OVERVIEW OF ENVIRONMENTAL PROTECTION AGENCY'S
NOX CONTROL TECHNOLOGY FOR STATIONARY COMBUSTION SOURCES
by
David G. Lachapelle
Joshua S. Bowen
Richard D. Stern
U. S. Environmental Protection Agency
Office of Research and Development
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, N. C. 27711
Presented at
67th Annual Meeting, AIChE
December 4, 1974
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGMENT iv
ABSTRACT v
LIST OF FIGURES vi
LIST OF TABLES vi i
INTRODUCTION 1
AUTHORITY FOR NOX STANDARDS 2
NATIONAL AMBIENT AIR QUALITY STANDARDS 3
NEW SOURCE PERFORMANCE STANDARDS 5
Steam Generators 5
Nitric Acid Plants 6
Gas Turbines 6
Other Sources 6
STATE AND LOCAL STANDARDS 8
STRATEGY FOR MEETING NAAQS 11
CSL NOX CONTROL TECHNOLOGY PROGRAM 12
Combustion Modification Technology 13
Field testing and survey studies 14
Fundamental combustion research studies 15
Fuel s research and devel opment 16
Process research and development 16
Maximum stationary source technology program planning-.. 17
Eff 1 uent Treatment Techno! ogy 19
SOURCES AND LEVELS OF NOX EMISSIONS 21
MECHANISMS OF NO FORMATION 25
-------
TABLE OR CONTENTS (Continued)
Page
NOX COMBUSTION CONTROL TECHNOLOGY FOR BOILERS 27
Low Excess Air Combustion 27
Flue Gas Recirculation .. . 28
Water or Steam Injection. 29
Staged Combustion 29
Reduced Air Preheat Temperature. 30
Load Reduction 31
Burner Design 31
NOX COMBUSTION CONTROL TECHNIQUES FOR STATIONARY I.C. ENGINES.... 33
NOX COMBUSTION CONTROL TECHNIQUES FOR STATIONARY GAS TURBINES.... 34
NOX COMBUSTION CONTROL TECHNIQUES FOR COMMERCIAL, RESIDENTIAL AND
PROCESS HEATING 7 35
COMBUSTION MODIFICATION EXPERIENCE 36
Utility Boilers ....... 36
Industrial Boilers. ;.'. 38
Coal fired boilers 39
Oil fired boi 1 ers 40
Gas fired boilers... 41
Stationary I.C. Engines 42
Gas Turbines 44
Commercial and Residential Heating 48
Commercial boilers. 51
Residential systems 51
Process Heating 53
EFFLUENT TREATMENT NOX CONTROL TECHNOLOGY....' 53
Current Research and Devel opment 54
NOX COMBUSTION CONTROL COSTS 56
Uti 1 i ty Boi 1 ers 57
Industrial Boilers 66
Diesel Engines 66
Spark Ignition Engines 67
Gas Turbines 67
n
-------
TABLE OF CONTENTS (Continued)
Page
SUMMARY 73
REFERENCES 75
CONVERSION TABLE 78
m
-------
ACKNOWLEDGEMENT
The authors gratefully acknowledge the helpful suggestions and
assistance of Messrs. R. E. Hall and W. S. Lanier of the U. S. Environ-
mental Protection Agency and Dr. H. B. Mason of the Aerotherm Division
of Acurex Corporation.
-------
... , ABSTRACT
An overview of'the Environmental Protection Agency's nitrogen
oxides control technology program is presented. Topics discussed
include the role of stationary sources and their NOX contribution,'
the state-of-the-art of NOX control including combustion modifica-
tion and flue gas cleaning technologies and implementation costs,
future goals and possible new approaches. Environmental Protection
Agency's standards and local regulations are reviewed.
-------
LIST OF FIGURES
Figure Title Page.
, * t.. T_ . , . r*_ _ i
1 Total NOX Emitted in the U. S. from Stationary Sources..., 22
2 Typical Gas Turbine Base Load NOX Emissions 44
3 Effectiveness of Water or Steam Injection in Reducing
NOX Formati on 47
4 Costs of NOX Control Methods for New Coal Fired Units 59
5 Costs of NOX Control Methods for Existing Coal Fired
Units 60
-------
LIST OF TABLES
Table Title Page
1 New Source Performance Standards (NSPS) For Stationary
Combustion Systems ,
2 Rule 68: Los Angeles County NOX Standards for Existing
Boilers Greater than 1,775 x 106 Btu/hr. Heat Input 9
3 Standards Proposed by EPA for Existing Boilers in Los
Angel es County 10
4 Summary of Total NOX Emissions from Fuel-User Sources
(1972) 23
5 Summary of Fuel Use by Combustion Source (1972) 24
6 Pollutant Control Techniques for Stationary I. C. Engines 33
7 Pollutant Control Techniques for Stationary Gas Turbines 35
8 NOX Levels and Typical NOx Reductions with Combustion
Modification to Utility Boilers 37
9 Effectiveness of Combustion Modification on Gas Turbines 49
10 Typical Baseline Emission Levels from Commercial and
Residential Heating 50
11 Effect on Mean Emissions of Identifying and "Replacing"
Residential Units in "Poor" Condition and Tuning 52
12 Estimated Investment Costs for Low Excess Air Firing on
Existing Utility Boilers Needing Modifications 61
13 Impact of NOX Control Techniques on Major Utility Boiler
Components 63
14 1973 Operating Costs of NOX Control Methods for New Coal
Fired Units 65
15 Water Injection Investment Cost (San Diego Gas and Electric). 68
vn
-------
Table Title Page
16 Water/Steam Injection Cost as a Function of Power Plant
Size 68
17 Cost of NOX Controls for Small Gas Turbines 70
18 Cost of Wet NOX Controls for Large Gas Turbines 71
19 Projected Short and Long Term Goals for NOX Control 74
VI11
-------
INTRODUCTION
Stationary combustion sources continue to contribute about half
of the total man-made nitrogen oxides (NOX) in the United States.
Recent estimates indicate that over 11 million tons* of NOX were emitted
from fossil fuel combustion in stationary sources in 1972. Because
of their quantity and potential for widespread adverse health effects,
NOX is among the atmospheric pollutants for which regulatory controls
have been established by the U. S. Environmental Protection Agency
(EPA). Authority for the promulgation and enforcement of source per-
formance standards and National Ambient Air Quality Standards (NAAQS)
came with passage of the 1970 Clean Air Act. The Act additionally
provides for an EPA research program to develop cost-effective, com-
mercially viable NOX control technology for both new and existing
stationary combustion sources . Toward that goal, the Combustion
Research Section of the Control Systems Laboratory of EPA is actively
engaged in a program of contract and in-house studies to develop NOX
combustion control technology for a variety of sources. EPA's current
activity in flue gas treatment for NOX is proceeding at a low level
and is primarily directed toward catalytic methods for NOX reduction.
*It is EPA policy to use Metric units; however, in this paper English
units are used for convenience. See attached conversion table.
-------
AUTHORITY FOR N0y STANDARDS . .
The Clean Air Act of 1970 required that the Administrator promul-
gate national ambient air quality standards for those pollutants "which
in his judgment have an adverse effect on public health and welfare."
In addition, the Clean Air Act directed the Administrator to promulgate
standards of performance .with respect to the emission of air pollutants
for those new stationary sources which he determines "may contribute
significantly to air pollution which causes or contributes to the eh-
dangerment of public health or welfare."
Prior to publication of a nitrogen oxides criteria document and
in response to the directive of the. Clean Air Act of 1970 that a reduc-
tion of nitrogen oxides emissions of at least 90 percent be achieved
for light duty motor vehicles by 1976, a nitrogen oxide Federal motor
vehicle emission standard of 0.4 g/mile was adopted. Subsequently,
the publication of the nitrogen oxides air quality criteria document
set in motion the mechanism for development of the National Ambient
Air Quality Standards (NAAQS) which were promulgated in 1971.^'
Nitrogen oxides were identified as an air pollutant having an
adverse effect on public health and welfare. Within this category of
chemical compounds, there are only two nitrogen oxides which are of
significance in the atmosphere. These are nitric oxide (NO) and ni-
trogen dioxide (N02); the combination of these two major species is
referred to under the general classification of nitrogen oxides, rep-
resented by the symbol NOX. NO is the predominant form emitted by
-------
most combustion processes, accounting for approximately 95% of the
total NOX from the combustion of fossil fuels. Some chemical processes
(such as nitric acid manufacture) and certain combustors (such as gas
turbines), however, may emit significant amounts of N02 directly.
Nitric oxide, once it is in the atmosphere, is converted to N02-
One mechanism of conversion is thermal oxidation which proceeds at a
slow rate. A second mechanism is the photochemical conversion involv-
ing N02, reactive hydrocarbons and ultraviolet energy from sunlight.
This latter reaction is rapid and produces undesirable oxidants, as
byproducts, which result in smog formation under certain conditions.
NATIONAL AMBIENT AIR QUALITY STANDARDS
In establishing a basis for ambient air quality standards, analy-
ses of available information on atmospheric chemistry and on the effects
of nitrogen oxides on man, plants and materials disclosed two main
points. The first is that NO is rapidly converted to N02 in the atmos-
phere, and secondly, that N02 is by far the more hazardous oxide form.
The adverse effects of N02 at ambient concentrations on human
beings were highlighted as the result of an epidemiological appraisal
(?)
of N0£ conducted by Shy in Chattanooga, Tennessee.v/ The results of
this study indicated adverse health effects contributable after expo-
sure to ambient levels of N02 at 0.1 ppm for 1 hour. This preliminary
indication provided the basis for the statutory mobile source standard
of 0.4 g/mile. Further analysis of the data led to the conclusion that
-------
an annual average N02 concentration greater than 113 yg/m3 (0.06 ppm)
has the potential for adverse effects on respiratory performance.
This became the basis for the NAAQS.
The NAAQS for N02 were set in 1971 as required by the Clean Air
Act of 1970. The primary standard is based on a level which is required
to protect the public health, and the secondary standard is based on a
level which is requisite to protect the public welfare from any known
or anticipated adverse effect associated with the presence of air pollu-
tants in the ambient air. Both standards were set at 100 micrograms
per cubic meter for N02-
Questions related to the reliability and accuracy of the analyti-
cal method used in the Chattanooga study have cast doubt on the 100
ug/m3 NAAQS, and the issue has not yet been resolved., Indications
are that the annual average NAAQS may be retained but that this may
be supplemented by a short term exposure standard as well. As a result
of revaluation of the levels of N02 based on more reliable analytical
methods, it appears that only four Air Quality Control regions (AQCRs)
(3)
now exceed the NAAQS. ' In addition another four AQCRs are near the
standard and remain a potential future problem. These problem areas
are all in regions of high population density so the NOX problem at
present is essentially an urban problem.
-------
NEW SOURCE PERFORMANCE STANDARDS
The Clean Air Act requires that emission control standards "reflect
the degree of emission reduction which (taking into account the cost of
achieving such reduction) the Administrator determines has been ade-
quately demonstrated." The approach to setting Federal standards for
stationary sources is to determine a priority ranking of sources to
be controlled based on: 1) magnitude of contribution to the total
emission of a given pollutant, 2) available information on control tech-
nology, 3) projected growth, and 4) impact of Federal standards rela-
tive to existing State standards. The sources have been grouped into
five priority categories to establish the order in which standards
would be set. The two sources of NOX which fell into the Priority I
category were steam generators of greater than 250 x 106 Btu/hr. input
and nitric acid plants. Other NOX sources to be regulated included
stationary gas turbines, stationary internal combustion engines, and
steam generators between 10 x 106 and 250 x 106 Btu/hr. input.
Steam Generators
The New Source Performance Standards (NSPS) for fossil fuel fired
steam generators over 250 x 106 Btu/hr. input cover any source on which
construction or modification was commenced after August 17, 1971. The
boilers which fall into this category are mostly utility boilers for
electric power generation although some large field erected industrial
steam-raising boilers are included. Along with the NOX standards, levels
for particulate matter and sulfur oxides were also included. The
-------
particulate standard is 0.1 Ib. per million Btu input (maximum 2 hr.
average) for filterable particulate. The sulfur oxides (SOX) standard
is 0.8 and 1.2 Ibs. S02 per million Btu for liquid and solid fuels,
respectively. The NOX standards are a function of fuel type as shown
in Table 1.
