xvEPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-78-215
November 1978
Assessment
of the Need for NOX
Flue Gas Treatment
Technology
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-600/7-78-215
November 1978
Assessment of the Need for NOX
Flue Gas Treatment Technology
by
W.E. Corbett, G.D. Jones, W.C. Micheletti,
R.M. Wells, and G.E. Wilkins
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
Contract No. 68-02-2608
Task No. 13
Program Element No. 1NE624
EPA Project Officer: J. David Mobfey
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
-------
ABSTRACT
The report gives results of a study to determine if and when the application
of NOX flue gas treatment (FGT) technology will be necessary in the U.S. It
addresses factors that will influence the levels of NOX emission control
needed to comply with both existing and future NOX standards. Topics trea-
ted include NOX emission sources, atmospheric transport and reactions, air
quality trends, regulations, and control strategies, and FGT methods. The
study concludes that the number of Air Quality Control Regions (AQCRs) with
NOX compliance problems can be expected to increase significantly in the
next decade. It further concludes that progressively larger reductions in
NOX emissions will be required in order to attain and maintain compliance
in "problem" AQCRs. The study does not establish conclusively whether or
not FGT will be required. However, current trends indicate that FGT may
be necessary in the future to achieve compliance with NOX standards in
certain AQCRs. This conclusion follows from the regionally specific nature
of U.S. NOX compliance problems, as well as uncertainties concerning both
future NOX emission reduction requirements and the ultimate effectiveness
of alternative NOX control methods, such as combustion modification.
ii
-------
CONTENTS
Abstract ...... ii
Figures iv
Tables iv
1.0 INTRODUCTION 1
2 . 0 SUMMARY AND CONCLUSIONS 3
3.0 THE NOX PROBLEM - SOURCES AND EFFECTS • 9
3.1 NOX Emission Sources 9
3.1.1 Nationwide Emission Trends 10
3.1.2 Regional Emission Profiles 12
3.1.3 Long-Tenn Trends in NOX Emission
Profiles 17
3.2 Atmospheric Reactions Involving NOX 19
3 .3 Atmospheric Transport of NOX 26
3.4 Air Quality . . 33
4.0 CURRENT NOX REGULATIONS AND TRENDS IN NOX
LEGISLATION 39
4.1 Current NOX Regulations 39
4.1.1 National Ambient Air Quality Stan-
dards 40
4.1.2 New Source Performance Standards 46
4.1.3 New Mobile Source Standards 47
4.2 Trends in NOX Legislation. 48
5 .0 CONTROL STRATEGIES 52
5 .1 Methods of Control 52
5.1.1 Stationary Sources 53
5.1.2 Mobile Sources.... 56
5.2 Attainment and Maintenance of Standards 57
5.2.1 .New Source Performance Standards 58
5.2.2 Annual Average Ambient Air Quality
Standard 59
5.2.3 Short-Term Standard 63
5.2.4 Prevention of Significant Deteriora-
tion 66
References 67
Appendix
111
-------
FIGURES
Page
Number
1 Observed Effects of N02 on Humans 42
2 Observed Effects of N02 on Animals 43
3 Variations in NO and N02 in Orange County, Cali-
fornia, October, 1974 64
TABLES
Number Page
1 Nationwide NOX Emission Estimates 1970-1976
(106 metric tons/yr) 11
2 Summary of NOX Emission Trends Projections 13
3 AQCR's Indicated as Potential N02 Problem Areas
by 1975 Monitoring Data 15
4 Contributions of the Various Sectors to 1975
NOX Emissions in Problem AQCR1 s 16
5 Mean Background Levels of Nitrogen O.xides 33
6 Number of AQCR's Reporting N02 and Oxidant Levels
in Excess of Standards 34
7 AQCR's Experiencing Violations of the One-Hour
NAAQS for Oxidants During 1973, 1974, and 1975... 36
8 State Ambient Air Quality Standards Which are
More Stringent than NAAQS for N02 41
9 Summary of NSPS for NOX Emissions from Ele'ctric
Utility Generating Stations 50
10 Comparative Costs of Stationary Source Controls.. 53
11 Comparative Costs of Mobile Source Controls 56
12 Typical Uncontrolled NOX Emissions from Large
Fossil Fuel-Fired Steam Generators 58
13 EPA R&D Program NOX Control Targets 60
IV
-------
1.0 INTRODUCTION
To date, government actions aimed at controlling the
impacts of ambient nitrogen oxides (NOX) emissions have focused
on combustion modifibation as the primary basis for the control
of both mobile and stationary sources. This approach has been
taken because combustion modification is the most cost effec-
tive approach to achieving initial reductions in uncontrolled
NOX emissions from all types of combustion sources. Now, how-
ever, a number of factors are providing incentives for a re-
examination of this basic control philosophy. In particular,
recent emission inventories and projections of air quality
trends indicate that NOx emissions from both stationary and mo-
bile sources are increasing. As a result, it appears doubtful
that NOX emission reductions currently attainable by use of
combustion modification alone can continue to provide the mar-
gin of control necessary to meet either existing or future,
more stringent, ambient air quality standards.
One approach which has considerable potential for re-
ducing NOX emissions beyond the levels currently attainable by
combustion modification is flue gas treatment. Whereas combus-
tion modification techniques appear to be capable of reducing
combustion source NOX emissions by approximately 50%, on the
order of a 9070 reduction in the NOX concentration of a typical
stationary source flue gas is possible with flue gas treatment.
The objective of this study which provided the basis for this
report was to determine if and when the need will exist for this
more stringent level of control.
This report, therefore, analyzes the factors which
will determine the level of NOX control necessary to comply x^ith
both existing and future standards. The conclusions of this
-------
analysis as they apply to the need for flue gas treatment tech-
nology are presented in Section 2.0. Background information on
the nature of the NOX emission problem including the sources of
atmospheric NOX emissions, the factors which control the trans-
port and conversion of NOX into a variety of other pollutants,
and the ambient levels of NOX related pollutants is presented in
Section 3.0. Section 4.0 addresses the current status of and
important trends in NOX regulations. The factors involved in de-
veloping cost effective compliance strategies for NOX are dis-
cussed in Section 5.0. Appendix A contains a description of
stationary source flue gas treatment methods.
-------
2.0 SUMMARY AND CONCLUSIONS
The inescapable conclusion of this study is that the
need for flue gas treatment as a control technique for nitrogen
oxides (NOX) from stationary combustion sources is not quantifi-
able at this time. There is no one factor or combination of fac-
tors, among those considered,, which can be said to demonstrate
that any given amount of flue gas treatment is or will be re-
quired for attainment or maintenance of present or future stand-
ards. This is due to the regionally specific nature of NOX com-
pliance problems, uncertainties that exist with respect to fu-
ture NOX emission reduction requirements for both stationary
and mobile sources, and uncertainties surrounding the develop-
ment and application of NOX control technologies to coal-fired
combustion sources. However, many if not all of the factors
analyzed in this study indicate that the number of Air Quality
Control Regions with NOX compliance problems will increase dra-
matically in the remainder of this century. They further indi-
cate that progressively larger and larger reduction in both sta-
tionary and mobile source emissions will be required to attain
and maintain compliance in these problem Air Quality Control Re-
gions. The following section summarizes the conclusions result-
ing from an analysis of the factors which affect whether or not
flue gas treatment will be required in the near term future.
NOX Emission Sources
On~a national basis, fossil fuel combustion accounts
for greater than 9070 of the NOX emitted from all man-made sources
Overall, these NOX emissions can be allocated as follows: mobile
sources O4070) electric utilities (^257,) , other stationary com-
bustion sources (^307,) , industrial processes and other miscella-
neous sources (^57.) . However, emission profiles for many Air
-------
Quality Control Regions which are currently experiencing prob-
lems with high ambient NOX levels differ widely from these norms.
Examples of both mobile source intensive and stationary source
intensive problem areas can be identified. This highlights the
regionally specific nature of the NOx control problem which, in
turn, supports the EPA policy of requiring each state to estab-
lish regulatory guidelines which are responsive to their localized
needs.
Atmospheric Reactions
Most of the NOX emitted from combustion sources is re-
leased as nitric oxide (NO), whereas, nitrogen dioxide (NOa) is
the criteria pollutant which is of concern from an ambient air
quality viewpoint. The oxidation of NO to produce N02 can occur
via a number of pathways. In polluted urban areas, this conver-
sion process appears to be controlled primarily by the reaction
of NO with oxidant free radicals. It follows that the maximum
impact of a given quantity of stationary source NOX emissions
upon ambient NOa levels would be felt in areas having simulta-
neously high oxidant levels.
Atmospheric Transport
Several recent studies have indicated that NOX emissions
from large stationary sources should not lead to problem ground
level concentrations of ambient N02 (relative to current stan-
dards) on an annual average basis. Achieving compliance with fu-
ture, more stringent standards (including short term standards)
may require the application of FGT, depending on the level of the
standard set.
-------
Ambient Concentrations
Although only a few air quality control regions in this
country are reporting violations of the current N02 National Am-
bient Air Quality Standard (100 yg/m3-annual average), on the or-
der of twenty AQCR's reported ambient N02 levels in excess of 80
ug/m3 in 1975. Because of this, the growth which is being ex-
perienced in all fossil energy consuming sectors may cause am-
bient NC>2 problems to be experienced in a significant number of
new AQCR's in the near-term future.
Standards
Standards which will determine the need for FGT are
Federal New Source Performance Standards, New Mobile Source Stan-
dards , State Implementation Plan Regulations, National Ambient
Air Quality Standards and Prevention of Significant Deterioration
requirements. Current Federal New Source Performance Standards
are based upon, and therefore require, only the application of
combustion modification. Recent health effects studies do not
point toward the need for a long-term ambient air quality stan-
dard for NO2 that is more stringent than that which currently
exists (100 yg/m3-annual average). However, the need for a short-
term N02 standard was recognized by Congress in the Clean Air Act
Amendments of 1977 which require that EPA establish a short-term
(1-3 hour average) N02 National Ambient Air Quality Standard un-
less the Administrator deems that this type of standard is not
necessary to protect public health. The level of control re-
quired to comply with this standard will not be known until the
standard is promulgated. However, recent studies by the Cali-
fornia Air Resources Board and EPA have indicated that flue gas
treatment may not be required in some Air Quality Control Regions
if the maximum allowable one hour average N02 concentration is
-------
less than 750 yg/m3. EPA is currently considering a level some-
where between 200 and 1000 ug/m3. California has already estab-
lished a one hour average standard with 500 yg/m3 as the maximum
allowable NOa concentration.
A number of states and/or air quality control districts
have adopted air quality standards which are more stringent than
current National Ambient Air Quality Standards for N02. Strict
enforcement of these standards may create a need for flue gas
treatment technology application, particularly in specific prob-
lem areas such as California's South Coast Air Basin.
Prevention of Significant Deterioration may require
flue gas treatment on stationary.sources located in Class I or
Class II areas even where New Source Performance Standards and
National Ambient Air Quality Standards would not indicate the
need.
Control Strategies
Unless a level of control requiring flue gas treatment
is adopted by some future New Source Performance Standard or re-
quired for Prevention of Significant Deterioration, the question
of "if and when" flue gas treatment will be necessary will ulti-
mately be decided in developing State Implementation Plans. Stra-
tegies for controlling N02 and other NOX related pollutants are
difficult to develop on a national level because each N02 problem
depends on such area specific factors as topography, meteorology,
and emission profiles.
All areas experiencing NOX compliance problems must at-
tain and maintain compliance by controlling mobile source emissions,
-------
stationary source emissions or both. Emissions from both sta-
tionary sources and mobile sources can be reduced either by ap-
plication of combustion modification or flue gas treatment tech-
niques .
Control Technology
Combustion modification techniques have been developed
and commercially applied that are capable of reducing stationary
source emissions to 86-260 g/GJ (0.2-0.6 lb/106 Btu), depending
on the type of fuel burned. Mobile source emissions can be con-
trolled to 0.62-3.7 g/km (1-6 g/mile), depending on the class of
vehicle. Flue gas treatment has been commercially applied in
Japan for gas- and oil-fired stationary combustion sources and
there are numerous programs underway to develop and improve flue
gas treatment and combustion modification techniques for coal-
fired stationary sources. In addition, post combustion cleanup
technology (catalytic converters) capable of reducing emissions
to 0.24 g/km (0.4 g/mile) is being developed for mobile sources.
The success or failure of these programs will greatly influence
the future course of NOX standards and compliance strategies .
Comparative costs indicate that when a control strategy
calls for gross reductions in NOX emissions from both mobile and
stationary sources, the most cost effective control approach in-
volves stationary source combustion modification. Mobile source
combustion modification is the next most cost effective NOX con-
trol technique on a cost per ton of NOX emissions reduced. Sta-
tionary source flue gas treatment and finally mobile source flue
gas treatment (catalytic converters) are the next most cost ef-
fective NOX control techniques.
-------
Secondary Pollutants
The extent to which ambient oxidant levels are affected
by changes in N02 concentrations remains to be established. Cer-
tainly, recent data indicate that oxidant levels are strongly af-
fected by changes in reactive hydrocarbon emission rates. This
finding has provided the basis for a number of oxidant control
strategies which are based upon hydrocarbon emission controls.
If similar links between ambient NO2 levels and problem concen-
trations of secondary pollutants such as peroxyacylnitrates,
ozone and atmospheric nitrates are established, this may justify
the use of controls which are more effective than conventional
combustion modification techniques. Research is currently under-
way to determine the role of NOX in the formation of oxidants.
Models which relate NOX emissions rates by source type to ambient
levels of N02 and other NOx related pollutants are currently
being developed.
Health Effects
Health effects of pollutants are major driving forces
behind the development of air quality standards. As such, this
area of research is critical to the issue of whether changes are
needed in the required levels of NOX control. Research is on-
going to determine the acute and chronic effects of both short-
and long-term exposures to N02, oxidants, and nitrates. In
addition, studies are underway to determine whether nitrosoamines
(known carcinogens) can be formed in the lung from potential
precursors such as NH3 and N02 in the ambient air. The results
of these studies may indicate the need for flue gas treatment if
tighter NOX control is judged to be necessary to protect human
health.
8
-------
3.0 THE NOX PROBLEM - SOURCES AND EFFECTS
This section summarizes four key aspects of the nitro-
gen oxides (NOX) emission problem. First in Section 3.1, the
sources of atmospheric NOX emissions are identified. The rela-
tive significance of the various emission sources and the re-
gional variations observed in NOX emission profiles are also dis-
cussed. The chemistry of atmospheric reactions involving NOX
is briefly reviewed in Section 3.2. This discussion is presented
to illustrate why it is necessary to consider more than just am-
bient N02 levels in assessing the possible need for flue gas
treatment (FGT) technology. Section 3.3 addresses the considera-
tions involved in assessing the relative impacts of point source
NOX emissions on ground level pollutant concentrations. In Sec-
tion 3.4 the region-specific nature of the NOX control problem
is highlighted. This is done by considering current air quality
I
trends for regions of the country which are either 1) already
out of compliance with respect to ambient N02 or oxidant levels
or 2) appear to have significant potential for the development
of future problems.
3.1 NOX Emission Sources
NOX is emitted by both natural and man-made sources.
Although natural sources contribute far greater amounts, man-
made sources of NOX are almost entirely responsible for the
high ambient NOX concentrations found in urban and industrialized
areas. Combustion of fossil fuels is by far the most significant
source of man-made NOX emissions.
On a worldwide basis, the major source of atmospheric
NOX is biologically produced NO. Globally, natural sources pro-
duce about 450 Tg (500 x 106 tons) of NO per year, while techno-
logical (man-made) sources emit on the order of 45 Tg (50 x 106
-------
Cons) per year of NOX (including both NO and N02) (Reference 1).
Estimates indicate that the total emission rate of NOX from tech-
nological sources in the United States is approximately 18-23 Tg
(20-25 x 106 tons) per year (Reference. 2) .
While natural source NOX emissions are much higher on
a global scale, they are also fairly evenly distributed. Natural
source NOX emissions contribute a low background concentration
as a result. Manmade NOX, while lower in global mass emissions,
can cause localized, high atmospheric concentrations.
3.1.1 Nationwide Emission Trends
U.S. emissions of NOX from man-made sources for 1970-
1976 are estimated in Table 1. As can be seen in the table,
the major contributor to technology-associated NOX in the U.S.
is fossil fuel combustion. The combined categories of transpor-
tation and stationary fuel combustion contribute about 95% of
the total U.S. emissions of man-made NOX. Stationary source
combustion is the single largest contributor, accounting for
51-54% of the total U.S. emissions. The electric utility con-
tribution ranged from an estimated 25% in 1970 to 29% in 1976.
By way of comparison, highway transportation emissions in-
creased from 31% in 1970 to 34% in 1976.
Industrial processes, solid wastes (incineration, pri-
marily) and miscellaneous sources are all shown in Table 1 to
be minor contributors to the total NOX emission picture. The po-
tential significance of these sources on a localized basis, how-
ever, should not be underestimated. In fact, some of the highest
ambient N02 levels ever recorded have been measured in eastern
Tennessee where high concentrations of nitric acid production
and utilization facilities exist.
