&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-79-147
June 1979
Assessment of Stationary
Source Npx Control
Technologies: Second
Annual Report
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-147
June 1979
Environmental Assessment of Stationary
Source NOX Control Technologies:
Second Annual Report
by
L R. Waterland, K. J. Lim, K. G. Salvesen,
R. M. Evans, E. G. Higginbotham, and H. B. Mason
Acurex/Aerotherm Division
485 Clyde Avenue
Mountain View, California 94042
Contract No. 68-02-2160
Program Element No. EHE624A
EPA Project Officer: Joshua S. Bowen
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
This report summarizes results of the second year of EPA Contract
68-02-2160: "Environmental Assessment of Stationary Source NOX
Combustion Modification Technologies." The EPA Project Officer is
J. S. Bowen and the Deputy Project Officer is R. E. Hall, both of the
Combustion Research Branch, IERL-RTP. This report was prepared by the
Energy and Environmental Division of Acurex Corporation. The Acurex
Project Manager is H. B. Mason; L. R. Waterland is the Chief Project
Engineer. Principal contributors to the effort, in addition to the report
authors, were: L. B. Anderson, C. Castaldini, Z. Chiba, E. Chu, M. A.
Herther, R. Ivani, R. J. Milligan, P. Overly, L. M. Schalit, A. B.
Shimizu, D. Smith, and J. Steiner. C. B. Moyer and G. R. Offen provided
technical review.
Additionally, subcontract support was provided by J. Thomasian of
Energy and Environmental Analysis, Inc., who authored portions of Section
3 of this report; A. Eschenroeder and A. Lloyd of Environmental Research
and Technology, who provided photochemical trajectory model calculations;
and J. Gabrielson and P. Langsjeon of KVB, Inc., A. Crawford and I. Manny
of Exxon Research and Engineering, and M. Hilt, L. B. Davis, and
N. Fitzroy of General Electric Corporation, who provided field test
support.
The contributions of the following individuals and organizations
are also gratefully acknowledged: R. P- Hangebrauck, J. S. Bowen, R. E.
Hall, D. G. Lachapelle, W. S. Lanier, G. B. Martin, J. H. Wasser,
G. R. Gill is, and R. B. Perry of the Energy Assessment and Control
Division, IERL-RTP, W. Axtman of the American Boiler Manufacturers'
Association, J. Crooks of the Tennessee Valley Authority, R. Lippeatt and
C. Jensen of Blueray Systems, Inc., E. Campobenedetto of the Babcock and
Wilcox Company, J. Vatsky of the Foster Wheeler Energy Corporation, W.
Barr, F. Strehlitz, and E. Marble of the Pacific Gas and Electric Company,
R. Meinzer of the San Diego Gas and Electric Company, and L. Robinson of
the Bay Area Air Pollution Control District.
n
-------
CONTENTS
Section Page
1 INTRODUCTION 1
1.1 Background 2
1.2 Program Overview 3
2 CURRENT PROCESS TECHNOLOGY BACKGROUND 9
2.1 Stationary Combustion Process Background 10
2.1.1 Utility and Large Industrial Boilers 10
2.1.2 Packaged Boilers 18
2.1.3 Warm Air Furnaces and Other Commercial and
Residential Combustion Equipment 18
2.1.4 Gas Turbines 24
2.1.5 Reciprocating 1C Engines 24
2.1.6 Industrial Process Heating 27
2.2 Stationary Source Fuel Consumption 27
2.2.1 Baseline Fuel Consumption 27
2.2.2 Projected Fuel Consumption 32
2.3 Trends in Equipment/Fuel Use 34
2.3.1 Utility Boilers 36
2.3.2 Packaged Boilers 36
2.3.3 Residential Heating Units 39
2.3.4 Gas Turbines 39
2.3.5 Reciprocating 1C Engines 41
2.3.6 Process Furnaces 41
2.4 Availability of Alternate Clean Fuels For Use
in Area Sources 42
2.4.1 Alternate Liquid Fuels -- Coal Liquids 43
2.4.2 Alternate Liquid Fuels Methanol 44
2.4.3 Alternate Gaseous Fuels Low Btu Gas 44
2.4.4 Alternate Gaseous Fuels Medium Btu Gas ... 45
2.4.5 Alternate Gaseous Fuels High Btu Gas .... 45
2.4.6 Summary 45
3 CURRENT ENVIRONMENTAL BACKGROUND 50
3.1 The Annual Average N02 Standard 51
3.2 Short Term N02 Standards 54
in
-------
TABLE OF CONTENTS (Continued)
Section Page
3.2.1 Causes of High Short Term N02 Levels 55
3.2.2 Potential Extent of Short Term N02
Violations 58
3.3 Other Clean Air Act Provisions 62
3.3.1 Prevention of Significant Deterioration .... 64
3.3.2 The Nonattainment Policy 65
3.4 Related Issues 68
3.4.1 The NOX-HC Relationship 68
3.4.2 Secondary Pollutants 69
3.4.3 Coal Utilization 70
3.5 The N02 Monitoring Network 70
3.6 Summary 72
4 ENVIRONMENTAL OBJECTIVES DEVELOPMENT 75
4.1 Source Analysis Models 76
4.1.1 SAM IA 77
4.1.2 SAM I 77
4.1.3 Extended SAM I 77
4.2 Process Impacts Evaluation 78
4.2.1 NOX Emissions Correlation 78
4.2.2 Process Analysis Procedures 81
4.2.3 Cost Analysis Procedures 86
4.3 Systems Analysis Methods 87
4.3.1 Preliminary Model 87
4.3.2 Advanced Models 91
5 ENVIRONMENTAL DATA ACQUISITION 94
5.1 Baseline Emissions 94
5.1.1 National Baseline Emissions Inventory 94
5.1.2 Projected National Emissions Inventories .... 101
5.1.3 Regional Emissions Inventories 101
5.2 Experimental Testing 108
-------
TABLE OF CONTENTS (Continued)
Section Page
6 CONTROL TECHNOLOGY OVERVIEW 121
6.1 Control Requirements 121
6.2 State-of-the-Art Controls 122
6.2.1 Low Excess Air 122
6.2.2 Off Stoichiometric Combustion (OSC) 122
6.2.3 Flue Gas Recirculation 123
6.2.4 Reduced Firing Rate 124
6.3 Advanced Controls 125
6.3.1 Low NOX Burners 125
6.3.2 Ammonia Injection 126
6.4 Other Control Methods 126
6.4.1 Reduced Air Preheat 127
6.4.2 Water Injection 127
6.4.3 Flue Gas Treatment 127
7 CONTROL TECHNOLOGY ASSESSMENT 130
7.1 Effectiveness of NOX Controls 130
7.2 Process Analysis of NOX Controls 133
7.2.1 Coal-Fired Boilers 136
7.2.2 Oil-Fired Boilers 141
7.2.3 Gas-Fired Boilers 144
7.3 Costs of NOX Controls 147
7.3.1 Retrofit Control Costs 147
7.3.2 Control Costs for New Boilers 151
7.4 Environmental Assessment of NOX Controls 152
7.5 Best Control Options 154
8 ENVIRONMENTAL ALTERNATIVES ANALYSIS 156
8.1 Baseline Impact Rankings 156
8.2 Air Quality Projections 159
8.2.1 Preliminary Model Results 166
8.2.2 LIRAQ Results 172
8.2.3 Photochemical Trajectory Model Results 176
-------
TABLE OF CONTENTS (Concluded)
Section Page
9 TECHNOLOGY TRANSFER 184
10 FUTURE EFFORTS 187
vl
-------
LIST OF TABLES
Table Page
2-1 Significant Stationary Fuel Combustion Equipment
Types/Major Fuels 12
2-2 Summary of Utility and Large Industrial Boiler
Characterization 15
2-3 Summary of Packaged Boiler Characterization 19
2-4 Summary of Warm Air Furnaces Characterization 21
2-5 Summary of Gas Turbine Characterization 25
2-6 Summary of Reciprocating 1C Engine
Characterization 26
2-7 Summary of Industrial Process Heating
Characterization 29
2-8 1974 Stationary Source Fuel Consumption (EG) 31
2-9a Stationary Source Fuel Consumption for the Year
2000: Reference Case Low Nuclear (EJ) 35
2-9b Stationary Source Fuel Consumption for the Year
2000: Reference Case High Nuclear (EJ) 35
3-1 Summary of Current NSPS & Mobile Emission
Standards; for NOX 52
3-2 AQCR's Recognized as Potential N02 Problem
Areas 52
3-3 Comparison of Estimated N02 Levels From Point
and Area Sources Under Different Meteorological
Conditions in Chicago 57
3-4 Estimated Point Source Related Violations of
Various One Hour N02 Standards 60
3-5 Estimated Number of AQCR's in Violation of One
Hour N02 Standard Based on Point Source Impact .... 61
3-6 Estimated Number of AQCR's in Violation of One
Hour N02 Standard Based on Area Source Impact 63
vn
-------
LIST OF TABLES (Continued)
Table Page
3-7 Simulation of Results for Attainment of a 150 g/m3
One Hour N02 Standard in the Chicago AQCR 68
3-8 The U.S. N0ฃ Monitoring Network 71
4-1 Field Test Program Data Compiled 79
4-2 Individual Test Points Correlated 80
4-3 Summary of Process Data Sources 82
4-4 Process Variables Investigated 84
5-1 Summary of 1974 Stationary Source NOX
Emissions By Fuel Type 97
5-2 Summary of Air and Solid Pollutant Emissions
From Stationary Fuel Burning Equipment 98
5-3 NOX Mass Emission Ranking of Stationary
Combustion Equipment and Criteria Pollutant and
Fuel Use Cross Ranking 99
5-4 Summary of Annual NOX Emissions from Fuel
User Sources (2000): Reference Scenario ~ Low
Nuclear 102
5-5 Summary of Annual NOX Emissions from Fuel
User Sources (2000): Reference Scenario -- High
Nuclear 103
5-6 Year 2000 NOX Mass Emissions Ranking for
Stationary Combustion Equipment and Criteria
Pollutant Cross Ranking 104
5-7 Distribution of Regional Uncontrolled NOX
Emissions (Gg/yr) 1974 107
5-8 NOX EA Field Test Program 109
5-9 Elemental Analysis: Species Determined 113
5-10 POM Analysis: Species Determined 113
viii
-------
LIST OF TABLES (Continued)
Table Page
5-11 Anion Analysis: Species Determined 113
5-12 Bioassay Analysis Protocol 119
7-1 Field Test Program Data Compiled 131
7-2 Individual Test Points Correlated 132
7-3 Effect of Low NOX Operation on Coal-Fired
Boilers 137
7-4 Effect of Low NOX Operation on Oil-Fired
Boilers . . . . 142
7-5 Effect of Low NOX Operation on Gas-Fired
Boilers 145
7-6 Summary of Retrofit Control Costs 149
7-7 Projected Retrofit Control Requirements for
Alternate NOX Emissions Levels 150
7-8 Evaluation of Incremental Emissions Due to NOX
Controls Applied to Boilers 153
8-1 Total Pollution Potential Ranking (Gaseous)
Stationary Sources in Year 1974 160
8-2 Average Source Pollution Potential Ranking (Gaseous)
Stationary Sources in Year 1974 162
8-3 NOX Pollution Potential Ranking Stationary
Sources in 1974 (N02 Basis) 164
8-4 AQCR's Investigated with Preliminary Model 166
8-5 Summary of Control Levels Required to Meet the
Annual Average N02 Standard in San Francisco,
AQCR 030 169
8-6 Definition of Stationary Source NOX Control
Levels 170
8-7 Summary of Control Levels Required to Meet the
Annual Average N02 Standard in St. Louis,
AQCR 070 171
IX
-------
LIST OF TABLES (Concluded)
Table Page
8-8 Effects of NOX and HC Reduction on One Hour
Peak Value of N02 174
8-9 Effect of NOX and HC Reduction on One Hour
Peak Value of 03 174
8-10 Results of Photochemical Trajectory Model
Calculations 179
-------
LIST OF ILLUSTRATIONS
Figure Page
1-1 NOX EA Approach 6
2-1 Sources of Nitrogen Oxide Emissions 11
2-2 Energy Scenarios 33
2-3 National Energy Consumption and Equipment Trends
For Utility Boilers 37
2-4 National Energy Consumption and Equipment Trends
For Packaged Boilers 38
2-5 National Energy Consumption and Equipment Trends
For Residential and Miscellaneous Combustion
Sources 40
4-1 Elements of the Systems Analysis Model 88
5-1 Distribution of Anthropogenic NOX Emissions For
the Year 1974 96
5-2 Regional Fuel Distributions 106
5-3 Analysis Scheme for SASS Train Samples 114
5-4 LC Separation Scheme 115
5-5 SASS Particulate Sample Combining Scheme 116
5-6 Analysis Scheme for Liquid/Solid Samples 117
7-1 Effect of Surface Heat Release Rate and Burner
Stoichiometry on NOX From Tangential Coal-Fired
Boilers 134
7-2 Effect of Heat Input and Burner Stoichiometry on
NOX From Tangential Coal-Fired Boilers 135
8-1 Comparison of Time History of N02 Concentration
at the San Jose Station for Various HC/NOX
Reductions 175
8-2 Map of an Air Parcel Trajectory for High
N02 Day 177
8-3 Comparison of the Effects of Sprawl and Control on
9 am to 12 noon N02 Maximum 182
xi
-------
SECTION 1
INTRODUCTION
This report summarizes the results of the second year of the
"Environmental Assessment of Stationary Source NOX Combustion
Modification Technologies" (NOX EA). The NOX EA is a three year
program to: (1) identify the multimedia environmental impact of stationary
conventional combustion sources and NOX combustion modification controls
applied to these sources; and (2) identify the most cost-effective,
environmentally sound NOX combustion modification controls for attaining
and maintaining current and projected N02 air quality standards to the
year 2000.
During the first year of the program, efforts were concentrated in
three areas:
Compiling background data on combustion source process
characteristics, multimedia pollutant emissions, and pollutant
environmental impacts
0 Developing methodologies for environmental assessment and
process engineering studies
t Setting program priorities on sources, controls, pollutants,
and impacts
Building upon this work, second year emphasis was placed on:
t Characterizing baseline (uncontrolled) combustion source impact
Developing fuels usage and emissions projections
Source testing to fill critical data gaps
Performing process analysis and environmental assessment
studies of NOX controls applied to utility boilers,
industrial boilers, and stationary gas turbines
Assembling and exercising reactive air quality models for
systems analysis applications
-------
Developing source analysis models for environmental impact
evaluation
This report summarizes the results of the second year activities and the
plans for the third year.
1.1 BACKGROUND
The 1970 Clean Air Act Amendments designated oxides of nitrogen
(NOX) as one of the criteria pollutants requiring regulatory controls to
prevent potential widespread adverse health and welfare effects.
Accordingly, in 1971, EPA set a primary and secondary National Ambient Air
Quality Standard (NAAQS) for N02 of 100 ug/m3 (annual average). To
attain and maintain the standard, the Clean Air Act mandated control of
new mobile and stationary NOX sources, each of which emits approximately
half of the manmade NOX nationwide. Emissions from mobile source, light
duty vehicles were to be reduced by 90 percent to a level of 0.25 g
N02/km (0.4 g/mile) by 1976. Stationary sources were to be regulated by
EPA standards of performance for new stationary sources (NSPS), which are
set as control technology becomes available. Additional standards
required to attain air quality in the Air Quality Control Regions could be
set for new or existing sources through the State Implementation Plans
(SIP's).
Since the Clean Air Act, techniques have been developed and
implemented that reduce NOX emissions by a moderate amount (30 to 50
percent) for a variety of source/fuel combinations. In 1971 EPA set NSPS
for large steam generators burning gas, oil, and coal (except lignite).
Currently, more stringent standards for coal-fired large utility steam
generators have been proposed, based on technology developed since 1971.
Standards have also been proposed for gas turbines and are being prepared
for reciprocating internal combustion engines and intermediate sized steam
generators. Local standards also have been set, primarily for new and
existing large steam generators and gas turbines, as parts of the State
Implementation Plans in several areas with NOX problems. This
regulatory activity has resulted in reducing NOX emissions from many
stationary sources by 30 to 50 percent. The number of controlled sources
is increasing as new units are installed with factory equipped NOX
controls.
Emissions have been reduced comparably for mobile source, light
duty vehicles. Although the goal of 90 percent reduction (to 0.25 g
N02/km) by 1976 has not been achieved, emissions were reduced by about
25 percent (1.9 g/km) for the 1974 to 1976 model years and now have been
reduced by 50 percent to 1.25 g/km. Achieving the 0.25 g/km goal has been
deferred indefinitely because of technical difficulties and fuel
penalties. Initially the 1974 Energy Supply and Environmental
Coordination Act deferred compliance to 1978. Recently, the Clean Air Act
Amendments of 1977 abolished the 0.25 g/km goal and replaced it with an
emission level of 0.62 g/km (1 g/mile) for 1981 and beyond. However, the
EPA Administrator is required to review the 0.25 g/km standard in 1980 and
report to Congress on the need for such a standard.
-------
Because the mobile source emission regulations have been relaxed,
stationary source NOX control has become more important for maintaining
air quality. Several air quality planning studies have evaluated the need
for stationary source NOX control in the 1980's and 1990's in view of
recent developments (References 1-1 through 1-4). These studies all
conclude that relaxing mobile standards, coupled with the continuing
growth rate of stationary sources, will require more stringent stationary
source controls than current NSPS provide. This conclusion has been
reinforced by projected increases in the use of coal in stationary
sources. The studies also conclude that the most cost-effective way to
achieve these reductions is by using combustion modification NOX
controls in new sources.
It is also possible that separate NOX control requirements will
be needed to attain and/or maintain additional N02 related standards.
Recent data on the health effects of N02 suggest that the current NAAQS
should be supplemented by limiting short term exposure (References 1-4
through 1-8). In fact, the Clean Air Act Amendments of 1977 require EPA
to set a short term N02 standard for a time period of less than three
hours unless no need for such a standard can be verified.
EPA is also continuing to evaluate the long range need for
additional NOX regulation as part of strategies to control oxidants or
pollutants for which NOX is a precursor, e.g., nitrates and nitrosamines
(References 1-4, 1-5, and 1-9 through 1-12). These regulations could be
source emission controls or additional ambient air quality standards. In
either case, additional stationary source control technology could be
required to assure compliance.
In summary, since the Clean Air Act, near term trends in NOX
control are toward reducing-stationary source emissions by a moderate
amount. Hardware modifications in existing units or new units of
conventional design will be stressed. For the far term, air quality
projections show that more stringent controls than originally anticipated
will be needed. To meet these standards, the preferred approach is to
control new sources by using low NOX redesigns.
1.2 PROGRAM OVERVIEW
Existing combustion modification techniques are increasingly being
used on stationary conventional combustion sources, and the prospects for
developing and using advanced techniques are good. Identifying combustion
generated pollutant species from these sources and evaluating their
potential environmental impacts have become increasing concerns. Thus, a
critical need exists to not only evaluate the baseline environmental
impact of conventional stationary source combustion, but also to evaluate
the environmental, economic, energy, and engineering implications of
combustion modification technology. The NOX EA was begun in June 1976
to provide such evaluations and specifically assess:
The impacts and potential correction measures associated with
using specific existing and advanced combustion modification
techniques, such as:
-------
The change in gaseous, liquid, and solid emissions to the
air, water, and land caused by NOX controls
-- The capital and operating cost of NOX controls per unit
reduction in NOX
The change in energy consumption efficiency
-- The change in equipment operating performance
The priorities and schedule for NOX control technology
development considering:
The above impacts for each source/control combination
-- The need for controls to attain and maintain the current
annual average N02 ambient air quality standard
The need for controls to attain and maintain a short term
N02 standard or other NOX related standards such as a
standard for oxidants
Alternate mobile source standards
Alternate energy and equipment use scenarios, to the year
2000, in the Air Quality Control Regions with potential
NOX problems
The first assessment concerns evaluating the net impacts from
specific combinations of stationary combustion source equipment and
control techniques. The NOX EA addresses this goal through a series of
coordinated efforts to evaluate the environmental impact and control
potential of multimedia effluents from current and emerging energy and
industrial processes. The assessment effort is focused in a major process
engineering and environmental assessment task. This task is supported by
additional tasks on emission characterization, pollutant impacts and
standards, and experimental testing. Results from these tasks will be
used to rank both current and emerging source/control combinations based
on overall environmental, economic, and operational impact. This
information is intended to help control developers and users select
appropriate control techniques to meet regulatory standards now and in the
future. It also will help define pollution control development needs and
priorities, identify economic and environmental trade-offs among
competitive processes, and ultimately guide regulatory policy. In this
respect, the NOX EA will contribute to the broad program of assessments
of energy systems and industrial processes being administered by EPA's
Office of Research and Development.
The second assessment above deals with specifying the best mix of
control techniques to meet air quality goals up to the year 2000. In the
NOX EA, this is done in a systems analysis task which projects air
quality in specific air quality control regions for various scenarios of
NOX control, energy growth, and equipment use. These projections,
-------
together with the control cost and impact data discussed above will
suggest the most cost-effective and environmentally sound controls.
Results from the analysis are used in the NOX EA program to set
priorities on both sources and controls. More importantly, this
information will help R&D groups concerned with providing a sufficient
range of environmentally sound techniques to meet the diverse control
implementation requirements. It will also help environmental planners
involved in formulating abatement strategies to meet current or projected
air quality standards.
The interrelationships and technical content of each of the tasks
in the NOX EA are shown in Figure 1-1. In this figure the arrows
indicate the sequence of subtasks and major interactions among tasks, the
boxes represent task efforts, and the ovals represent program outputs.
As noted above, second year efforts focused on characterizing
baseline source impact; developing fuels usage and emissions projections;
source testing; evaluating process, cost, and environmental impacts of
NOX controls applied to utility boilers; assembling and exercising
reactive air quality and systems analysis models; and source analysis
modeling. In this report, results are presented in terms of these areas
rather than on a task by task basis. This approach is consistent with
general environmental assessment annual report formats developed within
the lERL-RTP's Energy Assessment and Control Division. Thus, specific
task efforts are discussed herein as follows:
Task/Subtask Report Section
Emissions Characterization
Combustion source process/emissions
background 2.1
-- Stationary source fuel consumption 2.2
Equipment/fuels use projections 2.3
-- Multimedia emissions inventory 5.1
~ Baseline source impact ranking 8.1
Impacts and Standards
-- N02 and related standards projections 3
Experimental Testing
Sampling/analysis requirements;
field test program 5.2
t Source Analysis Modeling
Methodology development 4.1
Process Engineering and Environmental
Assessment
NOX control process background 6
-- Process engineering methodology
development 4.2
-- Detailed process studies (utility boilers) 7
Systems Analysis
Air quality model development 4.3
Control needs evaluation 8.2
-------
EMISSIONS
CHARACTERIZATION (B1)
COMPILE COMBUSTION
SOURCE PROCESS/EMISSIONS
BACKGROUND
GENERATE MULTIMEDIA
EMISSIONS INVENTORY
ASSEMBLE EMISSIONS
PROJECTIONS: COMPILE
REGIONAL VARIATIONS
COMPARE BASELINE
EMISSIONS TO MEGS
UPDATE EMISSIONS.
PROJECTIONS
IMPACTS AND
STANDARDS (B2)
EXPERIMENTAL
TESTING (B3)
SOURCE ANALYSIS
MODELING |D)
ASSESSMENT
CHARACTERIZE PRIMARY AND
SECONDARY MULTIMEDIA
POLLUTANTS
ASSEMBLE MULTIMEDIA
ENVIRONMENTAL GOALS
(MEGS)
PROJECT NO2
ENVIRONMENTAL GOALS
[ IMPACT CRITERIA: |
I STANDARDS PROJECTIONS i
Q1ASELINE IMPACT \ P
ISSESSMENT/RANKING 1 I
JUSTIFY AND
UPDATE MEGS
UPDATE NO2
GOALS
DEVELOP METHODS TO
COMPARE EFFLUENT
CONCENTRATIONS TO MEGS
FOR IMPACT EVALUATION
ASSEMBLE POLLUTANT
DISPERSION/DILUTION
FACTOR MODELS
OUTLINE PROCEDURES FOR
RAPID SCREENING.
SITE EVALUATION
FORMAT
.OPMENT: SAM IA
SAM I
PROCESS ENGINEERING AND
ENVIRONMENTAL ASSESSMENT (B5)
COMPILE NOl CONTROL
PROCESS BACKGROUND
EVALUATE INCREMENTAL
EMISSIONS DATA WITH
NOx CONTROLS
(PRELIMINARY SOURCE/ \
CONTROL/POLLUTANT L,
PRIORITIES i
DEVELOP PROCESS
ENGINEERING/EA METHODS
COMPILE DETAILED SOURCE/
CONTROL PROCESS DATA
CONDUCT DETAILED
PROCESS
STUDY FOR:
UTILITY BOILERS
INDUSTRIAL BOILERS
GAS TURBINES
RESIDENTIAL HEATING
1C ENGINES
INDUSTRIAL PROCESS
ADVANCED COMBUSTION
ENVIRONMENTAL
ASSESSMENT OF
COMBUSTION SOURCES
AND NO, CONTROLS
SYSTEMS
ANALYSIS (C)
DEVELOP PRELIMINARY
MODEL FOR
ENVIRONMENTAL
ALTERNATIVE
ANALYSES
SCREEN CONTROL
REQUIREMENTS FOR
ATTAINING/MAINTAINII
AIR QUALITY
SELECT AND ADAPT
REACTIVE AIR QUALITY
MODEL
PROJECT SOURCE OROWTH
AND AMBIENT STANDARDS
ASSESS CONTROL NEEDS
FOR VARIOUS REGULATORY
REQUIREMENTS
2ND YEAR
EFFORT
/MOST EFFECTIVE CONTROL
I OPTIONS. CONTROL AND
V REQUIREMENTS
, 3RD YEAR
EFFORT
Figure 1-1. NO EA approach.
-------
In addition to the above, results from another support task not
noted in Figure 1-1 are also summarized below. Updated conclusions from
this task, to survey the potential for alternate clean fuels use in area
sources, are reported in Section 2.4. Technology transfer activities
performed as part the general NOX EA program support task are summarized
in Section 9. Finally, third year plans are discussed in Section 10.
-------
REFERENCES FOR SECTION 1
1-1. Crenshaw, J. and A. Basala, "Analysis of Control Strategies to
Attain the National Ambient Air Quality Standard for Nitrogen
Dioxide," presented at the Washington Operation Research Council's
Third Cost Effectiveness Seminar, Gaithersburg, MD, March 1974.
1-2. "Air Quality, Noise and Health Report of a Panel of the
Interagency Task Force on Motor Vehicle Goals Beyond 1980,"
Department of Transportation, March 1976.
1-3. McCutchen, G. D., "NOX Emission Trends and Federal Regulation,"
presented at the AIChE 69th Annual Meeting, Chicago, November 1976.
1-4. "Air Program Strategy for Attainment and Maintenance of Ambient Air
Quality Standards and Control of Other Pollutants," Draft Report,
U.S. EPA, Washington, October 1976.
1-5. French, J. G., "Health Effects from Exposure to Oxides of
Nitrogen," presented at the AIChE 69th Annual Meeting, Chicago,
November 1976.
1-6. "Scientific and Technical Data Base for Criteria and Hazardous
Pollutants 1975 EPA/RTP Review," EPA-600/1-76-023, NTIS PB-253
942, January 1976.
1-7. Shy, C. M., "The Health Implications of Non-Attainment Policy,
Mandated Auto Emission Standards, and a Non-Significant
Deterioration Policy," presented to Committee on Environment and
Public Works, Serial 95-H7, February 1977.
1-8. "Health Effects for Short-term Exposure to Nitrogen Dioxide
(Draft)," EPA, Office of Research and Development, December 1977.
1-9. "Control Strategy for Nitrogen Oxides," Memo from B. J.
Steigerwald, Office of Air Quality Planning and Standards,
September 1976.
1-10. "Report on Air Quality Criteria: General Comments and
Recommendations," Report to the U.S. EPA by the National Air
Quality Advisory Committee of the Science Advisory Board, June 1976.
1-11. Personal communication with M. Jones, Office of Air Quality
Planning and Standards, Pollutant Strategies Branch, September 1976.
1-12. "Control of Photochemical Oxidants Technical Basis and
Implications of Recent Findings," EPA-450/2-75-005,
NTIS PB-242 428, July 1975.
-------
SECTION 2
CURRENT PROCESS TECHNOLOGY BACKGROUND
During the second year of the NOX EA the equipment
characterizations, fuels use compilations, and emissions inventories
developed during initial program efforts and documented in References 2-1
and 2-2 were further defined and updated. Results from this continuing
work are presented here in Section 2, and in Sections 5 and 8.
This section presents the combustion process technology background
used to order and simplify the NOX EA process engineering and
environmental assessment studies. The section characterizes equipment
designs according to characteristics that affect the formation and control
of multimedia pollutants. Although emphasis is on the stationary
combustion sources of NOX, other sources were also studied because the
need for stationary source controls depends on how well these other
sources can be controlled. The equipment categories described here are
used as the basis for the emissions inventory in Section 5.1 and the
source rankings discussed in Section 8. The source characterization
considered the following steps:
Identify significant sources of NOX; group sources according
to formative mechanism and nature of release into the atmosphere
t Categorize stationary combustion sources according to equipment
and fuel characteristics that affect the generation and/or
control of combustion generated pollution
Qualify equipment/fuel categories on the basis of current and
projected use and design trends; develop a list of
equipment/fuel combinations to be carried through subsequent
emission inventories, process studies, and environmental
assessments
Identify effluent streams from stationary combustion source
equipment/fuel categories which may be affected by using NOX
combustion modification controls
Identify operating modes (transients, upsets, maintenance) in
which emissions may be affected by NOX combustion
modification controls
-------
The significant sources of oxides of nitrogen emitted to the
atmosphere are shown in Figure 2-1. On a global basis, natural emissions
from biological decay and lightning make up about 90 percent of all NOX
emissions. However, in urban areas, up to 90 percent of ambient NOX may
be due to manmade sources, primarily combustion effluent streams. The
primary emphasis of the NOX EA is on the fuel combustion sources
bracketed at the top of the figure. Other sources are considered only to
gauge the relative emissions and impacts due to stationary fuel combustion.
The major stationary fuel combustion source classes have been
further categorized as shown in Table 2-1. This table lists the major
equipment designs and corresponding fuels fired, and was compiled from a
survey of installed sources, process characteristics, and emission data.
In the following, Section 2.1 discusses major stationary source
equipment designs which have a significant impact on NOX emissions. The
emphasis is on standard operating conditions for these sources, though
nonstandard conditions are given cursory evaluation. Stationary source
fuel consumption is characterized in Section 2.2. These data are
important inputs for the emission inventories in Section 5. In addition,
several energy scenarios through the year 2000, which bracket the
uncertainty in future conditions, are discussed. Section 2.3 considers
trends in equipment and fuels which are used to project energy use by
sector through 2000.
In addition to the source characterization, fuels use, and energy
projection efforts discussed in Sections 2.1 through 2.3, a related study
to characterize the potential for alternate clean fuel usage in area
sources continued during the second year of the NOX EA. Updated results
from this task are discussed in Section 2.4.
2.1 STATIONARY COMBUSTION PROCESS BACKGROUND
Stationary combustion sources, noted above as having a significant
impact on NOX emissions, are discussed in this section. Tables are
provided which list the major designs in each sector and the variations in
designs and fuels which are known to affect emissions. The primary design
types are those projected for widespread use in the 1980's and thus are
prime candidates for NOX control application. Secondary designs are
defined as those either diminishing in use or unlikely candidates for
NOX controls in the future. Secondary design types have not, and will
not be given further consideration in subsequent NOX EA studies.
2.1.1 Utility and Large Industrial Boilers
Utility and large industrial boilers are defined as field erected
watertube boilers with capacities greater than 73 MW heat input. These
boilers generally burn pulverized coal, residual oil, and natural gas.
Table 2-2 describes the variety of specific boiler designs and catalogs
multimedia pollutant emissions from these sources. Further discussion of
pollutant emissions from this source category under standard operation is
given in Section 5.1.
10
-------
Sources of
nitrogen _
ox1 des
Combustion
'effluent stream
emissions
Noncombustlon
effluent
stream
emissions
Fugitive
emlsslons-
rStatlonary-
rFuel
confcustlon
-Incineration
Mobile
rNatural-
- Anthropogenic
-Utility Boilers
-Packaged Boilers
-Warm A1r Furnaces
-Gas Turbines
-Reciprocating 1C Engines
-Industrial Process Combustion
-Advanced Combustion Processes
Emphasis
of
NOX E/A
-Nitric acid
HW1p1c acid
Explosives
-Fertilizer
-Nitration
Nitrogen cycle
Lightning
-Open burning
"Forest fires
-Structural fires
-Minor processes
Figure 2-1. Sources of nitrogen oxide emissions.
-------
TABLE 2-1. SIGNIFICANT STATIONARY FUEL COMBUSTION EQUIPMENT
TYPES/MAJOR FUELS
Utility Sector (Field Erected Watertubes) Fuel
Tangential
Wall Fired
PC, 0, G
PC, 0, G
Horizontally Opposed and Turbo Furnace PC, 0, G
Cyclone
Vertical and Stoker
Packaged Boiler Sector
Watertube 29 to 73 MWa (100
PC, 0
C
to 250 MBtu/hr) C, 0, G,
Watertube <29 MWa (<100 MBtu/hr) C, 0, G,
Firetube Scotch
Firetube HRT
Firetube Firebox
Cast Iron
Residential
Warm Air Furnace Sector
Central Heaters
Space Heaters
0, G, PG
C, 0, G,
C, 0, G,
0, G
C, 0, G
0, G
0, G
PG
PG
PG
PG
Other Residential Combustion 0, G
PC -- Pulverized coal
C -- Stoker coal or other coal
0 --Oil
G ~ Gas
PG Process gas
aHeat input
bHeat output
12
-------
TABLE 2-1. Continued
Gas Turbines
Large >15 MWb (>20,000 hp)
Medium 4 to 15 MWb (5,000 to 20,000 hp)
Small <4 MWb (<5,000 hp)
Reciprocating 1C Engines
Large Bore >75 kW/cylb (>100 hp/cyl)
Medium 75 kW to 75 kW/cylb (100 hp to
100 hp/cyl)
Small <75 kWb (<100 hp)
Industrial Process Heating
Glass Me Hers
Glass Annealing Kilns
Cement Kilns
Petroleum Refinery
Process Heaters
Catalytic Crackers
PC Pulverized coal
C -- Stoker coal or other coal
0 Oil
G Gas
PG Process gas
aHeat input
t>Heat output
Fuel
0, G
0, G
0, G
0, G
0, G
0, G
0, G
0, G
C, 0, G
0, G, PG
0, G, PG
13
-------
TABLE 2-1. Concluded
Brick and Ceramic Kilns
Iron and Steel
Coke Oven Underfire
Sintering Machines
Soaking Pits and Reheat Ovens
Fuel
0, G
G, PG
0, G, PG
C, 0, G, PG
PC Pulverized coal
C ~ Stoker coal or other coal
0 -- Oil
G -- Gas
PG -- Process gas
14
-------
TABLE 2-2. SUMMARY OF UTILITY AND LARGE INDUSTRIAL BOILER CHARACTERIZATION
Design Type
Tangential
Single Uall
Design
Characteristics
Fuel and air nozzles
In each corner of
the combustion
chamber are directed
tangent 1 ally to a
small firing circle
in the chamber.
Resulting spin
of the flames mixes
the fuel and air In
the combustion zone.
Burners mounted
on single furnace
wall up to
72 on single wall
Process Ranges
Input Capacity:
73 MW to 3800 MH
Steam Pressure:
18.6 MPa (subcritlcal)
26.2 MPa (supercritical)
Steam Temperature:
755k to 840k
Furnace Volume:
Up to 38,000 m
Furnace Pressure:
50 Pa to
1000 Pa
Furnace Heat Release:
Coal 104 to 250
kW/m
Oil, gas 208 to
518 kW/m
Excess Air:
25" X coal
10* oil
8* gas
Units typically limited
in capacity to about
400 MW (electric) because
of furnace area.
Fuel Consumption
(*)
67* coal
18* oil
15* gas
43* coal
22* oil
35* gas
Effluent Streams
Gaseous
Flue gas contain-
ing flyash, vola-
tilized trace
elements, SOj ,
NO, other
pollutants.
Liquid
Scrubber streams,
ash sluicing
streams, wet
bottom slag
streams.
Solid
Solid ash removal
Flyash removal
Gaseous
Flue gas contain-
ing flyash, vola-
tilized trace
elements, S02,
NO, other .
pollutants.
Liquid
Scrubber streams,
ash sluicing
streams, wet
bottom slag
streams.
Solid
Solid ash removal
Flyash removal
Operating
Modes
Soot blowing, on-
off transients,
load transients,
upsets, fuel
additives, rap-
ping, vibrating.
Soot blowing, on-
off transients
load transients,
upsets, fuel
additives, rap-
ping, vibrating.
Effects of Transient,
Nonstandard
Operation
During startup,
NO emissions are
low since flame
temperatures not
developed. Dur-
ing load reductions,
emissions of NOX
decrease because
of lower flame
temperatures.
NOX should de-
crease following
soot blow due to
Improved heat
transfer.
During startup,
NO emissions are
low since flame
temperatures not
developed. During
load reductions,
emissions of NO
decrease because
of lower flame
temperatures.
NOX should de-
crease following
soot blow due to
Improved heat
transfer.
Trends
Trend toward
coal firing in
new units; con-
version to oil
and coal in
existing units.
19.4* of current
installed units.
Trend toward
coal firing in
new units; wet
bottom units no
longer manufac-
tured due to
operational
problems with
low sulfur coals
and high combus-
tion tempera-
tures promoting
HOX.
59* of current
installed units.
Future
Importance
Primary
Primary
cn
-------
TABLE 2-2. Continued
Design Type
Horizontally
Opposed Wall
Turbo
Furnace
Design
Characteristics
Burners are mounted
on opposite furnace
walls up to 36
burners per wall.
A1r and fuel fired
down toward furnace
bottom using burners
spaced across
opposed furnace
walls. Flame propo-
gates slowly passing
vertically to the
upper furnace. NO
1s usually low due
to long combustion
time and relatively
low flaoe tempera-
ture.
i
Process Ranges
Units typically designed
In sizes greater than
400 MM (electric).
Units typically designed
In sizes greater than
400 MM (electric).
Fuel Consumption
(ซ)
32X coal
21% oil
47X gas
(Includes Turbo-
Furnace)
Included 1n
horizontally
opposed wall
Effluent Streams
Gaseous
Flue gas contain-
ing flyash,
volatilized trace
elements, SO ,
NO. other
pollutants.
Liquid
Scrubber streams.
ash sluicing
streams, wet
bottom slag
streams.
Solid
Solid ash removal
Flyash removal
Gaseous
Flue gas contain-
ing flyash.
volatilized trace
elements, SO ,
NO, other
pollutants.
Liquid
Scrubber streams,
ash sluicing
streams, wet
bottom slag
streams.
Operating
Modes
Soot blowing, on-
off transients,
load transients.
upsets, fuel
additives, rap-
ping, vibrating.
Soot blowing, on-
off transients,
load transients.
upsets, fuel
additives, rap-
ping, vibrating.
Effects of Transient,
Nonstandard
Operation
During startup.
NO emissions are
low since flame
temperatures not
developed. Dur-
ing load reductions.
emissions of NO
decrease because
of lower flame
temperatures.
NOX should de-
crease following
soot blow due to
improved heat
transfer.
During startup,
NO emissions are
low since flame
temperatures not
developed. Dur-
ing load reductions.
emissions of NO
decrease because
of lower flame
temperatures.
NO should de-
crease following
soot blow due to
Improved heat
transfer.
Trends
Trend toward
coal firing and
conversions to
oil and coal
firing; again,
wet bottoms
being phased
out.
8.2X of current
Installed units.
Trend toward
coal firing
(capacity In-
cluded with
opposed wall).
Future
Importance
Primary
Primary
' T-847
-------
TABLE 2-2. Concluded
Design Type
Cyclone
Vertical and
Stoker
Design
Characteristics
Fuel and air Intro-
duced clrcumferen-
tlally Into cooled
furnace to produce
swirling, high tem-
perature flame;
cyclone chamber
separate from main
furnace; cyclone
furnace must operate
at high temperatures
since It 1s a slag-
ging furnace.
