL BALANCE
>
u>
Pulverized Coal (High Sulfur Eastern)
Parallel Flow SCR - NOX/SOX
Intermediate Control
From
Economizi
T, °K
P, Pa
kR-mole/hr
H30
NOX
SOX
NHs
Ha
Baphtha
<^>
64a
98,600
271
1.57
7.99
-
-
-
<2>
648
96,150
282
0.31
1.20
0.07
-
-
<3>
283
615,000
-
-
-
1.88
-
-
<<>
289
752,000
-
-
1.88
-
-
<5>
429
552,000
1.11
-
-
-
-
-
<6>
408
310,000
9.40
-
-
-
-
-
<'>
789
665,000
48.0
-
-
-
-
-
<»>
666
860,000
-
-
-
-
-
2.23
<9>
704
274,000
16.8
-
-
-
47.9
-
<10>
700
177,000
64.7
-
6.79
-
-
-
<">
411
345,000
19.6
-
-
-
-
-
<12>
450
170,000
64.7
-
6.79
-
-
-
<">
300
276,000
50.4
-
-
-
-
<14>
346
276,000
50.4
-
<0.01
-
-
-
«»
389
165,000
64.7
6.79
-
-
-
<16>
389
•V.1.3X107
-
-
6.79
-
-
-
-------
MATERIAL BALANCE
I
h-'
N>
Pulverized Coal (Low Sulfur Western)
Parallel Flow SCR - NOX/SOX
i u e j. iiit; u -La L e
To /\
Preheater V'
Economizer v' (
UUH
1
I r'
1 |
1 i
I
1 i
1 1
1 1
I 1
1 1
1
f
1
_ruj.
Reactor
(processing)
Reactor
(regener-
ating)
« _
^
,-j
-"."•— S
T rn
Steam /V>
f I -^
-^* £L -^ar-^ V^
i i 1
A 1 D
1 Compressor/
Gasholder
1 V
Water
^ HZ tta^tL *^ Steam
Reformer *-^ Naphtha
?, Pa
k.E-inole/hr
HzO
NOX
SOX
SHs
Hj
Naphtha
98,600
325
1.86
1.69
-
-
-
<2>
96,150
340
0.37
0.25
0.07
-
-
<3>
615,000
-
-
-
2.23
-
-
752,000
-
-
-
2.23
-
V
552,000
1.32
-
-
-
-
-
<{>
310,000
11.2
-
-
-
-
-
<7>
665,000
10.2
-
-
-
-
-
<8>
860,000
-
-
-
-
-
0.47
<'>
274,000
3.55
-
-
-
10,1
-
<10>
177,000
13.7
-
1.44
-
-
-
<">
345,000
4.14
-
-
-
-
-
<12>
170,000
13.7
-
1.44
-
-
-
<">
276,000
10.6
-
-
-
-
-
<»>
276,000
10.6
-
<0.001
-
-
-
«»
165,000
13.7
-
1-44
-
-
-
<">
•vl. 3x10'
-
-
1.44
-
-
-
-------
APPENDIX 4
MATERIAL BALANCES FOR OIL-FIRED BOILERS
A4-1
-------
MATERIAL BALANCE
From
Economizer
I
To ^
Preheater
Reactor
Distillate Oil (4.4 MW)
Fixed Packed Bed SCR
Stringent Control
Steam
NH3
Vaporization
NH3
Storage
I, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
Nil 3
<>>
648
98,600
168
26.3
28.6
5.7
0.023
0.054
-
<2>
648
97,420
168
26.3
28.7
5.7
0.002
0.054
•c.OOl
&
283
615,000
-
-
-
-
-
-
0.020
<$>
289
752,000
-
-
-
-
-
-
0.020
<£>
429
552,000
-
-
0.011
-
-
-
-
<£>
408
310,000
-
-
0.101
-
-
-
-
-------
MATERIAL BALANCE
Distillate Oil (4.4 MW)
Fixed Packed Bed SCR
Moderate Control
From
Economizer
Steam
i
to
To
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg -mole/ hr
N2
C02
H20
02
N0x
SOX
NH3
<»
648
98,600
168
26.3
28.6
5.7
0.023
0.054
-
<2>
648
97,890
168
26.3
28.7
5.7
0.007
0.054
<.001
<3>
283
615,000
-
-
-
-
-
-
0.016
<£>
289
752,000
_
-
-
-
-
-
0.016
<£>
429
552,000
_
-
0.009
-
-
-
-
<£>
408
310,000
_
_
0.078
-
-
-
-
-------
MATERIAL BALANCE
Distillate Oil (44 MW)
Fixed Packed Bed SCR
Stringent Control
From
Economizer
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
N2
C0?
II ?0
02
HOX
S°x
N113
«>
1470
230
250
50.0
0.24
0.54
-
1470
230
252
50.0
0.02
0.54
<0.01
-
-
-
-
-
-
0.211
<£>
-
-
-
-
-
-
0.211
-
-
0.132
-
-
-
-
-
-
1.19
-
-
-
-
-------
MATERIAL BALANCE
Distillate Oil (44 MW)
Fixed Packed Bed SCR
Moderate Control
From
Economizer
Steam
To
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, "K
P, Pa
Nz
CO 2
I120
02
N0x
S0x
NII3
4>
1470
230
250
50.0
0.24
0.54
-
<2>
1470
2JO
252
50.0
0.08
0.54
<0.01
<3>
-
-
-
-
-
-
0.164
<$>
-
-
-
-
-
-
0.164
_
-
0.103
-
-
-
-
<6>
-
-
0.923
-
-
-
-
-------
MATERIAL BALANCE
Residual Oil (8.8 MW)
Parallel Flow SCR
Stringent Control
From
Economizer
-
XX
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
C02
H20
02
N0x
S0x
NH3
<$>
648
98,600
297
49.8
44.2
10.2
0.17
0.67
-
<§>
648
97,890
297
49.8
45.2
10.2
0.02
0.67
<0.01
<^>
283
615,000
_
_
-
-
-
-
0.16
<$>
289
752,000
_
-
-
-
-
-
0.16
<£>
429
552,000
_
-
0.09
-
-
-
-
<$>
408
310,000
_
_
0.78
-
-
-
-
-------
MATERIAL BALANCE
Residual Oil (8.8 MW)
Parallel Flow SCR
Moderate Control
From
Economizer
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
P, Pa
kg-mole/hr
N2
CO 2
H20
02
NOX
sox
NH3
<1>
648
98,600
297
49.8
44.2
10.2
0.17
0.67
-
<2>
648
98,180
297
49.8
45.0
10.2
0.05
0.67
<0.01
<3>
283
615,000
-
-
-
-
-
-
0.12
<$>
289
752,000
_
-
-
-
-
-
0.12
429
552,000
_
-
0.07
-
-
_
-
<£>
408
310,000
_
-
0.60
-
-
-
-
-------
MATERIAL BALANCE
Residual Oil (8.8 MW)
Moving Bed SCR
Stringent Control
From
Economizer
i
oo
To
Preheater
Steam
^1
Reactor
As
Catalyst
Handling
h
NH3
Vaporization
I, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
NH3
4>
648
98,600
297
49.8
44.2
10.2
0.17
0.67
"
648
98,160
297
49.8
45.2
10.2
0.02
0.67
<0.01
<^>
283
615,000
-
-
-
-
-
-
0.16
<£>
289
752,000
_
_
-
-
-
-
0.16
^
429
552,000
_
_
0.09
_
_
-
-
[ ^ |
408
310,000
0.78
_
_
-
-
NH3
Storage
-------
MATERIAL BALANCE
Residual Oil (8.8 MW)
Moving Bed SCR
Moderate Control
From X\ \
Economizer N/
To /^\
Preheater XX
Reactor
f
\
^
^
^ XI
^
/
^
"\
\^
J
As
Catalyst
Handling
^
P
c_^ r
vy ' v
f ~\ /T\
k^ \/
h
NH3 NH3
Vaporization Storage
T, °K
P, Pa
N2
C02
11 20
Oj
NOX
SOX
Nil 3
<^>
648
98,600
297
49.8
44.2
10.2
0.17
0.67
-
«>
648
98,340
297
49.8
45.0
10.2
0.05
0.67
<0.01
<3>
283
615,000
-
-
-
-
_
-
0.12
<£>
289
752,000
-
-
-
-
_
-
0.12
<5>
429
552,000
-
-
0.07
-
_
-
<6>
408
310,000
_
-
0.60
_
_
-
-------
MATERIAL BALANCE
Residual Oil (44 MW)
Parallel Flow SCR
Stringent Control
From
Economizer
Steam
>
I
l->
o
To
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
Nil 3
<»
648
98,600
1489
250
222
51.1
0.59
3.34
-
<3>
648
97,460
1490
250
226
51.0
0.06
3.34
0.02
<£>
283
615,000
-.
-
-
-
-
-
0.56
<§>
289
752,000
-
-
-
-
-
-
0.56
<£>
429
552,000
-
-
0.31
-
-
-
-
408
310,000
-
-
2.8
-
-
-
-
-------
MATERIAL BALANCE
Residual Oil (44 MW)
Parallel Flow SCR
Moderate Control
From
Economizer
Steam
To ^
Preheater
Reactor
i
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-mole/hr
Na
C02
U20
02
N0x
S0x
Nil 3
648
98,600
1489
250
222
51.1
0.59
3.34
-
<2>
648
97,910
1489
250
225
51.0
0.18
3.34
<.01
<3>
283
615,000
-
-
-
-
-
-
0.42
<$>
289
752,000
-
-
-
-
-
-
0.42
<5>
429
552,000
-
-
0.19
-
-
-
-
<^>
408
310,000
-
-
2.1
-
-
-
-
-------
MATERIAL BALANCE
From _____
Economizer
•
XX
To ^_
Preheater
Reactor
Residual Oil (44 MW)
Moving Bed SCR
Stringent Control
Steam
NJ
I
Ash
Catalyst
Handling
NH3
Vaporization
NH3
Storage
1, °K
P, Pa
kg-mole/hr
N2
CO 2
H20
02
N0x
S0x
NH3
0>
6^18
98,600
1489
250
222
51.1.