Nitric Acid Plants
The NSPS for NOX for nitric acid plants was set at 3 Ibs. (expressed
as N02) per ton of acid produced. A 10% opacity standard was also in-
cluded.
Gas Turbines
Publication of the New Source Performance Standards for stationary
gas turbines is planned for December, 1974. After comments are received
and the standards are finalized, promulgation is anticipated during the
first half of 1975. Although the original goal was 0.2 Ib. NOX (as N02)
per million Btu input for all fuels, it now appears that different stan-
dards may be established for gas and oil fired units. Current target
levels for stationary gas turbines of over 1.5 million Btu/hr. (about
125 hp) are 75 ppm NOX (15% excess oxygen) for oil fired units and 55
ppm NOX (15% excess oxygen) for gas fired units.
Other Sources
At least two other sources are under consideration for promulga-
tion of NSPS. These possible sources to be controlled include station-
ary internal combustion engines and steam generators (coal, oil and gas
-------
fired) between 10 x TO6 and 250 x 106 Btu/hr. input. Lignite fired
steam generators of over 250 x 106 Btu/hr. may also be added to the
list.
Table 1
NEW SOURCE PERFORMANCE STANDARDS (NSPS)
FOR STATIONARY COMBUSTION SYSTEMS
(Fossil Fuel Fired Steam Generating Units with Input Greater Than
250 x 106 Btu/hr.)
Fuel Type
Gas
Oil
Coal
Mixed*
Fuels
Standard
Ib. N02/106 Btu(2 hr. avg.)
0.2
0.3
0.7
X(0.2) + Y(0.3) + Z(0.7)
X + Y + Z
ppm (3% excess
02 dry basis)**
168
230
500
* X = % of total heat input derived from gas
Y = % of total heat input derived from oil
Z = % of total heat input derived from coal
** Based on typical average fuel heating values
-------
STATE AND LOCAL STANDARDS
The Clean Air Act provides for State or political subdivisions
to impose and enforce standards for control or abatement of air pollu-
tants except that these standards may not be less restrictive than the
Federal requirements if necessary to meet the NAAQS. The Los Angeles
Air Pollution Control District (LAAPCD) established Rule 68 which set
standards for existing gas and oil fired utility boilers of greater
than 1,775 x 106 Btu/hr. input and has been in effect since December
31, 1971. A more stringent Rule 68 level of control becomes effective
December 31, 1974. The Rule 68 levels are shown in Table 2. EPA has
proposed to supplement Rule 68 to cover the boilers which range between
250 x 106 and 1,775 x 106 Btu/hr. as shown in Table 3. Essentially
the same EPA regulations have been proposed for several other California
and New York Air Pollution Control Districts.
Perhaps the most restrictive regulation in existence anywhere is
the LAAPCD's Rule 67, which places a limit of 140 Ibs. N02 per hour
regardless of unit size. As an example under this rule, a 315 MW utili-
ty boiler can only emit 42 ppm NOx (corrected to 3% oxygen). The
Scattergood 3 unit of the Los Angeles Department of Water and Power
is currently undergoing initial testing and is projected to meet this
emission level.
-------
Table 2
RULE 68: LOS ANGELES COUNTY NOX STANDARDS FOR EXISTING BOILERS GREATER
THAN 1775 x 106 BTU/HR HEAT INPUT
Standard
Fuel Type In Effect N02, ppm*
Gas Before 12/31/74 225
After 12/31/74 125
Oil, Coal Before 12/31/74 325
After 12/31/74 225
Concentration is calculated at 3% excess 03, dry basis
-------
Table 3 J
STANDARDS PROPOSED BY EPA FOR EXISTING
BOILERS IN LOS ANGELES COUNTY*
Standard
Fuel Type
Gas
.;.
Oil
Boiler
(type, firing)
Tangential
Front Wall
Horizontally opposed
Tangential
Front Wall
Horizontally opposed
Ib. N02/106 Btu
0.2
0.4
0.6
0.3
0.4
0.6
N02, ppm**
168
336
504
230
307
460
*For boilers greater than 250 x 106 Btu/hr. but less than 1775 x 106
Btu/hr.
**Concentration is calculated at 3% excess 02, dry basis
10
-------
STRATEGY FOR MEETING NAAQS
Provision was made to apply available NOX control technology to
stationary sources in order to achieve the NAAQS if the reduction in
automotive NOX emission levels of 0.4 g/mile was insufficient to meet
the standards. It has been very difficult to meet this automotive stan-
dard and the control technology to achieve this level is costly and
has imposed a significant fuel penalty. Thus EPA's Office of Air Quality
Planning and Standards (OAQPS) has conducted a study analyzing the
most effective means for achieving the NAAQS considering control and
fuel penalty costs as well as air quality. As a result, OAQPS has recom-
mended that the automotive standard be revised to 2.0 g NOX per mile
and that more stringent NOX emission controls be placed on new and exist-
ing stationary sources as required in air quality control regions not
meeting NAAQS. This recommended strategy is currently undergoing review
prior to any formalized action in changing the earlier mandate.
The attainment of the NAAQS for nitrogen oxides will require achiev-
ing a balance of emissions control for both mobile and stationary sources,
If the recommended OAQPS strategy is adopted, it would require that
in certain AQCR's more stringent controls would have to be placed on
new and existing stationary sources to achieve the NAAQS. Considering
national annual growth rate patterns and future expansion of transporta-
tion and industry, it would be necessary to implement even more strin-
gent NSPS for major stationary NOX sources in the 1980-1985 period.
This will require increased R&D in both combustion control and effluent
treatment technology.
11
-------
It is assumed that present standards can be met with current sta-
tionary source technology (CSST). However, more effective NOX control
technology than that currently available will be required within the
next five years to ensure meeting the NAAQS is all AQCR's and the more
stringent NSPS which are anticipated. This need for a maximum station-
ary source technology (MSST) program for attaining adequately low levels
of NOX in the future provides the major basis for the EPA's Control
Systems Laboratory (CSL) NOX control technology program.
CSL NOX CONTROL TECHNOLOGY DEVELOPMENT PROGRAM
The current program for stationary NOX control technology has been
developed over the past seven years, and is based on growing information
relative to the definition of the problem, the most effective and eco-
nomical strategies to be followed in meeting the standards, and our
increased knowledge of the causes of the problem and the effectiveness
of various approaches to NOX emissions reduction. A systems study of
the NOX problem area.five years ago provided the preliminary background
and initial direction .to the program. This early investigation indi-
cated that 1) the major stationary sources of NOX emissions were fossil
fueled combustion systems and 2) the most effective means of control
for near-term solutions of the problem would be the application of com-
bustion modification techniques to minimize the formation of the nitro-
gen oxides. Although the recommended program was oriented strongly
to the development and evaluation of combustion modification techniques,
consideration was also given to flue gas cleaning methods. These latter
12
-------
methods were believed to hold promise for application to specialized
cases where high NOX concentrations are present or very low levels
of NOX emissions are required.
The CSL program is divided into two separate parts. The major
emphasis is directed to the development of combustion modification
techniques for controlling NOX emissions while a separate and much
smaller effort is investigating effluent treatment methods for re-
moval of NOX.
Combustion Modification Technology
The NOX combustion control program is a systems program con-
cerned with the control of combustion generated pollutants from all
types of stationary combustion sources. Although the primary pollu-
tant emphasis was defined as NOX control, it is known that NOX combus-
tion modification techniques, unless properly applied, could lead to
serious problems of emissions of products of incomplete combustion.
Some of the potential control techniques also appear to have the possi-
bility of decreasing process efficiency. Secondary emphasis therefore
is placed on the control of the products of incomplete combustion,
such as carbon monoxide (CO), unburned hydrocarbons (UHC), carbon
particulate, and smoke, and on process efficiency. A program guide-
line is that the NOX control technology should not produce an increase
in products of incomplete combustion, and should not adversely affect
efficiency but, if possible, should actually increase it.
13
-------
The program includes a number of separate projects which fall
generally into four separate areas:
1) field testing and survey activities,
2) fundamental combustion research studies,
3) fuels research and development, and
4) process research and development.
A general discussion of these various areas of activity is in-
cluded below to aid in understanding the program.
Field Testing and Survey Studies
The field testing and survey studies are designed to determine
what can be done currently to control NOx emissions. This work is
generally performed by R and D organizations on commercial equipment
and closely involves the cooperation of manufacturers, users and trade
associations. These groups provide technical and financial assistance,
make available test sites, and modify equipment or operating practices
and schedules to accommodate test programs. In addition to develop-
ing trends and providing directional recommendations for industry to
minimize emissions with today's technology, the work also defines
problem areas and delineates where the R&D efforts should be concen-
trated. These studies are the initial efforts in the development
of control technology and provide very near-term control technology.
They develop the information on available state-of-the-art NOX con-
trol technology to be passed along to other potential users who may
not yet have adopted these control measures.
,14
-------
Fundamental Combustion Research Studies
The fundamental combustion research studies are designed to pro-
vide basic combustion theory and a detailed understanding of the forma-
tion of pollutants and their reactions in the combustion process. The
studies consider the variations related to differences in fuels, combus-
tion systems, and operating conditions. The goal of these studies is
to provide a solid scientific basis for the rational development of
effective control technology so that reliance on empirical approaches
is minimized. As the role of chemical and physical factors and the
influence of combustion process variables on pollutant formation become
known, it should be possible to more readily postulate viable theories
for NOX control technology.
The fundamental combustion studies underway and planned for the
future generally fall into three categories, namely, the chemistry
of pollutant formation, the aerodynamics or physical factors affecting
pollutant formation, and mathematical simulations or modelling of
pollutant formation in combustion processes. Work in the first two
categories investigates the basic components involved in the forma-
tion and destruction of pollutants in combustion processes and flames.
These results provide the basis and input for the third category of
mathematical simulations which will be used ultimately in all program
areas.
15
-------
Fuels Research and Development
The fuels research and development studies are designed to de-
velop generalized combustion control technology which is applicable
to the optimum control of NOX emissions from the combustion of both
conventional fossil fuels and future fuels. These engineering R&D
studies examine the feasibility of the concepts which may be generated
as a result of the fundamental combustion research studies. They are
conducted on versatile experimental combustion systems and will define
the generalized combustion control concepts and technology for the
combustion of various fuels. This work will provide the optimum goal
for practical NOx control. When technically and economically feasi-
ble control techniques have been identified and generalized combus-
tion control technology has been developed as a result of these studies,
the technologies will be adjudged ready for application scale-up and
more thorough investigation in practical systems.
Process Research and Development
The process research and development portion of the combustion
modification program is planned to assess the application of optimum
NOX control technology to both traditional and advanced combustion
systems. The work generally involves studies of control techniques
as applied to commercial or prototype combustion systems and requires
close cooperation with industry to achieve the desired results. The
studies are planned to develop design information, to assess technical
16
-------
performance, to generate cost information, and to identify problems
relative to the application and use of the technology. Specific re-
sults of the work in this area include design and operational guidance
manuals that can be used by manufacturers and users to control NOX
emissions by combustion modifying techniques. The results of the
studies in this category provide the basis for the future demonstra-
tion of combustion control technology for major emission sources.
Maximum Stationary Source Technology (MSST) Program Planning
In view of anticipated changes in NOX control strategy based
on recommendations of the OAQPS to relax the automotive NOX emission
standards and to require more stringent NOX controls for stationary
sources, a project has been initiated to define the MSST program.
The Aerotherm/Acurex Corporation was given the task of assisting CSL
in developing a coordinated MSST program for NOX control technology
development. Three major objectives set forth for this task were:
1) to ensure that the CSL NOX program directly responds to the needs
of the OAQPS; 2) to delineate those sources which are being covered
by current stationary source technology (CSST) and to define the
reasonably expected controlled emissions obtainable through CSST,
and 3) to define additional efforts (tasks), potential further NOX
reductions, and sources covered by the MSST. Among several auxiliary
segments of this task, one involved estimating costs for a total
budget required for achieving the MSST program.