Vertical firing re-
sults from downward
firing pattern.
Used to a limited
degree to fire
anthracite coal.
Stoker projects fuel
Into the furnace
over the fire per-
mitting suspension
burning of fine
fuel particles.
Spreader stokers
are the primary
design type.
Process Ranges
Furnace Heat Release:
4.67 to 8.28 MM/m
Furnace Heat Release:
1.1 to 1.9 MW/m^ plan
area
Fuel Consumption
(X)
92% coal
ซ oil
4X gas
100X coal
Effluent Streams
Gaseous
Flue gas contain-
ing flyash, vola-
tilized trace
elements, SO ,
NO, and other
pollutants.
Liquid
Scrubber streams
Solid
Solid ash removal
Flyash removal
Gaseous
Flue gas contain-
ing flyash, vola-
tilized trace
elements, SO ,
NO, and other
pollutants.
Liquid
Scrubber streams
Solid
Solid' ash removal
Flyash removal
Operating
Modes
Soot blowing, on-
off transients,
load transients,
upsets, fuel
additives, rap-
ping, vibrating.
Soot blowing, on-
off transients
load transients.
upsets, fuel
additives, rap-
ping, vibrating.
Effects of Transient
Nonstandard
Operation
During startup.
NO emissions are
low since flame
temperatures not
developed. Dur-
ing load reductions,
emissions of NO
decrease because
of lower flame
temperatures.
NO should de-
crease following
soot blow due to
Improved heat
transfer.
During startup.
NO emissions are
low since flame
temperatures not
developed. Dur-
ing load reductions,
emissions of NO
decrease because
of lower flame
temperatures.
NO should de-
crease following
soot blow due to
Improved heat
transfer.
Trends
Two cyclone
boilers sold
since 1974
have not proven
adaptable to
emissions regu-
lations. Must
operate at high
temperatures re-
sulting In high
thermal NO
fixation; also
operational
problems with
low sulfur coal.
3.3% of Installed
units.
Since anthracite
usage has de-
clined, vertical
fired boilers are
no longer sold.
Design capacity
limitations and
high cost have
caused stoker
usage to diminish.
9.9% of current
Installed units.
Future
Importance
Secondary
Secondary
T-847
-------
Table 2-2 also lists the effects of nonstandard operating
conditions on emissions. Unfortunately, emissions during nonstandard
operation have not been extensively quantified. During startup NOX
emissions are generally low because flame temperatures have not
developed. However, particulate emissions may be high since precipitators
are generally not energized during startup. Also, unburned carbon may be
emitted due to poor mixing in the combustion region.
NOX emissions should decrease as furnace temperatures are lowered
during load reductions. However, if excess air levels are increased to
maintain steam temperatures, NOX emissions actually may increase.
Particulate emissions increase during soot blowing as the tube
surfaces are cleaned. However, NOX emissions should decrease after soot
blowing because of the lower gas temperatures caused by increased heat
transfer through the tube walls. Failure of equipment such as air
preheaters may also reduce NOX emissions by causing lower flame zone
temperatures.
2.1.2 Packaged Boilers
The packaged boiler category includes all industrial, commercial,
and residential packaged boilers. Generally, these boilers have
capacities less than 73 MW thermal input. There are only a few packaged
boilers with larger capacities and these are sufficiently similar to the
smaller units to be included in this category. Table 2-3 describes the
classes of packaged boilers and emissions from standard and nonstandard
operation. Further discussion of multimedia pollutant emissions under
standard operation is given in Section 5.1.
Since large packaged boilers (>29 MW or 100 MBtu/hr heat input)
operate much like utility boilers, the effects of transients and
nonstandard operations should be similar to those discussed in Section
2.1.1. For smaller packaged boilers, combustion characteristics are
significantly different. Although quantitative data for nonstandard
operating conditions are sparse, load changes are known to have a
relatively small effect on NOX emissions (Reference 2-3). However,
increasing the fuel preheat temperature of oil-fired boilers may increase
NOX emissions. At low preheat temperatures, the atomizing pressure is
not sufficient to properly atomize the colder, more viscous oil; this
results in lower atomization efficiency.
2.1.3 Warm Air Furnaces and Other Commercial and Residential
Combustion Equipment
This sector is made up of residential and commercial warm air
furnaces used for comfort heating, and miscellaneous commercial and
residential appliances used in cooking, refrigeration, air conditioning,
clothes drying, and the like. NOX EA emphasis is on space heaters,
where the unit is located in the room or area it heats, and central
heaters, which use ducts to transport and discharge warm air into the
heated space. Table 2-4 describes these equipment categories and the
effects of operating conditions on multipollutant emission levels.
18
-------
TABLE 2-3. SUMMARY OF PACKAGED BOILER CHARACTERIZATION
Design Type
Hater-tube
Scotch
Firetube
HRT Fire-
tube
Design
Characteristics
Combustion gases
circulate around
boiler tubes that
have water passing
through them.
Essentially the
only type of boiler
available above 29
MU (heat input).
Cylindrical shell
with one or more
furnaces 1n the
lower portion. Com-
bustion takes place
in front section.
Combustion products
flow back to rear
combustion chamber.
flow through tubes
to smoke box, then
discharge.
Hot gases pass to
back of unit, enter
horizontal tubes.
returning to front
of the boiler then
exit through smoke
box.
Typical
Operational
Values
Oil-Fired Watertube:
Capacity: 38 MU
Furnace volume:
123 m3
Heat release:
310 kw/m3
Burner type:
steam a torn iz at ion
Fuel preheat:
392K
Stack temperature:
422K
Excess oxygen: 5<
Scotch Firetube-Oil:
Capacity: Z.9 NU
Furnace volume:
2.5 m3
Heat release:
1190 kW/ป3
Operating pressure:
1030 Pa
Burners:
Air atomizing (2)
Fuel preheat:
37 IK
Excess oxygen:
4.91
Fuel Consumption
ซ)
41% coal
21* oil
38* gas
59X oil
41* gas
55* oil
35* gas
Effluent Streams
Gaseous
Flue gas
Part icu late
Liquid
Ash sluicing
water
Scrubber streams
Solids
Solid ash
removal
Flue gas
Bottom ash
Flue' gas
Bottom ash
Operating
Modes
Soot blowing,
on-off transients,
upsets, fuel ad-
ditives.
On-off transients,
load transients,
upsets, fuel ad-
ditives.
On-off transients,
load transients.
upsets, fuel ad-
ditives.
Effects of Transient,
Nonstandard
Operation
During startup,
low NOx emissions.
During load
reductions NOx
lowered.
Soot blowing
should cause lower
gas temperature
due to improved
heat transfer.
thus lowering NOx.
Changes in firing
rate have little
effect on NOx
emissions from
firetubes. Fuel
oil temperature
increases tend to
decrease NOx
emissions.
Changes in firing
rate have little
effect on NOx
emissions from
firetubes. Fuel
oil temperature
increases tend to
decrease NOx
emissions.
Trends
Pulverized coal
and stokers for
large watertubes.
Scotch firetubes
currently show
growth over other
firetube designs.
Trend toward de-
creasing use of.
HRT.
Future
Importance
Primary
Primary
Secondary
T-849
-------
TABLE 2-3. Concluded
Design Type
Firebox
Firetube
Cast Iron
Boilers
Steam and
Hot Hater
Units
ซ
Design
Characteristics
Combustion gases
enter front of first
tube pass, travel to
rear smoke box. re-
turn through second
pass to gas outlet
at the boiler front.
Gases rise through
vertical section,
and discharge
through the exhaust
duct. Water Is heat-
ed as It passes up-
wards through the
watertubes
Besides small resi-
dential units, shell
boilers, coop act,
locomotive, short
firebox, vertical
flretube, straight
tube, and coal re-
search designs are
grouped here.
Typical
Operational
Values
Cast Iron:
Distillate oil
Capacity: 0.38 MU
Furnace volume:
0.57 n3
Heat release:
673 kW/o3
Operating pressure:
1030 Pa
Burner type:
Pressure
atomizing (1)
Fuel preheat:
Hone
Excess oxygen:
4.4X
Fuel Consumption
(X)
53X oil
57X gas
59% oil
41X gas
1.5X coal
56X oil
42. 5X gas
Effluent Streams
Flue gas
Bottom ash
Flue gas
Bottom ash
Flue gas
Operating
Nodes
On-off transients,
load transients,
upsets, fuel ad-
ditives.
On-off cycling
transients
On-off cycling,
transients.
Effects of Transient,
Nonstandard
Operation
Changes 1n firing
rate have little
effect on NOx emis-
sions from flretubes.
Fuel oil temperature
Increase tends to de-
crease NOx emissions.
Trends
Decreasing use of
firebox flretubes
Future
Importance
Secondary
Secondary
Secondary
T-849
ro
o
-------
TABLE 2-4. SUMMARY OF WARM AIR FURNACES CHARACTERIZATION
Design Type
Commercial
and Resi-
dential
Central
Warm Air
Furnaces
Space
Heaters
Design
Characteristics
Furnaces in central
heaters enclosed in
steel casing; fuel
burned in combustion
space of heat ex-
changers. Heat ex-
changers have a
single combustion
chamber, either
cylindrical or di-
vided into indivi-
dual sections;
combustion gases
pass through secon-
dary gas passages of
the heat exchanger
and exit through
flue.
Room heaters self-
contained; equipped
with a flue. Heat
by radiation, or
natural or forced air
circulation.
Design Ranges
Typical Gas-Fired Forced
Air Furnace
Heat exchanger
area: 2.8 to 3.3 m2
Draft system:
natural
Excess combustion air:
20X to SOX
Overall heat
transfer coefficient:
11.3 to 17 W/mZK
Combustion chamber
pressure:
+ 49.8 Pa
ExTt flue gas
temperature:
506 to 617K
Overall efficiency:
75X to 80*
On-off operation
Fuel Consumption
(*)
31* distillate
oil
69X gas
(Miscellaneous
combustion fuels
such as wood, LPG,
etc. combined
with natural gas)
23% distillate
oil
73X gas
(Miscellaneous
combustion fuels
such as wood, LPG,
etc., combined
with natural gas)
(Includes other
residential com-
bustion).
Effluent Streams
Flue gas
Flue gas
Operating
Modes
On-off cycling
transients
On-off cycling
transients
Effects of Transient,
Nonstandard
Operation
NOx emissions levels
rise at a steady rate
after initial jump due
to ignition, drop off
quickly after the
burner is turned off.
NOx emissions increase
with on time of
burner. Improper
burner adjustment.
damaged components,
Increase NOx by as
much as SOX.
NOx emissions levels
rise at a steady rate
after initial jump due
to Ignition, drop off
quickly after the
burner is turned off.
NOx emissions increase
with on time of
burner. Improper
burner adjustment,
damaged components,
increase NOx fay as
much as SOX.
Trends
Oil firing in new
units, trend to
high efficiency In
new units.
General decline in
natural gas usage;
Increase in elec-
tric heat, in-
creased use of high
efficiency burners.
Oil firing in new
units, trend to
high efficiency 1n
new units.
General decline in
natural gas usage;
increase In elec-
tric heat, in-
creased use of high
efficiency burners.
Future
Importance
Primary
Secondary
T-BbO
ro
-------
TABLE 2-4. Concluded
Design Type
Other
Residential
Combustion
Design
Characteristics
Miscellaneous
equipment Includes
ranges and ovens,
clothes dryers.
fireplaces, swimming
pool heaters, re-
frigerating and air-
conditioning equip-
ment.
Design Ranges
Fuel Consumption
(X)
Included 1n
space heaters
Effluent Streams
Flue gas
Operating
Modes
On-off cycling,
transients
Effects of Transient,
Nonstandard
Operation
NOx emissions levels
rise at a steady rate
after Initial Jump due
to Ignition, drop off
quickly after the
burner 1s turned off.
NOx emissions Increase
with on time of
burner. Improper
burner adjustment,
damaged components,
Increase NOx by as
much as SOU.
Trends
Increased use of
electric heat;
high efficiency
burners 1n new
units.
Future
Importance
Secondary
T-850
PO
INJ
-------
Emissions inventory data under standard operating conditions are given in
Section 5.1.
The transient and nonstandard operations of warm air furnaces
include on-off cycling and out of tune or worn burner operations
(Reference 2-4).
The initial peak in emission levels at ignition is caused by the
inability of the cold refractory to support complete combustion. This
incomplete combustion produces peaks in the HC, CO, and particulate
emissions. As the refractory warms up, more complete combustion occurs,
thus decreasing combustible emissions. After shutdown, some fuel leaks
from the nozzle, which produces another peak in both the CO and HC
emissions (Reference 2-5). This can be controlled to some degree by using
a solenoid.
The transient emissions of NOX generally correspond to the
thermal history of the firebox. At startup, the emissions increase
rapidly as the temperature rises above the thermal NOX threshold.
During the cycle, the emissions continue to increase at a gradual rate as
the refractory firebox is heated causing a corresponding increase in the
temperature of the combustion gases. At shutdown, NOX emissions
decrease rapidly as the gas temperature is quenched by incoming air.
Transient emissions characteristics of gas burners should be very
similar to those of oil burners. However, the HC and CO emissions that
occur after shutoff in gas burners are probably not as high as those from
oil burners, since gas leaks are minimal after burner shutoff.
The duration of the "on" period within a cycle of a coal-fired warm
air furnace does not significantly affect polycyclic organic matter (POM)
and particulate emissions (Reference 2-6). However, particulate and POM
loadings generated during the "off" transient are higher than those
produced during the "on" transient for coals with volatile matter contents
greater than 20 percent. This phenomenon is caused by incomplete
combustion of tars emitted from the volatile coal. Data trends from two
samples show that NOX emissions increase as the "on" time of a cycle is
increased.
Improper burner adjustment, dirty burner cups or nozzles, or
damaged components can significantly increase pollutant emissions.
Extensive field testing of oil burners has been reported (References 2-7
and 2-8). This testing shows that, with proper maintenance, smoke, CO,
HC, and NOX emissions can reduced by over 50 percent, while filterable
particulate can be reduced by almost 25 percent.
For gas burners, tuning, cleaning, and replacement of worn burner
components should not have as dramatic an effect. Gas burners provide
much cleaner combustion, and can be expected to stay tuned for extended
periods with few maintenance problems.
23
-------
2.1.4 Gas Turbines
Gas turbines are rotary internal combustion engines fueled mainly
by natural gas, diesel or distillate fuel oils, and occasionally by
residual or crude oils. These units range in capacity from 30 kW to 100
MW power output and may be installed in groups for larger power output.
Table 2-5 discusses the gas turbine categories and the effects of
operating conditions on multimedia pollutant levels.
The transient and nonstandard operations of gas turbines can be
separated into three groups: operational variations, startup/shutdown,
and equipment failures. Operational variations include changes in load,
speed, power, ambient conditions, and fuel quality.
Generally, gas turbines are designed to operate most efficiently at
their rated capacity. However, deviations from these rated conditions are
often necessary, and can cause the gas turbines to lose efficiency as well
as change emissions characteristics. The most frequently changed
operational variables are load and/or speed. Two studies (References 2-9
and 2-10) have indicated that, generally, CO, NOX and HC emissions vary
with change in power or load. NOX emissions also increase with
increased compressor inlet temperature, whereas CO and HC decrease.
Few data presently are available on emissions characteristics
during startup/shutdown or equipment failures. However, CO, HC, smoke and
particulate emissions should increase during these periods because of
incomplete combustion. Under these conditions, air-to-fuel ratios are not
stable and combustion temperatures are low. NOX emissions diminish,
therefore, because of the lower combustor temperatures.
2.1.5 Reciprocating 1C Engines
Reciprocating 1C engines for stationary applications range in
capacity from 750 W to 48 MW power output. These engines are either
compression ignition (CI) units fueled by diesel oil or a dual fuel
combination of natural gas and diesel oil, or spark ignition (SI) engines
fueled by natural gas or gasoline. They are used for applications ranging
from shaft power for large electrical generators and pipeline compressors
to small air compressors and welders. Table 2-6 summarizes the
characteristics of these unit designs.
Nonstandard operating conditions include load change, startup and
shutdown transients, and upsets such as fuel or electrical system
failure. Large 1C engines used for power generation or pipeline
compression applications are generally well maintained for economy.
Moreover, they are run steadily for many hours at their most efficient
operating condition. However, smaller engines are not maintained as well,
and frequently are operated in transient modes. Transients affect
emissions largely through their influence on air-to-fuel ratios. For
example, NOX emissions peak near the air-to-fuel stoichiometric ratio.
24
-------
TABLE 2-5. SUMMARY OF GAS TURBINE CHARACTERIZATION
Design Type
Utility and
Industrial
Simple and
Regenera-
tive Cycles
Combined
Cycles,
Repowering
Design
Characteristics
Rotary Internal com-
bustion engines.
Simple gas turbine
consists of compres-
sor, combustion
chamber, and tur-
bine. Fuel is burn-
ed before quenching.
Hot gases quenched
by secondary combus-
tion air, expanded
through a turbine
providing shaft
horsepower.
Regenerative cycles
use hot gases to
preheat inlet air.
Combined cycle 1s a
basic simple cycle
unit exhausting to
a waste heat boiler
to recover thermal
energy. Repowering
adds a combustion
turbine to an exist-
ing steam plant, In-
volving the mechani-
cal or thermal
Integration of the
combustion and
steam cycles.
Typical
Operational
Values
Utility Gas Turbine
Simple Cycle
Capacity: 92.3 MW
Specific fuel
consumption:
11.67 MJAWh
Compression ratio:
10:1
Exhaust flow:
345 kg/s
Exhaust temperature:
822K
Utility Gas Turbine
Combined Cycle
Capacity:
364.5 MW (4 turbines)
Specific fuel consump-
tion:
8.56 MJ/kWh
Compression ratio:
10:1
Exhaust flow:
256 kg/s (1 turbine)
Exhaust temperature:
811K
Fuel Consumption
-ซ)
45X gas
55X oil
Negligible
Effluent Streams
Flue gas
Flue gas
Operating
Modes
On-off transient,
load following,
idling at spin-
ning reserve.
On-off transient,
load following.
idling at spin-
ning reserve.
Effects of Transient
Nonstandard
Operation
NOx emissions general-
ly increase with in-
creasing power
Increased turbine com-
pressor Inlet tempera-
tures cause NOx to
increase. Behavior of
NOx is directly
related to rpm when
corrected to a con-
stant percent Oj.
NOx emissions general-
ly increase with In-
creasing power.
Increased turbine com-
pressor inlet tempera-
tures cause NOx to
Increase. Behavior of
NOx is directly
related to rpm when
corrected to a con-
stant percent 02-
Trends
Trend to higher
turbine inlet
temperatures,
larger capacity
and oil firing
in new units;
rapid growth
projected.
Use of combined
cycles should in-
crease because of
improved heat rate
and fuel flexi-
bility of unit.
Future
Importance
Primary
Secondary
T-851
no
en
-------
TABLE 2-6. SUMMARY OF RECIPROCATING 1C ENGINE CHARACTERIZATION
Design Type
Compression
Ignition.
Turbo-
Charged.
Naturally
Aspirated
Spark
Ignition.
Turbo-
Charged,
Naturally
Aspirated
llwer
Scavenged
Design
Characteristics
Air or tn air and
gas mixture Is cow-
press loo heated In
cylinders. -Diesel
fuel Is then In-
jected Into the hot
gas. causing spon-
taneous Ignition.
Combustion Is spark
Initiated. Natural
gas or gasoline Is
either Injected or
prefixed with the
combustion air in
a carbureted system.
Air charging by
means of a loป
pressure blower.
vhlch also helps
purge eihaust gases.
Fuel Consumption
(ซ)
671 gas
151 diesel
lit gasoline
71 dual (oil
and gas)
(all 1C engines)
671 gas
151 diesel
111 gasoline
71 dual (oil
and gas)
(all 1C engines)
671 gas
1SS diesel
lit gasoline
71 dual (oil
and gas)
(all 1C engines)
Effluent Streams
Exhaust gas
Exhaust gas
Exhaust gas
Operating
Modes
On-off transients.
Idling, upsets
On-off transients.
Idling, upsets
On-off transients,
Idling, upsets
Effects of Transient,
Nonstandard
Operation
NOX emissions peak
near stolchiometrlc
air-to-fuel ratio.'
NO, emissions
diminish with decreas-
ing load, greater
speed and timing re-
tard.
NO), (Missions peak
near stolchlometrlc
air-to-fuel ratio.
NOX emissions
diminish with decreas-
ing load, greater
speed and timing re-
tard.
NOX emissions peak
near stolchiometrlc
air-to-fuel ratio.
NOX emissions
diminish with decreas-
ing load, greater
speed and timing re-
tard.
Trends
1C engines finding
use for compressor
applications on
pipelines; low
growth rate of
diesel units; 1C
engines increas-
ingly being re-
placed by gas tur-
bines for standby
applications In
buildings, hos-
pitals, etc., be-
cause of space.
weight, noise.
vibration.
1C engines finding
use for compressor
applications on
pipelines; low
growth rate of
diesel units; 1C
engines Increas-
ingly being re-
placed by gas tur-
bines for standby
applications in
buildings, hos-
pitals, etc., be-
cause of space.
weight, noise.
vibration.
New large units
tending toward
turbocharging
Future
Importance
Pr Imary
Pr Imary
Secondary
T-8W
ro
-------
Other operational variations such as load, engine speed, and spark
timing also affect pollutant emissions. In general, NOX emissions
diminish with decreasing load, greater speed, and retarded timing.
Variations in ambient temperature also affect emissions of pollutants.
Recent experiments on automotive gasoline engines indicate that ambient
temperature reductions increase HC and CO. However, NOX levels are not
greatly affected by changes in ambient temperature.
Most stationary engines burn No. 2 diesel fuel or natural gas. The
properties of pipeline quality natural gas are essentially constant, but
field gas can vary in composition and sulfur content. These variations
affect the emissions of all gaseous pollutants as well as the engine
performance. For diesel oils, the most important properties are
viscosity, cetane number, distillation point, and sulfur and ash content.
In general, only the sulfur content varies significantly in commercial
grade fuels, and hence only S02 emissions are affected noticeably by
normal fuel variations.
2.1.6 Industrial Process Heating
Significant quantities of fuel are consumed by industrial process
heating equipment in industries such as iron and steel production, glass
manufacture, petroleum refining, and brick and ceramic manufacture. In
addition, there are dozens of industrial processes such as coffee
roasting, drum cleaning, paint curing ovens, and smelting of metal ores
that burn smaller amounts of fuel. Fuels fired in these units include
oil, natural gas, producer gas, refinery gas, and occasionally coal.
Table 2-7 summarizes the industrial process heating characterization. Few
data are available which quantify the effects of nonstandard operations on
industrial process heating. Further testing of this equipment is
necessary before nonstandard conditions can be understood.
2.2 STATIONARY SOURCE FUEL CONSUMPTION
This section characterizes fuel consumption for equipment and fuel
combinations described in Section 2.1. These data are important input for
both the emissions inventories discussed in Section 5 and the impact
rankings discussed in Section 8.
2.2.1 Baseline Fuel Consumption
Since fossil fuels account for almost all of the energy consumed by
stationary combustion sources nationally, the survey performed included
only these fuels. Fuel consumption data were compiled for 1974, since
this was the most recent year for which comprehensive and complete
regional data were available. For comparative purposes it was important
that both the national and regional fuel consumption data represented the
same year. Table 2-8 summarizes total annual consumption for coal,
petroleum, and gas. These totals do not reflect total energy consumed by
stationary sources, because some of the process industries and nonfossil
fuel uses have not been included.
27
-------
TABLE 2-7. SUMMARY OF INDUSTRIAL PROCESS HEATING CHARACTERIZATION
ro
oo
Process Type
Cement
Kilns
Glass
Melting
Furnaces
Annealing
Lehrs
Coke Oven
Under f1 re
Design
Characteristics
Kilns are rotary
cylindrical devices
up to 230 m In
length. Feedstock
moves through kiln
1n opposite direc-
tion from products
of combustion
Continuous reverba-
tory furnaces; end
port or side port.
Flame burns over
glass surface; com-
bustion gas exits
through opposite end
exhaust stack after
heating the combus-
tion air.
Used to control the
cooling of gas to
prevent stains.
Lehrs fired by at-
mospheric, premlx,
or excess air
burners.
Produce metallurgi-
cal coke from coal
from the distil-
lation of volatile
matter producing
coke oven gas; done
in long rows of slot
type ovens; fuel gas
supplies required
heat. Spent combus-
tion gas heats inlet
air.
Process Ranges
K1ln product temperature:
1756K
Furnace temperatures:
1528 to 1583K
Flue temperature:
1500K
Fuel Consumption
W
45X gas
40X coal
15* oil
Natural gas- and
o1l-f1red; coal
1s unsuitable due
to Impurities.
Natural gas- and
oil-fired; coal
unsuitable
Blast furnace gas
and coke oven gas
are primary fuels
Effluent Streams
Combustion pro-
ducts and en-
trained substan-
ces from feed-
stock
Combustion pro-
ducts and en-
trained substan-
ces from feed-
stock
Combustion pro-
ducts
Combustion pro-
ducts
Operating
Modes
Charging opera-
tions, upsets,
starting tran-
sients
Charging opera-
tions, upsets.
starting tran-
sients
Upsets, transients
Charging opera-
tions, upsets,
starting tran-
sients
Trends
Coal firing In new
units; energy Im-
provements due to
grate preheaters
and shorter, less
energy Intensive
kilns.
Trend toward use
of electric
melters, or elec-
trically assisted
conventional melt-
ers; use of oil
Instead of gas in
fossil fuel units.
Projected fuel
consumption about
5X annual.
Future
Importance
Primary
Pr Imary
Primary
Pr Imary
T3553
-------
TABLE 2-7. Continued
ro
vo
Process Type
Steel
Sintering
Machines
Open
Hearth
Furnaces
Brick and
Ceramic
Kilns
Design
Characteristics
Used to agglomerate
ore fines, flue
dust, and coke
breeze for charging
of a blast furnace.
These products
travel on a travel-
ing grate sintering
machine; after ig-
nition Is forced up
through the mixture
causing fusion and
agglomeration.
The charge 1s melted
in a shallow hearth
by heat from a flame
passing over the
charge and radiation
from the heated dome.
Spent combustion
gases preheat the
inlet combustion
gases.
Tunnel or periodic
kiln used most often.
Periodic: hot gases
drawn over bricks,
down through them by
underground flues,
and out of the oven
to the chimney.
Tunnel: cars carry-
ing bricks travel by
rail through kiln at
about one car per
hour.
Process Ranges
Kiln product
temperatures: 1367 K
Fuel Consumption
(ซ)
Low Btu gas
Low Btu gas such
as blast furnace
gas
Oil, gas, or coal
(coal use less
common)
Effluent Streams
Combustion pro-
ducts and en-
trained substan-
ces from feed-
stock
Combustion pro-
ducts and en-
trained substan-
ces from feed-
stock
Combustion pro-
ducts and en-
trained subtan-
ces from driers
and feedstocks.
Operating
Modes
Upsets, starting
transients
Charging, upset-
Ing, starting
transients.
Charging, upsets.
starting
transients
Trends
Operation declin-
ing because of
system incompata-
biHty; pelletiz-
ing replacing
sintering lines
Basic oxygen
furnace 1n new
units; fuel con-
sumption decreas-
ing by 8X per year
Tunnel kilns in
new units; con-
tinuous produc-
tion with heat
recovery
Future
Importance
Primary
Primary
Pr 1mary
T-853
-------
TABLE 2-7. Concluded
CO
o
Process Type
Catalytic
Cracking
Process
Heaters
Refinery
and Iron
Steel
Flares
Design
Characteristics
Preheated gas and
oil 1s charged to a
moving stream of hot
regenerated catalyst.
The gas and oil 1s
cracked 1n the re-
actor; products pass
through cylone for
separation and are
then cut Into pro-
ducts 1n fractlon-
ator.
Two basic types --
natural draft and
forced draft. Con-
structed as either
horizontal box or
vertical cyHnderlcal
Used for the control
of gaseous combusti-
ble emissions from
stationary sources
Process Ranges
Process temperature:
840 to 922K
Fuel consumption:
829 kj/i feedstock.
Fuel Consumption
(X)
011, gas, or
electricity
70X process gas
Waste gas
Effluent Streams
Combustion pro-
ducts and volati-
lized products or
catalysts
Combustion pro-
ducts
Combustion pro-
ducts
Operating
Modes
Starting tran-
sients, charging
Upsets, start
transients
Upsets, transients
Trends
Growth about 2*
annually
New units are
mechanical draft
with combustion
air preheater
Future
Importance
Primary
Primary
Primary
T-853
-------
TABLE 2-8. 1974 STATIONARY SOURCE FUEL CONSUMPTION (EJ)'
Equipment
Sector
Utility Boilers
Packaged Boilersc
Warm Air Furnaces
and Miscellaneous
Combustion
Gas Turbines
Reciprocating
1C Engines
Total
Coal
10.833
3.470
14.303
Oil
3.483
5.780
2.132
0.844
0.328d
12.567
Gas
4.906
6.323b
5.542
0.681
0.9146
18.366
Total
Fuel
19.222
15.573
7.674
1.525
1.242
45.236
aEJ/yr = 10*8 J/yr
^Includes process gas
cThis sector includes steam and hot water units
dlncludes gasoline and oil portion of dual fuel
^Includes natural gas portion of dual fuel
31
-------
U.S. energy use in 1974 totaled about 77 EO (72 x 1015 Btu)
(Reference 2-11), of which 94 percent was supplied by the fossil fuels
coal, petroleum, and natural gas. Approximately 57 percent of the total
energy was used by stationary sources. Fossil fuels furnished 92 percent
of the energy for these stationary sources; the remainder was supplied by
nuclear, hydroelectric, and other miscellaneous sources such as waste
fuels, wood, and geothermal. Of the total amount of fossil fuels burned
in stationary sources, coal contributed 26 percent, natural gas
44 percent, and petroleum 30 percent. Unlike petroleum, which is also a
major source of energy for transportation, coal and natural gas are used
primarily in stationary applications.
2.2.2 Projected Fuel Consumption
Energy projections were next used to estimate the consumption
trends and order-of-magnitude potential environmental problems from
stationary source combustion. Since energy supply and demand can vary
greatly, several projections for energy growth and equipment/fuel use were
selected. These scenarios were selected to cover the range of probable
developments in energy supply and consumption.
Five different energy scenarios were examined. The main factors
considered in defining each alternative were: (1) the effect of
government regulations and policies on the rate of growth in demand for
energy resources; (2) the equipment additions, by fuel type, required to
meet demand and source attrition; and (3) the effect of oil to coal, gas
to coal, and gas to oil conversions on fuel consumption. The five energy
alternatives are:
Reference -- low nuclear
Reference high nuclear
Conservation
Electrification
Synthetics
Figure 2-2 shows the mix of fuels and equipment types for each scenario.
These alternatives encompass a variety of contingencies in both total
energy demand and demand for specific fuels which lead to important
differences in the type and quantity of pollutants released.
In selecting energy alternatives, background information was
obtained from the Department of Energy (DOE) and other sources (References
2-11 through 2-27). The DOE projections were used to take advantage of
the technical expertise and the wide circulation of their results.
The reference case, high nuclear scenario assumes that current
consumption patterns continue with no major design or efficiency
improvements in the residential, commercial, or industrial sectors. The
scenario assumes no new legislative mandates for energy conservation.
32
-------
OJ
U)
30 ^
60 -
o.
E
O
O
40 -
I
Gas
Oil
Coal
A -- Utility Roilers
B -- Packaged Boilers
C -- Warm Air Furnaces
D -- Gas Turbines u
Reciprocating 1C Engines
1985 2000
Reference high nuclear
1005 2000 1085 2000
Reference low nuclear Conservation
1935 2000
Electrification
o
o
1985 2000
Synthetics
Figure 2-2. Energy scenarios.
-------
However, the dependence of energy demand on energy cost is considered.
Coal and nuclear powerplants continue to expand to meet electricity
demand. Nuclear powerplants are projected to meet 65 percent of the
demand for new power generation by the year 2000. Other energy sources
such as geothermal, hydroelectric, and urban waste are projected to grow
as required to meet energy demand, without pushing the development of the
technology.
The reference case, low nuclear scenario also assumes that current
consumption patterns continue. Coal and nuclear powerplants continue to
meet new electricity capacity demand. However, this scenario assumes a
lower use of nuclear power and a higher use of coal. Nuclear power
accounts for 35 percent of new generating capacity through the year 2000,
whereas coal accounts for 65 percent. This scenario would occur if there
was increased pressure to use our coal resources to meet future energy
demand, and if the use of nuclear powerplants continues to be low because
of concerns about safety, waste disposal, safeguard costs, or uranium
costs.
The conservation scenario was developed to examine energy
conservation efforts such as improving energy conversion efficiency and
increasing the use of energy resources presently available. This means
increasing the recovery of gas and oil (secondary, tertiary recovery) and
using waste materials from recycling and energy conversion. Thus, energy
demand is effectively reduced, but the major sources of energy remain
essentially the same.
The electrification scenario maximizes potential end uses of
electricity and uses as much electric generating capacity as possible. In
addition, existing oil- and gas-fired equipment is converted to coal where
possible.
The synthetics scenario considers the effects of increased supply
of synthetic liquids and gaseous fuels. It evaluates the impact of
drawing on vast resources of coal and oil shale to produce liquid and
gaseous fuels as direct substitutes for petroleum fuels. The total energy
use projected is quite close to the reference scenario, although much less
oil and natural gas are consumed. This scenario also assumes that growth
in electric generating capacity is largely met by light water reactors, so
that new coal production can be used for synthetics.
Energy projections by specific equipment/fuel type were generated
for 1985 and 2000 for the five scenarios. The resulting projections were
carried through the emission projections and the pollution impact
evaluation. Summaries of energy consumption in the reference scenarios
are given in Table 2-9. Results for the other scenarios are documented in
Reference 2-28.
2.3 TRENDS IN EQUIPMENT/FUEL USE
From total energy consumption for 2000, energy consumption for
specific sectors and equipment/fuel combinations were determined by
34
-------
TABLE 2-9a. STATIONARY SOURCE FUEL CONSUMPTION FOR THE YEAR 2000:
REFERENCE CASE -- LOW NUCLEAR (EJ)
Equipment
Sector
Utility Boilers
Packaged Boilers
Warm Air Furnaces
and Miscellaneous
Combustion
Gas Turbines
Reciprocating
1C Engines
Total
Coal
24.398
2.763
-~
27.161
Oil
4.339
8.302
2.800
1.752
0.472C
18.165
Gas
--
6.949a
6.634
1.390
0.240d
15.213
Total
Fuel
28.737
18.514
9.434
3.142
0.712
60.539
alncludes process gas
This sector includes steam and hot water units
clncludes gasoline and oil portion of dual fuel
Includes natural gas portion of dual fuel
TABLE 2-9b. STATIONARY SOURCE FUEL CONSUMPTION FOR THE YEAR 2000:
REFERENCE CASE -- HIGH NUCLEAR (EJ)
Equipment
Sector
Utility Boilers
Packaged Boilers
Warm Air Furnaces
and Miscellaneous
Combustion
Gsi Turbines
Reciprocating
1C Engines
Total
Coal
42.697
4.835
~~
--
47.532
Oil
4.339
8.802
2.800
1.752
0.472C
18.165
Gas
--
6.949a
6.634
1.390
0.240d
15.213
Total
Fuel
47.036
20.586
9.434
3.142
0.712
80.910
Includes process gas
This sector includes steam and hot water units
Includes gasoline and oil portion of dual fuel
Includes natural gas portion of dual fuel
35
-------
evaluating trends in equipment sales and fuels usage. This section
discusses these trends for each stationary source sector.
2.3.1 Utility Boilers
Tangential, single and horizontally opposed firing, and
Turbofurnaces are the most common utility boiler designs. As shown in
Figure 2-3 these primary designs will continue to be used extensively
through the 1980's. Several recent design changes are being used on
tangential and wall fired boilers. New units use reduced heat release
rates to suppress slagging and tube wastage and modified combustion
conditions to lower NOX emissions.
Secondary designs cyclone, vertical, and stoker firing -- will
not be important equipment types because of inherent design limitations.
Stokers are limited to about 40 MW electrical output and have high initial
costs. Vertical furnaces were developed primarily for firing anthracite
coal, which is no longer used as a utility fuel. Cyclone furnaces have
not proved adaptable to emission reduction regulations. This slagging
furnace must operate at high combustion temperatures, which cause high
thermal NOX. It is a desirable choice, however, for high sodium lignite
applications.
Pressurized Fluidized Bed Combustion (PFBC) units are being
designed for use in combined gas turbine/steam cycles in which the PFBC
acts as both the external combustor for the gas turbine and a steam
generator for the steam turbine. However, PFBC's will not be commercially
developed until the late 1980's and are expected to have an insignificant
impact on national fuel usage through 2000.
2.3.2 Packaged Boilers
Industrial boiler manufacturers are stressing design flexibility to
adapt to changing fuels availability and impending emission limits.
Additional emphasis is placed on combustion controls, boiler safeguards
and heat recovery equipment for more efficient and reliable operation.
The trend toward fuel flexibility includes provisions for coal
firing in most large industrial units. In general, pulverized coal firing
is more efficient than stoker firing. In addition, combustion controls
may be used effectively on pulverized units while maintaining high
operating efficiency. However, as shown in Figure 2-4, stoker demand will
grow rapidly because the cost of small pulverized coal units (<60 MW
input) is higher for this industrial size category. In addition, stoker
fired boilers are able to burn a wide variety of solid byproduct wastes,
as well as residual refuse, with minimal preparation. Many solid wastes
burned in a pulverized unit would require careful pretreatment, as well as
an auxiliary burnout grate in the furnace.
For small applications (<30 MW), alternative steam sources such as
firetube, electric, and heat recovery boilers are finding increasing use.
Firebox firetube boilers can be designed to fire oil or gas now and coal
at some future date by leaving room to add an underfeed stoker.
36
-------
NOTE: Only coal consumption
is shown by equipment
types
c
o
3
to
c.
o
o
en
d)
c
Vertical and Stoker Boilers
5 _
1978
1985
2000
Time
Figure 2-3. National energy consumption and equipment trends
for utility boilers.
37
-------
NOTE: Only coal consumption
is shown by equipment
types
10
c
o
O.
3
(O
O
o
QJ
Pulverized Coal Firing
1978
1985
2000
Time
Figure 2-4. National energy consumption and equipment
trends for packaged boilers.
38
-------
Electric boilers have become economically attractive in some areas
during the past few years because of environmental pressures and the
increasing costs of petroleum fuels. Resistance type boilers are
typically limited to about 3 MW whereas electrode boilers are cheaper and
more practical at higher ratings.
2.3.3 Residential Heating Units
Trends in residential heating units are primarily toward optimized
burners that reduce emissions and increase fuel efficiency. Units are
presently being designed which use surface combustion of premixed fuel and
air on the furnace refractory. Combustion occurs without a visible flame,
and heat is transferred from the surface to an air cooled firebox wall by
radiation. The surface combustion concept allows operation at low excess
air, which improves the furnace efficiency. This unit design should be
commercialized in the early 1980's.
An advanced distillate oil burner has also been developed by EPA
which reduces NOX emissions and increases steady state furnace
efficiency by up to 10 percent (Reference 2-29). Field demonstrations of
a prototype burner installed in its integrated furnace have indicated
NOX reductions of 65 percent compared to conventional residential
burner/furnace systems. Furnace efficiencies of 83 to 84 percent were
achieved.
Another approach uses a thermal aerosol burner to fire No. 1 and
No. 2 fuel oils. This burner operates by heating the fuel and then
flashing it in the burner nozzle to produce a mixture of vapor and fine
droplets, and is commercially offered as part of the Blue Ray furnace
system. The manufacturer claims that clean, efficient combustion can be
achieved at low firing rates with excess air as low as 5 to 10 percent
resulting in a furnace efficiency of 83 percent. However practical, safe
home use may necessitate a much higher excess air level of 20 percent,
thus the high furnace efficiency may not be realized.
Figure 2-5 shows fuel use trends for residential heating use. As
the figure shows, natural gas is currently the major fuel. Distillate
oil, however, is increasing its share of the market and should be the
dominant fuel in the future. In addition, there is a continuing trend to
electricity for space heating applications.