0.59
3,34
-
648
97,850
1490
250
226
51.0
0.06
3.34
0.02
<£>
283
615,000
_
_
_
_
_
-
0.56
<2>
289
752,000
_
_
_
_
_
-
0.56
429
552,000
_
_
0.31
_
_
-
-
<^>
408
310,000
_
_
2.8
_
_
-
-
-------
MATERIAL BALANCE
Residual Oil (44 MW)
Moving Bed SCR
Moderate Control
From
Economizer
\f
To
Preheater
<^
Reactor
Steam
{
sh
As
Catalyst
Handling
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
kg-raole/hr
N2
CO 2
H20
02
S0x
Nil 3
648
98,600
1489
250
222
51.1
0.59
3.34
-
648
98,150
1489
250
225
51.0
0.18
3.34
<.01
<2>
283
615,000
-
-
-
-
-
-
0.42
<£>
289
752,000
-
-
-
-
-
-
0.42
<£>
429
552,000
-
-
0.19
-
-
-
-
<6>
408
310,000
-
-
2.1
-
-
-
-
-------
MATERIAL BALANCE
I
I—'
-P>
Residual Oil (44 MW)
Parallel Flow SCR - NOX/SOX
Intermediate Control
?, Pa
BjO
SO,
sox
NH3
Hj
Naphcha
/\
\/
98,600
222
0.59
3.34
-
-
-
/2\
\/
96,150
226
0.12
0.50
0.04
-
-
/x
\/
615,000
-
-
-
0.71
-
-
/TV
\/
752,000
-
-
-
0.71
-
-
/\
\_/
552,000
0.42
-
-
-
-
-
/\
\/
310,000
3.54
-
-
-
-
/\
\/
665,000
20.1
-
-
-
-
S?\
\/
860,000
-
-
-
-
-
0.93
/\
\/
274,000
7.01
-
-
-
20.0
-
S\
\/
177,000
27.0
-
2.84
-
-
-
/\
345,000
8.17
-
-
-
-
-
/-\
170,000
27.0
-
2.84
-
-
-
S\^
276,000
21.0
-
-
-
-
-
~/\
276,000
21.0
<0.01
-
-
-
S\^
165,000
27.0
-
2.84
-
-
-
~ ~ y'NT
11.3x10'
-
-
2.84
-
-
-
-------
APPENDIX 5
MATERIAL BALANCES FOR NATURAL GAS-FIRED BOILERS
A5-1
-------
MATERIAL BALANCE
Natural Gas (4.4 MW)
Fixed Packed Bed SCR
Stringent Control
From
Economizer
^N
Ul
K>
TO ^
Preheater
Steam
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P. Pa
kg-mole/hr
N2
CO 2
H20
02
N0y
S0x
t-IHj
648
98,600
183
21.4
46.6
6.3
0.026
Cr
-
<£>
648
97,350
183
21.4
46.7
6.3
0.003
tr
<.001
<£>
283
615,000
-
-
-
-
-
-
0.023
<3>
289
752,000
-
-
-
-
-
-
0.023
429
552,000
-
-
0.013
-
-
-
-
<£>
408
310,000
-
-
0.11
-
-
-
-
-------
MATERIAL BALANCE
Natural Gas (4.4 MW)
Fixed Packed Bed SCR
Moderate Control
From
Economizer
Ln
I
OJ
To ^
Preheater
Reactor
Steam
1
NH3
Vaporization
NH3
Storage
T, "K
P, Pa
N2
CO 2
H20
02
S0x
NH3
<1>
450
98,600
183
21.4
46.6
6.3
0.026
tr
-
<2>
450
97,850
183
21.4
46.7
6.3
0.008
tr
<.001
<5>
283
615,000
-
-
-
-
-
-
0.018
<£>
289
752,000
-
-
-
-
-
-
0.018
<£>
429
552,000
-
-
0.010
-
-
-
-
<£>
408
310,000
-
-
0.088
-
-
-
-
-------
MATERIAL BALANCE
Natural Gas (44 MW)
Fixed Packed Bed SCR
Stringent Control
From
Economizer
Steam
To +
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
N2
CO 2
H20
02
N0x
S0x
Nil 3
^_^_
1580
185
402
54.4
0.28
tr
-
1580
185
404
54.4
0.02
tr
<0.01
<&
_.
-
-
-
-
-
0.233
<$>
-
-
-
-
-
-
0.233
-
-
0.146
-
-
-
-
&
- " .
-
1.32
-
-
-
-
-------
MATERIAL BALANCE
Natural Gas (44 MW)
Fixed Packed Bed SCR
Moderate Control
From
Economizer
Steam
XX
Ln
Ui
To
Preheater
Reactor
NH3
Vaporization
NH3
Storage
T, °K
P, Pa
Nj
CO 2
1I20
02
N0x
S0x
Nil 3
<»
1580
185
402
54.4
0.28
tr
-
<>>
1580
185
404
54.4
0.08
tr
<0.01
-
-
-
-
-
-
0.181
<$>
-
-
-
-
-
-
0.181
<5^>
-
-
0.113
-
-
-
-
<£>
-
-
1.02
-
-
-
-
-------
APPENDIX 6
CAPITAL COST BREAKDOWNS
A6-1
-------
Table A6-1. CAPITAL COSTS
Boiler type: Underfeed
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Equipment cost
Basic equipment (includes freight) 29.270
Required auxiliaries 24.200
Total equipment cost 53.470
Installation costs, direct
Foundations and supports 4,290
Piping 14.080
Insulation 1,760
Painting 330
Electrical 1,350
Instruments 2,430
Installation labor 18,500
Total installation cost 42,740
Total Direct Costs (equipment + installation) 96.210
Installation costs, indirect
Engineering 43,892
Construction and field expense 9,621
Construction fees 9,621
Start-up 1,924
Performance tests 2.000
Total Indirect Costs 67,058
Contingencies 32,654
Total lurnkey Costs (direct+indirect+contingencies)195,922
Land 490
Working capital 16,675
GRAND TOTAL (turnkey + land + working capital) $213,090
A6-2
-------
Table A6-2. CAPITAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight)61,510
Required auxiliaries 81,000
Total equipment cost 142,510
Installation costs, direct
Foundations and supports 9,190
Piping 29,010
Insulation 3,900
Painting 7QQ
Electrical 2,570
Instruments 4,820
Installation labor 41,310
Total installation cost 82,310
Total Direct Costs (equipment + installation) 224,820
Installation costs, indirect
Engineering 43,892
Construction and field expense 22.482
Construction fees 22.482
Start-up 4.496
Performance tests 2rOQO
Total Indirect Costs 95,352
Contingencies 64,034
Total Turnkey Costs (direct+indirect+contingencies) 384,206
Land 961
Working capital 28,129
GRAND TOTAL (turnkey + land + working capital) $413,300
A6-3
-------
Table A6-3. CAPITAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Equipment cost
Basic equipment (includes freight) 49.880
Required auxiliaries 60.600
Total equipment cost 110,480
Installation costs, direct
Foundations and supports 7,430
Piping 23,470
Insulation 3.090
Painting 57Q
Electrical 2.100
Instruments 3r930
Installation labor 32,920
Total installation cost 73,510
Total Direct Costs (equipment + installation) 183,990
Installation costs, indirect
Engineering 43,892
Construction and field expense 18,399
Construction fees 18,399
Start-up 3,680
Performance tests 2,000
Total Indirect Costs 86,370
Contingencies 54,072
Total Turnkey Costs (direct+indirect+contingencies)324,432
Land 811
Working capital 24,294
GRAND TOTAL (turnkey + land + working capital) $349,540
A6-4
-------
Table A6-4. CAPITAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 42,000
Required auxiliaries 48,500
Total equipment cost 90,500
Installation costs, direct
Foundations and supports 6,150
Piping 19,270
Insulation 2,490
Painting 480
Electrical 1,780
Instruments 3,260
Installation labor 27,170
Total installation cost 60 .600
Total Direct Costs (equipment + installation) 151r100
Installation costs, indirect
Engineering 43,892
Construction and field expense 15,110
Construction fees 15,110
Start-up 3.022
Performance tests 2.000
Total Indirect Costs 79,134
Contingencies 46,047
Total Turnkey Costs (direct+indirect+contingencies) 276,281
Land 691
Working capital 21,894
GRAND TOTAL (turnkey + land + working capital) $298,870
A6-5
-------
Table A6-5. CAPITAL COSTS
Boiler type: Spreader Stoker
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Equipment cost
Basic equipment (includes freight) 74.330
Required auxiliaries 121.000
Total equipment cost 195r330
Installation costs, direct
Foundations and supports 10,260
Piping 32,610
Insulation 4.130
Painting 800
Electrical 2f950
Instruments 5,600
Installation labor 49 790
Total installation cost 106.140
Total Direct Costs (equipment + installation) 301.470
Installation costs, indirect
Engineering 43,892
Construction and field expense 30.147
Construction fees 30,147
Start-up 6.029
Performance tests 2.000
Total Indirect Costs 112.215
Contingencies 82.737
Total larnkey Costs (direct+indirect+contingencies) 496,422
Land 1,241
Working capital 36,958
GRAND TOTAL (turnkey + land -I- working capital) $534,620
A6-6
-------
Table A6-6. CAPITAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight)104,470
Required auxiliaries 188,000
Total equipment cost 292.470
Installation costs, direct
Foundations and supports 14,120
Piping 44,030
Insulation 5,560
Painting 1,090
Electrical 4,020
Instruments 7.770
Installation labor 69.860
Total installation cost 146,450
Total Direct Costs (equipment + installation) 438,920
Installation costs, indirect
Engineering 43,892
Construction and field expense 43,892_
Construction fees 43.092
Start-up 8.778
Performance tests 2rOOO
Total Indirect Costs 142.454
Contingencies 116,275
Total Turnkey Costs (direct+indirect+contingencies) 697.649
Land 1,744
Working capital 53,013
GRAND TOTAL (turnkey + land + working capital) $752,410
A6-7
-------
Table A6-7. CAPITAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 78,470
Required auxiliaries 113,000
Total equipment cost 191,470
Installation costs, direct
Foundations and supports 10,540
Piping 33.160
Insulation 4.130
Painting 830
Electrical
Instruments 6,110
Installation labor 49,780
Total installation cost 107.600
Total Direct Costs (equipment + installation) 299,070
Installation costs, indirect
Engineering 43.892
Construction and field expense 29.907
Construction fees 29.907
Start-up 5r981
Performance tests 2,000
Total Indirect Costs 111,687
Contingencies 82,151
Total Turnkey Costs (direct+indirect+contingencies) 492 ,908
Land 1.232
Working capital 31 ,757
GRAND TOTAL (turnkey + land + working capital) $531,900
A6-8
-------
Table A6-8. CAPITAL COSTS
Boiler type: Underfeed Stoker
Fuel: High Sulfur Eastern Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Total Direct Costs (equipment + installation) 1,284,000
Installation costs, indirect
Engineering 373,400
Construction and field expense 128,400
Construction fees 128.400
Start-up 25,700
Performance tests 4.000
Total Indirect Costs 660,000
Contingencies 389.000
Total Turnkey Costs (direct+indirect+contingencies) 2.333rOOP
Land 6.000
Working capital 64,000
GRAND TOTAL (turnkey + land + working capital) $2,403,000
A6-9
-------
Table A6-9. CAPITAL COSTS
Boiler type: Underfeed Stoker
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Total Direct Costs (equipment + installation) 648>800
Installation costs, indirect
Engineering 373,400
Construction and field expense 64,900
