i CONTROL OF NO EMISSIONS
I—x
for
Presentation to EPA
Region VIII
EPA Region ' \ LIBRARY by
Denver, Colorado
David G. Lachapelle
J. David Mobley
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
and
Robert Statnick
Office of Energy, Minerals, and Industry
Office of Research and Development
Washington, D.C. 20460
-------
CONTROL OF NO EMISSIONS
x
for
Presentation to EPA
Region VIII
by EPA Region VIII LIBRARY
Denver, Colorado
David. G. Lachapella
J. David Mobley
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
and
Robert Statnick
Office of Energy, Minerals, and Industry
Office of Research and Development
Washington, D.C. 20460
-------
TABLE OF CONTENTS
Section Number Page
1 Control of N0X Emissions 1-1
2 Combustion Modification 2-1
3 Assessment of Exxon's Thermal DeNOx Process 3-1
4 Assessment of N0X Flue Gas Treatment
Technology
5 Pilot - Scale Evaluation of Treatment Firing 5-1
6 Assessment Summary for Control of NOx
Emissions
iii
-------
INTRODUCTORY DISCUSSION
FOR CONTROL OF NOx
EMISSIONS
FOR
PRESENTATION TO EPA
REGION VIII
BY
ROBERT M. STATNICK
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
SECTION 1
INTRODUCTORY DISCUSSION FOR CONTROL OF NOx EMISSIONS
BACKGROUND
The purpose of this presentation is to provide EPA Region VIII
with an up-to-date assessment of the techniques that might be adapted
to minimize NOx emissions from utility plants using Western low
sulfur coal as fuel in Combustion Engineering (CE) tangentially—fired
boilers. The boilers were designed to comply with the 1971 EPA
regulation specifying a maximum N0X emission of 0.7 lb N0X/10®
Btu.
The subsequent discussions in this report pertain to both
existing and new CE tangentially-fired boilers; therefore it should
be noted that since 1970, all new tangentially-fired boilers have
been constructed with the necessary equipment for staged combustion,
namely, overfire air (OFA) addition. Older units of the same type
need to be retrofitted for overfire air operation.
Another factor that should be kept in mind throughout the fol-
lowing discussions is that boiler design has a major influence on the
baseline emissions level monitored prior to any applied NOx
reduction technologies. As shown in Figure 1, tangentially-fired
units have the lowest N0X emissions of all coal-fired utility
boilers.^
1-1
-------
Figure 1
Baseline Emissions
Coal Fired Utility
Boilers (1)
1600
1400
i
Wet Botto
m
1200
1000
Dry Bottom t
- Lte
600
400
ontally Opposed
^J^Tangential^xv
200
o L
0
_J I I—
200 400 600
MW
i
800 1000
-------
The longer gas residence times, uniform radiant heat transfer,
and ability to establish long diffusion flames tend to minimize N0X
formation from both thermal nitrogen fixation and fuel nitrogen
conversion. Therefore, it is not inconsistent that the existing CE
tangentially-fired boilers operate below 450 to 525 ppm (0.6 to 0.7
lb NOX/10*> Btu) baseline levels.
FORMATION OF N0X
Since all of the N0X control technologies in current commer-
cial use in U.S. coal fired utility plants are designed to minimize
N0X formation in the boiler, it is appropriate to devote a short
discussion to the formation mechanisms and factors involved.
The N0X generated by combustion of coal occurs through two
mechanisms. One mechanism is the oxidation of atmospheric nitrogen
(thermal N0X), while the other mechanism involves the conversion of
fuel-bound nitrogen (fuel-NOx). The two oxides of nitrogen that
are formed in this way are nitric oxide (NO) and nitrogen dioxide
(NO2). The NO is predominant and accounts for 90 to 95 percent of
the total N0X generated in a utility boiler. Once it enters the
atmosphere, NO is converted to NO2.
Formation of thermal N0X is known to be dependent on flame
temperatures, oxygen concentration in the combustion zone, and resi'
dence time at temperature. Fuel N0X, which constitutes a signifi-
cant portion of the total N0X, is relatively insensitive to
temperature compared to thermal N0X, and the formation is markedly
1-3
-------
affected by oxygen concentration in the combustion zone (See Figure
2). Fuel nitrogen conversion to NOx accounts for about 50 to 80
percent of total NOx in coal-fired boilers.
CONTROL/TREATMENT TECHNOLOGIES
Technologies for reducing NOx emissions from utility boiler
combustions may be divided into two categories; combustion modifica-
tions and flue gas treatment. These are pertinent to all boilers and
all coals.
COMBUSITON MODIFICATIONS
Demonstrated combustion controls pertinent to tangential-fired
boilers burning western coal include low excess air and overfire air
operation. Other technologies, such as the "rich fireball" concept,
are in the research and development stage.
LOW EXCESS AIR
Reducing the excess air level in the furnace has generally been
found to be an effective method of NOx control for all fuels (see
Figure 2). In this technique, the combustion air is reduced to the
minimum amount required to insure complete combustion, and to control
furnace slagging, reduce carbon carry over and maintain steam temper-
ature. With less oxygen available in the flame zone, both thermal
and fuel N0X formation are reduced. In addition, the reduced
airflow reduces sensible heat s-tack losses, resulting in an improve-
ment in overall boiler efficiency. Low excess air firing requires
minimal operational changes and, therefore, is the most convenient
control method to implement.
1-4
-------
Figure 2
N0X vs Theoretical Air (2)
Theoretical Air to Fuel Firing Zone, Percent
*NSPS— New Source Performance Standard
-------
OVERFIRE AIR
Staged combustion through overfire air seeks to control N0X by
carrying out initial combustion in a primary, fuel-rich, combustion
zone, then completing combustion, at lower temperatures, in a second,
fuel-lean zone. In tangentially-fired boilers, the second-stage air
is typically 15 to 20 percent of the total combustion air flow and is
introduced above the top burner row.
OTHER COMBUSTION MODIFICATIONS
There are various combustion modifications which are currently
in the research or development stage, including burner configuration
changes and the "rich fireball" concept.
Pilot-scale testing of new approaches to tangential firing is
currently underway on the EPA multi-burner furnace at Acurex
Corporation. The emphasis is on developing a low NOx tangential
firing system by controlling the fuel/air mixing characteristics of
the system. A number of configurations have been tested using a
matrix-type burner that permits placement of air and fuel nozzles in
different positions.
The most promising configuration tested to date has been one in
which a high percentage (60-80 percent) of the total combustion air
is directed parallel to the furnace wall while the fuel jet is
maintained at its normal yaw setting of 6 degrees from the furnace
diagonal. Thus, fuel and air diverge by about 39 degrees. This
results in an initially rich fireball with a subsequent reduced rate
1-6
-------
of £uel/air contacting. N0X emissions as low as 0.25 pounds per
million Btu have been achieved with this rich fireball approach.
Further optimization of this concept is continuing. Additionally,
tests are planned to evaluate this concept in conjunction with
overfire air to determine if further N0X reductions are achievable.
Although these technologies show promise for future N0X
reduction, they are several years away from commercialization.
POSTCOMBUSITON TREATMENTS
Although the degree of N0X reduction achievable in a full-
scale, pulverized coal fired power plant by combustion modifications
has not yet been established, it can be predicted that the ultimate
NOx reduction may be limited to the 60 to 80 percent range by these
methods.This still leaves a N0X content of 0.2 to 0.3 lb.
N0X/10^ Btu in the flue gas discharged. On the other hand, re-
ductions of up to 90 percent have been achieved in pilot plant
research efforts using postcombustion treatments. Therefore, the
postcombustion NOx treatment technologies are of particular
interest for achieving future EPA target emission levels.
Many postcombustion processes have been developed through the
research and pilot plant levels. These postcombustion prcesses can
be initially separated into two types -- wet or dry (see Figure 3).
In the wet processes, N0X is absorbed into an aqueous solution and
is subsequently converted to molecular N2, reduced N compounds, or
1-7
-------
-------
nitrite/nitrate. Moat of the wet scrubbing processes involve simul-
taneous removal of SO2 and N0X for economic, reasons. Because of
a number of problems, including the complex process chemistry and re-
quirements for a high L/G and a minimum S02/N0X ratio, and
liquid/solid waste disposal, wet processes are far less advanced than
dry processes.
Dry processes can be subdivided into four major categories:
catalytic reduction, noncatalytic reduction, adsorption, and radia-
tion. Adsorption and radiation processes are excluded in this pres-
entation because at present they appear comparatively less promising
than other competing processes.
Catalytic or noncatalytic reduction processes can be further
broken down into selective or nonselective processes depending on the
nature of the reducing agent used. Nonselective processes reducing
agents, such as CO or hydrocarbon fuels, reduce N0X and other
constituents at the same time, which may not be desirable for the
flue gas from a coal-fired power plant. If ammonia is used as a
reducing gas, N0X is generally reduced selectively to N2.
The following four presentations discuss combustion modification
and flue gas treatment technologies in detail. After these presenta-
tions, an attempt will be made to put these individual technologies
and their economic impacts into an overall perspective as it could
apply to any existing plant that might require consideration of its
N0X emissions and controls.
1-9
-------
REFERENCES
1. Thompson, R.E., Nitric Oxide Controls for Coal Fired Utilty
Boilers from an Applications Viewpoint, Proceedings: Second
NO-, Control Technology Seminar, Denver, CO, November 8-9, 1978,
EFKT FP-1109-SR, July, 1979.
2. Marshall, J.J. and Selker, A.P., The Role of Tangential Firing
and Fuel Properties in Obtaining Low N0X Operation for
Coal-fired Steam Generation, Proceedings: Second N0V Control
Technology Seminar, EPRI FP-1109-SR, July 1979. ~
3. Shah, N.D., EPRI's Program in the Postcombusiton N0X Control
Area, Proceedings: Second NO^ Control Technology Seminar, EPRI
FP-1109-SR, July 1979. ~
1-10
-------
SECTION 2
COMBUSTION MODIFICATION
FOR NOx CONTROL
FOR
PRESENTATION TO EPA REGION VIII
by
DAVID G. LACHAPELLE
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, NC 27711
-------
SECTION 2
COMBUSTION MODIFICATION
INTRODUCTION
Tangential firing is a unique fuel firing method utilized by
Combustion Engineering, Inc. (CE) since 1927 in its utility steam
generator design. Figure 1 shows a plan view of the characteristic
flame pattern that occurs in a tangentially-fired boiler. This
firing method has several advantages, especially in pulverized coal
systems, including:
• ability to fire a wide range of coal ranks
• uniform furnace heat absorption rates
• thorough mixing and long residence time of fuel and air
• low carbon loss
• ability to closely control superheat temperature by tilting
burners
• acceptable slag control over a relatively wide range of oper-
ating conditions
• wide turndown ratios
• characteristically lower N0X emissions than most wall-fired
designs
Figure 2 shows a unit side elevation of Columbia No. 1 which is
operated by Wisconsin Power and Light Company. This is a subbitumi-
nous coal-fired unit designed to meet the 1971 Utility Boiler NSPS
N0X emission limit of 0.7 lb. N0X/106 Btu heat input. This
unit uses an overfire air (OFA) system to limit N0X emissions.
2-1
-------
FIGURE 1
TANGENTIAL FLAME PATTERN
VIEWED FROM TOP OF FURNACE
2-2
-------
Figure 2
Unit Side Elevation,
Wisconsin Power and
Light Company,
Columbia Energy
£ Center No. 1
-------
It is one of several tangentially-fired units tested in EPA-sponsored
programs to determine the limits of the OFA N0X reduction
technique.