2.3.4 Gas Turbines
The growth in the use of gas turbines has been rapid since the
mid-1960's because of their low initial cost, ease of maintenance, high
power-to-weight ratio, reliability, and short delivery time. Gas turbines
are now being built in larger capacities with improved heat rates.
Moreover, combined cycle turbines are becoming the preferred future design
for intermediate and baseline applications because of their improved heat
rates and fuel flexibility. Present combined cycle plants are
economically feasible only for intermediate range systems. Increasing the
inlet temperatures to 2000K would improve unit efficiency to about 50
percent. However, these units will not be commercially available until
39
-------
7
6
!
O-
E
I
o
O)
1978
1985
2000
Time
Figure 2-5. National energy consumption and equipment trends for
residential and miscellaneous combustion sources.
40
-------
the mid 1980's and will have negligible fuel use impact nationally through
2000. Users predict that gas turbines will continue to supply about 10
percent of new generating capacity through at least 1985.
The growth in gas turbine use, however, is highly dependent on
their potential for burning coal derived fuels while maintaining a high
heat rate and competitive initial cost. Since many different
liquefaction, gasification, and other fuel cleanup processes are being
developed, future turbines must be able to burn a broad spectrum of fuels,
with a wide range of contaminant levels. Further, the energy loss in coal
conversion creates strong incentives to design more efficient gas turbines
and to use them in combined cycle systems for base and intermediate load
service. Pressurized fluidized bed combustor (PFBC) development for coal
and waste fuels firing is also proceeding. Combustors operating at high
pressure offer high efficiency in a combined cycle application. However,
substantial efforts are required before PFBC gas turbine combinations can
be commercialized. Thus, pressurized fluidized bed combustion will
probably not be commercialized until the late 1980's.
2.3.5 Reciprocating 1C Engines
Reciprocating internal combustion (1C) engines are available in a
wide range of sizes and configurations to serve an extremely varied set of
applications. The use of large 1C engines in baseload electric
generation, oil and gas production and transportation, and other such uses
should continue to remain strong. However, medium power engines face
competition from substitute power sources in nearly all applications,
particularly electricity generation. Direct purchase of electricity and
use of electric motors require less maintenance and lower initial and
operating costs for small general industrial and agricultural
applications. Thus, markets for medium power engines are declining except
where electricity is inaccessible or impractical. Low power engines are
also largely being replaced by electric motors.
Modified designs are being developed for reciprocating 1C engines
to increase efficiency and reduce emissions. Combustion chamber
modifications, and especially improvements to the fuel/air mixing process
in the cylinder or the use of a precombustion chamber, are the most
promising options. Some manufacturers, however, may elect to use exhaust
gas recirculation or water induction. An improved design for large
engines should be demonstrated by 1982. No other technology based
developments are expected in the foreseeable future for this source
category.
2.3.6 Process Furnaces
This source category is diverse, and trends in each segment are
unique to that industry. Therefore, the following discussion is organized
by industry.
Combustion sources in the iron and steel industry include sintering
lines, open hearth furnaces, soaking pits, reheat furnaces, and coke
ovens. Use of sintering lines is declining at the rate of about 3.4
41
-------
percent annually because they cannot accommodate rolling mill scale
contaminated with rolling oil. Open hearth furnaces are also diminishing
in importance, as old units are being replaced by basic oxygen furnaces.
The need for soaking pits and reheat furnaces is diminishing, too, because
continuous casting of molten metal is becoming the preferred method for
making iron and steel.
Overall, the growth of process fuel consumption in the iron and
steel industry is about 2.8 percent annually. This includes a projected
5.7 percent annual increase in fuel consumption for coke ovens.
The current trend in the glass industry is towards electric
melters. In addition, fuel oil is increasingly being used in place of
natural gas because of natural gas shortages and price increases. Coal,
for the most part, is an unacceptable fuel for the glass industry because
of its impurities. However, coal gasification may become a useful and
economically viable fuel source for the glass industry in the late 1980's.
It is expected that many cement industries will convert to coal
firing in the near future. According to current DOE statistics, 90
percent of all cement plants should be able to use coal by 1980, compared
to 76 percent today. The cement industry has reduced energy consumption
by using grate preheaters and quicker, less energy intensive kilns. One
further improvement may be to replace traditional rotary kilns with
fluidized bed kilns.
Cement industry figures show that the industry has grown at an
average rate of about 1.9 percent annually over the past 20 years.
Industry projections, however, predict a greater growth in the next few
years of between 2.6 to 4.1 percent per year.
Current trends in the petroleum refining industry are toward
mechanical draft process heaters with a combustion air preheater,
primarily because they conserve more energy than natural draft heaters.
Process heaters are fueled primarily (60 to 80 percent) by process gas, a
byproduct of the refinery process. The auxiliary fuel is generally oil.
However, oil consumption should decline as more process gas with a lower
sulfur content is used. Thus, oil consumption should decline by as much
as 28 percent. A 2.7 percent annual increase in process heating is
projected for 1980, and a 2.9 percent annual increase for 1985.
2.4 AVAILABILITY OF ALTERNATE CLEAN FUELS FOR USE IN AREA SOURCES
To better understand the environmental and economic aspects and the
near term potential for synthetic clean fuels use in area sources, a
separate study on this subject was included as a support task of the NOX
EA. The overall objectives of this study were to identify the scope and
timing of current R&D projects aimed at commercializing alternate/
synthetic fuels, to assess the potential for use of these fuels in area
sources, to evaluate potential impacts on combustion generated air
pollutant emissions deriving from their use, and thereby to anticipate
control development needs.
42
-------
The potential alternate fuels studied included low Btu, medium Btu,
high Btu (synthetic natural gas), and hydrogen gases; methanol and coal or
shale derived liquids; and solvent refined coal (SRC). The initial study
was conducted in the first year of the program and documented in
Reference 2-29. The conclusions and recommendations of this study were:
Significant commercialization of any synthetic clean fuel
process whose product would be used extensively for area
sources will not be realized within the next 10 to 15 years
Clean fuels use in area sources should be given only minor
emphasis in subsequent NOX EA efforts since pollutant control
development for candidate alternate fuels could be accomplished
in less than the available 10 to 15 year lead time
t An annual update of the clean fuels study should be made to
reevaluate the conclusions in view of the rapid state of flux
of synthetic fuels technology and the implications of a
national energy policy
The update to the original study discussed below is the result of a
continuing survey of alternate fuels developments that occurred during the
past year. In general, the conclusions of the original study have been
further substantiated.
2.4.1 Alternate Liquid Fuels -- Coal Liquids
The primary product of the most promising coal liquefaction
processes -- SRC II, H-Coal, and Exxon Donor Solvent -- is a synthetic
crude or a heavy oil similar to residual oil. Utility and large
industrial boilers are generally considered to be the prime candidate
users of these fuels. This equipment sector currently uses a large
portion of the petroleum based fuels that would be replaced by synthetic
liquids. Area sources are typically not designed for heavy oil firing
and, as such will probably not be capable of firing these coal derived
liquids.
However, even though area sources cannot fire these synthetic
liquids, widespread commercialization of these processes could indirectly
affect the area source fuel distribution. The increased use of synthetic
liquid fuels by large point sources will in turn increase the availability
of petroleum based distillate oils to area sources. For the purposes of
this study, combustion control evaluation activities for area sources
fueled by petroleum based fuels are already underway.
Of the three most promising liquefaction technologies, SRC II will
probably be first to reach the commercialization stage. A feasibility
study is now underway, funded by the Department of Energy to build a
230 Mg/h (6000 tpd) demonstration plant. If the decision is made to
build, this plant would be operating by 1985. If the demonstration is
successful, the process module would be duplicated four times resulting in
the first commercial scale plant with a 1.1 Gg/h (30,000 tpd) capacity.
Completion of this plant would be in the 1995 to 2000 timeframe.
43
-------
The H-Coal process is presently undergoing scaleup from a 110 Kg/h
(3 tpd) process development unit to a 9.5 Mg/h (250 tpd) pilot plant
expected to go online in late 1978. Completion of pilot plant evaluation
is scheduled for late 1980 so a demonstration scale process is not
expected prior to at least 1985. Consequently significant
commercialization of H-Coal liquids cannot be expected before 1995 to 2000.
The Exxon Donor Solvent (EDS) process is in the design and
procurement stage of a 9.5 Mg/h (250 tpd) pilot plant scheduled for
startup in early 1980. Pilot plant evaluation is scheduled to continue
through 1982. A demonstration plant would probably be operating prior to
the 1985 to 1990 timeframe. Commercialization of the EDS process is
therefore not expected to occur before about 2000.
Based on the above, the potential impact of coal derived liquids to
supplant a substantial portion of petroleum derived fuels prior to the
year 2000 is negligible. A recent study (Reference 2-30) estimates that
by 1990, only 1.3 to 3.3 Mฃ/h (200,000 to 500,000 bbl/day) of coal
derived liquids could be available from demonstration plants. This
quantity represents less than 10 percent of the projected 1985 and 2000
distillate oil consumptions of residential space heaters
(Reference 2-28). Accordingly, emission control research and development
directed at area sources firing synthetic coal liquids is not recommended.
2.4.2 Alternate Liquid Fuels Methanol
An extensive methanol fuels program is presently under
consideration by DOE. If the program is pursued, several use
demonstrations will be undertaken on highway vehicles, peaking gas
turbines, fuel cells, and utility boilers. Methanol is a particularly
good substitute for gasoline since it is almost directly interchangeable
in the internal combustion engine. As a result, high priority has been
placed on developing alcohol as a fuel for highway vehicles. Still, even
in stationary source applications, methanol and other alcohol fuels are
preferred for the larger combustion equipment and as such will probably
not see extensive use in area source equipment prior to 2000.
2.4.3 Alternate Gaseous Fuels -- Low Btu Gas
As stated in the original study (Reference 2-29), low Btu gas is
currently being pursued as a fuel source for large utility and industrial
combined cycle power plants and for utility boilers. The economics of
distribution for low Btu gas are such that only short range distribution
is feasible. As a result, low Btu gas use in area sources will probably
occur only in close proximity to either utility or industrial
installations.
Several low Btu gas generation installations are currently underway
or planned. Among these are six projects sponsored by DOE
(Reference 2-31) that will use the product gas for a variety of industrial
and/or commercial applications. In one project the fuel gas will be used
to produce hot water and steam for the heating and cooling needs of an
entire planned community of housing and associated industry. These DOE
44
-------
demonstration projects are scheduled for startup and testing during the
1979 to 1980 timeframe.
A number of other privately sponsored ventures are also underway.
These consist of either user demonstrations or equipment supplier pilot
plants. All of these ventures, however, are directed at the large
industrial user rather than the smaller commercial or residential
combustion equipment.
In summary, it appears that the impact of low Btu gas on the fuel
use patterns of area sources will be negligible. The only area source
equipment that might be fueled by low Btu producer gas would be at the
site or in the vicinity of the gasifier. Research and development work
for emission control of low Btu gas combustion specifically for area
sources is therefore not recommended.
2.4.4 Alternate Gaseous Fuels -- Medium Btu Gas
The obstacles confronting the widespread use of medium Btu gas in
area sources are essentially the same as those that confront low Btu gas
use in these sources. For medium Btu gases, the economics of distribution
are less discouraging but remain sufficently questionable such that under
current price and supply projections for petroleum fuels and natural gas,
widespread area source use of medium Btu gas is not anticipated.
The marginal economics of distribution for medium Btu gas require
that numerous district plants be built rather than a single central
plant. Two distinct disadvantages exist with this concept. Oxygen is
generally required for the gasification process thereby limiting plant
location to one where oxygen is available. Furthermore, the fl.ammability
limits of medium Btu gas require that extensive safety precautions be
exercised both in preparation and in even the shortest distribution.
At this time then, no emission control development directed at area
sources firing medium Btu gas is recommended.
2.4.5 Alternate Gaseous Fuels -- High Btu Gas
High Btu gas, sometimes referred to as pipeline quality gas or
synthetic natural gas (SNG) is essentially identical in composition to
natural gas. Thus as stated in the original report (Reference 2-29) high
Btu gas can use the same distribution and combustion systems now being
used for natural gas. Control techniques for natural gas should therefore
be directly applicable to this synthetic fuel.
2.4.6 Summary
The use of alternate clean fuels in area sources will be very
limited prior to the 1990 to 2000 timeframe. Even though synthetic clean
fuels will be available in increasing quantities after about 1985 to 1990,
the majority of these fuels will be used to fire larger point sources.
This displacement of conventional fuels in the utility and large
industrial systems will provide continued availability of petroleum based
45
-------
liquids and natural gas to area source combustion equipment. Consequently
technology development efforts directed at emissions control techniques
for clean fuels in area sources are not appropriate at this time.
46
-------
REFERENCES FOR SECTION 2
2-1. Mason, H. B., et aj_., "Preliminary Assessment of Combustion
Modification Techniques: Volume II - Technical Results,"
EPA-600/7-77-119b, NTIS PB-276 681/AS, October 1977.
2-2. Waterland, L. R., el; aj_., "Environmental Assessment of Stationary
Source NOX Control Technologies First Annual Report,"
EPA-600/7-78-046, NTIS PB-279 083/AS, March 1978.
2-3. Cato, 6. A., erb aj_., "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial
Boilers -- Phase I," EPA-650/2-74-078a, NTIS PB-238 920/AS, October
1974.
2-4. Offen, 6. R., et _ง_[., "Control of Particulate Matter from Oil
Burners and Boilers," EPA-450/3-76-005, NTIS PB-258 495/1BE, April
1976.
2-5. Hall, R. E., ^t a/L, "A Study of Air Pollutant Emissions from
Residential Heating Systems," EPA-650/2-74-003, NTIS PB-229 697/AS,
January 1974.
2-6. Giammar, R. D., et a]_., "Emissions from Residential and Small
Commercial Stoker-Coal-Fired Boilers Under Smokeless Operation,"
EPA-660/7-76-029, NTIS PB-263 891/4BE, October 1976.
2-7. Barrett, R. E., et _al_., "Field Investigation of Emissions from
Combustion Equipment for Space Heating," EPA-R2-73-084a,
NTIS PB-263 891/4BE, June 1973.
2-8. Copeland, J. E., et a/L, "Soiling Characteristics and Performance
of Domestic and Commercial Oil-Burning Units," APTIC Report 76132,
January 1968.
2-9. Roessler, W., et aj_., "Assessment of the Applicability of
Automotive Emission Control Technology to Stationary Engines,"
EPA-650/2-74-051, NTIS PB-237 115/AS, July 1974.
2-10. "Standards Support and Environmental Impact Statement, Volume I:
Proposed Standards of Performance of Stationary Gas Turbines,"
EPA-450/2-77-017a, NTIS PB-272 422/7BE, September 1977.
2-11. Dupree, W. G., and J. S. Corsentino, "Energy Through the Year 2000
(Revised)," U.S. Bureau of Mines, December 1975.
2-12. "1976 National Energy Outlook," Federal Energy Administration,
FEA/N/75/713, February 1976.
47
-------
2-13. "A National Plan for Energy Research, Development & Demonstration:
Creating Energy Choices for the Future," ERDA-48, Volume 2 of 2,
1976.
2-14. "The National Energy Plan," Executive Office of the President,
Energy Policy and Planning, 1977.
2-15. "Energy Perspectives 2," U.S. Department of the Interior, 1976.
2-16. "Energy Statistics," U.S. Senate, Finance Committee, 94:1,
July 1975.
2-17. Chapman, L. D., et al_., "Electricity Demand: Project Independence
and the Clean Air Act," Oak Ridge National Laboratory,
ORNL-NSF-EP89, November 1975.
2-18. "Proceedings of the Workshop on Analysis of 1974 and 1975 Power
Growth," Electric Power Research Institute, EPRI EA-318-SR,
December 1976.
2-19. "The National Power Survey Task Force Report: Energy Conversion
Research," Federal Power Commission, June 1974.
2-20. Benedict, M., "U.S. Energy: The Plan That Can Work," Technology
Review, May 1976.
2-21. "Resources for the Future Annual Report for the Year Ending
September 30, 1976," Resources for the Future, March 1977.
2-22. "An Integrated Technology Assessment of Electric Utility Energy
Systems, First Year Report (Draft), Volume 1: The Assessment,"
Teknekron, Inc., Berkeley, CA, 1977
2-23. Bomke, E. H., "A Forecast of Power Developments, 1975-2000," ASME
75-PWR-5, June 1975.
2-24. "The Potential for Energy Conservation ซ Substitution for Scarce
Fuels, A Staff Study," Executive Office of the President, Office of
Emergency Preparedness, January 1973.
2-25. Wright, R. R., "The Outlook for Petroleum Power Plant Fuels," ASME
76-1PC-PWR-6, April 1976.
2-26. "Status: Significant U.S. Power Plants in Planning or
Construction," Presidential Task Force on Power Plant Acceleration,
Federal Energy Administration, July 1976.
2-27. Gordon, R. L., "Historical Trends in Coal Utilization and Supply,"
U.S. Bureau of Mines, OFR 121-76.
2-28. Salvesen, K. G., et _ง]_., "Emission Characterization of Stationary
NOX Sources Volume 2, Appendices," EPA-600/7-78-120b, June
1978.
48
-------
2-29 Okuda, A. S., and L. P. Combs, "Field Verification of Low-Emission
Integrated Residential Furnaces," in Proceedings of the Third
Stationary Source Combustion Symposium: Volume 1,
EPA-600/7-79-050a, February 1979.
2-30. Shimizu A. B., "Identification and Characterization of Clean Fuels
for Area Sources," Acurex Final Report TR-77-57, Acurex
Corporation, Mountain View, CA, April 1978.
2-31. "O'Leary's Plan for Synthetic Fuel Advances," The Energy Daily,
March 13, 1978.
2-32. "Fossil Energy Research Program for the ERDA FY 78," ERDA 77-33,
April 1977.
49
-------
SECTION 3
CURRENT ENVIRONMENTAL BACKGROUND
As noted in Section 1, one of the major goals of the NOX EA
program is to identify the most cost-effective, environmentally sound
NOX control techniques to attain and maintain ambient air quality
standards to the year 2000, and, if adequate controls are unavailable, to
recommend R&D priorities to develop needed technologies. A key aspect of
satisfying this goal is to identify and incorporate into the analysis the
potential effects of evolving regulatory strategies which could impact the
need for NOX controls.
Thus, the primary purposes of this section are to summarize the
current regulatory activities that will, or could potentially, affect the
need for NOX controls, and to discuss the impacts of these activities.
This will provide the necessary perspective for the evaluation of NOX
control needs to be performed in the systems analysis activities of the
NOX EA.
The passage of the Clean Air Act Amendments of 1977 likely will
result in major changes in the strategy for controlling NOX emissions
from stationary sources. The four most significant changes required by
the act are the following:
Requiring EPA to determine whether a short term standard for
N02 ambient concentrations is necessary
Requiring EPA to include N02 within the Prevention of
Significant Deterioration provisions
Requiring EPA to promulgate, within five years, NSPS for all
major stationary sources, and fixed removal percentages for
emissions from fossil fuel fired combustion facilities
Establishing a 0.62 g/km NOX emission limit as the standard
for light duty vehicles, and relegating the 0.25 g/km NOX
emission limit to a research goal
The first three changes place increased regulatory emphasis on NOX
control. The fourth change essentially shifts a greater portion of the
burden of achieving any standard from mobile to stationary sources. The
50
-------
single most important change, however, is the requirement for EPA to make
a determination as to the need for a short term standard.
Section 3.1 below describes the current N0ฃ ambient standard, key
NOX emission regulations, and the status of standard attainment
throughout the country. Section 3.2 discusses the status of the N02
short term standard being developed, and its implications on AQCR
attainment/nonattainment throughout the country. Section 3.3 briefly
describes other regulatory provisions within the Clean Air Act that can
exert significant emphasis on stationary source NOX control via
mechanisms to ensure attainment or maintenance of ambient air quality.
Other issues, such as HC control, acid rain, and increased use of coal,
which relate to the need for NOX control are briefly discussed in
Section 3.4. In Section 3.5 the status and weaknesses of the present
N02 monitoring system in the U.S., primarily with regard to a short term
N02 standard are discussed.
3.1 THE ANNUAL AVERAGE N0ฃ STANDARD
In April 1971, EPA established N02 as one of the six criteria
pollutants to be regulated under the new Clean Air Act. A primary ambient
standard was set at 100 ug/m3, annual average, to provide protection
with an adequate margin of safety. The actual health effects for an
annual period were determined to occur at 150 yg/m3.
The existing NOX emission regulations designed to aid in
attaining and maintaining the ambient N02 annual standard comprise the
following: performance standards for mobile sources, New Source
Performance Standards (NSPS) for boilers with a firing rate in excess of
73 MW (250 x 106 Btu/hr) and nitric acid plants, and State
Implementation Plan (SIP) provisions covering existing stationary
sources. The promulgated standards are summarized in Table 3-1. In
addition to these, NSPS for NOX emissions from gas turbines and 1C
engines and more stringent standards for large steam generators have been
proposed or are in preparation. With the exception of motor vehicle
standards, few of these regulations present significant compliance
problems. Most of the regulations covering stationary combustion sources
can be met with modification of combustion practices. For automobiles,
the 1.2 g/km standard is being met through use of operational changes and
exhaust gas recirculation (E6R), with a conventional oxidation catalyst to
counteract resultant increases in HC and CO emissions. The 0.62 g/km
standard is to be met through use of a three way catalyst and EGR.
Of the 247 air quality control regions (AQCR's) designated in this
country, four are presently declared as nonattainment regions with respect
to the 100 yg/m3 annual average standard for NO?: (1) Metropolitan
Chicago, (067); (2) Metropolitan Denver (036); (3) San Diego (029); and
(4) Metropolitan Los Angeles (024) (Reference 3-1). However, the actual
number of AQCR's exceeding the standard could be greater, since many
AQCR's do not have sufficient data on N02 ambient levels to determine a
valid annual average (see Section 3.5). Table 3-2 shows 16 AQCR's which
have been informally considered as potential candidates for designation as
NOX Air Quality Maintenance Areas (Reference 3-2). Moreover, a recent
51
-------
TABLE 3-1. SUMMARY OF CURRENT NSPS & MOBILE EMISSION STANDARDS FOR NOX
Allowed Emission Levels
Motor Vehicles
Automobiles, 1978
Automobiles, 1981
Automobiles, (research goal)
Stationary Sources
Fossil Fuel Fired Steam Generators
(>73 MW, 250 x 106 Btu/hr)
Coal-fired (except lignite)
Oil-fired
Gas-fired
Nitric Acid Plants
g/km
1.2
0.62
0.25
301
129
86
(g/mi)
2.0)
1.0)
(0.4)
ng/J (Ib/I06 Btu)
(0.7)
(0.3)
(0.2)
1.5 g/kg (3 Ib/ton)
TABLE 3-2. AQCR'S RECOGNIZED AS POTENTIAL N02 PROBLEM AREASa
AQCR
Phoenix
Los Angeles
San Diego
San Francisco
Denver
Springfield, MA
New York City
Philadelphia
Atlanta
Chicago
Baltimore
Boston
Detroit
Canton, OH
Salt Lake City
Richmond, VA
AQCR Number
15
24
29
30
36
42
43
45
56
67
115
119
123
174
220
225
Reference 3-2
52
-------
study concluded that a greater than 50 percent chance exists that 34
AQCR's would be judged in nonattainment if sufficient data were available
(Reference 3-3).
Predictions of the number of AQCR's likely to be in violation of
the annual average standard in the future are hindered not only by the
lack of current data but also by inconsistent trends in monitored N02
concentrations for sites with several years of data (Reference 3-4).
However, a generalized assessment of the change in the number of violators
is possible by considering the contribution of point and area sources to
annual N0ฃ levels.
Recent work suggests that the major contributors to annual average
N02 violations are area sources (both mobile and dispersed stationary
sources, such as fossil fuel fired residential heating). Unlike point
sources, area sources emit pollutants near the ground, allowing little
effect from weather variations, and usually are concentrated within a
given region. The effect of area sources on local N02 concentrations is
fairly constant and, when a number of area sources are located close
together, high annual average concentrations can result. On the other
hand, point sources tend to emit NOX in concentrated form at higher
altitudes and point source emissions are thus susceptible to a great deal
of weather variations. Moreover, point sources often are more diffusely
sited than area sources.
Thus, based on the reasonable assumption that annual average N02
concentrations are due primarily to area sources, changes in the number of
violators can be approximated from changes in area source emission
patterns. In fact, area source emissions should be less than the 1975
values in the near term for several reasons:
Automobiles will be meeting more stringent emissions
requirements*
Fossil fuel fired residential heating units may be replaced by
electric heating**
t Growth in area sources will probably not occur in already high
emission areas***
*Based on reasonable assumptions for vehicle populations and use, control
deterioration factors, and source growth (~3 percent per year) the ratio
of future mobile source emissions to 1975 emissions is projected to be
0.80, 0.82, 1.1 in 1985, 1990 and 2000, respectively.
**This will increase point source emissions which will, in turn,
increase the point source contribution to the annual average.
***Based on space limitations for stationary sources and traffic density
limitations for mobile sources in high emission areas.
53
-------
If these trends occur, total reductions in area source NOX emissions
should occur through 1985, even though the number of stationary sources
may increase. Such conditions may improve or at least stabilize annual
average N0ฃ levels through 1985. However, beyond 1985, the growth in
area sources (and point sources) may more than compensate for decreased
mobile and stationary source emission factors, and aggravated N02 annual
average levels may result. Without additional controls, it is estimated
that a 25 percent increase in annual average N02 levels by the year 2000
will occur (see Section 8.2.1). Nevertheless, such predictions, although.
reasonable, remain conjectural considering the length of time over which
emissions projections must be made and the absence of more specific
modeling data on future ambient N02 annual levels.
3.2 SHORT TERM N02 STANDARDS
The existing ambient air quality standard for N02 was promulgated
in April 1971. The primary basis for this standard was epidemiological
evidence from 1968 to 1969 studies of school children and family groups
residing downwind from an explosives plant in Chattanooga, Tennessee.
These studies linked respiratory infection to annual N02 exposures of
about 150 yg/m3 (0.08 ppm) and higher. Based on later data, however,
and a better understanding of the role of elevated short term exposures in
the original Chattanooga study, it became evident that the annual standard
of 100 yg/m3 may not sufficiently protect public health.
Studies performed by the World Health Organization (WHO) and in
Japan had demonstrated that harmful effects can result from short term
exposures to N02- These studies in turn have led to the Japanese
Government's adoption of a one hour N02 standard which is effectively
six to seven times more stringent than the present EPA annual average for
N02- Aware of these events, EPA reexamined the Chattanooga data to
evaluate the study's validity as a basis for a short term N02 standard.
If short term exposures could be correlated with respiratory or other
health problems, a short term standard could be developed. Based on
available health study data, EPA concluded that if concentrations exceeded
200 ug/m3 no more than 10 percent of the time, then adverse effects on
human health would be prevented. Using statistical techniques, EPA then
concluded that if the 100 yg/m3 annual average standard is maintained,
the short term criterion (of no more than 10 percent of the measured one
hour concentrations in excess of 200 yg/m3) would be achieved in every
AQCR except Chattanooga.
However, in light of continuing studies on short term N02
exposures, it became clear that the necessity for a short term N02
standard needed to be further assessed. Accordingly, the Clean Air Act
Amendments of 1977 require EPA to promulgate, not later than August 1978,
a national primary air quality standard for N02 concentrations over a
period of not more than three hours, unless it is demonstrated that
sufficient evidence for such a standard does not exist.
Since the enactment of the Clean Air Act Amendments, EPA has
released a draft summary of the scientific basis for a short term N02
standard (Reference 3-5). A public meeting to receive comments from
54
-------
industry and the public was held in Washington, D.C. in April 1978. Based
on these, a standard in the range of 250 to 1000 yg/m^ for a one hour
average is being considered. However, no recommendation has yet been
submitted, and it now appears that proposal and promulgation will be
delayed until 1979.
In the following subsections causes of high short term N02 levels
are described, and the results of an analysis of predicted high short term
N02 levels in Chicago are presented. The potential for violation of
various short term standards is discussed. Potential point source
dominated violations are examined using a modeling technique, and
potential area source dominated violations are considered by evaluating
monitoring data.
3.2.1 Causes of High Short Term NO? Levels
Recent studies performed by EPA have shown that high short term
N02 concentrations come about through any one of several paths:
Area source emissions (both mobile and dispersed stationary
sources)
Isolated point sources with multiple combustors impacting on a
single site
Multiple point sources impacting on the same receptor
Both area and point sources with all sources contributing to
high concentrations
Terrain impaction by a plume from a large point source
The relative importance of these paths is highly dependent on both the
level which is established as the short term standard and the relative
contribution of each "source" to the short term N02 levels. Moreover,
the NOX control requirements may be significantly different for each
path.
An assessment of high short term NO;? concentrations must
therefore consider each type of source (point and area) and its respective
contribution to ambient concentrations. Although studies have shown that
either type can lead to relatively high concentrations, the nature of
their impacts is different. Point sources tend to produce infrequent and
spatially confined N02 peaks, although the slow formation rate of N02
smooths out these "hot spots" to some extent. Area sources, on the other
hand, are less varied in their impact on peak N02 levels in both time
and space.
Meteorological conditions can also determine the role of both point
and area sources in the short term buildup of pollutants. High ground
level concentrations from elevated point sources can be caused by surface
inversion breakup, fumigation (plume trapping where the plume is confined
beneath an elevated inversion), or plume downwash where the plume
55
-------
intersects the ground quickly. The meteorological parameters that
frequently characterize these conditions are an unstable atmosphere
(stability class of B or C) and moderate to high wind speeds. On the
other hand, the greatest impacts of ground level sources, such as vehicles
and other area sources (including point sources with short stacks), occur
when the atmosphere is quite stable (stability class of D or E*), wind
speeds are low, and mixing heights are small. The meteorological
conditions that maximize the impact of either point sources or ground
level area sources thus are at two opposite extremes, discouraging their
individual maximum impact from occurring simultaneously. However, the
difference in the impact of either source type at the maximum impact
condition and that of the other source type may not be very large.
For these reasons, studies performed in support of a short term
NOX standard have sought to model the NOX emissions of point and area
sources together to determine what conditions maximize the contribution of
one or the other, or both, to short term N02 concentrations. One such
study employed a multiple point and area source model (RAM) to model NOX
source contributions to the air quality in the Chicago AQCR
(Reference 3-6).
The RAM model is a Gaussian steady-state model capable of
predicting short term ambient concentrations of relatively stable
pollutants from multiple point and/or area sources. However, NO;? is
primarily a secondary pollutant formed by oxidation of NO. The initial NO
concentration present in exhaust gases, the plume diffusion and travel
time, and the ambient concentration of photochemical oxidants and reactive
hydrocarbons are some of the most important factors that affect conversion
of NO in a plume to N02- Consequently, a dynamic model of N02
formation from point source NOX emissions has been used in conjunction
with RAM to translate the predicted NOX concentrations at a receptor
arising from point source contributions to N0ฃ concentrations. For
estimating area source contributions to the background concentrations in
this study, a fixed N02/NOX ratio for each period of the day, based on
observed data at continuous monitoring sites in Chicago, was used to
translate ambient NOX concentrations derived from area sources into the
corresponding N02 levels.**
The results from this Chicago study (Reference 3-6) are summarized
in Table 3-3. Part (a) of the table shows the results for meteorological
conditions which maximize the point source impact relative to the area
source impact. The five highest N02 concentrations and the corresponding
*A11 AQCR's listed in Table 3-2 have predominantly D and E stability
classes
**An assumption used in the work of Reference 3-6 was that most monitors
in urban areas are sited to reflect contributions primarily from area
sources of NOX emissions. This assumption appears true based on
available evidence.
56
-------
TABLE 3-3. COMPARISON OF ESTIMATED N02 LEVELS FROM POINT AND AREA
SOURCES UNDER DIFFERENT METEOROLOGICAL CONDITIONS IN
CHICAGO* (yg/m3)(Reference 3-6)
en
Five Highest
Concentrations
1
2
3
4
5
Average concentration
for all receptors
above 200 pg/m3
Number of Receptors
above 200 yg/m3
(a) Meteorology for Maximum
Relative Point Source
Impact (yg/m3)
Total
509
589
348
348
342
277
Point
428
409
209
225
219
165
47
Area
81
81
139
123
123
111
(b) Meteorology for Maximum
Relative Area Source
Impact (yg/m3)
Total
568
479
472
472
472
371
Point
549
279
272
272
272
142
68
Area**
19
200
200
200
200
199
(c) Maximum Total Impact
(yg/m3)
Total
603
602
600
598
553
316
Point
493
434
407
430
383
142
67
Area
110
168
193
168
170
174
*Cook, Dupage, and portions of Will, Lake and Porter Counties
**The receptors used in the analysis were selected to record maximum total concentration.
Other receptors may reflect higher area source contributions, but lower total concentrations.
T-1175
-------
area and point source contributions to these concentrations, the number of
receptors reporting one hour levels above 200 yg/m^, and the average
concentration for these receptors are shown. Part (b) shows the results
for meteorological conditions which maximize the relative area source
impact. Results for meteorological conditions which lead to the maximum
total concentrations are shown in part (c). A summary of the conclusions
from this study is presented below:
t NOX emissions from either point or area sources can result in
high short term NO;? concentrations, although the point
sources are the major contributors*
Two distinct groups of point sources can be identified in terms
of their response (dilution and N02 formation rate) to
different meteorological conditions: (1) plants with tall
stacks such as utilities, and (2) plants with a large number of
short stacks such as steel mills and refineries
The diffusion characteristics of the second point source group
seem to be similar to those of the area sources
The meteorological conditions that maximize the impact of
sources with high effective stack heights are different from
the conditions that result in high concentrations from both
area sources or point sources with short effective stack heights
A set of meteorological conditions closer to the area source
maxima on the spectrum of diffusion conditions resulted in the
highest short term N0ฃ concentrations
These conclusions indicate that differing meteorological conditions can
maximize contributions of point and area sources, separately or
synergistically. For sources in urban areas, the multiple point and area
source influence is overriding ("maximum impact case" in Table 3-3). In
this case, point source influences are at a maximum simultaneously with
high area source contributions.
3.2.2 Potential Extent of Short Term NO? Violations
Of major concern from the perspective of NOX control requirements
is the extent (severity) of nonattainment on a nationwide basis for
various levels of short term N0ฃ standard. As noted above, ambient
concentration levels of N0ฃ are created by a mix of emissions from both
types of sources. However, the two categories of sources may be evaluated
separately to determine possible N0ฃ short term concentrations under
differing situations. Such a procedure has recently been employed to
determine the nationwide impacts of meeting a possible short term N02
*Note that discussion in the preceeding subsection attributed high annual
average N02 concentrations primarily to area sources.
58
-------
standard. Highlights of this work, reported in Reference 3-7, are
described in the following paragraphs.
Point Source Impacts
Because detailed modeling of all point sources in each AQCR is far
too ambitious for a nationwide study, a "model plant" technique was
devised using National Emission Data System (NEDS) data. This analysis
modeled a series of prototypic combustion plants which ranged in size and
operating parameters corresponding to various source categories (e.g.,
utility boilers, industrial boilers, and furnaces). The plants were
analyzed individually, using a simple Gaussian dispersion model and
meteorological conditions associated with ground level maximum N02, to
assess the air quality impacts of all respective NOX sources in the NEDS
file. The ground level NOX concentrations around each point source
characterized by the dispersion model were translated into N02 using the
dynamic NOX to N02 conversion model referred to in the discussion of
the RAM model.
Area source contributions to background N02 levels were
determined from N02 monitoring information within individual AQCR's.
Studies have shown that the annual average N02 concentrations in urban
areas are mainly due to area source influence and are relatively less
sensitive to point source impacts. Consequently, it is reasonable to use
observed annual average concentrations to quantify the area source
influence.*
The results of the point source analysis showed that a total of
4069 sources associated with 408 industries located in about 119 AQCR's
would produce violations of a 250 yg/m^ one hour standard. For a
500 yg/m^ standard, the number of affected sources and AQCR's decreases
significantly: 79 industries with about 1113 processes in about 30 AQCR's
would produce violations. Table 3-4 shows the types of processes
nationwide that are likely to be associated with violations of various
standard levels. Table 3-5 lists the corresponding number of AQCR's
projected to be in violation in 1975 and in 1982 for various short term
N02 standards.
Area Source Impacts
As a means to capture the maximum impact of area source emissions
directly, a simple modeling analysis (Reference 3-8) was used on monitored
N02 concentrations (and current NOX emission levels for mobile and
stationary sources) in those AQCR's which may experience future short term
problems. As previously stated, the N02 monitoring networks in most
*Based on the Chicago study, 1.5 times the highest recorded annual average
N02 was estimated as the area source background for the point source
analysis.
59
-------
TABLE 3-4. ESTIMATED POINT SOURCE RELATED VIOLATIONS OF VARIOUS ONE HOUR NOe STANDARDS (Reference 3-7)
Source Category
Utility Boilers -- Coal
Utility Boilers -- Oil
and Gas
Industrial Boilers Coal
Industrial Boilers Oil
and Gas
Gas Turbines
Reciprocating 1C Engines
Industrial Processes
0 Combustion
Nitric Acid
Municipal and Industrial
Incinerators
Total
Number of Sources Exceeding the Specified One Hour N02 Concentration (yg/m^)
250
350
599
300
742
268
698
1045
61
11
4074
500
42
7
72
207
19
516
235
19
1
TIT8
750
15
0
10
108
10
376
114
3
0
~636
1000
0
0
0
21
5
278
17
3
0
~^24
-------
TABLE 3-5. ESTIMATED NUMBER OF AQCR's IN VIOLATION OF ONE HOUR
N02 STANDARD BASED ON POINT SOURCE IMPACT (Reference 3-7)
Standard (yg/ttH)
1000
750
500
250
AQCR's in Violation
1975
6
11
30
119
1982a
6
11
28
116
1982, assumes 3 percent annual growth rate in VMT and expected
20 percent overall reduction in area source emissions due to mandatory
mobile source emission reduction requirements
61
-------
AQCR's are believed to reflect the impact of area, as opposed to point,
source emissions. Thus, using ambient air quality data to analyze short
term N02 pollution from area sources is justifiable.
The sample of AQCR's on which the area source analysis was based
included all those estimated to have current one hour N02 concentrations
above 200 yg/m3 (Reference 3-8). Table 3-6 summarizes the results of
the area source analysis in terms of two growth scenarios: 1) "low,"
assuming zero percent and one percent increases in stationary and mobile
sources, respectively; and 2) "high," assuming one percent and three
percent increases in stationary and mobile sources, respectively. Both
scenarios assume mobile source emission standards will remain as currently
mandated. Except for the 250 yg/m3 standard, only a few AQCR's are
estimated to be in nonattainment status due to area source emissions. The
current Federal motor vehicle control program is seen to effect a
considerable improvement in attainment status over time, although almost
70 AQCR's may still experience violations in 1990 for the 250 yg/m3
standard. It is important to note the major impact of vehicle emission
controls is realized in the late 1980's. This and the conservative growth
rates are the primary reasons that air quality is shown to improve in this
period. However, after 1990 air quality is projected to deteriorate (see
Section 8.2.1).
In summary, violations of possible short term standards may be
caused by either point or area source emissions and could occur for a
variety of meteorological conditions. Based on the results given in
Tables 3-5 and 3-6, it appears that at least 119 AQCR's, based only on
point source analyses, would be in violation of a 250 yg/m3 standard (in
1975) if sufficient monitors were available to record them (see Section
3.5). Moreover, it is unlikely that these 119 AQCR's include all 94
estimated in the area source analysis; therefore, the number in violation
is estimated to be 158. However, less than 100 AQCR's currently would be
in nonattainment based on N02 levels recorded at existing monitors,
which are predominately located to reflect area source impacts
(Table 3-6). By 1982, and without additional emission controls, the
violating AQCR's could be as few as 68 (current monitor placement) or 116
(ideal monitor placement). Considering probable duplication in violating
AQCR's in Tables 3-5 and 3-6, the number of nonattainment AQCR's in 1982
for a 250 yg/m3 standard is estimated to be 145. Of course this
estimate is again based on conservative growth rates and a successful
mobile source task NOX control program. Thus, the number of violations
is projected to continue to decrease until the mid to late 1980's and then
start to increase.