Construction fees 64,900
Start-up 13,000
Performance tests 4.000
Total Indirect Costs 520.2QO
Contingencies 233,800
Total Turnkey Costs (direct+indirect+contingencies) 1,403,000
Land 4,QQQ
Working capital 43.QQQ
GRAND TOTAL (turnkey + land + working capital) $1,450,000
A6-10
-------
TABLE A6-10. CAPITAL COSTS
Boiler type: Pulverized Coal
Fuel: High Sulfur Eastern Coal
Control technique: Parallel Flew SCR (NOX/SOX)
Control level: Intermediate
Total Direct Costs (equipment + installation) 3,734,000
Installation costs, indirect
Engineering 373,400
Construction and field expense 373,400
Construction fees 373,400
Start-up 74,700 -
Performance tests 4,000
Total Indirect Costs 1,199,000
Contingencies 987'000
Total Turnkey Costs (direct+indirect+contingencies)5,920,OOP
Land 15,000
Working capital 180,000
GRAND TOTAL (turnkey + land + working capital) $6,115,000
A6-11
-------
TABLE A6-11. CAPITAL COSTS
Boiler type: Pulverized Coal
Fuel: LOW Sulfur Western Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Total Direct Costs (equipment + installation) 1,793,000
Installation costs, indirect
Engineering 373,400
Construction and field expense 179,300
Construction fees 1 79,300
Start-up 3539QO
Performance tests 4.QQQ
Total Indirect Costs 772.000
Contingencies 513,000
Total Turnkey Costs (direct+indirect+contingencies) 3,078,OOP
Land 8,000
Working capital 78,000
GEAND TOTAL (turnkey + land + working capital) $3.164,000
A6-12
-------
Table A6-12. CAPITAL COSTS
Boiler type: Firetube (4.4 MWfc)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 8,230
Required auxiliaries 8,580
Total equipment cost 16,810
Installation costs, direct
Foundations and supports 850
Piping 2,990
Insulation 360
Painting 60
Electrical 580
Instruments 470
Installation labor 3.910
Total installation cost 9,220
Total Direct Costs (equipment -1- installation) 26,030
Installation costs, indirect
Engineering 43.892
Construction and field expense 2.603
Construction fees 2.603
Start-up 521
Performance tests 2fOOP
Total Indirect Costs 51,619
Contingencies 11,647
Total Turnkey Costs (direct+indirect+contingencies) 89.296
Land 223
Working capital 9,892
GRAND TOTAL (turnkey + land + working capital) $99,410
A6-13
-------
Table A6-13. CAPITAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 6,230
Required auxiliaries 5f 120
Total equipment cost 11,350
Installation costs, direct
Foundations and supports _ 620
Piping 2.270
Insulation _ 280
Painting _ 40
Electrical _ son
Instruments
Installation labor 2.980
Total installation cost 7,040
Total Direct Costs (equipment + installation) 18,390
Installation costs, indirect
Engineering 43.892
Construction and field expense 1.839
Construction fees 1,839
Start-up 368
Performance tests 2,000
Total Indirect Costs 49,938
Contingencies 10,248
Total Turnkey Costs (direct+indirect+contingencies) 78.576
Land 196
Working capital 9,402
GRAND TOTAL (turnkey + land + working capital) $88,170
A6-14
-------
Table A6-14. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 34,030
Required auxiliaries 86,140
Total equipment cost 120,170
Installation costs, direct
Foundations and supports 3,280
Piping 10,420
Insulation 1,150
Painting 220
Electrical 1.660
Instruments ]_ 690
Installation labor 21,300
Total installation cost 39,720
Total Direct Costs (equipment + installation) 159,890
Installation costs, indirect
Engineering 43,890
Construction and field expense 15.990
Construction fees 15.990
Start-up 3.200
Performance tests 2.000
Total Indirect Costs 81.070
Contingencies 36,140
Total Turnkey Costs (direct+indirect+contingencies) 277,1 DO
Land 690
Working capital 27,810
GRAND TOTAL (turnkey + land + working capital) 305,600
A6-15
-------
Table A6-15. CAPITAL COSTS
Boiler type: Watertube (44 MWfc)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 25,400
Required auxiliaries 51,680
Total equipment cost 77,080
Installation costs, direct
Foundations and supports 2.500
Piping 8.000
Insulation 900
Painting 180
Electrical 1 ,280
Instruments 1,320
Installation labor 14,930
Total installation cost 29,110
Total Direct Costs (equipment + installation) 106,190
Installation costs, indirect
Engineering 43,890
Construction and field expense 10.620
Construction fees 10.620
Start-up 2.120
Performance tests 2rQOO
Total Indirect Costs . 69.250
Contingencies 26.320
Total Turnkey Costs (direct+indirect+contingencies)201.760
Land 500
Working capital 21,510
GRAND TOTAL (turnkey + land + working capital) 223,800
A6-16
-------
Table A6-16. CAPITAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 29.100
Required auxiliaries 23,800
Total equipment cost 52.900
Installation costs, direct
Foundations and supports 4 339
Piping 14,080
Insulation i R7Q
Painting 460
Electrical 1.240
Instruments 2r350
Installation labor 18r580
Total installation cost 43 000
Total Direct Costs (equipment + installation) 95,900
Installation costs, indirect
Engineering 43.890
Construction and field expense 9.590
Construction fees 9.590
Start-up 1.920
Performance tests 2.000
Total Indirect Costs 67»QOQ
Contingencies 24,400
Total Turnkey Costs (direct+indirect+contingencies)187,300
Land 500
Working capital _ 15,100
GRAND TOTAL (turnkey + land + working capital) $202,900
A6-17
-------
Table A6-17- CAPITAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 20,910
Required auxiliaries 14,300
Total equipment cost 35,210
Installation costs, direct
Foundations and supports 2.830
Piping 9.850
Insulation 1.3QQ
Painting 240
Electrical I(QIQ
Instruments 1.670
Installation labor 12.820
Total installation cost 29.720
Total Direct Costs (equipment + installation) 64,930
Installation costs, indirect
Engineering 43,890
Construction and field expense 6,490
Construction fees 6,490
Start-up 1,300
Performance tests 2,000
Total Indirect Costs 60,170
Contingencies 18.770
Total larnkey Costs (direct+indirect+contingencies)143.900
Land 400
Working capital 13.800
GRAND TOTAL (turnkey + land + working capital) $158.100
A6-18
-------
Table A6-18. CAPITAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: stringent
Equipment cost
Basic equipment (includes freight) 21.340
Required auxiliaries 15r740
Total equipment cost 37,080
Installation costs, direct
Foundations and supports 3,070
Piping 9,870
Insulation 1,290
Painting 250
Electrical 1,010
Instruments 1,710
Installation labor 12.100
Total installation cost 29.300
Total Direct Costs (equipment + installation) 66,380
Installation costs, indirect
Engineering 43,890
Construction and field expense 6.640
Construction fees 6. 640
Start-up 1.330
Performance tests 2.000
Total Indirect Costs 60,500
Contingencies 19,030
Total Turnkey Costs (direct+indirect+contingencies)145,910
Land 4QQ
Working capital 20,350
GRAND TOTAL (turnkey + land + working capital) $166,700
A6-19
-------
Table A6-19. CAPITAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 17,440
Required auxiliaries 9,440
Total equipment cost 26,880
Installation costs, direct
Foundations and supports 2,540
Piping 8.090
Insulation 1.Q60
Painting _ 200
Electrical _ 870
Instruments i
Installation labor 9,810
Total installation cost 23.960
Total Direct Costs (equipment + installation) 50,840
Installation costs, indirect
Engineering 43,890
Construction and field expense 5, 080
Construction fees 5,080
Start-up 1.020
Performance tests 2,000
Total Indirect Costs 57,070
Contingencies 16,190
Total Turnkey Costs (direct+indirect+contingencies)124 , 100
Land 300
Working capital 19,190
GRAND TOTAL (turnkey + land + working capital) $144,600
A6-20
-------
Table A6-20. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 74,670
Required auxiliaries 113.000
Total equipment cost 187,670
Installation costs, direct
Foundations and supports 10, 77Q
Piping 34.420
Insulation 4.620
Painting 820
Electrical 3 ISO
Instruments 5 66D
Installation labor ^0 7?n
Total installation cost 110.190
Total Direct Costs (equipment + installation) 297,860
Installation costs, indirect
Engineering 43, 892
Construction and field expense 29,786
Construction fees ?q a 786
Start-up 5,957
Performance tests 2tQQQ
Total Indirect Costs 111.421
Contingencies 61,392
Total Turnkey Costs (direct+indirect+contingencies) 470,673
Land 1,177
Working capital 31,290
GRAND TOTAL (turnkey + land + working capital) $503.140
A6-21
-------
Table A6-21. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 54,980
Required auxiliaries 68.100
Total equipment cost 123.080
Installation costs, direct
Foundations and supports 7,900
Piping 25,320
Insulation 3.310
Painting 600
Electrical 2.360
Instruments 4r310
Installation labor 35 890
Total installation cost 79,690
Total Direct Costs (equipment + installation) 202,770
Installation costs, indirect
Engineering 43,892
Construction and field expense 20,277
Construction fees 20,277
Start-up 4.055
Performance tests 2,000
Total Indirect Costs 90,501
Contingencies 43,991
Total T -rnkey Costs (direct+indirect+contingencies) 337.262
Land 843
Working capital 23,456
GRAND TOTAL (turnkey + land + working capital) $378.190
A6-22
-------
Table A6-22. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 34,650
Required auxiliaries 89,550~
Total equipment cost 124,200
Installation costs, direct
Foundations and supports 6,620
Piping 20,870
Insulation 2.650
Painting 510
Electrical 2,050