Combustion Engineering, Inc. has been utilizing staged combus-
tion to assess the effectiveness of technology to control N0X
emissions since promulgation of the 1971 N0X NSPS. This is
accomplished by means of an OFA system. Figure 3 shows a windbox
assembly of a modern CE pulverized coal-fired unit equipped with
overfire air. The OFA system permits the introduction of about 20
percent of the full-load combustion air requirements above the top
level of fuel nozzles. The overfire air enters the furnace tangen-
tially through two compartments located in the extended windbox in
each furnace corner. Control dampers regulate the overfire air rate.
The angle'of entry (up to +30°) of the OFA is controlled by a tilting
mechanism which is independent of the fuel and secondary air tilt
controller. In CE units, fuel and secondary air can also be tilted
up to +30° and are primarily used for superheat temperature control.
Figure 4 shows how the OFA system was retrofitted on the Barry No. 2
unit of Alabama Power Company as part of an EPA-sponsored study to
determine the limits of N0X control achievable using OFA. In this
case, the OFA ports were offset from the corners because of struc-
tural and pressure part considerations.
Staged combustion is generally used in conjunction with low (or
reduced) excess air firing. The effectiveness of these combusti
:ion
2-4
-------
Figure 3
Tangential Firing
System
Incorporating
Overfire Air
for NOx Control
Coal Firing
OVERFIRE AIR
NOZZLES
Oil GUN
SIDE IGNIJOR
NOZZLE
SECONDARY
AIR NOZZLES
COAL NOZZLES
-------
Figure 4
Schematic Overfire
Air System,
Barry Station No. 2
K>
I
0
1
F- FUEL AND AIR
A- AIR
O-OVERFIRE AIR
-------
modification techniques results from the lowering of oxygen availa-
bility in the fuel firing zone and, thus, both fuel and thermal N0X
are reduced. The balance of the excess air requirements are provided
above the fuel firing zone to promote carbon burnout and minimize
emissions of hydrocarbons and smoke. When using these techniques,
the lowest practical levels of staging are generally dictated by the
need to limit these products of incomplete combustion or to prevent
operating problems such as boiler vibrations, furnace wall slagging,
and fireside corrosion. In general, staged combustion plus low
excess air firing is capable of achieving 30 to 50 percent NCU
reductions from uncontrolled levels on coal-fired units.
2-7
-------
NO-y EMISSION DATA
Data on NOx emissions have been reported for several
tangentially coal-fired boilers. Figure 5 shows measured N0X
emissions for a range of tangential furnace sizes firing bituminous
coals. The EPA standard shown is the 1971 standard of 0.7 lb NO2/
106 Btu. These units were pre-NSPS designs and did not have OFA
systems. Despite this, nearly all these units were capable of
meeting a standard of 0.7. Similarly, Figure 6 shows measured N0X
emissions for a range of tangential furnace sizes firing
subbituminous coals. Again, the 1971 standard is shown. These units
were also pre-NSPS and did not have OFA systems and were still
capable of operating at or below the 0.7 limit.
More recently, NOx emissions data have become available on a
number of tangentially coal-fired boilers equipped with OFA systems.
The bulk of data, as shown in Figure 7, were generated during
EPA—sponsored programs of the Combustion Research Branch of IERL—RTP.
The EPA standard shown is still the 1971 standard since these units
were designed for that emission limit. (The 130—MW unit was the
Barry No. 2 unit retrofitted with an OFA system.) Note that for the
case of OFA operation, these N0X emission levels are at or below
the 1979 NSPS of 0.6 lb NO2/106 Btu for bituminous coal and 0.5
for subbituminous coal.
2-8
-------
Figure 5
Measured NOx Emissions
vs.
Furnace Rating-
Bituminous Coal-Fired Units with
C-E Tangential Firing
1971 EPA STANDAR^-^
NORMAL OPERATION
20 30% EXCESS AIR
NO OVERFIRE AIR
s
o
N
to
CM
3
8 § S
«- CM CM
O O ©
8 8 8
o
fv
n
© o
«- (M
I
o
in
s
FURNACE RATING. MEGAWATTS
-------
The units firing bituminous coal are:
o Barry No. 2 (Alabama Power Co.), 130 MW, eastern bituminous
o Huntington Canyon No. 2 (Utah Power and Light Co.), 430 MW,
western bituminous
Those firing subbituminous coal are:
o Hayden No. 2 (Lower Colorado River Authority), 280 MW
o Comanche No. 1 (Colorado Public Service Co.), 360 MW
o Columbia No. 1 (Wisconsin Power and Light Co.), 525 MW
o Navaho No. 2 (Salt River Project), 800 MW
It should be noted that the NOx emission data are based on
full-load short term tests (typically 1 to 4 hour duration), under
optimized furnace operation and do not reflect adverse conditions
such as variable coal quality or heavy slagging.
There are several important factors that must be appreciated
when reviewing N0X emission test results. First, ash fusion tem-
perature and resultant furnace wall slagging bear very heavily on how
a boiler must be operated if load requirements are to be met.
Second, the method available for slag control is to increase excess
air in the furnace firing zone; however, this causes N0X to
increase. Third, the data shown are for optimized operating condi-
tions which include relatively clean and nonslagging furnace walls.
This is possible for short-term tests. The real-life situation is
somewhat different under routine 0FA operation. For example, furnace
walls at times slagged heavily at Huntington and Columbia. When this
occurred, the operator would increase excess air to the furnace
2-10
-------
Figure 6
Measured NOx Emissions
vs.
Furnace Rating—
Subbituminous Coal-Fired Units with
C-E Tangential Firing
N>
I
3
H
03
CM
P>
O
rv
m
o
8
FURNACE RATING, MEGAWATTS
-------
Figure 7
Measured N0X Emissions
vs.
Furnace Rating
of Tangentially Fired Coal Units
Equipped with Overfire Air
FURNACE RATING, MEGAWATTS
-------
firing zone to shed slag. During these episodes, N0X emissions
could increase by about 20 to 50 percent from the optimized values
shown. Generally, however, 30-day averages of 0.5 and 0.6 for
subbituminous and bituminous coals, respectively, appear to be
achievable. These types of considerations were factored into the
1979 NSPS, and emission limits for new units are to be based on 30-
day rolling averages.
The Federal Register (Vol. 44, No. 113, June 11, 1979) cites
data submitted on the Colstrip No* 1 and 2 units. It states that
these units do not consistently achieve the current regulatory level
of 0.5 on a 24-hour basis. These units, however, were designed to
meet a 0.7 standard. However, on a 30-day rolling average basis, it
appears that a limit of 0.5 would be achievable.
Colstrip units 3 and 4 are also designed to meet the 0.7 stan-
dard. We are not aware of any design differences that would preclude
these units from attaining the same emission limits as units 1 and 2.
However, there is also no evidence currently available that would
indicate that these units could consistently achieve N0X levels
significantly below 0.5 lb NO2/I06 Btu.
2-13
-------
SECTION 3
ASSESSMENT OF EXXON'S
THERMAL DeNOx PROCESS
FOR
PRESENTATION TO
EPA REGION VIII
BY
DAVID G. LACHAPELLE
COMBUSTION*RESEARCH BRANCH
ENERGY ASSESSMENT AND CONTROL DIVISION
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
BACKGROUND AND EXPERIENCE
The Exxon proprietary Thermal DeNOx Process was discovered in
August 1972 by Dr. Richard K. Lyon of Exxon Research and Engineering
Co. U.S. Patent No. 3,900,554 was issued on August 19, 1975.
The Process is based on the noncatalytic homogeneous selective
gas-phase reaction of nitric oxide with ammonia. Thermal DeNOx has
been commercially demonstrated in gas- and oil-fired boilers, process
furnaces and on a municipal incinerator. Use of the Process with
coal-firing has been performed only on a pilot scale.
Since 1972, a number of laboratory, pilot, and full-scale tests
have been conducted with the Thermal DeNOx Process. The tests have
been structured to get a better understanding of the critical process
parameters that affect performance. Essentially all of the tests
have been done by Exxon. The bulk of the tests have been performed
on gas- and oil-fired systems except for the full-scale test on a
solid-waste incinerator. Recently, KVB, Inc., under contract to
Exxon and the Electric Power Research Institute studied the use of
Thermal DeNOx on a 3 x 10^ Btu/hr coal-fired boiler. The most
current studies have been sponsored by EPA and were designed to
develop cost estimates and evaluate potential environmental
limitations/concerns with the Process. Final reports from these
EPA-sponsored studies are now available.
3-1
-------
The Thermal DeNOx Process proceeds by a complex homogeneous
noncatalytic gas-phase reaction of NH3 with NO in the presence of
02. The reaction can be represented globally by
4N0 + 4NH3 + 02 > 4N2 + 6H20
(1)
However, the reaction is in competition with
4NH3 + 502 > 4N0 + 6H20
(2)
These reactions are sensitive to a number of parameters, the most
important of which is temperature. For example, in typical flue gas
environments, reaction (1) dominates at temperatures of about 1740°F
(950°C). At higher temperatures, reaction (2) becomes significant
and dominates at temperatures above about 1900°F (1Q38°C). As tem-
peratures are reduced below about 1650°F (900°C), the rates of both
reactions slow, the NO reduction falls off, and NH3 flows through
unreacted. Consequently, the Thermal DeNOx Process must be
operated over a relatively narrow temperature "window" if maximum NO
reductions are to be achieved. Performance is also sensitive to the
initial NO concentration, NH3/NO ratio, excess oxygen, available
residence time at reaction temperature, and mixing efficiency. Also,
the use of H2 injection along with NH3 can lower the optimum re-
action temperature "window." The magnitude of this shift is related
to the amount of H2 injected relative to NH3. For example, at
H2/NH3 ratios of about 2:1, the NO reduction reaction can be
forced to proceed rapidly at 1290°F (700°C).
Figures 1 to 5 illustrate some of these more important factors
affecting performance of the Thermal DeNOx Process.
3-2
-------
EFFECT OF FLUE GAS TEMPERATURE ON THERMAL
DeNOx PERFORMANCE (REFERENCE 1)
1.0
0.8
2 0.6
"c
o~
2
73
c
o"
z
0.2
0
700 800 900 1000 1100
TEMPERATURE, °C
Figure 1. Effect of flue gas temperature on Thermal DeNO^
performance (Reference 1).
3-3
-------
EFFECT OF NH3 INJECTION RATE ON NO EMISSIONS
(REFERENCE 1)
(NH3)/(NO)
Figure 2. Effect of NHg injection rate on NO emissions
(Reference 1).
3-4
-------
EFFECT OF INITIAL NITRIC OXIDE CONCENTRATIONS
ON REDUCTIONS WITH AMMONIA INJECTION
(REFERENCE 1)
1.0
0.8
•f 0.6
C
o
2
*3
| 0.4
O
z
0.2
0
0 1 2 3 4 5
(NH3)/(NO)
Figure 3. Effect of initial nitric oxide concentrations on
reductions with ammonia injection (Reference 1).
3-5
-------
EFFECT OF NH3 INJECTION RATE ON NH3 CARRYOVER
EMISSIONS (REFERENCE 1)
(NH3)/(NO)
Figure 4. Effect of NH3 injection rate on NH3 carryover
emissions (Reference 1).
3-6
-------
THERMAL DeNOx REACTION PRODUCTS AS FUNCTIONS
OF TEMPERATURE WITH AND WITHOUT HYDROGEN
INJECTION (REFERENCE 2)
FLUE GAS TEMPERATURE, °C
Figure 5. Thermal DeNOx reaction products as functions of
temperature with and without hydrogen injection
(Reference 2).