17
-------
The status of this effort is well advanced but results were
not complete (draft final report) at the time this paper was prepared.
Although most of the effort to date has emphasized the combustion
of fossil fuels in conventional combustion systems identified as the
major emission sources (e.g., utility boilers, industrial boilers,
commercial and residential heating systems), future effort is expected
to expand into new areas. Work will be increased in the area of
utilization of mixed fossil and waste fuels and of new alternate
fuels as they are developed. Consideration will be given to a wider
range of conventional combustion equipment, including stationary gas
turbines, stationary internal combustion engines, and industrial
process furnaces. In addition, newer concepts of combustion will
be evaluated for their potential for low NOX emissions. This work
will include investigations of catalytic combustion systems and of
advanced power cycles. In general, the program should provide the
combustion control technology to minimize NOX and other combustion-
related emissions from a broad range of existing and new point and
area combustion sources while simultaneously maintaining acceptable
or improved combustion process efficiency. The effort, while stress-
ing the development of control technology, will, also provide assess-
ments of the effect on the environment of practical applications of
these controls.
18
-------
Effluent Treatment Technology
The Esso Systems Study^5' on NOX control identified a number of
candidate effluent treatment techniques (e.g., sulfuric acid scrubbing,
alkaline scrubbing, molecular sieve absorption and catalytic reduction).
Early preliminary studies of sulfuric acid scrubbing and alkaline
scrubbing, considering such variables as NO concentration, NO to N0£
ratio, and scrubbing agent, indicated that these scrubbing processes
alone were not feasible methods. ' With sulfuric acid scrubbing,
mass transfer rates were quite low leading to very large contact equipment
and high costs. N0£ absorption with alkaline scrubbing appeared feasible
but mass transfer rates with sorbents for flue gas desulfurization were
relatively slow, indicating similar problems. In addition, preliminary
studies regarding oxidation of NO to N02 indicated that ozone production
capital and operating costs would be quite expensive and require high
power consumption. Being explored at present are catalytic NOX control
with the use of reductants and molecular sieve absorption. Demonstration
of the technique of molecular sieve absorption is underway at 2 nitric
acid plants. This technology is not applicable to combustion flue gases
due to preferential sorption of the moisture in the flue gas.
Future efforts in the development of NOX effluent treatment are
anticipated to continue in three major areas relating to NOX reduction:
Evaluation of the selective reduction of NO with platinum
catalyst and ammonia as a reductant will be continued on a utility
boiler firing low-sulfur distillate oil. The effects of operating
time on catalyst performance will be evaluated and methods of
extending the life of the catalyst will be investigated.
19
-------
The technology of non-noble metal catalysts for NOV reduction
A
will be investigated further for applicability to combustion units
firing SOX producing fuels. This activity will include additional
laboratory scale evaluations to provide design criteria and cost
information. These evaluations will include selective NOX reduction
with NH3 and nonselective reduction with ^ and CO. Based on
results, process evaluations may be conducted at larger scales.
Feasibility studies and experimental evaluations will be
conducted on promising processes which simultaneously remove
SOX and NOX from combustion flue gases. Processes with possible
application include Esso/B&W, Shell/UOP and Chiyoda Thoroughbred 102.
In addition, applications of effluent treatment technology will
continue to be surveyed with particular attention paid to Japan
where efforts have been intense.
20
-------
SOURCES AND LEVELS OF NOX EMISSIONS
(A\
A recently completed EPA-sponsored study v ' has updated and
refined prior 1968 estimates by Essov ' of NOX emissions from station-
ary sources. Figure 1 shows the major stationary sources and their
NOX contribution. Utility boilers are the major NOX source, account-
ing for over 48% of the nearly 11.7 million tons emitted. Station-
ary reciprocating internal combustion (I.C.) engines and gas turbines
collectively accounted for over 21% of the total. Industrial boilers
contributed over 18% and residential and commercial heating accounted
for slightly over 7%. Another NOX source was process heating at
over 3%. Collectively, these five sources account for over 98% of
the man-made NOX from stationary sources, the remainder coming from
non-combustion sources and incineration. The significance of the NOX
emission update is that since the 1968 studies, stationary engines
(I.C. and gas turbines) have moved from third to second place as a
result of refined emission factors and better estimates of installed
capacity. Additionally, the total stationary source NOX has increased
by over two million tons with the majority of the increase coming
from utility boilers.
Table 4 is a summary of total NOX emissions from stationary
fuel -user sources. It shows that the major NOx contributors are coal
fired utility boilers. The split between total NOX from gas and coal
combustion is about equal, even though the total usage of gas is sig-
nificantly higher than coal. This is due primarily to the influence
21
-------
Incineration 0.4%
oncombustion 1.3%
Gas Turbine 2.5%
Industrial Process
Heating 3.3%
Commercial/
Residential
Space
Heating 7.1!
Industrial
Boilers
18.1%
Utility Boilers
48.6%
Reciprocating
I.C. Engines
18.
Source
Utility Boilers
Reciprocating I.C. Engines
Industrial Boilers
Commercial/Residential Heating
Industrial Process Heating
Gas Turbines
Noncombustion
Incineration
TOTAL
Estimated NOX Emissions
Tons/Year
5,670,000
2,
2,
189,000
108,000
826,800
390,200
291 ,000
149,000
41,000
11,665,000
Fig. 1. Total NOX emitted in the U.S. from stationary sources, 1972.
(4)
22
-------
Table 4
PO
SUMMARY OF TOTAL NOX EMISSIONS FROM FUEL-USER SOURCES (1972)^
NOX Production, 106 tons/yr (percent of
Source
Utility Boilers
I.C. Engines
Reciprocating
Gas Turbines
Industrial Boilers
Commercial /Resi-
dential Heating
Process Heating
Non-Combustion
Incineration
Totals by Fuel
Gas
1.114(9.55)
1.873(16.06)
0.172(1.47)
0.495(4.24)
0.3308(2.84)
0.1855(1.59)
-
-
4.1703(35.75)
Oil
0.768(6.58)
0.316(2.71)
0.119(1.02)
1.098(9.41)
0.467(4.00)
0.149(1.28)
-
-
2.9174(25.01)
Coal
3.788(32.47)
-
0.515(4.41)
0.029(0.25)
0.0553(0.47)
-
-
4.3873(37.61)
total )
Totals
by Source
5.670(48.61)
2.189(18.77)
0.291(2.49)
2.108(18.07)
0.8268(7.09)
0.3902(3.35)
0.149(1.28)
0.041(0.35)
11.665(100)
Cumulative
Percentage
48.61
67.38
69.87
87.94
95.03
98.38
- 99.66
100
-------
of fuel nitrogen conversion to NOX in coal (and oil). This phenome-
non will be discussed later.
Table 5 provides a perspective summary of fuel use by combustion
sources. The fuel usage for process heating (furnaces, kilns, etc.)
is a crude estimate. However, the overall distribution of fuel use
by source is reasonably accurate. It shows that gas usage was nearly
twice that of coal. It is expected that coal usage will increase
significantly in future years as a result of our national energy
situation. Utility boilers were the largest users of fuel with com-
mercial and residential heating being a very close second.
Table 5
SUMMARY OF FUEL USE BY COMBUSTION SOURCE (1972)
(4)
Source
Utility Boilers
Commercial and
Residential Heating
Industrial Boilers
I.C. Engines and
Gas Turbinss
Process Heating
Totals by Fuel
*Includes Liquified Gas
Fuel Used
Gas*
3.8(8.6)
7.9(17.9)
4.5(10.2)
0.7(1.6)
18.6(42.2)
, 10lb Btu/yr
Oil
2.6(5.9)
5.9(13.3)
5.5(12.4)
Oc M ~\ \
0.6(1.4)
15-. 1(34.1) 1
(percent of
Coal
8.4(19.0)
0.2(0.4)
1.8(4.1)
0.1(0.2)
0.5(23.7)
Total )
Totals
by Source
14.8(33.5)
J
14.0(31.6)
11.8(26.7)
o of c n^
1.4(3.2)
44.2(100)
24
-------
Despite its second place position in terms of fuel use, NOX
emissions from commercial and residential heating were ranked fourth.
This is due primarily to the comparatively low emissions from this
source category and the fact that they generally fire the cleaner fuels
(gas and No. 2 oil). Gas turbines and I.e. engines, despite their
low ranking in terms of fuel use, were second in terms of NOX produc-
tion due to their high emission levels and widespread use.
?
It is important that when analogies are drawn between fuel use
and source that full consideration be given to population factors
(i.e., number of units) and relative emissions between sources and
fuels.
This type of information is especially helpful in planning R&D
programs for combustion control of NOX from stationary sources. The
overall NOX ranking provides direction as to the major source categor-
ies requiring control. Further breakdown by fuel type gives additional
insight for control strategies. Still further refinement by equip-
, ^
ment type and fuel allows individualized assessment of control tech-
nology.
MECHANISMS OF NO FORMATION
Before discussing the types of NOX control technologies that
are presently available, it is worth reviewing the mechanisms by which
NO is formed. An evaluation of the background data has led to the
conclusion that NO formed in the combustion of fuels can occur by two
-------
principal mechanisms: (1) high temperature fixation of molecular
nitrogen in the combustion air to yield "thermal NO", and (2) conver-
sion of chemically bound nitrogen in the fuel to yield "fuel NO".
The important reactions and kinetic rate constants associated with
thermal NO were initially identified by Zeldovich in 1948^ ' and have
subsequently been studied by other investigators. The rate of forma-
tion of thermal NO is exponentially dependent on temperature and has
a one-half power dependence on oxygen concentration. In practice,
the actual levels of thermal NO ultimately formed are a function of
additional factors, e.g., air/fuel mixing patterns, heat release/
removal rates, and fuel injector design.
Recently, the importance of fuel NO has become apparent. Initial
investigations by Shaw^ ' indicated that nitrogenous compounds added
fo\
to gaseous fuels could be converted to NO. Work by Martinv ' using
distillate oil doped with varying concentrations of several nitrogen-
ous compounds gave insight as to the degree of nitrogen conversion to
NO. This work showed that the fraction of bound nitrogen converted
to NO decreased with increasing nitrogen content although the absolute
magnitude of the NO increased. Additionally this work showed that
the degree of conversion increased with increasing excess air. it was
postulated that fuel nitrogen was relatively insensitive to tempera-
(9)
ture and strongly dependent on oxygen availability. Turner v 'ob-
tained similar results in a modified package boiler using residual
oil with naturally occurring fuel nitrogen. Investigations by Heap
26
-------
et al.' ' attributed NO formation in coal flames to conversion of
volatile nitrogen compounds present in coal. Experiments by Pershing^ '
confirmed the dominance of fuel NO in residual oil and coal flames by
substituting a 21% oxygen/79% argon atmosphere for the combustion air.
This recent work additionally served to confirm the temperature in-
sensitivity of fuel nitrogen conversion.
As a consequence of prior experimental work it becomes important
to give due consideration to the type of fuel being burned before
selecting a combustion modification technique for NOX control. A
discussion of available combustion modification techniques follows.
NOy COMBUSTION CONTROL TECHNOLOGY FOR BOILERS
Low Excess Air Combustion
Low excess air combustion is one of the most promising and widely
applicable combustion modification techniques for reducing NOX. It
has been used successfully with all fuels and employed on a variety
of furnace types. By reducing oxygen availability at the burner(s),
both thermal and fuel NO can be reduced. The actual degree of NOX
reduction is a function of furnace design, fuel type and overall excess
air. When employing this technique, the lowest practical excess air
levels are generally dictated by a need to limit products of incomplete
combustion (CO, hydrocarbons and smoke) or to prevent operating prob-
lems such as boiler vibrations, slagging and fireside corrosion.
27
-------
Flue Gas Recirculation
The use of flue gas recirculation (FGR) in combustion processes
is not a new technique. Its principal use, at least in utility boilers,
is for steam temperature control. For that application, the flue gas
is recirculated to the furnace bottom. To be useful as a NOX control
technique, FGR must be directed through the burners into the primary
combustion zone. Its effect on reducing NOx is two-fold: (1) the flame
zone temperature is reduced by the recirculating flue gases, and (2)
the concentration of oxygen available for NO production is reduced.
The thermal effect is generally considered to be more important.