Instruments 3.620
Installation labor 26,110
Total installation cost 62.430
Total Direct Costs (equipment + installation) 186.630
Installation costs, indirect
Engineering 43.892
Construction and field expense 18,663
Construction fees 18.663
Start-up 3.733
Performance tests 2,000
Total Indirect Costs 86,951
Contingencies 41.037
Total Turnkey Costs (direct+indirect+contingencies) 314.617
Land 787
Working capital 31,621
GRAND TOTAL (turnkey + land + working capital) $347,020
A6-23
-------
Table A6-23. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 26,170
Required auxiliaries 57.550
Total equipment cost 83,720
Installation costs, direct
Foundations and supports 4.740
Piping 14.930
Insulation 1.880
Painting 370
Electrical 1,510
Instruments 2,650
Installation labor ia 570
Total installation cost 44.650
Total Direct Costs (equipment + installation) 128.370
Installation costs, indirect
Engineering 43,892
Construction and field expense 12.837
Construction fees 12.837
Start-up 2.567
Performance tests 2.000
Total Indirect Costs 74,133
Contingencies 30,376
Total Turnkey Costs (direct+indirect+contingencies) 232,883
Land 582
Working capital 26.223
GRAND TOTAL (turnkey + land -I- working capital) S259.690
A6-24
-------
Table A6-24. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Total Direct Costs (equipment + installation) 2,323,000
Installation costs, indirect
Engineering 373r400
Construction and field expense 232.300
Construction fees 232.300
Start-up 46.500
Performance tests 4.000
Total Indirect Costs 889,000
Contingencies 482.000
Total Turnkey Costs (direct+indirect+contingencies) 3.693.000
Land 9.000
Working capital 99,000
GRAND TOTAL (turnkey + land + working capital) $3.801.000
A6-25
-------
Table A6-25. CAPITAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 8,530
Required auxiliaries 8.900
Total equipment cost 17.430
Installation costs, direct
Foundations and supports _ 870
Piping 3,050
Insulation
Painting
Electrical
Instruments _ 470
Installation labor 3.99Q
Total installation cost 9,390
Total Direct Costs (equipment + installation) 26,820
Installation costs, indirect
Engineering 43,892
Construction and field expense 2,682
Construction fees 2,682
Start-up 536
Performance tests 2,000
Total Indirect Costs 51,792
Contingencies n. 792
Total '- jrnkey Costs (direct+indirect+contingencies) 90 , 404
Land 226
Working capital 9,893
GRAND TOTAL (turnkey + land + working capital) $100,520
A6-26
-------
Table A6-26. CAPITAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight) 6,580
Required auxiliaries 5. 320
Total equipment cost 11,900
Installation costs, direct
Foundations and supports 650
Piping 2.330
Insulation 280
Painting 4Q
Electrical 510
Instruments 360
Installation labor 5r59Q
Total installation cost 9,760
Total Direct Costs (equipment + installation) 21,660
Installation costs, indirect
Engineering 43.892
Construction and field expense 2.166
Construction fees 2.166
Start-up 433
Performance tests 2,000
Total Indirect Costs 50,657
Contingencies 10,848
Total Turnkey Costs (direct+indirect+contingencies) 83,165
Land 208
Working capital 9,391
GRAND TOTAL (turnkey + land + working capital) $92,760
A6-27
-------
Table A6-27. CAPITAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 33,410
Required auxiliaries 87,700
Total equipment cost 121,110
Installation costs, direct
Foundations and supports 3,310
Piping 10.480
Insulation lr160
Painting 220
Electrical 1,670
Instruments 1.710
Installation labor 21.380
Total installation cost 39r 930
Total Direct Costs (equipment + installation) 161.040
Installation costs, indirect
Engineering 43.890
Construction and field expense 16.100
Construction fees 16.100
Start-up 3.220
Performance tests 2. OOP
Total Indirect Costs 81,310
Contingencies 3.6,350
Total Turnkey Costs (direct+indirect+contingencies)278.700
Land 700
Working capital 27,440
GRAND TOTAL (turnkey 4- land + working capital) 306.800
A6-28
-------
Table A6-28. CAPITAL COSTS
Boiler type: Watertube (44 MWfc)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Equipment cost
Basic equipment (includes freight)25,760
Required auxiliaries 52,580
Total equipment cost 78.34Q
Installation costs, direct
Foundations and supports 2,520
Piping 8,050
Insulation 900
Painting 180
Electrical ir28Q
Instruments i 340
Installation labor 1 S OOP
Total installation cost 29,270
Total Direct Costs (equipment + installation) 107,610
Installation costs, indirect
Engineering 43,890
Construction and field expense 10,760
Construction fees 10, 760
Start-up 2.150
Performance tests 2.000
Total Indirect Costs 69r560
Contingencies 26,580
Total Turnkey Costs (direct+indirect+contingencies)203,750
Land 510
Working capital 19,470
GRAND TOTAL (turnkey + land + working capital) 223.700
A6-29
-------
APPENDIX 7
ANNUAL COST BREAKDOWNS
A7-1
-------
Table A7-1. ANNUAL COSTS
Boiler type: Underfeed
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Direct costs
Direct labor $15,780
Maintenance labor 28.830
Materials 3.527
Catalyst 14.501
Electricity 854
Steam 775
Ammonia 2,431
Total direct cost 66,699
Overhead
Payroll 4,734
Plant 12,516
Total overhead cost 17,250
Capital Charges
G&A, taxes & ins. 7,837
Capital recovery 25,750
Total capital charges 33,595
TOTAL ANNUALIZED COSTS $117,540
A7-2
-------
Table A7-2. ANNUAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
Direct costs
Direct labor $15.780
Maintenance labor 28.830
Materials 6*916
Catalyst 48.590
Electricity 3.101
Steam 2.272
Ammonia 7 _
Total direct cost 112,519
Overhead
Payroll 4.734
Plant 13,396
Total overhead cost 18,130
Capital Charges
G&A, taxes & ins. 15,368
Capital recovery 50,512
Total capital charges 65,880
TOTAL ANNUALIZED COSTS $196,530
A7-3
-------
Table A7-3. ANNUAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Direct costs
Direct labor $15.780
Maintenance labor 28.830
Materials 5.840
Catalyst ^6
Electricity 2r345
Steam i
Ammonia & n n
Total direct cost 97.177
Overhead
Payroll 4,734
Plant 13.117
Total overhead cost 17.851
Capital Charges
G&A, taxes & ins. 12,977
Capital recovery 42,653
Total capital charges 55,630
TOTAL ANNUALIZED COSTS $170,660
A7-4
-------
Table A7-4. ANNUAL COSTS
Boiler type: Chaingrate
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Moderate
Direct costs
Direct labor $15.780
Maintenance labor 28,830
Materials 4,973
Catalyst 29.129
Electricity 1.892
Steam 1.713
Ammonia q 256
Total direct cost 87,573
Overhead
Payroll 4.734
Plant 12.892
Total overhead cost 17,626
Capital Charges
G&A, taxes & ins. 11,051
Capital recovery 36,323
Total capital charges 47,374
TOTAL ANNUALIZED COSTS $152,570
A7-5
-------
Table A7-5. ANNUAL COSTS
Boiler type: Spreader Stoker
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Intermediate
Direct costs
Direct labor $15,780
Maintenance labor 28,830
Materials 8,936
Catalyst 72,600
Electricity 5,847
Steam 3.814
Ammonia 12.023
Total direct cost 147,830
Overhead
Payroll 4,734
Plant 13,922
Total overhead cost 18,656
Capital Charges
G&A, taxes & ins. 19,857
Capital recovery 65,265"
Total capital charges 85,122
TOTAL ANNUALIZED COSTS $251,610
A7-6
-------
Table A7-6. ANNUAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
Direct costs
Direct labor $ 15,780
Maintenance labor 28.830
Materials 12.558
Catalyst 112.800
Electricity 12,417
Steam 7.133
Ammonia 22.535
Total direct cost 212,053
Overhead
Payroll 4,734
Plant 14,864
Total overhead cost 19,598
Capital Charges
G&A, taxes & ins. 27,906
Capital recovery 91,720"
Total capital charges 119,626
TOTAL ANNUALIZED COSTS $351,280
A7-7
-------
Table A7-7. ANNUAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Moderate
Direct costs
Direct labor $15,780
Maintenance labor 28,83C)
Materials 8,872
Catalyst 67,800
Electricity 7,490
Steam 5.371
Ammonia 16.885
Total direct cost 151,028
Overhead
Payroll 4,734
Plant 13,905
Total overhead cost 18,639
Capital Charges
G&A, taxes & ins. 19.716
Capital recovery 64.803
Total capital charges 84,519
TOTAL ANNUALIZED COSTS $254,190
A7-8
-------
Table A7-8. ANNUAL COSTS
Boiler type: Underfeed Stoker
Fuel: High Sulfur Eastern Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Direct costs
Direct labor 31,590
Maintenance labor 76,900
Materials 70,000
Catalyst 12.450
Electricity 19,730
Steam 26,100
Fuel 33,790
Boiler feed water 39,520
Ammonia 3,030
Heat credit (-32.270)
By-product credit (-26,050)
Total direct cost 254.800
Overhead
Payroll 9.480
Plant 46.410
Total overhead cost 55.900
Capital Charges
G&A, taxes & ins. 93,320
Capital recovery 306,700
Total capital charges 400,000
TOTAL ANNUALIZFJ) COSTS $710,700
A7-9
-------
Table A7-9. ANNUAL COSTS
Boiler type: Underfeed Stoker
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Direct costs
Direct labor 31,590
Maintenance labor 76,900
Materials 42,090
Catalyst 2,620
Electricity 6,990_
Steam 6,400
Fuel 7,140
Boiler feed water 8,350
Ammonia 3,710
Heat credit (-7,780)
By-product credit (-5,490)
Total direct cost 172,500
Overhead
Payroll 9,480
Plant 39,150
Total overhead cost 48,600
Capital Charges
G&A, taxes & ins. 56,120
Capital recovery 184,450
Total capital charges 240,600
TOTAL ANNUALIZED COSTS $462.000
A7-10
-------
Table A7-10. ANNUAL COSTS
Boiler type: Pulverized Coal
Fuel: High Sulfur Eastern Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Direct costs