3.3 OTHER CLEAN AIR ACT PROVISIONS
Overall, promulgation of the short term N0ฃ ambient standard may
be viewed as the most important immediate regulatory development relating
to NOX controls needs. However, other significant Clean Air Act (CAA)
provisions also govern (or will govern) the need for NOX controls.
These include New Source Performance Standards (standards of performance
for new stationary sources, NSPS) governing the emissions of NOX from
specific sources; Prevention of Significant Deterioration (PSD) provisions
62
-------
TABLE 3-6. ESTIMATED NUMBER OF AQCR's IN VIOLATION OF ONE HOUR N02
STANDARD BASED ON AREA SOURCE IMPACT^ (Reference 3-7)
Standard (yg/m^)
1000
750
500
250
Number of AQCR's in Violation
1975
0
2
17
94
1982
High
Growth13
0
2
10
84
Low
Growthb
0
0
4
68
1990
High
Growth
0
0
7
73
Low
Growth
0
0
2
45
aBased on 150 AQCR's recording (or estimated to exhibit) second highest
one hour N02 levels of 200 yg/m^ or more in 1975
''"Low growth" assumes a one percent annual growth rate for VMT and a
zero percent annual growth rate for stationary area sources. "High
growth" assumes a three and one percent annual growth rate for VMT and
stationary area sources, respectively. Statutory mobile source emission
standards are also assumed.
63
-------
governing both NOX emissions and ambient concentrations; and the
nonattainment policy governing both NOX emissions and ambient
concentrations.
New Source Performance Standards are technology based emission
standards. Development of NSPS will affect specific technologies at
different times; in general, however, the implications of NSPS will be
straightforward, requiring available control of NOX emissions from
individual source categories at a cost determined by EPA to be
appropriate.
The implications of the PSD and nonattainment provisions, on the
other hand, are not so straightforward. The nature and the stringency of
either provision as applied to stationary sources will depend on the short
term NOg standard promulgated. In essence, the short term N02
standard provides only a foundation for the establishment of appropriate
PSD and nonattainment regulations, which actually impact sources of NOX
through implementation of the individual SIP's.
In this section the PSD and nonattainment provisions of the CAA as
they relate to a short term N02 standard are discussed. It is important
to note that SIP regulations arising from either provision affect
technology cost, through NOX emission control requirements, on a
regional basis, depending on the short term N02 air quality of that
region.
3.3.1 Prevention of Significant Deterioration
Prevention of Significant Deterioration provisions (Sections 160
through 169 of the CAA) are designed to protect air quality in areas now
meeting all ambient standards. PSD regulations perform three interrelated
functions: (1) they limit the degradation of air quality in "clean air"
areas; (2) they provide a mechanism to regulate pollutant emissions from
new sources; and (3) they allow the individual states to determine the
degree of new source growth desired in clean air areas.
The PSD provisions outlined in the CAA allow for three area
classification categories: Class I, where practically any air quality
deterioration would be precluded; Class II, where deterioration in air
quality arising from moderate growth would not be considered significant;
and Class III, where intensive and concentrated industrial growth can
occur while not departing from the intent of the PSD regulations. The
area classification plans are to be executed and enforced through revised
SIP's.
Specific ambient pollutant increment concentrations are assigned to
each classification category which, when added to the determined
"baseline" pollutant concentrations in a given area, prescribe the maximum
allowable air quality degradation for that area. The number of new
sources or expansions allowed in a given area are regulated through the
preconstruction permitting process. This process requires a new source to
demonstrate its strategy for compliance with the PSD increments and, among
a number of specific stipulations, requires new sources to employ "best
64
-------
available control technology" (BACT). BACT, as defined by the Act, means
an emission limitation based on the maximum degree of pollutant reduction
available, taking into account energy, environmental, economic, and other
costs. In no event can BACT mean an emission limitation less stringent
than that allowed under the NSPS for a particular source. The most
important aspect of BACT is that the States are empowered to determine it
on a case by case basis.
The Clean Air Act stipulates that the pollutants sulfur dioxide
(S02) and particulates presently be covered by PSD regulations within
SIP's. By 1980 the EPA is to conduct a study to determine whether and how
other pollutants also are to be covered by PSD. The pollutants to be
studied include nitrogen oxides, hydrocarbons, carbon monoxide, and
photochemical oxidants. The regulations, if and when the EPA does
promulgate them, must provide specific measures at least as effective as
the increments established for S02 and particulates. Such measures may
include air quality increments and specific numerical measures against
which permit applications may be evaluated.
Implications of PSD
If a short term N02 standard is promulgated, PSD provisions will
affect the initial siting and/or the expansion (or addition) of major
stationary sources with respect to NOX. Depending on the level of the
N02 ambient standard, N02 PSD provisions may establish lower levels of
allowed N02 ambient degradation to protect the air quality in different
PSD classification regions. In all cases, PSD provisions will enforce the
use of BACT on stationary sources as a mechanism to ensure compliance with
allowed short term ambient concentrations of N02.
The type and level of control established under BACT can vary
according to different regions, since it is a case by case determination
allowing states to choose the amount of new source growth desired. In
essence then, the implications of any PSD N02 regulations are economic;
they concern the cost of controlling new sources in a region so as to
ensure compliance with an established level of N02 ambient degradation.
Unfortunately, it is difficult to predict the regional impacts of a NOX
PSD regulation since: (1) the level of N02 ambient degradation allowed
under PSD is unknown, and (2) the nature of BACT and the amount of local
growth desired cannot be assumed.
3.3.2 The Nonattainment Policy
The nonattainment provisions of the Clean Air Act Amendments of
1977 outline regulations governing the introduction of new sources in
regions which have been shown by monitoring data (or calculated by air
quality modeling) to exceed any national ambient air quality standard.
Under the CAA, revised SIP's for these regions must assure attainment of
primary air quality standards for NOX and the other criteria pollutants
no later than December 31, 1982; with respect to especially severe oxidant
and carbon monoxide problems, the deadline may be extended to December 31, 1987,
65
-------
Before July 1, 1979, the interpretive EPA regulation published
December 21, 1976, governing nonattainment regions shall apply to new
sources wishing to enter such regions. The EPA regulations specify an
emissions "trade-off" policy which requires:
"emission reductions from existing sources in the area of a
proposed source (whether or not under the same ownership) such
that the total emissions from the existing and proposed sources
are sufficiently less than the total allowable emissions from
the existing sources under the SIP prior to the request to
construct or modify, so as to represent reasonable progress
toward attainment of the applicable NAAQS."
"Trade-off" may occur only if the state has an enforceable SIP which
requires new sources to meet Lowest Achievable Emission Rate (LAER).*
After July 1, 1979, the state must have a revised and approved
implementation plan, assuring attainment as a precondition for the
construction or modification of any major stationary source. This plan
must include a permitting process for construction or modification of
major stationary sources in nonattainment areas. A permit may only be
granted if the following conditions are met:
Total emissions in the proposed modification/construction area
must be significantly less after the modified or new facility
is in operation than before
The proposed source is in compliance with LAER
t The owner or operator of the proposed new or modified source
has demonstrated that all major stationary sources owned or
operated by such person in such state are in compliance, or on
a schedule for compliance with all applicable emission
limitations and standards under the Act
Implications of the Nonattainment Policy
The implications of the nonattainment policy, much like those
arising from PSD, primarily concern the economics of facility siting and
operation. Unlike PSD, which governs pollutants in areas now meeting
ambient standards, the nonattainment provisions provide mechanisms
designed to ensure attainment in those areas presently violating standards
with regard to a particular pollutant. To progress toward attainment, the
nonattainment provisions stipulate that emissions from existing sources be
reduced accordingly. If new sources are to be added in the region, the
*LAER is defined as the lowest applicable emission rate contained in any
State Plan or the lowest emission rate achievable in practice by that
category of source, whichever is lower.
66
-------
nonattainment provisions require that emissions from existing sources be
reduced so that resulting total emissions represent progress toward
attainment, and that new sources meet very strict emission limitations,
essentially regardless of the cost.
If a short term N02 standard is promulgated, the states must
revise their SIP's within nine months. The SIP revisions must provide
attainment of the standard within three years. In this regard, attainment
of a one hour N0ฃ standard may be required by mid-1982. Depending on
the level of standard set, the number of regions placed in violation of
the standard and the cost to attain the standard will vary.
A primary element of the studies supporting EPA's development of a
short term N02 standard has been to estimate the cost of
attainment/compliance with various standard levels being considered. In
one such study (Reference 3-7), the cost to attain a 250yg/m3 ambient
one hour N02 level in the Chicago AQCR was assessed for three different
control approaches. The control strategies considered were:
Least cost controls applied to specific sources only as
necessary to reduce ambient concentrations below the required
level for the meteorological conditions leading to maximum
ambient levels*
0 RACT with least cost all point sources initially are
required to implement controls that have been demonstrated and
are reasonably economical (Reasonable Available Control
Technology). Additionally, incremental controls that are
needed to meet the standard after RACT implementation are
imposed in a cost minimizing manner.
Maximum feasible control -- all sources implement the greatest
degree of NOX control available. This is comparable to a
situation in which all sources are required to reduce emissions
by 90 percent, regardless of their impact on air quality
In all three cases, the control approaches were designed to achieve the
standard at all receptors based on existing sources; they did not consider
new sources and the associated cost of achieving LAER and purchasing
offsets.
*Under the least cost solution, at each receptor, each contributing source
is controlled to the level at which its marginal cost of control per unit
reduction in ambient N02 concentration is less than that for any other
source contributing to the same receptor. Once the source with the lowest
marginal cost is controlled to this level, other sources are controlled in
sequence, starting with the source with the next lowest cost, until the
standard has been achieved at all receptors.
67
-------
TABLE 3-7. SIMULATION OF RESULTS FOR ATTAINMENT OF A 250 yg/m3 ONE HOUR
N02 STANDARD IN THE CHICAGO AQCR (Reference 3-7)
Pure Least Cost
RACT w/Least Cost
Maximum Feasible Control
Sources
Controlled*
94
797
797
Emissions
Reduction
(Mg/hr)
2.3
18.2
48.1
Annual Control
Cost to Emitters
(106 $/yr)
21
53
588
aNo new sources considered
The estimated costs of attainment for each of these cases are shown
in Table 3-7. The pure, least cost solution provides attainment without
necessarily requiring controls for all existing sources. The RACT with
least cost option affords attainment through implementation of available
control to all existing sources and any additional control thereafter
required to attain the standard. The maximum feasible control option
"penalizes" all point sources regardless of their contribution to ambient
standard violations. This latter situation may depict a nonattainment
AQCR implementing a control strategy that provides for the maximum growth
allowance attainable. As expected, this strategy results in the greatest
cost impact. It is also quite clear that the control approach to
attainment will have a tremendous impact on the number of sources
controlled, the amount of pollutant removed, and the technical
requirements placed on the control technology itself.
3.4 RELATED ISSUES
In this section three additional issues associated with NOX
control needs are briefly discussed: the interrelationship of NOX and
HC control strategies on both N02 and oxidant, other secondary
pollutants related to NOX, and the increased utilization of coal.
3.4.1 The NOy-HC Relationship
The interrelationship of NOX and HC emissions in affecting N0ฃ
and oxidant ambient concentrations is a well known fact; although the
specific details are not that clearly understood. Smog chambers and
analytical photochemical models have been used to study the chemistry of
the NOX-HC system. The necessity to consider this connection in the
evaluation of control strategies has been recognized by EPA and is
68
-------
contained in the isopleth method for assessment of control needs for
meeting ambient oxidant levels (Reference 3-9).
It is now clear that any control strategy for one of these
pollutants must consider the consequent impact on ambient levels of both
N02 and oxidant. This is made more complex because the impact depends
on the existing ambient concentrations, the spatial scale of interest, and
the time duration of interest. For example, control measures to maximize
improvement in urban N02 may result in an increase in rural (downwind)
oxidant levels. As another example, control strategies to reduce one hour
peak concentrations may only shift the occurrence of the peak (in time or
space) or may reduce the peak but have no effect on the annual average.
Extensive study of the oxidant problem in the San Francisco AQCR
(Reference 3-10) has shown the detrimental effects of NOX control on
attainment of ambient oxidant goals. However, it is also recognized that
current, or future, violations of N02 standards must be anticipated.
Generally, it appears that the best approach will be simultaneous control
of both NOX and HC with the particular mix being determined by AQCR
specifics. (Results for such strategies for the San Francisco AQCR are
presented in Section 8.2.2.)
3.4.2 Secondary Pollutants
Acid rain, nitrate aerosols, organic aerosols, sulfate aerosols,
PAN (peroxyacetylnitrate), and nitrosamines are either known or thought to
be secondary pollutants of NOX. Presently, EPA does not regulate these
pollutants. However, studies to determine the necessity for their
regulation are being conducted.
Promulgation of a short term N02 standard could result in a
reduction in the occurrence of these pollutants, although, in many cases,
their link to NOX has not been verified. How a short term standard will
affect possible regulation of these secondary pollutants is unknown. Of
the pollutants mentioned, acid rain is the closest to being regulated.
EPA is conducting continuing studies on acid rain and hopes to regulate it
by 1981. Studies on sulfate aerosols also are being conducted; whether
they will be regulated, however, is unclear. If sulfate aerosols are
regulated, it may be to protect visibility, and not as a requisite to
protecting public health. Presently, EPA cannot ascertain whether NOX
is implicated in sulfate aerosol formation.
Nitrate aerosols, organic aerosols, and PAN most likely will not be
regulated for 10 years or more. Presently, EPA has tremendous difficulty
measuring nitrate and organic aerosols. These problems must be resolved
to better understand the role of NOX in the formation of these
pollutants. With regard to PAN, EPA will examine the role and formation
of this pollutant when it next reassesses the existing annual N02
standard.
How these secondary pollutants will be regulated is not clear. It
is possible that, in some cases, regulation would take the form of further
restrictions on NOX emissions beyond those required to comply with a
short term N02 standard for certain regions. However, until the role of
69
-------
N02/NOX in the formation of these pollutants is better understood, the
interaction of the short term N0ฃ standard with the occurrence of and
possible regulation of secondary pollutants cannot be evaluated.
3.4.3 Coal Utilization
In general, coal combustion results in higher emissions of NOX
than comparable combustion processes utilizing oil or gas. Consequently,
depending on the standard level chosen, attainment of a short term N02
standard may be difficult and/or costly for those areas having substantial
numbers of coal based emission sources. In some respects, promulgation of
a short term NC>2 standard may have the most significant policy
implication on the role of coal in the National Energy Plan.
Unfortunately, studies supporting EPA's development of a short term N02
standard have examined the standard's implication primarily with regard to
the existing fuel use structure of industries and utilities and have not
treated potential increased coal usage. Thus, any conclusion drawn here
would be premature.
3.5 THE N02 MONITORING NETWORK
The discussion in the preceeding sections focused on the
implications of various regulatory activities in driving the need to
develop and implement stationary source NOX control techniques.
Implicit in all this discussion is the fact that NOX control needs are
really defined by the extent of the potential violation problems
associated with any given regulation or standard. Of course, the number
and extent of standards violations can only be determined from readings
obtained through an air quality monitoring network. Thus, it seems
appropriate here to briefly discuss the status, and potential shortcomings
of the existing N02 monitoring network.
Two types of ambient N02 monitors are currently in use in the
U.S.: 24 hour bubblers and continuous monitors. The 24 hour bubblers,
most of which use the sodium arsenite method, can be used to determine
annual average N02 concentrations. The continuous monitors, most of
which use the chemiluminescence or the Saltzman method, are used to
measure both one hour and annual average N02 levels. The sodium
arsenite, chemiluminescence, and continuous Saltzman method all are
considered acceptable monitoring methods by EPA.
In 1975 and 1976, approximately 1613 to 1740 N02 monitors
operated in the U.S.; of these, only about 260 were continuous monitors
(as shown in Table 3-8). EPA considers the existing 24 hour monitoring
network, designed to record annual average N02 concentrations, as too
extensive. Consequently, EPA is recommending that the number of 24 hour
bubblers be reduced, as more continuous monitors come into use, since
continuous monitors are capable of supplying both short term and annual
average N02 concentration measurements.
Although most of the continuous monitors are placed in large
cities, a number of cities with populations greater than 200,000 have no
continuous monitors. Thus, if a one hour N02 standard is promulgated,
70
-------
TABLE 3-8. THE U.S. N02 MONITORING NETWORK
Monitors in Operation
Year
1975
1976
Continuous
258
260
24 Hour
1355
1480
Total
1613
1740
Monitors Recording
Valid Annual Averages
715
1123
71
-------
each state would have to assess the adequacy of its continuous monitoring
network, and upgrade it, if necessary, as part of its State Implementation
Plan revision in response to the new standard. The adequacy of the
monitoring network will thus vary from region to region and will be
evaluated by EPA on a regional basis in its review of SIP's.
In any event, if continuous N02 monitors were common in all
AQCR's the likelihood of any AQCR violating the various suggested short
term N02 standards could be determined with ease. Unfortunately,
continuous monitors are not common. As a consequence, the potential for
violation must be judged from 24 hour readings or annual averages. This
can be done by establishing peak to mean ratios for representative
continuous monitors and using these values to extend the annual average
values determined from the 24 hour monitors.
Evidence now available suggests that the ratio of one hour peak
readings to annual average levels for area source dominated monitors is
less than six to one (References 3-3 and 3-8). Moreover, it has been
reported that the average peak to mean ratio lies between six and seven
for continuous monitors in central urban commercial and residential areas
(Reference 3-11). Area sources are undoubtedly the major contributors to
the N02 concentrations at these sites, although point sources in the
region do have some impact. However, the ratio due to the area source
impact alone should be below this six or seven peak to mean value. The
peak to mean range of four to six, therefore, seems to be associated with
sites impacted predominantly by area sources, and urban area monitors
reporting peak to mean ratios of over six are believed to be significantly
impacted by point sources.
As a general rule, annual average N0ฃ values can be extended to
peak one hour values by assuming a peak to mean ratio of six. It should
be noted, however, that monitors intended to provide annual average data
may not be properly located to record maximum one hour values and that
locations heavily impacted by point sources may have peak-to-mean values
as large as 12. Thus estimates of the impact of various short term
standards using extensions of existing annual average data at a ratio of
six to one should be considered as conservative.
3.6 SUMMARY
In this section, a variety of environmental issues related to
assessing present and future NOX control requirements have been
discussed. Many points were brought out which deserve summary and
reiteration. Therefore, the main points of the discussion are summarized
and the major conclusions are as follows:
Promulgation of a short term N02 standard may have major
impact on NOX control needs and NOX control strategies
Area sources (mobile and dispersed stationary) appear to be the
primary contributors to high annual average N02 levels
72
-------
Large point sources or concentrated smaller point sources
appear to be the major contributors to high short term N02
levels, although, area sources may also be significant
contributors
Four AQCR's are presently in nonattainment with respect to the
100 yg/m3 annual average, based on current monitoring data.
It is estimated that 30 more would be in violation if
sufficient monitoring data were available.
Approximately 100 AQCR's would presently be in violation of a
short term standard of 250 yg/irn, based on estimates from
current monitoring station data
t The number of AQCR's in violation of short term or annual
average standards will probably decrease in the near term (1980
to 1990) but increase in the long term (2000) without
additional stationary source control beyond current and
projected NSPS
PSD and nonattainment regulations in conjunction with a short
term standard, will have major impacts on NOX control
requirements, the number of sources controlled, and the cost of
control
Simultaneous control of NOX and HC must be considered if both
N02 and oxidant ambient goals are to be met
The current monitoring network reflects the impact of area
sources and, even so, does not adequately measure the extent of
violation of the annual average
If a short term standard is promulgated, many more continuous
monitors will be required to adequately measure short term
N02 levels. It will be necessary to site these monitors to
record the impact of point sources.
73
-------
REFERENCES FOR SECTION 3
3-1. The Federal Register, Vol. 43, No. 43, March 3, 1978.
3-2. "Preliminary Evaluation of Potential NOX Control Strategies for
the Electric Power Industry, Vol. 1," EPRI Report FP-715, March
1978.
3-3. Personal communication with R. Morris, Office of Policy Analysis,
Department of Energy.
3-4. "National Air Quality and Emission Trends Report -- 1975,"
EPA-450/1-76-002, NTIS PB-263 922/7BE, November 1976.
3-5. "Health Effects for Short-Term Exposure to Nitrogen Dioxide
(Draft)," EPA, Office of Research and Development, December 1977.
(Incorporated into "Air Quality Criteria for Oxides of Nitrogen"
(Draft), Environmental Criteria and Assessment Office, Office of
Research and Development, November 1978.)
3-6. "NOX Source Assessment of the Impacts of the Chicago
AQCR - Volume III," Draft report prepared for EPA by Energy and
Environmental Analysis, Inc., Arlington, VA, December 1978.
3-7. "Estimated Cost of Meeting Alternative Short-Term N02
Standards - Volume II," Draft report prepared for EPA by Energy and
Environmental Analysis, Inc., Arlington, VA, December 1978.
3-8. Thuilliers, R. W., and W. Viezee, "Air Quality Analysis in Support
of a Short-Term Nitrogen Dioxide Standard," Draft report prepared
for EPA by SRI International, Menlo Park, CA, December 1977.
3-9. Freas, W. P., et ^1_., "Uses, Limitations, and Technical Basis of
Procedures for Quantifying Relationships between Photochemical
Oxidants and Precursors," EPA-450/2-77-021a, NTIS PB-278 142/5GI,
November 1977.
3-10. "Environmental Management Plan for the San Francisco Bay Region -
Draft Air Quality Maintenance Plan," Prepared by the Association of
Bay Area Governments, the Bay Area Air Pollution Control District,
and the Metropolitan Transportation Commission, December 1977.
3-11. Trijonis, J., "Empirical Relationships between Atmospheric N02
and its Precursors," EPA-600/3-78-018, NTIS PB-278 547/AS, February
1978.
74
-------
SECTION 4
ENVIRONMENTAL OBJECTIVES DEVELOPMENT
Addressing the goals of the NOX EA program, as stated in Section
1, requires performing impact assessments of NOX sources and
source/control combinations of three general types:
Multimedia environmental impact assessments of individual
sources under both baseline and controlled operation
t Operational and cost impact evaluations of applying NOX
combustion modification controls to individual sources
Air quality impact assessments of applying different NOX
control strategies to the accumulation of sources on a regional
basis
Thus, at the individual combustion source category level,
evaluations are needed of the environmental impact of the multimedia
pollutant emissions from a given source under both uncontrolled (or
baseline) and controlled (for NOX) operation. Such evaluations are
needed not only to guide the setting of priorities for control development
recommendations. They are also needed to allow overall impact comparisons
between baseline operation and the application of various NOX control
options to ensure that the NOX control techniques are environmentally
sound, and to provide a basis for identifying preferred means of control.
Impact assessments of this type require the development of Source
Analysis Models (SAM's) which translate multimedia pollutant emissions
data into measures of potential hazard to health and welfare. Thus, such
SAM's will take emissions data, compare these to health or ecological
effects indicators, and output quantitative indicators of potential for
environmental harm.
Also needed at the individual source level are procedures for
assessing the effects of NOX control application on source efficiency,
operation, and costs of operation. These evaluations are needed to flag
potential adverse operational impacts of NOX control and to evaluate
their cost effectiveness and economic soundness. To perform this kind of
assessment requires detailed process and cost analysis methods.
75
-------
Finally, evaluations are needed of the effects on ambient air
quality of applying various NOX control strategies on a regional basis.
These are required so that the preferred, environmentally sound, and
cost-effective control strategies can be identified and, if found
insufficient, control R&D needs and recommendations can be formulated.
These assessments require the development of ambient air quality models
which translate source emissions data to ambient pollutant levels on a
regional basis.
The development of methodologies to address each of the above
assessment needs is described in this section. Thus, Section 4.1
describes the form of several Source Analysis Models developed for
pollutant impact assessments; Section 4.2 describes the process and cost
analysis methods used to evaluate the operational and cost impacts of
applying NOX controls to a given source; and Section 4.3 discusses the
systems analysis models used to evaluate the effects of various control
strategies on regional ambient N02 and 03 levels.
4.1 SOURCE ANALYSIS MODELS
As noted above, Source Analysis Models (SAM's) are required in
environmental assessment activities to treat source emissions data by
comparing them to health/ecological effects indicators and thereby
translate them into quantitative measures of potential environmental
hazard. In the NOX EA, SAM's have been, or are being, developed for
performing these comparisons in three levels of mathematical detail.
These SAM's are intended for use not only within the NOX EA, but in
other IERL environmental assessments as well. The three levels of SAM's
currently defined are:
SAM IA designed for rapid screening
SAM I designed for intermediate screening
t SAM II designed for regional site evaluation
All SAM's developed will use, as the requisite health/ecological
effects indicators, the set of Multimedia Environmental Goals (MEG's)
developed elsewhere (Reference 4-1). These MEG values represent either
defined maximum allowable effluent stream pollutant concentrations based
on acute toxicity considerations, or maximum allowable ambient pollutant
levels based on chronic exposure considerations. MEG's of the first type
(allowable effluent concentrations) are termed Minimum Acute Toxicity
Effluent (MATE) values. MEG's of the second type (allowable ambient
levels) are termed Estimated Permissible Concentrations (EPC's) or Ambient
Level Goals (ALG's).
Thus, each SAM developed will provide structured comparisons
between source pollutant emissions levels and a given set of MEG's to
produce the desired measures of multimedia environmental impact of a
pollutant source. To date procedures for SAM IA and SAM I have been
developed. In addition, an extended form of SAM I representing a
projected approach to SAM II has been defined. An overview of each of
76
-------
these methodologies follows. The detailed formulation of the SAM II model
will be performed in future efforts; thus it is not discussed below.
4.1.1 SAM IA
SAM IA was the first of the models developed and was designed for
rapid screening applications. In this relatively simple model,
comparisons between discharge stream pollutant concentrations are made
directly to corresponding MATE values. Individual pollutant Potential
Degrees of Hazard (PDOH) are defined as the ratio of an undiluted
pollutant concentration to its MATE value. A further impact indicator,
the Potential Toxic Unit Discharge Rate (PTUDR), is defined as the product
of the PDOH with discharge stream flow rate. PDOH's and PTUDR's are then
summed over all pollutants emitted in a given stream to yield the desired
measures of potential environmental impact. Details of the model are
explained more fully in Reference 4-2.
It should be noted here that efforts are currently underway to
incorporate the results of bioassay testing (Reference 4-3) into SAM IA so
that chemical analysis results from emissions testing can be qualitatively
compared to corresponding bioassay results. These efforts are not yet
complete.
4.1.2 SAM I
The next most sophisticated model developed was the SAM I model,
developed for intermediate screening purposes. SAM I comparisons
incorporate ambient level MEG's. Thus, in SAM I ambient MEG values are
translated to pollutant emission level concentration goals through the use
of dilution factors. Dilution factors were thus defined for a set of
discharge stream/receiving medium combinations (e.g., gaseous stream
discharge to the atmosphere, liquid stream discharge to a river, solid
stream discharge to a waste pile, etc.) based on the application of
dilution models.
A given pollutant emission level concentration goal is defined as
the product of an appropriate dilution factor with the pollutant ambient
MEG value. From this the pollutant species Potential Degree of Hazard is
defined as the ratio of the effluent stream pollutant concentration and
its emission level concentration goal. The corresponding PTUDR is defined
as the product of this PDOH with the given pollutant species mass
discharge rate.
As in SAM IA, PDOH's and PTUDR's are summed for each discharge
stream to provide the desired overall measures of potential environmental
impact. Further details of this model are given in Reference 4-4.
4.1.3 Extended SAM I
An extended form of SAM I was also developed to form a more
fundamental basis for the baseline combustion source impact rankings
discussed in Section 8.1. In this model a more detailed treatment is
77
-------
given to gaseous stream emissions to the atmosphere, while the SAM IA
methodology is retained for liquid and solid effluent streams.
In the treatment of gaseous effluents, the extended SAM I model
explicitly applies mathematical dispersion models in a continuous fashion
in contrast to the discrete treatment adopted in SAM I. Point source
emissions are treated using the Gaussian dispersion model tabulations of
Turner (Reference 4-5). Area sources are treated using the dispersion
model of Holzworth (Reference 4-6).
The environmental impact indicator defined in this extended SAM I
model is termed a potential impact factor and represents the ratio of
resultant ground level ambient pollutant concentration to the
corresponding MEG value integrated over exposed population. The model
also incorporates differing urban and rural population densities, and
installed pollutant source densities, and factors in corrections for
ambient background pollutant concentrations. Details of this model are
documented in Reference 4-7.
4.2 PROCESS IMPACTS EVALUATION
Evaluating the effectiveness and impacts of NOX combustion
controls applied to stationary sources requires assessing their effects on
both controlled source performance, (especially as translated into changes
in operational limitations, operating costs, and energy consumption) and
on incremental emissions of pollutants other than NOX. In this section,
the methods developed for use in the NOX EA to evaluate process impacts
~ correlation of NOX emissions with boiler/fuel variables, detailed
process analysis procedures, and cost analysis of controls -- are
discussed. The discussion centers on utility and large industrial
boilers, the largest source of stationary combustion source NOX, and the
source category treated in detail in second year efforts. The results
from applying the methodologies presented below to data assembled for
utility boilers are discussed in Section 7.
4.2.1 NOy Emissions Correlation
The key boiler/burner design and operating variables and fuel
properties affecting NOX formation were identified by performing
statistical correlations of NOX emissions data with these parameters.
Thus, the basis and effectiveness of control techniques which modify these
parameters were assessed. A second order regression model was used to fit
uncontrolled and controlled NOX emission data from field tests on a
total of 61 boiler firing type/fuel combinations. These combinations, the
controls applied, and the individual test points correlated are summarized
in Tables 4-1 and 4-2. The correlation parameters considered in the
analysis were:
t Boiler operating variables:
~ Overall furnace fuel/air stoichiometry
Stoichiometry at active burners
78
-------
TABLE 4-1. FIELD TEST PROGRAM DATA COMPILED
Fuel
Coal
Oil
Natural Gas
Total
Firing Type
Tangential
13
2
1
16
Opposed Wall
6
7
8
21
Single Wall
10a
7
7b
24
Total
29
16
16
61
'includes two wet bottom furnaces
Includes one unit originally designed for coal firing with a
wet bottom furnace
79
-------
TABLE 4-2. INDIVIDUAL TEST POINTS CORRELATED
Firing Typซ
Tangent 1il
Opposed
Uall
Single Wall
Tangential
Opposed
Hall
Single Uall
Tangential
Opposed
Uall
Single Uall
All Boilers
Fuel
Coal
Coal
Coal
Oil
Oil
Oil
Hat gas
Hat gas
Nat gas
All fuels
Baseline0
21
8
18
1
6
4
1
7
5
71
Single Controls
LEAC
29
11
23
5
6
1
9
4
as
ose"
46
11
29
1
11
S
--
18
9
130
F6Rซ
7
~
2
4
2
--
2
17
Low
LoacT
24
7
19
1
7
e
2
13
7
88
Cnbtned Controls*
Loo load
* OSC
27
5
19
1
7
6
1
13
7
86
Low Load
+ FGR
1
1
5
10
5
3
3
28
OSC *
FGR
--
2
2
10
1
3
4
22
Low Load +
OSC < FGR
1
11
8
8
5
33
Total
147
52
108
6
56
61
13
74
46
563
aLow excess air also generally e*>loyed
Baseline no controls applied; boiler load near or at MI Inure rating; excess air at
nonul or above nontal settings
CLEA low excess air setting
"OSC off stolchiooetrlc combustion (Includes: biased burner firing, burners out of
service, overflre air)
*F6R flue gat reclrculatlon; generally Includes low excess air setting
Load less than 80 percent of awxlMW continuous rating (NCR)
T-807
80
-------
-- Percent flue gas recirculated
-- Firing rate
Percent burners firing
Heat input per active burner
Boiler design variables:
-- Maximum continuous rating
-- Volumetric heat release rate
Surface heat release rate
Heat input per active burner
~ Number of burners
-- Number of furnaces
-- Number of division walls
t Fuel properties:
Fuel type (coal, oil, and natural gas)
-- Fuel nitrogen
Fuel moisture
Heating value
A multiple regression procedure was developed which statistically
correlated NOX emissions with the key parameters. This regression model
served as a predictive tool in estimating the emissions impact of NOX
controls, as well as identifying the important design and operating
parameters affecting NOX formation.
Results from applying this correlation model to utility boiler
field test data are discussed in Section 7.1.
4.2.2 Process Analysis Procedures
To evaluate the impact of controls on process operation, detailed
process variable data compiled for baseline and for low NOX operation
were analyzed and compared. Significant changes in the process variables
were noted, and these were highlighted as real or potential problems. A
summary of field test programs used as sources of process data is given in
Table 4-3. Specific test report references can be found in Reference
4-8. Process variables investigated are itemized in Table 4-4.
81
-------
TABLE 4-3. SUMMARY OF PROCESS DATA SOURCES
00
INJ
Furnace
Type
Tangential
Opposed Hall
Single Nail
Tangential
Opposed Hall
Single Hall
Turbo Furnace
Fuel
Coa?
Coal
Coal
Oil
on
011
011
Boiler
Barry Ho. Z
Barry No. 4
Hunttngton Canyon No. 2
Columbia No. 1
Navajo No. 2
Comanche No. 1
Kingston No. 6a
Harllee Branch No. 3
Four Corners No. 4
Hatfleld No. 3
E.C. Gaston No. 1
BM Units Nos. 1 * 2"
"FW Unit No. l"a
Widows Creek No. 5
Widows Creek No. 6
Crist Station No. 6
Nercer No. 1 ,
"FH Unit No. Z"a
"FH Unit No. 3"a
South Bay No. 4*
Plttsburg No. 7
Moss Landing Nos. 6 & 7*
Ormond Beach Nos. 1 4 Z
Sewaren Station No. 5
Enclna Nos. 1. 2 t, 3a
South Bay No. 3a
Potrero No. 3-1
Manufacturer
CE
CE
CE
CE
CE
CE
CE
BM
BM
BM
BM
BiW
FW
BM
BM
FH
FH
FH
FH
CE
CE
CE
BH
FH
BM
BM
BM
RS
RS
Utility Company
Alabama Power
Alabama Power
Utah Power and Light
Wisconsin Power I Light
Salt River Project
Public Service of Colorado
Tennessee Valley Authority
Georgia Power
Arizona Public Service
Allegheny Power Service
Southern Electric Generating
Tennessee Valley Authority
Tennessee Valley Authority
Gulf Power
Public Service Electric & Gas
San Diego Gas I Electric
Pacific Gas ft Electric
Southern California Edison
Pacific Gas ft Electric
Southern California Edison
Southern California Edison
Public Service Electric & Gas
San 01 ego Gas ft Electric
San Diego Gas ft Electric
Pacific Gas & Electric
NOX Control
Technique
BOOS. OFA
LEA, BOOS
OFA
OFA
LEA. BOOS, OFA
OFA
LEA. 8BF, BOOS
LEA, BOOS
BOOS, HI
BOOS, FGR
LNB, LEA, BOOS
LNB
LEA, OFA, LNB
LEA. BOOS
LEA, BOOS
LEA, BOOS
LEA, BBF
LEA, OFA, LNB
LEA. OFA, LNB
LEA. BOOS. RAP
OFA, FGR
FGR. BOOS
OFA. FGR
FGR. OFA, BOOS, HI
FGR. OFA. BOOS
LEA. BOOS
LEA. BOOS
A1r adjustment,
WI, RAP
OFA, FGR
New or
Retrofit
Retrofit
New, NSPS
New. NSPS
New, NSPS
New
New
Retrofit
New. NSPS
Retrofit
Retrofit
New, NSPS
Retrofit
Retrofit
OFA New
FGR Retrofit
OFA New
FGR Retrofit
--
Retrofit
Retrofit
Denotes new results or previously unreported data
-------
TABLE 4-3. Concluded
Furnace
Type
Tangential
Opposed Wall
Single Hall
Turbo Furnace
Fuel
Gas
Gas
Gas
Gas
Boiler '
South Bay No. 4*
Pitts burg No. 7
Moss Landing Nos. 6*7*
Ptttsburg Nos. 5(6
Contra Costa Nos. 9 t 10
Enclna Nos. 1. 2(3*
South Bay No. 3a
Potrero No. 3-1
Manufacturer
CE
CE
BftW
BiU
BiW
KM
RS
RS
Utility Company
San Diego Gas & Electric
Pacific Gas 4 Electric
Pacific Gas I Electric
Pacific Gas t Electric
Pacific Gas I Electric
San Diego Gas & Electric
San Diego Gas t Electric
Pacific Gas I Electric
NOX Control
Technique
LEA, BOOS
OFA, FGR
OFA, FGR
OFA, FGR
OFA, FGR
BOOS
Air adjustment
HI, RAP
OFA, FGR
New or
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
~
Retrofit
Retrofit
'Denotes new results or previously unreported data
00
CO
-------
TABLE 4-4. PROCESS VARIABLES INVESTIGATED
Process Variables
Boiler Load
Furnace Excess Air
Excess Air at Firing Zone
Percent Oxygen in Flue Gas
Percent Oxygen in Windbox
Furnace Cleanliness Condition
Percent Overfire Air
Percent Flue Gas Recirculation
Burners Out of Service
Damper Positions
Burner Tilt
Flowrates:
Superheater Steam
Reneater Steam
SH Attemperator Spray
RH Attemperator Spray
Airflow
Fuel Flow
Pressures:
Steam Drum
SH Steam Outlet
RH Steam Outlet
Furnace
Windbox
Fan Inlet
Fan Discharge
Temperatures:
Superheater Steam
Reheater Steam
Air Heater Air In/Out
Air Heater Gas In/Out
Furnace Gas Outlet
Stack Gas Inlet
Heat Absorption:
Furnace
Superheater
Reheater
Economizer
84
-------
TABLE 4-4. Concluded
Fan Power Consumption
Gas Emissions:
NOX
sox
Carbon Monoxide
Hydrocarbons
Polycyclic Organic Matter
Particulate Loading
Particulate Size Distribution
Ringleman Smoke Density
Carbon/Unburned Fuel Loss
Additional Factors Considered:
Corrosion Rates
Slagging and Fouling
Flame Instability
Furnace Vibration
Fan and Duct Vibrations
85
-------
Wherever possible, comparisons of baseline and controlled operation
were made on tests which were similar in the general operating
characteristics tested. Steam flow and load conditions, overall excess
air levels, furnace conditions, etc., were matched for the baseline and
controlled tests selected for comparison.
In certain tests, where process data were sufficiently detailed,
overall mass and energy balances were conducted. The mass balances were
used to determine the amount of gaseous pollutants and particulate and
solid matter emitted by the boiler under baseline and low NOX
conditions. Overall energy balances were used to check boiler
efficiencies. Energy balances on individual boiler components established
the distribution of heat absorption in the boiler. Attemperator spray
flowrates were checked by heat and mass balances on superheater and
reheater sections. Air and gas volume flowrates were calculated to
determine the effect of changed operating conditions on fan draft and
power requirements.
For coal-fired tests, data were collected on carbon loss in flyash,
furnace slagging, and water wall tube corrosion. Data were also obtained
from some tests on coal- and oil-fired boilers on particle loading and
size distribution. Some data, mainly for oil and gas fuels, were also
available on flame stability, furnace vibrations, superheater tube
temperatures, and flame carryover to the convective section.
Comparisons of the process data were made for baseline and low
NOX modes of operation. Significant changes in the process variables
were noted and evaluated for their impact on emissions and boiler
operation and maintenance. The results from applying this analysis to the
utility boiler data on units noted in Table 4-3 are discussed in
Section 7-2.
4.2.3 Cost Analysis Procedures
Representative control costs were generated for typical
boiler/control combinations using regulated utility economics. First,
typical boilers were identified using the EPA's Energy Data System
(Reference 4-9). The cases selected were a tangential coal-fired unit to
power a 225 MW turbine generator, a 540 MW opposed wall coal-fired unit,
and a 90 MW front wall oil- and gas-fired unit. Primary considerations in
making these selections included:
The trend toward coal firing, particularly in larger size
units, emphasizes tangential and opposed wall firing designs
t Many units are capable of burning both oil and gas, especially
in the smaller size ranges. Single wall (front or rear) fired
units are common in this application.