Since FGR will reduce thermal NO primarily, its use is generally
restricted to gas fired systems or to very low nitrogen fuels (e.g.,
distillate oils) where fuel nitrogen conversion is not particularly
significant. The technique can be expensive since a high temperature
fan and additional ductwork is required. Typical FGR rates of 15
to 30% have been employed. Burner stability problems can occur at
higher rates. Application of FGR to existing boilers can adversely
affect steam side performance because of the increased gas velocities
through the convective passes of the boiler and load reduction may be
necessary to maintain gas velocities within design limits. Additional-
ly, space limitations on existing boilers may make installation of FGR
difficult.
28
-------
Hater or Steam Injection
The use of water or steam injection into the combustion zone
can reduce thermal NO since its effect is to reduce flame temperature.
Its use would be restricted to gas or distillate oil fired systems
where thermal NO predominates. It may be somewhat more effective on
a mass basis than flue gas recirculation for reduction of thermal NO.
The installation cost is quite low especially on gas fired systems
equipped for standby oil firing since the oil guns provide a conven-
ient point for injection and only a pump and piping is required.
However, in the case of boilers, a severe penalty occurs in operating
costs because of the increase in gas flow through the boiler and latent
heat losses in the stack. It could be used as a "trimming" technique
to maintain acceptable NOX levels for gas and distillate oil fired
systems. The maximum amount of water injection that could be used
would be about 50 pounds of water per million Btu fired. This would
impose an efficiency penalty of at least 5%.
Staged Combustion
Staged combustion for utility boilers was developed in the
late 1950's cooperatively by Southern California Edison (SCE) and the
Babcock & Wilcox Company. The study led to the investigation of burners
operating with 90 to 95% of stoichiometric air through the burners with
the balance of the combustion air admitted through "NO" ports located
above the burners. Later work by SCE and KVB Engineering produced a
modification to staged combustion known as "off-stoichiometric firing".
29
-------
The technique involves firing lower sets of burners fuel rich and
upper burners either fuel lean or on air only. A variation of this
technique, "biased firing", operates burners in staggered configura-
tions of fuel rich and either fuel lean or on air only. Staged com-
bustion is effective in reducing both thermaland fuel NOX emissions
because of limited oxygen and lower flame temperatures in the primary
combustion zone, and lower effective temperatures in the secondary,
air-rich combustion zone. Unless the active burners can carry the
fuel requirement.for.full load, use of off-stoichiometric firing and
biased firing can bri.ng about a load reduction. This is often true
for coal fired boilers where pulverizer capacity may be a limiting
factor. New .unit designs are being built with overfire air ports so
that all burners are active. This technique is applicable for all
fuels especially for coal where the bulk of the NO results from fuel
nitrogen conversion. Two potential problem areas with coal fired
boilers (i.e.., slagging and fireside corrosion) are being addressed
in EPA-sponsored studies.
Reduced Air Preheat Temperature
Reduction of combustion air preheat temperature has the effect
of lowering combustion zone peak temperatures with a subsequent reduc-
tion in thermal NOX. It is not considered a practical control tech-
nique because of its adverse impact on thermal efficiency.
30
-------
Load Reduction
Operating combustion processes under reduced load generally
brings about reductions in thermal NOX because of the decrease in
combustion intensity and peak temperatures. Load reduction is not
consistent with meeting today's energy needs especially for electric
generation.
Burner Design
Experimental work being conducted under EPA contracts is show-
ing good promise for low-NOx burner designs applicable to all fuels
and a variety of systems. Hork performed for EPA by the International
Flame Research Foundation^ ' in IJmuiden, Holland has had primary
emphasis on pulverized coal flames with a secondary emphasis on resid-
ual oil. In addition to input/output testing to evaluate promising
low-NOx conditions, the work also included detailed mapping of se-
lected flames for definition of important aerodynamic effects and
pollutant specie concentrations. Several promising low-NOx (150-300 ppm)
burner configurations for use in wall fired pulverized coal boilers
have been identified. Future R&D will develop scale-up criteria for
pulverized coal burners to a size of about 108 Btu/hr. to validate
emission and combustion performance. Following this effort, a proto-
type burner will be installed in an industrial boiler. Later, the
design will be tested in a multi-burner utility boiler.
31
-------
EPA in-house studies of a number of commercially available dis-
(12)
til late oil fired burnersv ' concluded that an optimum burner design
could control NO with smoke-free low excess air operation. An EPA-
(13)
sponsored study with the Rocketdyne Division of Rockwell Internationar '
is developing an optimum distillate oil burner for residential and
commercial applications. To date, the optimized design has produced
NO levels from 50 to 65% of those from conventional burners and is
capable of smoke-free operation at 10% excess air.
Burner design studies of gas fired industrial burners are being
performed for EPA by the Institute of Gas Technology. The first phase
of work consisted of input/output testing of five burner types with
(14)
detailed flame probing of selected burners under low-NOx conditions. '
On-going work is evaluating three specific burner types (kiln, baffle
and boiler burners) for input/output testing and detailed flame analy-
sis. NO reductions of 40-50% have been achieved through burner/fuel
injector changes and are showing the importance of aerodynamic effects
on pollutant formation.
Corporate-sponsored burner design work is also being performed
by such companies as TRW, Babcock & Wilcox, Exxon R&E, and Foster
Wheeler.
In the long term, improved burner design may well replace the
external combustion modifications now in use and achieve NOX levels
significantly below today's state-of-the-art.
32
-------
NOy COMBUSTION COriTROL TECHNIQUES FOR STATIONARY I.C. ENGINES
Most of the control technology for stationary I. C. engines has
been developed by the automotive industry. Table 6 lists some of the
techniques for pollutant control for stationary I.C. engines.
Table 6
POLLUTANT CONTROL TECHNIQUES FOR STATIONARY I.C. ENGINES
(15)
Technique
Major Effect
Potential Problems
Speed vs stoichiometry
Decreased torque load
(at constant speed)
Decreased air manifold
temperature and exhaust
back pressure increase
Increased valve overlap
Exhaust gas recircula-
tion
With speed, NO increases
under fuel rich and de-
creases under fuel lean
conditions
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Pre-combustion chamber Two-stage combustion
Water injection
Fuel injection/ignition
timing
Variable compression
ratio
Exhaust thermal reactor
and catalytic converter
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Retrofit difficulties;
inability to meet load
demand
Retrofit difficulties;
inability to meet load
demand
Efficiency reduction
Fuel economy reduction.
Applicable only to 4
cycle engines
Additional controls;
intake manifold fouling;
operating difficulties;
efficiency reduction
Primarily for liquid
fueled engines; higher
first cost
Decreased efficiency;
higher operating and
maintenance costs
Could reduce power out-
put and efficiency
Higher first cost and
maintenance
Oxidation of CO and HC. Higher first costs, main-
Reduction of NOX with NH3 tenance and operating
or natural gas cost. Catalyst develop-
mental problems
-------
The optimum choice for any engine is dependent on the effective-
ness of the control technique weighted against its effect on engine
reliability, durability and life. A study by McGowin' ' recommends
the following NOX control techniques for stationary diesel and natural
gas engines: "
ENGINE . ., .. SHORT & IMMEDIATE TERM LONG TERM
Diesel Water Injection Precombustion Catalytic NOX
Chamber Reduction
Natural Gas. Water Injection. Increased Catalytic NOX
Valve Overlap for 4 cycle Reduction
naturally aspirated engines
NOy COMBUSTION CONTROL TECHNIQUES FOR STATIONARY GAS TURBINES
Most of the control techniques for stationary gas turbines are
being developed by the engine manufacturers. Additionally, NOX con-
trol for aircraft engines is under development by the manufacturers
and NASA. Some of the technology for flight engines could be applicable
to stationary engines. Table 7 lists some of the techniques for pollu-
tant control for stationary gas turbine engines.
It appears that for the short term, water injection will be used
to control NOX from stationary gas .turbines. For the long term, dry
control technology based on combustor design concepts will replace
water injection.
34
-------
Table 7
POLLUTANT CONTROL TECHNIQUES FOR STATIONARY GAS TURBINES
(15)
Technique
Major Effect
Potential Problems
Primary zone leaning by
modified combustion
chamber design
Water injection
Exhaust gas recircula-
tion
Reduced turbine inlet
temperature
Peak temperature reduc-
tion. Reduced residence
time of fuel at peak tem-
perature
Peak temperature reduc-
tion ,
Peak temperature reduc-
tion
Peak temperature reduc-
tion
Less control over flame
stabilization. Less
control over lower lean
extinction performance
Reduced efficiency. In-
creased maintenance.
Additional equipment for
handling demineralized
water. Higher operating
costs
Reduced efficiency. Addi-
tional controls. Opera-
tional problems
Reduced efficiency. Re-
quires larger turbine
size for a given power
output
NOy COMBUSTION CONTROL TECHNIQUES FOR COMMERCIAL. RESIDENTIAL AND PROCESS
HEATING
Some of the techniques described for boilers have application to
commercial, residential and process heating. A prime consideration for
these classes of equipment is cost since external combustion modifica-
tions (e.g., staged combustion or flue gas recirculation) could markedly
increase the first cost of these systems. Presently, low excess air
firing and improved burner designs appear to be the most viable techniques
for this equipment class. EPA-sponsored studies of improved burner designs
for process furnaces and residential heaters appear promising as a cost-
effective method of substantially reducing NOX emissions from these systems.
35
-------
COMBUSTION MODIFICATION EXPERIENCE
Utility Boilers
Utility boilers encompass the size range of 250 million Btu/hr.
and above. Uncontrolled NOx emissions from these boilers cover a rather
wide range. Table 8 gives typical levels of NOX from utility boilers
together with the levels of NOX reduction achieved with combustion
modifications. Uncontrolled values shown parenthetically are instances
of high emissions but are not necessarily typical. The percent NOX
reductions shown for low excess air (LEA) firing, LEA + staging and
flue gas recirculation (FGR) were attained under full load (90 to
105% of namep.late rating) conditions. The data shown under "25% load
reduction" are presented to give an indication of the degree of NOX
reduction achieved with load reduction alone.
Low excess air firing and combinations of low excess air and
some form of staged combustion are strong tools for NOX reduction and
are applicable to virtually all classes of utility boilers and fuel
types. Flue gas recirculation has not been as widely practiced and
its efficacy is most pronounced for natural gas where the NOX is ther-
mally produced. Data on combustion modifications to cyclone fired
boilers is limited. This equipment class is the most difficult of all
to control by conventional combustion modifications because of its
comparatively limited flexibility. ,..
All utility boiler manufacturers are now guaranteeing that NOX
emissions on new units will not exceed the NSPS limits for the particular
36
-------
oo
Table 8
NOX LEVELS AND TYPICAL NOX REDUCTIONS WITH COMBUSTION MODIFICATION TO UTILITY BOILERS (17»18»19v2°)
Boiler Type
Tangential
Tangential
Tangential
Front Wall
Front Wall
Front Wall
Horizontally
Opposed
Horizontally
Opposed
Horizontally
Opposed
Cyclone
Cyclone
% NOX
Fuel Uncontrolled NOX, ppm* LEA
Gas
Oil v
\
Coal
Gas
Oil
Coal
Gas
Oil
Coal
Oil
Coal
100-350
100-350 (600)
300-600
130-700 (1500)
250-550 (650)
425-850
400-950 (1500)
200-500
400-900
300-500
500-1200
37
28
27-49
15-23
19-35
11-29
15-21
3-9
21-28
10-16
-
REDUCTION WITH COMBUSTION MODIFICATION
LEA +
Staging** FGR 25% Load Reduction
24-72 67
Typical NOX reduc-
20-55 10-40
tions for boilers
24-59
with a 25% load re-
23-72
duction are:
26-45 31-46
34-53 - Gas: 30%
Oil: 10%
34-71
Coal: 18%
19-40
34-48
-
-
*ppm @ 3% 02, dry basis
**Variations of staged combustion. Includes two-stage, off-stoichiometric firing and/or biased firing.
-------
fuel being burned. Their approach is to use one or more of the com-
bustion modification techniques available in their new unit designs.