10
-------
TABLE 1. NATIONWIDE NOx EMISSION ESTIMATES
1970-1976 (106 metric tons/yr)
Emission Source
Transportation
Highway
Non-highway
Stationary fuel combustion*
Electric Utilities
Other
Industrial processes
Chemicals
Pettoleua refining
Mineral products
Solid waate
Mlacellaneous
Forest wildfires and
managed burning
Coal refuse burning
Totalb
Year
1970
8.4
6.3
2.1
10.9
5.1
"5.8
0.6
0.2
0.3
0.1
0.3
0.2
0.1
0.1
20.4
(41Z)
(31Z)
(10Z)
(53Z)
(25Z)
(28Z)
( 3Z)
( ID
( ID
1971 1972
8.9
6.7
2.2
11.2
5.4
5.8
0.6
0.2
0.3
0.1
0.3
0.3
0.2
0.1
21.3
(42Z) 9.4
(31Z) 7.1
( 9Z) 2.3
(53Z)' 11.7
(25Z) S.9
(27Z) 5.8
( 3Z) 0.7
0.3
0.3
0.1
(••11) 0.2
( 1Z) 0.2
0.1
0.1
22.2
(42Z)
(32Z)
(10Z)
(53Z)
(27Z)
(26Z)
( 3Z)
( U)
( 1Z)
1973 1974 1975 1976
9.7
7.3
2.4
12.1
6.3
5.8
0.7
0.3
0.3
0.1
0.2
0.2
0.1
0.1
22.9
(42Z) 9.6
(32Z) 7.3
(10Z) 2.3
(53Z) 11.9
(28Z) 6.2
(25Z) 5.7
( 3Z) 0.7
0.3
0.3
0.1
( 1Z) 0.2
( 1Z) 0.2
0.1
0.1
22.6
(42Z) 9.9
(32Z) 7.6
(10Z) 2.3
(531) 11.2
(27Z) 6.1
( 5Z) 5;1
( 3Z) 0.7
0.3
0.3
0.1
( 1Z) 0.2
( 1Z) 0.2
0.1
O.'l
22.2
(45Z) 10.1
(34Z) 7.8
(10Z) 2.3
( OZ) 11.8
(27Z) 6.6
(23Z) 5.2,
( 3Z) 0.7
0.3
0.3
0.1
{ 1Z) 0.1
( 1Z) 0.3
0.2
0.1
23.0
(«*D
(34Z)
(10Z)
(51Z)
(29Z)
(23Z),
( 3Z)
( 1Z)
( 1Z)
Includes both area and point sources.
b
Totals may not add due to rounding.
Source: Reference 2.
-------
The increase in total NOX emissions which occurred
during the years 1970-1976 is attributable mainly to an increase
in fuel consumed by electric utilities and an increase in high-
way vehicle travel.
Emission projections are difficult to make because of
a variety of unknowns in both emission factors and growth trends.
A summary of several recent projections of NOX emission trends
is presented in Table 2. While the bases for these projections
vary widely, all of the results indicate very clearly that signi-
ficant levels of NOX control will have to be achieved in order
to maintain total NOX emissions at their current levels. A
more detailed discussion of how these controls might be imple-
mented is presented in Section 5.0.
3.1.2 , Regional Emission Profiles
While the preceding discussion does provide useful
insight into the relative significance of the various NOX emis-
sion sources on a nationwide basis, NOX emission control stra-
tegies should be developed on a regional basis. For this rea-
son, data on NOX emission profiles which are typical of regions
of the country which are currently having NO2 or oxidant com-
pliance problems are considered in this section.
Table 3 lists the ambient air monitoring stations
which recorded annual average N02 levels in excess of the 100
yg/m3 standard in 1975. Also listed are stations which recorded
ambient N02 levels greater than 80% of the standard.
The relative contributions of the various NOX emission
sources located within each of these problem Air Quality Control
Regions (AQCR's) are shown in Table 4. An obvious observation
12
-------
TABLE 2.
SUMMARY OF N0x EMISSION TRENDS PROJECTIONS
Source and Ytdi
Basis for I'rojectiu
S Ignlfleant Ru9ulLa
i' I us I tins/rnmmi-u t H
ytMiey Co nun I Lt L
(Kef. 3-3)
Greenfield, At taway,
and Tyler - 1977
(Ref. 3-4)
National increases modelled.
Base year 1970
Emission factor ratlos for
a tat lonary sources 0.9 for
i9BO. 0.7 for 1990, 0.5 for 1999.
sources, 2TL giowl ti fur mob I I L- bonn:i-->
Mobile Sourrt? st ;jnJai ds - :;i:v^r;il
different scenar I us I IK- ludliig one
for current standards.
Emission factor rut Ius for.mob Ilu
sector 0.73 for 1980. 0.5'i for 1990,
and 0.52 fur 1999. Vehicle turnover
rate 13 yt-ara for LDV.
Twenty problem AQCK's modelled.
Base y^itr 1975 for utilities and
transi>orLatlon; 1973 for other sectors.
5-61 growth in electrical generation.
Four fuel ut*e scenarios - Maximum
coal, high coal - low nuclear;
nominal growth; low coal - high
nuclear.
Three emiaalnnu restrict Ions scenarlos-
The must stringent of emission
restrictions (either stale or federal),
BACT applied to all sources, projected
NSFS.
^Mobile source emission factors for~LDV-
low altitude Ig/mlle In 1985 and 0.4g/
mile In 2000; high altitude 0.97 g/tulle
ID 1985 and 0.4 g/mlle in 2000;
California 0.85 g/mlle in 1985 and
0.4g/mile in 2000.
Mobile source-emlbslons baaed on
average per capita ownership projections.
Turn over rates for power plants taken
into account.
Relative to 1970 levela. diij with
current standards» NO? concent rat Ions
in the air will be 4Z higher In 1980.
31 higher in 1990. and 82 higher in
2000.
Arh I rv Ing l IK.' j.-.Miimrd til .it lou.'iry uttiirci;
.<-ml:tshm lutlur r;.t loi; would n^ulri.- «!.!<•-
aprr.id ;i|i|tl U-;it Ion ot controls (Including
F(.T). Tin- Krowlh r;iU-s .if-^um^d arc-
c r 11 (f.i I .UK! a n- Jndgt'd i (> be too Low,
did K'Jit I ng uvcn oiori* IIL'L.'I| for cont rot u .
In 19B5 6 AQCR's are predicted to bv
out of compliance under all emissions
restrictions. By 2000 10 AQCI^s are
projected exceeding the standard under
low coal, most stringent emissions
restriction scenario; 19 AQCR's arc
projected to be out of .compllance under
the maximum coal, least stringent
emission restrictions.
There are .1 t<:w AO/'K*s which will require
alI ;wa11 able cuntrol methodb Lo stay
In cont|>L lance (including KCT) .
-------
TABLE 2. SUMMARY OF NO EMISSION TRENDS PROJECTIONS (Continued)
Source and Yeur
fur Pro |eclI
r 1 ti.i I«HIS /f!«imill":
Ar.urux. Corpora t Ion/
Acrot herma I 1)1 vision
1977 (Ref. 3-5)
Argunnt* Labs, TKC -
1976 (Kef. 3-6)
Interagency Commit tee-
1976 (Ref. 3-7)
Chicago and Loa Ange 1 i-a mode 1 1 ed.
Base year concent cat Ions 132 ug/m'
160 pg/tn for Lou Angeles;
96 pg/m1 and 120 |ig/m3 far Chicago.
Three growth scenarios - nural n;i I
growth, low mobile, and high
atat iunury. NormlnaI growth ueburocH
moderate growth for stationary source,
1 g/mlle mobile standard beyond I960.
Low mobile scenarlo assumes 0.4 g/cnlle
moblle standard beyond 1981 .
Source weighting factnra used to show
sensitivity to stack height.
National Emissions modelled.
1975 base year.
General scenarios for applying
N5PS at different rates. Mobile
source emissions assumed to be 0.4
g/nlle beginning in 1978.
National emissions modelled.
1972 base year.
toblle source growth rate 3Z.
Emission factor ratios given for
several mobile standards. For
current standard, emission factor
ratios are 0.69 for 1980, 0.30
for 1990r 0.29 for 1999.
Stationary source growth rate 3.9X.
Emission factor ratios for
stationary sources 0.90 for 1980.
0.70 for 1990. 0.40 for 1999.
A 13 yr turnover rate assumed
for mobile sources.
Tin- rout roI I i-vcl i L-I(II I r<-il In tn.c li
clli.-s Is douiln.iUMl hy mol> I 1 •• source roiiiru
assumpL Inns . The low uuib 11 u scenario serins
unreal!si K-. K baued on the high b/ise year.
In Chicago, control of statlonary
sources Is required in all cases
except for 1985 if the low base year
Is used. For both cities In 2000,
alI combust Ion modification technology
will he required and ammonia Injection
will be requIred in some cases.
Under present control levels emlsslons The renults lud icate the need for further
will Increase 662 by 1985. If BACT Is development of control technologies such
applied to all new sources, the Increase as FCT, so that more stringent NSPS could
will be 24Z. Only a few stationary
sources are capable of significant
quantities of emission control.
Total emissions will Increase through
1990 In spite of the most rapid appli-
cation of NSPS. Stationary source con-
trol cannot compensate for mobile source
growth.
Predictions are for 12-17 of AQCR'a to
be out of compliance by 1980. Increase
in NOx emissions will average 12-241.
be applled.
Emission fac tor ratlos used Imply at rIngent
control methods for stationary sources,
Including FCT by 1999. The assumed growth
rates are thought too low.
Source: References 3, 4, 5, 6, 7 .
-------
TABLE 3. AQCR'S INDICATED AS POTENTIAL N02
PROBLEM AREAS BY 1975 MONITORING DATA
AQCR £
207
24
67
43
119
42
45
36
226
70
123
47
15
18
103
78
131
174
79
229
Monitoring site reporting
highest annual average
Kingsport, Tennessee
Pasadena, California
Chicago, Illinois
New York, New York
Boston, Massachusetts
Springfield, Massachusetts
Philadelphia, Pennsylvania
Denver, Colorado
Vinton, Virginia
St. Louis, Missouri
Southfield, Michigan
Seven Corners , Virginia
Phoenix, Arizona
Memphis, Tennessee
Ashland, Kentucky
Louisville, Kentucky
Minneapolis, Minnesota
Cleveland, Ohio
Cincinnati, Ohio
Seattle, Washington
Annual average
(Mg/m3)
186
153
111
102
102
100
98
96
96
94
94
88
85
85
85
84
84
84
82
82
Source: Reference 8.
15
-------
TABLE 4. CONTRIBUTIONS OF THE VARIOUS SECTORS TO 1975
NOX EMISSIONS IN PROBLEM AQCR'Sa
AQCK
Niunlii-r
207
24
67
43
1 19
42
45
36
226
70
123
47
15
18
IOJ .
78
1.11
1/4
79
229
Total U.S.
AIJCK
Local Ion
Kln^spurt, Tennessee
Pasadena, California
CliLr.tRo, I 1 1 Inols
Neu Yoik, Neu York
Boston, Massachusetts
Spr 1 iigf 1 e 1 d , Massachusetts
Ph 1 1 ade 1 ph i a , Pennsylvania
louver, Colorado
Vlnton, Virginia
St. Louis. Missouri
Soullifleld. Michigan
Seven Corners, Virginia
I'lioenl x , Ar Izona
Memphis, Tennessee
Ashland, Kentucky
Louisville, Kentucky
Ml nneapn 1 Is , Ml nneso ta
Cleveland, Ohio
Clnrlnn.lt 1 , Ohio
Seattle, Washington
2«,
Kle. Irli-.il
III 1 1 II li-s
(7. ol Tol.il )
45
1 3
24
II
10
12
12.6
30
8.5
41
21
23
8.5
47
75
4J
29
25
27
-
h
IndiiMi lal
(7, ol IV, lal )
1 7
1 1
29
6.4
7.0
13
20
8.1
42
16
17
2.5
4.9
8.4
6.5
16
9.1
25
18
46
-
Cminm
(7. ..I
1 .
1.
3.
9.
12
7.
4.
3.
3.
. 2.
4.
9.
1.
2.
0.
3.
3.
3.
4.
2.
c
•r. lal
Total )
H
3
•>
9
4
6
2
3
6
3
8
9
4
9
0
6
8
9
9
!<••:: 1
a. of
(1
2
)
4
2
4
3
2
I
2
3
3
1
I
0
1
2
2
3
1
e
deni lal
Tolal )
.9
.6
. 1
.0
.9
_ 2
.0
.6
.9
.1
.5
.2
.8
.8
.8
.9
.7
.1
.2
.7
Tfaiispoi --
l.ll I.HI
(Z ol Tolal)
32
60
14
63
64
59
54
47
39
35
41
56
67
36
15
)3
51
39
43
39
44Z
Ml-i. .
(/. ol T«l al )
'1.6
10
I.. I
'..H
1.5
4 . 1
5.7
9.1
6.0
3.3
13
5.7
16
4
1.6
2.2
4.7
4.2
4.2
9.9
e
Total
(Ions NO,)
2 1 7 ,OOO
730,000
685,000
1,330,000
280,000
143,000
489,000
1 35,000
75,000
263.000
403,000
186,000
108,000
94 ,OOO
190,000
114 ,OOO
165 ,1100
331 ,000
146,000
200,000
24,200,000
M.ix In
annual .ivr
NO. con,-.'
(nr./m1)
IHl,
r. i
1 1 1
10'
102
100
98
96
96
94
94
88
85
8'j
H5
K4
84
84
82
H2
J'sonrce: Ki> Terence 4 .
^Includes combust Ion and process emissions.
IjKi-oiii Table 3-3.
_From Tiili I e 3-1 .
''Commercial iind KeNldunU.il and Mlar.. = M*'..
-------
which can be made from these data is the fact that there is no
"typical" problem AQCR emission profile. About half of the AQCR's
listed have higher NOX emissions from the transportation sector
than the U.S. average contribution of 44%. There is also consid-
erable variation in the electric utility contributions listed.
In two of the AQCR's,. electric utility sources contributed only
8.570 of the total emissions of NOX. The maximum utility sector
contribution seen in Table 4 was 75% (AQCR 103 - Ashland, KY).
Seven of the 20 AQCR's listed had higher NOX emissions from the
utility sector than the' U.S. average of 28%.
These data .clearly indicate the site specificity of
the NOX emission problem. Unfavorable meteorological conditions
coupled with high NOX emissions from any one of a number of
potential sources can create N02 compliance problems. These
trends illustrate why compliance strategies for the attainment
of ambient air quality standards should be developed on a re-
gional basis.
Another factor which is not reflected in the data shown
in Table 4 but which should be considered in this assessment is
the impact of seasonal variations in fuel consumption and there-
fore, NOX emission profiles.
3.1.3 Long-Term Trends in NOX Emission Profiles
The hazards of forecasting the effects of technologi-
cal and economic changes are well known. If future air quality
standards ar.e to be attained, however, general growth trends must
be recognized and considered in planning recommendations.
There are several current trends which will undoubted-
ly impact the relative significance of large point sources as
NOX emitters. Some of these trends include:
17
-------
in the short-term, an increase in the use of coal,
primarily in the electric utility and industrial
sectors,
an increase in the degree of electrification,
primarily in the residential and commercial
sectors,
a decrease in natural gas consumption in all
sectors, and
a decrease in the NOX emissions from mobile
sources as older cars are replaced with newer
models having lower emissions.
Some of these effects will be offset by trends such as
an increase in the use of nuclear and solar
energy, and
an increase in the application of NOX control
techniques to existing sources and the replace-
ment of old "dirty" units with new ones.
In addition, the overall energy use picture is clouded
by the uncertain effects of future conservation measures such as
decreased residential/commercial energy usage
for heating, cooling, lighting, etc., and
energy recovery schemes (e.g., combustion air
preheating) which will continue to be widely
applied in the industrial sector.
18
-------
The net effects of these trends are difficult to pre-
dict, primarily because of the uncertainties which exist in the
projected growth rates of the various energy consuming sectors.
It is expected, however, that the significance of large point
sources as N(X emitters will increase.
X
3.2 Atmospheric Reactions Involving NOX
N0x emitted from both stationary and mobile sources
(mainly in the form of NO) may undergo considerable transforma-
tion while being transported from an emission source to a sink
and/or receptor. .In particular, the chemical reactions which
result in the formation of urban smog are very relevant to the
NO emission problem.
X
The role of NOX in urban smog formation is very complex,
Although much work has been done recently to develop a better
understanding of the mechanisms involved in the formation of
various atmospheric pollutants, considerable work remains to be
done, particularly in the areas listed below.
quantification of the various reactive chemical
species present in urban air masses, and particu-
larly the reactive intermediates which participate
in rate controlling steps for key component forma-
tion reactions
identification of the specific chemical mechanism(s)
whereby important chemical species are formed
quantification of important variables which affect
the rates of formation and disappearance of
important chemical species
19
-------
Much information of this nature has been developed,
primarily in "smog chamber" studies using synthetic mixtures of
polluted air. Because of several obvious problems, all of this
work may not be directly applicable to "real world" situations.
It is true, however, that this work has led to a good understand-
ing of many of the important reactions involved in the formation
of urban smog.
Researchers are currently applying this information
in the development of a variety of air quality simulation models.
Hopefully, these models will be capable of relating emissions to
air quality, taking into account meteorological variables, photo-
chemical reactions, and pollutant transport. A model which ac-
curately predicts the levels of all significant pollutants at
their points of maximum impact is the ultimate goal. It should
be able to predict air quality in the vicinity of the emissions,
as well as in adjacent areas and should be applicable to any
area. Such a model would serve as a vital tool for developing
control strategies. No such model is available currently. A
good description of the various types of models and their status
of development may be found in Reference 9.
The role of NOX in the formation of photochemical smog
is briefly summarized below. This discussion is presented to il-
lustrate the fact that the potential impacts of atmospheric NOX
emissions cannot be assessed only in terms of their effects upon
ambient N02 levels. For a more comprehensive discussion of this
subject, the reader should refer to the recent reviews found in
References 10 through 14.