The average control cost on a per unit output basis is not a
strong function of unit size. Hence a representative unit size
was judged adequate.
86
-------
Preliminary engineering designs of the NOX controls that would be
required for the selected boilers were prepared. This design work
provided an estimate of the hardware and installation requirements for
applying retrofit controls. Up to date vendor quotations were obtained.
In addition, the design effort permitted estimating the actual engineering
time required for implementing controls.
The cost analysis was based on the annualized revenue approach,
adapted from that used by the Tennessee Valley Authority in evaluating the
cost of powerplant projects for EPA (Reference 4-10) and EPRI (Reference
4-11). Details of the cost analysis algorithm adopted are given in
Reference 4-8.
The use of accepted estimation procedures for costing NOX control
implementation in current dollars was employed, with heavy reliance on
discussions with boiler manufacturers, equipment vendors, and utilities.
Use of the annualized cost methodology then permitted a systematic, well
documented, up to date cost analysis of typical controls for
representative boiler design/fuel classifications. In this manner, the
cost effectiveness of controls was compared from boiler to boiler on a
consistent basis.
Results from applying this procedure to costing NOX controls for
the aforementioned utility boiler cases are given in Section 7.3.
4.3 SYSTEMS ANALYSIS METHODS
The purpose of the systems analysis is to provide a quantitative
basis for identifying the future needs (when, where, how much, and what
kind) for NOX controls to satisfy the requirements of the Clean Air
Act. This information will be used in the program to recommend R&D
directions and schedules for developing necessary controls.
In the systems analysis, uncontrolled emissions projections,
controls cost and effectiveness data, fuel costs, and ambient air quality
goals are combined to evaluate the control needs for a particular Air
Quality Control Region (AQCR). The elements of the systems analysis model
developed are shown in Figure 4-1. The specific air quality issues which
must be assessed were discussed in Section 3. In this section, the
methods used for the systems analysis are briefly described. Results of
applying the analysis are presented in Section 8.2.
4.3.1 Preliminary Model
The most critical element in the systems analysis is the air
quality model. Candidate models differ not only in their degrees of
sophistication, but also in their resolution and versatility. Usually,
the sophisticated models require more elaborate input data than the
simpler models, a significant amount of calibration, and considerable
experience to use them intelligently. On the other hand, the simpler
models, which try to model the atmospheric processes in an integral
manner, are based on many correlations of the available data and lack the
resolution of the sophisticated models.
87
-------
AIR QUALITY
MODEL
PRIORITIZATION
OF
CONTROLS
CONTROL REQUIREMENTS
CONTROL COSTS
AMBIENT AIR QUALITY
Figure 4-1. Elements of the systems analysis model
88
-------
During the first year of the NOX EA, a systems analysis model was
needed to provide a preliminary priority ranking of control methods. A
modified form of rollback was selected to reduce the amount of emission
data needed, minimize computational costs, and provide maximum flexibility
in the initial phases of the analysis. Furthermore, only the NOX/N02
relationship was considered, and thus, HC emissions data did not need to
be collected.
This same model has also been used during the past year as a
screening method applied to a wide variety of AQCR's. Approximately 30
AQCR's have been identified (Reference 4-12) as exceeding or possibly
exceeding the annual average N02 standard between now and 1985 (see
Section 3.1). The flexibility and minimum data requirements of the
preliminary model allowed us to examine 8 representative AQCR's for each
of over 20 different emissions/control scenarios.
The rollback model used is given by:
AC = k Z. (1 - R.) E.W.) + BG
where AC = ambient concentration (N0ฃ)
Ei = uncontrolled emissions from source i
R-J = fractional emissions reduction by control of source i
W-j = weighting factor for source i
BG = background concentration (the background concentration has been
assumed to be 10 yg/m3 for all cases)
The calibration constant, k, is determined by evaluating the equation at
some "base year" for which the ambient concentration and emissions are
known (R-j = 0).
Although factors such as stack height and relative position of
source and receptor are not explicitly included in this model, they are
implicitly included because the model is essentially a correlation between
existing emission patterns and the resulting ambient air quality
conditions. Moreover, in the formulation employed it is possible to
specify the relative importance of each source type by using the weighting
factors. For example, in an AQCR characterized by large mixing heights,
emissions from elevated sources are widely dispersed and, therefore, do
not have the same impact on ground level concentration as the same amount
of ground level emissions. Thus, a source weighting factor less than 1.0
could be assigned to the elevated sources (e.g., powerplants) to account
for the effects of stack height. (Each choice of weighting factors is
equivalent to choosing a different model for the AQCR. In all cases the
model must be calibrated for the base year before future year projections
are made.)
89
-------
The utility of the preliminary model has been further increased by
testing the sensitivity of the results to the input values. This ensured
that the predicted control requirements would be responsive to the
majority of NOX critical situations which might develop. Control
strategies were developed for numerous combinations of stationary and
mobile source growth, base year calibration, and source weighting factors
for each AQCR.
Growth scenarios that represented reasonable bounds for both mobile
and stationary sources were selected. Generally, growth rates apply to an
end use sector such as industrial or residential; however, in this
analysis they have been extended to each source within the sector.
Whenever possible, growth rates specific to an AQCR were used. If
specific AQCR rates were not available, state, regional, or national rates
were used. In addition, the influence of population growth and any local
limitation on new source growth were considered. Two basic scenarios were
selected for stationary sources. One case represented a moderately
conservative growth influenced by conservation measures and rising energy
costs. The other represented a higher growth rate closer to historical
patterns. This case represented a reasonable upper boundary on stationary
source growth.
The growth rates of emissions from mobile sources were treated
differently, since a detailed investigation of mobile source control
options was not of direct interest to this study. However, the emissions
contributions of the mobile sources were needed; therefore, two
representative scenarios were used. One scenario (the nominal case) was
selected to reflect historical growth in vehicle population and miles
traveled, as well as the current mobile emission standards* (0.62 g/km for
light duty vehicles in 1981). The alternate, or low, case was for a
reduced growth rate (closer to the population growth rate) and an emission
standard reflecting the research goal of 0.25 g/km in 1985 for light duty
vehicles. This case was selected to represent the most optimistic mobile
emissions scenario, which is least demanding of stationary source control.
Two values of the N02 annaal average ambient concentration were
selected for each AQCR for the base year calibration. These reflected the
high and low of the AQCR maximum annual average recorded from 1972 to
1975, or, in some cases, a reasonable variation in the recorded or
estimated maximum annual average over the same time period.
The preliminary model has proven to be very useful for its intended
purpose preliminary priority ranking of NOX control needs based only
on consideration of N02 air quality. Several conclusions derived from
model calculations are discussed in Section 8.2. However, more
sophisticated models, as outlined in the next section, are needed to
explore the many complexities of the N02 air quality problem.
*For California AQCR's the current California schedule for mobile source
emission standards was used.
90
-------
4.3.2 Advanced Models
As the preliminary modeling efforts progressed it became
increasingly apparent that there were many questions regarding the
air quality problem that could not be answered by the simple air quality
models such as modified rollback. This was further accented by increased
interest in the impact of a one hour N02 standard on the need for NOX
controls. (Most of the issues of interest concerning a one hour standard
were discussed in Section 3.) To meet the needs of a more sophisticated
analysis in a variety of AQCR's at a realistic cost it was decided to use
or extend the results of previous analyses.
Nonreacting dispersion modeling had been done for the Chicago and
Baltimore AQCR's. In addition, detailed photochemical modeling had been
done for Los Angeles, San Francisco and Denver, and was planned for
St. Louis. In all cases, this meant that emissions and meteorology data
bases had already been created, and appropriate computer models were
operational. The results of these air quality modeling efforts were thus
used to verify the assumptions of the rollback model, to guide the
modification of the air quality model in the systems analysis, and to
examine specific emission/air quality issues. Major advantages of this
approach are the ability to include specific source/receptor
relationships, meteorology, mixing, kinetic reactions, and the
HC/NOX/N02/03 interactions.
Efforts to date have concentrated on two models and two AQCR's.
The LIRAQ model, developed by the Lawrence Livermore Laboratory, has been
applied to the San Francisco AQCR and an advanced version of the DIFKIN
model, developed by ERT, to Los Angeles. These two models are briefly
described below.
4.3.2.1 ERT Model
The ERT model, a successor to DIFKIN, is a photochemical,
Lagrangian trajectory model which tracks an air mass throughout the region
of interest based on a prescribed wind field. The model considers
advection of pollutants (motion relative to the air mass being tracked)
and assumes horizontal diffusion of pollutants to be negligible. These
assumptions reduce the species continuity equations to a vertical
diffusion equation similar to the heat equation. Embedded sources in the
vertical cells represent both emissions (from elevated sources) and
chemical transformations. Sources at the lower boundary represent
emissions from area emitters as expressed by time and space varying flux
schedule boundary conditions. In the detailed chemical mechanism a
distinction is made among the effects of five classes of reactive
hydrocarbon substances: paraffins, olefins, aromatics, formaldehyde, and
higher aldehydes. Although a high degree of lumping of parameters occurs
within each class, distinguishing five separate categories permits
relatively specific treatment of different levels of reactivity.
Several trajectory analysis simulations were performed using this
model to address questions appropriate to the systems analysis effort.
91
-------
Model simulations performed, and their results, are briefly discussed in
Section 8.2.
4.3.2.2 LIRAQ Model
The LIRAQ model was developed by the Lawrence Livermore Laboratory
with the support of the National Science Foundation and in cooperation
with the Bay Area Air Pollution Control District (BAAPCD) (Reference
4-13). This model is a photochemical Eulerian model designed to treat
most of the important factors of interest in the San Francisco Bay Area.
The complex topography and changing meteorology are treated on one of
several available grid scales. Mass consistent windfields, based on real
or hypothetical meterological situations, are provided by an auxiliary
program. Reactive hydrocarbons are divided into three characteristic
types: alkenes, alkanes and aromatics, and aldehydes. The model computes
pollutant concentrations at all grid points in the region at each time
interval and gives resultant concentration contours for each hour.
Computational scenarios and results of calculations obtained using
this model are also discussed in Section 8.2.
4.3.2.3 Short Term - Annual Average Correlation
Both of the models described above provide results in the form of
one hour concentrations. To extrapolate such results to an impact on an
annual average N02 level, some relationship between annual average and
one hour values is necessary. This is usually provided by the
concentration frequency data from monitoring stations in the region of
interest. In both San Francisco and Los Angeles the concentration
frequency data were found to be approximately log-normal. The implication
of this is that the same percentage change calculated for one hour values
can also be applied to an annual average. This assumes that the slope (on
a log probability scale) does not change as a result of the control
strategy.
There are no direct data to support the above assumption; however,
comparison of data from several monitoring stations in Los Angeles for the
period 1970 to 1973 indicated a relatively constant slope. This is
slightly misleading since only minor changes in the emissions patterns
occurred during this time period. Furthermore, as will be shown in
Section 8.2, the response of the N02 one hour peak and the 24 hour
average (a better measure of the long term average response) may be very
different depending on both the NOX and HC emissions changes. It does
appear that for those cases where the ratio of HC to NOX emissions
remains relatively constant (i.e., simultaneous control of NOX and HC)
the one hour and 24 hour average values change by approximately the same
percentage. Therefore, a constant relationship between the N02 one hour
and annual average, subject to simultaneous NOX and HC control was
assumed.
92
-------
REFERENCES FOR SECTION 4
4-1. Cleland, J.G., and G.L. Kingsbury, "Multimedia Environmental Goals
for Environmental Assessment," Volumes I and II,
EPA-600/7-77-136a,b, NTIS PB-276 919/AS, November 1977.
4-2. Schalit, L.M., and K.J. Wolfe, "SAM IA: A Rapid Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
EPA-600/7-78-015, NTIS PB-277 088/AS, February 1978.
4-3. Duke, K.M., et a/L, "IERL-RTP Procedures Manual: Level 1
Environmental Assessment Biological Tests for Pilot Studies,"
EPA-600/7-77-043, NTIS PB-268 484/3BE, April 1977.
4-4. Anderson, L.B., et _ง_]_., "SAM I: An Intermediate Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
Acurex Draft Report TR-79-154, Acurex Corporation, Mountain View,
CA, December 1978.
4-5. Turner, D.B., "Workbook of Atmospheric Dispersion Estimates," U.S.
Public Health Service, AP-26, 1970.
4-6. Holzworth G., "Mixing Heights, Wind Speeds, and Potential for Urban
Air Pollution Throughout the Contiguous United States," Office of
Air Programs, U.S. EPA, January 1972.
4-7. Salvesen, K.G., et_ al., "Emissions Characterization of Stationary
NOX Sources," EPA-600~/7-78-120a,b, NTIS PB-284 520, June 1978.
4-8. Lim, K.J., ejt afL, "Environmental Assessment of Utility Boiler
Combustion Modification NOX Controls," Acurex Draft Report
TR-78-105, April 1978.
4-9. Energy Data System, U.S. Environmental Protection Agency, Office of
Air and Waste Management, Office of Air Quality Planning and
Standards, Strategies and Air Standards Division.
4-10. McGlamery, G.G., et aJL, "Detailed Cost Estimates for Advanced
Effluent Desulfurization Processes," EPA-600/2-75-006,
NTIS PB-242 541/AS, January 1975.
4-11. Waitzman, D.A., et ajk, "Evaluation of Fixed-Bed, Low-Btu Coal
Gasification Systems for Retrofitting Power Plants," EPRI Report
No. 203-1, February 1975.
4-12. Water land, L. R., et al, "Environmental Assessment of Stationary
Source NOX Control~Tech~Viologies -- First Annual Report,"
EPA-600/7-78-046, NTIS PB-279 083/AS, March 1978.
4-13. MacCracken, M.C., and G.D. Sauter, "Development of an Air Pollution
Model for the San Francisco Bay Area," Lawrence Livermore
Laboratory Report No. UCRL-51920, Volume 1, NTIS N76 33720,
October 1975.
93
-------
SECTION 5
ENVIRONMENTAL DATA ACQUISITION
This section describes the updated multimedia emissions inventory
compiled to support the source impact ranking and environmental impact
assessment efforts in the NOX EA. Emissions data available as of 1976
were previously discussed in the Preliminary Environmental Assessment
report (Reference 5-1) and summarized in the NOX EA first annual report
(Reference 5-2). The results presented herein incorporate additional data
and augment the earlier work with projections. These updated results are
discussed more fully in Reference 5-3. Updating of the inventories
assembled will continue in the third year.
Based on the emission inventory work, together with preliminary
process analysis and environmental assessment efforts, numerous emissions
data gaps were identified in the inventories. To address these a field
test program was defined and is now underway. This test program,
developed with emphasis on clarifying the incremental effects of NOX
control application on pollutant emissions other than NOX, is also
described in the following.
5.1 BASELINE EMISSIONS
The national and regional multimedia emissions inventories are
presented below for the stationary NOX sources and fuels identified in
Section 2.
5.1.1 National Baseline Emissions Inventory
A baseline multimedia emissions inventory was produced for all
significant stationary NOX sources. This inventory was then extended to
include all other sources of NOX (mobile, noncombustion, fugitive) to
compare emissions from stationary combustion sources with those from other
sources. Multimedia pollutants inventoried included the criteria
pollutants (NOX, SOX, CO, HC, particulates), sulfates, polycyclic
organic matter (POM), trace metals, and liquid and solid effluents.
This inventory was compiled to provide the basis for weighing the
incremental emissions impact of using NOX controls. In addition, the
inventory also serves as a reference for projections to the year 2000 for
anticipated trends in fuels, equipment, and stationary source emissions.
94
-------
Data gaps identified in compiling the emission factors highlight areas
where further testing is needed.
The emissions inventory was performed in the following sequence:
Compile fuel consumption data for the categories of combustion
sources specified in Section 2. Subdivide fuel consumption
data based on fuel bound pollutant precursor composition.
Compile multimedia emission data
-- Base fuel dependent pollutant emission factors on the trace
composition of fuels
Base combustion dependent pollutant emission factors on
unit fuel consumption for specific equipment designs
Survey the degree to which NOX, SOX, particulates are
controlled
t Produce emissions inventory
Rank sources according to emission rates
Although detailed breakdowns of fuel consumption, emission factors,
and total emissions for each equipment/fuel combination were developed,
only emission totals for each sector are summarized here.
The distribution of anthropogenic NOX emissions is shown in
Figure 5-1 for the year 1974. The estimates of utility boiler emissions
account for the reduction resulting from using NOX controls. From a
survey of boilers in areas with NOX emission regulations, it was
estimated that using NOX controls in 1974 resulted in a 3.0 percent
reduction in nationwide utility boiler emissions. This corresponds to a
1.6 percent reduction in stationary fuel combustion emissions. Reductions
from using controls on other sources were negligible in 1974.
Stationary source NOX emissions are subdivided by sector and fuel
type in Table 5-1. The emission inventory summaries for other pollutants
are shown in Table 5-2.
Data for the criteria pollutants were generally good and the
results of these current inventories are in reasonable agreement with
other recent inventories. Data for the noncriteria pollutants and liquid
or solid effluent streams, however, were sparse and scattered. For
example, emission factors for POM varied by as much as two orders of
magnitude; thus, Table 5-2 shows a range for total POM emissions.
Table 5-3 ranks equipment/fuel combinations by annual, nationwide
NOX emissions, and lists corresponding rankings for these combinations
by fuel consumption and emissions of criteria pollutants. Although there
were over 70 equipment/fuel combinations inventoried, the 30 most
significant combinations account for over 80 percent of NOX emissions.
The ranking of a specific equipment/fuel type depends both on total
95
-------
Noncombustion 0.9%
Fugitive 2.3%
Incineration 0.2%
Stationary fuel combustion
52.1%
Mobile sources
44.5%
Stationary Fuel Combustion
Fugitive Emissions
Noncombustion
Incineration
Mobile Sources
Total
eg
11,297
498
193
40
9,630
21,658
1,000 Ton
12,437
548
212
44
10,600
23,841
Percent Total
(52.1)
(2.3)
(0.9)
(0.2)
(44.5)
100
Figure 5-1. Distribution of anthropogenic NOX emissions for the year
1974.
96
-------
TABLE 5-1. SUMMARY OF 1974 STATIONARY SOURCE NOY EMISSIONS BY FUEL TYPE
ซ
Sector
Utility Boilers
Packaged Boilersb
Warm Air Furnaces
Gas Turbines
Reciprocating 1C
Engines
Industrial Process
Heating
Non combust ion
Incineration
Fugitive
Total
NOX Production Gg/yr Total by
(% of total) Sector
Coal
3808
(31.7)
781
(6.5)
--
4589
(38.2)
Oil
848
(7.0)
886
(7.4)
130
(1.1)
308
(2.6)
457C
(3.8)
--
2629
(21.8)
Gas
1152
(9.6)
779
(6.5)
190
(1-6)
132
(1.1)
140Qd
(11.6)
--
3653
(30.4)
(% of total)
5808
(48.3)
2446
(20.3)
320
(2.7)
440
(3.7)
1857
(15.4)
426
(3.5)
193
(1.6)
40
(0.3)
498
(4.2)
12,028
(100.0)
Cumulative
(*)
48.3
68.6
71.3
75.0
90.4
93.9
95.5
95.8
100.0
3N02 basis
^Includes steam and hot water commercial and residential heating units
clncludes gasoline
^Includes dual fuels (oil and gas)
97
-------
TABLE 5-2. SUMMARY OF AIR AND SOLID POLLUTANT EMISSIONS FROM STATIONARY
FUEL BURNING EQUIPMENT (Gg/yr)
CO
Sector
Utility Boilers
Packaged Boilers
Warm Air Furnaces
& Misc. Comb.
Gas Turbines
Recip. 1C Engines
Process Heating
TOTAL
N0xb
5808
2446
320
440
1857
426
11,297
sox
16,768
6,405
232
10.5
19.6
622
24,057
HC
29.5
72.1
29.7
13.7
578
166
889
CO
270
175
132.6
73.4
1,824
9,079
11,554
Part.
5,965
4,930
39.3
17.3
21.5
4,766
15.739
Sul fates
231
146
6.4
a
a
a
383
POM
0.01 - 1.2
0.2 - 67.8
0.06
a
a
a
69
Dryc
Ash Removal
6.2
1.1
7.3
Sluiced0
Ash Removal
24.8
4.4
--
29.2
t
No emission factor available
Controlled NOX> N02 basis
cBased on 80 percent hopper and flyash removal by sluicing methods; 20 percent dry solid removal
-------
TABLE 5-3. NOX MASS EMISSION RANKING OF STATIONARY COMBUSTION
EQUIPMENT AND CRITERIA POLLUTANT AND FUEL USE CROSS RANKING
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Sector
Utility Boilers
Reciprocating 1C
Engines
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Reciprocating 1C
Engines
Utility Boilers
Packaged Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Packaged Boilers
Equipment Type
Tangential
>75 kW/cylc
Wall Firing
Cyclone
Wall Firing
Wall Firing
Horizontally Opposed
75 kW to 75 kW/cylc
Horizontally Opposed
Wall Firing WTd >29 MW^
Stoker Firing WTd <29 MW&
Wall Firing WT29 MW*>
Tangential
Scotch FTe
Single Burner WTd <29 MW&
Horizontally Opposed
Single Burner WTd <29 MWb
Fuel
Coal
Gas
Coal
Coal
Gas
Oil
Gas
Oil
Coal
Gas
Coal
Oil
Oil
Oil
Gas
Oil
Coal
Annual NOX
Emissions
(Mg)
1,410,000
1,262,000
1,137,000
848,300
646,800
458,300
352,200
325,000
324,500
318,500
278,170
232,480
205,100
203,990
180,000
168,900
164,220
Cumulative
(Mg)
1,410,000
2,672,000
3,809,000
4,657,300
5,304,100
5,762,400
6,114,600
6,439,600
6,764,100
7,082,600
7,360,770
7,593,250
7,798,350
8,002,250
8,182,250
8,351,150
8,515,370
Cumulative
(Percent)
11.7
22.2
31.7
38.7
44.1
47.9
50.8
53.5
56.2
58.9
61.2
63.1
64.8
66.5
68.0
69.4
70.8
Fuel
Rank
1
21
3
6
4
8
14
>30
23
16
7
26
12
11
5
>30
>30
SOX
Rank
1
>30
2
3
>30
9
>30
>30
5
>30
4
16
10
11
>30
17
8
CO
Rank
7
4
6
12
13
17
24
3
>30
29
11
>30
27
>30
>30
>30
>30
HC
Rank
16
1
23
9
28
27
>30
3
>30
19
4
26
>30
>30
22
>30
>30
Part.
Rank
2
>30
5
13
>30
18
>30
26
7
>30
1
22
19
16
>30
27
9
aN02 basis
bHeat input
cHeat output
dWatertube
eFiretube
EE-007
-------
TABLE 5-3. (Concluded)
Rank
18
19
20
21
22
23
24
25
26
27
28
29
30
Sector
Industrial
Process Comb.
Packaged Boilers
Utility Boilers
Packaged Boilers
Gas Turbines
Packaged Boilers
Warm Air Furnaces
Packaged Boilers
Packaged Boilers
Gas Turbines
Reciprocating 1C
Engines
Industrial
Process Comb.
Utility Boilers
Equipment Type
Refinery Heaters
Forced & Natural Draft
Firebox FTe
Tangential
Stoker Firing WTd
4 to 15 MWC
Single Burner WTd <29 MW&
Central
Stoker Firing FTe <29 MWฐ
Scotch FTe
>J5 MWC
>75 kW/cylc
Refinery Heaters
Forced & Natural Draft
Vertical and Stoker
Fuel
Oil
Oil
Gas
Coal
Oil
Oil
Gas
Coal
Gas
Oil
Oil
Gas
Coal
Annual NOX
Emissions
(Mg)
147,350
139,260
137,900
125,350
118,500
116,430
106,300
102,040
98,010
97,400
94,000
92,608
88,500
Cumulative
(Mg)
8,662,720
8,801,980
8,939,880
9,065,230
9,183,730
9,300,160
9,406,460
9,508,500
9,606,510
9,703,910
9,797,910
9,890,518
9,979,018
Cumulative
(Percent)
72.0
73.1
74.3
75.3
76.3
77.3
78.2
79.0
79.8
80.6
81.4
82.2
82.9
Fuel
Rank
>30
17
13
>30
30
27
2
29
19
>30
>30
15
>30
SOX
Rank
29
13
>30
7
>30
15
>30
6
>30
>30
>30
>30
12
CO
Rank
>30
>30
>30
28
15
>30
10
>30
>30
>30
22
>30
>20
HC
Rank
18
>30
>30
29
14
>30
8
10
>30
30
13
7
530
Part.
Rank
21
20
>30
8
>30
23
25
6
>30
>30
>30
30
10
o
o
basis
bHeat input
CHeat output
dWatertube
eFiretube
EE-007
-------
installed capacity and emission factors. A high ranking, therefore, does
not necessarily imply that a given source is a high emitter; large
installed capacity may offset a low emission factor to give the high
ranking. In general, coal-fired sources rank high in SOX and
particulate emissions, while 1C engines rank high in emissions of CO and
hydrocarbons.
As noted above, inventory results presented are for 1974 data, the
most recent when the effort was initiated. However, future NOX EA
efforts will update this national emissions inventory to 1977 using
improved emission factors, 1977 utility boiler fuel consumption, and
updated fuel projection data for other equipment.
5.1.2 Projected National Emissions Inventories
Emissions inventories assembled for the year 2000 for NOX are
presented here for the two reference scenarios described in Section 2.2.
These emissions inventories are a culmination of the projected fuel
consumption data presented in Section 2.2 and control projections.
Tables 5-4 and 5-5 summarize total NOX emissions from fuel user
sources for the year 2000 for the reference scenarios. The NOX mass
emissions ranking of stationary combustion equipment is presented in
Table 5-6 for the year 2000 high nuclear reference scenario. Tangential
boilers appear to be the most significant NOX source through 2000 if
projected trends are realized. Coal-fired sources should increase their
share of NOX emissions and dominate the highest rankings. Natural
gas-fired sources show lower NOX emissions rankings due to decreased
fuel consumption and implementation of controls. In 2000, the highest
natural gas source is tenth on the ranking, compared to second in 1974.
Oil-fired sources also show a gradual decrease in NOX emissions due to
their attrition and replacement with coal-fired sources.
5.1.3 Regional Emissions Inventories
This section presents regional emissions inventories for combustion
related pollutants from stationary sources. Figure 5-2 shows the
distribution of fuels by region in 1974. This distribution formed the
basis for the regional emissions inventory generated. The figure shows
that oil is the major fuel used in the East Coast region. The West Coast
and Southwest are supplied largely by natural gas, and the Midwest relies
primarily on coal for its fossil fuel requirements. Table 5-7 summarizes
NOX emissions for the nine regions shown in Figure 5-2. These estimates
are for uncontrolled NOX only since the impact of NOX control
implementation on a regional basis was small in 1974.
Over 40 percent of all NOX emissions from utility boilers are
from the East-North-Central and the South Atlantic regions. The New
England region accounts for less than 5 percent of utility boiler NOX
emissions. However, areas such as New England and the Far West may be
most strongly affected by fuel switching to coal since they are heavily
101
-------
TABLE 5-4. SUMMARY OF ANNUAL NOX EMISSIONS FROM FUEL USER SOURCES
(2000): REFERENCE SCENARIO -- LOW NUCLEAR
Sector
Utility Boilers
Packaged Boilers
Warm Air Furnaces
Gas Turbines
Reciprocating
1C Engines
Process Heating
Noncombustion
Incineration
Total by Fuels
NOX Production Gg
(% of Total)
Gas
657.2
(4.84)
178.8
(1.32)
156.9
(1.15)
248.3
(1.83)
--
1,241.2
(9.13)
Coal
7,951.9
(58.51)
898.2
(6.61)
8,850.1
(65.12)
Oil
763.5
(5.62)
1,064.4
(7.83)
124.4
(0.92)
249.7
(1.84)
610.3
(4.49)
2,812.2
(20.69)
Total
By Sector Gg
(% of Total)
8,715.3
(64.13)
2,619.8
(19.28)
303.1
(2.23)
406.5
(2.99)
858.6
(6.32)
289.5
(2.13)
322.0
(2.37)
76.0
(0.56)
13,590.8
Cumulative
(%)
64.13
83.40
85.63
88.62
94.94
97.07
99.44
100.0
aN02 basis
102
-------
TABLE 5-5. SUMMARY OF ANNUAL NOX EMISSIONS FROM FUEL USER SOURCES
(2000): REFERENCE SCENARIO -- HIGH NUCLEAR
Sector
Utility Boilers
Packaged Boilers
Warm Air Furnaces
Gas Turbines
Reciprocating
1C Engines
Process Heating
Noncombustion
Incineration
Total by Fuels
NOX Production -- Gg
(% of Total)
Gas
657.2
(6.22)
178.8
(1.69)
156.9
(1.49)
248.3
(2.35)
--
1,241.2
(11.75)
Coal
5,197.0
(49.21)
622.1
(5.89)
5,819.0
(55.10)
Oil
763.5
(7.23)
1,064.4
(10.08)
124.4
(1.18)
249.7
(2.36)
610.3
(5.78)
--
2,812.2
(26.63)
Total
By Sector -- Gg
(% of Total)
5,960.5
(56.44)
2,343.7
(22.19)
303.1
(2.87)
406.5
(3.85)
858.6
(8.13)
289.5
(2.74)
322.0
(3.05)
76.0
(0.72)
10,599.9
Cumulative
(%}
56.44
78.64
81.51
85.36
93.49
96.23
99.28
100.0
aN02 basis
103
-------
TABLE 5-6. YEAR 2000 NOX MASS EMISSIONS RANKING FOR STATIONARY COMBUSTION
EQUIPMENT AND CRITERIA POLLUTANT CROSS RANKING
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Sector
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Utility Boilers
Packaged Boilers
Packaged Boilers
Reciprocating 1C
Engines
Reciprocating 1C
Engines
Packaged Boilers
Packaged Boilers
Packaged Boilers
Reciprocating 1C
Engines
Packaged Boilers
Equipment Type
Tangential
Wall Firing
Horizontally Opposed
Cyclone
Wall Firing
Scotch FTd <29 MWป
Tangential
Stoker Firing WTC<29 MWป
Wall Firing WK > 29 MWa
Clf 75 kW to 75 kW/cylb
Sie >75 kW/cylb
Wall Firing WTC>29 MWป
Single Burner WT<29 MW*
Firebox FT75 kW/cylb
Single Burner WTC<29 MWa
Fuel
Coal
Coal
Coal
Coal
Oil
Oil
Oil
Coal
Oil
Oil
Gas
Gas
Gas
Oil
Oil
Oil
Annual
NOX Emissions
(Mg)
2,586,100
1,634,800
472,400
450,300
378,100
267,500
236,200
221,600
212,700
202,600
201,800
189,300
184,700
184,300
161,200
150,200
aHeat input
bHeat output
cWatertube
dFiretube
eSpark ignition
'Compression ignition
104
-------
TABLE 5-6. Concluded
Rank
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Sector
Utility Boilers
Gas Turbines
Reciprocating 1C
Engines
Packaged Boilers
Gas Turbines
Packaged Boilers
Reciprocating 1C
Engines
Packaged Boilers
Packaged Boilers
Warm Air Furnaces
Packaged Boilers
Gas Turbines
Packaged Boilers
Warm A*ir Furnaces
Equipment Type
Horizontally Opposed
Simple Cycle >15 MWb
Sie 75 kW to 75 kW/c.ylb
Wall Firing WTC >29 MWa
Simple Cycle 4 MW to 15 MWb
HRT Boiler
CIf >75 kW/cylb
Scotch FTd <29 MWa
Stoker WTC >29 MWa
Warm Air Central Furnace
Firebox FTb < 29 MWซ
Simple Cycle >15 MWb
Single Burner WT < 29 MWa
Warm Air Central Furnace
Fuel
Oil
Oil
Gas
Coal
Oil
Oil
Dual
(Oil and Gas)
Gas
Coal
Gas
Gas
Gas
Coal
Oil
Annual
NOX Emissions
(Mg)
139,400
137,700
136.900
130,700
114,600
113,900
110,500
100,500
99,800
95,600
93,100
99,800
85,200
81,900
aHeat input
bHeat output
cWatertube
' dFiretube
eSpark ignition
^Compression ignition
105
-------
43% C
18% 0
39% G
28% C
45% 0
27% G
SOUTH
ATLANTIC
?ฃฃ&? EAST2
36% C
39% 0
25% G
6% C
77% 0
17% G
(Except South Atlantic,
where oil represents 39ฐฃ
total fuel consumption.)
Figure 5-2. Regional fuel distributions.
-------
TABLE 5-7. DISTRIBUTION OF REGIONAL UNCONTROLLED NO a EMISSIONS
(Gg/yr) ซ 1974
Sector and Equipment
Type
Utility Boilers
Tangential
Wall Fired
Horizontally
Opposed
Cyclone
Vertical and Stoker
Subtotal
Packaged Boilers
Commercial and
Residential Furnaces
Gas Turbines
1C Engines
Process Heating
Subtotal
Total
Fuel
Coal
Oil
Gas
Coal
Oil
Gas
Coal
Oil
Gas
Coal
Oil
Gas
Coal
All
All
All
All
All
All
All
New
England
7.5
30.1
0.4
6.0
67.5
1.8
1.7
58.5
2.0
1.5
2.4
0.1
0.5
180.0
146.3
9.5
131.0
11.7
0.5
299.0
479.0
Middle
Atlantic
161.7
54.8
1.8
130.1
122.6
8.4
37.1
35.7
4.5
97.4
4.3
0.2
10.1
668.7
372.2
31.2
66.8
60.7
63.8
594.7
1263.4
E-N-
Central
477.6
10.2
4.8
385.7
22.8
22.8
110.0
8.4
12.3
288.7
0.7
0.5
30.0
1374.5
621.3
65.5
19.3
248.5
87.6
1042.2
2416.7
W-N-
Central
132.8
1.4
14.1
107.3
3.0
66.4
30.5
1.1
36.2
80.2
0.1
1.4
8.4
482.9
180.4
22,7
36.7
360.6
25.4
625.8
1108.7
South
Atlantic
281.5
60.2
8.9
227.4
134.6
41.4
64.8
35.2
22.5
170.1
4.8
0.9
17.6
1069.9
412.6
56.5
33.8
79.7
18.2
600.8
1670.7
E-S-
Central
220.3
3.8
2.1
178.0
8.3
9.5
50.8
3.0
5.2
133.1
0.3
0.2
13.8
628.4 .
171.5
22.9
9.4
130.0
27.3
361.1
989.5
W-S-
Central
18.6
8.9
85.3
15.0
19.8
400.0
4.3
7.3
217.1
11.3
0.7
9.2
1.2
798.7
250.6
42.6
83.9
684.2
149.9
1211.2
2009.9
Mountain
97.8
4.4
8.2
78.9
9.8
38.4
22.5
3.6
20.9
59.1
0.4
0.9
6.1
351.0
96.1
25.4
52.3
206.9
2.9
383.6
734.6
Pacific
11.4
31.3
12.3
9.3
69.9
58.1
2.6
16.1
31.5
6.9
2.5
1.3
0.8
254.0
195.6
44.4
7.3
74.7
50.0
372.0
626.0
Total
1409.2
205.1
137.9
1137.7
458.3
646.8
324.3
168.9 '
352.2
848.3
16.2
14.7
88.5
5808.1
2446.6
320.7
440.5
1857.0
425.6
5490.4
11298.5
aN02 basis
T-861(a)
-------
dominated by oil and gas firing. The East-North-Central and South
Atlantic regions also account for over 40 percent of the NOX emissions
from packaged boilers. But, considering all stationary sources, the
East-North-Central and the West-South-Central regions of the nation
generate the highest levels of NOX representing about 40 percent of the
total emissions.
The regional inventories developed here show significant localized
variations of NOX emissions by fuel/equipment type. These variations
result from both the regional fuel mix variations and the distribution of
stationary source types. Thus, a national policy of NOX control must be
broad enough to encompass these regional variations in developing
strategies for future NOX emissions reductions.
5.2 EXPERIMENTAL TESTING
During compilation of the baseline emissions inventory discussed in
Section 5.1 and in the preliminary evaluation of the incremental effects
of NOX controls on pollutant emissions other than NOX (Reference 5-1),
it became apparent that data were lacking in several key areas. Most
noteworthy was the virtual absence of data on the effects of NOX
combustion controls on emission levels of noncriteria flue gas pollutants
and liquid and solid effluents. To address these data needs a field test
program was defined and is currently underway.
Based on the results of the preliminary source impact ranking
performed in the first year of the NOX EA (Reference 5-2) a series of 19
candidate field tests were identified. From this set of 19 potential
tests, 7 were selected and scheduled. A summary of these seven tests is
given in Table 5-8.
A prerequisite for selecting a test to be performed was that,
whenever possible, field testing was to be performed as a subcontracted
addition to planned or ongoing tests. This represented the most cost-
effective manner to obtain needed data. Thus, collaborating test
contractors are also listed in Table 5-8. Of course where add-on testing
was not feasible, new tests were initiated as indicated for two tests.
As noted in Table 5-8 the sampling program followed for each test
incorporated:
Continuous monitoring of flue gas NOX, S02, CO, C02, and
02
Flue gas Source Assessment Sampling System (SASS), EPA Method 5
particulate load, and EPA Method 8 (or equivalent) sulfur
species sampling; both upstream and downstream of the
particulate collector, if applicable
Flue gas grab sampling and onsite gas chromatographic analysis
for CI-GS hydrocarbons; both upstream and downstream of the
particulate collector, if applicable
108
-------
TABLE 5-8. NOX EA FIELD TEST PROGRAM
Source Category
Coal -fired
Utility Boiler
Coal-fired
Utility Boiler
Description
Kingston 16; 180 MM
tangential; twin
furnace, 12 burners/
furnace, 3 elevations;
cyclone, 2 ESP's for
par ticu late control
Crist 17, 500 MW
opposed wall fired; 24
burners, 3 elevations;
ESP for part icu late
control
Test Points
(Unit Operation)
Baseline
Biased Firing (2)
BOOS (2)
Baseline
BOOS (2)
Sampling Protocol
Continuous NOX, S02, CO,
C02, 02
Inlet to 1st ESP:
-- SASS
Method 5
- Method 8
Gas grab (Ci-C6 HC)
Outlet of 1st ESP:
~ SASS
Method 5
Method 8
Gas grab (Ci-Cs HC)
Bottom ash
Hopper ash (1st ESP,
cyclone)
Fuel
Operating data
Continuous NOX, S02, CO
C02, 02
ESP inlet
~ SASS
- Method 5
- Method 8
~ Gas grab (Ci-C6 HC)
ESP outlet
- SASS
Method 5
Method 8
Gas grab (Ci-Cs HC)
Bottom ash
ESP hopper ash
Fuel
Operating data
Bloassay
Test
Collaborator
TVA
Exxon
Status
Complete,
August 1977
Complete,
June 1978
-------
TABLE 5-8. Continued
Source Category
Oil-fired
Utility Boiler
Coal -fired
Industrial
Boiler
Coal -fired
Industrial
Boiler
Description
Moss Landing #6; 740 MW
opposed wall fired; 48
burners, 4 elevations
Traveling grate spreader
stoker, 38 kg/s
(300,000 Ib/hr);
ESP for part icu late
control; wet scrubber
for SOX control
Traveling grate
spreader stoker,
25 kg/s (200,000 Ib/hr)
ESP for part icu late
control
Test Points
(Unit Operation)
Baseline
FGR
F6R + OFA
Baseline
LEA + high OFA
Baseline
LEA + High OFA
Sampling Protocol
Continuous NOX, S02, CO,
C02, 02
SASS
Method 5
Method 8
Gas grab (Ci-Cfi HC)
Fuel
Operating data
Bioassay
Continuous NOX, S02, CO,
C02, 02
Boiler exit:
- SASS
- Method 5
Shell -Emeryville
~ Gas grab (Ci-Ce HC)
ESP outlet
SASS
- Method 5
-- Shell-Emeryville
Gas grab (CrC6 HC)
Bottom ash
Cyclone hopper ash
Fuel
Operating data
Continuous NOX, S02, CO,
C02, 02
Boiler exit:
- SASS
Method 5
She 11 -Emeryville
Gas grab (Cj-Cs HC)
ESP Outlet
SASS
Method 5
Shell -Emeryville
Gas grab (Ci-C6 HC)
Bottom ash
ESP hopper ash
Fuel
Operating data
Bioassay
Test
Collaborator
New test
start
KVB
KVB
Status
Complete,
September 1978
Complete,
October 1977
Complete,
February 1978
-------
TABLE 5-8. Concluded
Source Category
Oil-fired
Gas Turbine
Oil-fired
Residential
Heating Unit
Description
T.H. Wharton Station,
60 MW GE MS 7001 C
machine
Blue Ray low NOX
furnace, Medford,
New York
Test Points
(Unit Operation)
Baseline
Maximum water
injection
Continuous
Cycling
Sampling Protocol
Continuous NOX, SO?, CO
C02, 02
SASS
Method 5
Method 8
Fuel
Water
Operating data
Bioassay
Continuous NOX, S02, CO
C02, 02
SASS
Method 5
Method 8
Fuel
Test
Collaborator
General
Electric
New test
start with
EPA-RTP
Status
Complete,
April 1978
Complete,
November 1977
-------
t Bottom ash slurry sampling
Participate collector hopper ash (slurry) sampling
t Fuel and fuel additive, if applicable, sample collection
Operating data collection
Also, as noted in Table 5-8, the test program was conducted, as a minimum,
for at least two conditions of source operation: baseline (uncontrolled)
and low NOX operation. In several instances, operation at intermediate
levels of NOX control was tested. In addition, replicate testing was
performed in selected cases.