-------
The Thermal DeNOx Process may form by-product pollutants
directly or indirectly from the presence of NH3 in the combustion
gas. Potential by-product emissions suggested by Exxon include
NH3, CO, HCN, and N2O. Also, when a sulfur-containing fuel is
burned, NH3 and SO3 combine with water vapor to form ammonium
bisulfate, NH4HSO4.
Ammonium bisulfate is a viscous liquid from 147°C to about 450°C
(300°-840°F). It has been known to cause corrosion of metal
surfaces. Thus far, however, no increase in metal corrosion
attributable to ammonium bisulfate has been identified with the
Thermal DeNOx Process. We are aware of some problems with air
heater plugging in Japan due to bisulfate deposition. At one site,
plugging occurred within 4 months when DeNOx rates over 40 percent
were attempted. Daily soot blowing and water washing twice a year,
combined with restricting DeNOx rates to about 40 percent, have
made the problem manageable. The formation of NH4HSO4 can be
controlled by limiting the NH3 carryover. This can be accomplished
by NH3 injection at a slightly higher than optimum temperature or
by using H2 additive. In general, unreacted NH3 and ammonium
bisulfate are considered the most serious byproducts and the ones
that could most adversely affect the use of the Thermal DeNOx
Process.
Carbon monoxide emissions may also be promoted by ammonia injec-
tion because the Thermal DeNOx reaction inhibits the oxidation of
CO to C02« Thus, if there is unburned CO at the point of NH3
3-8
-------
injection, the CO tnay not be oxidized, but will be discharged to the
atmosphere. Under normal operating conditions, CO levels are not
usually significant in steam generators. Using hydrocarbons as addi-
tives to control the NH3-NO-O2 reaction increases the concentra-
tion of CO in the flue gas. Exxon reported that as much as 50
percent of the hydrocarbons may be oxidized to CO. This CO may then
be emitted to the atmosphere because the ammonia inhibits the O2 +
CO CO2 reaction.
HCN is found only if hydrocarbons are present in the region in
which NH3 is injected. Under normal boiler operation, gaseous
hydrocarbons are not present unless they are injected along with the
NH3. KVB Inc. reported that for gas, oil and coal firing, HCN was
present in the untreated flue gas at 3 to 10 ppm concentration,
depending oh excess air levels. Injection of NH3 did not
measurably affect the HCN level.
The reaction of NO with NH3 and O2 forms N2O as a minor
by-product. However, less than 2 moles are generated for every 100
moles of NO reduced, according to Exxon experimental data* All the
available evidence indicates that N2O is relatively harmless at
those levels, and does not represent an environmental concern.
Thermal DeNOx has been demonstrated on twelve boilers or fur-
naces to date. The bulk of these have been in Japan where N0X
standards are more stringent than in the U.S. The capability of the
Process frequently represents a compromise between the limitation of
3-9
-------
the Process chemistry and cost-effectiveness. Usually performance is
maximized for full-load operation, and smaller N0X reductions are
accepted at reduced loads where reaction zone temperatures are lower.
Total N0X emissions are generally at target levels over the full
range of operating conditions, since N0X emissions are usually
lower at reduced load. Results from six of these demonstrations are
shown in Figure 6. Most of the data shown below about 900°C were pre-
sumably obtained using H2 along with NH3.
Recently, KVB, Inc. completed a pilot-scale investigation of the
use of Thermal DeNOx with coal combustion (4). The objective of
the investigation was to determine the degree of N0X reduction
achievable in flue gas resulting from coal combustion., The primary
parameters evaluated were injection temperature and location,
NH3/NO ratio, hydrogen addition and coal type (three bituminous and
one subbituminous). Additionally, by-product emissions were measured
at various NH3 injection rates.
The combustion facility consisted of a 0.9 MTt (3 x 10^
Btu/hr) firetube boiler equipped with a ring-type natural gas burner
and a geometrically scaled version of a utility-type coal burner.
Combustion air preheat of 600°F was used. The NH3 injection system
consisted of five movable injectors located at the end of the
firetube section distributing the NH3 plus N2 carrier gas
counterflow to the flue gas stream. The injectors could be
positioned axially along the firebox length to evaluate the effect of
the temperature at the plane of injection.
3-10
-------
THERMAL DeNOx SYSTEM PERFORMANCE ON COMMERCIAL
UNITS AS FUNCTIONS OF TEMPERATURE
(REFERENCE 3)
70
60
50
g 40
cc
Ui
&
z
o
p 30
u
o
Q
Ui
ax 20
O
10
0
700 800 900 1000
FLUE GAS TEMPERATURE, °C
A
A-
o «
a
%
o
~
SIZE DESCRIPTION
^ 25 t/hr Package Boiler
• 120 t/hr ^ Industrial Boiler
D 100 MWei _ ..
¦ 100MWe } Ut,l,tv Bo,ler
A 150 kbbl/d , u-fltPM
~ 150 kbbl/d * Heaters
L I
Figure 6. Thermal DeNOxsystem performance on commercial
units as functions of temperature (Reference 3).
3-11
-------
The major results to be discussed will center on:
• reaction temperature
• ammonia injection rate
o hydrogen addition and
o byproduct emissions
Reaction Temperature
As discussed previously, reaction temperature is the primary
variable affecting performance of the Thermal DeNOx Process.
Figure 7 shows this effect for natural gas and the four coals tested.
Since temperature gradients as large as 1208C (216°F) occurred radi-
ally on any one plane, an average radial temperature was determined.
Note that the maximum NO reduction varied from 940° to 1000°C (1724°
to 1832°F). However, excluding the Illinois coal, maximum NO
reductions occured within a narrower range of about 940° to 970°C
(1724° to 1778°F). It was initially speculated that the higher
sulfur content of the Illinois coal might have caused this
temperature shift. Consequently, a special test series was conducted
in which a sulfur compound was injected into a flue gas generated
from distillate oil combustion. These results given in Figure 8 show
that sulfur dioxide has essentially no effect on the reaction
temperature for nunr-i'mnm NO reduction. Exxon feels that the
difference is attributable to artifacts in the average temperature
readings obtained with the Illinois coal. Exxon is conducting
laboratory tests to confirm this theory.
3-12
-------
EFFECT OF TEMPERATURE ON NO REDUCTIONS
COAL AND NATURAL GAS FIRING
(REFERENCE 4)
(NH3/N00 * 1.0, EXCESS 02~5.0%)
NATURAL GAS
— — UTAH COAL
NAVAHOCOAL
PITTSBURGH COAL
ILLINOIS COAL
815 870 925 980 1035 1090
AVERAGE RADIAL TEMPERATURE, °C
Figure 7. Effect of temperature on NO reductions, coal
and natural gas firing (Reference 4).
/
//
./
//
3-13
-------
EFFECT OF SULFUR ON NO REDUCTION, OIL FIRING
(REFERENCE 4)
1.0
EXCESS OXYGEN ~ 5%
INITIAL NO(NOq): 350 ppm
NH3/NOo 1
0.8
0.6
I
N
n &
Jv. A
0.4
^—'8
O SO2 = 2900 ppm
0.2
~ SO2 3 1100 ppm
A SO2 = 120 ppm
»
I l l
I I
0.25 0.5 0.75 1.0 1.25
BACK PLANE OF NhU INJECTION, METERS FROM BACK FURNACE WALL
WALL J
815 870 925 980
APPROXIMATE AVERAGE TEMPERATURE, °C
Figure 8. Effect of sulfur on NO reduction, oil firing
(Reference 4).
3-14
-------
Ammonia Injection Rate
KVB Inc. conducted a number of tests to evaluate the effect of
various molar ratios of NH3/NO for natural gas and each of the four
coals. As shown in Figure 9, NO emissions decrease as NH3/NO
ratios were increased over the range of injection temperatures
studied. Generally, 65 percent NO reductions were achieved at
equimolar concentrations of NH3 to NO. A comparison of NO
reductions at the optimum temperature condition for each fuel
indicates that NH3/NO molar ratios of 1.0 to 1.5 gave NO reductions
of 65 to 75 percent. Higher molar ratios yielded only minimal
further reductions in NO.
Hydrogen Addition
KVB conducted a limited number of tests with combined NH3-H2
injection. These were done on the Pittsburgh No. 8 coal. This use
of H2 addition produced higher NO reduction at temperatures lower
than the optimum 950°C (1742°F). Maximum DeNOx rates were main-
tained over a temperature range of 750°G to 950°C (1400° to 1742°F)
by injecting H2 at rates ranging from 0.2 to 0.94 measured as the
molar ratio of Hj to NH3. The injection of Hj also contributed
to lower NH3 breakthrough. KVB also found that at high hydrogen in-
jection rates the NO levels began to increase, while the NH3 levels
in the combustion products continued to decrease.
By-product Emissions
Ammonia emissions are an important consideration when evaluating
the Thermal DeNOx Process for coal combustion* Ammonia emissions
3-15
-------
COMPARISON OF NO REDUCTIONS AT THE OPTIMUM
TEMPERATURE CONDITION (REFERENCE 4)
u>
On
1.0
0.8 —
o 0.6 —
O
z
o
z 0.4 —
0.2 —
0
O NATURAL GAS
O UTAH COAL
A NAVAHO COAL
• ILLINOIS COAL
¦ PITTSBURGH COAL
0
0.5
1.0
1.5
2.0
(nh3)/(no0)
Figure 9. Comparison of NO reductions at the optimum
temperature condition (Reference 4).
-------
are also related to the amount of NH^HSO^ which may be formed by
the NH3 + SO3 + H20 reaction. Coal type and excess air levels
have a bearing on the amount of SO3 formed. Subbituminous coals,
such as the Navaho, contain calcium, and yield strongly basic flyash
which tends to retain SO3. In fact, KVB, Inc. reported only about
5 ppm SO3 with this coal compared to the bituminous coals like
Illinois and Pittsburgh which contained up to 21 ppm SO3. Thus,
the high sulfur, low alkaline coals may be more prone to form
NH4HSO4. Ammonia emissions are tied in to initial NH3/NO
injection rates. For the coals tested (except Illinois), NH3
emissions began to increase at injection rates above 0.5 NH3/NO
(Figure 10). No explanation was given for the lower NH3 emissions
with Illinois coal. However, the higher optimum injection temperature
could account for lower NH3 breakthrough. Hydrogen injection can
be used to control NH3 breakthrough. In fact, the KVB Inc. tests
showed that NH3 emissions could be decreased nearly 90 percent when
H2 was injected at a rate of 2.0 (measured as molar H2/NH3).
The NH3 injection rate (measured as molar NH3/NO) was
approximately 1.5 during these tests. In general, NH3 emissions of
10 to 40 ppm seem to be characteristic, based on the commercial gas-
and oil-fired installations and these coal-fired tests.
Cyanide and nitrate emissions were measured by KVB, Inc. during
the tests on natural gas and coal. Emission levels averaged about 2
ppm for cyanide and about 10 ppm for nitrates at both baseline and
3-17
-------
COMPARISON OF THE NH3 EMISSIONS FOR ALL
FUELS TESTED AT THE PEAK NO REMOVAL
TEMPERATURE (REFERENCE 4)
-------
reduced N0X conditions. There appears to be no correlation between
these emissions and the amount of NH3 injected. Consequently,
cyanide and nitrate emissions are not considered to be a problem with
the Thermal DeNOx Process.
Mixed results were obtained with regard to sulfate emissions,
increasing on two tests and decreasing on two tests. No conclusions
could be drawn. Slight decreases in SO3 emissions were observed
during NH3 injection, which would indicate that the Process does
not increase SO3 emissions. If anything, SO3 would react with
NH3 and HjO to form NH4HSO4.