As discussed previously, the bulk of the NOX emitted
from combustion sources is released in the form of nitric oxide,
NO. Since the conversion of NO to N02 is a critical first step
20
-------
in the sequence of reactions leading to the buildup of high con-
centrations of N02 as well as other pollutants, this reaction will
be discussed first.
NO may react with molecular oxygen according to Re-
action 1 to form N02.
2ND + 02 -> 2N02 (1)
However, the rate of this reaction is not sufficient to account
for che rapid conversion of NO to N02 which is observed in polluted
urban atmospheres. The key to this overall conversion of NO to
N02 is now considered .to be the reaction of NO with peroxy free
radicals. In particular, the hydroperoxy radical is felt to be
an important species
H02 + NO - N02 + OH ; (2)
although, alkylperoxy
R02 + NO + RO + OH (3)
~.nd acylperoxy
0 0
ii n
RC-02 + NO * RCO + N02 (4)
radicals can also participate in NO oxidation reactions. Acyl-
peroxy radicals can also react with N02 to form peroxyacyl ni-
trates (PAN), an important class of eye irritants.
A variety of paths exist for the formation of these
highly reactive intermediates. Hydroxyl radicals and CO for
example can react to form hydroperoxy radicals according to the
following sequence:
21
-------
OH + CO ->• H + C02 (5)
H + 02 + M + H02 + M, (6)
where M is a third body capable of absorbing excess vibrational
energy. Hydroxyl radicals also can react with a variety of
hydrocarbons (such as aldehydes) to produce hydroperoxy radicals.
Ozone formation via N02 photolysis provides another
important mechanism whereby a number of important species are
formed.
N02 + sunlight (2900-4300A) -> 0* + NO (7)
M + 0* + 02 + 03 + M (8)
In these equations, 0* is an activated oxygen atom and M is any
third body which is capable of absorbing the excess energy re-
leased in reaction 8.
The ozone thus formed may react with a variety of at-
mospheric pollutants. Among them are NO and N02 .
NO + 03 -" N02 + 02 (9)
N02 + 03 -»• N03 + 02 (10)
This last reaction is important because of the subsequent possi-
bility of NOa reaction with other species such as NOa and H20 to
form nitric acid thus providing a mechanism for N02 removal from
the atmosphere. Ozone can also react with the hydrocarbons pres-
ent in the air to form a wide range of reactive species or under-
go photolysis,
22
-------
03 + sunlight -»• 0* + 02, (11)
to form an activated oxygen atom which can subsequently partici-
pate in important photochemical reactions such as hydroxyl radi-
cal formation,
0* + H20 -»• 20H, (12)
or hydrocarbon oxidation.
In addition to the gas phase pollutants described above,
smoggy urban air contains aerosols and particulates which cause
light dispersion and haze. Much work has been done in analyzing
data from the Los Angeles Basin area concerning secondary aero- •
sol and particulate formation and characterization. It has been
shown that secondary aerosols (sulfates, nitrates, and organics)
can be formed by reactions involving primary gaseous pollutants
(SC>2 , NOX, and hydrocarbons) (Reference 15). It has been fur-
ther shown that under some circumstances aerosol formation from
gas-to-particulate conversions may equal or exceed that due to
primary emissions (Reference 16).
'when organic compounds are broken up into smaller more
reactive fragments as a result of atmospheric reactions, these
fragments may polymerize into higher molecular weight compounds
which can condense to form aerosol droplets. Sulfate aerosols
and particulates, can result from the nucleation and hydration
of sulfuric acid in the atmosphere (Reference 17). The formation
of nitrate particulates and aerosols is more complex, however.
The photochemically produced precursor of atmospheric nitrates
is nitric acid, HN03, which has a relatively high volatility at
trace concentrations. As a result, HN03 does not tend to con-
dense as sulfuric acid does. Rather, HN03 must react with other
23
-------
species or dissolve in a condensed phase already present. One
obvious possibility is the reaction of HN03 with NH3 to produce
ammonium nitrate.
NH3 + HN03 -" NH4N03(s) (13)
Nitrate particulates do, in fact, appear to exist primarily as
ammonium salts (Reference 15); although, the existence of this
reaction as a significant mechanism in the atmosphere remains
to be confirmed.
Many gaseous nitrogen species such as N02 , N205 and
HNOs can dissolve and/or react in an aqueous phase, but the role
of these reactions in atmospheric aerosol formation remains to
be established (Reference 18) .
The reactions discussed to this point are only elemen-
tary examples of the incredibly complex series of reactions which
occur in sunlight irradiated atmospheres. Ultimately, these re-
actions result in the conversion of NO into other substances which
must subsequently be removed from the air by adsorption or absorp-
tion.
The existence of the chemical reactions which were just
described means that the control of NOX may have to be considered
in developing effective control strategies for both ambient oxi-
dants and particulates. Whether these considerations will require
more stringent levels of NOX control that those currently re-
quired to satisfy ambient N02 standards remains to be established.
To date, no oxidant- or particulate-related NOX emission controls
have been promulgated (Reference 19) .
24
-------
Recent studies of this subject have shown that the sig-
nificance of NOX as a precursor for the formation of other pollu-
tants appears to be dependent not only on the HC:NOX ratio, but
also on the transport mechanisms which determine how long a given
pollutant mix remains in the air. A study by Trijonis (Reference
20), comparing changes in air quality to changes in emissions >in
Southern California, indicates that this is so, at least for Sou-
thern California. A decrease in the HC:NOX emission ratio in the
western portion of the South Coast Air Basin caused by reducing
HC emissions while allowing NOX emissions to rise has resulted in
decreased oxidant levels in western and central Los Angeles County
but increased oxidant levels .in the eastern South Coast Air Basin.
This finding along with indications that large stationary sources
located in the western part of the basin may be contributing to
high NOz levels in the eastern portions supports the conclusion
that NOX emission controls should be incorporated into the oxi-
dant control strategy for the South Coast Air Basin.
The problems associated with devising oxidant control
strategies are becoming more and more complex. It has been recog-
nized that oxidant problems in outlying areas may be caused by
the transport of air parcels from urban or industrialized areas
(References 19 and 21). It is also thought that the age of the
pollutant mix (length of time in the air) affects the atmospheric
chemistry and, therefore, the severity of the oxidant problem.
For these reasons, control strategies for urban, suburban, in-
dustrial, and rural areas cannot be dissociated. Oxidant control
strategies developed for a specific area should, therefore, ad-
dress not only local emissions, but also existing background
levels of key pollutants (Reference 19). All of the factors
which influence the levels of ambient oxidants (length of time
pollutants remain in the air, distance of pollutant transport,
sources of natural emissions, etc.) are very area-specific and
25
-------
will demand the development of a control strategy which is tail-
ored to the specific requirements of the area.
In addition to the documented role of NOX in oxidant
formation, it has been suggested that nitrosoamines could be
formed in the atmosphere through the reaction of NOa and amines.
This hypothesis generated much concern because of animal tests
which show nitrosoamines to be powerful carcinogens.
The reaction of NOz and amines has been observed in
the lab under acid conditions. The presence of ozorte has fur-
ther been shown to accelerate nitrosoamine formation. In theory,
the reaction should proceed in the atmosphere. To date, however,
there have been no monitoring data collected to support the hy-
pothesis of nitrosoamine formation in the atmosphere. All nitro-
soamines detected in air have been traced to specific point source
emissions (References 22 and 23). Inhalation of precursors has
also been suggested as a possible mechanism leading to nitroso-
amine synthesis in the lung. This mechanism is not thought to
be significant, however (Reference 22).
Confirmation of the atmospheric formation of nitroso-
amines or of nitrosoamine formation in the lung could eventually
provide a basis for stringent NOX controls. Monitoring work in.
this area is continuing at EPA, the National Institute for Occu-
pational Safety and Health, and at the University of California.
3.3 Atmospheric Transport of NOX
Stationary point sources are emission sources which can
be approximated by single points as opposed to area sources which
are characterized by a fairly uniform distribution of emissions
over a broad area. In order to assess the impacts of NOX emis-
sions from large stationary point sources upon the concentrations
26
-------
of N0x-related pollutants at ground level, the factors which re-
late to the physical transport of large point source plumes need
to be considered. In addressing this problem, two limiting cases
can be proposed.
The first involves a situation such as that which fre-
quently occurs during periods of air stagnation in the South Coast
Air Basin (Los Angeles area) in Southern California. In this
situation, it is probably reasonable to allocate ambient NOX levels
to the various source categories in proportion to the emission
rates of NOX from those sources. However, the need to handle mo-
bile source NOX emission spikes (caused by peak traffic periods)
must be recognized.
The second case, which is more commonly encountered,
involves the impact of stationary source plumes at points which
are downwind from the emission source. The considerations in
this case are the factors which affect the ambient levels of NOX
related pollutants in the vicinity of a ground level receptor.
The effects of emissions from large stationary sources
on air quality depend not only on the chemical phenomena which
were discussed in the previous section, but also on physical
transport of the plume. The behavior of a plume from a large
stationary source can be considerably modified by meteorological
parameters such as cloud cover, wind patterns, temperature and
relative humidity which result in part from spatial differences
in the earth's ability to absorb and reradiate the energy re-
ceived from the sun. In addition, the ability of the atmosphere
to mix and dilute pollutants is highly dependent not only on tem-
perature, but also on the spatial variability of temperature,
particularly in the vertical direction. Local topographical fea-
tures have the effect of modifying large-scale weather patterns
27
-------
which control pollutant transport over large (regional) areas.
In particular, topography can also influence the wind patterns
which govern the dispersion of pollutants in the atmospheric
boundary layer.
The important meteorological variables which affect the
transport and dispersion of pollutants from large elevated point
sources include wind speed, wind direction and wind direction
variability, and vertical temperature structure (stability and
mixing depth).
The effect of an increase in wind speed is to increase
the degree of mixing of the dispersing material in the horizontal
downwind direction. In addition, the rise of the plume from the
source is diminished with increasing wind speed.
Spatial and temporal variations in the wind direction
affect not only the general direction of pollutant transport, but
also the horizontal dilution of the plume. During periods of
fairly persistent spatial and temporal wind directions, plumes
from large elevated point sources may be identified up to 500-
1000 kg (310-620 miles) downwind (Reference 24). However, during
periods of fluctuation in the wind direction the horizontal
crosswind dispersion of the plume is enhanced and maximum down-
wind concentrations may drop off rapidly with distance.
The temperature structure of stability of the atmosphere
also defines the dispersive capability in the vertical and hori-
zontal crosswind directions. As the atmosphere becomes more
stable, plumes emitted from large elevated point sources assume
a relatively compact shape. However, as the air becomes less
stable, plume spread in the vertical and horizontal crosswind
direction increases.
28
-------
Highest ground level concentrations in .the vicinity of
large stationary point sources occur during conditions which tend
to minimize lateral, vertical, and downwind dispersion while at
~he same time allowing the plume to reach ground level. For low-
level area sources or "urban"-type sources-, maximum ambient NOX
concentrations very near the source occur during stable, nighttime,
light wind conditions. However, because these same conditions
cause emissions from elevated point sources to be confined to
elevated stable layers, they do not give rise to high ground
level concentrations for these types of emissions.
Two conditions recognized for producing adverse dis-
persion conditions for large elevated point sources are limited
mixing and coning (Reference 25). The limited mixing or trapping
condition is most often associated with the presence of a large
slowly moving high pressure system approximately centered over
the area. At night, a strong ground based inversion develops
beneath an overlying stable layer, and emissions from a large
elevated point source become embedded in this stable layer and
do not reach the ground. During the midmorning, ground based
turbulent mixing reaches the base of the plume, dispersing it
rapidly to the ground with resultant high peak concentrations
(up to about 30 minutes duration). If the mixing depth continues
to develop slowly beneath the overlying stable layer, relatively
high ground level concentrations may persist for several hours.
During periods of limited mixing, high ground level
concentrations are also associated with low wind speeds and the
resultant decrease in horizontal downwind dispersion. However.
if the wind speeds are too low, considerable meander in the plume
may occur due to wind direction variability, and ground level
concentrations will drop.
29
-------
The phenomenon of coning, i.e., plume dispersion during
periods of near neutral stabilities and moderate-to-strong winds,
gives rise to relatively high concentrations although short-term
coning concentrations are typically not as high as short-term
limited mixing concentrations for large, elevated point sources.
The magnitude of coning concentrations is very sensitive to the
wind speed. If the speed is low the plume rise will be large and
ground level concentrations will be diminished. In addition,
for low wind speeds, plume meander may increase and resultant
pollutant concentrations, especially for longer averaging times
(5-24 hours), may decrease. If the wind speed is high, the
horizontal dilution of the plume will increase and ground level
concentrations will drop.
Often the highest ground level concentrations occur at
great distances from large, elevated point sources during coning
periods, because during such conditions, wind direction variabil-
ity (temporal and spatial) is often minimized. As a result the
point of maximum plume impact may remain near the same location
for long periods of time (up to 24 hours). This type of plume
behavior may impact the required controls for meeting a short
~erm standard.
The interaction between complex terrain and emissions
from large point sources can influence ground level concentration
patterns. The increased roughness of the ground surface in areas
of complex terrain may generate additional turbulence and enhance
plume dilution. Topographical features such as narrow valleys
or elongated bluffs can induce channeling of the plume, limiting
v.ts spread in the horizontal crosswind direction. In addition,
•_ne vertical temperature structure within narrow valleys may be
such that the mixing layer may not develop as rapidly as it does
over flat terrain. As a result, periods of limited mixing may
be more severe, especially for low-level point source releases.
30
-------
The impingement of plumes from elevated point sources
against elevated terrain can often result in very high ground
level concentrations (Reference 26). For periods during which
the plume is embedded in a stable layer and the wind direction
is fairly persistent, the plume may travel for great distances
(>25 km) before impinging against vertical terrain features.
After traveling these distances, ground level concentrations may
be relatively high because of the•limited vertical and horizontal
crosswind dispersion of the plume.
The transport phenomena and dispersion associated with
Large point sources differ from those for low-level area sources
in several ways which are summarized below.
Peak ground level concentrations occur closer to
the source for area sources and farther downwind
for large point .sources. This observation is a
reflection of the difference in the heights of
the plumes.
Because of the larger horizontal crosswind extent
of area sources and because the source density is
often fairly homogeneous through the area, peak
ground level concentrations resulting from area
sources often cover a large area. On the other
hand, the point of maximum impact for point
sources usually covers a small area.
For large area sources, peak concentrations are
mainly a function of the wind speed and the mag-
nitude and extent of the vertical dispersion.
Because of their large horizontal dimensions,
the effect of horizontal crosswind dispersion
31
-------
is minimized in comparison to the effect of
horizontal crosswind dispersion on point source
emissions.
Highest ground level concentrations associated with
emissions from large point sources typically occur
during periods of limited layer mixing and coning.
Highest levels resulting from low-level area source
emissions typically occur during periods of light
winds, stable conditions and low-mixing depths.
Consideration of transport processes and plume behavior
not only indicates another difficult complexity affecting air
quality, but also shows some of the reasons that pollution prob-
lems may exist in certain regions and not in others. The dis-
cussion above emphasizes the fact that meteorological and
geographical conditions determine the location of impact of a
plume and its degree of dispersion. If conditions exist which
are conducive to worst-case behavior described in this section,
stringent controls on large sources of NOX may be indicated,
even as stringent as FGT, depending on the severity of the
problem.
Another facet to the problem of plume transport is the
fact that a plume may travel for long distances, creating a
longer reaction time for gases, aerosols, and particulates in
the mixture. Because the age of a pollutant mixture is an impor-
tant factor in both NO ->• NOz conversion and oxidant formation,
plumes from large stationary sources can contribute to oxidant
problems downwind of the source. Confirmation of this phenomenon
could give greater impetus to stringent controls for NOX which
could potentially include FGT if site specific oxidant problems
were severe enough.
32
-------
3.4 Air Quality
Because of transport phenomena and physical and chemi-
cal transformation processes occurring in the atmosphere, NOX
emissions are not directly relatable to the levels of NOz measured
in the atmosphere. Most nitrogen oxides are emitted as NO, chang-
ing rapidly in the atmosphere until an equilibrium between NO and
NO? is established. Most monitoring data are collected for NOz,
che criteria pollutant. Thus, while emissions are measured in
•terms of nitrogen oxides (NOX) , air quality is measured in terms
of NOz concentration. In this section, the levels of NOz and
oxidants found in the atmosphere in air quality problem areas in
the U.S. are discussed.
Nitrogen oxides are not evenly distributed globally.
Urban and industrial areas have much higher atmospheric concen-
trations than nonurban areas. Background levels are very diffi-
i
cult to determine; however, concentrations estimated from several
sources are presented in Table 5.
TABLE 5. MEAN BACKGROUND LEVELS OF NITROGEN OXIDES
NO 2.5 yg/m3 (2 ppb) land areas between 65°N and 65°S
0.25 yg/m3 (0.2 ppb) all other areas
N02 7.5 yg/m3 (4 ppb) land areas between 65°N and 65°S
0.94 yg/m3 (0.5 ppb) all other areas
Source: Reference 1.
Concentrations of N02 measured in urban and industrial
areas may reach levels; several hundred times higher than the
background levels presented here.
33
-------
Levels of N02 in the atmosphere vary with time as well
as with geographical location. Diurnal variations are typical,
due in large part to cycles in human activity. Vehicular traffic
is a notable cause of increased N02 levels. During low traffic
periods, the NC>2 may be dispersed or converted, thereby reducing
the concentrations. Seasonal patterns which must also be con-
sidered are caused by variations in the temperature, prevailing
winds, and solar radiation intensity, as well as by variations
in the amount of heating fuel combusted.