A key part of the test program involved close monitoring of source
operating data. This was done not only to ensure that test conditions
remained constant and representative of acceptable source operation over
the duration of sample collection, but also to provide the necessary input
to further process analysis efforts analogous to those described in
Section 6.
Subsequent laboratory chemical analyses of samples collected
generally followed IERL-RTP defined Level 1 procedures (References 5-4,
5-5). A specific exception dealt with liquid and solid sample trace *
element analysis. Here, instead of assaying for trace elements by spark
source mass spectroscopy, atomic absorption spectroscopy was employed to
determine the 23 more commonly occurring elements listed in Table 5-9.
Another exception dealt with organic analyses of flue gas (XAD-2 extract),
particulate and liquid/solid samples. Here the analyses were extended,
when feasible, to the determination of the 11 polycyclic organic compounds
(POM) listed in Table 5-10. Other minor exceptions were:
The SASS particulate combining scheme shown in Figure 5-3 was
employed to maximize the usefulness of analysis results
t Analyses for the ionic species listed in Table 5-11 were
performed using specific ion electrodes instead of test kits
The specific analysis procedures followed are indicated
schematically in Figures 5-3 through 5-6. Following these procedures the
Level 1 analysis data listed below can be obtained for each test point:
Continuous flue gas NOX, S02, CO, C02, and 02
Flue gas S02, $03, and speciated C\-CQ hydrocarbon
Flue gas particulate load and size distribution
Flue gas vapor phase trace element composition for the 23
elements listed in Table 5-9
112
-------
TABLE 5-9. ELEMENTAL ANALYSIS: SPECIES DETERMINED
Antimony (Sb)
Arsenic (As)
Barium (Ba)
Beryllium (Be)
Bismuth (Bi)
Boron (B)
Cadmium (Cd)
Chromium (Cr)
Cobalt (Co)
Copper (Cu)
Iron (Fe)
Lead (Pb)
Manganese (Mn)
Mercury (Hg)
Molybdenum (Mo)
Nickel (Ni)
Selenium (Se)
Tellurium (Te)
Thallium (Ti)
Tin (Sn)
Titanium (Ti)
Vanadium (V)
Zinc (Zn)
TABLE 5-10. POM ANALYSIS: SPECIES DETERMINED
Anthracene
Anthanthrene
Benz(a)anthracene
Benzo(g,h,i)pery1ene
Benzo(a)pyrene
Benzo(e)pyrene
Coronene
Fluoranthene
Phenanthrene
Perylene
Pyrene
TABLE 5-11. ANION ANALYSIS: SPECIES DETERMINED
Chloride (C1-)
Fluoride (F~)
Nitrate (N03~)
Cyanide (CN-)
Sulfate (S042-)
Ammonia (NH4+)
113
-------
HOtto.1 WAAw
1
DRป AND
WEIGH
1
DESICCATE
AND WEIGH
WASH
1
0*1 AND
WfJOH
1
DESICCATE
AND WHOM
1
DRY AMD
WtlOH
1
DESICCATE
AND WCKiH
1
DRt AND
WEIGH
1
DESICCATE
AND WEIGH
ACID OR
PAM
OlOiSTION
r^;
rss
ME
AC D OH
PAH*
nOIITION
|
DILUTE
10 cm u
TO
DM)
1
AA
ANAIVBU
Nt
D*.
El
TtON
1)
|
COMBINE
m
COMBINE ip
ANO ML1ER
PROCEED AS 1
INDICATED 1
II
1
AGIO OR
PARH
DtGESTlON
1
DILUTE TO
10 OR tDOM
|
AA
ANALYSIS
AMI
COUaiNE
EICESS FILlEf)
HI- ASM WITH
DRlCD MASH
tit
AปH
I
AClU DH
PAHM
DIOEBTION
AA
fit TEH
|
PARK
DIGESTION
|
FlLIEft
COMBINE
DILUTE 10
10 OH lODmi
HOMOQCMUt
<
MUAJMOCA
I
SOKHLiT
EXTRACTION
CHlCll
|
MIASUHI
VOLUMt
TCO
ANALTStt
Cr-Ca
COM0INC
COW
t-4gMMOViO
1
FARM AC4D
(MOUTtON
1
AA
ANALTIW
EMIRATE
ECCSMRV
K-DI
ono
m
COWBIME AND
UKABUM VOLUHE
TCO
ANALYSIS
Cr-C>ซ
-
GRAY
ALIQUOT
FORiH
OEPObll ON
SALl PLATE
1
ANAiTSlS
ALIQUOT
FOR LC Hi
1C
SLfARAOON
lil
AA ANAttSIS
Figure 5-3. Analysis scheme for SASS train samples.
-------
YES
GRAV
ALIQUOT
FOR LC
SOLVENT
EXCHANGE
TO HEXANE
LC
i r \ r
1234
XAD: ฃ.0.5mg(f)/m3]
ASH: [>1mg(f)/Kg]
(8)
I I I
567
NO
STOP
GRAV
XAD: [>0.5mg(f)/m3]
ASH:[>1mg(f)/Kg]
(8)
YES
LRMS
NO
STOP
Figure 5-4. LC separation scheme.
-------
300 MUD ซig
FOR AA
ANALtSlS
( iUO 1WM) H IN PROPORTION TO
THE SAMPLE WEIGHTS FOR ORGANIC ANALYSES
COMBtNE ASH IN PROPORTION TO
THE SAMPLE WEIGHTS FOR ORGANIC ANALYSE
IF THE WtiGHT IS ซ-10g.
USE ENURE SAMPLE FOR
ORGANIC ANALYSES OTHER-
WISE USt ^10 0 KEEPING
REMAiMCEFI FOR FUTURE
ORGANIC AHALYSCS
IF THE WEIGHT IS -^ 10 g
USE ENTIRE SAMPLE FOR
ORGANIC ANALYSES OTHER-
WISE USE -I0g KEEPING
REMAINDER FOR FUTURE
ORGANIC ANALYSES
Figure 5-5. SASS particulate sample combining scheme.
-------
SOLID OR
SOLIDS
PORTION OF
SLURRY
PARR OR
ACID DIGEST
ELEMENTAL
ANALYSIS
BY AA
SOXHLET
EXTRACTION
WITH CH2CI2
LC SEPARATION
INTO 8
FRACTIONS
ORGANIC
ANALYSIS
BY IR AND
LRMS
TCO
Cr-Cie
ORGANICS
ACID
LEACHATE
GENERATION
SOi"
NEUTRAL
CN
cr
Figure 5-6. Analysis scheme for liquid/solid samples.
-------
Flue gas < Cy organic composition in terms of seven compound
polarity fractions and flue gas POM composition for the 11 POM
species listed in Table 5-10
Particulate composition for the 23 elements listed in Table 5-9
and the six ionic species listed in Table 5-11, as a function
of particulate size
t Particulate organic composition for seven polarity fractions,
and for the 11 POM species listed in Table 5-10, as a function
of particulate size
t Liquid/solid stream (bottom, hopper ash) composition for the 23
elements listed in Table 5-9 and the six ionic species listed
in Table 5-11
Liquid/solid stream (bottom, hopper ash) organic composition
for seven polarity fractions and for the 11 POM species listed
in Table 5-10
t Particulate and ash C, H, 0, N, and S content
t Fuel proximate and ultimate analysis (heating value, and water,
C, H, 0, N, and S content)
t Fuel trace element content for the 23 elements listed in
Table 5-9
The above data satisfy the specific needs identified in earlier
NOX EA efforts (Reference 5-2). Specific attention was focused on
obtaining data on emitted POM, $03 and condensed sulfate, and trace
element levels as a function of particulate size, especially as these are
affected by NOX control applications.
In addition to the chemical analysis program, bioassay testing in
accordance with IERL-RTP guidelines (Reference 5-6) will also be performed
on samples collected during the gas turbine, oil-fired utility boiler,
second coal-fired utility boiler, and second coal-fired industrial stoker
tests. The general bioassay protocol to be followed is indicated in
Table 5-12.
As Table 5-8 indicated, all seven planned tests have been
completed. Sample chemical analyses, bioassay testing, and test data
reduction are currently underway and will be available in the near future.
118
-------
TABLE 5-12. BIOASSAY ANALYSIS PROTOCOL
Sample Type
Bioassay Test Protocol
Sample Size
Requirements
SASS cyclones,
SASS cyclones,
ly + filter
XAD-2 extract
Bottom ash
ESP Hopper ash
Microbial Mutagenesis
Cytotoxicity, RAM
Microbial Mutagenesis
Cytotoxicity, RAM
Microbial Mutagenesis
Cytotoxicity, WI-38
Microbial Mutagenesis
Cytotoxicity, RAM
Rodent Acute Toxicity
Freshwater Algal Bioassay
Freshwater Static Bioassay
Microbial Mutagenesis
Cytotoxicity, RAM
Rodent Acute Toxicity
Freshwater Algal Bioassay
Freshwater Static Bioassay
l.Og
0.5g
l.Og
0.5g
50
50
l.Og
0.5g
lOOg
50 kg
(200 fc if
sluiced)
l.Og
0.5g
lOOg
50 kg
-------
REFERENCES FOR SECTION 5
5-1. Mason, H.B., et _al_., "Preliminary Environmental Assessment of
Combustion Modification Techniques: Volume II, Technical
Results," EPA-600/7-77-1195, NTIS PB-276 681/AS, October 1977.
5-2. Waterland, L.R., et ฃ]_., "Environmental Assessment of Stationary
Source NOX Control Technologies ~ First Annual Report,"
EPA-600/7-78-046, NTIS PB-279 083/AS, March 1978.
5-3. Salvesen, K.G., et ^L, "Emissions Characterization of Stationary
NOX Sources: Volume I. Results," EPA-600/7-78-120a, NTIS
PB-284 520, June 1978.
5-4. Hamersma, J.W., et ^L, "IERL-RTP Procedures Manual: Level 1
Environmental Assessment," EPA-600/2-76-160a, NTIS PB-257 850/AS,
June 1976.
5-5 Lentzen, D. E., et _al., "IERL-RTP Procedures Manual: Level 1
Environmental Assessment (Second Edition)," EPA-600/7-78-201,
January 1979.
5-6 Duke, K.M., et a].., "IERL-RTP Procedures Manual: Level 1
Environmental Assessment Biological Tests for Pilot Studies,"
EPA-600/7-77-043, NTIS PB-268 484/3BE, April 1977.
120
-------
SECTION 6
CONTROL TECHNOLOGY OVERVIEW
The control technology assessments in the NOX EA will compile and
evaluate process data to provide environmental assessments of combustion
modification control technologies. The overall objectives of the
assessments are to:
Characterize current and advanced NOX combustion process
modifications and project schedules for applying them
Assess the technical and environmental soundness of these
control technologies
Recommend R&D for filling technological gaps and producing
needed data
t Provide objective evaluations of important aspects of NOX
control systems
The results will be documented in a series of reports covering the seven
major stationary source equipment categories.
The main efforts in the second year focused on the assessment of
NOX control techniques for the utility boiler source category. Results
from this study were recently documented (Reference 6-1) and are briefly
summarized in this section, which presents an overview of utility boiler
NOX control techniques, and in Section 7, which presents the detailed
results of the environmental assessment of applying the more promising
current technology controls.
Modifying the combustion process conditions is the most effective
and widely used technique for achieving moderate (20 to 60 percent)
reduction in combustion generated oxides of nitrogen from utility
boilers. This section reviews the combustion modification techniques
either demonstrated or currently under development. The review begins
with a discussion of the status and prospects of control requirements.
6.1 CONTROL REQUIREMENTS
The incentive
separate mechanisms:
for developing NOX controls derives from two
the Federal Standards of Performance for New
121
-------
Stationary Sources (NSPS) and the State Implementation Plans (SIP's). The
NSPS are intended largely to assist in maintaining air quality by
offsetting increases due to source growth. By law, EPA reviews, revises,
and sets NSPS as advanced control technology is developed and
demonstrated. If emission standards in addition to the NSPS are required
to attain and/or maintain the National Ambient Air Quality Standards in
Air Quality Control Regions within the jurisdiction of the states, these
standards are set through SIP's.
In the following sections, present and developing control
techniques that can help meet projected standards for utility boilers are
reviewed.
6.2 STATE-OF-THE-ART CONTROLS
There are several effective combustion modification techniques that
may be used singly or in combination on utility boilers. These techniques
include low excess air firing, biased burner firing, burners out of
service, overfire air, flue gas recirculation, and reduced firing rate.
These methods for controlling NOX may be used on existing boilers
although modifications to the units may be necessary.
6.2.1 Low Excess Air
Reducing the excess air level in the furnace is an effective method
of NOX control. In this technique, the combustion air is reduced to a
minimum amount required for complete combustion, maintaining acceptable
furnace cleanliness, and maintaining steam temperature. With less oxygen
available in the flame zone, both thermal and fuel NOX formation are
reduced. In addition, the reduced airflow lowers the quantity of flue gas
released resulting in an improvement in boiler efficiency.
Low excess air firing is usually the first NOX control technique
applied. Reductions in NOX emissions of 10 to 20 percent can be
expected. It may be used with virtually all fuels and firing methods.
However, furnace slagging and tube wastage considerations may limit the
degree of application. Low excess air may also be employed in combination
with the other NOX control methods.
6.2.2 Off Stoichiometric Combustion (OSC)
Off Stoichiometric, or staged, combustion seeks to control NOX by
carrying out initial combustion in a primary, fuel-rich, combustion zone,
then completing combustion, at lower temperatures, in a second, fuel-lean
zone. In practice, OSC is implemented through biased burner firing (BBF),
burners out of service (BOOS), or overfire air injection (OFA).
Biased Burner Firing, Burners Out Of Service
Biased burner firing consists of firing the lower rows of burners
more fuel rich than the upper rows of burners. This may be accomplished
by maintaining normal air distribution to the burners while adjusting fuel
flow so that a greater amount of fuel enters the furnace through the lower
122
-------
rows of burners than through the upper rows of burners. Additional air
required for complete combustion enters through the upper rows of burners
which are firing air rich.
In the burners out of service mode, individual burners,' or rows of
burners, admit air only. This reduces the airflow through the fuel
admitting, or active, burners. Thus, the burners are firing more fuel
rich than normal, with the remaining air required for combustion being
admitted through the inactive burners.
These methods reduce NOX emissions by reducing the excess air
available in the active burner zone. This reduces fuel and thermal NOX
formation. These techniques are applicable to all fuels and are
particularly attractive as control methods for existing units since few,
if any, equipment modifications are required. Average NOX reductions of
30 to 50 percent can be expected. In some cases, however, derating of the
unit may be required if there is too limited extra firing capability with
the active burners. This is most likely to be a problem with pulverized
coal units without spare pulverizer capacity.
Overfire Air
The overfire air technique for NOX control involves firing the
burners more fuel rich than normal while admitting the remaining
combustion air through overfire air ports.
Overfire air is very effective for NOX reduction and may be used
with all fuels. Reductions in NOX of 30 to 50 percent can be expected.
However, there is an increased potential for furnace tube wastage due to
local reducing conditions when firing coal or high sulfur oil. There is
also a greater tendency for slag accumulation in the furnace when firing
coal. In addition, with reduced airflow to the burners, there may be
reduced mixing of the fuel and air. Thus, additional excess air may be
required to ensure complete combustion. This may result in a decrease in
efficiency.
Overfire air is more attractive in original designs than in
retrofit applications for cost considerations. Additional duct work,
furnace penetrations, and extra fan capacity may be required. There may
be physical obstructions outside of the boiler setting making installation
more costly. Or, there may also be insufficient height between the top
row of burners and the furnace exit to permit the installation of overfire
air ports and the enlarged combustion zone created by the off
stoichiometric combustion technique.
6.2.3 Flue Gas Recirculation
Flue gas recirculation for NOX control consists of extracting a
portion of the flue gas from the economizer outlet and returning it to the
furnace, admitting the flue gas through the furnace hopper or through the
burner windbox or both. Flue gas recirculation lowers the bulk furnace
gas temperature and reduces oxygen concentration in the combustion zone.
123
-------
Flue gas recirculation through the furnace hopper and near the
furnace exit has long been used for steam temperature control. Flue gas
recirculation through the windbox and, to a lesser degree, through the
furnace hopper is very effective for NOX control on gas- and oil-fired
units. However, it has been shown to be relatively ineffective on
coal-fired units.
Flue gas recirculation for NOX control is more attractive for new
designs than as a retrofit application. Retrofit installation of flue gas
recirculation can be quite costly. The fan, flues, dampers, and controls
as well as possibly having to increase existing fan capacity due to
increased draft loss, can represent a large investment. In addition, the
flue gas recirculation system itself may require a substantial maintenance
program due to the high temperature environment experienced and potential
erosion from entrained ash. Thus, the cost effectiveness of this method
of NOX control has to be examined carefully when comparing it to other
control techniques.
As a new design feature, the furnace and convective surfaces can be
sized for the increase in mass flow and change the furnace temperatures.
In contrast, in retrofit applications the increased mass flow increases
turbulence and mixing in the burner zone, and alters the convective
section heat absorption. Erosion and vibration problems may result.
Flame detection can also be difficult with flue gas recirculation through
the windbox. In addition, controls must be employed to regulate the
proportion of flue gas to air so that sufficient concentration of oxygen
is available for combustion.
Limited data indicate that F6R alone reduces NOX by about 15
percent for coal, 20 to 30 percent for oil, and 30 to 60 percent for gas.
For oil and gas firing, FGR is more effective when combined with off
stoichiometric firing.
6.2.4 Reduced Firing Rate
Thermal NOX formation generally increases as the volumetric heat
release rate or combustion intensity increases. Thus, NOX can be
controlled by reducing combustion intensity through load reduction, or
derating, in existing units and by enlarging the firebox in new units.
The reduced heat release rate lowers the bulk gas temperature which in
turn reduces thermal NOX formation.
The overall heat release rate per unit volume is generally
independent of unit rated power output. However, the ratio of primary
flame zone heat release to heat removal often increases as the unit
capacity is increased. This causes NOX emissions for large units to be
generally greater than for small units of similar design, firing
characteristics, and fuel.
The increase in NOX emissions with increased capacity is
especially evident for gas-fired boilers, since total NOX emissions are
due to thermal NOX. However, for coal-fired and oil-fired units the
effects of increased capacity are less noticeable, since the conversion of
124
-------
fuel nitrogen to NOX for these fuels represents a major component of
total NOX formation. Still, a reduction in firing rate will affect
firebox aerodynamics which may, consequently, affect fuel NOX
emissions. But such effects on fuel NOX production are less significant.
Analyses of test data show that for coal firing, an average of 15
percent reduction in NOX resulted from a 28 percent reduction in firing
rate. For oil firing, an average of 30 percent reduction in NOX
resulted from a 42 percent reduction in firing rate. For gas firing, an
average of 44 percent reduction in NOX resulted from a 44 percent
reduction in firing rate. Thus, reduction of NOX with lowered firing
rate is most evident with gas-fired boilers.
Reduced firing rate often leads to several operating problems.
Aside from the limiting of capacity, low load operation usually requires
higher levels of excess air to maintain steam temperature and to control
smoke and CO emissions. The steam temperature control range is also
reduced substantially. This will reduce the operating flexibility of the
unit and its response to changes in load. The combined results are
reduced operating efficiency due to higher excess air and reduced load
following capability due to a reduction in control range.
When the unit is designed for a reduced heat release rate, the
problems associated with derating are largely avoided. The use of an
enlarged firebox produces NOX reductions similar to load reduction on
existing units.
6.3 ADVANCED CONTROLS
Two advanced control techniques hold special promise for the
future: low NOX burners with near term applications, and ammonia
injection with possible widespread application in 1985 and beyond.
6.3.1 Low NOX Burners
Several utility boiler manufacturers have recently been active in
the development of new burners designed to reduce NOX emissions from
coal-fired units. Although other techniques such as low excess air, off
stoichiometric combustion, and flue gas recirculation have been shown to
be effective in reducing NOX levels, there has been some concern as to
the efficacy of those techniques and the adverse side effects resulting
from their application. Consequently, low NOX burners are being
installed in many new wall fired units either as the primary NOX control
device or for use in conjunction with other NOX reduction methods.
Most low NOX burners designed for utility boilers control NOX
by reducing flame turbulence, delaying fuel air mixing, and establishing
fuel-rich zones where combustion initially takes place. This represents <
departure from the usual burner design procedures which promote high
turbulence, high intensity, rapid combustion flames. The longer, less
intense flames produced with low NOX burners result in lower flame
temperatures which reduce thermal NOX generation. Moreover, the reduced
availability of oxygen in the initial combustion zone inhibits fuel NOX
125
-------
conversion. Thus, both thermal and fuel NOX are controlled by the low
NOX burners.
These new, optimized design burners are capable of reducing NOX
emissions 40 to 60 percent with coal firing (References 6-2 and 6-3). New
wall fired boilers designed to meet current NSPS now come equipped with
low NOX burners. Retrofit application, however, is still in the
demonstration stage.
In addition to the new burner designs being developed by utility
boiler manufacturers, EPA-IERL/CRB is also conducting a development
program which seeks to demonstrate an advanced low emission pulverized
coal burner design on both utility and industrial boilers. Pilot scale
prototype burners have been shown capable of reducing NOX emissions
below 100 ppm (Reference 6-4). Demonstration programs are currently being
initiated.
Based on all this work, low NOX burners appear to be a very
promising control technology, with fewer potential problems than most
traditional combustion modification techniques.
6.3.2 Ammonia Injection
The selective, noncatalytic reduction of NOX via ammonia
injection has received increasing attention as a possible means to reach
quite stringent levels of control in utility boilers. In this technique
ammonia is used to reduce nitric oxide, in the presence of oxygen, to
nitrogen in a series of gas phase reactions occurring in the temperature
range of 980 to 1310K (1500 to 1900ฐF) (Reference 6-5). Demonstration
tests in Japan on oil- and gas-fired sources have shown the technique
capable of achieving 40 to 60 percent NOX reductions at optimum
temperatures in the 1200 to 1250K (1700 to 1800ฐF) range
(Reference 6-5). Further demonstrations in the U.S., including tests on
coal-fired sources, are planned.
Based on results to date, ammonia injection can be considered as
available control technology for gas- and oil-fired sources, but must be
treated as still in the development stage for coal-fired boilers. In all
applications, though, many practical problems remain to be solved. One
problem is the precise residence time/temperature conditions required for
the process. Other concerns include the effect of high dust loadings and
sulfur oxide concentrations on the effectiveness of ammonia injection in
coal-fired applications. A related problem concerns the fate, and
potential effects of any ammonium bisulfate formed from excess ammonia
present in S02/S03 containing flue gases. In any event, projected
applications of the technique have focused on reducing the NOX remaining
after other combustion modifications have been applied.
6.4 OTHER CONTROL METHODS
There are several other possible control techniques for reducing
utility boiler NOX emissions. However, they have less promise for
widespread application than those described earlier, for such reasons as
126
-------
energy penalties, high cost, or technical difficulties. These are briefly
discussed below.
6.4.1 Reduced Air Preheat
Reduced combustion air preheat (RAP) lowers the peak temperatures
in the combustion zone, thus lowering thermal NOX emissions. However,
with the associated severe loss in boiler efficiency, RAP is not
considered a practical control technique.
6.4.2 Water Injection
Water injection reduces flame temperature, and hence lowers thermal
NOX. However, boiler efficiency losses of the order of 10 percent have
been reported. Thus, water injection is not seen as a feasible NOX
reduction technique for utility boilers based on the large energy penalty
incurred.
6.4.3 Flue Gas Treatment
While combustion modification techniques seek to lower NOX
emissions by minimizing NO formation, flue gas treatment (FGT) processes
involve post-combustion NOX removal from the flue gas. Flue gas
treatment has potential for use combined with combustion modifications
when very high removal efficiencies are required.
FGT has been applied to only a few commercial oil- and gas-fired
boilers in Japan. No FGT installation for NOX control on utility
boilers exists in the United States as combustion modifications represent
the most cost effective approach to achieving moderate NOX reductions.
However, combustion modifications alone may not be able to provide the
degree of control necessary to meet future N02 ambient air quality
standards. Thus EPA has initiated several demonstration projects to
investigate the use of FGT in the U.S. (Reference 6-6).
FGT processes can be divided into two main categories: dry
processes and wet processes. Dry processes reduce NOX by catalytic
reduction and generally operate at temperatures between 570 to 700K (570
to 800ฐF). Wet systems are generally either oxidation/absorption or
absorption/reduction processes, both operating in the 310 to 320K (100 to
12QOF) range.
Among the many dry process variations, selective catalytic
reduction (SCR) using ammonia has been perhaps the most successful. Over
50 percent NOX, and often up to 90 percent reductions have been claimed
using such processes. However, plugging of the catalyst bed and fouling
of the catalyst itself are major operational concerns, especially with
coal firing. Moreover, use of SCR has raised concerns in that any ammonia
left in the flue gas may combine with existing S03/S02 to produce a
visible plume, and byproducts, such as ammonium bisulfate, which are
corrosive to boiler equipment.
127
-------
Wet FGT processes utilize more complex chemistry than dry
processes. In the oxidation/absorption processes, strong oxidants such as
ozone or chlorine dioxide are used to convert the relatively inactive NO
in the flue gas to N02 or ^5 fฐr subsequent absorption. In the
absorption/oxidation processes, chelating compounds, such as ferrous
ethylenediaminetetracetic acid are required in the scrubbing solution to
trap the NO. However, because wet processes rely on absorption, most of
them create troublesome byproducts such as nitric acid, potassium nitrate,
ammonium sulfate, calcium nitrate, and gypsum which may have little
commercial value. In addition, the high cost of an absorber and an
oxidant or chelating agent is likely to be prohibitive for flue gases with
high NOX concentrations.
In general, the dry FGT techniques used in Japan can probably be
applied to gas- and oil-fired sources in the U.S. However, the
applicability of dry processes to coal-fired boilers remains to be
determined. Wet processes are less well developed and costlier than dry
FGT processes; however, wet processes have the potential to remove NOX
and SOX simultaneously. In any case, more field tests are needed to
determine the costs, secondary effects, reliability, and waste disposal
problems. Flue gas treatment holds some promise as a control technique
for use when high NOX removal efficiencies are necessitated by stringent
emission standards. However, compared to combustion modifications FGT is
considerably more expensive.
128
-------
REFERENCES FOR SECTION 6
6-1. Lim, K.J., et al., "Environmental Assessment of Utility Boiler
Combustion MocTTFication NOX Controls," Acurex Draft Report
TR-78-105, Acurex Corporation, Mountain View, CA, April 1978.
6-2. Campobenedetto, E.J., "The Dual Register Pulverized Coal Burner --
Field Test Results," presented to Engineering Foundation Conference
on Clean Combustion of Coal, New Hampshire, August 1977.
6-3. Vatsky, 0., "Attaining Low NOX Emissions by Combining Low
Emission Burners and Off-Stoichiometric Firing," presented at the
70th Annual AIChE Meeting, New York, November 1977.
6-4. Gershman, R., et_ a\_., "Design and Scale-up of Low Emission Burners
for Industrial and Utility Boilers," in Proceedings of the Second
Stationary Source Combustion Symposium: Volume V,
EPA-600/7-77-073e, NTIS PB-274 897, July 1977.
6-5. Bartok, W., "Non Catalytic Reduction of NOX with NHs," in
Proceedings of the Second Stationary Source Combustion Symposium:
Volume II, EPA-600/7-77-073b, NTIS PB-271 756/9BE, July 1977.
6-6. Mobley, J.D., and R.D. Stern, "Status of Flue Gas Treatment
Technology for Control of NOX and Simultaneous Control of NOX
and SOX ," in Proceedings of the Second Stationary Source
Combustion Symposium: Volume III, EPA 600/7-77-073c.
NTIS PB-271 757/7BE, July 1977.
129
-------
SECTION 7
CONTROL TECHNOLOGY ASSESSMENT
As noted in Section 1, the key objectives of the NOX EA are to
identify the environmental impact of combustion modification NOX
controls applied to stationary combustion sources and to specify the most
cost-effective and environmentally sound NOX controls to attain and
maintain N02 air quality goals. To satisfy these goals, a major aim of
the program is to extend the control technology process background
presented in Section 6 to include detailed evaluations of the emissions,
source performance, and cost impacts of applying these controls.
Second year results from the assessment of NOX combustion
modification controls applied to utility boilers are presented in this
section. The basis and effectiveness of these controls and their process
operational, cost, and environmental impacts are discussed.
7.1 EFFECTIVENESS OF NOX CONTROLS
Combustion modification techniques control NOX formation by
decreasing primary flame zone 02, lowering peak flame temperature, and
shortening the flame zone residence time. The percentage reductions in
NOX that can be expected with application of the various techniques were
briefly discussed in Section 6. To reiterate, fine tuning and application
of low excess air (LEA) can reduce NOX emissions 10 to 20 percent. Off
stoichiometric combustion (OSC), biased burner firing (BBF), burners out
of service (BOOS), and overfire air (OFA) can lower NOX emissions 30 to
50 percent. Low NOX burners show great promise, reducing NOX 40 to 60
percent for new coal-fired boilers, with retrofit application feasible.
Flue gas recirculation (FGR) is an effective technique for oil and gas
firing, especially when combined with OSC, lowering NOX by 30 to 75
percent.
The above control performance expectations were quantitatively
derived by applying the NOX emissions correlation model outlined in
Section 4.3.1. to an emission data base assembled from the results of 61
NOX control application field test programs and including 563 individual
test points. The data base included test programs on coal-, oil-, and
gas-fueled tangential, opposed wall, and single wall fired boilers as
shown in Table 7-1. Controls tested included LEA, OSC, FGR, load
reduction, and combinations of these, as shown in Table 7-2. Both
published, and previously unreported data were included.
130
-------
TABLE 7-1. FIELD TEST PROGRAM DATA COMPILED
Fuel
Coal
Oil
Natural Gas
Total
Firing Type
Tangential
13
2
1
16
Opposed Wall
6
7
8
21
Single Wall
10a
7
7b
24
Total
29
16
16
61
Includes two wet bottom furnaces
Includes one unit originally designed for coal firing with a
wet bottom furnace
131
-------
TABLE 7-2. INDIVIDUAL TEST POINTS CORRELATED
Firing Type
Tangential
Opposed
Hall
Single Hall
Tangential
Opposed
Wall
Single Will
Tangential
Opposed
Wall
Single Hall
All Boilers
Fuel
Coal
Coal
Coal
Oil
011
Oil
Nat gas
Nat gas
Nat gas
All fuels
Bซe11neb
21
e
18
1
6
4
1
7
5
71
Single Controls
LEAC
29
11
23
~
5
6
1
9
4
88
oscd
46
11
29
1
11
5
--
18
9
130
FGR'
7
2
4
2
2
17
Low
Loadf
24
7
19
1
7
8
2
13
7
88
Combined Controls*
Ion load
* OSC
27
5
19
1
7
6
1
13
7
86
Low Load
+ FGR
--
1
1
5
10
5
3
3
28
OSC *
FGR
2
ซ
--
2
10
1
3
4
22
Low Load +
OSC + FGR
~
1
11
8
--
8
5
33
Total
147
52
108
6
56
61
13
74
46
563
'LOU excess air also generally employed
Baseline no controls applied; boiler load near or at axiBum rating; excess air at
normal or above normal settings
CLEA low excess air setting
OSC ซ off stoichlometrlc combustion (Includes: biased burner firing, burners out of
service, overflre air)
*FGR flue gas reclrculatlon; generally Includes low excess air setting
Load less than 80 percent of Mxiftum continuous rating (MCR)
T-B07
132
-------
The NOX correlation model showed that the key boiler/burner
design and operating variables affecting NOX emissions were:
Heat input per active burner
Stoichiometry to active burners
t Firing rate
t Number of burners firing (degree of BOOS)
t Surface heat release rate
Furnace Stoichiometry
Percent flue gas recirculation
Number of furnace division walls
The only fuel property statistically required for use was the fuel type:
coal, oil, or natural gas. Thus the effective NOX controls were the
ones that controlled an "optimal combination" of selected variables.
Correlation results were obtained for seven of the nine possible
firing type/fuel combinations. Data for controls applied to gas- and
oil-fired tangential boilers were too scattered to give good results.
As an example, for coal-fired tangential boilers, NOX emissions
were predicted (with a correlation coefficient of 0.87) as:
y = 389.4 + 1.962 x 10-7(xi)(x2) - 3.017 x 10-5(X1)
+ 3.249 x 10-6(x3)(x4) + 1.57 x 10-3(xi)2
where
y = NOX emissions (ppm dry at 3 percent 02)
xi = Heat input per active burner (W)
X2 = Stoichiometry to active burners (percent)
X3 = Surface heat release rate (W/m^)
X4 = Furnace Stoichiometry (percent)
Figures 7-1 and 7-2 are graphical presentations of these results. The
correlation fit is good considering that the data were from 147 tests
carried out on a total of 13 boilers, in several different test programs.
Results for other boiler/fuel classifications were comparable, and are
reported elsewhere (Reference 7-1).
7.2 PROCESS ANALYSIS OF NOX CONTROLS
This section summarizes the major impacts of combustion
modification controls on boiler operation and incremental emissions. The
discussion is organized by fuel type and control technique, the dominant
factors.
133
-------
OJ
OJ
X
O
700_
600
500
400
300
200
100
O
Q
8
ฉ
20
40
60
80
100
120
140
3 O
Surface heat release rate (10' Bt.u/hr-ft )
160
Stoichiometry to active
burners (percent)
0 140
0 120
X 100
2 so
I
500
100
200 300
kW/ni2
400
Figure 7-1. Effect of surface heat release rate and burner Stoichiometry
on NO from tangential coal-fired boilers.
n.
-------
Stoichiometrv to active
burners (oerccnt.)
CJ
cn
700
600
o
o->
CL
CL
C
o
200 -
100
Heat innut/active burner (10 Btu/hr)
I i i
10
20
50
120
O 140
D 120
X 100
Z 80
C<
o
ci
MVJ
Figure 7-2. Effect of heat input and burner stoichiometry on
NO from tangential coal-fired boilers.
-------
7.2.1 Coal-Fired Boilers
The effects of low NOX operation on coal-fired boilers are
summarized in Table 7-3. The most commonly applied low NOX techniques
for coal-fired boilers are low excess air and off stoichiometric
combustion. Low NOX burners are also being installed on some new units
and have been found to be effective. Other techniques which have been
tested but are less commonly employed are flue gas recirculation, which
has been found to be relatively ineffective, and water injection (WI),
which is not preferred because of efficiency losses.
The major concerns regarding low NOX operation on coal-fired
boilers have been the effects on boiler efficiency, load capacity, water
wall tube corrosion and slagging, carbon loss, particulate loading and
size distribution, heat absorption profile, and convective section tube
and steam temperatures.
In most past experience with OSC, optimal excess air levels have
been comparable to those used under baseline conditions. In these cases
the efficiency of the boiler would remain unaffected if unburned carbon
losses do not increase appreciably. However, in some cases when, due to
nonuniform fuel/air distribution or other causes, the excess air
requirement increases substantially with OSC, a significant decrease in
efficiency may occur. From Table 7-3, it is seen that efficiency
decreases up to 1 percent may occur under OSC. It is also seen that the
same boiler (Widows Creek No. 5) tested at a different time under LEA and
BOOS showed an average increase in efficiency by 1 percent.
Many new boilers now come factory equipped with OFA ports. Older
boilers can be retrofitted with OFA ports or operate with minimal hardware
changes under BOOS or biased firing. BOOS firing is normally accomplished
by shutting off one or more pulverizers supplying the upper burner
levels. If the other pulverizers cannot handle the extra fuel to maintain
the total fuel flow constant, boiler derating will be required. From
Table 7-3, it is seen that boiler derating of 10 to 25 percent is not
uncommon with BOOS firing. Biased firing (reducing but not shutting off
completely, fuel flow to upper burner levels) may reduce or eliminate the
amount of derating a boiler has to suffer. However, this type of firing
has not been tested sufficiently to establish its effectiveness as a NOX
control technique.
The possibility of increased corrosion has been a major cause for
concern with OSC operation. Furnaces fired with certain Eastern U.S.
bituminous coals with high sulfur contents may be especially susceptible
to corrosion attack under reducing atmospheres. Local reducing atmosphere
pockets may exist under OSC operation even when burner stoichiometry is
slightly over 100 percent. The problem may be further aggravated by
slagging as slag generally fuses at lower temperatures under reducing
conditions. The sulfur in the molten slag may then readily attack tube
walls. Still, it has been found in general that no significant
acceleration in corrosion rates occurs under OSC conditions. More recent
experience has substantiated this conclusion (Reference 7-2).
Nevertheless, the issue cannot be considered resolved until definitive
136
-------
TABLE 7-3. EFFECT OF LOW NOU OPERATION ON COAL-FIRED BOILERS
Boiler
Tangential
Barry Ho. 2
Colwfcla
No. 1
Hunt Ing ton
Canyon No. 2
Barry No. 4
Navajo No. 2
Conanche No. 1
Opposed Hall
Harllee Branch
No. 3
Four Corners
No. 4
Hitfltld No. 3
low NO
TechnlqOe
BOOS
OTA
or*
OFA
LEA. BOOS
LEA, BOOS, OFA
OFA
LEA. BOOS
LEA. BOOS
BOOS
Efficiency
Unaffected
Unaffected
Unaffected
Unaffected
Unaffected
Unaffected
Unaffected
0.6X average
decrease
0.61 Increase
0.3X decrease
Corrosion*
Measured 75X
Increase, but
within normal
range
Measured 70S
Increase, but
within normal
range
No change
Measured 251
decrease, but
within normal
range
No significant
change
No significant
change
No significant
change
Slight Increase
No significant
change
Load
Capacity
201 derate
Unaffected
Unaffected
Unaffected
20X or more
derate with
BOOS
Unaffected
Unaffected
Up to 17X
derate
with BOOS
Up to 25X
derate
with BOOS
10X derate
Carbon Loss
In Flyash
Slight Increase
Sllghi Increase
Slight Increase
Slight Increase
-SOX average
decrease
No change
-30X average
decrease
-130X average
Increase
-SOX average
decrease
-30X average
Increase
Dust Loading*
-100X Increase
-100X Increase
..
..
-SOX average
Increase
-40X average
Increase
-20X average
decrease
-10X average
Increase
-1SX average
decrease
Unaffected
Part. Size
Distribution*
..
--
_.
No change
No significant
change
__
_
Other Effects.
Comments
Minor changes In heat
absorption profile
SH attemperation
Increased by 70X
Minor changes In heat
absorption profile
SH attemperation
Increased over 200X
Minor changes In heat
absorption profile
SH attemperation
Increased by 70X
Minor changes In heat
absorption profile
No SH attemperation
required
No slagging or foul-
ing. No significant
Increase In tube tem-
peratures.