Direct labor 31,590
Maintenance labor 76,900
Materials 106.600
Catalyst 83.100
Electricity 128,100
Steam 175,200
Fuel 226,500
Boiler feed water 269.500
Ammonia 24,100
Heat credit (-223,000)
By-product credit (-173,900)
Total direct cost 724,700
Overhead
Payroll 9,480
Plant 55.920
Total overhead cost 65.400
Capital Charges
G&A, taxes & ins. 236,800
Capital recovery 778,300
Total capital charges 1.010,000
TOTAL ANNUALIZED COSTS $1,805,000
A7-11
-------
Table A7-11. ANNUAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Direct costs
Direct labor 31,590
Maintenance labor 76,900
Materials 55,400
Catalyst 17.520
Electricity 43,560
Steam 44.140
Fuel 47,620
Boiler feed water 55,820
Ammonia 29,780
Heat credit (-51,930)
By-product credit (-36,680)
Total direct cost 313,700
Overhead
Payroll 9,480
Plant 42,610
Total overhead cost 52,100
Capital Charges
G&A, taxes & ins. 123,100
Capital recovery 404,700
Total capital charges 527,800
TOTAL ANNUALIZED COSTS $893,600
A7-12
-------
Table A7-12. ANNUAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Direct costs
Direct labor 11,835
Maintenance labor 21,623
Materials 1,205
Catalyst 3,861
Electricity 547
Steam 301
Ammonia -\ qy
Total direct cost 39,569
Overhead
Payroll 3,551
Plant 9,012
Total overhead cost 12,563
Capital Charges
G&A, taxes & ins. 3,571
Capital recovery 11.740
Total capital charges 15,311
TOTAL ANNUALIZED COSTS $67,440
A7-13
-------
Table A7-13. ANNUAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Direct costs
Direct labor $11,835
Maintenance labor 21.623
Materials 1P061
Catalyst 2,304
Electricity
Steam _ ?QI
Ammonia -\ ^R
Total direct cost 37,607
Overhead
Payroll 3.551
Plant 8.975
Total overhead cost 12,526
Capital Charges
G&A, taxes & ins. 3,143
Capital recovery 10,331
Total capital charges 13,474
TOTAL ANNUALIZED COSTS $63,610
A7-14
-------
Table A7-14. ANNUAL COSTS
Boiler type: Watertube (44 MW )
Fuel: Distillate Oil t
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Direct costs
Direct labor 14,480
Maintainence labor 26,440
Materials 8,310
Catalyst 47,380
Electricity 8.500
Steam 3.740
Ammonia 2, 470
Total direct cost 111,270
Overhead
Payroll
4,340
Plant 12.800
Total overhead cost 17.140
Capital Charges
G&A, taxes & ins. 11.080
Capital recovery 36.430
Total capital charges
47,510
TOTAL ANNUALIZED COSTS 175>90Q
A7-15
-------
Table A7-15. ANNUAL COSTS
Boiler type: Watertube (44 MW )
Fuel: Distillate Oil
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Direct costs
Direct labor 14 f 480
Maintainence labor 26,440
Materials 6,050
Catalyst 28.420
Electricity 5,140
Steam 3
Ammonia -\ Q20
Total direct cost 86,020
Overhead
Payroll 4.340
Plant 12,210
Total overhead cost 16,550
Capital Charges
G&A, taxes & ins. 8.070
Capital recovery 26.530
Total capital charges 34,600
TOTAL ANNUALIZED COSTS 137,200
A7-16
-------
Table A7-16. ANNUAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Stringent
Direct costs
Direct labor 14,480
Maintenance labor 26,430
Materials 3,090
Catalyst 13,090
Electricity 740
Steam 590
Ammonia 1,840
Total direct cost 60,260
Overhead
Payroll 4,340
Plant 11,440
Total overhead cost 15,780
Capital Charges
G&A, taxes & ins. 7.490
Capital recovery 24,620
Total capital charges 32,110
TOTAL ANNUALIZED COSTS $108,200
A7-17
-------
Table A7-17. ANNUAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Moderate
Direct costs
Direct labor 14,480
Maintenance labor 26,430
Materials 4,320
Catalyst 7.870
Electricity 460
Steam 440
Ammonia 1,370
Total direct cost 55.370
Overhead
Payroll 4,340
Plant 11.760
Total overhead cost 16,100
Capital Charges
G&A, taxes & ins. 5,750
Capital recovery 18.910
Total capital charges 24,660
TOTAL ANNUALIZED COSTS $96.100
A7-18
-------
Table A7-18. ANNUAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Stringent
Direct costs
Direct labor 14,465
Maintenance labor 52,855
Materials 2,410
Catalyst 8,660
Electricity 570
Steam 590
Ammonia 1,840
Total direct cost 81.390
Overhead
Payroll 4,340
Plant 18,130
Total overhead cost 22,470
Capital Charges
G&A, taxes & ins. 5.840
Capital recovery 19,180
Total capital charges 25,020
TOTAL ANNUALIZED COSTS $129,900
A7-19
-------
Table A7-19. ANNUAL COSTS
Boiler type: Watertube (8.8 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Moderate
Direct costs
Direct labor 14,465
Maintenance labor 52,855
Materials 2,050
Catalyst 5,190
Electricity 400
Steam 440
Ammonia 1,370
Total direct cost 76,770
Overhead
Payroll 4.340
Plant 18,040
Total overhead cost 22,380
Capital Charges
G&A, taxes & ins. 4,960
Capital recovery 16,320
Total capital charges 21,280
TOTAL ANNUALIZED COSTS $120.400
A7-20
-------
Table A7-20. ANNUAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Stringent
Direct costs
Direct labor $14,465
Maintenance labor 26.428
Materials 7.766
Catalyst 62,150
Electricity 5.697
Steam 2.089
Ammonia 6.565
Total direct cost 125,160
Overhead
Payroll 4.340
Plant 12,651
Total overhead cost 16,991
Capital Charges
G&A, taxes & ins. 18,827
Capital recovery 61,879
Total capital charges 80,706
TOTAL ANNUALIZED COSTS $222,860
A7-21
-------
Table A7-21. ANNUAL COSTS
Boiler type: Watertube (44 MWfc)
Fuel: Residual Oil
Control technique: Parallel Flow SCR
Control level: Moderate
Direct costs
Direct labor $14,465
Maintenance labor 26,428
Materials 5.565
Catalyst 37,455
Electricity 3,457
Steam 1.516
Ammonia 4.938
Total direct cost 93,324
Overhead
Payroll 4.340
Plant 12.079
Total overhead cost 16,419
Capital Charges
G&A, taxes & ins. 13.490
Capital recovery 44,340
Total capital charges 57,830
TOTAL ANNUALIZED COSTS $181,180
A7-22
-------
Table A7-22. ANNUAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Stringent
Direct costs
Direct labor $14,465
Maintenance labor 52.855
Materials
Catalyst 41.510
Electricity 3 820
Steam 7,080
Ammonia 6.560
Total direct cost 126,481
Overhead
Payroll 4.340
Plant 18.853
Total overhead cost 23,193
Capital Charges
G&A, taxes & ins. 12,585
Capital recovery 41,363
Total capital charges 53,948
TOTAL ANNUALIZED COSTS $203,620
A7-23
-------
Table A7-23. ANNUAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Moving Bed SCR
Control level: Moderate
Direct costs
Direct labor $14.465
Maintenance labor 52,855
Materials 3,842
Catalyst 24.900
Electricity 2.350
Steam lr54Q
Ammonia
Total direct cost 104,892
Overhead
Payroll 4.340
Plant 18,502
Total overhead cost 22,842
Capital Charges
G&A, taxes & ins. 9,315
Capital recovery 30,617
Total capital charges 39,932
TOTAL ANNUALIZED COSTS $167,670
A7-24
-------
Table A7-24. ANNUAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Residual Oil
Control technique: Parallel Flow SCR (NOX/SOX)
Control level: Intermediate
Direct costs
Direct labor 28,960
Maintenance labor 70,490
Materials 110,790
Catalyst 31,900
Electricity 53,330
Steam 66,780
Fuel 86,490
Boiler feed water 102,980
Ammonia 8,710
Heat credit (-86,530)
By-product credit (-78,290)
Total direct cost 395,600
Overhead
Payroll 8,690
Plant 54.660
Total overhead cost 63,350
Capital Charges
G&A, taxes & ins. 147,720
Capital recovery 485,520
Total capital charges 633,240
TOTAL ANNUALIZED COSTS $1,092,000
A7-25
-------
Table A7-25. ANNUAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Direct costs
Direct labor $11,835
Maintenance labor 21,623
Materials 1,220
Catalyst 4,QQ5
Electricity 600
Steam 70
Ammonia 220
Total direct cost 39,573
Overhead
Payroll 3,551
Plant 9,016
Total overhead cost 12,567
Capital Charges
G&A, taxes & ins. 3,616
Capital recovery 11,885
Total capital charges 15,501
TOTAL ANNUALIZED COSTS $67,640
A7-26
-------
Table A7-26. ANNUAL COSTS
Boiler type: Firetube (4.4 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Direct costs
Direct labor $11,835
Maintenance labor 21,623
Materials 1,123
Catalyst 2,394
Electricity 370
Steam 50
Ammonia 170
Total direct cost 37,565
Overhead
Payroll 3,551
Plant 8,991
Total overhead cost 12,542
Capital Charges
G&A, taxes & ins. 3,327
Capital recovery 19,934
Total capital charges 14,261
TOTAL ANNUALIZED COSTS $64,370
A7-27
-------
Table A7-27. ANNUAL COSTS
Boiler type: Watertube (44 MWt)
Fuel: Natural Gas
Control technique: Fixed Packed Bed SCR
Control level: Stringent
Direct costs
Direct labor 14.480
Maintainence labor 26.440
Materials 8.360
Catalyst 48r24Q
Electricity 8,630
Steam 880
Ammonia 2.730
Total direct cost 109,760
Overhead
Payroll 4.340
Plant 12.810
Total overhead cost 17,510
Capital Charges
G&A, taxes & ins. 11.150
Capital recovery 36,640
Total capital charges 47,790
TOTAL ANNUALIZED COSTS 174,700
A7-28
-------
Table A7-28. ANNUAL COSTS
Boiler type: Watertube (44 MW )
Fuel: Natural Gas t
Control technique: Fixed Packed Bed SCR
Control level: Moderate
Direct costs
Direct labor 14.430
Maintainence labor 26.440
Materials 6.110
Catalyst 28.920
Electricity 5.230
Steam 680
Ammonia 2,120
Total direct cost 77.870
*
Overhead
Payroll 4,340
Plant 12.230
Total overhead cost 16.570
Capital Charges
G&A, taxes & ins. 8.150
Capital recovery 26.790
Total capital charges 34,940
i 29 400
TOTAL ANNUALIZED COSTS '
A7-29
-------
APPENDIX 8
SAMPLE CALCULATIONS
A8-1
-------
An example calculation is shown below to illustrate how the energy
vaues were arrived at. The example illustrates the case of a Pulverized
Coal standard boiler with a Parallel Flow reactor and stringent control.