The CO emissions increased when NH3 was injected to reduce NO,
but this increase was limited to 50 ppm for Pittsburgh coal. The
results indicate that NH3 injection inhibits the oxidation of CO to
CO2. However, these CO levels are considered to be neither
environmentally significant nor detrimental to the efficiency of a
coal-fired utility boiler.
3-19
-------
APPLICABILITY ASSESSMENT AND COST ESTIMATES
Two EPA-sponsored paper studies have been conducted to assess
the applicability of the Thermal DeNOx Process and to develop cost
estimates for coal-fired utility boilers# These studies have been
performed by Acurex and Exxon. Final reports are available.
Eight typical coal-fired utility boilers representative of U.S.
boiler population were selected for study by Exxon (Table 1). These
boilers were chosen to permit an evaluation of the Thermal DeNOx
Process on different utility boiler types, sizes, firing methods and
coals. Thermal DeNOx performance and costs were determined for two
N0X reduction targets:
a. A trim case that meets the proposed New Source Performance
Standards (NSPS) of 0.5 lb NOX/10*> Btu (375 ppm @ 3%
O2) for boilers fired with subbituminous coal.
b. Deep reduction in N0X levels to 0.4 lb N0x/10^ Btu
(300 ppm) for boilers fired with bituminous coal and lignite
and 0.3 lb/10^ Btu (225 ppm) for subbituminous coal-fired
boilers.
Also considered was the:
c. Maximum practical reduction in N0X levels attainable with
the Thermal DeNOx Process.
Two initial N0X levels were considered for each of the above
targets: (1) uncontrolled and (2) initial N0X reduction using
combustion modification (CM). Each boiler was assumed to be equipped
with two ammonia injection grids to permit load following. In addi-
tion to these six cases, special analyses were performed for flue gas
temperature nonuniformity and the use of hydrogen with ammonia to
3-20
-------
TABLE I. BOILERS SELECTED FOR STUDY
BOILER
MANUFACTURER
FIRING METHOD
BOILER
SIZE, MW
COAL TYPE
BABCOCK
WILCOX
FRONT WALL
130
SUBBITUMINOUS
BABCOCK
WILCOX
HORIZONTALLY
333
BITUMINOUS
BABCOCK AND
WILCOX
CYCLONE
400
LIGNITE
COMBUSTION
ENGINEERING
TANGENTIAL
SINGLE FURNACE
350
BITUMINOUS
COMBUSTION
ENGINEERING
TANGENTIAL
800
SUBBITUMINOUS
FOSTER WHEELER
FRONT WALL
330
BITUMINOUS
FOSTER WHEELER
HORIZONTALLY
OPPOSED
670
SUBBITUMINOUS
RILEY STOKER
TURBO FURNACE
350
BITUMINOUS
3-21
-------
permit load following. Exxon utilized a proprietary Performance
Prediction Procedure to project Thermal DeNOx performance. Addi-
tionally, the Thermal DeNOx costs for reaching N0X levels of 0.3
to 0.4 lb/10^ Btu were compared to the costs of CM alone. Table 2
gives the predicted thermal DeNOx performance achievable at full
boiler load and 1.5 NH3/NO ratio without combustion modifications.
Tables 3 and 4 present the thermal DeNOx cost comparison summary
for Cases 3 and 6 of the Exxon study.
The major findings from the Acurex and Exxon studies (5,6) are:
• DeNOx performance was projected to be essentially equiva-
lent for the eight boilers evaluated.
• Significant differences in flue gas temperature and flow path
configuration will result in significantly different injec-
tion grid locations among the boilers of different manufac-
turers.
• Injector grid location would not be significantly affected by
assuming a 50°C larger temperature range in the injection
plane than that used in the Performance Prediction Procedure.
However, a temperature range increase of this size could re-
duce DeNOx performance by 5 to 10 percentage points.
• Flue gas temperatures gradients of 110°C (200°F) can often
occur across an injection plane even at constant heat input.
• A dual grid approach is desirable for load following.
• All units could reach the proposed N0X NSPS using Thermal
DeNOx alone. However, five of eight units could also reach
this level using CM alone.
• All units except one could meet the deep N0X reduction
target when Thermal DeNOx is used in combination with CM.
The one exception is the cyclone boiler fired with lignite.
3-22
-------
BOILER
Manufacturer
TABLE 2 - Predicted Thermal DeNo Performance Achievable
At Full Boiler Load anl 1.5 NH_/NO Ratio
without Conventional Modification ^
Type
Size, MW
COAL
Type
N0x (lb/10 BTU)
Initial Final
Level Level Achievable
**
**
%
Reduction Achievable
B&W
Front wall
130
Subbituminous
0.67
0.33
51
B&W
Horizontally
opposed
333
Bituminous
0.93
0.39
58
B&W
Cyclone
400
Lignite
1.33
0.57
57
CE
Tangential
Single furnace
350
Bituminous
0.67
0.28
58
CE
Tangential
800
Subbituminous
0.71
0.30
58
FW
Front wall
330
Bituminous
1.13
0.52
54
FW
Horizontally
Opposed
670
Subbituminous
0.93
0.37
60
RS
Turbo Furnace
350
Bituminous
0.93
0.39
58
Selected by Exxon from combination of internal data and data supplied by boiler manufacturers. (Reference)
Estimated by Exxon from combination of fundamentals and pilot plant and commercial scale experience (Reference -).
-------
TABLE 3 - THERMAL DeNO COST COMPARISON SUMMARY **
x
(Deep Reduction Target:
NH^ Injection Without Combustion Modification)
Initial Target Reagent Coat - aills/kW-hr Carrier Cost - milla/kU-hr On-Slte Coat Total Cost
Unit HO^ * NO^ * Operating Capital Total Operating Capital Total mllls/kW-hr milla/kW-hr
B&W 130 MW Subbltunlnoua 0.67 0.30 ______
333 Ml Bituminous 0.93 0.40 0.41 0.40 0.45 0.07 0.06 0.13 0.05 0.63
400 MW Lignite 1.33 0.40 ______
CE 350 MH Bituminous 0.67 0.40 0.17 0.03 0.20 0.07 0.06 0.13 0.05 0.38
800 MH Subbltuminoua 0.71 0.30 0.43 0.03 0.46 0.09 0.03 0.12 0.04 0.62
to
I
N>
¦O
fW 330 MW Bituminous 1.13 0.40 ______
670 MW Subbltuminous 0.93 0.30- -- - -
RS 350 MM Bituminous 0.93 0.40 0.58 0.05 0.63 0.09 0.15 0.15 0.05 0.83
* - lb. MO /106 Btu
** *
-Thermal DeMO^ costs do not include licensing fees and charges for preliminary engineering and testing.
-------
TABLE 4 - THERMAL DENO COST COMPARISON SUtttARY**
jc '
Maximum DeNO at NU-/NO /NO. «* 1.5 With Coabustlon Modification
x 3xi
Unit
Initial
NO*
Target
MO**
Reagent Cost - ailla/kW-hr
Operating Capital Total
Carrier Coat - nllla/kU-hr
Operating Capital Total
On-Slte Coat
allls/kU-hr
Coabustlon
Modification
Coabustlon
Modification Coat
Total Thermal
DeNO Cost
¦llls^kU-hr
0.55
Total Cost
B&U 130 MW Subbltualnoua
— x .
0.40
0.20
0.22
0.06
0.28
0.08
0.11
0.19
0.08
Technique
LNB
ailla/kW-hr
0.06
allls/kH-br
0.61
330 MJ Bltualnous
0.56
0.23
0.28
0.04
0. 32
0.07
0.06
0.13
0.05
LNB
0.06
0.50
0.56
400 W Lignite
1.20
0.52
0.86
0.07
0.92
0.10
0.06
0.16
0.05
LEA
0.0
1.13
1.13
CE 350 MU Bltualnous
0.60
0.25
0.31
0.04
0.35
0.07
0.06
0.13
0.05
OFA
0.08
0.53
0.61
800 MU Subbltualnous
0.50
0.21
0.31
0.02
0.33
0.09
0.03
0.12
0.04
OFA
0.08
0.49
0.57
FU 330 HU Bituminous
0.68
0.31
0.35
0.04
0.39
0.07
0.06
0.13
0.05
LNB
0.06
0.57
0.63
670 MU Subbltualnous
0.56
0.22
0.39
0.03
0.42
0.10
0.04
0.14
0.04
LNB
0.06
0.60
0.66
RS 3S0 MU Bltualnous
0.56
0.23
0.35
0.04
0.39
0.09
0.06
0.15
0.05
OFA
0.10
0.59
0.69
*
After combustion modification reductions
*lb NO /106 Btu
Thermal DeNO^ costs do not include licensing fees and charges for preliminary engineering and testing
-------
Exxon-developed cost: estimates averaged 9 percent lower than
those developed by Acurex and are probably more optimistic
than those by Acurex. Neither the Acurex nor the Exxon cost
estimates include Exxon's licensing fees which cannot be
disclosed except by Exxon.
Exxon—projected costs to reach the proposed NSPS from an
uncontrolled base level ranged from 0.25 mills/kWh for the
250 MW CE boiler to a high of 1.17 mills/kWh for the lignite
fired cyclone boiler. The average cost for all boilers
considered was 0.57 mills/kWh, or 0.49 mills/kWh not
including the cyclone boiler.
Four of the eight boilers could reach the deep reduction
target using Thermal DeNOx alone. Exxon's costs ranged
from 0.38 mills/kWh to 0.83 mills/kWh for these boilers.
Exxon—projected costs to reach the deep reduction target
using Thermal DeNOx with combustion modifications ranged
from 0.38 mills/kWh to 0.51 mills/kWh, with the average being
0.44 mills/kWh for the seven boilers reaching the target
level.
The N0X reductions ranging from 62 to 76 percent and
averaging 70 percent relative to an uncontrolled base case
could be achieved using Thermal DeNOx at a maximum
practical level in combination with combustion modifications,
Ihe NOx levels in the 0.20 to 0.23 lb N0X/10 Btu range
could be realized for most of the boilers. Exxon's costs
ranged from 0.55 to 1.14 mills/kWh and averaged 0.68
mills/kWh for all boilers studied. With the lignite boiler
excluded, the range was 0.55 to 0.67 mills/kWh and the
average was 0.61 mills/kWh.
The Exxon costs for onsites and the carrier were found to be
proportional to boiler size.
The total ammonia reagent costs for all cases, normalized for
the amount of N0X removed ( N0X) expressed as NO2, were
nearly equal for all eight units studied at $0.09/lb N0X.
This parameter was considered to be a good overall judgment
criterion of the chemical efficiency and economic efficacy of
the Thermal DeNOx Process.
The Exxon Thermal DeN0x Process was considered to be
equally amenable to all units studied.
3-26
-------
• The costs for reaching N0X levels in Che 0.3 to 0.4
lb/10** Btu range were compared for Thermal DeNOx and
combustion modifications. The costs of most conventional
combustion modifications and combinations thereof were lower
than that of Thermal DeNQx. Extreme N0X reduction
methods such as derating or staged combustion that incurred
derating would be more expensive.
• There still exist several operational and environmental areas
of concern with the Thermal DeNOx Process, especially for
coal-fired application. These include:
- Effect of coal type on reaction temperature
- Flue gas temperature fluctuations and load changes
- NH3 emissions and effect on corrosion, fouling, SO3
depletion and electrostatic precipitator performance
- NH3 and by-product emissions.
RECOMMENDATIONS
Thermal DeNOx should be used only as a supplement to combus-
tion modifications.