-------
TABLE 7-3. Concluded
Boiler
E.C. Gaston
No. 1
Single wall
Widows Creek
No. 5 (TVA
test)
Widows Creek
No. 5 (Exxon
test)
Widows Creek
No. 6
Mercer Station
No. 1 (wet
bottoaj
Crist Station
No. 6
Low NOX
Technique
FSR
LNB. LEA. BOOS
BOOS
LEA, BOOS
LEA. BOOS
LEA. Biased
firing
LEA. BOOS
Efficiency
0.4S decrease in
boiler effi-
clency. Sane
decrease in cycle
efficiency due to
RH attcMperation
0.3S decrease
on average (LNB
baseline)
IS decrease
IS average
Increase
Unaffected
Unaffected
0.4S decrease
Corrosion*
No significant
increase
Results of tests
inconclusive
No significant
Increase
--
No significant
Increase
--
Load
Capacity
Unaffected
Up to 30S
derate
(LNB with
BOOS)
Unaffected
Unaffected
Unaffected
Unaffected
Up to 15S
derate
Carbon Loss
In Flyash
-120X average
increase
-130S average
increase (LNB
baseline)
30X increase
301 average
decrease
70S average
Increase
SOS average
increase
60S Increase
Oust Loading1
Unaffected
-IBS average
increase (LNB
baseline)
No significant
Increase
15S average
decrease
20S average
decrease
IDS average
increase
SOS increase
Part Size
Distribution*
Shift towards
waller par-
ticles (LNB.
with or with-
out BOOS)
--
No significant
change
--
Other Effects.
Consents
Stable flames and
uniform coBbustton.
Increase in RH
attemperation. No
significant increase
in tube temperatures.
Unit retrofitted
with low NO, burners.
Baseline. LEA and
BOOS tests with LNB
compared to baseline
tests on sister
boiler with no LNB.
1 , ... T-833
Denotes not Investigated
co
00
-------
results from long term tests with measurements on actual water wall tubes
are available. Insofar as slagging is concerned, short term tests
performed to date indicate no increase in slagging or fouling of tubes
under OSC operation.
Increased carbon loss in flyash may occur with OSC if complete
burnout of the carbon particles does not occur in the furnace. High
carbon loss will result in decreased boiler efficiency and may also cause
electrostatic precipitator (ESP) operating problems. From Table 7-3, it
is seen that increases in carbon loss vary over a wide range and can be as
high as 70 to 130 percent in some cases. However, increased carbon loss
is not perceived as one of the major problems associated with OSC
operation. If the carbon content in flyash increases to levels where it
threatens to impair the operation of dust collection systems, the unburned
carbon can usually be easily controlled by increasing the overall excess
air level in the furnace. Although this will tend to increase stack heat
losses, the decrease in boiler efficiency will be partially compensated
for by reduced unburned carbon losses.
Increased particulate loading with OSC may be a source of problems
if baseline loadings are close to acceptable limits. Installing larger or
more efficient dust removal devices may be necessary. The problem can be
particularly severe if the particle size distribution shifts toward
smaller sizes because the efficiency of many dust collectors, such as
ESP's, decreases in the 0.1 to 1.0ym range. From Table 7-3 it is seen
that dust loading changes can vary widely. In some cases, dust loading
may double with OSC operation, although from the few size distribution
data available no shift in distribution is evident. It should be noted,
however, that most of the particulate loading measurements were carried
out at the economizer outlet and do not necessarily reflect stack outlet
conditions.
Extension of the combustion region to higher elevations in the
furnace may result in potential problems with excessive steam and tube
temperatures. However, among the numerous short term OSC tests conducted
no such problems have been reported. In some tests where furnace and
convective section tube temperatures were measured directly, no
significant increase was found. Changes in heat absorption profiles were
also found to be minor, thus indicating no need for addition or removal of
heat transfer surfaces. Superheater attemperator spray flowrates tripled
in one case under OFA operation, but in all cases were well within spray
flow capacities of the unit. Reheater attemperator spray flowrates did
not show any increase due to OSC operation, thus cycle efficiencies were
not affected.
Many new wall fired coal boilers are being fitted with low NOX
burners (LNB). These burners are designed to reduce NOX levels to meet
statutory requirements either alone or in some cases in combination with
OFA ports. Using LNB has the advantage of eliminating or decreasing the
need for reducing or near reducing conditions near furnace walls.
Corrosion problems associated with reducing atmospheres should thus not
arise with these systems. Although the LNB flames can be expected to be
less turbulent and hence longer than flames from normal burners, the
139
-------
combustion zone will probably not extend any farther up the furnace than
with OSC. Potential changes in heat absorption profile and excessive
steam and tube temperatures are, therefore, less likely to occur.
As fuel and airflows are controlled more closely in LNB equipped
systems, nonuniform distribution of fuel/air ratios leading to excessive
CO generation or high excess air requirements should be eliminated.
Boiler efficiencies should, therefore, not be affected by installation of
LNB. However, Table 7-3 shows that the efficiency of one boiler decreased
slightly when retrofitted with LNB. The decrease in efficiency was mainly
due to the large increase in unburned carbon loss. Particulate loading
also increased slightly with LNB, and there was a distinct shift towards
smaller size particles. Still, more testing is required to check whether
these changes were isolated instances or whether they form a pattern with
LNB operation. It should be noted that the decrease in efficiency and
increases in carbon loss and particulate loading were not greater than
those encountered with OSC operation. Corrosion rates inferred from tests
with corrosion coupons showed no significant increase with LNB. Some BOOS
tests were also carried out on the LNB equipped boiler. A substantial
decrease in NOX emissions resulted, below those already achieved with
LNB alone. However, the boiler was derated by up to 30 percent. Other
potential problems associated with OSC could also arise with this type of
firing.
F6R to the windbox has been tested as a NOX control technique for
coal-fired boilers. FGR inhibits thermal NOX formation but is not very
effective in controlling fuel NOX. Thus, the technique has not been
used widely on coal-fired units as it is not very effective in these
applications. The tests on Hatfield No. 3 showed that OSC was indeed much
more effective in controlling NOX than FGR. Table 7-3 summarizes some
of the effects of FGR operation on that unit. The increase in carbon loss
averaged 120 percent, although there were wide variations in the measured
values. Load capacity and dust loading remained unaffected. There was a
slight decrease in boiler efficiency attributable to the power consumption
by the FGR fans. There were no significant increases in tube temperature
and stable flames and uniform combustion were observed throughout the
tests, even at high recirculation rates (up to 15 percent at full load and
34 percent at reduced loads). Reheat steam spray attemperation increased
at high recirculation rates which could result in a loss in cycle
efficiency. No corrosion measurements were made so that the effects of
FGR on corrosion are not known. Corrosion due to chemical attack is not
expected to be a major problem with FGR. However, tube erosion may
increase as the higher gas velocities may result in greater particle
impact on exposed surfaces.
Some data were available on the effect of water injection on NOX
emissions. Water injection, however, results in a significant
deterioration of boiler performance. It has therefore not been
recommended as a NOX control measure for coal-fired boilers.
It should be emphasized that the effects of NOX control, in many
cases, will be critically dependent on boiler operating conditions.
Still, with proper design of retrofit systems and adequate maintenance
140
-------
programs, low NOX operation should not result in a substantial increase
in operational problems over normal boiler operation. Moreover, when
NOX controls are designed into new units, potential problems can be
anticipated and largely corrected.
7.2.2 Oil-Fired Boilers
The effects of low NOX operation on oil-fired boilers are
summarized in Table 7-4. The most common low NOX techniques tested for
oil-fired boilers are low excess air (LEA), off stoichiometric combustion
(OSC), and flue gas recirculation (FGR). Other techniques which have been
tested are water injection (WI) and reduced air preheat (RAP). However,
these have found little application due to attendant efficiency losses.
The major concerns regarding low NOX operation on oil-fired
boilers are effects on boiler efficiency, load capacity, vibration and
flame instability, and steam and tube temperatures.
OSC operation generally increases the minimum excess air
requirements of the boiler, which may result in a loss in boiler
efficiency. In extreme cases when the boiler is operating close to the
limits of its fan capacity, boiler derating may be required. Derates of
as much as 15 percent have been reported due to the lack of capability to
meet the increased airflow requirements at full load.
In many cases, BOOS operation in oil-fired boilers has been found
to be more effective in controlling NOX than OFA firing. Under BOOS
firing the fuel flow to the active burners must be increased if load is to
remain constant. In some cases, it has been necessary to enlarge the
burner tips in order to accommodate these increased flows.
Other potential problems attendant with applying OSC in oil-fired
boilers have concerned flame instabilities, boiler vibrations, and
excessive convective section tube temperatures. However, in past
experience, none of these problems has been significant. Staged operation
does usually result in hazy flames and obscure flame zones. Thus new
flame scanners and detectors are often required in retrofit applications.
In addition, because OSC produces an extended flame zone, flame carryover
to the convective section may occasionally occur. However, in one case
where intermittent flame carryover occurred, no excessive tube
temperatures were, recorded.
Similarly there are a number of potential problems which can occur
in retrofit FGR applications. The most common problems, such as FGR fan
and duct vibrations, can usually be avoided by good design. Other
problems such as flame instability, which can lead to furnace vibrations,
are caused by the increased gas velocity at the burner throats.
Modifications to the burner geometry and design such as enlarging the
throat, altering the burner tips, or adding diffuser plates or flame
retainers, may then be required. These modifications are usually made by
trial and error for each boiler and are often very time consuming. If the
problems of excessive boiler vibration and flame instabilities persist at
high loads, the boiler may have to be derated.
141
-------
TABLE 7-4. EFFECT OF LOW NO OPERATION ON OIL-FIRED BOILERS
Boiler
Tanojentlal
South Bay No. 4
rittsburg No. 7
SCE tangential
boilers
Noss Landing
Bos. 6 and 7
Ormond Beach
Nos. 1 and 2
see m units
Sewaren Station
No. S
Low NO,
Technique
LEA
BOOS
RAP
OF* and FGR
BOOS and FGR
OFA and FGR
BOOS and FGR
Mater Injection
BOOS and FGR
LEA. BOOS
Efficiency*
5t Increase
Decrease In efficiency
compared to LEA due to
Increased excess air
requirements
Unaffected due to
special preheat er
design
Increased excess air
requirements resulting
In decreased efficiency
Increased excess air
requirements resulting
In decreased efficiency
Increased sensible and
latent stack losses
FGR reduced minimum
excess air require-
ments Increasing
unit efficiency
Load
Capacity*
--
--
Slower startups
and load changes
--
10 to 15* derate
due to maxed FD
fan capacity
Vibration and
Flame Instability1
-
FGR fan vibration
problems
FGR fan and duct
vibration, furnace
vibration problems.
Associated flame
Instability
Flame Instability
and associated
furnace vibration
Boiler vibration
problems
Steam and Tube
Temperatures
--
--
High water wall tube
temperatures
Other Effects, Comments
No adverse effects reported.
Fan power consumption
reduced.
No other adverse effects
reported
Limited tests. NOx
control effectiveness not
demonstrated.
No adverse effects reported
High furnace pressures.
Increased FGR and forced
draft fan power consumption,
Flame detection problems
due to change In flame
characteristics
Limited tests carried out
with HI at partial loads.
Excess air requirements
increased
Flame detection problems
due to change In flaw*
characteristics
Tests carried out at part 1 a
loads. No adverse effects
reported. Partlculate load-
Ing and size distribution
unaffected.
T-831
-pi
ro
-------
TABLE 7-4. Concluded
Boiler
Slnqle Hall
Enclna Nos. 1,
Z and 3
Turbo
South Bay No. 3
Potrero Wo. 3-1
Low NO
Technique
LEA and BOOS
(2 burners
on air only)
BOOS
(3 burners on
air only)
Airflow
adjustments
Hater Injection
Reduced air
preheat
OFA and FGR
Efficiency*
Increased unit effi-
ciency. Some adverse
effect on cycle effi-
ciency due to lower
steam temperatures
Increased excess air
requirements resulting
In reduced efficiency
Slight reduction In
EA resulting In slight
Increase In efficiency
6S decrease at full
load
Reduction In effi-
ciency greater than
that with water
Injection
Higher excess air re-
quirements, but addi-
tion of economizer
surface expected to
Improve efficiency
Load
Capacity"
SX derate due to
maxed ID fan
capacity
5X derate due to
excessive tube
temperatures
Vibration and
Flame Instability"
In most tests no
flame Instability
or blowoff noted
No flame Instability
noted even at high
rates of MI
Side to side
wlndbox oxygen
cycling
Steam and Tube
Temperatures
Decrease In SH I RH
steam temperature
Intermittent flame
carryover to SH
Inlet but tube
temperature limits
not exceeded
--
--
Tube and steam tem-
perature limits ap-
proached. Increased
SH tube failures.
Other Effects. Comments
No other adverse effects
reported
No abnormal tube fouling,
corrosion or erosion noted.
Increased tendency to smoke
and obscure flame zone.
No adverse effects reported
No other adverse effects
reported
Limited tests
Increased tendency to smoke
required higher minimum ex-
cess 0? levels. RH surface
removed to avoid excessive
RH steam attemperatton.
Larger economizer Installed
to compensate for RH surface
removal.
T-831
CO
'Denotes npt Investigated
-------
Another potential problem associated with F6R is high tube and
steam temperatures in the convective section. The increased mass
velocities which occur with FGR cause the convective heat transfer
coefficient to rise. This, coupled with reduced furnace heat absorption,
can give rise to high convective section temperatures leading to tube
failures, exceeding attemperator spray flow limits, or loss in cycle
efficiency due to excessive reheat steam attemperation. Increased mass
flowrates in the furnace may also cause furnace pressures to increase
beyond safe limits. FGR usually, however, has an advantage of not
increasing minimum excess air levels. Boiler efficiency is therefore
relatively unaffected except for the power consumed by the FGR or booster
fans.
The combination of OSC and FGR is very effective in reducing NOX
emissions. However, the problems associated with each technique are also
combined. Tube and steam temperature problems in the upper furnace are
particularly exacerbated, as both OSC and FGR tend to increase upper
furnace temperatures and convective section heat transfer rates. In
addition, boiler efficiencies usually decline slightly with combined OSC
and FGR firing due to higher excess air requirements and greater fan power
consumption.
As with coal-fired boilers, before low NOX techniques are applied
to an oil-fired boiler, it is important to assure that it is in good
operating condition. Uniform burner air and fuel flows are essential for
optimal NOX control. Retrofit NOX control systems must be designed
and installed properly to minimize potential adverse effects. Despite
these precautions, in some cases inevitable problems will occur, such as
flame instability or high tube temperatures. In some of these cases,
certain hardware modifications will be required to resolve the problems.
In other cases, increased vigilance will be needed on the part of the
boiler operator, and an accelerated schedule of maintenance and overhaul
may be required. Many of the problems experienced in the past can now be
avoided because of hindsight and experience. Thus, retrofit systems can
now be designed and installed with care to avoid any potential adverse
effects. New units with built-in OFA and FGR systems or LNB should
function without problems.
7.2.3 Gas-Fired Boilers
The effects of low NOX operation on gas-fired boilers are
summarized in Table 7-5. The low NOX techniques used and their effects
are very similar to those for oil-fired boilers. Usually, there is no
distinction between oil- and gas-fired boilers as they are designed to
switch from one fuel to the other according to availability. Since boiler
design details, NOX control methods, and the effects of low NOX
operation are similar for gas- and oil-fired units, most of the above
discussion of applicable NOX control, measures to oil-fired boilers and
potential problems resulting applies. Some effects specific to gas-fired
boilers alone are treated briefly below.
NOX emissions oftentimes are difficult to control after switching
from oil to gas firing. Residual oil firing tends to foul the furnace due
144
-------
TABLE 7-5. EFFECT OF LOW NO OPERATION ON GAS-FIRED BOILERS
A
Boiler
Tangential
South Bay No. 4
Plttsburg No. 7
Horizontally
Opposed
Moss Landing
Nos. 6 and 7
Plttsburg
Nos. 5 and 6
Contra Costa
Nos. 9 and 10
Single Will
Enclna Nos. 1,
2 and 3
Low NOX
Technique
LEA
BOOS
OFA and FGR
OFA and FGR
OFA and FGR
OFA and FGR
BOOS
(2 arid 3
burners out
of service)
Efficiency*
2 to 3% Increase
Decrease In efficiency
compared to LEA due to
Increased excess air
requirements
__
0.8* decrease In cycle
efficiency due to RH
steam attemper at I on
Low EA levels were
possible even with
BOOS, resulting In
increased efficiency
Load*
Capacity
25* derate due to
excessive steam
temperatures ,
slower load
change response
Load curtailment
to SOX after oil
burns due to SH
tube temperature
limits being
exceeded
No derate. Load
pickup response
not affected
Vibration anda
Flame Instability
-
Fan and duct
vibration problems
Furnace and duct
vibration problems.
Flame Instability.
FGR fan and duct
vibrations. Flame
Instability problems.
FGR duct vibrations
Some pressure
pulsing at
corners of
firebox
Steam and Tube*
Temperatures
High tube and RH
steam temperatures
RH spray and SH tube
temperature limits
approached after oil
burns upper wall tube
failures
Upper water wall
tube failures
High SH and RH steam
temperatures. SH
tube temperature
limits being
approached.
Some flame carryover
to SH but no
problems with high
tube temperature or
tube wastage
Other Effects, Comments
No adverse effects reported
No other adverse effects
reported
Furnace pressure limit
approached. FGR fan power
requirements Increased by
as much as 661. Problems
associated with switching
to gas after oil burning
could be eliminated only
with complete water washing
of furnace.
Boiler initially restricted
to manual operation due to
problems with flame Insta-
bility on automatic control
Furnace pressure limits
approached after oil firing.
FGR fan preheating required
to reduce vibrations on cold
boiler startups.
No other adverse effects
reported
Ul
*0enotes not investigated
-------
TABLE 7-5. Concluded
Boiler
Turbo
South Bay No. 3
Potrero No. 3-1
low NO,
Technique
Airflow
adjustments
Water Injection
OFA and FGR
Efficiency*
Slight reduction (n
EA resulting In slight
Improvement 1n
efficiency
101 decrease at full
toad
Installation of larger
economizer expected to
Improve efficiency
Load"
Capacity
"
-
5i derate due to
problems with high
temperatures
Vibration and*
Flame Instability
"
No fine Instability
noted even at high
rates of U!
Side to side
wtndbox oxygen
cycling
Steam and Tube"
Temperatures
"
--
Tube metal and steam
temperature limits
reached at high
loads
Other Effects, Comments
No adverse effects reported
No other adverse effects
reported
Hardware modifications
Included partial RH surface
removal to avoid excessive
RH steam attemperatlon.
Larger economizer then
Installed to compensate for
smaller RH surface.
en
'Denotes not Investigated
T-Big
-------
to the oil ash content. Thus, NOX control measures which have been
tested on a clean furnace with gas may be found inadequate after oil
firing due to the changed furnace conditions. These problems can be
resolved by complete water washing of the furnace after any oil burns.
This is not very practical, however, especially if oil to gas fuel
switching occurs frequently.
Boilers fired with gas usually have higher gas temperatures at the
furnace outlet than when fired with oil. Gas flames are less luminous and
therefore radiate less energy to the furnace walls than oil flames. The
upper furnace and convective section inlet surfaces are thus subject to
higher temperatures with gas firing. These temperatures may increase
further when the combustion zone is extended due to OSC. Furthermore,
heat transfer rates in the convective section will rise with increased
mass velocities due to FGR. Upper furnace and convective section tube
failures and excessive steam temperatures are therefore more likely to
occur with OSC and FGR applied to gas-fired boilers. The situation may be
aggravated further if switching from gas fuel occurs after an oil burn, as
fouling will further reduce furnace absorption and, hence, increase gas
temperatures. Excessive tube temperatures will usually result in a
derating of the system. However, problems with gas firing are not
commonly encountered at present due to the scarcity of natural gas fuels,
and that trend is likely to continue in the future.
7.3 COSTS OF NOX CONTROLS
In the detailed environmental assessment of NOX controls applied
to utility boilers, representative control costs were prepared for the
following typical boiler/control combinations:
Boiler/Fuel Type NOX Control
Tangential/Coal OFA
Opposed Wall/Coal OFA
Opposed Wall/Coal Low NOX Burners
Opposed Wall/Coal BOOS
Single Wall/Oil and Gas BOOS
Single Wall/Oil and Gas OFA and FGR
Overfire air and low NOX burners were selected as the retrofit
control methods for coal firing. Burners out of service was not
necessarily recommended for coal-fired units, but was included to
demonstrate the prohibitively high cost of derating a unit, as is often
the case for pulverized coal units. Burners out of service, and flue gas
recirculation through the burners combined with overfire air were selected
as the retrofit control methods for the single wall oil- and gas-fired
unit. These methods have been shown to be effective in retrofit
applications, as discussed earlier in this report.
7.3.1 Retrofit Control Costs
Based on the cost analysis methodology presented in Section 4.3.3,
typical retrofit control costs (1977 dollars) are summarized in
147
-------
Table 7-6. It is assumed here that low excess air represents standard
operating procedure. Any investment costs for this control are usually
offset by savings in operating efficiency. All other assumptions and
detailed cost input data are summarized elsewhere (Reference 7-1). It
should be emphasized here that the control costs shown in Table 7-6 are
only representative typical retrofit control costs. They represent
retrofitting relatively new boilers, say 5 to 10 years old with at least
25 years of service remaining. With the exception of BOOS for coal-fired
units, and F6R/OFA for oil- and gas-fired units, annualized control costs
generally fall in the $0.50 to 0.70/kW-yr, based on a 7000-hour operating
year. For comparison, the cost of operating a power plant is
approximately $175/kW-yr.
Burners out of service was treated in the cost analysis not as a
recommended control technique for coal firing but to show the
prohibitively high cost of derating. This high cost was due principally
to the need to purchase make up power from elsewhere and to account for
the lost capacity of the system through a capital charge.
Table 7-7 presents projected retrofit control requirements for
alternative NOX emissions levels. Control techniques are also
recommended to achieve each given NOX emission level. These
requirements and techniques, combined with the cost to control column,
complete the cost effectiveness picture.
Based on the favorable process analysis results presented in
Section 7.2, it is evident from an examination of Tables 7-6 and 7-7 that
OFA and low NOX burners (LNB) are the preferred, cost-effective NOX
controls for coal firing. For very high levels of NOX control for
coal-fired units (170 ng/J), both OFA and LNB would be required. For more
moderate levels of control, LNB are less expensive and more cost-effective
than OFA in reducing NOX in wall fired units. However, the use of LNB
technology in retrofit application is still a few years away. Thus, LNB
is not recommended now for moderate levels of control in retrofit
applications in spite of the fact that the technology is potentially less
expensive than OFA.
As far as moderate control for oil- and gas-fired units, off
stoichiometric combustion via BOOS appears to be the preferred route, as
indicated in Tables 7-6 and 7-7. Initial investment is minimized since
there are no associated major hardware requirements, only engineering and
startup costs. To reach the next level of NOX control (86 ng/J for oil,
43 ng/J for gas), FGR with OFA would seem to be in order. However, the
increase in cost from $0.49/kW-yr for BOOS to $3/kW-yr for FGR and OFA
does not make the option attractive. Besides, from a regulatory point of
view, requirement of the emission level achievable with FGR and OFA would
not be particularly attractive since oil- and gas-fired units with BOOS
would already have very low NOX emissions (129 ng/J for oil, 86 ng/J for
gas) compared to coal-fired units.
148
-------
TABLE 7-6. SUMMARY OF RETROFIT CONTROL COSTS
to
Boiler/Fuel Type
Tangent 1 a 1 /Coa 1 -F 1 r ed
OFA
Opposed Wall /Coal -Fired
OFA
LNB
BOOS
Single Wall/011- and Gas-Fired
BOOS
FGR/OFA
Initial
Investment
(SAW)
0.90
0.62
2.03
0.08
0.30
5.71
Annuallzed Indirect
Operating Cost
($/kW-yr)
0.21
0.16
0.34
5.34
0.05
1.14
Annuallzed Direct
Operating Cost
($/kW-yr)a
0.32
0.52
0.06
24.78
-
0.44
1.91
Total to Cost
Control
(J/kW.-yr)a
0.53
0.69
0.40
30.12
0.49
3.05
aBased on 7000-hour operating year. Typical cos.ts only.
T-870
-------
TABLE 7-7. PROJECTED RETROFIT CONTROL REQUIREMENTS FOR ALTERNATE
NOX EMISSIONS LEVELS
in
o
Fuel/N0x Emission Level
ng/J (lb/106 Btu)
Coal
301 (0.7)
258 (0.6)
215 (0.5)
172 (0.4)
Oil
129 (0.3)
86 (0.2)
Gas
86 (0.2)
43 (0.1)
Recommended
Control3
OFA
OFA
LNBC
OFA + LNBC
BOOS
F6R + OFA
BOOS
F6R + OFA
Cost to Control
$/kW-yrb
0.50 to 0.70
0.50 to 0.70
0.40 to 0.50
0.95 to 1.20
0.50 to 0.60
~3.00
0.20 to 0.30
~3.00
Cost Effectiveness
$/kg NOX Removed
0.03 to 0.04
0.02 to 0.03
0.01 to 0.02
0.02 to 0.03
0.04 to 0.05
-0.16
0.03 to 0.04
-0.12
aLEA considered standard operating practice.
bTypical installation only; could be significantly higher.
cTechnology not thoroughly demonstrated for retrofit yet.
-------
7.3.2 Control Costs for New Boilers
Estimating the incremental costs of NOX controls for NSPS boilers
is in some respects an even more difficult task than costing retrofits.
Certain modifications on new units, though effective in reducing NOX
emissions, were originally incorporated due to operational considerations
rather than from a control viewpoint. For example, the furnace of a
typical NSPS unit has been enlarged to reduce slagging potential and allow
the burning of process quality fuels. But this also reduces NOX due to
the lowered heat release rate. Thus, since the design change would have
been implemented even without the anticipated NOX reduction, the cost of
that design modification should not be attributed to NOX control.
Babcock & Wilcox has estimated the incremental costs of NOX
controls on an NSPS coal-fired boiler (Reference 7-3). The two units used
in the comparison were identical except for NOX controls on the NSPS
unit which included:
Replacing the high turbulence, rapid-mixing cell burner with
the limited turbulence dual register (low NOX) burner
Increasing the burner zone by spreading the burners vertically
to include 22 percent more furnace surface
Metering and controlling the airflow to each row of burners
using a compartmented windbox
To provide these changes for NOX control, the price increase was about
$1.75 to $2.50/kW (1977 dollars). If these costs are annualized according
to the format of Section 4.3.3, they translate to 0.28 to 0.40 $/kW-yr.
In addition, Foster Wheeler has performed a detailed design study
aimed at identifying the incremental costs of NOX control to meet 1971
NSPS (Reference 7-4). Foster Wheeler looked at three unit designs with
the following results:
Boiler Design Relative Cost
Unit 1: Pre-NSPS base design 100
Unit 2: Enlarged furnace, no 114
active NOX control
Unit 3: NSPS design; enlarged 115.5
furnace, new burner
design, perforated hood,
overfire air, boundary air
o
Assuming the cost of a pre-NSPS coal-fired boiler to be about
$100/kW in 1969, or $180/kW in 1977 construction costs, the incremental
cost of active NOX controls (NBD plus OFA) is $2.78/kW, or about
$0.44/kW-yr annualized. The Foster Wheeler estimate which includes both
151
-------
NBD and OFA, thus agrees quite well with the Babcock & Wilcox estimate,
which includes only NBD and associated equipment.
Comparing these costs with the retrofit costs (0.40 to 0.70 $/kW-yr
for LNB or OFA) presented in Table 7-7 and considering the better NOX
control anticipated with NSPS units, it is certainly more cost effective
to implement controls on new units. Furthermore, fewer operational
problems are expected with factory installed controls.
7.4 ENVIRONMENTAL ASSESSMENT OF NOX CONTROLS
Modification of the combustion process in utility boilers for NOX
control reduces the ambient levels of N02, which is both a toxic
substance and a precursor for nitrate aerosols, nitrosamines, and
photochemical smog. These modifications can also cause changes in
emissions of other combustion generated pollutants. If unchecked, these
changes, referred to as incremental emissions, may have an adverse effect
on the environment, in addition to effects on overall system performance.
However, since the incremental emissions are sensitive to the same
combustion conditions as NOX, they may, with proper engineering, also be
held to acceptable levels during control development so that the net
environmental benefit is maximized. In fact, control of incremental
emissions of carbon monoxide, hydrocarbons, and particulate has been a key
part of all past NOX control development programs. In addition, recent
control development has been giving increased attention to other potential
pollutants such as sulfates, organics, and trace metals.
Unfortunately, previously developed incremental emissions data for
other than the criteria pollutants are quite scarce. Thus, NOX EA
evaluations to date have relied heavily on combustion fundamentals and
pollutant formation theory to postulate expected changes in emission
levels of noncriteria pollutants with NOX control application. Of
course, NOX EA field testing efforts are underway to resolve these data
insufficiencies. However, results from the utility boiler tests performed
are still not complete.
Table 7-8 shows the cumulative evaluation of the potential effects
of NOX control application on incremental emissions from utility boilers
of CO, vapor phase hydrocarbon (HC), sulfate, particulate, organics (POM),
and trace metals. Entries in the table are based on actual data, where
available, or fundamental hypothesis, where data were insufficient.
As Table 7-8 illustrates, using preferred NOX combustion controls
on boilers should have few adverse effects on incremental emissions of CO,
vapor phase hydrocarbons, or particulates. Indiscriminately lowering
excess air can drastically affect boiler CO emissions, and particulate
emissions can increase with off stoichiometric combustion and flue gas
recirculation. However, with suitable engineering during development and
careful implementation, these incremental emissions problems can be
minimized.
In contrast, applying almost every combustion control has
intermediate to high potential impact on incremental emissions of sulfate,
152
-------
TABLE 7-8. EVALUATION OF INCREMENTAL EMISSIONS DUE TO NOX CONTROLS APPLIED
TO BOILERS
NO Control
^
Low Excess A1r
Staged
Combustion
Flue Gas
Redrculatlon
Reduced A1r
Preheat
Reduced Load
Water
Injection
Ammonia
Injection
Incremental Emission
CO
++
0
0
0
0
0
0
Vapor Phase
HC
0
0
0
0
0
0
0
Sulfate
+
+
+
+
+
+
++
Particulate
0
+
+
0
0
+
+
Organlcs
++
++
+
+
+
+
0
Segregating
Trace Metals
+
4-
f
0
0
0
+
Nonsegregatlng
Trace Metals
0
0
+
+
0
0
0
01
CO
Key: +'+ denotes having high potential emissions impact
+ denotes having intermediate potential emissions impact, data needed
0 denotes having low potential emissions impact
-------
organics, and trace metals. For trace metal and organic emissions,
substantiating data are largely lacking, but fundamental formation
mechanisms cause justifiable concern. In the sulfate case, fundamental
formation mechanisms suggest that these emissions remain unchanged or
decreased with all controls except ammonia injection. However, complex
interactive effects are difficult to clarify, and this pollutant class is
considered sufficiently hazardous to justify some concern in the absence
of conclusive data.
As quantitative data on incremental emissions become available from
the NOX EA field testing program, the impacts of NOX controls on the
environment will be better characterized through application of source
analysis models (References 7-5, 7-6, and 7-7).
7.5 BEST CONTROL OPTIONS
Pending final data resolution of incremental emissions, combustion
modification NOX controls are deemed to be environmentally sound,
cost-effective means of reducing NOX emissions. As discussed in Section
7.3 and summarized in Table 7-7, off stoichiometric combustion and low
NOX burners are the preferred techniques for both retrofit and new
application. For coal firing, OFA or LNB are both cost effective, with
LNB the preferred route for new wall fired boilers. Low NOX burners
still require further full scale demonstration and development before
retrofit application can be considered routine. For more stringent
control, both OFA and LNB may be required. For oil and gas firing, BOOS
is the recommended technique. It is not cost effective to install FGR and
OFA for stringent control for oil and gas.
With proper design and implementation of the recommended controls,
there should be minimal impact, in general, on boiler operation. Long
term operation testing should be continued to confirm this assessment.
154
-------
REFERENCES FOR SECTION 7
7-1. Lim, K. J., ฃt aj_., "Environmental Assessment of Combustion
Modification NOX Controls," Acurex Draft Report TR-78-105, April
1978.
7.2 Manny, E. H., et a\_., "Field Testing of Utility Boilers and Gas
Turbines for Emission Reduction," in Proceedings of the Third
Stationary Source Combustion Symposium; Volume I,
EPA-600/7-79-050a, February 1979.
7.3. Personal communication from E. 0. Campobenedetto, Babcock & Wilcox
Company, Barberton, Ohio, November 1979.
7-4. Vatsky, J., "Effectiveness of NOX Emission Controls on Utility
Steam Generators," Report to Acurex Corporation from Foster Wheeler
Energy Corporation, Livingston, NJ, November 1978.
7-5. Salvesen, K.G., iet aK, "Emissions Characterization of Stationary
NOX Sources: Volume I. Results," EPA-600/7-78-120a, NTIS PB-284
520, June 1978.
7-6. Schalit, L.M., and K.J. Wolfe, "SAM IA: A Rapid Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
EPA-600/7-78-015, NTIS PB-277 088/AS, February 1978.
7-7. Anderson, L.B., et. a]_., "SAM I: An Intermediate Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
Acurex Report TR-79-154, Acurex Corporation, Mountain View, CA,
December 1978.
155
-------
SECTION 8
ENVIRONMENTAL ALTERNATIVES ANALYSIS
As noted in Section 4, impact assessments of three general types
are being performed in the NOX EA:
t Baseline and controlled multimedia environmental assessments of
stationary combustion sources
t Operational and cost impact evaluations of NOX combustion
modification control applications
Systems analysis assessments of applying NOX control
strategies on a regional basis
Second year results of the process evaluations of applying NOX controls
to utility boilers were discussed in Section 7. This section discusses
results to date from applying the methodologies outlined in Section 4 to:
Evaluating baseline combustion source pollutant impact
potential and source ranking based on this evaluation
(Section 8.1)
t Projecting the air quality implications of enforcing various
control strategies on a regional basis and identifying
N0x/hydrocarbon/oxidant control strategy interactions
(Section 8.2)
8.1 BASELINE IMPACT RANKINGS
During the second year, the Source Analysis Model discussed in
Section 4.1.3 (extended SAM I) was used to identify and rank potential
environmental problems from stationary combustion sources due either to
specific pollutants from a single effluent stream or from the entire
source. The model was used to calculate impact potential either for a
single source or the nationwide aggregate of sources considering
population proximity to the source.
Data for use in the model were compiled for source emissions, human
health impact threshold criteria, population densities near source
concentrations, and emission growth rates. Emissions data and growth
projections were discussed in Section 2. Population densities and
156
-------
urban/rural region designations were defined from EPA and Bureau of Census
data. Urban/rural equipment populations and regional fuel consumption
data were obtained from the National Emissions Data System (NEDS).
Multimedia Environmental Goals (Reference 8-1) were used for health impact
threshold criteria. Although these data are not as complete as desired,
they were used with the SAM model to obtain a tentative indication of
potential problem areas. The following list summarizes capabilities of
the SAM model and notes specific cases which were evaluated. Detailed
discussion of data sources and cases evaluated appears in Reference 8-2.
Source Analysis
Model Capabilities
Total nationwide potential
impact factors for specific
source types, considering
population exposure and all
pollutants inventoried for
gaseous effluent streams
Total nationwide potential
impact factors for all
pollutants inventoried for
liquid and solid effluent
streams
Projections of total
nationwide impact factors
Single source, single
pollutant potential impact
not considering population
exposure
Calculations Performed
Total gaseous effluent stream
pollution potential ranking
for 1974
Average gaseous effluent stream
pollution potential ranking
for 1974
Total liquid and
stream pollution
ranking for 1974
solid effluent
potential
Total gaseous effluent stream
pollution potential ranking for
1985 and 2000
Total gaseous effluent stream
pollution potential cross
ranking for 1974, 1985 and 2000
NOX single source pollution
potential ranking for
stationary sources
Pollution potential of single
pollutants from utility boilers,
packaged boilers, gas turbines,
1C engines and industrial
process heating
Additional results are tabulated in the Appendices of Volume II of
Reference 8-2.
Although the potential impact factor results generated in this
study were useful for detecting gross qualitative trends, firm
157
-------
quantitative conclusions were precluded by inadequacies in the data and
the uncertainties in projected energy usage. Key data needs identified
are as follows:
Multimedia source emissions data
Most of the noncriteria pollutant emissions data are for
compound classes or sample fractions; species
concentrations are needed for compound classes showing
pollution potential
-- POM and trace element data are sparse and exhibit large
scatter from different sampling programs. Emissions of
these pollutants are highly dependent on the origin of the
fuel and the specific stationary source and effluent stream
from which the data were obtained.
-- Data on emissions during transient or nonstandard operation
are virtually nonexistent. New tests are needed if these
effects are to be considered.
Liquid and solid emissions data are only quantified for the
utility and large industrial boiler equipment sector.
Although this sector represents the major portion of liquid
and solid pollution potential, further study of packaged
boilers and industrial process heating effluent streams
should be pursued. In addition, the fractions of total ash
which are emitted as bottom ash and flyash vary with boiler
type. However, sufficient data were not available to
estimate this effect.
t Health impact threshold criteria
The Multimedia Environmental Goals (MEG's) employed
(Reference 8-1) are preliminary, and designed for screening
purposes only. They are not ambient standards, but rather
indications of ambient concentrations at which health
effects from continuous exposure should be investigated.
In addition, compounds were not speciated. Since one
health effects value was used to represent an entire
pollutant class, various individual species were not
considered.
Population exposure to source emissions
-- Specific values for average source size and urban/rural
splits were in many cases based on poor quality data. For
utility and large industrial boilers, and most packaged
units, the data were adequate. However, for internal
combustion engines and industrial process heating, data
exhibited a wide range of values making specification
difficult.
158
-------
Given the above qualifications, selected results from the potential impact
factor calculations and projections are discussed briefly below.
The 1974 total pollution potential rankings are shown in
Table 8-1. The table indicates that coal-fired utility and industrial
sources have the largest total potential impact factors. Small stoker
fired boilers rank highest, primarily because:
They have high particulate (trace element) emission factors
(low degree of particulate control application)
They have low stacks
They are located in urban (high population density) regions
Utility boilers rank next highest in the total pollution potential ranking
because of the shear quantity of emissions from these sources.
This point is illustrated in Table 8-2 which shows average source
gaseous pollution potential impact ranking. As indicated, opposed wall
coal-fired boilers have the highest average source pollution potential.
This potential impact value was obtained by dividing the total impact
factor by the total number of sources of a specific equipment type and
thus represents a measure of the impact of a single typical source.
Opposed wall fired units are used in the larger capacity ranges (>400 MW
electric output). Thus because of their large size and resulting high
fuel consumption, opposed wall boilers have a high average source
pollution, potential. However, this result must be used with care since
the ranking is not normalized for energy consumption. For example, a 600
MW (electrical output) opposed wall fired boiler may have less pollution
potential than three 200 MW (electrical output) single wall fired boilers
required to supply the same power. This ranking is primarily intended to
assess characteristic average source impacts. Stokers are lower in the
ranking because their impact is a result of many smaller sources rather
than a few large single sources.
Table 8-3 shows the results of potential impact calculations
considering NOX emissions only. The table illustrates that cyclone
boilers have the highest single source NOX impact. This is primarily
because uncontrolled NOX emissions from cyclone (coal-fired) boilers are
more than double the emissions from tangential units and about 75 percent
higher than from wall fired units. However, the total nationwide
pollution potential of cyclone boilers should decline in the future since
the use of this unit type is projected to decrease.