Although Source Performance Standards have not yet been promul-
gated for existing utility boilers, it is significant that the data
developed thus far has been obtained from existing units. Consequently,
these techniques could be applied to a large portion of the existing
utility boiler population.
Industrial Boilers
Industrial boilers include the size range of 10 x 106 to 500 x
106 Btu/hr. Field testing of industrial boilers is being conducted
el'.
under EPA contract by KVB Engineering, Inc. The first phase of work
has been completed. During this phase, 47 boilers were tested. Since
some of these boilers were tested with more than one fuel and/or burner,
a total of 75 sets of test data on different boiler/burner/fuel combina-
tions were obtained.
Typically, existing industrial boilers in the range from 10 x
105 to 250 x 106 Btu/hr. have limited flexibility to permit combustion
modifications. This is due primarily to the relatively small size and
simple construction of these boilers especially those with single bur-
ners, fixed air swirl, and unsophisticated control systems. However,
in most cases, NOX reductions could be achieved by off-stoichiometric
combustion for multiple burner units or by changes in excess air level,
burner adjustments or other operational parameters. Industrial boilers
greater than 250 x 106 Btu/hr. generally have the same flexibility
38
-------
as utility boilers, and thus are more adaptable to combustion modifi-
(21)
cation. Pertinent results of the Phase I work are summarized below.
NOX emissions are corrected to 3% 02» dry basis.
Coal Fired Boilers
Base load NOX emissions for these boilers varied from 224 to
800 ppm for a variety of underfed and spreader stokers, pulverized
coal units and one cyclone unit. Excess air was found to be a strong
factor affecting NOX emissions, decreasing with decreasing excess air.
For the watertube boilers, NOX decreased on the average by 50 ppm for
each one percent decrease in 02. Boilers with lower volumetric heat
release rates (Btu/hr.-ft.3) had lower NOX emissions. Similarly a
strong dependence on burner heat release rates was shown. Burners
in the 8 to 30 MBtu/hr. size range had NOX emissions between 200 and
400 ppm. Burners in the 30 to 100 MBtu/hr. range had NOX emissions
between 370 and 600 ppm. The highest NOX emissions of 800 ppm occurred
with a 255 MBtu/hr. cyclone coal combustor.
One of the most successful tests on coal fired units was con-
ducted on a spreader stoker traveling grate boiler. Staged combus-
tion was simulated by utilizing auxiliary oil burner ports as air
injection ports. Diverting some of the air from the grates to these
burner ports resulted in lower NOX emissions and higher grate tem-
peratures, so a compromise between NOX reduction and grate temperature
was required. The resulting NOX emissions were reduced by 20-25%
over the boilers operating load range.
39
-------
Oil Fired Boilers
Base load NOX emissions for these boilers varied from 50 to 200
ppm with No. 2 oil and 150 to 619 ppm for No. 5 and No. 6 oils. The
influence of fuel nitrogen content was clearly shown. Base load NOX
emissions ranged from about 105 ppm for 0.01% nitrogen fuel oils to
about 400 ppm for 0.5% nitrogen fuel oils. This corresponds to an
average fuel nitrogen to NO conversion of about 46%.
The oil fired firetube boilers tested during the program with No.
2 and No. 5 oils showed little dependence of NOX emissions on excess
air and load. All the firetube boilers used ambient temperature com-
bustion air. The oil fired watertube boilers with and without pre-
heated combustion air showed decreasing NOX with decreasing excess
air. The No. 2 oil tests were not as sensitive to excess air levels
as were the No. 5 and No. 6 oil tests. Boilers with lower volumetric
heat release rates (Btu/hr.-ft.3) had lower NOX emissions. Similarly,
burners with lower heat release rates had lower NOX emissions.
Multiburner oil fired units were successfully operated under
off-stoichiometric conditions by terminating fuel flow to individual
burners and using that burner port as an air injection port. NOX
reductions of 17 to 49% were achieved by this technique. Generally,
removing a top burner from fuel service gave a higher NOX reduction
than removing a lower burner.
For properly atomized oils, the type of atomizer (air, steam
or mechanical) did not significantly affect NOX emissions for a particu-
lar fuel oil.
40
-------
Gas Fired Boilers
Base load NOX emissions for these boilers ranged from 50 to 375
ppm. Combustion air temperature strongly affected NOX emissions as
expected since the NOX from gas firing is thermally produced. To
further illustrate, base load NOX emissions for boilers using ambient
temperature combustion air varied from 55 to 116 ppm (70 to 116 ppm
for watertube boilers and 55 to 107 for firetube boilers). Base
load NOX emissions for watertube boilers with preheated combustion
air ranged from 90 to 374 ppm. The firetube boilers were not equipped
with air heaters.
NOX emissions decreased with decreasing burner heat release
rates. This relationship was closely coupled to combustion air tempera-
ture as was the effect of excess air. For example, the decrease in
NOX with decreasing excess air was more pronounced when preheated com-
bustion air was being employed.
Multiburner natural gas fired boilers were successfully operated
under off-stoichiometric conditions by terminating fuel flow to indi-
vidual burners and using that burner port as an air injection port.
NOX reductions of 12 to 40% were achieved by this technique. In one
instance, a 24% NOX reduction was obtained by only adjusting the fuel/
air mixture ratio at the burners. No burners were taken out of service
for this test.
The ability to reduce NOX emissions from industrial boilers is
dependent on the design of the boiler and the fuel being fired. The
41
-------
importance of this field testing work is to illustrate that NOX emis-
sions from this equipment class can be significantly reduced. New
unit designs could incorporate features for staged combustion and
controlled air/fuel mixing at nominal cost.
Future field testing of industrial boilers will serve to refine
the data obtained thus far and explore in greater detail the effective-
ness of combustion modifications.
Stationary I. C. Engines
Stationary I. C. engines are built in a wide range of sizes
depending on their "application. Generally, spark-ignition engines
range from fractional horsepower single cylinder units to large multi-
cylinder units of over 1000 horsepower. Diesel engines range in size
from under 100 horsepower to over 1000 horsepower.
NOX emissions from this equipment category cover a wide range
of about 100 to 3000 ppm depending on engine type, size, design, fuel,
load, fuel/air stoichiometry, etc. In fact, NOX emissions from sta-
tionary I.C; engines are subject to as many, if not more, variables
as other combustion systems. The bibliography on this subject is
quite extensive and it is beyond the scope of this overview paper
to delve into this broad area in great detail. The subject has been
treated in depth most recently in EPA-sponsored studies conducted by
/i c \ /op \
Shell^ .and Aerospace. . Consequently, only a brief summary of
emission control techniques for stationary I.C. engines is presented.
Derating of stationary engines, although effective for NOX control,
is economically unattractive because of the increase in the cost of
42
-------
the engine per unit horsepower output. However, NOX reductions of
40% have been obtained with 30% derating of diesel engines.
Changes to fuel injection timing or ignition timing retard are
applicable to diesel and spark ignition engines and require no hard-
ware changes or additions. Injection/ignition retard is not particu-
larly cost effective because the technique can result in high specific
fuel consumption and higher exhaust gas temperatures. For example,
at 12 degrees ignition retard, NOX reductions of 50% have occurred
but fuel consumption increased 12%. A limited amount of injection
retard combined with intake air cooling could be used on diesel engines
with virtually no loss in fuel economy and achieve NOX reduction of
about 25%.
Exhaust gas recirculation and water injection can be employed
on stationary engines. NOX reductions of about 60% have been realized
with a slight increase in specific fuel consumption.
Turbocharging of naturally aspirated diesel engines combined
with retarded injection timing and interceding can give NOX reduc-
tions of up to 35%. Specific fuel consumption did not seem to be
affected.
Thermal reactors and catalytic converters for exhaust gas treat-
ment have been developed for automotive application. Their effective-
ness has not been adequately demonstrated for stationary engines. These
techniques could control HC and CO but would not effect NOX unless in-
jection of a reducing species like Nf^, H2 or natural gas were used for
catalytic reduction of NOX.
43
-------
Gas Turbines
Gas turbines for stationary application range in size from 40
to 87,000 horsepower (about 65 MW) although larger machines are being
designed. Base load NOX emissions vary widely and are a function of
turbine size and fuel type. Under base load conditions NOX emissions
increase with increasing turbine size primarily due to the higher
combustor pressures and turbine inlet temperatures employed for in-
creased efficiency, and increased residence time at high temperature.
NOX emissions from oil fired turbines are higher than those from gas
fired turbines due in part to fuel nitrogen conversion. NOX emissions
are referenced to 15% 02 because this is generally the 02 level in a
gas turbine exhaust. Typical base load NOX levels from uncontrolled
gas and oil fired turbines are shown in Figure 2.
LO
CL
Q.
X
o
280
240
200
160
120
80
40
0
SIMPLE CYCL
0. 2 OIL
I
SIMPLE CYCLE
NATURAL GAS
I
I
1
I
I
I
10 20
30 40 50
BASE LOAD,
I
60 70 80
840
720
600
480
360
240
120
0
Fig. 2. Typical gas turbine base load NOX emissions.
(22)
o
X
o
ro
44
-------
The levels are also shown corrected to 3% 02 so that they may
be compared to other conventional combustion equipment. It can be
seen that NOX emissions from gas turbines are comparable to those from
utility boilers. The variance in NOX levels between turbines of equal
size is due in part to design differences and/or fuel characteristics.
Combustion modification experience for gas turbines has centered
chiefly around schemes for peak temperature reduction. This is a
generally viable approach since the principal fuels used in these ma-
chines are natural gas and distillate oils and the bulk of the NOX
produced is thermal. However, gas turbines are being built to fire
residual oils. In these instances, conversion of fuel nitrogen to
NO could be significant.
Tests conducted with water and steam injection have shown sig-
nificant reductions in NOx although turbine efficiency is affected.
Since the heat of vaporization is about one-third of the total heat
absorbed by water, steam injection requires greater flow rates than
water to achieve the same flame temperature reduction. However, the
better mixing achievable with steam can tend to reduce the difference
in the flow rates. Since the heat required to vaporize water is not
recovered in the turbine cycle, the useful heat input to the turbine
is reduced and turbine efficiency drops. However, with steam injec-
tion, the mass flow through the turbine is increased without additional
compressor work and the thermal efficiency can improve. Typically
one to three percent decreases in thermal efficiency occur when water
is injected at a rate equal to one percent of the turbine air mass
45
-------
flow rate (one percent of the air mass flow rate is equivalent to
about 60 percent of the fuel flow rate or a water/fuel ratio of 0.6).
The actual degree of NOX reduction is strongly influenced by
the water or steam to fuel ratio. Equally important is the injec-
tion technique. A properly injected engine can achieve higher NOX
reductions at equivalent or lower water/fuel ratios than can an im-
properly injected engine.
Figure 3 shows a broad data band indicating the effectiveness
of water or steam injection. The data band includes a variety of
turbine sizes firing natural gas and oils. Generally, for water or
steam injection ratios in the i?ange of 0.6 to 1.0 NOX reductions
of about 50-75% have been achieved with oil firing and about 60-90%
reductions have been achieved with gas firing. The proposed gas
turbine New Source Performance Standards of 55 ppm (gas) and 75 ppm
(oil) at 15% Q£ should be achievable on an individual basis by adjust-
ments in injection ratios. Experience to date does not indicate any
adverse effects of injection techniques on turbines, provided the
dissolved solids in the water are below 5 ppm or that steam is kept
superheated prior to injection.
Turbine manufacturers are continuing development work on low
NOX combustors since it is felt that "dry" controls will be more
economical and attractive to users. Techniques being evaluated in-
clude primary zone leaning, and exhaust gas recirculation both of which
serve to suppress peak temperature. Advanced concepts being studied
46
-------
include premixed, prevaporized, and well stirred external combustors.
These combustors could offer more flexibility toward reducing emis-
sions since, being external, they would not have to be designed to
fit between the compressor and turbine as are typical combustors.
Additionally, advanced combustors could incorporate variable geometry
features to optimize air/fuel stoichiometry over the turbine's load
range.
Table 9 summarizes various combustion modifications applicable
to gas turbines and gives an indication of the degree of pollutant con-
trol achievable with these techniques.