A full-scale demonstration on a coal-fired utility boiler would
provide the most realistic tests and permit an evaluation of all
areas of concern. Having selected a test boiler, detailed tempera-
ture and velocity profile measurements must be made in the convective
passages, at various loads, to verify grid placement and performance
estimates. A complete evaluation program would serve to verify
DeNOx performance, costs, operational problems and other side ef-
fects.
3-27
-------
REFERENCES
1. Muzio, L. J., Homogeneous Gas Phase Decomposition of <^des °J
Nitrogen. EPRI Report FP-253, National Technical Inform
Service PB 257 555, August 197 6.
2. Bartok, W., Noncatalytic Reduction of N0X with NH3, in
Proceedings of the Second Stationary Source. C?nlb"S"^r!^®
Volume II. EPA-600/7-77-073b, National Technical Information
Service PB 271 756 98E, July 1977.
3. Performance of the Thermal DeNOx Process in Commercial
Applications. Exxon's Sales Brochure. 19/ •
4. Muzio, L. J. et al., Noncatalytic NO Removal with Ammonia. EPRI
Final Report FP-735, Research Project 835-1. April iy/o.
5. Varga, G. M. et al., Applicability of the Thermal DeNOx Process
to Coal-Fired Utility Boilers. In Proceedings of the Third
Stationary Source Combustion Symposium: Volume I.
EPA-600/7-79-050a. February 1979.
6. Castaldini, C. et al., Technical Assessment of Exxon's Thermal
DeNOx Process. EPA-600/7-79-117. May 1979.
3-28
-------
SECTION 4
ASSESSMENT OF NOx FLUE
GAS TREATMENT TECHNOLOGY
for
Presentation to EPA Region VIII
by
J. David Mobley
Process Technology Branch
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
-------
ASSESSMENT OF NOx FLUE GAS TREATMENT TECHNOLOGY
INTRODUCTION
Nitrogen oxides (NOx) and sulfur oxides (S0X) in the
atmosphere have been determined to have adverse effects on human
health and welfare. To aid in preventing these adverse effects, the
Industrial Envirnmental Research Laboratory at Research Triangle
Park, N.C. (IERL-RTP) is leading the U.S. Enviornmental Protection
Agency's (EPA's) efforts to develop and demonstrate N0X and SOx
control technologies for stationary combusiton sources.
Flue gas desulfurizaton technology has progressed to commercial
application and has achieved 90% control of S02. Although NOx
control by combustion modification technology has been applied
commercially, NOx control by flue gas treatment technology has not
been utilized in the U.S. Combustion modification technology reduces
N0X emissions by approximately 50% in a relatively cost effective
manner. Flue gas treatment technology should be able to reduce NOx
emissions by 90% and has the potential for 90% control of both N0X
and S0X emissions.
EPA is proceeding with small scale N0X and N0x/S0x flue
gas treatment experimental projects in parallel with technology
assessment and control strategy studies. To save both development
time and money, EPA is investigating Japanese technology for
potential application to the U.S. coal-fired situation. Through
these actions, the basic foundation will be established if the
4-1
-------
technology is required in the U.S. and acceleration of the
development program becomes necessary.
NCU FLUE GAS TREATMENT TECHNIQUES
Due to stringent emission standards, Japanese technology for
control of NOx and simultaneous control of N0X and S0X by flue
gas treatment techniques is more advanced than any other country's.
An overview of current Japanese technology, both dry and wet
processes for N0X and NOx/SOx control, follows.
DRY NO.,- PROCESSES
Numerous dry process types are being developed and are being
applied at full scale. Selective catalytic reduction (SCR) processes
are the only ones that have achieved notable success in treating
combustion flue gas for removal of N0X and have progressed to the
point of being commercially applied. SCR processes are based on the
preferential reaction of ammonia (NH3) with N0X rather than other
flue gas constituents. Since oxygen (O2) enhances the reduction,
the reactions can best be expressed as:
4NH3 + 4N0 + 02 * 4N2 + 6H20 (1)
4NH3 + 2N02 + 02 >3N2 + 6H20 (2)
Reaction 1 predominates since approximately 95% of the N0X in
combustion flue gas is in the form of nitric oxide (NO). Therefore,
a stoichiometric amount of ammonia can be used to reduce N0X under
ideal conditions to harmless molecular nitrogen
4-2
-------
(N2) water vapor (H2O). An NHjtNO mole ratio o£ about 1:1
has typically reduced N0X emissions by 90% with a leak ammonia
concentration of less than 20 ppm.
The SCR processes are relatively simple, requiring a reactor, a
catalyst, and an ammonia storage and injection system. Due to
increased pressure drop across the SCR reactor some increase in
boiler fan capacity, or possibly an additional fan, may be necessary.
The optimum temperature for the reaction is about 1000°C
(1830°F). However, the catalyst effectively reduces the reaction
temperature to the 300® ~ 450°C (570-845°F) range. To obtain flue
gas temperatures in this range and to avoid the requirement for large
amounts of reheat, the reactor is usually located between the boiler
economizer and the air preheater. A typical flow diagram is shown in
Figure 1. Obviously, the reactor and catalyst are the critical
elements of the process.
Catalysts with vanadium compounds were found to be resistant to
attack from SO3 and promote the reduction of N0X with ammonia.
Titanium dioxide was found to be an acceptable carrier* Therefore,
many S0X resistant catalysts are based on Ti02 and V2O5.
There are many other possible combinations, but constituents and
concentrations of most catalysts are proprietary.
Reactor and catalyst configurations also vary with the
application, primarily to accommodate the different particulate
concentrations. The various types are shown in Figure 2 and
discussed below.
4-3
-------
FIGURE 1
TYPICAL FLOWSHEET FOR NOx SELECTIVE
CATALYTIC REDUCTION PROCESSES. (4)
-------
Parallel gj:
Flow
^.Catalyst
ilr
s \
si s
Moving
Bed
Flue
Gas ^
>
ir~
n-
-y
> Catalyst
J
Fixed
Packed
jrjs *
jTv s
JCk *
1
:-W>
^3gp
^Catalyst
FIGURE 2
TYPES OF REACTORS AND CATALYSTS FOR NOx SELECTIVE
CATALYTIC REDUCTION PROCESSES. (3)
4-5
-------
SCR catalysts and reactors designed for use with
natural—gas—fired boilers are very similar to typical fixed packed
bed reactor designs used in the chemical industry. Catalysts are
produced as spherical pellets, cylinders, or rings (similar to
Raschig rings) and contained within reactor vessels as fixed packed
beds. Gas distributors and catalyst supports are also similar to
those used in typical industrial reactors. However, designs for use
with oil- and coal-fired boilers have to be capable of tolerating
particulates (fly ash) in the flue gas stream. If a fixed packed bed
is used to treat a gas containing particulates, the catalyst bed will
become plugged and the process must be shut down in order to screen
the catalyst. For particulate laden flue gases, special catalyst
shapes and reactors have been designed that will operate without
plugging.
One such reactor design is the moving bed type in which
particulate laden catalyst is removed from the reactor, screened to
remove the fly ash, and returned. This arrangement can treat gases
with particulate concentrations up to 20 mg/Nm? (0.009 gr/scf or
0.0013 lb/ft3). The catalyst shapes employed are similar to those
used with fixed packed beds; the ring type is preferred where
particulate loadings approach 20 mg/Nm3 (0.0013 lb/ft3).
For applications with high particulate concentrations, a
parallel flow type catalyst is preferred. With these catalysts, the
gas flows straight through the open channels parallel to the catalyst
coated on the walls of the metallic or ceramic carrier. The
-------
particulates in the gas remain entrained while N0X reaches the
catalyst surface by turbulent convection and diffusion. An
alternative to the parallel flow catalysts is the parallel passage
reactor. In this design, the catalysts is the parallel passage
reactor. In this design, the catalyst material is arranged in
channels and held in place by a metal screen. The operating
principle is similar to that of the parallel flow catalyst.
The various parallel flow catalyst shapes are shown in Figure 3.
The parallel flow catalysts are normally amnufactured in a unit cell
configuration about 1 as shown in Figure 4. The cells are
stacked in banks in the reactor as shown in Figure 5.
The choice of one reactor type over another is based on the
characteristics of the flue gas to be treated which is, in turn, a
function of the fuel fired. Gas-fired applications can use the fixed
packed beds which are less expensive than the others. Oil-fired
applications may be able to use fixed packed beds if the particulate
concentration is low (<<20 mg/Nm^); however, it is likely that a
parallel flow or moving bed system will be necessary. Most processes
designed for use with coal-fired boilers require parallel flow
systems unless there is efficient upstream particulate removal. In
systems with hot-side particulate removal, it may be feasible to use
a moving bed system. Where it is possible to use more than one type
of system, the decision will be primarily one of economics. However,
the trend in Japan is toward the parallel flow catalysts.
4-7
-------
F
Metal Honeycomb
Tubular
Parallel Passage
8 mm
A A A!\J,
12 mm
i
t'i***** ''ii.'.'i'.'i'i'r.'r.'i'.'.'ri'i'i
7 mm
z:
6 mm
20 mm
Parallel Plate
i
I
00
t' '• '' ¦' C ' i:\- » ¦ , » '(!'kl"* •?'-'ir'j ;«
10 mm
10 mm
10 mm
Ceramic Honeycomb
10 mm
FIGURE 3
TYPES OF PARALLEL FLOW CATALYST SHAPES FOR N0X SELECTIVE
CATALYTIC REDUCTION PROCESSES. (3)
-------
Tubular configuration
Metallic honeycomb configuration
FIGURE 4
UNIT CELLS OF PARALLEL FLOW CATALYSTS FOR NOx SELECTIVE
CATALYTIC REDUCTION PROCESSES. (3)
4-9
-------
Catalyst Layer
FIGURES
PARALLEL FLOW REACTOR FOR NOx SELECTIVE
CATALYTIC REDUCTION PROCESSES. (3)
4-10
-------
ASSESSMENT OF NO„ FGT
Even though much progress has been made in catalysts and reactor
design, some problems still remain. The catalysts may not be
resistant to all contaminants in flue gas or be able to tolerate high
particulate loadings. In addition, fine particulates, smaller than
about 1 micrometer, may blind the catalyst surface. Catalyst life
also needs to be extended from the current guarantees of 1 to 2 years
for applications with S0X and particulates in the gas stream.
One of the major concerns with SCR processes is the formation of
solid ammonium sulfate [ (NH^^SO^,] and liquid ammonium
bisulfate (NH^HSO^) downstream of the reactor. The formation
conditions are difficult to avoid since some unreacted ammonia from a
SCR system and some SO3 from combustion of sulfur containing fuels
is expected. The biggest problem seems to be deposition of
(NH^)2SO4 and NH4HSO4 on the air preheater. These
compounds are highly corrosive and interfere with heat transfer. The
problem appears to be most severe with high sulfur oil firing.
Other concerns and potential problems include: emission of
NH3 and NH3 compounds; causing or increasing the emission of
undesirable compounds such as SO3; affecting the performance of
downstream pollution control equipment such as flue gas
desulfurization processes, electrostatic precipitators, and
baghouses; lack of proven NH3 and N0X analytical control systems;
sensitivity of the process to temperature changes due to boiler load
swings; and reliability of the process and its affect on the boiler
systems availability.