Calculations for the other standard boilers were performed in a similar
manner.
Sample Calculation—
First, it is necessary to perform a combustion calculation to charac-
terize the flue gas.
130% excess air
Basis: Coal analysis: Ib/lb fuel fired
C = 0.5760
H2 = 0.320
02 = 0.1120
N2 = -O'.OIZO
S = 0.0060
H20 = 0.2080
Ash = 0.0540
1.0000
The calculation is based on a method presented in Steam1 and the values
shown here are documented in the reference.
A8-2
-------
02 and Air required for combustion
c
H2
02
N2
S
H20
Ash
0.5760
0.0320
0.1120
0.120
0.0060
0.2080
0.0540
Total 1.00
Less 02 in the fuel
Requirement
Excess (30%)
Total
0 Ib
2' Ib fuel fired
x 2.66 = 1.532
x 7.94 = 0.254
x 1.00 = 0.006
1.792
0.112
Air,
Ib
1.680
0.504
2.184
Ib fuel fired
x 11.53 = 6.641
x 34.34 = 1.099
x 4.29 = O.C26
7.766
0.482
7.284
2.185
9.469
Products of combustion
C02 0.5760 x 3.66
H20 (0.032 x 8.94) + (0.2080) + (0.013 x 9.469)
02 excess
N2 9.469 x (0.7685 + 0.0120)
NO (specified by Acurex)
SO (specified by Acurex)
CO (specified by Acurex)
HC as CHit (specified by Acurex)
Fly ash (specified by Acurex)
17 T f j in /9 tons 2000 Ib . Ib fuel
Fuel feed rate = 10.42 ,___ x —^^— = 20,840
Ib/lb fuel
2.108
0.617
0.504
7.391
0.0090
0.0114
0.0005
0.0001
0.0432
10.68
hr
ton
hr
A8-3
-------
Flue Gas Composition:
N2
C02
H20
02
S02
N0x
CO
HC (as
Fly ash
Ib/hr
154,000
43,900
12,900
10,500
238
188
10
3
900
222,639
222 639
Average molecular weight = —z—-—
—
moles/hr
5,496
998
716
328
4
4
= 29.5
= 29.5
7,546
Ib
Ib mole
g
g mole
mole %
72.8
13.2
9.5
4.3
0.1
0.1
100.0
Flue gas flow rate
Gv - 73»20°
Reactor Sizing
1.698 Nm3/hr
scfm
= 75,500
Nm3
Next, it is necessary to size the reactor so the pressure drop across
the reactor may be calculated. For the stringent level of control, a large
reactor size and bed depth are used to ensure 90% NO reduction.
Basis: 2 Space velocity = 3000 hr x (based on catalyst volume: 3000
catalyst volumes of flue gas per
hour)
Bed depth = 4.5 m
r <- i <- i 75,500 Nm3/hr oc „ 3
Catalyst volume= hr-'l = 25.2 m3
A8-4
-------
To calculate the reactor volume, the specific surface areas of the pure
catalyst and of the catalyst packed in a reactor are needed.
601 m2/m3 catalyst
(20mm parallel plate)
194 m2/m3 packed reactor volume
r> <_ ^ i /oc o s i \/601 m2/m3 catalyst volume \
Reactor volume = (25.2 m3 catalyst)/^--—2 / 3 ;—T—' —^
1194 m /m packed reactor volumey
= 78 m3
= width2 x depth (square reactor)
Therefore, width = 4.16 m
Pressure Drop
Now that the reactor geometry has been defined, the pressure drop across
the reactor can be determined. For this calculation the following equation
3
is used.
2 f G2L (l-£)3~n
AP =
where AP = pressure drop across bed of granular solids, lb^/ft
f = friction factor : a function of modified Reynold's
number (N' ) , dimensionless
G = gas superficial mass velocity, Ib /ft sec
L = bed depth, ft
e = void fraction, dimensionless
n = exponent: a function of modified Reynold's number (N^£)
dimensionless
D = average particle diameter : diameter of a sphere of the
same volume as the non-spherical particle, ft
A8-5
-------
e = dimensional constant, 32.2
°
Ib ft
- z-rr
sec lb
p = gas density, Ib /ft3
(j) = shape factor of the solid : quotient of the surface area
of a sphere of equivalent volume divided by the actual
surface area of the non-spherical particle, dimensionless
(The modified Reynolds number, N^ , is defined as D G/y,
Ke p
where y = gas viscosity, Ib /ft sec)
Parallel Flow Catalyst (a square passage was assumed for ease of
calculation)
t
b
4
T
a = 20 mm
b = 14 mm
Cell length = 1 m (assumed, a common commercial cell length )
In order to calculate a shape factor it is necessary to calculate the
diameter of a sphere that has a volume of catalyst equivalent to a
single passage of the square honeycomb catalyst.
Catalyst volume per passage = [(20 mm)2 - (14 mm)2] 1000 mm
= 204,000 mm3
or 7.21 x 10~3 ft3
V sphere = ~ n r3 = 7.21 x 10~3
J P
r = 0.120 ft
P
D = 0.240 ft
P
The shape factor can now be calculated
A8-6
-------
-------
where T = critical temperature of component i, °K
V = critical volume of component i, cm3/g-mole
c
T = reduced temperature = ratio of gas temperature to critical
temperature (T/T ), dimensionless
/(T ) = gas viscosity temperature function, dimensionless
Values for y. were calculated using data from Smith & VanNess6
N2
C02
H20
02
T °k
V fc
126.2
304.2
647.1
154.6
T
r
5
2
1
4
,°k
.33
.21
.04
.35
c'g:
89
94
56
73
cm
mole
.5
.0
.0
.4
f(l
3
1
0
2
•33Tr)
.07
.65
.862
.68
M
28
44
18
32
.02
.01
.02
.00
y
2
2
1
2
Ib
i'ft-sec
.04xlO~5
.07xlO~5
.42xlO~5
. 41x10" 5
The following data were used with equation (A8-2) to calculate the gas
viscosity
Ib
N2
C02
H20
02
y±
0.728
0.132
0.095
0.043
Mi'ft-sec
2. 04xlO~ 5
2.07xlO~5
1.42xlO~5
2.41xlO~5
11.
i
28.02
44.01
18.02
32.00
with the following result
Ib
= 2.00x10
mixture ' ft-sec
Gas Superficial Mass Velocity, G
From the results of the combustion calculation (total mass flow of flue
gas) and the reactor sizing calculation (reactor width), the superficial
mass velocity can be found.
A8-8
-------
Ib 1 hr
G = v ^L./_V^"",^V = o 332 2
(4.16m)2 (10.76 ^-] ft sec
Modified Reynolds Number, N'
Using the results of the catalyst characterization, gas mixture viscosity,
and mass velocity calculations the modified Reynolds number can be found.
D G (0.240 ft) (0.332
-sec .
'Re o -5 lb - 398°
^ ' X ft-sec
Knowing the modified Reynolds number the friction factor and exponent
can then be determined.
f = 0.7
m
n = 1.97
Flue Gas Density, p
Reaction T = 750°F = 1210°R
Volumetric flue, gas flow = 73,200 acfm @ 350°F (PedCo)
From combustion calculation: mass flow = 222,600 —
mass
Density
volume
lb
73 200
73,200 min35o+460°R/ hr
A8-9
-------
All of these terms are then substituted into equation (A8-1) to determine
the pressure drop.
2f G2L (l-£)3~n
AP = —^ (A8-1)
~
2s—) f(4--
sec/LI
°-332 - (4-5m)
• 9 7
Ib ft v / Ib
(0.240ft) [32.2 m2l, 1(0.0339 7-^)(0.301)3 1-97 (0.677)3
sec -Lb,. / \ it
JU>" ft*
AP = 0.210 psi = 148 mm H20
Now the energy consumption of the various process steps can be calcu-
lated. The energy consuming items considered in this case are
flue gas fan,
liquid NHs pump,
NHs vaporization, and
NH3 dilution.
Flue Gas Fan
The gas side horsepower (hp) can be calculated from7
(hp)gas = 1-57xlO"4 Q AP (A8-4)
Q = ft3/min
AP = in H20
(hp)a.r = (1.57x10-) (73,200^) (148mm H20)
= 67.0 hp
18-10
-------
Fan efficiencies typically range from 40 to 70%. If an efficiency of
55% is assumed the shaft hp can be calculated
(hp)shaft = tftj = 122 hp
In terms of electrical usage, this is
(122hP)(°-74gkW)=90.8kW
Liquid NH3 Pump
For 90% removal an NH3:NO mole ratio of 0.95 is typical2
AIQQ lb NOxVl mole NOA/.95 mole NH3\A? lb NH3\
NH3 requirement = (188 —^ ( 46 lb N0 1 mole N0 lb mole )
\ ' \ x/\ x/\ /
= lb NH3
66 hr
or 0.21 gpm
The following pump curve8 indicates that a 0.5 hp centrifugal pump
operated at 1750 rpm can supply 28 ft of NH3 head (7.6 psi). This is
adequate to transfer sufficient NH3 to the vaporizer.
A8-11
-------
120
100
„ 80
_j
•5 60
\L
< .„
o
«
X
|2
0
^
^
/
34,
**
7,*>0
"N
w/-
N
#0
V
"X
WF
X
3/77
X,
Pi//l
N
X,
X,
N
p c
X,
N
•s. J
/I
T"-^.
r
hat
5 In
^^
,
J
/
/
/
>
^
Kit
Ctl 1
f
f
1^
y/
t
5^
rist
mpi
f
7
/
^
^
^,
e c
Her
int.
»
UAKi
'we
?03T
Vh
^
/
^
/*
"•^
s
1S
/-.
•£
^
'
^/
7
t
' III
79;
"V
/^
X
\
S2*
*A
?«-
/
!'
^
s
""
s.
\
^>
P-
%
\
^^
1 '
1
j
X
~^hp.
l|
J
|^-4.W.
i
0 40 SO 120 160 2OO Z4O 28O it
Gallons Per Minute
In terms of electrical usage this amounts to
(0.5 hp)(°-7^7kW)= 0.373 kW
NHa Vaporization
Looking at the worst case for NH3 vaporization, a cold winter day at
a Midwest location, the ambient air temperature might be, say, -10°F.