Commercial and Residential Heating
Boilers for commercial heating cover a capacity range of 10 to
300 boiler horsepower (about 0.30 to 10 million Btu/hr.). Residen-
tial heating units generally include a size range of 75,000 to 300,000
Btu/hr. The most recent and comprehensive studies of emissions from
these classes of equipment have been performed by Battelle' ' '
(12)
and EPA.V ' Typical baseline emission levels of pollutants from
commercial boilers and residential heating systems are presented in
Table 10.
This emission data is derived from field tests performed by
Battelle. Thirty-five residential systems (33 on oil and two on gas)
and 13 commercial boilers firing oil and/or gas (33 boiler/fuel com-
binations) were tested. Obviously variations occurred for all pollu-
tants due to boiler size, design, age, burner type, fuel, operating
48
-------
TOO
80 __
o
II
d
2 60
Q
UJ
o:
x
o
o 40
20
0.4 0.6 0.8
WATER/FUEL RATIO
Fig. 3. Effectiveness of water or steam injection in reducing NOX
(23)
formation.v '
-------
conditions, etc. The data presented is only meant to give a general
indication of emission levels for 'a broad range of equipment. The
effect of fuel type on NOX emissions is most pronounced. Conversion
of 40 to 60% of the fuel nitrogen to NOX was indicated in these studies.
Table 10
TYPICAL BASELINE EMISSION LEVELS FROM COMMERCIAL AND"RESIDENTIAL HEATING
Unit
Residential
Residential
Commercial
Commercial
Commercial
Commercial
Commercial
Commercial
Emission Concentration, ppm 3% 02, dry basis
NOX as Bacharach
Fuel N0£ CO HC Smoke
Gas
No. 2 Oil
Gas
No. 2 Oil
No. 4 Oil
LSR*
No. 5 Oil
No. 6 Oil
70
115
80
100
390
260
290
415
15
65
20
4
7
3
16
10
3
13
9
3
3
5
4
5
0
3
0.2
0.9
2.6
2.9
3.0
3.9
*Low Sulfur Residual Oil H% S)
Combustion modification of this class of equipment is quite lim-
ited. Equipment in this size range is all of the single burner type, and
is of relatively simple design. Operating variables are generally lim-
ited to excess air variations, load, atomization method and/or pressure.
Generally, the range within which operating variables can be operated
50
-------
Table 9
EFFECTIVENESS OF COMBUSTION MODIFICATION ON GAS TURBINES
(22)
Combustion
Modification
NOv
Emissions
CO
HC
Smoke
Interim
Primary Zone
Leaning
10 - 30%
reduction
Small
reduction
Smal 1
reduction
Reduction
Exhaust Gas
Recirculation
^30%
reduction
Negligible
effect
Negligible
effect
Negligible
effect
Water/Steam
Injection
50 - 75%
reduction
(oil)-
60 - 90%
reduction-
(gas)
Some-
reduction
or in-
crease
Smal 1
reduction
or small
increase
Small
increase
possible
Advanced Combustors
Premixed,
prevaporized,
well stirred,
variable geometry,
external combustors
20 ppm 40 ppm 5 ppm Invisible
achievable achievable achievable
49
-------
is quite narrow and adjustments of burners for these systems is often
based on acceptable smoke and C02 performance.
Commercial Boilers
Emissions from commercial boilers can be reduced by controlling
excess air levels to optimize all gaseous emissions. Reduced excess
air could reduce NOX emissions by about 10% for the higher,nitrogen
fuel oils; i.e., No. 4 and above.
An obvious method of reducing NOX and smoke is to select a fuel
oil having a lower nitrogen content. Economics and supply factors
will dictate the extent to which this technique can be practiced.
Improved burner design can ultimately provide significantly lower
levels of NOX without adversely affecting CO, HC, and smoke. By
providing aerodynamic staging and internal recirculation of flue gases,
both fuel and thermal NOX can be suppressed. EPA-sponsored R&D in
this area is showing promising results. Low emission burner designs
will probably be the most cost effective method for reducing combus-
tion related pollutants from commercial boilers.
Residential Systems
Emissions from residentia-1 systems-can be reduced by proper
tuning and maintenance. This is effective in reducing smoke, particu-
1 ate,,. CO. and HC emissions although NOX is unaffected. Tuning can
improve.efficiency by about.2%. and subsequently result in lower emis-
sions as a result of fuel,savings.
51
-------
Identification and replacement of units in poor condition would
reduce area-wide emissions. Newer burners typically have lower emis-
sions and higher fuel efficiency than older burners.
Methods for reducing mean pollutant emissions and estimates of
the degree of reduction suggested in the Battelle study are shown
in Table 11.
Table 11
EFFECT ON MEAN EMISSIONS OF IDENTIFYING AND "REPLACING"
RESIDENTIAL UNITS IN "POOR" CONDITION AND TUNING
Pollutant
Reduction in Emissions, percent
Step 1
Identifying and
Replacing of Poor Units
Step 2
"Replacement"
Plus Tuning
Smoke
CO
HC
NOX
Filterable Particulate
_
>65
87
No change
17
59
>81
90
No change
24
Again, improved burner design will ultimately provide for low
emission residential heating systems. Since these units are fired
almost exclusively with gas and No. 2 oil the NOX to be controlled
is thermally formed. Burners having internal flue gas recirculation,
52
-------
controlled mixing of air and fuel, and ability to operate at low
excess air levels should reduce thermal NOV by 60% or more.
A
Process Heating
A comprehensive study of emissions from process heating together
with an estimate of potential N0₯ control techniques has recently been
/\
initiated as an EPA contract. It is expected that this study will
provide definitive guidance for future R&D in combustion control of
NOX from process furnaces together with cost estimates for implementing
such control.
EFFLUENT TREATMENT N0y CONTROL TECHNOLOGY
The Esso System Study(5) identified a number of candidate NOX control
processes based on treatment of process effluent gases. Alkaline scrubbing,
molecular sieve absorption, and catalytic reduction were considered as
possible NOX control techniques. Based on this evaluation, effluent
treatment appeared to hold some promise primarily for specialized cases
where high NOX concentrations are present or very low levels of NOX emissions
are required.
An EPA in-house study examined the feasibility of aqueous alkaline
scrubbing in some detail. The variables consider included NO concentration,
NO to N02 ratio, and scrubbing agent. It was concluded that alkaline
scrubbing was not feasible for the removal of NO alone or equimolar
concentrations of NO and N02- However, it was confirmed that N02 itself
could be absorbed by several scrubbing solutions.^31) However, oxidation
of NO to N02 appeared to be quite expensive.
Molecular sieve absorption is applicable to nitric acid plants
where the NOX emissions are a result of process conditions and can
not be appreciabley changed by process modification. Demonstration
53
-------
of this technology is underway on 2 nitric acid plant tail gases. This
technology does not appear to be applicable to combustion flue gases because
of preferential sorption of the moisture in the flue gas.
EPA activity in effluent treatment has been proceeding at a fairly
low level but may be accelerated in view of delayed automobile compliance
schedules and the need to meet NAAQS.
Current Research and Development
In a contract with Environics, Inc., selective noble metal reduction
of NOX with ammonia is being explored on a pilot scale. The experimental
system utilizes a side stream equivalent to 2 MW from a gas fired
utility boiler as a flue gas source. The flue gas, which contains
around 225 ppm of NO, is mixed with ammonia and then contacted with a
catalyst bed at elevated temperature. NO reductions of 90% have
X
been achieved; however, equal or greater amounts of NH3 (e.g., 25-30 ppm)
are also present in the exhaust. The pilot scale unit has accumulated
about 2000 hours of testing with natural gas firing with essentially no
performance degradation of the catalyst. Current;plans, necessitated
by the recent natural gas curtailment for utility use in the L.A. area,
are to test with sulfur-containing flue gas produced by combustion of oil,
starting in the winter of 1974. In preparation for testing with flue
gas from oil combustion, and electrical preheater has been installed
to raise the flue gas temperature prior to entering the catalyst bed.
A report on the pilot plant results with gas firing and oil firing is
expected to be available in the spring of 1975.
54
-------
TRW Systems has conducted a technical and economic assessment
of various catalytic schemes for NO control for stationary power
A
plants. The study purpose was to review catalyst technology for NO
/\
abatement, assess technical and economic feasibility of various schemes,
evaluate the applicability to power plants, and recommend the most
promising schemes. A comprehensive data bank was developed containing
approximately 250 articles and patents on pertinent NOX catalysts
and catalytic processes. On a lab scale with simulated flue gas (with-
out flyash), catalytic reduction of NO was investigated over a range
of space velocities from 5,000 to 20,000 hr.~^ (standard temperature
and pressure) and temperatures from 200 to 450°C. Approximately 45
catalysts were screened to determine: ndnselective NOX reduction with
\\2 and CO; and selective reduction with Nl^, h^, and CO. Identified as
being most promising (because of their activity) for selective reduction
with NH3 are: iron-chromium, vanadium, copper-lead, molybdenum, and
platinum catalysts. Screening of Fe-Cr and vanadia indicated that the
presence of S02 did not affect the catalyst.
Parametric investigations of best catalyst systems, with major
emphasis on selective NOX-NH3 non-noble systems have been completed.
These parametric studies indicated NO conversions of about 60-95%
with NH3 stoichiometry about 1.0, temperature of 400°C, space velocities
of 5000 to 20,000 hr."1 and inlet NO concentrations of 250-1000 ppm.
A draft report covering the TRW program is undergoing review
and is expected to be available in the very near future.
55
-------
Under a research grant, Montana State University is conducting a
detailed thermodynamic study on the use of metal sulfides for reducing
NO via the following reaction: MeS + 4NO -> MeSCty + 2N2. The reduction
proceeds with all 17 of the different sulfides tested to date, but over a
wide temperature range: from 650°C for CdS to about 100°C for I^S. Operation
in a typical flue gas temperature regime of about 150-175°C is desired.
Certain compounds appear to promote or catalyze the reduction reaction.
Tests with 10 different dry sulfides and promoter compounds such as CoCl2>
NiC^, FeCl2» I^CrFs, and KsFeFs showed significantly lower temperatures
at which the reduction proceeded compared with reaction temperature require-
ments without the promoter. For example, using NaCl as a promoter, the
react tan tpr?ftceejleji^i30Qj°Q jt&fl&5n
-------
Utility Boilers
The most recent and comprehensive study of NOX control costs
for utility boilers was performed by Combustion Engineering, Inc.
under EPA contract.^6) The five combustion modification techniques
investigated were:
1. Introducing 20 percent of the total combustion air over the
fuel firing zone as overfire air.
' 2. 'Introducing 30 percent flue gas recirculation through the
secondary air ducts,and windbox compartment.
3. Combining the 20 percent overfire air and 30 percent flue
gas recirculation of 1 and 2.
;''' ' . ' '
4. Introducing 17 percent flue gas recirculation through the
transport air (primary air) of the coal pulverizer (mills)
system.
5. Introducing water injection into the fuel firing zone at a
rate of 5 percent of total evaporation (equivalent to about
50 pounds of water per million Btu's fired).
Two levels of cost were established; the first being for new unit
designs with heating surfaces adjusted to compensate for changes in
heat transfer distribution and rates resulting from the use of these
techniques. The second level of cost applied to existing units with
no changes to heating surface. For both cases, the costs shown are
in 1973 dollars, and except where noted otherwise are estimated on
a - 10 percent basis.
57
-------
Figures 4 and 5 show the cost ranges for the techniques consi-
dered. It can be seen that the cost ranges for existing units vary
more widely than for new units. This is mainly due .to variations
in unit design and construction which can either hinder or aid the
installation of a given NOX control system. As an example, an over-
fire air system could be designed as a windbox extension unless ex-
isting structural requirements and/or obstructions necessitated in-
stallation of a more costly system Including extensive ductwork and
individual air injection ports. Similarly, water injection could
require changes in unit ducting to maintain unit capacity. The cost
range for the combined overfire air and windbox gas recirculation
is the sum of the cost ranges for the. individual, systems. The cost
ranges for existing units do not include any changes to heating sur-
faces since these changes must be made on an individual unit basis.
Heating surfaces could increase, decrease or remain the same because
of variations in existing designs.
