&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

-------