The pressure in the vaporizer is
Saturated vapor pressure (-10°F) =23.7 psia
+ Pump head = 7 ^
31.3 psia
The normal boiling point of NH3 is -28.0°F. 9 In order to determine
the actual boiling point, and thus the heat load on the vaporizer,
it is necessary to evaluate it at the higher pressure. This can be
accomplished by use of the Clausius-Clapeyron equation.10
A8-12
-------
^
AHvap = -Rd(VT) (A8-5)
At -28.0°F and 1 atm. AH = 589.3
vap lb
The Clausius-Clapeyron equation is used to determine the boiling temp-
erature at the higher pressure of 31.3 psia. Use of the equation assumes
a constant AH , however, this is not strictly true. For this reason,
vap' ' y '
two iterations are calculated. The first uses AH at 14.7 psia and
vap
calculates an elevated boiling temperature at P = 31.3 psia. The AH
at this temperature is found from thermodynamic tables. Then the two
AH values are averaged to determine a pseudo constant AH . This
vap 6 ^ vap
value is then used in the. second iteration to determine a new and more
accurate elevated boiling temperature.
Solving equation (A8-5) gives
(A8-6)
The data used to calculate T2 are
P! =14.7 psia
P2 = 31.3 psia
T! = -28.0°F + 460 = 432°R
AH = 589.3 ^
vap lb
Btu
R = 1.986
lb mole °R
Solution of equation (A8-6) for T2 gives
T2 = 462°R = 2°F
AH at 2°F is 567.3 . 9
vap lb
AS-13
-------
The two AH values are averaged to obtain a pseudo constant AH
vap vap
-TTT _ 589.3 + 567.3 _ „„ ,, Btu
AH. — „ — J/O.J -r-:
vap 2 lb
Solving equation (A8-6) a second time using this new AH indicates
Vcip -n j_
a boiling temperature of 3°F. At this temperature, AH is 566.5 —n~-
vap J_D
Now the energy requirement for NH3 vaporization can be calculated. The
heat capacity of NH3 at this temperature is11
C [NH3U)1 = 1.10
P
Q . . = 66 "V"3 1.10 f£S? [3°F-(-10°F)] + 566.5 ..
xvaporization I hr /|\ lb Fj lb
= 38,300
NH3 Dilution
The NH3 is diluted with 30 psig steam prior to injection. A 5:1 mole
ratio is used.12 The heat of vaporization of 30 psig steam is 929.0
T
c _ . . , /,, lb NH3\/ mole NH3 \ /5 moles steam\
Mass of steam injected = 66 — r - -)(-,-, n ,, „,, \ I - ; - ^^ -
J I hr l\ 17.0 lb NH3J I mole NH3 I
(18 lb steam\ _ lb steam
mole steam I hr
Q = (350 lb) 929-b° BtU) = 325,000
A8-14
-------
Summarizing the energy consuming steps,
Item
Flue gas fan
Liquid NH3 pump
NH3 vaporization
NH3 dilution
Total
Energy usage
90.8 kW n
elec
0.373 kW .
elec
38,300 Btu/hr
325,000 Btu/hr
Btu/hr
MW
thermal
908,000
3,730
38,300
325,000
0.266
0.00109
0.0112
0.0953
1,275,030
0.374
The example calculation is continued below to show the methods used to
arrive at annual cost figures for NOX control systems. The case is the
Parallel Flow SCR system applied to the Pulverized Coal standard boiler and
operated at the stringent level of control. The material balance and pro-
cess flow diagram, as it appears in the Appendix, are presented in Figure
A8-1 on the following page. First, each of the pieces of equipment, in
succession, will be sized (including any necessary design calculations) and
the F.O.B. costs determined. These results will then be utilized to deter-
mine the installed costs. Then, the direct operating costs are calculated.
These costs are combined via the recommended format using the appropriate
load factor to arrive at the annual costs.
From Figure A8-1, the NH3 flow is shown to be 1.76 g^ . Assuming
the plant maintains a 15-day storage supply (large enough to survive delay
in deliveries due to bad weather, strikes, etc.), the required NHs storage
tankage is determined.
p-tc. n ^ kg-moleVl7.0 kgV . Ib
Gal Storage = 11.76 —^r (-, ^~ I \ re./ -\—
& \ hr / \ kg-mole/ \.454 kg
ft3
17
7.48
hr
i.O lb/ \ ft3 ) \ day
x (15 day supply) (A8-7)
= 4600 gal
.Exponential Factor
F.O.B. Equipment Cost = Base Cost (Unit Cost) x Pressure
Factor
A8-15
-------
From
Economizer
.
XX
To
Preheater
OD
I
Reactor
NH3
Vaporization
T, °K
P, Pa
Nz
CO 2
H20
02
HOX
S0x
NH3
<*>
648
98,600
2495
453
325
149
1.86
1.69
-
0
648
97,150
2497
453
336
149
0.19
1.69
0.03
<3>
283
615,000
_
_
_
_
-
-
1.76
289
752,000
_
„
_
_
-
-
1.76
429
552,000
_
_
0.98
_
_
-
-
<">
408
310,000
_
_
8.8
_
_
-
-
NH3
Storage
Figure A8-1.
Material balance.
Pulverized Coal
Parallel Flow SCR
Stringent Control
-------
F.O.B. NH3 Tank Cost (mid-1970) 1 5 = 10, 000 f 1 ' x 1.38 (A8-8)
= $8,000
Next, the two liquid NH 3 pumps (one for a spare in case of failure) are
examined. The volumetric liquid NH3 flow rate can be determined from:
Liq. NH3 Vol. Flow = I 1.76 ^—-i i 17.0 kg 1 / Ib
ft3 ) / 7.48 gal] / hr ) (A8-9)
hr / \kg-mole / \.454 kg
\ / \ /
_ I f 7.48 gal) / hr
39.0 Ib / \ ft' / \ 60 min
= 0.21 GPM
At this flow rate, a 0.5 hp centrifugal pump is adequate.8
F.O.B. Equipment Cost = H Pumps x Base Cost
S ^00
F.O.B. Pump Cost (mid-1970)16 = 2 pumps x-
^ pump
= $600
To size the NH3 vaporizer the sensible heat and the heat of vaporization
for NH3 is required. For the worst case at -10°F and 31 psia tank pressure,
the heat required = 581 .
The heat load on the vaporizer is shown to be:
^.e_J____ 17.0 kg\ / Ib \ / 581 BTU required
Q I1'76 hr }[ kg-molej \ .454 kg / \ Ib NH3 vaporized
A8-17
-------
The heat transfer area required for the NH3 vaporizer can be calculated from:
A = ^ CA8-10)
c n
iy
where A = heat transfer surface area, ft
q = heat transferred, BTU/hr
F = safety factor (assume 2.0)
U = heat transfer coefficient, BTU/hr ft2 °F
c
T = temperature of heat medium, °F
(800 psia steam = 312°F)
T = fluid temperature, °F
(worst case: NH3 = -10°F)
(38,300 ~)(2.0)
. _ __ hr
A — / T-mTT VTTS
300 hr_f™°F J (312 - (-10) "F
0.79 ft2
The smallest commercially available doube pipe heat exchanger is 1 ft2 and
the F.O.B. Vaporizer Cost (mid-1970) = $300.
1 9
Next, the reactor size is determined from the volumetric flue gas flow
rate and the reactor space velocity.
Volume Flue Gas -
Nm3\/60 min
"~
sf h
= 75,500 - (A8-11)
hr
20 75,500 Nm3/hr
Catalyst Volume = •
3000 hr 1 space velocity
(A8-12)
A8-18
-------
Reactor Volume21 = (25.2 m3 catalyst) x
/601 m2 surface area/m3 catalyst _ \
\194 m surface area/m packed volume/
= 78.0 m3 packed volume (A8-13)
Reactor Length21 = 4.5 m ( ^ Q } = 14.8 ft
\ . j04om/
Square Reactor Volume = 78.0 m3 = W2L = 4.5 W2
2 78.0 m3 _ ,
W = —. — = - 3- = l/.j m
4.5m
W = 4.16 m = 13.6 ft
F.O.B. Reactor Material Cost (mid-1970) = $16,00022
Finally, the draft fan motor drive must be determined.
0.000157 Q Ap 7 /,o A
Motor hp. - 55% efficiency (A8"
ft
0.000157 ^/J'*UU min/ (148 mmH20) \25.4 mm
.55
= 122 hp
/h \ °'77
Motor Drive - F.O.B. Motor Cost (mid-1970)23 = 5800(^j (A8-14)
^122\°-77
= 580° 70 /
= $8,900
A8-19
-------
Each piece of equipment is factored by its respective escalation index to
give a 1978 F.O.B. cost and an 8 percent freight charge for delivery is added
to it. 4 The direct installation costs are determined by the appropriate
factor multiplied times the 1970 F.O.B. equpment cost and that category's
respective escalation index. These analyses for each equipment item are
presented on the following pages.