At about 600 MW, single cell furnaces reach a practical size
limit and divided furnace designs are employed. Since a divided
tangentially fired furnace has twice the firing corners of a single
cell furnace, the costs increase significantly.
For new units, flue gas recirculation has the highest invest-
ment cost and overfire air has the lowest investment cost. For ex-
isting units, flue gas recirculation is again the most expensive
control technique. Although water injection on existing units is
58
-------
100
Fig. 4.
WIN DBOX GAS RECIRCULATION
OVERFIRE AIR
^BINED
RFIRE AIR AND WINOBOX
GAS RECIRCULATION
RECIRCULATION THRU
MILLS
WillDBOX WATER INJECTION
200
300
600
700
800
400 500
UNIT SIZE, MW
Costs of NOX control methods for new coal fired units
(included in initial design).(26)
59
-------
V)
8
a:
o
UJ
£
^
w
O
O
a:
O
LJ
NDBOX GAS RECIRCULATION
RFIRE AIR
COMBINED ;, '
OV0RFIRE AIR AND WINDBOX
RECIRCULATION
RECIRCULATION THRU MILLS
ER INJECTION INCLUDING FAN
$DUCT CHANGES
R INJECTION WITHOUT FAN
DUCT CHANGES
100 200 300 400 500 600 700 SCO
UNIT SIZE, MH
Fig. 5. Costs of NOX control methods for existing coal fired units
(heating surface changes not included).
60
-------
slightly cheaper than overfire air, its initially low cost is ulti-
mately overshadowed by its high operating cost as will be later shown,
The cost of low excess air firing-.was
-------
These costs can vary to some degree depending on the actual
extent of modification required and are only provided as guidelines.
Conversion of coal units is more expensive because of the added com-
plexity and controls needed for proper air and fuel mixing. As unit
size increases, the cost per KW decreases since the larger units
typically have inherently greater flexibility and may require less
extensive modification.
The use of low excess air firing reportedly increases boiler
efficiency by 0.5 to 2%, in addition to savings resulting from de-
creased maintenance and operating costs especially for gas and oil
fired units. Consequently, the investment costs, when required,
can be offset in fuel and operating expenses. Thus, minimum or low
excess firing can be the cheapest combustion control method avail-
able.
Table 13 gives the impact on major system components, efficiency
and capacity when employing the NOX control techniques considered.
The relative changes in unit design or efficiency are shown to increase
(or require addition) by a plus (+) or decrease by a minus (-). If
the item is unchanged or changes to a negligible extent it is indicated
as a zero (0). It must be remembered that for this comparison, heat-
ing surfaces are unchanged on existing units.
Thus, new unit designs can be built with full consideration of
the impact of combustion modifications on heat transfer and operational
aspects of the boiler. Capacity and boiler efficiency would not suffer
except in the case of water injection where latent heat losses are not
recoverable.
62
-------
Table 13
IMPACT OF NOX CONTROL TECHNIQUES ON MAJOR UTILITY BOILER COMPONENTS
en
GO
System
Component
Forced Draft
Fan Size
Secondary Air
Ducts
Wind box Size
FGR Fan
FGR Ducts
Dust
Collectors
Coal
Pulverizers
Convective
Surface
Superheat
Surface
Reheat
Surface
Economizer
Surface
Boiler f
Efficiency
Capacity
NEW UNIT DESIGN
Sec. Prim.
O.A.a FGRb a+b FGRC
+ 0 + +
0 + + 0
0 + + +
N/Ae + + +
N/A + + +
0 + + +
0 00 0
0 + + +
0
_ _ _ _
0 + + +
0 0 00
0 0 00
Water
Inj.d
0 or +
0
0
N/A
N/A
0
0
+
_
_
+
_
0
O.A.
0 or
0
0 or
N/A
N/A
0
Sec
a FGR
+ 0
+
+ +
+
+
+
0 or + 0
N/A
N/A;
N/A
N/A
o-.
0
a. Overfire air system d.
b. Flue gas recirculation through the secondary e.
air duct and windbox compartments f.
c. Flue gas recirculation to the transport air
(primary air) of the coal pulverizers (mills)
N/A
N/A
N/A
N/A
0
-
EXISTING UNITS
Prim.
b a+b FGRC
0 or + +
+ 0
+ +
+ +
+ +
+ +
0 or + 0
N/A . N/A
N/A N/A
N/A N/A
N/A N/A
0 0
.-
Water injection to the firing
Not applicable
Average heat rate, Btu/KWH
Water
lnj.d
0
0
0
N/A
N/A
0
0
N/A
N/A
N/A
N/A
-
-
zone
-------
For existing units, boiler efficiency is essentially unchanged
(except for water injection) although unit capacity will drop for all
cases except overfire air operation unless compensation is made to
heating surfaces.
Table 14 summarizes the operating costs for new units using the
NOX control techniques considered (in 1973 dollars). Divided furnaces
were not considered in these estimates. Water injection is the most
expensive system to operate due primarily to increased fuel costs
resulting from losses in unit efficiency. The least expensive to
operate is overfire air.
To put these operating costs in perspective, they can be compared
to the "average" generating costs shown at the bottom of Table 14.
Except for the case of water injection, the differential in operating
cost is below 1% even with flue gas recirculation. The differential
operating cost for overfire air is even lower, ranging from about
0.02 to 0.05%.
It is encouraging that the control techniques having the widest
applicability for NOX control from utility boilers (low excess air
firing and staged combustion) also have the lowest investment and
operating costs.
Although the cost data for utility boilers were developed for
tangentially coal fired utility boilers, it is felt that the range of
costs presented should be generally applicable to wall fired boilers
burning coal. The cost of combustion modification to wall fired
64
-------
cr>
en
Table 14
1973 OPERATING COSTS OF NOX CONTROL METHODS FOR
NEW COAL FIRED UNI
(Single Furnace)
Control Method
MW Rating
Equipment Costs,
Annual Fixed Charge,
Additional Annual
Fuel Cost,
Additional Annual
Fan Power Cost,
Total Annual Cost,
1. Overflre
Air (20%)
103$
103$
103$
103$
103$
Operating Cost, mllls/KWHR
100
31
5
...
5
0.009
450
63
10
-
10
0.004
750
90
14
....
14
0.003
2. Windbox
Flue Gas
Recirc. (30%)
100
350
56
21
77
0.143
450
1185
190
95
285
0.117
750
1650
264
158
422
0.104
3. Combination
of 1 and 2
100
375
60
21
81
0.150
450
1248
200
95
295
0.121
750
1800
288
158
446
0.110
4. Coal Mill
Flue Gas
Recirc. (17%)
100
300
48
22
70
0.130
450
1015
162
100
262
0.108
750
1425
228
166
394
0.097
5. Water
Injection
100
160
26
147
13
186
0.344
450 750
560 825
90 132
660 1099
58 97
808 1328
0.332 0.327
(g)
Based on: a. Delivered and erected equipment costs (- 10% accuracy). Excluding contingency and Interest during construction.
b. 5400 hr/yr at rated MW and net plant heat rate of 9400^Btu/KWHR.
c. 50i/106 Btu coal cost.
d. $250/hp fan power cost, or $40/hp per year.
e. Annual fixed charge rate, of 16%.
f. Operating costs are t 10%.
g. Does not include cost of water piping in plant or cost of makeup water.
Base unit operating costs* for coal fired power plants excluding S02 removal systems.
Unit Size, MW 100 450 750
Operating Cost, mills/KWHR 16.2 13.5 12.6
*Includes 1973 capital costs, labor, maintenance, fuel costs, 20% contingency, 17% Interest during construction.
-------
boilers is being addressed as part of an EPA study being conducted
by the Tennessee Valley Authority. Additionally, it is intuitively
felt that the cost for similar combustion modification on gas and
oil fired utility boilers should be no higher than for the coal fired
units.
Industrial Boilers
Cost data for combustion modification of industrial boilers
is virtually non-existent. Research and development, including field
testing and application of NOX control to this equipment category
is in its early stages. Low excess air firing and some form of staged
combustion afppear to be widely applicable. It is expected that sig-
nificant modifications are required; they will represent a higher
fraction of the base unit cost than utility boilers. Cost estimates
for NOX control of industrial boilers will be developed as part of
on-going and planned EPA studies.
Diesel Engines
Because of the relatively limited emission control work performed
to date, it is difficult to make accurate estimates of cost and main-
tenance requirements for the various control options. Generally, most
engine manufacturers feel that the basic engine cost would be increased
by 10 to 50% and operating cost would increase by about 15% depending
on the degree of NOX reduction, control method and increase in specific
fuel consumption.
66
-------
Spark Ignition Engines
Many of the emission control systems utilized in automobile
engines have application to stationary spark-ignition engines. The
cost of these devices for automotive engines are known but it is
difficult to estimate the cost per unit horsepower when applied to
stationary engines. It is quite possible that similar devices could
be several times more costly for stationary engines. Since these
engines generally operate at constant speed, leaner air/fuel mixtures
can be tolerated. Consequently, the fuel penalty could be lower than
in the case of automotive engines.
Gas Turbines
Costs for NOX control based on "wet" methods have most recently
been estimated in Aerospace and EPA studies. Input to these studies
have come from several gas turbine manufacturers and from utilities
employing water or steam injection as a NOX control technique.
Investment costs for water injection as provided by San Diego
Gas and Electric are shown in Table 15.
For this example, the investment cost for water injection was
about ten percent of baseline cost for 20 MW plants and about six
percent for 29 and 81 MW plants.
67
-------
Table 15
WATER INJECTION INVESTMENT COST
(SAN DIEGO GAS AND ELECTRIC)^22^
Control System
Gas Turbine Size, MW
20 49
81
Combustor modifications
including water injection $1.00/KW $0.86/KW $1.04/KW
nozzles
Water injection pumps
and water regulation $3.54/KW $2.88/KW $3.10/KW
system
Associated piping and
water storage facilities $1.72/KW $1.05/KW $0.87/KW
Water treatment equipment $0.90/KW $0.47/KW $0.47/KW
General expenses includ-
ing engineering, adminis- $1.15/KW $0.82/KW $0.57/KW
tration, testing, taxes
Total $8.31/KW $6.07/KW $6.05/KW
Table 16 illustrates how the investment for water injection and
increased operating cost is prohibitive for the small gas turbines.
Table 16
WATER/STEAM INJECTION COST AS A FUNCTION OF POWER PLANT SIZE
(22)
MW Output
0.26 (350 hp)
2.90 (3900 hp)
20.00
33.00
65.00
Investment Cost,
Percent Baseline
Water Steam
100.0
18.0
10. 0
7.3
7.3
150.0
24.0
12.0
10.6
10.6
Operational Cost,
Percent Baseline
Water Steam
55.0
6.5
6.0
5.7
5.7
165
32
--
--
*For peaking gas turbine, 1000 hr./yr.
68
-------
However, since NOX emissions for the smaller turbines are some-
what lower, they may not require water injection for NOX control.
Instead, they may be able to meet emission levels through combustion
modifications.
Since the operating cost for water injection-decreases with
increasing turbine usage, the operating cost could drop to about
2.5% of basic operating costs for turbines operating 6000 to 8000
hours per year. Thus, NOX control by water injection becomes more
cost effective with increasing load factors.
Similar estimates for NOX control of gas turbines have been
prepared by EPA to support proposed emission standards. Table 17
shows the cost of NOX control for small ,gas turbines using dry and
wet techniques. It serves to illustrate the higher cost for control
on small engines and the reduction in control cost when engines are
run at higher load factors.
Table 18 provides cost estimates for water injection on large
gas turbines. The costs for various load factors and water injection
rates compare quite well to other reported data.