A number of areas in the United States are experiencing
high atmospheric levels of N02 and photochemical oxidants. Table
6 summarizes the number of AQCR's reporting N02 and oxidant stan-
dard violations for the years 1970-1974. Of the 247 total AQCR's
only 4 were exceeding the N02 annual average standard at that
time, but the oxidant standard was being exceeded in 76 AQCR's.
Even though only a few AQCR's were actually exceeding thep N02
standard, a study of 1975 monitoring data resulted in a list of
20 AQCR's containing 30 percent of the U.S. population which have
potential N02 problems. This list was shown in Table 3.
TABLE 6. NUMBER OF AQCR'S REPORTING N02 AND OXIDANT LEVELS
IN EXCESS OF STANDARDS
Oxidant
1970 1971 1972 1973 1974
AQCR's reporting at
Least minimal data
1-hour standard exceeded
Nitrogen dioxide AQCR's
reporting at least one
station-yr.
Annual standard exceeded
17 30 51 77 86
14 24 31 65 76
11
2
12
3
15
2
28 101
4 4
Minimal data consists of at least 3-24 hour samples or 400
hourly values.
Source: Reference 2.
34
-------
A list of. potential N02 problem areas compiled from
1974 monitoring data contained a few AQCR's which did not appear
as problem areas in the list contained in Table 4. They are
San Diego (#29), San Francisco (#30), Atlanta (#56), Baltimore
(#115), McLean-Mercer-Oliver Co. (#172), Canton-Cleveland (#174),
Salt Lake City (#220), and Richmond (#225) (Reference 2). If
these AQCR's are considered to remain potential problem areas
and added to the list of 20 shown in Table 4, a total of 28 of
the 247 areas might have an NOa problem now or in the near future.
A list of AQCR's in which oxidant levels were exceeding
the one-hour standard in 1973 and 1974 is included in Table 7.
Although control strategies for photochemical oxidants do not
include NOX emission limitations at this time, this is a possi-
bility for future control strategies.
These air monitoring data clearly show that the control
of NOi and oxidants is an urban problem. High levels of these
pollutants...are not being measured uniformly throughout the United
States. Comparatively, the oxidant problem appears much worse
than the N02 problem; 1974 data show that oxidant levels were
exceeding the air quality standard in 76 AQCR's, but N02 levels
were exceeding the standard .in only 4 AQCR's.
Left unchecked, the numbers of AQCR's out of compliance
with respect to N02 would be expected to increase with growth,
however. An EPA study predicts that 12-17 AQCR's will be exceed-
ing the current N02 standard by 1980 (Reference 7). Such esti-
mates emphasize the importance of including growth projections
in control strategies and also point to the fact that the prob-
lem of high levels of pollutants may become more prevalent in
the future.
35
-------
TABLE 7. AQCR'S EXPERIENCING VIOLATIONS OF THE ONE-HOUR NAAQS
FOR OXIDANTS DURING 1973, 1974, and 1975
004
005
007
013
015
018
024
025
028
029
030
031
032
033
036
043
045
047
049
050
051
052
055
056
060
067
069
070
072
075
077
078
079
080
085
086
088
092
094
095
099
AQCR
Birmingham
Mobile
Huntsville
Las Vegas
Phoenix - Tuscon
Memphis
Los Angeles
N. Central Coast
Sacramento Valley
San Diego
San Francisco
San Joaquin Valley
S. Cent. Coast. Cal.
S.E. Desert, Cal.
Denver
Metro New York
Philadelphia
Washington, D.C.
Jacksonville
S.E. Florida
S.W. Florida
West Central Florida
Chattanooga
Atlanta
Hawaii
Chicago
Metropolitan Quad Cities
St. Louis
Paudcah, Ky.
Springfield, 111.
Owensboro, Ky.
Louisville
Cincinnati
Indianapolis
Omaha
Sioux City
Northeast Iowa
S. Cent. Iowa
Kansas City
Topeka
S. Cent. Kansas
1973
As of 4/7/75
2nd high
Vig/m3
435
206
-
438
372
196
1156
293
431
587
509
509
254
548
548
456
744
744
-
270
215
_
_
195
166
568
_
764
101
-
_
362
333
225
205
_
_
225
-
190
170
1974
Partial year
2nd high
yg/m3
280
280
-
300
-
240
660
210
170
340
170
230
170
600
500
350
720
300
-
-
-
_
_
10
_
810
- -
500
. 200
225
200
200
310
330
180
_
_
500
-
-
460
1975
2nd high
yg/m3
269
245
196
200
255
255
784
216
412
372
392
372
216
372
349
510
625
451
451
196
_
274
427
324
_
427
210
862
204
—
214
461
412
245
225
173
176
196
160
200
560
(Continued)
36
-------
TABLE 7. AQCR'S EXPERIENCING VIOLATIONS OF THE ONE-HOUR NAAQS
FOR OXIDANTS DURING 1973, 1974, and 1975 (Continued)
102
103
106
107
115
117
118
119
120
121
122
123
124
127
128
129
131
151
152
153
158
159
160
161
162
164
167
171
173
174
176
178
181
184
186
193
195
196
197
200
AQCR
Lexington
Ashland, Ky.
S. La. - S.E. Texas
Berlin, N.H.
Baltimore
Pittsfield, Mass.
Wo r Chester
Boston
Providence
Nashua, N.H.
Central Michigan
Detroit
Toledo
Central Minnesota
S.E. Minnesota
Duluth
Minn. - St. Paul
Scranton
Albuquerque
El Paso
Syracuse
Glen falls, N.Y.
Rochester
Schenectady
Buffalo
Elmira
Charlottee, N.C.
Asheville, N.C.
Dayton
Cleveland
Columbus
N.W. Penn. - Youngstown
Steubenville
Okla. City
Tulsa
Portland (Ore. -Wash.)
Cent. Pennsylvania
S. Cent. Pennsylvania
S.W. Pennsylvania
Columbia, S.C.
1973
As of 4/7/75
2nd high
Ug/m3
146
636
195
450
_
_
409
274
194
_
208
235
_
—
_
186
_
244
_
344
233
231
356
280
_
288
210
245
352
274
313
_
400
176
246
_
_
—
155
1974
Partial year
2nd high
yg/m3
370
390
110
_
310
350
400
410
230
_
180
_
_
_
_
130
500
180
160
250
220
170
270
270
_
220
210
160
130
90
130
_
310
160
260
250
380
180
280
1975
2nd high
Ug/m3
231
_
369
127
372
323
308
376
329
247
372
514
265
343
333
274
231
482
245
321
225
194
_
284
404
263
320
120
250
451
306
496
343
239
190
294
325
370
416
245
(Continued)
37
-------
TABLE 7. AQCR'S EXPERIENCING VIOLATIONS OF THE ONE-HOUR NAAQS
FOR OXIDANTS DURING 1973, 1974, and 1975 (Continued)
207
208
212
214
215
216
217
220
223
225
229
239
240
AQCR
E. Term. - S.W. Virginia
Mid-Tennessee
Austin-Waco
Corpus Christi
Dallas
Houston
San Antonio
Salt Lake City
Hampton Roads
Richmond
Puget Sound
S.E. Wisconsin
S. Wisconsin
2nd high
yg/m3
568
205
313
219
248
484
-
217
215
245
78
450
211
2nd high
yg/m3
230
340
230
250
290
340
170
240
350
270
260
230
130
2nd high
yg/m3
323
392
206
241
323
588
296
296
251
353
235
425
™
Source: References 7, 27.
38
-------
4.0 . CURRENT NOX REGULATIONS AND TRENDS IN NOX LEGISLATION
The heed to utilize FGT technology to control NOX
emissions from large stationary point sources will depend
heavily on current and future air pollution legislation. Air
quality standards and the federal and state regulations enacted
to achieve those standards may directly or indirectly dictate
the use of specific NOx control technologies. Therefore, the
purpose of this chapter is to review those NOx regulations
currently in effect and to examine the prevalent trends in NOX
legislation. Since NOX is a participating species in reactions
leading to the formation of photochemical oxidants, references
will also be made to oxidant legislation.
4.1 Current NOX Regulations
The enactment of the Clean Air Act in 1963 provided
the foundation for a series of other federal and state govern-
ment actions designed to limit atmospheric emissions of specific
compounds designated as pollutants. Since that time a number of
both national and regional air quality and pollutant emission
standards have been promulgated. The most notable regulations
currently in effect for NOX are the National Ambient Air Quality
Standards (NAAQS), the State Implementation Plans (SIP's) regula-
tions, the New Source Performance Standards (NSPS), and the New
Mobile Source Standards. In this section, the purpose, general
level of regulation, and any regional variations seen in these
standards are discussed. The implications of more recent require-
ments (e.g., Prevention of Significant Deterioration) will be dis-
cussed in the section dealing with trends in NOX legislation.
39
-------
4.1.1 National Ambient Air Quality Standards
Although NAAQS do not regulate emissions directly, they
do exert an indirect influence on emission controls. Depending
upon the allowable atmospheric concentration for N02, regulatory
guidelines within a specific region may or may not require FGT.
The Clean Air Act Amendments of 1970 established two types of
air quality standards, primary and secondary.
Primary standards are those set for the protection of
the public health with an adequate margin of safety. For N02
this standard is based on health effects studies conducted in
1968-69 and in 1972. The primary ambient air quality standard
for N02 is 100 yg/rn3 measured as an annual arithmetic mean.
Secondary ambient air quality standards are set for
the protection of the public welfare. These standards address
impacts upon soils, water, crops, vegetation, manmade materials,
animals, wildlife, weather, visibility, climate, and personal
comfort and well being. For N02, the secondary standard is
equivalent to the primary standard, 100 yg/m3.
According to the Clean Air Act, a state may choose to
enact ambient air quality standards which are stricter than the
corresponding federal standards. For N02, four states have done
so: California, Hawaii, New Mexico, and North Dakota. These
standards are summarized in Table 8.
N02 is considered the most toxic and potentially dan-
gerous oxide of nitrogen found in the atmosphere. It has been
studied in more detail than any of the other nitrogen oxide com-
pounds. The existing data on health effects due to N02 exposure
have been summarized several times. Two such summaries may be
40
-------
TABLE 8. STATE AMBIENT AIR QUALITY STANDARDS WHICH ARE
MORE STRINGENT THAN NAAQS FOR N02
State Ambient Air Quality Standard
100 ug/m3 annual arithmetic mean
California 3 (Q 25 ppm) f()r a 1_hour average
TT . . 70 ug/m3 annual arithmetic mean
150 ug/m3 for a 24-hour average
New Mexico 10° yg/m3 annual arithmetic mean
200 ug/m3 for a 24-hour average
North Dakota 10° ^/m* annual arithmetic mean
200 yg/m3 maximum 1-hour concentration
Source: Reference 28.
found in References 29 and 30. Criteria documents for the short-
term and annual average standard for N02 are in preparation and
will include summaries of health effects data also. The graphi-
cal summaries presented in Figures 1 and 2 show the levels of
N02 at which health effects have been observed and their rela-
tion to the level dictated by the National Ambient Air Quality
Standards for N02 . Most of the effects reported have been on
the respiratory system, but some N02 health system effects have
also been reported. The lowest reported level of N02 which has
been associated with human health effects is 200 ug/m3 (0.1 ppm) .
This level reportedly caused bronchioconstriction in 13 out of 20
asthamatics tested (Reference 31). The work is highly controver-
sial, however, and will have to be confirmed by further studies.
41
-------
N>
1000
Fatal
100
Edema
Eye and itose
Irritation
10
Decreased Pulmonary)
Function /
(Minutes to Hours) )
ACUTE
CHROMIC
Increased Protein
Level In Urine
0.1
Concentration of
NO- In ppm
0.01
'//////
1000
minutes
10
Occupational Standard
5 ppm
Highest Level
Recorded In I. A.
]
0.1
Primary Air Quality Standard (Annual Mean)
0.05 ppm
2 iv nl
Reduced Clearance
Capacity for Bacteria
I Increased Incidence of
Lower Respiratory Infection
Figure 1. Observed Effects of NOj on Humans.
Source:- Reference 30. Reprinted by permission from Electric. Power Research Institute
-------
ACUTE
co
1000 r-
100
Fatal (Monkeys)
Epithelium Changes (Mice).... .
Bronchlolltls Lesions .
.(Hamsters)
10
Reduced Mucoclllary Transport)
(Rats) '
Increased Susceptibility to)
Pneumonia (Mice) I
Bactericidal Dysfunction -
hours (Mice)
Increased Llpoperoxlda--
tion
0.1
Concentration of NO,
In ppra
0.01
777///.
hours
weeks
days
hours
hours
liypcrplasla of Type II
Cells - Days (Guinea
Hours
1000
CHRONIC
100
Emphysema Lesions'
(Rabbits)
Lung Height Increase.
Temporary Increase In I
Type II Cells, and Em-|
physerna (Rats)
1
Increased Breath-/*
Ing Rate Sustained
(Rats)
Monthly Average 1n-
Southern California
0.1
Primary Air Quality Standard for Annual Mean
O.OS ppm
'Natural Background Level'
of NO,
0.01
'Pulmonary Edema (Rats. Rabbits. Guinea
,Pigs)
— Emphysema In Terminal Bronchioles
' .(Rats)
atal (Newborn Rats)
Fibrosls (?)
Decreased Compliance and Increased Lung
Weight (Rats)
lypcrtrophy of Bronchial Epithelium
(Monkeys)
Slight Emphysema (Combined Challenge with
Flu Virus - Monkeys)
Reduced Resistance to Infection; After
One Year. Emphysema and Alveolar Extension
(Mice)
Source:
Figure 2. Observed Effects of N02 on Animals.
Reference 30. Reprinted by permission from Electric Power Research Institute
-------
Although the present state of knowledge indicates that
NOX acts as a precursor in the formation of photochemical oxidants,
as of yet, there has been no attempt to control ambient oxidant
levels by limiting ambient N02 levels (Reference 32). For the
protection of the public health, the Primary Ambient Air Quality
Standard for oxidants has been set at 160 yg/m3 (0.08 ppm) for a
one-hour exposure. This standard is based on a series of health
studies, a good summary of which can be found in Reference 33.
Additionally, the new criteria document for photochemi-
cal oxidants, now in preparation, will contain a health effects
summary. As is the case with NOa, most of the reported oxidant
effects are seen in the respiratory system. Some extrapulmonary
effects have also been reported, however. Dose response curves
constructed by recent human exposure studies show pulmonary sys-
tem effects for healthy subjects exposed to 03 at levels of 720
ug/m3 (0.37 ppm) and higher for 2 hours (References 34, 35 and
36). However, there is some evidence of health effects in
healthy human subjects at a level as low as 490 yg/m3 (0.25 ppm)
(Reference 37). There are indications that exposure to photo-
chemical oxidants mixtures may cause effects at lower levels of
03 than the 490-720 yg/m3 (0.37 to 0.25 ppm) figure derived from
03 alone (References 21, 38 and 39). Recent studies of exposure
of healthy humans to PAN showed no physiological effects from
levels commonly found in the environment. However, PAN and
similar, compounds are notorious eye irritants at ambient levels
(References 40 and 41).
Results of related, ongoing studies of both N02 and
oxidant health effects have the potential for effecting tighter
NOX controls if they are shown necessary to protect the public
health.
44
-------
State Implementation Plans
State Implementation Plans (SIPs) are the regulatory
mechanisms used to achieve and maintain compliance with ambient
air quality standards within a state's boundaries. The Clean
Air Act requires that each state design its own control strategy
~o accomplish this.
Although all states are required to comply with NAAQS,
each state is confronted with a different set of emission sources
and resulting air pollution problems. Therefore, SIPs can vary
widely from state to state. The primary objective of each SIP
is to alleviate pollution problems either by controlling or eli-
minating specific emission sources, including both new and exist-
ing sources.
In almost every instance, only two industrial opera-
tions, combustion units and nitric acid plants, have NOX emission
limits set by SIPs. In 35 of 50 states, existing combustion units
are unregulated with respect to NOX emissions. In the remaining
15 states, the NOX regulations for existing combustion units vary
considerably. Most of these states, however, require a unit to
have a heat input of at least 73 MW (250 million Btu/hr) before
coming under regulation. Usually, these regulations are based on
the type of fuel burned. Typical ranges for NOX emissions limits
are as follows: gas: 86-130 g/GJ (0.2-0.3 lb/106 Btu); oil:
130-260 g/GJ (0.3-0.6 lb/106 Btu); coal: 300-390 g/GJ (0.7-0.9
lb/106 Btu). For new, modified, or expanded combustion unit, 32
states regulate the maximum allowable NOX emissions. For the
most part, these states have adopted Federal NSPS (see Section
4.1.2) as a part of their SIPs (Reference 42).
45
-------
Regulation of NOX emissions from nitric acid plants is
also somewhat varied among the states. Less than half (21) of
the states have SIP regulations for existing nitric acid plants.
These regulations range in the severity of their emission limits.
from 1.5-20 g N02/kg (3.0 to 40.0 Ib per ton) of acid produced.
•In some instances, the nitric acid plant must also meet an opa-
city requirement of from 5-2070. For new, modified, or expanded
nitric acid plants, almost two-thirds (32) of the states have
imposed SIP regulations. For the most part, these states have
adopted Federal NSPS (see Section 4.1.2) and tightened the.opa-
city requirement to less than 1070 (Reference 42) .
Currently there are no NOX limitations in any SIPs
aimed at the control of photochemical oxidants. Control of
hydrocarbon emissions aimed at NAAQS attainment has been empha-
sized for oxidant control.