A8-20
-------
NH3 Storage Tank
F.O.B. Equipment Cost (mid-1970) = 10,000
x 1.38
= $ 8,000
Equipment Cost (1978)
Basic equipment
Freight
Required auxiliaries
June - 1978 Costs
= F.O.B. 1970 x Escalation Index
= $8000 x 1.91
= 0.08 x Basic Equipment
Total Equipment Cost
.1 5
Installation Costs, Direct = F.O.B. 1970 x Installation Cost
Fraction x Escalation Index
Foundation and supports = F.O.B. 1970 x 0.080 x 2.11
Ductwork
Stack
Piping
Insulation
Painting
Electrical
Instruments
Installation Labor
= F.O.B. 1970 x,0.153 x 2.02
= F.O.B. 1970 x 0.012 x 2.11
= F.O.B. 1970 x 0.007 x 1.70
= F.O.B. 1970 x 0.118 x 1.63
= F.O.B. 1970 x 0.352 x 1.37
Total Installation Cost
1,400
N/A
N/A
2,500
N/A
200
100
1,500
3,900
9,600
Total Direct Cost
AS-21
-------
Liquid NH3/Pumps
2 - 0.5 hp Centrifugal Pumps (1 spare)
Pump and motor - F.O.B. Equipment Cost (mid-1970) = $300 x 2 = $ 600
Equipment Cost (1978)
Basic equipment
Freight
Required auxiliaries
Installation Costs, Direct
1 6
Foundation and supports
Ductwork
Stack
Piping
Insulation
Painting
Electrical
Instruments
Installation labor
June - 1978 Costs
= F.O.B. 1970 x Escalation Index
= $600 x 2.08
= 0.08 x Basic equipment
Total Equipment Cost
= F.O.B. 1970 x Installation Cost
Fraction x Escalation Index
= F.O.B. 1970 x 0.039 x 2.11
= F.O.B. 1970 x 0.293 x 2.02
= F.O.B. 1970 x 0.028 x 2.11
= F.O.B. 1970 x 0.008 x 2.11
= F.O.B. 1970 x 0.303 x 1.70
= F.O.B. 1970 x 0.029 x 1.63
= F.O.B. 1970 x 0.679 x 1.37
Total Installation Cost
50
N/A
N/A
360
40
10
310
30
560
1,360
Total Direct Cost
A8-22
-------
NH3 Vaporizer
1 ft2 Double-Pipe Heat Exchanger
(minimum size available)
Vaporizer - F.O.B. Equipment Cost (mid-1970) = $300
Equipment Costs (1978)
Basic equipment
Freight
Required auxiliaries
Installation Costs, Direct
1 9
June - 1978 Costs
= F.O.B. 1970 x Escalation Index
= $300 x 1.91
= 0.08 x Basic equipment
Total Equipment Cost
= F.O.B. 1970 x Installation Cost
Fraction x Escalation Index
Foundation and supports = F.O.B. 1970 x 0.038 x 2.11
Ductwork
Stack
= F.O.B. 1970 x 0.213 x 2.02
= F.O.B. 1970 x 0.022 x 2.11
= F.O.B. 1970 x 0.002 x 2.11
= F.O.B. 1970 x 0.010 x 170
= F.O.B. 1970 x 0.048 x 1.63
= F.O.B. 1970 x 0.467 x 1.37
Total Installation Cost
Piping
Insulation
Painting
Electrical
Instruments
Installation labor
$ 570
50
N/A
620
50
N/A
N/A
270
30
10
50
400
810
Total Direct Cost
A8-23
-------
Reactor
Reactor - F.O.B. Equipment Cost (mid-1970): Material
: Fabricated (2x
Material) =
June - 1978 Costs
Equipment Cost (1978)
Basic equipment = F.O.B. 1970 x Escalation Index
= $32,000 x 1.91
Required auxiliary:Catalyst = $212 x ft3 catalyst
Freight = 0.08 x Basic equipment =
Total Equipment Cost =
Installation Costs, Direct22 = F.O.B. 1970 x Installation Cost
Fraction x Escalation Index
Foundation and supports = F.O.B. 1970 x 0.176 x 2.11
Ductwork =
Stack
= F.O.B,
= F.O.B.
= F.O.B.
= F.O.B.
= F.O.B.
$ 16,000
32,000
Piping
Insulation
Painting
Electrical
Instruments
Installation labor
1970 x 0.595 x 2.02
1970 x 0.080 x 2.11
1970 x 0.013 x 2.11
1970 x 0.049 x 1.70
1970 x 0.114 x 1.63
= [F.O.B. 1970 x 0.972 x 1.37
+ (Catalyst x 0.10) ]
Total Installation Cost
61,100
188,000
4,890
254,000
11,800
N/A
N/A
38,400
5,400
880
2,600
6,000
61,400
= 126,480
Total Direct Cost
$380,470
A8-24
-------
Draft Fan Motor Drive
Motor - F.O.B. Equipment Cost (mid-1970) = 5,800
= $ 8,900
Equipment Cost (1978)
Basic equipment
Freight
Required auxiliaries
Installation Costs, Direct 5
Foundation and supports
Ductwork
Stack
Piping
Insulation
Painting
Electrical
Instruments
Installation labor
June - 1978 Costs
= F.O.B. 1970 x Escalation Index
= $8900 x 2.08
= 0.08 x Basic equipment
Total Equipment Cost
= F.O.B. 1970 x Installation Cost
Fraction x Escalation Index
= F.O.B. 1970 x 0.043 x 2.11
= F.O.B. 1970 x 0.141 x 2.02
= F.O.B. 1970 x 0.005 x 2.11
= F.O.B. 1970 x 0.068 x 1.70
= F.O.B. 1970 x 0.013 x 1.63
= F.O.B. 1970 x 0.295 x 1.37
Total Installation Cost
Total Direct Cost
18,500
1,500
N/A
20,000
820
N/A
N/A
2,500
90
N/A
= 1,000
190
= 3,600
= 8,200
= $28,200
A?-2 5
-------
The direct operating costs are shown below.
Ammonia
kg mole\ /J130W1 ton
1'/b hr ; V ton A 2000
17 kg \ / Ib
Ib \kg mole
= $22,535
Electricity
/1000_g\ ( hr)(0.6)
\ kg ^
= $12,429
Steam
9.78 kg mole\ / 18 kg
- h? - -
= $7,133
I Ib \ $3.50
5T7 1000 lb
,__,. , wn ,.
(8760 hr) (0.6)
m
The individual equipment costs and installation costs are summed and the
totals entered in Table A8-1. The direct operating costs are entered in
Table A8-2.
A8-26
-------
TABLE A8-1. CAPITAL COSTS
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
Equipment cost
Basic equipment (includes freight) 104,470
Required auxiliaries 188,000
Total equipment cost 292,470
Installation costs, direct
Foundations and supports 14,120
Piping 44,030
Insulation 5,560
Painting 1,090
Electrical 4,020
Instruments 7,770
Installation labor 69,860
Total installation cost 146,450
Total Direct Costs (equipment + installation) 438,920
Installation costs, indirect
Engineering 43^9-7
Construction and field expense A
Construction fees 4
Start-up 8,778
Performance tests 2 , 000
Total Indirect Costs 142,454
Contingencies 116.275
Total Turnkey Costs (direct-t-indirect+contingencies)597,649
Land 1,744
Working capital 54,288
GRAND TOTAL (turnkey + land + working capital) $753,620
A8-27
-------
TABLE A8-2. ANNUAL COSTS
Direct costs
Overhead
Direct labor
Maintainence labor
Materials
Catalyst
Electricity
Steam
Ammonia
Total direct cost
Payroll
Plant
Boiler type: Pulverized Coal
Fuel: Low Sulfur Western Coal
Control technique: Parallel Flow SCR
Control level: Stringent
17,279
31.569
12.884
113,081
12,417
7,133
??.53S
5.184
16.050
Total overhead cost
Capital Charges
G&A, taxes & ins. 27,Q06
Capital recovery q-\ 770
71ft QOO
21.234
Total capital charges
TOTAL ANNUALIZED COSTS
1 1 Q
$357.760
A8-28
-------
REFERENCES
1. Babcock & Wilcox. Steam, Its Generation and Use. 39th Edition. 1978,
p. 6-11.
2. Ando, Jumpei. "NOX Abatement for Stationary Sources in Japan." EPA
report currently in preparation, April 1978.
3. Perry, Robert H. Chemical Engineers Handbook. 5th Edition. 1973.
McGraw-Hill, pp. 5-52, 53.
4. Ibid. , p. 3-249.
5. Ibid. , p. 3-247, 3-248.
6. Smith, J.M. and VanNess, H.C. Introduction to Chemical Engineering
Thermodynamics. Third Edition. 1975. McGraw-Hill. p. 570.
7. Perry, R.H., op. ait. , p. 6-21.
8. Ibid. , p. 6-7
9. Smith, J.M., op'.cit., p. 6-21
10. Ibid. , p. 185.
11. Ibid. , p. 113.
12. Marcos, Chemico Air Pollution Control Corporation. Telephone Conversa-
tion. 29 September 1978.
13. Smith, J.M., op. ait. , p. 576.
14. Weast, R.C. Handbook of Chemistry and Physics. 56th Edition. 1975-
1976. p. E-28.
15. Guthrie, Kenneth M. Process Plant Estimating. Craftsman. 1974.
pp. 349, 350.
16. Ibid. , pp. 159, 163.
17. Perry, R.H., op.cit., p. 10-36.
18. Ibid. , p. 10-39.
A8-29
-------
19. Guthrie, K.M. , op.cit. , pp. 144, 145.
20. Ando, J., op.cit., p. 3-31.
21. Ibid., p. 3-30.
22. Guthrie, K.M., op.oit. , pp. 150-154.
23. Woods, Donald R. Financial Decision Making in the Process Industry.
Prentice-Hall. 1975. p. 301.
24. Guthrie, K.M. "Capital Cost Estimating." Chemical Engineering.
March 24, 1969. p. 122.
25. Ibid. , p. 174.
A8-30
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/7-79-178g
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Technology Assessment Report for Industrial Boiler
Applications: NOx Flue Gas Treatment
5. REPORT DATE
December 1979
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Gary D. Jones and Kevin L. Johnson
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
10. PROGRAM ELEMENT NO.
INE624
11. CONTRACT/GRANT NO.
68-02-2608, Task 45
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PE RIOD .CO VERED
Task Final; 6/78 - 11/79
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer is J. David Mobley, Mail Drop 61, 919/
541-2915.
is.
gjves results of an assessment of the applicability of NOx flue
gas treatment (FGT) technology to industrial boilers and is one of a series of tech-
nology assessment reports to aid in determining the technological basis for a New
Source Performance Standard for Industrial Boilers. The status of development and
performance of alternative NOx FGT control techniques were assessed and the cost,
energy, and environmental impacts of the most promising processes were identified.
It was found that processes utilizing selective catalytic reduction (SCR) of NOx with
ammonia can achieve 90% reduction of NOx emissions, and that these processes are
the nearest to commercialization in the U.S. Cost estimates of applying SCR proces-
ses in the U.S. indicated that the cost effectiveness varies significantly depending on
the fuel fired, boiler size, and control level. However, boiler size is the most signi-
ficant factor affecting cost effectiveness with the economy of scale causing control of
large sources to be the most effective. The energy impact of applying SCR processes
averaged about 0. 5% of boiler capacity. No adverse environmental impacts were ap-
parent TaftMugft emissions, liquid effluents , and solid wastes must be controlled.
For regulatory purposes this assessment must be viewed as preliminary, pending
results of the more extensive impact; studies required by Clean Air Act Sect. 111.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Nitrogen Oxides
Flue Gases
Assessments
Industrial Processes
Boilers
Catalysts
Ammonia
Fossil Fuels
Pollution Control
Stationary Sources
Flue Gas Treatment
Selective Catalytic Re-
duction (SCR)
13B
07B
2 IB
14B
13H
ISA
07D
2 ID
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
581
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
A8-31
------- |