8.2 AIR QUALITY PROJECTIONS
The goals of the systems analysis task of the NOX EA, and the
methodologies assembled were described in Section 4.3. In this section,
results from applying the preliminary model and the two advanced models
are presented. Results in three main areas are discussed. The
preliminary model with source weighted rollback was used to examine NOX
159
-------
TABLE 8-1. TOTAL POLLUTION POTENTIAL RANKING (GASEOUS)
STATIONARY SOURCES IN YEAR 1974
en
o
Rank
1
2
3
4
5
6
7
6
9
10
11
12
13
14
15
16
17
Sector
Packaged Boilers
Packaged Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Packaged Boilers
Equipment Type
Stoker Firing WTC <29 MWa
Stoker Firing FTd <29 MWa
Tangential
Wall Firing
Wall Firing WTฐ >29 MWa
Stoker Firing WTC >29 MWa
Vertical & Stoker
Cyclone
Horizontally Opposed
Tangential
Wall Firing
Horizontally Opposed
Wall Firing WT0 >29 MWa
Scotch FTd <29 MWa
Firebox FTd <29 MWa
Tangential
Scotch FT*
Fuel
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Oil
Oil
011
Oil
Oil
Oil
Gas
Gas
Total Impact -Factor
6.73 x 1011
5.59 x 1011
1.42 x 1011
1.09 x 1011
7.78 x 1010
7.64 x 1010
5.69 x 1010
4.12 x 1010
2.10 x 1010
2.65 x 109
2.22 x 109
1.13 x 109
7.02 x 108
5.50 x 10ฎ
3.64 x 108
3.20 x 108
2.88 x 108
T-6ZZ
-------
TABLE 8-1. Concluded
Rank
18
19
20
21
22
23
24
25
26
27
28
29
30
Sector
Ind. Process Comb.
Reciprocating 1C
Engines
Packaged Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Ind. Process Comb.
Packaged Boilers
Gas Turbines
Ind. Process Comb.
Equipment Type
Coke Oven Underfire
SIe >75 kW/cylb
Single Burner WTฐ <29 MWa
HTR Boiler <29 MWa
Brick & Ceramic Kilns
Horizontally Opposed
Wall Firing
Cyclone
Wall Firing WT0 >29 MWa
Cement Kilns
Cast Iron
Simple Cycle >15 MWb
Refinery Htr. Nat. Draft
Fuel
Process Material
Gas
Oil
Oil
Process Material
Gas
Gas
Oil
Gas
Process Material
Oil
Oil
Gas
Total Impact Factor
2.84 x 108
2.30 x 108
2.28 x 108
2.25 x 108
2.01 x 108
1.61 x 108
1.28 x 108
1.27 x 108
2.72 x 107
2.71 x 107
2.47 x 107
2.39 x 107
2.22 x 107
aHeat Input
bHeat output
cWatertube
dFiretube
eSpark ignition
-------
TABLE 8-2. AVERAGE SOURCE POLLUTION POTENTIAL RANKING (GASEOUS)
STATIONARY SOURCES IN YEAR 1974
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Sector
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Packaged Boilers
Packaged Boilers
Utility Boilers
Packaged Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Packaged Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Equipment Type
Horizontally Opposed
Cyclone
Tangential
Wall Firing
Mall Firing HT0 >29 HWa
Stoker Firing WT0 <29 HWa
Stoker Firing WT0 >29 MHa
Vertical and Stoker
Stoker Firing FT* <29 MWa
Horizontally Opposed
Tangential
Cyclone
Wall Firing
Horizontally Opposed
Wall Firing WT0 >29 HW8
Wall Firing
Tangential
Cyclone
Fuel
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Oil
Oil
Oil
Oil
Oil
011
Gas
Gas
Gas
Average Impact Factor
4.26 x 108
3.52 x 108
3.11 x 108
1.76 x 108
1.21 x 108
8.45 x 107
8.35 x 107
7.34 x 107
2.29 x 107
1.52 x 107
1.39 x 107
3.27 x 106
2.21 x 106
1.76 x 106
7.71 x 105
2.49 x 105
1.54 x 105
9.55 x 104
ro
T-621
-------
TABLE 8-2. Concluded
CT>
CO
Rank
19
20
21
22
23
24
25
26
27
28
29
30
Sector
Gas Turbines
Ind. Process Comb.
Ind. Process Comb.
Gas Turbines
Packaged Boiler
Packaged Boiler
Ind. Process Comb.
Ind. Process Comb.
Ind. Process Comb.
Packaged Boilers
Ind. Process Comb.
Packaged Boilers
Equipment Type
Simple Cycle >15 HWb
Refinery Htr. Nat. Draft
Refinery Htr. Forced Draft
Simple Cycle >15 HUb
Wall Firing WT0 >29 MWa
Single Burner MTฐ <29 MWd
Refinery Htr. Forced Draft
Refinery Htr. Nat. Draft
Coke Oven Under fire
Scotch FTd <29 MHa
Cement Kilns
Scotch FTd <29 MWa
Fuel
Oil
011
Oil
Gas
Gas
Oil
Gas
Gas
Process Material
Gas
Process Material
Oil
Total Impact Factor
8.70 x 104
6.60 x 104
5.81 x 104
5.80 x 104
5.26 x 104
3.21 x 104
2.73 x 104
2.09 x 104
1.92 x 104
1.26 x 104
l.?4 x 104
1.20 x 104
T-621
aHeat input
bHeat output
cUatertube
dF1retube
-------
TABLE 8-3. N0ฅ POLLUTION POTENTIAL RANKING STATIONARY SOURCES IN 1974 (N0? BASIS)
/\ ฃ
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Sector
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Utility Boilers
Gas Turbines
Gas Turbines
Ind. Process Comb.
Equipment Type
Cyclone
Horizontally Opposed
Horizontally Opposed
Horizontally Opposed
Cyclone
Tangential
Horizontally Opposed
Tangential
Tangential
Hall Firing
Tangential
Wall Firing
Wall Firing
Cyclone
Simple Cycle >15 MWb
Simple Cycle >15 MWb
Refinery Htr. Forced Draft
Fuel
Bituminous
Lignite
Gas
Bituminous
Lignite
Bituminous
Oil
Lignite
Gas
Lignite
Oil
Bituminous
Gas
Gas .
Oil
Gas
Oil
NO Impact Factor
4.97 x 109
3.40 x 109
2.80 x 109
2.78 x 109
2.44 x 109
9.82 x 108
9.21 x 108
8.22 x 108
3.79 x 108
2.88 x 108
2.55 x 108
2.43 x 108
2.30 x 108
1.37 x 108
1.24 x 108
5.30 x 107
5.14 x 107
T-613
-------
TABLE 8-3. Concluded
CTi
CJ1
Rank
IS
19
20
21
22
23
24
25
26
27
28
29
30
Sector
Utility Boilers
Utility Boilers
Ind. Process Comb.
Packaged Boilers
Packaged Boilers
Ind. Process Comb.
Packaged Boilers
Ind. Process Comb.
Packaged Boilers
Reciprocating 1C
Engines
Reciprocating 1C
Engines
Packaged Boilers
Reciprocating 1C
Engines
Equipment Type
Wall Firing
Cyclone
Refinery Htr. Nat. Draft
Hall Firing WT0 >29 MWa
Wall Firing WTC >29 MWa
Refinery Htr. Forced Draft
Wall Firing WTC >29 MWa
Refinery Htr. Nat. Draft
Stoker Firing WTฐ >29 MWa
Cle >75 kW/cylb
SIf >75 kW/cylb
Stoker Firing WTC <29 MWa
CIe >75 kW/cylb
Fuel
Oil
Oil
Oil
Oil
Bit./Lig. Coal
Gas
Gas
Gas
Bit./Lig. Coal
Oil
Gas
Bit./Lig. Coal
Dual (Oil + Gas)
NO Impact Factor
4.81 x 107
4.07 x 107
3.89 x 107
2.59 x 107
2.59 x 107
2.45 x 107
2.25 x 107
1.26 x 107
6.00 x 106
4.09 x 106
3.51 x 106
2.47 x 106
1.97 x 104
T-613
"Heat Input
bHeat output
cWatertube
dFiretube
Compression ignition
Spark ignition
-------
control needs for eight AQCR's for a variety of growth and source
weighting cases. The Eulerian photochemical model, LIRAQ, was used to
investigate one hour N0ฃ levels for the San Francisco AQCR. And,
specific questions related to stack height effects and urban sprawl were
examined with a photochemical trajectory model.
These results are not intended to indicate what specific sources
should be controlled but, rather, are used to examine the impact of
various levels of control. Control needs must be specifically
investigated on an individual AQCR basis.
8.2.1 Preliminary Model Results
Over 20 different emissions growth/source weighting combinations
for eight AQCR's, listed in Table 8-4, were considered. The eight AQCR's
were selected to represent a variety of source category, fuel use, and
mobile/stationary source mixes. Results for Chicago and Los Angeles were
discussed in Reference 8-3. Results for two additional AQCR's, St. Louis
and San Francisco, and composite results are presented below.
TABLE 8-4. AQCR's INVESTIGATED WITH PRELIMINARY MODEL
Los Angeles (024)
Chicago (067)
Philadelphia (045)
New York City (043)
Denver (036)
San Francisco (030)
Pittsburgh (197)
St. Louis (070)
Low ~ Recorded
Annual Average N02,
1972-1975
(vg/m3)
132
96
83
99
88
76
62
76
High Rolling
Quarter Average N02,
1972-1975
(yg/m3)
182
121
121
113
110
101
98
85
Composite Results
The variety of emissions source growth scenarios, described in
Section 4.3, resulted in predicted uncontrolled* NOX emissions changes
*No stationary source controls beyond 1971 NSPS. Each scenario does
assume a specified level of mobile source control.
166
-------
relative to 1973 emissions of -6 percent to +3 percent in 1985 and of
+5 percent to +50 percent in 2000. The relatively small spread in 1985
emissions reflects the impact of currently planned mobile source
controls. The very large spread in 2000 reflects the different projected
impacts from low stationary source growth with very strict mobile control
and high stationary source growth (~3 percent per year) compounded over 28
years.
Changes in ambient concentration corresponding to the above
emissions changes, ranged between -12 and +3 percent for 1985 and zero to
+43 percent in the year 2000. Since these results are from a variety of
AQCR's, they are representative of the range of expected change in ambient
concentrations for all AQCR's. These calculated changes in ambient
concentration do not exactly follow the percent changes in emissions since
the use of the source weighting factors in the preliminary model can
reduce or increase the impact of emissions growth of selected sources.
In the present analysis, the most significant use of source
weighting factors was to reduce the relative impact of powerplants to 20
percent of their emissions. This was to account for the dispersion of
powerplant emissions prior to impact on urban receptors, and represents a
reasonable lower limit for powerplant impact. Furthermore, as discussed
in Section 3, area sources are thought to have a more significant impact
on annual average N02 levels than do powerplants, thus use of a
weighting factor less than unity for powerplants would seem appropriate.
The level of NOX control required to offset the increase in
ambient concentration depends on the initial value (1973) of the annual
average NOg concentration. As discussed in Section 3, the current
number of nonattainment (for NOg) AQCR's may be between 4 and 30. A
conservative estimate is that at least four AQCR's will need significant
application of combustion modification NOX controls to attain the annual
average N02 standard by 1985. By the year 2000 this number will
conservatively increase to 15. (Assuming an average 25 percent increase
in ambient concentration*, any AQCR with a 1973 annual average greater
than 80 yg/m^ would exceed 100 yg/m^ by 2000.) In addition, one-half
of these would also need implementation of advanced controls such as
ammonia injection and possibly flue gas treatment.
It should be emphasized that these conservative estimates are used
to compensate for the extreme uncertainty in the monitoring data and the
inherent errors in the assumptions of the model. It should also be noted
that conservative growth rates and a successful mobile control program are
"built-in" to the estimates. All of this is to say that the above should
*A 25 percent increase in annual average N02 level by the year 2000 is
most representative of expected changes. The zero percent lower limit
referred to above applies only to heavily mobile dominated AQCR's with
an extremely effective mobile control program.
167
-------
be considered an optimistic view of future NOX control needs for the
compliance with the current annual average standard.
San Francisco, AQCR 030
The San Francisco AQCR is representative of those AQCR's which are
not currently nonattainment areas for NO? but which could become so in
the future based on the current annual average air quality standard.
Emissions of NOX in the region are heavily dominated by mobile sources
(-70 percent), and the region is a nonattainment area for oxidant.
Furthermore, if a one hour N02 standard of 500 yg/m3 were promulgated,
San Francisco would be in violation. An additional reason for presenting
results for San Francisco is for comparison with the photochemical model
results discussed in Section 8.2.2.
Table 8-5 shows the results of the matrix of calculations performed
for San Francisco. The growth scenarios are described in Section 4.3 and
Reference 8-3 and are briefly defined at the bottom of the table. Both
the high and low base year concentrations (Table 8-4) are used for the
calculations, and results for 1985 and 2000 are shown for each case. The
required NOX control levels, indicated by 0, 1, 2, 3 and V, are
described in Table 8-6. Controls are applied in the most cost-effective
manner*, and new controls are introduced, if required, at the time (year)
they are assumed to be developed.
The results shown in Table 8-5 are representative of mobile source
dominated AQCR's with 1973 concentrations in the range of 75 to
100 yg/m3- The dominance of mobile sources is clearly indicated by the
significant reduction in predicted 1985 concentrations in all cases and by
the results for the low mobile cases in both 1985 and 2000. The results
for the high base year concentrations (BYR = 101 yg/m3) show that a
mobile dominated AQCR that is near nonattainment now will need the maximum
amount of combustion modification NOX control by the year 2000.
Furthermore, if powerplants are not a significant contributor to the
annual average (PP = 0.2), then even more stationary source NOX control
will be required.
St. Louis. AQCR 070
St. Louis is representative of an industrialized region dominated
by stationary sources (75 percent) with coal firing as a significant
source of NOX (72 percent of stationary source NOX is from coal-fired
sources). Results for St. Louis are shown in Table 8-7. The relatively
minor impact of different mobile source growth and control scenarios is
illustrated by comparing the nominal growth and low mobile cases.
Similarily, high stationary growth begins to have significant impact by
*When required, a control type is applied to all sources in a given source
category.
168
-------
TABLE 8-5. SUMMARY OF CONTROL LEVELS REQUIRED TO MEET THE ANNUAL AVERAGE
N02 STANDARD IN SAN FRANCISCO, AQCR 030
IO
Case
Nominal Growth6
Low Mob11ef
High StationaryQ
BYRa = 76
PPฐ = 1.0 PP = 0.5 PP = 0.2
MSC = 1.0 MS = 1.2 MS = 1.0
0(67)
0(64L
'0(80)
0(65).
0(68)
)(82)
BYR = 101
PP = 1.0 PP = 0.5 PP = 0.2
MS = 1.0 MS = 1.2 MS = 1.0
0(79),
'(M 88)
0(91)
0(75L
'0(84)
T-1293
Base year ambient concentration for calibration
bpp Powerpi ant weighting factor
CMS Mobile source weighting factor
dNumbers in parentheses indicate annual average concentration in ^g/m3. if no number
given, annual average equals or exceeds 100 (jg/m3.
eStationary source growth less than historical, mobile sources grow at 3.5X/yr, 0.62 g/km in 1981
fStationary source growth less than historical, mobile sources grow at 1.0%/yr, 0.25 g/km in 1985
9Stationary source growth at approximately historical rates, mobile sources grow at 3 5X/vr 0 62
g/km in 1981 ' '
-------
TABLE 8-6. DEFINITION OF STATIONARY SOURCE NOX CONTROL LEVELS
Level3
0 No controls (assumes 1971 NSPS for large
boilers is met)
40-80 percent control of new residential and
commercial furnaces
t 6-16 percent control by low excess air for
industrial and utility boilers
t Off stoichiometric combustion, flue gas
recirculation, low-NOx burners and other
advanced designs for boilers
t Operating adjustments and new design for 1C
engines
Water injection and new designs for gas turbines
Ammonia injection for boilers (50 percent
reduction in NOX beyond Level 2 controls)
V t Combustion control limits exceeded, flue gas
treatment required
aControl levels are ordered by increasing cost of the controls.
Within each level controls are applied in order of increasing cost.
170
-------
TABLE 8-7. SUMMARY OF CONTROL LEVELS REQUIRED TO MEET THE ANNUAL AVERAGE
N02 STANDARD IN ST. LOUIS, AQCR 070
Case
Nominal Growths
Low Nobilef
High Stat1onary9
BYRซ = 76 pg/mS
ppb ป l.o PR = 0.5 PP - 0.2
MSC = 1.0 MS = 1.2 MS = 1.0
0(77)d/
y&(96)
0(80) S
/*
wy
/0(98)
v
/
0(79) /
/'
0(76J/
/ '
/^
0(70)//
y6(84)
/
0(80) /
/ 2
BYR => 85 ug/m3
PP ป 1.0 PP ป 0.5 PP = 0.2
MS =1.0 MS - 1.2 MS = 1.0
0(86)/
/ 2
I/
0(90) /
/3
0(84j/
/*
S
' /
0(88) /
/'
0(85) /
/*
s <
wy
/0(9t)
/
0(92) /
/'
T-1292
ซBYR Base year ambient concentration for calibration
bpp Powerpiant weighting factor
CMS Mobile source weighting factor
lumbers 1n parentheses Indicate annual average concentration In pg/m3. If no number
given, annual average equals or exceeds 100 pg/m3.
^Stationary source growth less than historical, mobile sources grow at 3.5t/yr, 0.62 g/km In 1981
fStationary source growth less than historical, mobile sources grow at l.OX/yr, 0.25 g/km in 1985
9Stationary source growth at approximately historical rates, mobile sources grow at 3.5X/yr, 0 62
g/km In 1981
-------
the year 2000, even for the low base year concentration (76yg/m3). The
difference in control requirements between mobile and stationary source
dominated AQCR's is further illustrated by comparison of the low base year
cases for St. Louis and San Francisco. This base year concentration is
the same, but the ambient levels or control requirements in the year 2000
are much different.
A significant feature of these results, which are typical of
stationary source dominated AQCR's, is that although no NOX controls are
needed in 1985, considerable NOX control is required by the year 2000,
even though the present ambient level is well below the annual average
standard. It is expected that a one hour N02 standard less than
500yg/m3 would necessitate even further NOX controls.
8.2.2 LIRAQ Results*
The Association of Bay Area Governments has sponsored an extensive
study of predicted future air quality in the San Francisco AQCR to provide
the air quality maintenance portion of an environmental management plan
for the San Francisco Bay Region. A variety of source growth and control
scenarios has been examined using the LIRAQ photochemical model, described
in Section 4.3.2. Some of the details and results of the study are
presented in Reference 8-4. Of primary interest to the NOX EA are the
various NOX/HC control strategies and their impact on both N02 and
03 ambient levels.
All the model calculations obtained were made for the meterology of
a high oxidant day, since oxidant is presently the most pressing air
quality issue in the San Francisco AQCR. For the purposes of the NOX EA
a high NO;? day would be preferred; the meteorology of a high N0ฃ day
is significantly different from a high oxidant day. However, the primary
use of the results was to make relative comparision of the effects of
control strategies rather than specific predictions of one hour N02
maximum. That is, the percent change in the predicted area wide one hour
maximum value with respect to percent reductions in HC and NOX emissions
was examined. This is perhaps the best way to use the high oxidant day
results to infer the impact on N02 one hour peaks, 24 hour averages, and
annual averages.
The cases considered were the following combinations of percent HC
reduction and percent NOX reduction: 80/0, 40/0, 40/20, 80/40, and
0/40. In all cases the emissions reductions were applied uniformly
throughout the region. Results from these cases were used to examine the
effect of NOX and HC controls, separately and in combination, on the
region wide one hour maximum of N02 and ozone. The impact of these
*LIRAQ output was provided by Mr. Lewis Robinson, of the Bay Area Air
Pollution Control District, from work performed for the Association of
Bay Area Governments.
172
-------
control strategies on the annual average N02 concentration was also
inferred from the results. Of particular interest were the interactive
effects of the control strategies.
Tables 8-8 and 8-9 show the percent reduction in one hour peak
values of N0ฃ and 03 respectively for these cases. For a 40 percent
reduction in NOX the peak N0ฃ value was reduced only 14 percent. This
is a very different result from what would be expected from a rollback
model (40 percent). However, several additional factors need to be
considered.
The location of the predicted maximum one hour N02 level in the
region shifted in both time and Tocation. At the location of the original
peak value a 20 percent reduction in peak N02 occurred. For the
controlled case the peak value occurred earlier in the day (by about an
hour) and decayed faster. The N02 concentration behavior at the
location of the original N02 peak is illustrated in Figure 8-1. At this
location concentrations throughout most of day are 30 to 50 percent less
than the uncontrolled case. The 24 hour average N02 concentration is
reduced by 35 percent, which indicates that the reduction in annual
average would probably be much closer to 40 percent than that suggested by
the peak reduction.
Figure 8-1 also shows the time history of N02 concentration for
the case of 20 percent NOX reduction and 40 percent HC reduction. The
N02 peak is reduced by 35 percent; whereas, reductions for other times
are much closer to the 20 percent NOX emissions reduction. The 24 hour
average N02 concentration is reduced by 25 percent, which is more
representative of the impact of the annual average than is the reduction
in peak value. This clearly illustrates the influence of HC controls on
the N02 peak, and the difference in control strategies that may be
required to meet annual average one hour standards.
The interactive effects of NOX and HC controls on maximum one
hour ozone concentrations are illustrated by the results shown in
Table 8-9. In all cases, control of NOX increases the peak ozone levels
and partially offsets the gains from HC control. However, such a
combination control strategy will be necessary to meet both N02 and
ozone ambient goals.
In conclusion, these results support the use of a rollback type
model to examine the effect of controls on the annual average N02 if the
controls are applied uniformly throughout the region of interest. It is
also apparent from the difference in the response to controls of the one
hour peak and the annual average (as implied by the 24 hour average) that
the ratio of one hour pe^ak to annual average may significantly change as a
result of large scale control implementation programs. Therefore, care
should be taken in using values of this ratio calculated from existing
monitoring data to represent future conditions. This difference in
response also indicates that NOX control requirements may be different
for meeting annual average or one hour standards. Furthermore, the
combination of NOX and HC control strategies is very significant to
173
-------
TABLE 8-8. EFFECTS OF N0y AND HC REDUCTION ON ONE HOUR PEAK VALUE OF N02
% NOX Reduction
% HC Reduction
0
40
80
0
30b
55
20
__a
35
40
14
60
aData not available
bpercent reduction in one hour N0ฃ peak
TABLE 8-9. EFFECT OF NOX AND HC REDUCTION ON ONE HOUR PEAK VALUE OF 03
% NOX Reduction
% HC Reduction
0
40
80
0
55b
80
20
__a
36
40
(33)c
70
aData not available
bpercent reduction in one hour 03 peak
C0zone peak increased by 33 percent
174
-------
OJ
o
OJ
CD
XJ
Ol
(O
o
Baseline, uncontrolled
40% HC/20% N0v reduction
A
40% NOV reduction
X
10 12
Time of Day
Figure 8-1. Comparison of time history of NO^ concentration at the San Jose station
for various HC/NO reductions.
X
-------
meeting both a short term and annual average N02 standard and an oxidant
standard.
8.2.3 Photochemical Trajectory Model Results
A photochemical trajectory model, described in Section 4.3.2, was
selected to investigate four specific issues regarding NOX controls:
The influence on peak N02 levels of a large powerplant upwind
of a metropolitan area
t The influence of the stack height of the powerplant for the
above case
The effectiveness of NOX control applied to the powerplant
compared to the effectiveness of area source control on the
peak N02 level
t The impact of simultaneous urban sprawl and area source control
on peak N02 concentrations
For the purposes of the present investigations, a trajectory
through downtown Los Angeles on the third day of an N02 episode was
selected as representative of severe N02 meteorological conditions.
This trajectory is shown in Figure 8-2. The location of the air mass of
interest at each hour of the day from 1 am to 2 pm is shown on the
trajectory map. In those cases with a powerplant on the trajectory, it
was located at 2 am on the trajectory. It should be emphasized that this
trajectory was selected to provide a representative case. No further
relationship to specific sources on the Los Angeles area should be
inferred.
Meteorological conditions (wind speeds, mixing heights, etc.) and
emission fluxes were prepared by Environmental Research and Technology
Inc. (ERT) from data available for Los Angeles. Meteorological conditions
were characterized by:
High, 50 kPa pressure surface heights
Low wind speeds both at the surface and aloft
Weak offshore surface pressure gradient
Fog, low clouds, or haze along the coast during night and
morning hours
Early morning inversion height of 213 m, lifting to 700 m by
mid-morning
A series of computer simulations using these meteorological
conditions and several combinations of source emission strengths
(corresponding to growth and/or control) were made. It should be
emphasized that the purpose of these runs was not to verify the model for
this day, but to provide a relative comparison of the air quality for
176
-------
14 / Pasadena
Santa Monica
Mountains
Los
Anqeles
October 16, 1973
Trajectory
Puente Hills
Chi no
Hills
Powerplant added here
Torrance
Redondo
Beach
Huntington
Beach
Figure 8-2. Map of an air parcel trajectory for high NCL day.
-------
different emissions patterns. No powerplant was on the trajectory;
therefore, for the purpose of simulation, one was added as described
above. The following cases were examined:
t Baseline emissions, no large point sources
Baseline emissions with a large powerplant situated on the
early part of the trajectory, emissions from a 200 m stack
Baseline emissions, with a large powerplant at ground level
Reduction of 40 percent in NOX, 50 percent in HC from
baseline case (no powerplant emissions)
t Baseline emissions with increased area sources in early portion
of trajectory (no powerplant emissions)
The results presented here are somewhat restricted in that only
limited meteorological conditions and initial conditions were examined.
(Initial conditions were rolled back for those calculations where
emissions levels were reduced). The results from all the calculations are
summarized in Table 8-10. They are presented in terms of the change in
the 9 am to 12 noon maximum N02 concentration relative to the baseline
case. The 9 am to 12 noon maximum was used since the N0ฃ one hour
maximum usually occurred within this time period and the air parcel was in
the downtown Los Angeles area (the location of the measured N02 maximum
for the day) during this time period.
The effects of of a large upwind point source are shown by
comparing the first two cases in Table 8-10. As shown in Figure 8-2, a
large powerplant is located near the coast, approximately 10 hours upwind
of the downtown Los Angeles monitoring station (DOLA). The total
emissions introduced into the air mass from the powerplant are 267 kg*
(assumed to be 90 percent NO) compared to total area source NOX
emissions prior to 10 am of 55 kg and prior to noon of 88 kg. The
presence of the powerplant increases the morning maximum by 30 percent,
regardless of whether the emissions are introduced at the surface or at
200 m. Examination of the time history of the surface concentrations of
the air mass reveals that about 4 hours downwind of the powerplant there
is almost no difference between the two cases. This, at first surprising,
result is a consequence of the low early morning inversion height (213 m)
which does not begin to break up until 9 am. The powerplant emissions are
trapped below the inversion layer and diffuse vertically to fill the
entire layer. Therefore, the height of the emissions source has almost no
impact on the morning maximum value. This situation (a trapped plume)
represents a "worst case" for the impact of elevated emissions.
The influence of NOX control on the powerplant can be
approximated by comparing the baseline and the two powerplant cases
*A11 emissions are expressed as N02.
178
-------
TABLE 8-10. RESULTS OF THE PHOTOCHEMICAL TRAJECTORY MODEL CALCULATIONS
Percent Change in 9 am to 12 noon
Max NO? Concentration from
Baseline Case with no Powerplant
Powerplant at 200 m
Powerplant at surface
40% NOX, 50% HC reduction
40% NOX reduction
Increased upwind emissions
Increased upwind emissions
with 40% NOX, 50% HC
reduction
+ 30
+ 30
- 45
- 30
+ 17
- 35
179
-------
described above except that the baseline case is now treated as the
complete NOX control case. Complete NOX control of the powerplant,
i.e., 83 percent reduction in upwind emissions (267/322), would result in
a 23 percent reduction (30/130) in peak N02. This is not really
surprising since emissions from this major NOX source have ample time to
disperse prior to the point of maximum ground level impact and since a
significant portion of the N02 concentration comes from initial N02
levels in the air mass*. However, this is not the only consequence of
controlling the powerplant. The concentrations of NOg aloft are also
reduced by approximately 23 percent. This serves to reduce afternoon
N02 by as much as 50 percent. Furthermore, in cases where emissions are
carried to sea by late night winds and returned to land in the morning,
the presence of the powerplant could greatly change the early morning
initial concentrations in the returning air mass, which is known to have a
significant impact on N02 concentrations later in the day.
The impacts of both NOX and HC controls on the N02 peak
concentration are illustrated by the two emissions reduction cases, 40
percent NOX and 50 percent HC, and 40 percent NOX only, shown in
Table 8-10. Combined NOX and HC controls are more effective in reducing
the N02 peak than control of NOX only (Table 8-10, 45 percent
reduction compared to 30 percent reduction**). It should also be noted
that the late afternoon ozone peak was reduced by 40 percent for the
combined NOX/HC reductions but was increased by almost 40 percent with
only NOX reduction. This further emphasizes the necessity of combined
NOX/HC control strategies.
The impact of urban sprawl with and without controls is shown by
comparing the baseline case and the last two cases in Table 8-10. In the
first of these cases, NOX and HC emissions in the pre-1 am portion of
the trajectory*** were assumed to increase by 20 percent of the prenoon
emissions in the baseline case. This was used to simulate extensive urban
growth into rural areas or increased industrialization of previously
underdeveloped areas. The second case has 40 percent reduction in NOX
and 50 percent reduction in HC applied to this first case. This
represents outward growth combined with controls.
* Another factor which must also be considered is the percent conversion
to N02 of the NO from the powerplant and the area sources.
** Control of NOX only was more effective in this simulation than in the
LIRAQ one because here the initial concentration of N02 which
numerically represents 30 percent of the morning is also reduced
by 40 percent maximum value. Therefore 12 percent of the
reduction in the morning peak results from the change in initial
conditions and the remaining 18 percent from reduction in emissions.
This compares well with the LIRAQ results shown in Table 8-8.
***The fact that the trajectory is over the ocean prior to 1 am is of
no consequence at this point since we are not interested in the
specifics of the trajectory.
180
-------
Figure 8-3 shows the percent change in the 9 am to 12 noon NO;?
maximum as a function of the percent change in the prenoon NOX emissions
for cases where the HC emissions are also changed. These cases represent
sprawl, sprawl and control, and control only. The approximate linearity
shown in Figure 8-3 suggests that the morning maximum N02 is primarily a
function of the total prenoon emissions (and the initial conditions) and
is not significantly influenced by the exact upwind distribution of those
emissions. Thus growth in emissions either by sprawl or increased density
can be offset by an equivalent emissions reduction by control. This
linearity also lends support to the use of rollback to approximate changes
in ambient concentration.
Before concluding, two very important facts should be noted
relative to the above discussion of the results shown in Figure 8-3.
First, there was no horizontal dispersion of the upwind emissions since it
was assumed that growth occurred uniformly. This is much different than
if a single point source were located on the trajectory. Second, HC
emissions were adjusted in 'nearly the same ratio as NOX emissions in all
cases. The significance of such simultaneous NOX/HC reduction to the
results has been previously discussed.
Based on these limited calculations, the following conclusions can
be made:
t Stack height is not a significant factor on far downwind
concentrations for low inversion heights
The mid-morning N02 peak is a linear function of upwind NOX
emissions provided the ratio of NOX to HC emissions is
constant; therefore, use of a rollback model is justified in
these cases
Stringent control of NOX emissions from a large point source
located several hours upwind does not have a proportional
effect on maximum N02 concentration
Combined NOX/HC reduction has much more impact on the N02
peak than NOX control only
Additional calculations are planned to investigate the impact of
changing the initial conditions and meteorology. Calculations
representative of St. Louis may also be made since the model has been
applied to St. Louis and meteorological and emissions data have been
prepared.
181
-------
CM
O
c
o
o
CM
O
HI
cn
c
o
CJ
o
-------
REFERENCES FOR SECTION 8
8-1. Cleland, J. G., and G. L. Kingsbury, "Multimedia Environmental
Goals for Environmental Assessment," EPA-600/7-77-136a,b,
NTIS PB-276 919/AS, November 1977.
8-2. S.alvesen, K. G., et jH_., "Emissions Characterization of Stationary
NOX Sources: Volume I. Results," EPA-600/7-78-120a, NTIS PB-281
520, June 1978.
8-3. Mason, H. B., et a]_., "Preliminary Environmental Assessment of
Combustion Modification Techniques: Vol. II, Technical Results,"
EPA-600/7-77-119b, NTIS PB-276 68I/AS, October 1977.
8-4. "Environmental Management Plan for the San Francisco Bay Region --
Draft Air Quality Maintenance Plan," prepared by the Association of
Bay Area Governments, the Bay Area Air Pollution Control District,
and the Metropolitan Transportation Commission, December 1977.
183
-------
SECTION 9
TECHNOLOGY TRANSFER
Results of the first year of the NOX EA were previously
documented in the first annual report on the program (Reference 9-1).
Results and findings from various second year task efforts were discussed
in detail in the preceeding sections of this report. Of course, a key aim
of this program is to ensure that these conclusions are widely
disseminated to the pollution control development, industrial user, and
regulatory communities at large. Therefore, in addition to this report,
program outputs were documented in several other reports issued in the
past year. These include:
Results of the Emissions Characterization Task to determine the
baseline multimedia emissions from stationary combustion
sources, to rank sources by pollution potential, and to
establish priorities for developing and implementing NOX
controls (Reference 9-2)
Results from the Alternate Clean Fuels task to evaluate the
differential environmental impact and costs of potential
alternate clean fuels use in area sources (Reference 9-3)
Results from the Process Engineering and Environmental
Assessment task to evaluate the operational, cost, and
environmental impacts of applying combustion modification NOX
controls to utility boilers (Reference 9-4)
Results of the Source Analysis Modeling task to develop rapid
and intermediate screening models (References 9-5 and 9-6)
Further technology transfer activities were performed as part of
the General Program Support task of the NOX EA. Perhaps the most
significant among these was the preparation of a Standards Support Plan
for Stationary Conventional Combustion Sources (Reference 9-7). This
report documents the activities of each EPA office responsible for
providing regulatory support information (Office of Research and
Development), and for setting standards (Office of Air Quality Planning
and Standards, Office of Water Planning and Standards, Office of Toxic
Substances, and Office of Enforcement). The focus of the report is on
documenting the schedules which show the interrelationships among
regulatory mandates, EPA Program Office plans, and the IERL R&D plans.
184
-------
Other general program support technology transfer activities
included:
Continuing publication of "NOX Control Review," a quarterly
technology status report on NOX control development and
implementation, and regulatory strategy
t Coordination of the Second Stationary Source Combustion
Symposium, held in New Orleans in August 1977, and publication
of the symposium proceedings (Reference 9-8)
t Presentation of the keynote paper at the above symposium
Updated documentation of the status of IERL developmental
programs in combustion modification controls, for use in the
IERL annual report (Reference 9-9)
Participation in various IERL-EACD environmental assessment
coordination meetings
Support of the Conventional Combustion Environmental Assessment
efforts
185
-------
REFERENCES FOR SECTION 9
9-1. Waterland, L. R., et a]_., "Environmental Assessment of Stationary
Source NOX Control Technologies First Annual Report,"
EPA-600/7-78-046, NTIS PB-279 083/AS, March 1978.
9-2. Salvesen, K. G., et _al_., "Emissions Characterization of Stationary
NOX Sources," Vols. 1 and 2, EPA-600/7-78-120a,b, NTIS PB-284
520, June 1978.
9-3. Shimizu, A. B., "Identification and Characterization of Clean Fuels
for Area Sources," Acurex Final Report TR-77-57, Acurex
Corporation, Mountain View, CA, April 1978.
9-4. Lim, K. J., et al., "Environmental Assessment of Utility Boiler
Combustion Modification NOX Controls," Acurex Draft Report
TR-78-105, Acurex Corporation, Mountain View, CA, April 1978.
9-5. Schalit, L. M., and K. J. Wolfe, "SAM IA: A Rapid Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
EPA-600/7-78-015, NTIS PB-277 088/AS, February 1978.
9-6. Anderson, L. B., et aK, "SAM I: An Intermediate Screening Method
for Environmental Assessment of Fossil Energy Process Effluents,"
Acurex Report TR-79-154, Acurex Corporation, Mountain View, CA,
December 1978.
9-7. Offen, 6. R., et_ ^L, "Standards Support Plan for Stationary
Conventional Combustion Sources," Acurex Draft Report TR-78-107,
Acurex Corporation, Mountain View, CA, May 1978.
9-8. "Proceedings of the Second Stationary Source Combustion Symposium,
Vols. I through V, EPA-600/7-77-073a,b,c,d,e, NTIS PB-270 923/6BE,
271 756/9BE, 271 757/7BE, 274 029, 274 897, July 1977.
9-9. "Industrial Environmental Research Laboratory RTP, Annual Report
1976," Environmental Protection Agency, Research Triangle Park, NC.
186
-------
SECTION 10
FUTURE EFFORTS
The focus of the technical effort performed in the first two years
of the NOX EA program was described in the introductory discussion of
this report in Section 1. Future efforts in the third and final year of
the program will bring together and extend the methodology development,
data compilation, process evaluation, and systems analysis results of
these past activities to address the stated objectives of the program.
Specifically, third year emphasis will be placed on:
Updating the baseline source evaluation and emissions
inventories to 1977
Completing the process analysis and environmental assessment
studies of NOX controls applied to industrial boilers, gas
turbines, residential and commercial heating systems, and
internal combustion engines
Evaluating the effectiveness, cost, operational and
environmental impacts of advanced NOX control concepts
Identifying preferred NOX control strategies and future
control R&D priorities
In updating the baseline emissions characterization to 1977, more
recent equipment population, fuel consumption, and emission factor
information will be incorporated. In completing the process engineering
studies, the same methods used for the utility boiler effort, described in
Section 4.2, will be extended to the remaining source categories to be
treated. In addition, results for the gas turbine, 1C engine, and
residential/commercial heating systems studies will be reported in
IERL-EACD Environmental Assessment Report format. In both the above
efforts, newly available results from the field test programs discussed in
Section 5.2 will be factored in.
All of the above efforts will be integrated in the systems analysis
task to specifically give stated program outputs. Specific attention will
be afforded to the factors discussed in Section 3 in deriving
recommendations for regional NOX control strategies and scoping control
R&D needs.
187
-------
TECHNICAL REPORT DATA .
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-147
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE Environmental Assess me nt of
Stationary Source NOx Control Technologies:
Second Annual Report
5. REPORT DATE
June 1979
6. PERFORMING ORGANIZATION CODE
7.AUTHOR(S) L.R. Water land, K.J.Lim, K.G.Salvesen,
R. M. Evans, E. G. Higginbotham, and H. B. Mason
8. PERFORMING ORGANIZATION REPOI
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex/Aerotherm Division
485 Clyde Avenue
Mountain View, California 94042
10. PROGRAM ELEMENT NO.
E HE 62 4 A
11. CONTRACT/GRANT NO.
68-02-2160
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 PER
Annual; 6/77 - 6/78
ERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES JERL-RTP project officer is Joshua
919/541-2470. EPA-600/7-78-046 was the first annual
S. Bowen, Mail Drop 65,
report.
i6.
summarizes results of the 2nd year of an environmental assess-
ment of stationary source NOx control technologies. The 2nd year effort focused on:
(a) characterizing the baseline (uncontrolled) environmental impact of stationary
combustion sources; (b) developing fuel consumption and NOx emission inventories
and projecting these to the year 2000; (c) field testing selected stationary combustion
sources to determine multimedia pollutant emissions under both baseline and con-
trolled (for NOx) operation; (d) performing process engineering and environmental
assessment studies of NOx controls applied to utility and industrial boilers and to
gas turbines; (e) assembling and exercising reactive air quality models in systems
analysis applications; and (f) developing source analysis models for environmental
impact evaluation. The report summarizes program results in each of these areas.
Preliminary NOx control technology analysis for utility boilers indicates that off-
stoichiometric combustion and low NOx burners (LNB) are the preferred techniques
for both retrofit and new applications . For coal firing , overf ire air operation and
LNB are both cost effective; LNB is preferred for new wall-fired boilers. For oil
and gas firing, staged combustion with burners out of service is recommended.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Pollution
Nitrogen Oxides
Assessments
Fossil Fuels
Fuel Consumption
Combustion
Emission
Testing
Boilers
Gas Turbines
Mathematical
Models
Pollution Control
Stationary Sources
Environmental Assess-
ment
Emission Inventories
13B
07B
14B
21D
21K
21B
13A
13G
12A
8. Ol
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
21. NO. OF PAGES
199
22. PRICE
EPA Form 2220-1 (9-73)
188
------- |