The long term effect of water or steam injection on long term
engine life is still unknown. However, no major problems have been
reported to date when the systems are properly operated. One of the
major disadvantages of water injection is that siting flexibility is
limited for simple cycle and regenerative gas turbines which would
fiQ
-------
Table 17
COST OF NOX CONTROLS FOR SMALL GAS TURBINES
Size, hp
Purchase cost (PC), uncontrolled, $
Total installed cost (TIC)J.SxPC
Total capital investment (TCI)
1.25 TIC, $
Control increment, percent
TCI, controlled, $
Unit investment, controlled, $/hp
Heat rate, B.tu/hph
Equivalent hours duty per year
Fuel @ $0.91/MBtuf, $
Fixed charges, uncontrolled9, $ 2
Total annual cost, uncontrolled, $ 3
Utilities'1, $
Incremental fixed charges9, $
Total annual cost, controlled, $ 3
Control cost, percent
(27)
a Per vendorv
b Assumed
c Combustor modifications per vendor
d As in emergency service, including
e As in pipeline service
f In the pipeline application, fuel
g Carrying charges 17%j maintenance
h Raw water, regeneration chemicals,
8
11
14
350
,800a
,440
,300
Dry ,20b
17
12
100d
380
,600
,000
--
520
,500
17
(27)
>
,200
49
,000
8
30
2
33
33
,000e
,600
,600
,200
--
520
i700
1.6
3
110
143
178
Dry
200
11
100
3,500
32,200
35,700
3,900
3-9,600
11
,500
,000a
,000
,800
12a
,000
57
,000
8
280
32
312
3
316
,000
,300
,200
,500
,900
,400
1.3
3
no
143
178
Wet
224
11
100
3,500
32,200
35,700
12
8,000
43,700
22
,500
,oooa
,000
,800
25C
,000
64
,000
8
280
32
312
.1
8
321
,000
,300
,200
,500
,000
,000
,500
2.9
water treatment system per Crits * '
readiness
from
1%
and
the 1
power
tests
ine would
together
be much less expensive
assumed $1/1000
gallon
-------
Table 18
COST OF WET NOX CONTROLS FOR LARGE GAS TURBINES
(23)
Size,
25
4x65
Capital Costs,
Total Capital Investment (TCI) ,
Uncontrolled
Equivalent Hours Duty per Year
Water/Fuel Ratio ,
Control Increment , percent
TCI, Controlled
Unit Investment, Controlled, $/KW
8000
0.5
10.0
3120
125
2800
1000
0.5
8.5
3040
121
1000
0.8
9.5
3070
123
4000
0.5
3.9
27000
104
26000
1000
0.5
3.5
26900
103
1000
0.8
3.9
27000
104
Annualized Costs, 103$
Heat Rate, Btu/KWH
Fuel @ $0.91/MBtu .
Fixed Charges, Uncontrolled
Total Annual Cost, Uncontrolled
Utilitiese
Incremental Fixed Charges
Total Annual Cost, Controlled
Incremental Annual Cost, percent
2260
504
2764
9
57
2830
2.3
12400
282
504
786
1
43
830
5.6
282
504
786
2
49
837
6.5
11070
4670
15740
40
180
15960
1.4
11700
2770
4670
7440
10
160
7610
2.3
2770
4670
7440
16
180
7640
2.6
d
e
Applying to the 25 MW case, the 1970 Federal Power Survey datum of $85/KW, escalating from 1968
to 1974 at 5% compounded, and assuming a weak economy of scale for the larger case.
Wet controls include an injection system sized for peak injection rate, based on owner's
and a treatment and storage system sized conservatively for peak daily usage, based on Crits.
In the 8000-hour case representing a pipeline compressor, fuel from the line would be much less
expensive.
Carrying charges 17%, maintenance 1%.
Raw water, regeneration chemicals, power and sewerage together at $1/1000 gallons.
-------
normally require no supply of water. This becomes an important consi-
deration in areas of limited water supply or subfreezing ambient tem-
peratures. The decrease in engine efficiency also becomes of prime
importance with todays rising fuel costs.
It is expected that by the period 1979-1985, advanced combustor
designs having very low emissions could be fully developed. These would
ultimately replace "wet" control technology. Development costs would
be in the range of 4 to 10 million dollars including the demonstration
stage.
72
-------
SUMMARY
New Stationary Source Performance Standards for NOX that have
been promulgated or are scheduled for very near term promulgation are
based on current state-of-the-art combustion control technology. Be-
cause of the expected growth in stationary sources, delays in achiev-
ing automotive standards and the need to maintain or improve ambient
air quality, it is expected that more stringent NOX standards and/or
regulations of additional stationary sources may be required. The
need therefore exists for continued R&D to develop the technology
that will be needed in future years. 'Control Systems Laboratory's
R&D posture will be to achieve future NOX goals without adverse im-
i ... '''. i
pact on other combustion related pollutants, i.e., particulate, CO
and UHC. Concurrently, future R&D will strive to maintain or improve
process efficiency to the greatest extent possible.
Projected short and long term R&D goals for NOX control are
shown in Table 19. The 1980 goals are based on extension of existing
technology coupled with new burner designs. It is felt that these
goals can generally be attained with present combustor/furnace con-
figurations. The 1985 goals could possibly entail the redesign of
some equipment types or require the use of novel combustion and/or
effluent treatment schemes.
Obviously, the achievement of these goals is heavily dependent
on the level of funding that is ultimately provided.
73
-------
Table 19
PROJECTED SHORT AND LONG TERM GOALS FOR NOX CONTROL
(ppm (<> 3% O2)
Source
Utility Boilers
Gas
Residual oil
Coal
Industrial Boilers
Gas
Residual oil
Coal
Commercial Boilers
Gas
Distillate oil
Residual oil
1980 Goal
100
150
200
80
125
150
50
70
150
1985 Goal
50
90
100
50
90
100
30
50
90
Spark Ignition Engines
Gas 1200 400
Compression Ignition Engines
Diesel Oil 1200 800
Gas Turbines
Gas 90 90
Distillate Oil 165 165
74
-------
REFERENCES
1. "National Primary and Secondary Ambient Air Quality Standards,"
Federal Register Vol. 36, No. 84, April 30, 1971.
2. EPA, Air Quality Criteria for Nitrogen Oxides, AP-84, January 1971.
3. Crenshaw, John and Allen Basala, "Analysis of Control Strategies
to Attain the National Ambient Air Quality Standard for Nitrogen
Dioxide." Presented at the Washington Operation Research Council's
Third Cost-Effectiveness Seminar, Gaithersburg, Md., March 18-19,
1974.
4. Mason, H. B. and A. B. Shimizu, "Definition of the Maximum Sta-
tionary Source Technology Systems Program for NOX Control."
Draft of final report, October 1974.
5. Bartok, W. et al., "Systems Study of Nitrogen Oxide Control
Methods for Stationary Sources, Vol. II." Prepared for National
Air Pollution Control Administration, NTIS Report No. PB 192-789,
Esso, 1969.
6. Zeldovitch, Y. B., P. Y. Sadonikov, and D. A. Frank-Kamenetskii,
"Oxidation of Nitrogen in Combustion." Academy of Sciences of
USSR, Institute of Chemical Physics, Moscow-Leningrad, 1947.
7. Shaw, J. T., and A. T. Thomas, NTIS No. PB 229-102/AS, "Oxides
of Nitrogen in Relation to Combustion of Coal." 7th International
Conference on Coal Science, Prague, Czechoslovakia, June 1968.
8. Martin, G. B. and E. E. Berkau, "An Investigation of the Conver-
sion of Various Fuel Nitrogen Compounds to Nitrogen Oxides in
Oil Combustion." AIChE/Symposium Series, Air Pollution and Its
Control, Volume 68 (1972).
9. Turner, D. W., R. L. Andrews, and C. W. Siegmund, "Influence of
Combustion Modification and Fuel Nitrogen Content on Nitrogen
Oxides Emissions from Fuel Oil Combustion." Presented at 64th
Annual AIChE Meeting, San Francisco, Calif., November 1971.
10. Heap, M. P., T. M. Lowes, R. Walmsley, and H. Bartelds, "Burner
Design Principles for Minimum NOX Emissions." EPA Coal Combus-
tion Seminar, Research Triangle Park, N. C., June 1973.
11. Pershing, D. W., G. B. Martin, and E. E. Berkau, "Influence of
Design Variables on the Production of Thermal and Fuel NO from
Residual Oil and Coal Combustion." Presented at the 66th Annual
AIChE Meeting, Philadelphia, Pa., November 1973.
-------
12. Hall, R. E., J. H. Wasser and E. E. Berkau, "A Study of Air
Pollutant Emissions from Residential Heating Systems." EPA-
650/2-74-003, January 1974.
13. Dickerson, R. A. and A. S. Okuda, "Design of an Optimum Dis-
tillate Oil Burner for Control of Pollutant Emissions." EPA-
650/2-74-047, June 1974.
14. Shoffstall, D. R. and D. H. Larson, "Aerodynamic Control of
Nitrogen Oxides and Other Pollutants from Fossil Fuel Combus-
tion." EPA-650/2-73-033a, October 1973;
15. Brown, R. A., H. B. Mason, and R. J. Schreiber, "Systems Analysis
Requirements for Nitrogen Oxide Control of Stationary Sources."
EPA-650/2-74-091, September 1974.
16. McGowin, C. R., "Stationary Internal Combustion Engines in the
United States." EPA-R2-73-210, April 1973.
17. Bartok, W., et al., "Systematic Field Study of NOX Emission Con-
trol Methods for Utility Boilers." Esso Research and Engineering
Co., Linden, New Jersey. Report GRU.4GNOS.71, prepared for
Office of Air Programs, Environmental Protection Agency, Research
Triangle Park, North Carolina, December 31, 1971.
18. Jain, L. K., E. L. Calvin, and R. L. Looper, "State of the Art
for Controlling NOX Emissions, Part I: Utility Boilers." Final
Report, Catalytic, Inc., EPA-R2-72-072a, September 1972.
19. Berkau, E. E. and D. G. Lachapelle, "Status of EPA's Combustion
Program for Control of Nitrogen Oxide Emissions from Stationary
Sources." Presented at the Southeast APCA Meeting, Raleigh, N.C.,
September 19, 1972.
20. Crawford, A. R. et al., "Field Testing: Application of Combustion
Modification to Control NOX Emissions from Utility Boilers."
EPA-650/2-74-066, June 1974.
21. Cato, G. A., et al., "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial
Boilers - Phase I. EPA-650/2-74-078a, October 1974.
22. Roessler, W. U. et al., "Assessment of the Applicability of
Automotive Emission Control Technology to Stationary Engines."
EPA-650/2-74-051 , July 1974.
76
-------
23. Durkee, K., E. A. Noble, Frank Collins and Dave Marsland, "Draft
of Standards Support Document for An Investigation of the Best
System of Emission Reduction for Stationary Gas Turbines." EPA,
Office of Air Quality Planning and Standards, Research Triangle
Park, N.C., August 1974.
24. Levy, A., et al., "A Field Investigation of Emissions from Fuel
Oil Combustion for Space Heating." Conducted by Battelle Columbus
Laboratories for the American Petroleum Institute. API Publica-
tion 4099, November 1, 1971.
25. Barrett, R. E., et al., "Field Investigation of Emissions from
Combustion Equipment for Space Heating." EPA-R2-73-084a, June 1973.
26. Blakeslee, C. E. and A. P. Selker, "Program for Reduction of NOX
from Tangential Coal-Fired Boilers, Phase I." EPA-650/2-73-005,
August 1973.
27. General Motors Corporation, "Response to Draft Standards for
Control of Air Pollution from Stationary Gas Turbines," March 21,
1973.
28. Grits', G. J., "Economic FactorsaifnjWater^Treatment," Ind. WaterrM
Eng., 8(8), 22 (1971).
29. San Diego Gas and Electric Co., Letter from W. A. Zitlau to
D. R. Goodwin, EPA/OAQPS, dated October 1972.
30. Tyco Laboratory Final Report, "Development of the Catalytic Chamber
Process", EPA-R2-72-038 dated September 1972.
31. Esso Research Reports, "Development of the Aqueous Processes for
Removing NOX from Flue Gases" EPA-R2-72-051, dated September 1972
and Addendum, EPA-R2-73-051a dated June 1973.
77
-------
Conversion Table
To convert from
To
Multiply by
Btu
Btu
Btu/ft3
Cubic feet
Horsepower
Lbs/106 Btu
Mile
Pounds
Tons (short)
Calories, gram
Kilocalories, gram
Calories, kg/m3
Cubic meters
Horsepower (metric)
Grams/106 calories
Kilometers
Grams
Kilograms
251.98
0.25
8.899
0.0283
1.0139
1.80
1.6094
453.5924
907.19
78
------- |