4-11
-------
Despite these potential problems, the processes have been
successfully installed and operated in Japan. As shown in Table 1
there are at least 13 process developers in Japan. In combination,
they have installed and operated 62 applications of N0X selective
catalytic reduction technology plus numerous pilot plants. Each of
the 62 applications is larger than 10,000 Nm^/hr ( 3.3 MWe). Most
of the installations are on gas- or oil-fired sources -- 37
installations are on clean gas streams and 25 installations are on
gas streams with S0X and particulates present. Of special
significance are the applications to utility boilers in Japan which
are shown in Table 2. The last two entries in Table 2 are being
planned for coal-fired sources. A 90 MW installation is expected to
start up in October 1980 at Hokkaido Electric's Tomakomai Station,
and a 250 MW installation is expected to start up in mid-1981 at the
Electric Power Development Company's Takehara Station.
DRY SIMULTANEOUS N0v/S0„ PROCESSES (3)
The Shell Flue Gas Treating process, a unique variation of
selective catalytic reduction technology, can simultaneously remove
both N0X and S0X from combustion flue gas. The process uses
copper oxide supported on stabilized alumina placed in two or more
parallel passage reactors. The reactions can be expressed as
follows:
4-12
-------
TABLE 1. SCR PLANTS (LARGER THAN 10,000 NmS/hr)
Plant Capacity Gaa Type of
Process Developer User Site Gas Source Fuel - (Nm'/hr) Pretreatment Keactor Completion
Hitachi, Ltd.
Mltaublalil R.I.
Ishikawajlma H.I.
Hitachi Zosen
Kawatetsu Chemical
Clilba
Coke Oven
COG, BFG*
300,000
ESPb,
Uc
1MB
Nov
1976
Chlyoda Kenzal
Kaizuka
Boiler
UO(IIS)*
15,000
None
1MB
Oct
1977
Kansal Palnc
Amagasakl
Boiler
Kerosene
16,000
II"
FBf
1977
Nlsahln Steel
Aabigaaakl
Boiler
110 (LS)*
20,000
U
FB
Aug
1977
Hlashln Steel
Amagasakl
Boiler
UO(LS)
19,000
11
FB
July 1977
Kansal Electric
Kainan
Boiler
Crude Oil
300,000
None
FB
1977
Chubu Electric
Chita
Boiler
LNG
2,000,000x2
None
FB
Apr
1978
Nippon Oils h Fats
Amagasakl
Boiler
110
20,000
None
1MB
Apr
1978
Hissliln Steel
Sakai
Boiler
Kerosene
30,000
None
n I
Dec
1978
Company A
—
Boiler
UO(LS)
490,000
None
FBP
June
1978
Company B
—
Boiler
IIO(LS)
550,000
None
FBP
June'
•978
Sumitomo Chemical
Sodegaura
Boiler
UO(LS)
300,000
ESP,
HE1
1MB
Sept 1976
Osaka Gas
Takaishi
Boiler
LNG
15,000x2
None
FB
Dec
1976
Tokyo Electric
Yokosuka
Boiler
110 (LS)
40,000
None
FBPJ
Mar
1977
Fuji Oil
Sodegaura
Boiler
110 (LS)
200,000
None
FBUk
Jan
1978
Kyushu Electric
Kokura
Boiler
LNG
1,690,000x2
None
FB
Oct
1978
Company C
—
Boiler
IIO(LS)
1,010,000
None
FBP
Feb
1978
Company D
—
Boiler
UO(LS)
490,000
None
FBP
July 1978
Chubu Electric
Chita
Boiler
H0(LS)
1,920,000
None
FBII
Feb
1980
Chubu Electric
Taketoyo
Boiler
Crude Oil
20,000
None
FBII
Apr
1977
AJ lnomoto
Kawasaki
Boiler
IIO(LS)
180,000
None
FBII
Jan
1978
Company E
—
Boiler
1I0(LS)
960,000
None
FBU
Apr
1978
Company F
—
Boiler
U0(LS)
480,000
None
FBII
June
1978
Cltugoku Electric
Kudamatsu
Boiler
Crude Oil
1,000,000
None
FBII
Apr
1979
Cliugoku Electric
Kudamatsu
Boiler
l»(LS)
1,900,000
None
FBU
July 1979
Tolioku Electric
Nligata
Boiler
IIO(LS)
1,660,000
None
FBII
Aug
1981
Idemltsu Kossn
Chlba
Boiler
CO
350,000
H
FB
Oct
1975
Shlndnlkyowa P.C.
Yokkaichl
Boiler
IIO(US)
440,000
ESP, DS |
, he, n
FB
Nov
1975
ToshIn Steel
llyogo
Furnace
Kerosene
71,000
None
FB
May
1976
Kawasaki Steel
Chlba
sm"
Coke, Oil
762,000
QS, ESP.
IIK, U
FB
Nov
1976
Nippon Satetsu
Chlba
Furnace
110
10,000
None
FB
Dec
1976
(Continued)
® f 1
.Coke oven g*s, blast furnace gaa Fixed bed JFixed bed parallel plate
Electrostatic precipitator flteavy oil (low sulfur) ^Fixed bed honeycomb
*jlleatlng Heater ^Uesulfurination
Intermittent moving bed llent exchanger "Sintering machine
'Heavy oil (high sulfur)
-------
TABLE 1. (Continued)
Proceao Developer
User
Plant
Site
Cm Source
Fuel
Capacity
(Ma'/lir)
Cam
Pretreatment
Type of
Reactor
CoapletIon
Sumitomo Cheaical
Sodegaura
Boiler
ltO(HS)a
30,000
ESFb
PBe
July 1973
lllgaatilnlbon H.
Sodegaura
Furnace
1>FC
200,000
Hone
Fft
Hay 1974
HIVvon tmonl*
Sodegaura
Furnace
tPG
250,000
None
n
Jan 1975
Sualtoao Cite* leal
Nlihaaa
Furnace
I.PG
200,000
None
FB
Apr 1975
SialtoM Cheaical
Anegaaakl
Furnace
l-PO
200,000
None
FB
Apr 1975
Suaitoao Cheaical
Anegasaki
Furnace
LPG
100,000
None
FB
Maruichi Kok«n
Sakal
Furnace
LPG
10,000
None
FB
Sua it onto Cheaical
Sodegaura
Boiler
IIO(HS)*,
240,000
Met ESP, HE, }fi
FB
Mar 1976
Sumitomo cliealcal
Sodegaura
Boiler
IW(MS)
300,000
ESP
1MB
Oct 1976
Sumitomo Chemical
Hltaul Toatsu
Mitsui Toatsu
Takalahi
Furnace
Off Caa
87,000
None
FB
Feb
1976
Osaka Pet. Chea.
Takalahl
Furnace
Off Caa
91,000
None
Ffl
Sept
1976
Mitsui Toatsu
Takolshi
Furnace
Off Caa
170,000
None
FB
Jan
1977
MlsfiJnilion H.
Takaialii
Furnace
Off Caa
363,000
None
FB
June
1977
Ideaitau Pet.
Chea.
Ichihara
Boiler
Off Caa
300,000
None
FB
Oct
1977
Mitsui
Mitsui Pet. Chew.
Chiba
Boiler
CO
200,000
F.SP
FB k
Oct
1975
Engineering
Ukisliiaa Pet.
Chea.
Chiba
Boiler
IIO(LS)
220,000
None
FBT"
Apr
1978
Sumitoho Chen.
Toho Can
Soraal
Boiler
Naphtha
31,000x2
None
fb
Oct
1977
Engineering
Toho Gas
Soraal
Boiler
Naphtha
23,000
None
FB
Dec
1977
Toho Caa
Soraal
Boiler
Naphtha
23,000
None
FB
June
1978
Tolio Caa
Soraal
Boiler
Naphtha
19,000
None
FB
July
1978
Mitsubishi Kakoki
Nippon Yakin
Kawasaki
Boiler
MO(IIS)
14,000
II8
FB
July
1978
Toho Caa
Soraal
Furnace
Naphtha
19,000
None
FB
Oct
19/8
Toho Caa
Km/aaakl
Boiler
Kerosene
10,000
None
FB
Nov
1978
Toho Caa
Chita
Bolter
Kerosene
30,000
None
FB
Oct
1977
Toho Caa
Chita
Boiler
Kerosene
43,000
None
FB
Oct
1978
J. C. Corp.
taahtaa Oil
Kaettima
Furnace
II0(MS)
50.000
None
ppl
Nov
1975
Fuji Oil
Chiba
Boiler
CO
70,000
None
PP
July
1976
Nippon Steel
Kiaitsa
Coke Oven
COG
150,000
None
PP
Mar
1977
Aiiah 1 Clasa
Asahl Claaa
Furnace
70,000
None
1MB1
Kobe Steel
Kurabo
Kansal N.K.
Kurabo
toatDMkl Coke Oven COG 104,000
HJrakata Boiler UO(IIS) 30,000
*IIeavjr oil (high sulfur)
Electrostatic precipitator
®rued bed
Heavy oil (medium sulfur)
pleavy oil (low aulfur)
Haat exchanger
^Heater
'fixed bed tubular utaljrat
None
None
1MB
QIB
Aug 1971
Aug 1975
Interaittent moving bed
'Parallel passage
Continuous Moving bed
-------
TABLE 2. TEST AND-COMMERCIAL PLANTS FOR UTILITY BOILERS (LARGER THAN 280,000 NmS/hr)
Capacity
1000
Plant
HOx Removal
Power Company
Plant Site
Fuel
HW
Nm'/hr
Constructor
Procesa
Completion
(X)
Ciuibu Electric
Chita
Low-S
on
375
1036
Mill*
SNRb
Feb
1977
45
Ctmbu Electric
Chita
LNG
700
1910
Ultachl,
Ltd.
SCR*
Apr
1978
80
Chubu Electric
Chita
LMC
700
1910
llltachl.
Ltd.
SCR
Sept
1978
80
Cliubu Electric
Chita
Low-S
Oil
700
1920
Mill
SCR
Feb
1980
80
Kanaal Electric
Kalnpu
Low-S
Oil
115
300
Hitachi,
Ltd.
SCR
June
1977
80
Company A
—
Low-S
Oil
156
490
Hitachi,
Ltd.
SCR
June
1978
50*
Company B
--
Uw-S
011
175
550
Hitachi,
Ltd.
SNR+SCR*
June
1978
50®
Kyualai Electric
Kltakyuslm
LNC
600
1610
MUI
SCR
July
1978
80
Kyualiu Electric
Kltakyusltu
LNG
600
1610
Mill
SCR
Dec
1978
80
Company C
—
Low-S
on
375
1010
Mill
SCR
Feb
1978
40®
Company D
—
Low-S
Oil
156
490
Mill
SCR
July
1978
40®
Company E
—
Low-S
on
350
1020
mia
SNR+SCR
Apr
1978
50®
Company F
—
Low-S
Oil
156
490
HU
SNR+SCR
June
1978
60®
CliuRoku Electric
KuJamatsu
Low-S
Oil
350
1000
mi
SCR
Apr
1979
80
Cliugoku Electric
Kudamatsu
Low-S
on
700
1900
mi
SCR
July
1979
80
Tolioku Electric
Nllgata
Low-S
Oil
600
1660
mi
SCR
Aug
1981
60
Electric Tower P.C.
Takaliara
Coal
250
800
—
SCR
June
1981
(80)'
Ilokkaldo Electric
Tomakomal
Coal
350x»tf
280
Hitachi,
Ltd.
SCR
Oct
1980
(90)'
^Htsublalil lleavy Industries.
Selective noncatalytlc reduction.
^Selective catalytic reduction.
^Ishlkawajlma-llarlma Heavy Industries.
^Combination of SNR and SCR
Ouu-Courtk of the gas from a 350 HU boiler.
iTIie low removal efficiency Is due to the use of a small amount of plate catalyst In an existing duct,
uealgn value.