Unlike most states, California has an implementation
plan for each county. The California NOX regulations are usually
more stringent than those found in most SIPs. For instance, in
the Southern California Air Pollution District, NOX emissions
from existing steam generators are limited to 160 mg/m3 (125 ppm)
(gas) and 280 mg/m3 (225 ppm) (liquid and solid), regardless of
size. New units have even stricter NOX limitations: 100 mg/m3
(80 ppm) (gas), 200 mg/m3 (160 ppm) (liquid), and 280 mg/m3 (225
ppm) (solid) (Reference 43). Currently, a proposed regulation
requiring 907, reduction from present NOX emission levels is being
considered for all utility boilers in the South Coast Air Basin.
4.1.2 New Source Performance Standards
The purpose of New Source Performance Standards (NSPS)
is to prevent the degradation of existing air quality. The aim
46
-------
is to avoid future air pollution problems by establishing stan-
dards of performance for new stationary:sources and modified or
expanded existing sources. The "standard of performance" for a
new stationary source is based on the best emissions control
system, which is both proven and available at a reasonable cost.
The degree of emission reduction which can be achieved with this
system is'designated by EPA as the "standard of performance".
Therefore, NSPS are a direct result of technology availability
and cost.
Currently only steam generators with a heat input
greater than 73 MW (250 million Btu/hr) and nitric acid plants
are affected by NSPS regulations. For steam generators, NOX
emissions are regulated according to fuel: gas, 86 g/GJ (0.2
lb/106 Btu); oil, 130 g/GJ (0.3 lb/106 Btu); coal, 300 g/GJ (0.7
lb/106 Btu). For nitric acid plants, NSPS guidelines limit NOX
emissions to 1.5 g/kg (3.0 Ib per ton) of 1007» acid produced
(References 44, 45 and 46).
4.1.3 New Mobile Source Standards
As is the case with New Source Performance Standards,
mobile source standards are designed to prevent the degradation
of existing air quality. All NOX emission standards for motor
vehicles have proven technology as their basis. The 1977'amend-
ments to the Clean Air Act have established NOX mobile source
standards for both light- and heavy-duty gasoline-fueled vehicles
or engines. For light-duty vehicles these standards set NOX emis-
sions limits at 1.2 g/km (2.0 g/mile) for 1977-1980 models and
0.62 g/km (1.0 g/mile) for 1981 and thereafter. For heavy-duty
vehicles, the standards require a reduction of at least 7570 from
the average of the actually measured emissions from a vehicle or
engine manufactured during the baseline model year (Reference 47)
47
-------
California has adopted more stringent mobile source standards
for light-duty vehicles. Currently, California limits NOx emis-
sions from new light-duty vehicles to 0.93 g/km (1.5 g/mile).
By 1982 this NOX limit will be further reduced to 0.24 g/km (0.4
g/mile). A waiver is granted for light-duty vehicles using
diesel engines. Diesel light-duty vehicles are limited to 0.93
g/km (1.5 g/mile) for the years 1981-1984.
4.2 Trends in NOX Legislation
Future trends in NOX legislation will continue to be
driven by the results of health effects research and advances in
emission control technologies. In addition, the link between
ambient NOx and the production of photochemical oxidants may '
lead to NOx legislation aimed at oxidant control. This will
depend on the research findings concerning the mechanism of
photochemical oxidation. At present, no NOX regulations are
specifically designed for that purpose. The possibility that1
several AQCRs are approaching noncompliance status with respect
to ambient NO? levels may also provide incentives for future,
more stringent NOX legislation.
The Clean Air Act Amendments of 1977 will have a sig-
nificant impact on future NOX regulations in that they require
each state to submit a revised SIP that provides for the attain-
ment of primary NAAQS for N02 by December 31, 1982. The new SIP
must also provide for the attainment of NAAQS for areas experienc-
ing severe oxidant problems by December 31, 1987. Revised SIPs
will also have to address two new source regulatory policies :
emissions offset and prevention of significant deterioration
(PSD).
The offset policy requires that a permit program be
established for new or modified major facilities. Under this
48
-------
program, emissions from new sources must be either within new
growth allowances built into the revised SIP or be offset by
a reduction in emissions from another source within the area.
Cost will be a determining factor, but will be given less empha-
sis than in the case of NSPS.
The PSD policy provides for the protection of public
health and welfare and the preservation, protection, and enhance-
ment of all Class I areas (national parks, monuments, forests,
preserves, and recreation areas). Significant deterioration is
measured by pollutant levels and visibility. New stationary
sources seeking to locate in Class I areas must meet preconstruc-
tion 'requirements ensuring that NOX emissions from the source
will not exceed an allowable increment for NOX. New sources are
also required to use the best available control technology. The
visibility stipulations include color as well as visibility re-
duction. Nitrogen dioxide contributes to both conditions since
it is both a precursor to photochemical smog and a brown gas.
Preconstruction requirements must also be met for new
or modified sources in non-attainment areas. In these cases, the
source must comply with the Lowest Achievable Emission Rate (LAER)
before being permitted for construction and operation.
Based on data which indicate that short-term exposures
to N02 may cause adverse health effects, the 1977 CAA amendments
may require EPA to set a short-term N02 ambient air quality stan-
dard. Discussions with EPA's Office of Air Quality Planning and
Standards indicate that this standard will probably range from
200-1000 yg/m3 for a one to three hour average.
As technological advances are made, regulations reflect-
ing the state-of-the-art of control technology will change. These
49
-------
regulations include New Source Performance Standards and New Mo-
bile Source Standards. Currently, NSPS regulations for NOX from
electric utility generating stations are being considered for
revision. The new standards proposed on September 19, 1978 apply
to electric utility steam generating units capable of firing more
than 73 MW (250 x 10s Btu/hr) heat input of fossil fuel. A sum-
mary of the provisions applicable to the most common fuels are
summarized in Table 9.
TABLE 9. SUMMARY OF NSPS FOR NOX EMISSIONS FROM
ELECTRIC UTILITY GENERATING STATIONS
Fuel Proposed NSPS
natural gas 86 g/GJ (0.2 lb/106 Btu)
oil 130 g/GJ (0.3 lb/106 Btu)
coal 260 g/GJ (0.6 lb/106 Btu)
lignite 210 g/GJ (0.5 lb/106 Btu)
Source: Reference 48.
Several new emission source categories are also being
investigated. Screening studies are underway for adipic acid
manufacturing, dimethylterephthalate/terephthalic acid plants,
explosives (high and low), fiberglass, textile, and wool manu-
facturing, to determine the need for NOX NSPS. Technical studies
are either in progress or completed for stationary internal com-
bustion engines (diesel and gasoline), stationary gas turbines,
and steam generators with a heat input of 0.09-73 MW (0.3-250 x
106 Btu/hr) (gas, oil, and coal) to provide detailed information
for NOX NSPS (References 44 and 45).
Mobile source standards may change somewhat. Current-
ly, the 1982 California standard of 0.24 g/km (0.4 grams per
50
-------
mile) appears to represent the best level achievable with avail-
able control technology.
In summary, it appears that current trends in NOX regu-
lations are toward more stringent control of all sources.
51
-------
5.0 CONTROL STRATEGIES
Unless a level of control requiring FGT is adopted
by some future NSPS, the question of "if and when" FGT will be
necessary will ultimately be decided in the development of NOX
control strategies such as the ones included in State Implementa-
tion Plans. Strategies for control of NOX cannot be developed
on a national level because the NOX problem is a very area-
specific one which depends on topography, meteorology, and
emission source characteristics.
The purpose of this section is not to develop control
strategies for individual AQCR's but to determine how strategies
for attaining and maintaining compliance with various sets of
standards affect the need for FGT technology. The various tech-
nologies for controlling NOX emissions from stationary sources
and mobile sources are considered. The manner in which these
control techniques might be applied to attain or maintain com-
pliance with various standards is discussed.
5.1 Methods of Control
Techniques that will reduce NOX emissions from both
stationary and mobile source fossil fuel combustion fall into
two general categories: combustion modification (CM) which
limits NOX formation and post-combustion flue gas treatment (FGT)
A variety of techniques and processes are available or under
development in each of these areas. In the following subsec-
tions the status of development, cost effectiveness and tech-
nical limitations of both of the generic classes of processes
are briefly discussed. Further information on specific techni-
ques and processes is given in Appendix A.
52
-------
5.1.1
Stationary Sources
Major sources of concern in the development of an NO
control strategy for stationary sources are utility boilers and
large industrial boilers and furnaces. In 1975, contributions
from this class of sources to total NO emissions in each of the
twenty problem AQCR's previously discussed ranged from 10-8070.
In five of the AQCR's, utility boilers alone accounted for 4070
or more of the total N0x emissions.
The control techniques that have been developed or are
currently undergoing development for those sources are: combus-
tion modification, fluidized bed combustion, and flue gas treat-
ment. Other stationary sources which have been indicated as
major contributors to N0x emissions in some areas are commercial
and residential furnaces (Reference 49). Since control techni-
ques for these small sources are being studied and/or developed
at this time, definitive cost data are unavailable. Comparative
costs for CM and FGT as applied to large combustion sources are
shown in Table 10. The economics of fluidized bed combustion
are not yet well established, and therefore, the cost effective-
ness of this approach to NOX control cannot be assessed at this
time.
TABLE 10. COMPARATIVE COSTS OF STATIONARY SOURCE CONTROLS
Source
Type of Control
Level of
Control
Cos t
Utility Boilers
Combustion modification:
New Boilers 30-50% $110/Mg NOX ($LOO/ton NOX)
Retrofit 30-50% $248/MR N0y (S225/ton NOX)
Flue Gas Treatment
Selective catalytic
reduction 50-90% $1320/Mg NOX ($!200/ton N(K)
Fluidized Bed Combustion 40-70°;
Industrial Boilers Combustion Modification 25-65% $165/Mp NOX ($1.50/ton NOX)
Source: Reference 50.
53
-------
As indicated in Table 10, in most stationary source
applications, combustion modifications will be the most cost
effective technique for reducing NOX emissions. There are
several combustion modification techniques available for control
of stationary sources. These include low excess air (LEA) fir-
ing, staged combustion, flue gas recirculation (FGR), burner
modification and steam injection. All of these techniques re-
duce the formation of NOX in a boiler or furnace by either lower-
ing the flame temperature or lowering the oxygen concentration
in the flame front, or both. The best technique or combination or
techniques will be a function of the degree of removal required
and the effect of the modification on process economics. Cur-
rently, the most significant potential disadvantage of combustion
modification is that some techniques create a reducing atmosphere
in the lower section of the boiler which may lead to accelerated
tube wastage. Tests are currently underway to resolve the issue
of tube wastage during substoichiometric firing. If these tests
indicate no significant wastage, then new boilers can be expected
to be equipped with provisions for LEA and staged combustion.
The additional need for and value of FGR has not yet been defi-
nitely established. Combustion modifications will generally be
the first retrofit technique applied to existing boilers in
areas where ambient air quality standards are being exceeded.
This is already the case in California (Reference 51) .
Fluidized bed combustion (FBC) is an emerging technology
specifically applicable to coal. The feature of FBC which affects
NOX emissions is the relatively homogeneous temperature of about
840°C (1550°F) which exists throughout the bed. Conventional
boilers have a temperature gradient from the flame to the wall
with flame temperatures on the order of 1370°C (2500°F). The
lower combustion temperature of FBC kinetically inhibits forma-
tion of thermal NOX and, thereby, leads to lower NOX emissions.
FBC is currently being tested on a large scale (30 MWe) test
54
-------
unit and is expected to be available commercially within the
next decade (Reference 52). However, the extent to which FBC
will be applied cannot be determined at,this time. Currently,
the impetus for development of FBC is not its NOX reduction capa-
bility, but rather its combination of advantages over a conven-
tional boiler followed by a flue gas desulfurization (FGD) unit.
Flue gas treatment differs from combustion modification
and fluidized bed combustion in that N0x is reduced or removed
from the flue gas after formation. Generally, FGT processes
will be significantly more expensive than combustion modification.
As a result, it is likely that FGT processes will be applied only
in addition to combustion modification in situations where ad-
ditional control is necessary.
The NO -only FGT processes, both catalytic and non-
catalytic selective reduction, will soon be in a stage of develop-
ment which will allow commercial application. In fact, demonstra-
tion of this technology has been ordered by GARB.
Current economics seem to indicate that the most pro-
mising combination for maximum N0x and SOa control from a con-
ventional coal-fired boiler appears to be application of 1) LEA
and staged combustion, 2) a selective catalytic or noncatalytic
FGT system, and 3) a conventional FGD unit. However, in every
application, there will be specific technical and economic factors
which will determine the best combination of control options.
As further development of simultaneous N0x/S02 FGD processes
continues, it may be demonstrated that it is more economical
to operate a single simultaneous NO /S02 removal process than
two individual processes in situations where removal of both
pollutants is required (see discussion in Appendix A).
55
-------
5.1.2
Mobile Sources
The major mobile sources of concern in the development
of an N0x control strategy are light duty vehicles (passenger
cars), light duty trucks (pick-ups and vans), and heavy duty
vehicles (trucks and buses). In 1975, contributions from this
class of sources to total NO emissions in each of the twenty
problem AQCR's previously discussed ranged from 15-617o. In
i
eight of those AQCR's mobile source emissions accounted for
more than 5070 of the total NO emissions. The NO control tech-
X X
niques that have been developed or are currently undergoing
development for these sources are 'combustion modification and
catalytic reduction. Comparative 'costs of mobile source emis-
sion controls as a function of source type and level of control
achievable are shown on Table 11.
TABLE 11. COMPARATIVE COSTS OF MOBILE SOURCE CONTROLS
Source
Type of Control NOX emission reduction
From To
Cost
Light duty vehicles
and
Light duty trucks
Light duty vehicles
Heavy duty diesel
Heavy duty gasoline
Combustion
Modification
Catalytic
Reduction
Combustion
Modification
Combustion
Modification
1.2 g/km 0.62 g/km. $500/Mg NOX
(2.0 g/mile) (1.0 g/mile) ($450/ton NOX)
0.62 g/km 0.24 g/km $2530/Mg NOX
(1.0 g/mile) (0.4 g/mile) ($2300/ton NOX)
12.6 g/km 3.7 g/km
(20.3 g/mile)(6 g/mile)
$450/Mg NOX
($4107ton NOX)
8.6'g/km 1.9 g/km $740/Mg NOX
13.8 g/mile (3 g/mile) ($669/ton NOX)
Source: Reference 5G-.
56
-------
Combustion modification for NO control has been ap-
plied to light and heavy duty vehicles for several years. The
techniques employed (air/fuel ratio control, exhaust gas recycle,
stratified charge, etc.) are very similar, both in concept and
in practice, to the combustion modification techniques previously
discussed for stationary sources. However, the cost of combus-
tion modifications is roughly three to four times higher for
mobile sources per ton of NO removed than for stationary sources
X
Exhaust gas treatment (use of catalytic converters)
has not yet been commercially applied to light duty vehicles.
(Catalytic converters for N0x control should not be confused
with those currently in use for hydrocarbon control). The costs
for this post combustion cleanup are estimated to be approxi-
mately twice the estimated costs for stationary source FGT on
a per ton of NO removal basis. Considerable research and
development is underway to develop this technology in antici-
pation of tighter federal standards. Regardless of the final
decision on Federal mobile source standards, exhaust gas treat-
ment will be required in some states (notably California) in
order to meet state standards.
In addition to the high costs associated with post-
combustion N0x control of mobile sources, there are other mobile
source control problems to be considered. The number of sources
to be controlled is huge relative to the number of large sta-
tionary sources, making retrofit to existing sources virtually
impossible. In addition, consideration must be given to guaran-
teeing proper maintenance and inspection of controls.
5.2 Attainment and Maintenance of Standards
The N0x standards which impact the need for FGT can
be classified into two groups : emission standards and ambient
57
-------
air quality standards. The emission standards of interest are
Federal NSPS and any state regulations adopted in SIPs.
The ambient air quality standards of interest are the
Annual Average N02 NAAQS, the recently mandated Short Term N02
NAAQS individual State, short term NCh standards, and PSD re-
quirements . The degree to which attainment and maintenance of
each of these standards affects the need for FGT is discussed in
the following subsections.
5.2.1
New Source Performance Standards
NO emissions vary with the type of fuel used and the
method of firing the fuel. Typical uncontrolled emissions for
various fuels are shown in Table 12. Also shown are the emis-
sion limits specified by current NSPS and the percent NO reduc-
X
tion corresponding to those limits.
TABLE 12. TYPICAL UNCONTROLLED NOX EMISSIONS FROM LARGE FOSSIL
FUEL-FIRED STEAM GENERATORS
Fuel
NOX Emissions
g/GJ
(lb/106 Btu)
NSPS
g/GJ
(lb/106 Btu)
% Reduction Required
to meet NSPS
Natural Gas
(1050 Btu/ft3)
Fuel Oil
(150,000 Btu/gal)
Coal (11,900 Btu/lb)
27
(0.062
37
(0.13
149
(0.32
- 200
- 0.47)
- 350
- 0.81)
- 470
- 1.1)
86
(0.2)
130
(0.3)
300
(0.7)
0-57
0-63
0-36
Source: Reference 53.
58
-------
Congress in the 1977 Clean Air Act Amendments man-
dated that the current NSPS be reviewed and revised as appro-
priate. FGT is not being required by the revised Federal NSPS
for large steam generators. Although the NSPS are being tightened,
they specify a level of emissions attainable through use of com-
bustion modification techniques alone. This means that for the
near term, new and expanding sources will not be required to
use FGT to comply with the Federal NSPS. However, it should be
noted that FGT technology is being evaluated to determine its
feasibility as the basis for a NSPS for steam generators with
less than 73 MW (250 Btu/hr) heat input.