-------
CuO + 1/2 Oj + S02
^ CuS04
(3)
4NO + 4NH3 + 02 CuS04 ^ 4 N2 + 6H20 (4)
CuSO^ + 2H2
^Cu + S02 + 2H20 (5)
Cu + 1/2 02
^ CuO
(6)
Flue gas is introduced at 385°C into (725°F) one of the reactors
where the S02 reacts with copper oxide to form copper sulfate. The
copper sulfate and, to a lesser extent, the copper oxide act as
catalysts in the reduction of N0X with ammonia. When the reactor
is saturated with copper sulfate, flue gas is switched to a fresh
reactor for acceptance of the flue gas, and the spent reactor is
regenerated. In the regeneration cycle, hydrogen is used to reduce
the copper sulfate to copper, yielding a S02 stream of sufficient
concentration for conversion to sulfur or sulfuric acid. The copper
is the reactor is oxidized, preparing the reactor for acceptance of
the flue gas again. Between acceptance and regeneration, steam is
injected into the reactor to purge the remaining flue gas or hydrogen
to eliminate any possibility of combustion. The process can also be
operated in the NOx-only mode by eliminating the regeneration
cycle, or in the SOx-only mode by eliminating the ammonia
injection.
4-16
-------
The process has been installed in Japan on a heavy-oil-fired
boiler treating 120,000 Nm-tyhr of flue gas. The unit has
demonstrated 90% SO2 removal and 70% N0X removal. UOP Process
Division is the licensor of the process in the United States.
WET PROCESSES (3)
The wet NOx and simultaneous NOx/SOx processes developed
to date cannot complete economically with dry processes for removal
of N0X from combustion flue gas. This is primarily due to the
complexity, limited applicability, and water pollution problems
associated with wet processes.
ECONOMIC. ENERGY, AND ENVIRONMENTAL IMPACTS
A major concern in applying any pollution control technology is
the associated economic, energy, and environmental impacts.
Preliminary estimates have been made to predict what the impacts
would be for applying selective catalytic reduction technology to a
hypothetical site. National impacts have not been determined.
ECONOMIC IMPACTS (4)
Preliminary economic estimates for application of selective
catalytic reduction processes to a coal-fired boiler were made by the
Tennessee Valley Authority. The basis for this estimate was a new,
500 MW boiler firing Eastern bituminous coal with a heating value of
24.4 MJ/kg (10,500 Btu/lb), a sulfur content of 3.5%, and an
4-17
-------
ash content of 16%. The study assumed operation of the boiler for
7000 hr/yr and the selective catalytic reduction process for 90%
reduction of N0X. The estimates were based on a capital investment
in mid-1979 and an annual revenue requirement in mid-1980* Results
of that study, which are summarized in Tables 3 and 4, indicate that
the capital costs of selective catalytic reduction process will be
about $36/kW or $8.6/ACFM and that annual revenue requirements will
be about 2.6 mills/kWh. (Since the volume of gas treated is the
dominating parameter in determining the cost of installing a
selective catalytic reduction system, the $/ACFM factor may be the
most meaningful in determining representative costs of firing various
coals in various sized plants.) The study also indicated that the
dry simultaneous N0x/S0x process would be competitive with these
costs if the cost of a flue gas desulfurization process for SO2
control was added to the cost of the N0X selective catalytic
reduction process, but that the cost of wet simultaneous N0x/S0^
processes would be significantly higher.
ENERGY IMPACTS (4, 5)
The SCR processes are projected to require about 0.2% of the
boiler capacity. This assumes that flue gas reheat will not be
required and does not include an energy requirement represented by
raw materials.
4-18
-------
TABLE 3. ESTIMATED CAPITAL INVESTMENT FOR EMISSION CONTROL SYSTEMS (4)
Process Types
Selective Catalytic Reduction (SCR)
SCR, FGD, ESP
Dry Simultaneous NOx/SOx
Wet Simultaneous NOx/SOx/PM
Control Equipment Capital Costs ($/kw)
N0„ SOo PM TOTAL
100
42
164
163^
200
Basis for the Estimate:
Particulate Matter (PM) Control System
FGD System
S02 Removal Efficiency
NOx Removal Efficiency
Particulate Removal Efficiency
Boiler Size
Fuel
Heating value
Sulfur content
Ash content
Operation
Capital Investment
Annual Revenue Requirement
SCR Processes
Dry Simultaneous N0x/S0x
Wet Simultaneous N0x/S0x/PM
ESP for dry systems;
wet scrubber for wet simultaneous systems
Limestone
90%
90%
99.5%
500 MW, new
Coal
24.4 MJ/kg (10,500 Btu/lb)
3.5%
16%
7000 hr/yr
mid-1979
mid-1980
Average of UOP-Shell, Hitachi
Zosen, Kurabo Process Costs
UOP-Shell
Average of Moretana Calcium and
Asahi CIO2 Process Costs
-------
TABLE 4. ESTIMATED ANNUALIZED COST FOR EMISSION CONTROL SYSTEM (4)
Process Types
Selective Catalytic Reduction
(SCR)
SCR.FGD, ESP
Dry Simultaneous NOx/SOx
Wet Simultaneous NOx/SOx/PM
Control Equipment Annualized Costs (roills/kWh)
NOy SO? PM TOTAL
2.7 — — 2.7
2.7 4.2 0.7 7.6
6.4 6.4 0.9 7.3
11.3 11.3 11.3 11.3
Basis for the Estimate
(See Table 3)
-------
ENVIRONMENTAL IMPACTS (5)
The major concerns from an environmental impact viewpoint appear
to be emissions of NH3 and NH3 compounds and disposal of spent
catalysts* Operational techniques to limit NH3 emissions and
methods for regenerating or reclaiming the catalysts are under
development to minimize problems in this regard.
IMPACT OF AMMONIA UTILIZATION (6)
The economic sensitivity of the SCR sysems to the cost of
ammonia has been the subject of some concern. It was found that the
annual ammonia cost was about 10% of the annual revenue requirements
for the dry systems. Therefore, the average annual requirement may
be expected to increase about 10% if the ammonia cost is doubled.
A major concern of the widespread utilization of ammonia-based
N0X control systems has been the impact on the domestic ammonia
market. A study of this situation indicated that the need for NH3
for NOx control apparently would not have an abrupt adverse impact
on the availability and price of NH3. Under the assumptions of the
study, the primary impact on the domestic NH3 market would be to
cause the U.S. NH3 demand to increase at 4.5%/yr during the period
1985-2000. This increase could be met with the addition of one
ammonia plant per year.
4-21
-------
DEMONSTRATION PROJECTS
To determine the actual performance and cost of applying
selective catalytic reduction processes in the U.S., several hardware
projects have been planned as summarized in Table 5. These projects,
in conjunction with technology assessment and control strategy
studies, will enable an assessment of the feasibility of applying
N0X and simultaneous NOx/SOx flue gas treatment processes in
the U.S.
However, if the coal-fired pilot plants and oil-fired
demonstration scale plants are successful, then a coal-fired
prototype or demonstration scale plant (10-100MW) will be desirable.
It is conceivable that such a plant could be built in conjunction
with a new or expanding facility in an area with nonattainment,
prevention of significant deterioration, visibility or other
environmental constraints. In this manner, the environmental impact
of the facility could be minimized, and selective catalytic reduction
technology for N0X control could be advanced to the commercially
demonstrated status.
4-22
-------
TABLE 5. PLANNED DEMONSTRATION PROJECTS OF NOx SELECTIVE CATALYTIC
REDUCTION TECHNOLOGY IN THE U.S.
Sponsor
Contractor
Location
Planned
Size N0X
(MWe) Fuel Reduction Start-Up
Environmental Protection
Agency
Environmental Protection
Agency
Electric Power Reaearch
Institute
Southern California
Edison2
Loa Angeles Department2
of Water and Power
(LADWP)
Hitachi Zoaen
Georgia Power Co.
UOP Process Division Tampa Electric Co.
Kawasaki Heavy
Industries
Kawasaki Heavy
Industries
Undetermined
0.5 Coal 90Z
0.5 Coal 90Z'
Public Service Co.
of Colorado
Arapahoe Station
Southern California
Edison
Huntington Beach Station
Undetermined
2.5 Coal
100
Oil
90Z
901
Summer 1979
Summer 1979
Winter 1980
October 1981
>1002 Oil 90%2 October 19812
1) 901 Reduction of 8O2 also planned by the dry, simultaneous N0x/S0x process.
2) Required to meet regulations of the California Air Resources Board.
-------
REFERENCES
Mobley, J.D., "Flue Gas Treatment Technology for N0X Control,"
Proceedings of the Third Stationary Source Combustion Symposium,
Volume II, Advanced Processes and Special Topics, Report No.
EPA-600/7-79-05Ob, NTIS No. PB292-540/AS, pp. 245-281, February
1979.
Jones, 6.D., and J.D. Mobley. "Technical and Economic Assessment
of N0X Flue Gas Treatment Technology." Presented at the
American Institute of Chemical Engineer's 87th National Meeting,
August 1979, Boston, MA.
Ando, Jumpei, NO^Abatement for Stationary Sources in Japan.
EPA-600/7-79-205, August 1979, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
Maxwell, J.D., T.A. Burnett, and H.L. Faucett. Preliminary
Economic Analysis of N0V Flue Gas Treatment Processes.
Tennessee Valley Authority, November 1979, Final Report. EPA
Interagency Agreement D8-E721-FU, U.S. Environmental Protection
Agency, Research Triangle Park, NC and EPR.I Contract RP783-3,
Electric Power Research Institute, Palo Alto, CA.
Jones, G.D., and K. Johnson. Technology Assessment Report for
Industrial Boiler Applications: N0y Flue Gas Treatment.
Radian Corporation, June 1979, Draft Report. EPA Contract No.
68-02-2608, Task 45. U.S. Environmental Protection Agency,
Research Triangle Park, NC.
Burnett, T.A. and H.L. Faucett. Impact of Ammonia Utilization by
N0V Flue Gas Treatment Processes. Tennessee Valley Authority,
EPA-600/7-79-011, January 1979. U.S. Environmental Protection
Agency, Research Triangle Park, NC.