Table 13 shows the IERL/RTP research goals for station-
ary source NO emission reduction. It is quite likely that as
these goals, are achieved, the NSPS will again be reviewed.
If the technical and economic feasibility of applying
new control technology can be demonstrated, it is very possible
that the NSPS will be reviwed to require emission limits sub-
stantially lower than those currently established or proposed.
The 1985 goal for coal-fired utility boilers of 125 mg/m3 (10.0
ppm) M0x emitted at 3% 02 is roughly equivalent to 64 g/GJ (0.15
lbs/106 Btu). This level of control would either require FGT or
a substantial improvement in the current state-of-the-art of
combustion modification.
5.2.2 Annual Average Ambient Air Quality Standard
Currently the only oxide of nitrogen for which air
quality standards exist is nitrogen dioxide (N02). In Section
3.0 it was shown that NO emissions from direct fossil fuel
X
combustion account for more than 9070 of the total U.S. anthropo-
genic emissions of N0x. As a result, any strategy for controlling
the ambient concentration of N02 or secondary NO pollutants
59
-------
TABLE 13. EPA R&D PROGRAM NOX CONTROL TARGETS
Utility boilers
Gas
Oil
Coal
Industrial boilers
Gas
Residual oil
Coal
Reciprocating engines
Spark ignition-gas
Compression ignition-oil
Gas turbines
Gas
Oil
188
280
688
188
406
560
3750
3125
500
188
750
280
0
(150)*
(225)*
(550)a
(150)
(325)
(450)
(3,000)
(2,500)
(400)
(150)b
(600)
(225)D
125
188
250
100
156
188
1500
1500
94
156
(100)
U50)
(200)°
(80)
U25)
(150)°
d
(1,200)°
(l,200)d
(75)d
d
(125)a
62
112
125
62
112
125
500
1000
31
31
(50)
(90)
(100)
(50)
(90)
(100)
(400)
(800)
(25)
(25)
Current NSPS.
Estimated achievable with wet control technology.
c
Developed and field-applied technology.
Developed technology.
Source: Reference 54.
60
-------
must concentrate or reducing NO emissions from fossil fuel com-
X
bustion.
As shown in Section 3.0, there are only a few AQCR's
currently out of compliance with the National Ambient Air Quality
Standard for N02 of 100 yg/m3 measured as an annual arithmetic
mean. Furthermore, the margin by which the air quality standard
is exceeded is small. Because of these facts, stringent NO
X
emission controls such as FGT have to date not been required in
order for AQCR's to attain compliance. There has been concern
expressed, however, that growth will cause many more AQCR.'s to
be out of compliance with the annual average standard and by a
wider margin than those seen today (Reference 55). If the N02
problem worsens, there may be more impetus for implementing
more stringent control measures such as FGT.
Twenty AQCR's which potentially have a problem attain-
ing or maintaining compliance with N02 ambient air quality stan-
dards were identified in Section 3.0. The emissions profiles in
1975 for these twenty problem AQCR's are presented in Table 4.
It is obvious from an analysis of this data that identical con-
trol strategies would not be appropriate for all twenty AQCR's.
Some of the AQCR's are dominated by stationary sources, some by
mobile sources. Strategies for controlling ambient concentra-
tions of NO2 and secondary NOX pollutants in any given AQCR
should, therefore, be developed on a case by case basis consid-
ering not only the emissions inventory for the specific AQCR un-
der consideration, but also such diverse factors as the topo-
graphy and meteorology of the area.
If the decision is made that N0x emissions must be
reduced to attain or maintain the annual average air quality
standards in a given AQCR there are several options to be con-
sidered in development of a viable control strategy.
61
-------
Tightening mobile source emission standards is one
control option for AQCR's experiencing difficulty in meeting the
annual average standard. The Federal mobile source emission stan-
dard has recently been lowered. Emissions for 1977-1980 models
are limited to 1.2 g/km (2.0 g NO /mile), and for the 1981 and
later models they are limited to 0.62 g/km (1.0 g N0x/mile).
Control strategies for problem areas that are mobile source domi-
nated might include lowering the standards even more until the
practical limit of 0.4 g N0x/mile is reached. As was shown in
Section 5.1.2 the cost of this control option is very high and
many technical and administrative problems have to be considered.
Another option for AQCR's experiencing difficulty in
attaining or maintaining the annual average N02 standard, espe-
cially one in which the NO emissions are stationary source
dominated, would be to tighten stationary source emission regu-
lations. As was discussed in Section 5.1.1, if it is considered
necessary to tighten controls on NO emissions from large sta-
X
tionary sources, CM techniques are the usual first step because
of the high cost (per ton N0x removed) of FGT. However, applica-
tion of FGT could be considered if the practical limit of control
by combustion modification had been reached and additional re-
ductions in NOX emissions were required for attainment or main-
tenance of the standard.
In light of the uncertainties discussed in Section 3.0,
with respect to the exact relationship between emissions of NO
and ambient concentrations of N02, it is highly probable that the
most common control strategy that would be applied to a problem
AQCR would represent a combination of the two options discussed
above. Using the comparative costs previously presented as a
guide to cost effectiveness, the first controls to be applied
would be combustion modification of stationary sources. This
62
-------
would be followed by combustion modification of mobile sources.
If additional reductions were required, FGT would be applied to
stationary sources and finally to mobile sources.
5.2.3 Short-Term Standard
Congress has directed EPA to set a short-term standard
for N02 unless it is deemed unnecessary for the protection of
public health. It is not clear how many AQCR's would have trouble
complying with this type of standard, but it is generally assumed
that a short-term standard would be a stricter standard than
the current annual average standard.- A number of AQCR's that
have potential problems complying with the annual average standard
can be expected to also have problems complying with the short-
term standard.
The control options for attaining and maintaining com-
pliance with a short-term standard would be the same as those
discussed in the previous section on the annual average standard.
However, it should be noted that FGT might be required in some
AQCR's to meet a short-term standard, even though it might not
have been required to meet the annual average standard.
Other areas of uncertainty exist in the comparative
contributions of stationary and mobile sources to the high short-
term levels. It is possible that both types of sources could
contribute to peak concentrations. Conditions under which plumes
might impact ground level with very little dispersion are pre-
sented in Section 3.3. However, rush hour traffic has been
demonstrated as a cause of high N02 levels from mobile sources
in many cities and mobile sources are thought to be major of-
fenders, as illustrated in Figure 3 for Southern California.
Strategies for lowering peak NC-2 levels have not yet been de-
veloped. Attempts may be made to lower peaks by reducing back-
63
-------
2SO
200
ISO
100
SO
NO
0400 0800 1200 1600 2000 2400 0400
TItt
Figure 3. Variations in NO and NO2 in Orange
County, California, October, 1974.
Source: Reference 29.
64
-------
ground levels. If this is the case, stationary source controls
such as FGT may be indicated.
Several states have existing short-term standards.
California is a noteable example because of the stringency of
the standard and because of the severity of the N02 problem in
the South Coast Air Basin (SCAB). The California standard is
500.]jg/ra3, one hour maximum. In an effort to achieve the stan-
dard, mobile source emission standards are being lowered to
0.24 g/km (0.4 g/mile). Combustion modification techniques
have already been implemented in the SCAB. The California Air
Resources Board (GARB) believes that in addition to these con-
trols, FGT systems will be required for large combustion sources
The NO problem is a severe one in California. GARB
estimates that even if all stationary source emissions were
eliminated and the 0.24 g/km (0.4 g/mile) mobile standard were
implemented, compliance with the 500 yg/m3, one-hour N02 stan-
dard would only be marginally achievable in the South Coast Air
Basin.
In a related study (Reference 56) calculations for the
Chicago AQCR have indicated that FGT may be required to attain
and maintain compliance with a short-term N02 standard, depend-
ing on the level and averaging time selected for the standard.
With a one-hour standard of 500 yg/m3 some FGT would be required,
If the standard were 250 yg/m3 (one-hour maximum), considerable
FGT would be required. However, for Chicago, the study showed
that compliance with a one-hour standard of 750 yg/m3 could be
attained by use of combustion modification alone.
Whether other areas in the U.S. will have similar prob-
lems complying with a short-term standard remains to be seen.
65
-------
The range of concentrations currently under consideration by EPA
for the one-hour maximum is 200-1000 ug/m3. Obviously the impact
of the proposed short-term N02 standard cannot be fully assessed
until there is final resolution of the level and the averaging
time allowed.
5.2.4 Prevention of Significant Deterioration
The previous discussion has centered on devising con-
trol strategies for bringing air quality in noncompliance or
near noncompliance areas within an acceptable range. There is
another factor to consider in control strategy development for
all AQCR's. The Clean Air Act Amendments of 1977 have specified
that control strategies for prevention of significant deteriora-
tion (PSD) of air quality must be developed. N02 is specifically
mentioned with allowable increments of increased N02 concentra-
tions to be set in the near future. It is conceivable that under
these provisions new fossil fuel-fired power plants and other
large fossil fuel burning stationary sources might have to em-
ploy stringent NO emissions controls to be allowed to locate
in certain areas. As previously discussed, if the emission con-
trols needed for stationary sources require more than 5070 NO
emissions reductions, application of FGT would be indicated.
Wilderness and national park areas have been given special em-
phasis with respect to nondegradation provisions. Protection
of air quality in these Class I areas is considered of primary
importance. Because of this, PSD may require FGT on any station-
ary fossil fuel burning source located in such an area in order
to protect the pristine quality of the air even where NSPS and
NAAOS would not indicate the need for FGT.
66
-------
REFERENCES
1. Robinson, E. and R. C. Robbins. Gaseous Atmospheric Pollu-
tants from Urban and Natural Sources. J. APCA 20, 303-306
(1970).
2. National Air Quality and Emissions Trends Report, 1976.
U.S. Environmental Protection Agency, Office of Air and
Waste Management, Office of Air Quality Planning and Stan-
dards. EPA-450/1-77-002. Research Triangle Park, North
Carolina. December, 1977.
3. An Analysis of Alternative Motor Vehicle Emission Standards.
United States Department of Transportation, U.S. Environ-
mental Protection Agency, U.S. Federal Energy Administration.
May 19, 1977.
4. Greenfield, Stanley, et'al. A Preliminary Evaluation of Po-
tential NOX Control Strategies for the Electric Power Indus-
try. Draft report. EPRI. April 30, 1977.
5. Preliminary Environmental Assessment of the Application of
Combustion Modification Technology to Control Pollutant
Emissions from Major Stationary Combustion Sources. Draft
Report TR-77-28. Mountain View, Calif. Acurex Corporation/
Aerotherm Division. March, 1977.
6. Habegger, Loren J., et al. Priorities and Procedures for De-
velopment of Standards of Performance for New Stationary
Sources of Atmospheric Emissions. EPA-450/3-76-020. Argonne
National Laboratory, Argonne, 111. May, 1976.
7. Air Quality Noise and Health, Report of a Panel on the Inter-
agency Task Force on Motor Vehicle Goals Beyond 1980. Office
of the Secretary of Transportation. Washington, D.C. March,
1976.
8. Henderson, V. and A.S. Hoffman. The Status of 1965 N0? Air
Quality Data, National Aerometric Data Bank. U.S. Environ-
mental Protection Agency Monitoring Section. 1976.
9. Ozone and Other Photochemical Oxidants. National Academy of
Sciences. Washington, D.C. 1977.
10. Altshuller, A.P. and J.J. Bufalini. Photochemical Aspects of
Air Pollution, A Review. Photochem. Photobiol. 4:97-146.
1965.
67
-------
11. Altshuller, A.P. and J.J. Bufalini. Photochemical Aspects
of Air Pollution, A Review. Environ. Sci. Technol. 5:39-64.
1971.
12. Leighton, P.A. Photochemistry of Air Pollution. New York
Academic Press. 1961. 300 pp.
13. Levy, H. II. Photochemistry of the Troposphere. Adv.
Photochem. 9:369-524. 1974.
14. Pitts, J.N.J. and B.J. Finlayson. Mechanisms of Photochemi-
cal Air Pollution. Angew. Chem. (Engl.) 14:1-15. 1975.
15. Pitts, James N., Jr., Robert L. Metcalf and Alan C. Lloyd,
eds. Advances in Environmental Science and Technology. Vol.
4. N.Y. Wiley. 1974.
16. England, Christopher and William H. Corcoran. The Rate and
Mechanism of the Air Oxidation of Parts-Per-Million Concen-
trations of Nitric Oxide in the Presence of Water Vapor.
I & EC, Fundam. 14 (1), 55. (1975).
17. Dodge, M.C. and T.A. Hecht. Rate Constant Measurement Needed
to Improve a General Kinetic Mechanism for Photochemical Smog,
In Int. J. Chem. Kinetics, Symp. No. 1, Chemical Kinetics
Data for the Upper and Lower Atmosphere, Proc. , Warrenton,
Va. , Sept. 15-18, 1974. Sidney W. Benson, et al. eds. N.Y.
Wiley. 1975. pp. 155-163.
18. Grosjean, Daniel, et al. Concentration Size Distribution
and Modes of Formation of Particulate Nitrate, Sulfate and
Ammonium Compounds in the Eastern Part of the Los Angeles
Basin. Presented at the 69th Annual APCA Mtg. Portland,
Oregon. June 27-July 1, 1976.
19. Dimitriades, Basil and A. Paul Altshuller. International
Conference on Oxidant Problems: Analysis of the Evidence/
Viewpoints Presented, Part I. Definition of Key Issues.
Journal of the Air Pollution Control Association 27: 4
299-307 (1977).
20. Control Strategies for Oxidant and Nitrogen Dioxide, A Re-
port on the ARE Photochemical/Transport Workshop at UCLA.
January 6-7, 1977. State of California Air Resources Board.
January 25, 1977.
21. Pitts, James N., Jr. Keys to Photochemical Smog Control.
Env. Sci. Tech. 11 (5), 456 (1977).
68
-------
22. Scientific and Technical Assessment Report on Nitrosoamines.
EPA-600/6-77-001. U.S. EPA. Office of Research and Develop-
ment. Washington, B.C. November, 1976.
23. Fine, David, et al. N-Nitrose Compounds. Detection in Am-
bient Air. Science 192. 1328-1330. (25 June 1976).
24. Finlayson, Barbara J. and James N. Pitts, Jr. Photochemis-
try of the Polluted Troposphere. Science 192 (4235), 111
(1976).
25. Ozone and Other Photochemical Oxidants. National Academy of
Sciences. Washington, D.C. 1977.
26. Hidy, G.M. and C.S. Burton. Atmospheric Aerosol Formation by
Chemical Reactions. In Int. J. Chem. Kinetics, Symp. No. 1,
Chemical Kinetics Data for the Upper and Lower Atmosphere,
Proc., Warrenton, Va., Sept. 15-18, 1974. Sidney W. Benson,
et al., eds. N.Y., Wiley. 1975. pp. 509 ff.
27. Air Quality Data-1975 Annual Statistics, U.S. Environmental
Protection Agency, Office of Air and Waste Management, Office
of Air Quality Planning and Standards. EPA-450/2-77-002.
Research Triangle Park, North Carolina. May 1977.
28. Personal Communication with State Air Control Agencies.
29. Nitrogen Oxides, National Academy of Sciences. Washington,
D.C. 1977.
30. Ziskind, Richard and Donald Hausknecht. Health Effects of
Nitrogen Oxides. 571-1A. Electric Power Research Institute.
Palo Alto, California. February 1976.
31. Orehek, J., et al. Effects of Short-term, Low-level Nitro-
gen Dioxide Exposure on Bronchial Sensitivity of Asthematic
Patients. J. Clin. Invest. 57: 301-307. 1976.
32. Dimitriades, Basil and A. Paul Altshuller. International
Conference on Oxidant Problems: Analyses of the Evidence/
Viewpoints Presented, Part I. Definition of Key Issues.
Journal of the Air Pollution Control Association. 27:4 299-
307 (1977).
33. Ozone.and Other Photochemical Oxidants. National Academy of
Sciences. Washington, D.C. 1977.
69
-------
34. Hackney, J., et al., Experimental Studies on Human Health
Effects of Air Pollutants I. Design Considerations. Arch
Environ Health 30: 373-378. 1975.
35. Hackney, J., et al. Experimental Studies on Human Health
Effects of Air Pollutants. III. Two-Hour Exposure to Ozone
Alone and in Combination with Other Pollutant Gases. Arch.
Environ. Health 30: 385-390. 1975.
36. Hackney, J., et al. Experimental Studies on Human Health
Effects of Air Pollutants II. Four-hour Exposure to Ozone
Alone and in Combination with Other Pollutant Gases. Arch.
Environ. Health 30: 379-384. 1975.
37. Hazucha. M. Effects of Ozone and Sulfur Dioxide on Pulmo-
nary Function in Man. Ph.D. Thesis. McGill University.
Montreal, Canada. 1973. 233 p.
38. Schoettlin. C.E. and E. Landau. Air Pollution and Asthma
Attacks in the Los Angeles Area. Public Health Rep. 76:
575-578. 1961.
39. U.S. Department of Health, Education, and Welfare. Public
Health Service, National Air Pollution Control Administration
Summary and Conclusions, pp. 10-1-10-13. In Air Quality
Criteria for Photochemical Oxidants. NAPCA Publ. No. AP-63.
Washington, D.C.: U.S. GPO. 1970.
40. Drinkwater, B.L., et al. Air Pollution, Exercise, and Heat
Stress. Arch, Environ. Health 28: 171-181. 1974.
41. Raven, et al. Effect of Carbon Monoxide and Peropyacity
Nitrate on Man's Maximum Aerobic Capacity. J. Appl. Physiol.
36: 288-293. 1974.
42. Bureau of National Affairs. Environment Reporter. State
Air Laws. Sections 200-556. Washington, D.C.
43. Personal Communication with John Danielson, Southern Cali-
fornia Air Quality Management Department. March 13, 1978.
44. McCutcheon, G,.D/. NOX Emission Trends and Federal Regulation.
Chemical Engineering Progress. August, 1977. p. 58.
45. Ricci, Larry J. EPA Sets Its Sights on Nixing CPI's NOX
Emissions. Chemical Engineering. February 14, 1977.
p. 34.
70
-------
46. U.S. Environmental Protection Agency. Draft-Standards of
Performance for New Stationary Sources; Electric Utility
Steam Generating Units. November 29, 1977. p. 4.
47. The Clean Air Act of 1977. The Bureau of National Affairs,
Inc. Washington, B.C. September 9, 1977.
48. Standards of Performance for New Stationary Sources, Electric
Utility Steam Generating Units. Federal Register, 43 (182):
42170, 1978.
49. Preliminary Environmental Assessment of the Application of
Combustion Modification Technology to Control Pollutant Emis-
sions from Major Stationary Combustion Sources. Draft Re-
port TR-77-28. Mountain View, CA. Acurex Corporation/
Aerotherm Division. March, 1977.
50. Air Quality, Noise and Health, Report of a Panel of the In-
teragency Task Force on Motor Vehicle Goals Beyond 1980.
Office of the Secretary of Transportation, Washington, D.C.
March, 1976.
51. Barr, W.H., et al. Modifying Large Boilers to Reduce Nitric
Acid Emissions. CEP 73 (7), 59-68 (1977).
52. Foster Wheeler Energy Corporation. Conceptual Design for an
Atmospheric Fluidized Bed Steam Generator. ERDA Contract
No. EF-77-C-01-2583. November 1977.
53. Shimizer, A.B., et al. NOX Combustion Control Methods and
Costs for Stationary Sources - Summary Study. Final Report.
EPA 600/2-75-046. EPA Contract No. 68-02-1318, Task 2.
Mountain View, California. Acurex Corporation/Aerotherm
Division. September, 1975.
54. News Features. Chemical Engineering 84 (4) 36 (1977).
55. McCutcheon, G.P. NOX Emission Trends and Federal Regulation.
Chemical Engineering Progress. August, 1977. p. 58.
56. Eppright, B.R., et al. Impact of Point Source Control
Strategies on N02 Levels. Draft Report. EPA Contract
68-02-2608, Task 14. February 1978.
71
-------
APPENDIX A
STATIONARY SOURCE NOX CONTROL
TECHNOLOGY DESCRIPTIONS
A-l
-------
COMBUSTION MODIFICATION
In order to understand how combustion modifications
reduce N0x it is useful to examine some of the chemistry of NOX
formation. Only a brief description is given here. To begin
with, there are two mechanisms of NOV formation. In one mechan-
X
ism, NO is formed by reaction of oxygen with chemically bound
nitrogen in the fuel. The amount of NO formed in this manner
is roughly a function of fuel nitrogen and excess air concentra-
tions with higher levels increasing NO formation. In the other
mechanism, NO is formed through reaction of oxygen with nitrogen
in the combustion air. This reaction is governed primarily by
the flame temperature with increased NO formation occurring at
higher temperatures. NO formed by these two mechanisms is called
"fuel NOX" and "thermal NOX," respectively. The contribution
of each reaction to the total amount of NO formed varies with
fuel type and boiler configuration.
Combustion modification techniques that reduce these
emissions do so by altering one or both of the primary NOX for-
mation mechanisms. This means either reducing the oxygen con-
centration, lowering the flame temperature or both. In the sec-
tions that follow, each of the currently viable techniques is
examined with respect to:
principle of operation,
status of development,
operating experience,
advantages/disadvantages, and
costs.
Both new and retrofit situations are considered.
A-2
-------
Low Excess Air
Operating at low excess air (LEA) involves providing
combustion air at an air/fuel ratio close to the theoretical air
requirement for stoichiometric combustion. Without LEA firing,
boilers are operated with ^30% excess air to assure complete
fuel combustion. Higher excess air levels are not efficient since
they lead to excessive heat loss in the flue gas. The level to
which excess air can be reduced and still maintain complete com-
bustion is a variable that must be determined empirically for
each specific application. For example, less excess air is re-
quired when LEA is used singly than when it is used in conjunction
with other combustion modification techniques. LEA firing is
achieved by closing down on the combustion air dampers until the
desired flue gas Oa concentration is attained. When used in con-
junction with staged combustion, a portion of ,the excess air is
supplied by overfire air nozzles. In this situation, the over-
fire air dampers serve to regulate the amount of excess air.
LEA firing, while practiced since the 1950's, was not
applied to any great extent, until recently, due to the disadvan-
tages associated with this mode of operation. Running the boiler
at minimum excess air levels has the advantages of some increase
in boiler efficiency and decreased corrosion. On the other hand
there are the disadvantageous effects of increased hydrocarbon
and carbon monoxide emissions, increased ash generation with oil
fuels and decreased'carbon conversions with coal fuels.
LEA has been applied recently on many utility boilers
since it is one of the'most economical means of achieving a re-
duction in NOX emissions. LEA firing is usually applied in con-
junction with staged combustion -because this modification offsets
A-3
-------
the disadvantages of LEA firing. With LEA alone, NOX removals
on the order of 3070 have been experienced with all fuel types
(Ref. A-l).
The cost of installing LEA on various boiler sizes is
shown in Figure 1. The plot shows that there are economies of
scale in which costs are lower for larger units. It also shows
similar costs for gas and oil-fired boilers (about $l/kW). LEA
costs for coal-fired units are higher (about $1.25/kW).
Staged Combustion
The creation of a reducing atmosphere by staged com-
bustion involves a first stage of combustion in which the fuel
is incompletely burned by a substoichiometric amount of air. The
combustion is completed above the burners where air is supplied
through overfire air ports. Conversion of fuel bound nitrogen to
NOX is reduced via the reducing atmosphere in the substoichiome-
tric stage. In the reducing atmosphere, the fuel bound nitrogen
tends to form N2 rather than NO. In addition, the flame tempera-
ture is lower in the fuel-rich zone which reduces formation of
thermal NO.
Staged combustion cannot be retrofitted to all existing
boilers. Application to new boilers involves installation of
overfire air ports above the combustion zone, however, this is
often not possible with existing boilers due to structural con-
straints. In retrofit situations there are several ways to
approach staged combustion. One approach is to take burners out
of service and use them purely as air injection ports with the
remaining burners operating substoichiometrically. Alternatively,
some of the burners can be operated fuel-rich and the remainder
fuel-lean. A typical staged combustion configuration is shown in
Figure 2.
A-4
-------
KEY: 0 - Reported cost data
B - Assumed data which are representative of the
range of cost data reported
GAS
OIL
COAL
10
,-'1975 S^
\ Kw )
10'
1
.6
.4
.2
.1
0431
002
000
0
MW
4
0
0
8
0
0
MW
1
2
0
0
4
0
0
8
0
0
MW
1
2
0
0
Figure 1. Cost of low excess air firing.
(Source: Reference A-l)
A-5
02-3637-1
-------
Figure 2. An example of a staged combustion configuration.
(Source: Reference A-2)
A-6
02-3631-1
-------
Like LEA, staged combustion for N0x control was devel-
oped in the 1950's, but has not seen widespread use until recent
years. It is becoming a popular control technique when used in
conjunction with LEA firing, since these two modifications have
offsetting effects on boiler operation. There are some poten-
tial problems when modifying an existing boiler, however. If the
total fuel flow cannot be delivered with the air injection nozzles
taken out of fuel service, then boiler capacity may be limited.
Staged combustion will, most likely, be retrofitted only to
large boilers which have many burners, flexible fuel flow and
adjustable secondary air registers. Emission reductions with
staged combustion alone range from 50% with gas to.30% with coal
(Ref. A-l).
Advantages of staged combustion are that it requires
only minor equipment modifications and that the cost is low if
boiler capacity is not affected. Disadvantages are .that it can-
not be applied to all boilers, it can affect flame characteris-
tics, and it creates a reducing atmosphere in the substoichiome-
tric combustion zone. This corrosive reducing atmosphere occurs
even when staged combustion is used in conjunction with LEA and
is currently a topic of industrial concern, particularly with
respect to tube wastage (tube wastage rates are potentially
higher in a reducing atmosphere). Tests to determine these rates
have so far been inconclusive with some investigators finding
increased wastage (Ref. A-3) and others finding no effect (Ref.
A-2). The results of some long-term (one year) tests (Ref. A-2)
should be available in the near future to help resolve this
question.
Costs for implementing staged combustion are shown in
Figure 3. As with LEA, costs are similar for gas and oil, about
$1.50/kW but higher for coal, about $2/kW.
A-7
-------
1975 S
Kw
KEY: 0 - Reported cost data
10
,6
.4-
.06
- Assumed data which are representative of the
range of cost data reported
GAS
4
0
0
8
0
0
1
2
0
0
OIL
4
0
0
3
0
0
1
2
0
0
COAL
O
4
0
0
8
0
0
1
2
0
0
10
.6
.2
,06
MW
MW
MW
Figure 3. Cost of low excess air firing.
( Source: Reference A-l)
A-8
02-3635-1
-------
Flue Gas Recirculation
Flue gas recirculation (FGR), as its name implies, in-
volves injection of flue gas into the primary combustion zone.
This affects both mechanisms of NOX formation since both the
flame temperature and the 02 concentration are reduced. FGR
does require some major equipment modifications to duct work,
dampers, controls and fans. FGR for temperature control has been
used for the past 15-20 years, however, when used for temperature
control, FGR will not reduce NOX emissions since the flue gas is
injected into the bottom of the boiler rather than into the com-
bustion zone. FGR alone appears to be capable of achieving up
to a 307o (Ref. A-l) NOX emissions reduction with gas and oil
fuels, with coal, it tends to be less effective. However, as
with all combustion modifications, the actual control level
achieved will vary due to conditions unique to each particular
application. Potential disadvantages of FGR are increased pro-
cess control complexity and vibration of recirculation ducts and
fans (Ref. A-4). FGR is not being considered for NOX control as
strongly as staged combustion and LEA. This is probably due to
the relative complexity of FGR with no additional benefit in NOX
reduction over other techniques.
Costs of FGR are shown in Figure 4. The equipment
modifications required are reflected in the fact that FGR is
more costly than the combustion modification techniques previously
discussed. The costs are essentially the same for all fuels,
about $3/kW.
Burner Modification
As was discussed earlier, the flame temperature can
affect N0x formation. Since different flame temperatures occur
with different burner types, the firing pattern used in a given
boiler is a factor in the formation of thermal NC> . Table 1
X
A-9
-------
V
Kw
10
.6
.4-
.2
.1
.06
KEY: 0 - Reported cost data
B - Assumed data which are representative of the
range of cost data reported
4
0
0
GAS
•i 1-
8
0
0
1
2
0
0
OIL
4
0
0
8 1
0 2
0 0
COAL
4
0
0
S 1
0 2
0 0
0
10
.6
.4
.2
.06
MW
Figure 4. Cost of fl'je gas recirculation.
(Sourcs: Reference A-l)
A-10
02-3636-1
-------
shows the effect of firing type on emissions from coal-fired
power plants. Cyclone boilers generate the highest temperatures
and correspondingly emit the most NOX. Other firing methods have
lower flame temperatures and NOX emissions are correspondingly
lower.
TABLE 1. TYPICAL NOX EMISSIONS AS A FUNCTION
OF METHOD OF FIRING
Typical NOX Emissions
Method of Firing g/GJ lb/106 Btu
Vertical 137-188 0.32-0.44
Horizontally Opposed 206-227 0.48-0.53
Spreader Stoker 244-287 0.57-0.67
Tangential (corner) 253-304 0.59-0.71
Front Wall 236-364 0.55-0.85
Cyclone 471-728 1.1 -1.7
Source: Reference A-5.
Variation of burner operating parameters have been
found to have an effect on NOX emissions. The parameters of
interest include burner swirl, throat velocity and flame pattern.
A number of boiler manufacturers have responded to NOX regula-
tions by designing new boilers to incorporate established low-NOx
burner designs (Ref. A-6). Low-N0x burners have the advantage
of adjusting to load changes more easily than other systems com-
bustion modification techniques. No combustion modification tech-
niques cost data were available for low-NQx burners.
Water/Steam Injection
Water or steam can be injected into the flame zone with
the effect of lowering the flame temperature. This decreases the
formation of thermal NOV. It is a very effective means of reduc-
X
A-ll
-------
reducing N0x emissions; however, water injection is not widely
applied due to excessive thermal efficiency losses.
Summary
Of the combustion modification techniques presented,
the most promising are LEA and staged combustion. These tech-
niques are especially attractive when used in combination since,
in this mode, there is no loss in boiler efficiency (Ref. A-2).
This combination will allow a coal-fired boiler to meet
the current NSPS of 300 g/GJ (0.7 lb/106 Btu) (Ref. A-6) .* On
large boilers, it may be difficult to implement combustion modi-
fications without drastically modifying the fuel supply system
or reducing the capacity of the system.
-The proposed new NSPS are lower: 260 g/GJ (0.6 lb/106 Btu)
for coal and 210 g/GJ (0.5 lb/106 Btu) for lignite. This
combination should be able to meet the new standards as well
A-12
-------
FLUIDIZED BED COMBUSTION
Fluidized bed combustion (FBC) is treated separately
because it is a new and unique method of combustion. This
approach involves burning coal in a bed of limestone particles
fluidized by an air stream which also serves as the source of
combustion air. The bed contains about 0.5 wt 7» coal which is
continuously supplied to make up for that consumed by combustion.
A typical unit is shown in Figure 5.
The feature of FBC which affects NOX emissions is the
relatively homegeneous temperature of about 843°C (1550°F) which
exists throughout the bed. Conventional boilers have a tempera-
ture gradient from the flame to the wall with flame temperatures
on the order of.1371°C (2500°F). The lower combustion tempera-
ture of FBC kinetically inhibits formation of thermal NOX and,
thereby, leads to lower NOX emissions. FBC is currently being
tested on a large scale (30 MWe) test unit and'is expected to be
available commercially within the next decade (Ref. A-7). Typi-
cal NOX emissions from the development unit range from 86-120 g
N02/GJ (0.2 - 0.4 lb/106 Btu). This emission rate is well under
the current NSPS of 300 g N02/GJ (0.7 lb/106 Btu).* The economics
of FBC are not yet well established, and therefore, the cost ef-
fectiveness of this approach to NOX control cannot be assessed.
'The proposed new NSPS are lower: 260 GJ (0.6 lb/106 Btu) for
coal and 210 g/GJ (0.5 lb/106 Btu) for lignite. The emission
rate for FBC is well within these emission limits.
A-13
-------
FLUE
FUEL
INJECTION
PIPES
AIR.
DISTRIBUTION
GRID
Figure 5. Fluidized Bed Steam Generator
(Source: Reference A-8)
A-14
-------
FLUE GAS TREATMENT (NOx-ONLY)
Flue gas treatment (FGT) for NOX control differs from
combustion modification in that it involves removal of NOX after
it has been formed rather than by limiting its formation. There
are over 20 developers, of processes that remove NOX from flue
gas. In addition, there are equivalent numbers of processes
available for simultaneous removal of S02 and NOX. These simul-
taneous SOa/NOx processes will be described in a subsequent sec-
tion.
Essentially all of the feasible N0x-only FGT processes
can be divided into two categories.
Selective Catalytic Reduction
Noncatalytic .Selective Reduction
In the following sections, both of these process types are dis-
cussed with respect to their
principles of operation,
status of development,
operating experiences,
advantages/disadvantages, and
costs.
A-15
-------
Selective Catalytic Reduction
Selective catalytic reduction (SCR) processes reduce
the flue gas NOX concentration by reacting NOX with NH3 accord-
ing to the following reactions
4NH3 + 4NO 4- 02 + 4N2 + 6H20 (1)
4NH3 + 2N02 + 02 -»• 3N2 + 6H20 (2)
A generalized SCR process flow diagram is shown in Figure 6. In
this process, flue gas is injected with NH3 in an amount equimolar
with the NOX and fed to a catalytic reactor where the NOX is re-
duced to N2. The treated gas is then passed through the combus-
tion air preheater and on to the stack.
The system shown in Figure 6 will work well with gas-
or oil-fired units using conventional catalysts and fixed bed
reactor designs. With coal firing, potential dust plugging prob-
lems dictate the use of one or more of the following approaches :
operation of the NOX control system down-
stream of an efficient and reliable particu-
late removal device
a moving bed design which permits the
periodic removal of catalyst for cleaning
a catalyst shape that does not collect the
entrained particulates present in the flue
gas as they pass through the catalyst bed.
Diagrams which illustrate the process configurations involved in
each of these cases are shown in Figure 7.
A-16
-------
BOILER
FUEL
COMBUSTION AIR
Q BOILER
'D CATALYTIC REACTOR
(D COMBUSTION AIR PREHEATER
© STACK
Figure 6. Typical configuration - selective .catalytic
reduction process for MOX only.
A-17
02-3632-1
-------
Option 1. Use of a special catalyst shape that is not affected
by flue gas grain loadings
NH-
COAL-
COMBUSTION AIR
Option 2. Use of a moving bed reaction design which permits
the removal of the catalyst for cleaning
COAL
COMBUSTION AIR
Figure 7. N0x-only control alternatives involving selective catalytic
reduction processes - application to coal-fired boilers.
(Option 3 on next page)
02-3633-1
A-18
-------
Option 3. Removal of the flue gas participates upstream of
the NO FGT reactor
COAL
COMBUSTION AIR
Option 3a. Use of a *hot-side* ESP
COAL-
Option 3b. Use of a *cold side* ESP or baghouse
1J Coal fired boiler
vCatalyic1 reactor
30 Combustion%ir preheater>
) Stack.
Catalyst regeneration system
Participate removal device
iHeater
) ;Feed/product ;heat exchanger
Figure 7 (Continued)
A-19
02-3634-T
-------
Of the various configurations which are illustrated in
Figure 7, options 1 and 3a are preferred because generally they
will have the lowest capital and operating costs. In a retrofit
situation, however, it will not always be feasible to install a
reactor upstream of the combustion air preheater. This situation
may therefore, require an arrangement similar to that shown as op-
tion. 3b.
The advantages and disadvantages of SCR processes are
presented in Table 2.
TABLE 2. ADVANTAGES AND DISADVANTAGES OF SELECTIVE
CATALYTIC REDUCTION PROCESSES
Advantages
Disadvantages
Achieves excellent NOX removal
(usually 90% or greater).
Most process-options create no
wastes or by-products (except
spent catalyst).
Demands lower capital investment
and revenue requirements than
wet processes.
Involves gas-phase reactions and,
therefore, requires less complex
operating steps than wet pro-
cesses .
May require reheating of the flue
gas to attain and control reaction
temperature (however, some develop-
mental schemes involve locating the
reactor in such a position that re-
heat may not be required) .
Emits NHs and under certain condi-
tions, NHijHSOit as a particulate.
Some systems are sensitive to par-
ticulates and SOs in the flue gas.
Difficult to retrofit in some appli-
cations.
Source: Reference A-9.
The catalysts used in most SCR processes are oxides of
non-noble metals. These have shown the best combination of high
reactivity and resistance to SOX poisoning. Most SCR processes
are still in the development stage (<50 MW) although some have
been applied commercially. Commercial applications are a recent
A-20
-------
occurrence, however some performance data are available on these
units. NOX removals in the 80-90% range and residual NH3 levels
of less than 10 ppm are typical for Japanese installations. A
major problem area associated with these processes, however, is
the precipitation of NH^HSCU produced by the following reaction
NH3 + S03 + H20 + NH^HSCU (3)
The bisulfate forms a liquid precipitate upon formation and de-
composes to the reactants when heated. The rate of NH^HSO^ for-
mation is a function of reactant concentration and temperature.
Precipitation can be avoided if high temperatures are maintained.
While the cost of a specific SCR process will vary de-
pending on a number of site-specific , factors , published economic
data can be used to develop an expected range of costs. Generally,
the capital investment required will range from $15 to $45/kW.
Operating costs are similar for the various processes at about
1.5 mills/kWh. These costs are primarily for NH3 and catalyst
which do not change significantly with process type or application.
These figures are given in 1976 dollars and are based on a 200-500
MW boiler producing 300 ppm NOX (Ref. A-10) .
Selective Noncatalytic Reduction
Selective noncatalytic reduction, also known as ammonia
injection, utilizes the same reaction of NO and NH3 as does SCR.
The major difference is the temperature at which the reaction
occurs. Whereas with an SCR process the reaction will proceed
at about 400°C (780°F) , without a catalyst, a temperature of about
1000°C (1860°F) is required. In this process, NH3 injection ports
are located in the boiler after the combustion zone. A signifi-
cant control problem with the process is that the reaction is ex-
tremely temperature sensitive. A temperature that is 100 °C (212 °F)
A-21
-------
higher or lower can result in increased NO or NH3 emissions, re-
spectively. The optimum reaction temperature can be lowered by
adding H2 with the NH3. The process can be controlled therefore
by adjusting the H2 injection rate to respond to temperature
changes at the injection point.
Since NH3 is injected into a high temperature region of
the boiler with this process, it is sometimes considered to be a
combustion modification technique. However, since its principle
of operation is so similar to that of an SCR process, it is dis-
cussed here for comparison purposes.
Selective noncatalytic reduction has been reported to
give a 70% reduction in NOX emissions at an NH3 stoichiometry of
1. Since its demonstration, this process has been installed on a
number of Japanese industrial boilers. This process is sensitive
to boiler type and configuration with less than ideal configura-
tions giving poorer removals. Advantages and disadvantages of se-
lective noncatalytic reduction processes are presented in Table 3.
TABLE 3. ADVANTAGES AND DISADVANTAGES OF SELECTIVE,
NONCATALYTIC REDUCTION PROCESSES
Advantages
Disadvantages
Requires no reheat.
Requires no catalyst.
Lower capital investments than
SCR processes.
Generates no liquid or solid waste
or by-product streams.
Uses homogeneous, gas-phase reac-
tions and, thus, requires the
least complicated operating steps
of any process.
Can achieve only about 60-70% NOX
removal.
Needs large amounts of reductant
(NH3) in order to achieve high re-
movals; NH3:NOX mole ratio greater
than 3:1.
Sensitive to temperature variations
with potential for significant emis-
sions of NH3 or NO during boiler
load variations.
Emits NH3 and, under certain condi-
tions, NHi+HSOi* particulate.
Source: Reference A-9.
A-22
-------
The operating cost of a typical selective noncatalytic
reduction process is reported to be in the range of 0.7 to 1.6
mills/kWh (Ref. A-10). The cost range reported is for a retrofit
application and therefore, it is expected that it can be reduced
in new boilers if design provisions are made to accommodate the
process.
Summary
Comparison of the two process types just discussed in-
dicates that both are capable of controlling power plant NOX emis-
sions. Selective noncatalytic reduction is the simplest of the
two, however, it uses large quantities of ammonia. Since ammonia
is currently made from natural gas and is in demand as a fertili-
zer feedstock, its future availability for use in NOX control may
be questionable.
Selection of a particular control technique will depend
OP. the technical and economic factors associated with each speci-
fic application. A major problem with all of these processes is
that they do not provide any SC>2 removal capability. There are
many cases in which SOa removal is or will be required in addition
to NOX removal. In these cases, it may not be technically or
economically attractive to have an SOa removal system installed
in series with an NOX removal system since there are control op-
tions available which allow for the simultaneous removal of both
S02 and NOX. These processes are discussed in the following sec-
tion.
A-23
-------
SIMULTANEOUS NOX/S02 FLUE GAS TREATMENT
A single process that removes both NOX and S02 may be
more economical than using separate processes for NOX and S02
removal. Several FGT techniques for simultaneously removing NOX
and SO2 from a flue gas stream have been developed in the last
few years. These can be divided into two categories: wet and
dry. The wet processes utilize a scrubber to absorb the pollu-
tants, whereas the dry processes utilize a combination of adsorp-
tion and SCR. As before, the technologies will be discussed with
respect to the following items.
Principle of operation
Status of development
Operating experience
Advantages and disadvantages
Costs
Wet Processes for NOX/S02 Removal
There are two predominant process types in this cate-
gory: oxidation-absorption/reduction and direct absorption/re-
duction. Oxidation-absorption/reduction processes use a gas
phase oxidant such as ozone or chlorine dioxide, to convert NO
to N02. N02 and S02 are subsequently scrubbed by a circulating
sulfite solution. Absorbed N02 reacts with the sulfite to form
N2 or N20 and the absorbed S02 reacts with water to form sulfite,
thereby replenishing the sulfite solution. The oxidation step
is necessary since NO is relatively insoluble in and does not re-
act readily with a sulfite solution. A generalized flow diagram
for this process is shown in Figure 8. Prototype scale (10-40
MW) processes of this type have been tested on oil-fired boilers.
A-24
-------
TUEL
I
NJ
Ul
FLUE GAS
FROM BOILER
60* C
ABSORBER
150*C
03 or C102
80° C
*• TO STACK
NaOH or
Na2C03
to
BY-PRODUCT
TREATMENT
Figure 8. Generalized flow diagram for wet NOX/S02 process.
02-3638-1
-------
An alternative to the oxidation-absorption/reduction
system is one in which NO is absorbed and reduced withour prior
oxidation. Processes of this type use an iron chelate complex
to facilitate absorption of NO. The absorbed NO is then reduced
to N2 or N20 by reaction with sulfite. Absorbed S02 produces the
sulfite used in this reaction. Absorption/reduction processes
are less developed than the oxidation-absorption/reduction pro-
cesses primarily due to the fact that they require a more com-
plex by-product treatment scheme. More specifically, it is very
difficult to recover the expensive iron chelate from the spent
scrubbing liquor in these processes. Absorption/reduction pro-
cesses are currently at the pilot scale of development. Plans
for prototype scale units are being discussed. Both types of
wet processes have exhibited 80-90% NOX removal and 95+% S02
removal. The advantages and disadvantages for both process
types are shown in Table 4. An additional potential advantage of
the wet processes is that an existing FGD system could be modified
to accommodate the NOX/S02 technology.
Costs for these processes vary from $65 to 134/kW and
from 4.8 to 8.9 mills/kWh. These costs are based on 1976 dollars.
Dry Processes for NOX/SO? Control
Like the dry N0x-only processes, the most promising dry
process for simultaneous NOX/S02 removal is based upon the reduc-
tion of NOX with NH3 by SCR. The primary difference is that the
copper-based catalyst used functions as an S02 adsorbent as well
as a NOX reduction catalyst.
A generalized flow diagram for this type of process is
shown in Figure 9. For economic reasons the configuration which
involves locating the reactor upstream of the combustion air pre-
heater is favored (flow scheme 9a). However, in a retrofit situa-
A-26
-------
AIR.
OFFGAS TO
SULFUR
RECOVERY
HOT
REDUCING
GAS
9a. *Hot-side* configuration
9b. *Cold side'" configuration
OFFGAS TO
*• SULFUR
RECOVERY
HOT
REDUCING
GAS
^) BOILER
J) COMBUSTION AIR PREHEATER
'T) PARTICULATE REMOVEAL
^ SYSTEM(e.g. ESP or baghouse)
T) REACTOR CONTAINING ADSORBENT/SCR
CATALYST IN ACCEPTANCE MOLD
,5) REACTOR WITH SPENT CATALYST
IN REGENERATION MODE
© STACK
® FEED/PRODUCT EXCHANGER
3D REACTOR FEED HEATER
Figure 9. Processing scheme used with dry simultaneous NOx/SO?
removal system
A-27
-------
TABLE 4. ADVANTAGES AND DISADVANTAGES OF WET
PROCESSES RELATIVE TO DRY PROCESS ALTERNATIVES
Advantages Disadvantages
Oxidation-Absorption/Reduction Systems
Capable of removing significant Requires an expensive gas-phase
amounts of both NOX (85-90%) and oxidant.
SOa (95+%) simultaneously.
Requires a large absorber.
Treatment/disposal of blowdown
waste streams.
Requires flue gas reheat.
Involves a series of complicated
processing steps.
Requires specific ranges of flue
gas constituents.
Absorption/Reduction Systems
Capable of removing significant Requires the use of an expensive
amounts of both NOX (85-90%) and chelating compound.
S02 (95+%) simultaneously.
Requires a large absorber.
Requires flue gas reheat.
Complex reaction chemistry and sor-
bent regeneration scheme.
Involves a series of complicated
processing steps.
Requires specific ranges of flue
gas constituents.
Source: Reference A-9.
A-28
-------
tion, this arrangement may not be possible. For this reason, both
possible configurations are shown in Figure 9.
The most noteworthy feature of this type of process re-
actor is the regenerator configuration. In the reactor, S02 reacts
with CuO and oxygen to form CuSCU . CuSOi, then promotes the NO re-
duction reaction. Several reactors are operated in "swing" opera-
tion, that is, when one reactor is saturated it is taken off line
for regeneration and a freshly regenerated reactor is brought on
line. The loaded acceptor/catalyst is regenerated by reacting the
CuSOi* with a hot reducting gas to produce CuO and S02. The regen-
erator offgas composition will vary depending on the reducing gas
used, but, in general, this stream would be sent to a RESOX, modi-
fied Glaus, liquefaction, or oxidation process for sulfur recovery.
Potential products here are H2S04, S, or S02.
The NOX/SOX version of this process is currently at the
pilot-scale level of development. Tests on coal-fired flue gas
should begin in the near future as part of an EPA demonstration
program. Generally, the advantages and disadvantages are the
same as those of the N0x-only adsorption processes except that
the NOX/S02 processes have the additional advantage of removing
S02 as well as NOX. Advantages and disadvantages are listed in
Table 5.
Costs for this type of process are estimated at 130/kW
and 6.0 mills/kWh. Economics vary depending on the characteris-
tics of the flue gas being treated.
Summary
Where simultaneous NOX/S02 removal is required, selec-
tion of a particular process will depend almost entirely on the
A-29
-------
TABLE 5. ADVANTAGES AND DISADVANTAGES OF THE
DRY NOX/SOX PROCESS
Advantages
Disadvantages
Capable of removing significant
amounts of both NOX (90%) and
SOX (90%) simultaneously.
SO x is converted into a marketable
by-product.
There are no significant waste
streams (<10 ppm NH 3 in treated
gas) .
Involves gas-phase reactions and,
therefore, requires less complex
operating steps than wet pro-
cesses.
Is not sensitive to particulates.
Requires use of a reducing gas.
Is operationally more complex than
throwaway systems.
A-30
-------
specific application. The costs for both wet and dry processes
cover the same range and therefore, as a class, the two types
of systems are economically competitive. Process selection, then,
will depend on how well the particular processes meet the needs
of the application.
A-31
-------
REFERENCES
A-l Blythe , Gary M. and William E. Corbett. A State-of-the-
Arc Survey of Control Techniques for the Abatement of
NOz Emissions from Stationary Combustion Sources.
Radian Project No. 200-045-47, TN No. 200-045-47-02.
Austin, Texas. Radian Corporation. July 1976.
A-2 Bowen, Joshua S. and Robert E. Hall, Chairmen. Pro-
ceedings of the Second Stationary Source Combustion
Symposium. 5vols. EPA 600/7-77-073a-d. Washington,
D.C. EPA. Office of Research and Development.
July, 1977.
A-3 Hollinden, Gerald A. and R. L. Zielke. Evaluation of
the Effects of Combustion Modifications in Controlling
NOX Emissions at TVA's Widows Creek Steam Plant.
Presented at the American Power Conference 38th Annual
Meeting. Chicago. April 1976.
A-4 Barr, W.H., F.W. Strehlitz, and S.M. Dalton. Modifying
Large Boilers to Reduce Nitric Acid Emissions. CEP 73
(7) , 59-68 (1977).
A-5 Driscoll, J.N. Flue Gas Monitoring Techniques and
Manual Determination of Gaseous Pollutants. Ann Arbor
Science Publishers. Ann Arbor, Michigan.. 1974.
pp. 219-263.
A-6 Graham, John. Combustion Optimization. Electr. World
183 (6), 43-58 (1976).
A-7 Foster Wheeler Energy Corporation. Conceptual Design
for an Atmospheric Fluidized Bed Steam Generator.
ERDA Contract No. EF-77-C-01-2583. November 1977.
A-8 Proceedings of the NOX Control Technology Seminar,
San Francisco, February 1976. Palo Alto, California.
EPRI. 1976.
A-9 Faucett, H.L., J.D. Maxwell, andT.A. Burnett.
Technical Assessment of NO Removal Processes for
Utility Application. Final Report. EPA-IAG No.
D7-E721-FU, EPA 600/7-77-127, EPRI RP 783-1, EPRI
F-568, TVA Bull. Y-120. Muscle Shoals, Alabama.
Tennessee Valley Authority. Office of Agricultural
and Chemical Development. November 1977.
A-10 A Way to Lower NOX in Utility Boilers. Env. Sci.
Tech. 11 (3), 226 (1977).
A-32
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing/
1. REPORT NO.
EPA-600/7-78-215
3. RECIPIENT'S ACCESSION- NO.
4. TITLE AND SUBTITLE
Assessment of the Need for NOx Flue Gas Treatment
Technology
5. REPORT DATE
November 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W.E.Corbett, G.D.Jones, W. C. Micheletti,
R. M.Wells, and G.E.Wilkins
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
10. PROGRAM ELEMENT NO.
1NE624
11. CONTRACT/GRANT NO.
68-02-2608, Task 13
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 PERIOD COVERED
Task Final; 3/77 - 10/78
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES IERL_RTP project officer is J. David Mobley, MD-61, 919/541-
2915.
. ABSTRACT
repOrt gives results of a. study to determine if and when the application
of NOx flue gas treatment (FGT) technology will be necessary in the U.S. It addres-
ses factors that will influence the levels of NOx emission control needed to comply
with both existing and future NOx standards. Topics treated include NOx emission
sources, atmospheric transport and reactions, air quality trends, regulations, and
control strategies, and flue gas treatment methods. The study concludes that the
number of Air Quality Control Regions (AQCRs) with NOx compliance problems can
be expected to increase significantly in the next decade. It further concludes that
progressively larger reductions in NOx emissions will be required in order to attain
and maintain compliance in 'problem' AQCRs. The study does not establish conclu-
sively whether or not FGT will be required. However, current trends indicate that
FGT may be necessary in the future to achieve compliance with NOx standards in
certain AQCRs. This conclusion follows from the regionally specific nature of U.S.
NOx compliance problems , as well as uncertainties concerning both future NOx
emission reduction requirements and the ultimate effectiveness of alternative NOx
control methods , such as combustion modification.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Nitrogen Oxides
Flue Gases
Assessments
Air Pollution Control
Stationary Sources
Flue Gas Treatment
13 B
07B
2 IB
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
108
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
A-33
------- |