4-24
-------
SECTION 5
PILOT-SCALE EVALUATION
OF TANGENTIAL FIRING
FOR
PRESENTATION TO EPA REGION VIII
by
DAVID G. LACHAPELLE
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, NC 22711
-------
Pilot Scale Evaluation
of Tangential Firing
-------
Objectives
• Develop a Better Understanding of Processes Which Control
NOx Formation During Combustion of Pulverized
Coal in Tangentially-Fired Furnaces
• Develop NOx Combustion Concepts for
Retrofitting on New Design of Tangentially-
Fired Boilers
-------
Major Tasks
"A": Coal-Fired Tangential System Definition and Evaluation
"B": Optimization of Near Burner and Intermediate Zone NOx
"C": Optimization of Coal-Fired Tangential System Low-NOx
Concepts
Ln
I
N>
-------
"A" Series Tasks
Test Plan
Chemical/Physical Process Definition
Fireball Characterization
Fiame/Flame Interaction (Vertical)
Fiame/Flame Interaction (Circumferential)
Early Mixing Tests
-------
"B" Series Tasks
Test Plan
Hardware Design
Early Mixing/02 Availability
Early Mixing/Hot FGR
Early Mixing/Temperature Effects
Intermediate Zone/02 Availability
Intermediate Zone/FGR
Intermediate Zone/Active Cooling
-------
"C" Series Tasks
Test Plan
Fireball Simulations (Water Modeling)
Hardware Burner Testing
Fuels Testing
-------
Facility
• 39" Cube Refractory Firebox
• Four Refractory-Lined Heat Exchange Sections
• Front Wall or Tangential Burners
• Flow Control and Measurement
• Secondary/Stage Air Preheat to 800° F
I
• Multi-Fuel Capability (Coal, Oil, Gas)
• Firing Rate: 1.0 to 3.0 x106BTU/Hr
-------
Figure 1
Furnace
Cross Section
-------
Coal Analysis (as Rec'd)
Utah Bituminous
Starpoint No. 1, Hiawatha Seam
In
I
CO
Carbon 65.97
Hydrogen 4.86
Nitrogen 1.21
Sulfur 0.92
Chlorine 0.01
Moisture 5.64
Ash 9.38
Oxygen (Diff) 12.01
Heating Value, BTU/Lb 11,842
-------
Figure 2
Tangential System
Characteristics
Ol
I
Primary Air plus Coal
Slow-mix Vertical Jets
Vertical Stacking off Burners
Serial Impingement of Adjacent Jets
Convection of Burnt Gases Past
Several Burner Levels
Asymmetric Combustion
Large-Scale Vortex and Bulk
Recirculation of Combustion
Products
bulk of combustion
air
Top View ( air
Side View
air and fuel registers
-------
Table 1
Potential NOx Formation/Reduction Process
System
Characteristics
Process
NOx Impact
Primary air plus coal
Early fuel-lean zone
Negative
Slow mix vertical jets
Fuel-rich near burner zones
Positive
Vertical stacking of burners
Slower mixing, fuel-rich zones
Positive
Serial impingement of adjacent
jets
• Positve ignition, serial flame
processing
• Breakup of jet, rapid mixing
Positive
Negative
Convection of burnt gases
past several burner levels
• Serial flame processing
• Breakup of jet, rapid mixing
Positive
Negative
Asymmetric combustion
• Mixing of products on
furnace side
• Breakup of jet, rapid mixing on
furnace side
• Cooling of gases on wall side
Positive
Negative
9
Large-scale vortex and bulk
recirculation of combustion
products
• Mixing of products with burning
jets, serial flame processing
• Breakup of jet, rapid mixing
Positive
Negative
Fuel Characteristics
Process
NOx Impact
Volatilization
Fuel N evolution
?
Nitrogen
Fuel N conversion
Negative
Nitrogen/sulfur
• Fuel-rich enhancement
• Fuel-lean inhibition
Negative
Positive
Nitrogen/sulfur/oxygen
Fuel N conversion by O
Negative
CharN
Char NOx
7
-------
Figure 3
Burner Configurations
Ui
I
CE-Type *
<500 ppm)
Configuration C
(525 ppm)
o
o o
O <3 O
O O
O
Firing Rate = 1 x 106 Btu/hr.
Overall excess air = 15%
NO corrected to 0% 02
Configuration A
(800 ppm)
Configuration B
(700 ppm)
Vortex Direction
Configuration 0
(815 ppm)
Primary Air
and Fuel (0.75" Dia)
Secondary
Air (1.75" Dia)
Primary Air
and Fuel (0.87" Dia)
Secondary
Air (0.87" Dia)
-------
Figure 4
Burner Configurations (cont.)
I
ro
ft
p
Secondary & .
ajr Primary air
n. and fuel
1.25" dia —-Q
Configuration E
(550 ppm)
,2.25"
h—^
Secondary
airX) . .
v-^ ^^Pnmary air
and fuel
2" dia
.-O
Configuration F
(450 ppm)
Vortex Direction
Firing rate: 1 x 106 Btu/hr.
Overall excess air: 15%
NO corrected to 0% 02
1.25"
dia.-
Secondary^pQ ^prjmary ajf
and fuel
O —Wall air (2" dia. jet)
Configuration G
(325 ppm)
-------
in
I
Co
Figure 5
Configuration G (Top View)
-------
Figure 6
Burner Configurations
Vortex
direction
I 3"
N—~
/V 1.25" dia.
Primary fuel
Secondary 3/3" jVa"
air L
ui
Wall air
(6" x 3/8" slot)
Q\®C and alr
_____ 0.87" dia.
O*— Secondary
•
Fuel and
^ primary
air
Configuration H
(260 ppm)
Wall air
(6" x 3/8" slot) 3"
Configuration I
(270 ppm)
Firing Rate: 1 x 10® Btu/hr
Overall excess air: 15%
NO corrected to 0% 02
Wall air
(6" x 3/8" slot)
Configuration J
(300 ppm)
Fuel and
primary air
-------
Ln
t
u*
Figure 7
Wall Air Configurations
Configuration H
(Top View)
Configuration I
(Top View)
-------
Near-Term Plans
• Continue Testing of Configurations
• Possibly Evaluate Shielding of Fuel and Air with Flue Gas
• Optimize Most Promising Concepts
• Build Optimum Burners
• Extend Tests to Other Coal Types
• Possibly Evaluate at Two Burner Elevations
-------
Far-Term Plans
Explore Possibility of Scale-Up with CE
Conduct Tests @ 12 to 25 x 10® BTU/Hr per Burner
Explore Possibility of Field Demonstration
-------
SECTION 6
ASSESSMENT SUMMARY
FOR CONTROL OF NOx
EMISSIONS
FOR
PRESENTATION TO EPA
REGION VIII
-------
ASSESSMENT SUMMARY FOR CONTROL OF NOx EMISSIONS
In the preceding presentations we have defined the Colstrip
plant situation regarding NOx emissions; including boiler design,
type of coal used, and the effect of these variables on baseline
N0X levels. We have also discussed pertinent control and treatment
technologies to reduce the N0X emission levels, including process
and equipment description, performance, costs, and impacts on the
environment and energy usage. What remains is to summarize this
information and to present an overview, fitting the pieces together
and presenting conclusions.
Applicability Ranges for Control Technologies
One of the first conclusions drawn from the preceding
discussions is that no one currently available N0X control or
treatment technology is preferable to all others over the entire
range of N0X control levels desired.
Figure 1 summarizes the applicable range for N0X reduction
technologies for tangentially-fired boilers such as used in the
Colstrip plant. Normally, these facilities can achieve 0.7 lb.
N0x/10^ Btu emission levels without any additional control or
treatment. By using low excess air operating conditions, N0X
levels of 0.5-0.6 lb/10^ Btu may be achieved. Since low excess air
technology requires little, if any, equipment modifications or
operating costs, and may even give improved boiler efficiency, this
is a simple economic approach within its specified range of
effectiveness.
6-1
-------
ON
I
N>
Figure 1
Applicability of N0X Reduction Technologies
for Tangentially Fired Boilers
No Low Excess
Treatment Air (LEA)
I
Overfire LEA + OFA +
Air (OFA) Thermal
+ LEA DeNOx
Rich
Fire Ball
Dry
Catalytic
Combustion Modifications
j i i
Post-Combustion
Treatments
-------
Windboxes for new tangentially-fired boilers (also all made
since 1970) include provision for injection of overfire air as
standard equipment. Use of overfire air as a N0X reduction
technology is usually applicable down to approximately 0.4-0.5 lb.
NOX/106 Btu remaining after treatment. The staged combustion
made possible by overfire air addition is more effective than low
excess air technology alone, and extends the applicability range as
shown in Figure 1. Costs involved are primarily those for windbox
modifications for older boilers. The 0.5 lb N0X /10^ Btu
emission level represents the best estimate of current C'olstrip plant
capabilities for both existing and new units.
As shown on Figure 1, other combustion modification techniques
are in the research or development stages. These techniques such as
the "rich fireball" concept and new burner designs may make it
possible to reduce N0X emissions well below the current 0.4-0.5
lb./10^ Btu level without flue gas treatment.
Unlike combustion modification controls, the flue gas treatment
technologies are applicable over a wide range of N0X concentra-
tions. Their particular attractiveness is in their ability to reduce
N0X emissions to levels significantly below those for presently
demonstrated combustion modifications. Applicable flue gas treatment
technologies include the Exxon thermal (non-catalytic) DeNOx pro-
cess and dry catalytic DeN0x processes as discussed in the preced-
ing presentations.
6-3
-------
The Exxon Thermal DeNOx process, which uses ammonia injection
to reduce the nitric oxide in the flue gas to nitrogen, can be used
to treat flue gases containing even larger initial N0X concentra-
tions than shown in Figure 1. Flue gases with N0X levels well
above 1.0 lb N0x/10^ Btu are estimated to be reduced to 40 to 50
percent of their original levels (1). The same, or greater (up to 90
percent) reductions have been achieved in pilot plant experiments for
flue gases with much lower initial NO concentrations (0.4 lb
NOx/10^Btu) (2). Therefore, the Exxon thermal DeNOx process
can be applied to flue gases containing almost any level of N0X,
but to achieve a final N0X level of approximately 0.2 lb
N0x/106 Btu it is necessary to start with an initial N0X
concentration of 0.6 lb/10^ Btu or less. This requirement makes
the combination of combustion modifications with thermal DeN0x an
attractive option.
Dry catalytic postcombustion treatments were reported in the
previous presentation to give up to 90 percent N0X removal for the
SCR process (catalytic reduction of NO with ammonia) and 70 percent
for the Shell Flue Gas Treating process (simultaneous removal of
N0X and S0X). Although it is possible that these dry catalytic
postcombustion processes may be used alone, it is likely that they
too will be used in combination with combustion modification con-
trols.
6-4
-------
Technology Status
Table 1 and Figure 2 summarizes the.status of the N0X
reduction technologies for coal-fired boilers.
Combustion modifications by use of low excess air and overfire
air techniques have been demonstrated for coal-fired U.S. utility
plant boilers in a variety of plants. Overfire air operation has
proven to be a successful and practical supplementary control for
limiting N0X formation on coal-fired steam generators.^)
The Exxon thermal DeNOx and the dry catalytic postcombustion
processes:, on the other hand, have not been demonstrated on flue
gases from coal-fired boilers beyond the pilot plant stage. One
pilot plant demonstration of the Exxon thermal DeNOx process has
been carried out in the U.S. using a 3 x 10^ Btu/hr (nominally 250
lb./hr coal feed) coal-fired boiler. The thermal DeNOx process has
also been demonstrated on various commercial gas and oil-fired boil-
ers and furnaces in the U.S. and -Japan (see Table 2).
Dry catalytic post combustion processes have been developed
primarily in Japan. Many pilot plant demonstrations, including a
number on coal-fired power plants (see Table 3) have been carried
out. In addition, various commercial plants using dry catalytic
N0X removal processes are in operation, or being installed (see
Table 4). The presence of more particulates and S0X in the flue
gases from coal-fired boilers as compared to those from gas- or
oil-fired boilers has thus far slowed the commercial use of dry
6-5
-------
I
ON
I
(J>
Figure 2
Status of N0X Control
Technology
1.1
1.0-
0.9-
0.8-
c o.7 -|
O (O
w °
« r
\
C vx0.6 H
O
uj ~
ra n 0.5 —
CD S
0)
-2 0.4 H
0.3 H
0.2-
0.1 -
\
LEA-^l
OFA + LEA
Thermal DeNO
Legend
LEA Low Excess Air
OFA Overfire Air
FGT Flue Gas Treatment
RFB Rich Fire Ball
-] 1 1 1 1 i i i i i i i i i i r
1971 '72 '73 '74 '75 '76 '77 '78 '79 '80 '81 '82 '83 '84 '85 '86
-------
TECHNOLOGY
TABLE 1 - STATUS OF NO REDUCTION TECHNOLOGIES FOR COAL-FIRED BOILERS
x
DEMONSTRATION LEVEL
ESTIMATED YEARS TO U.S.
COMMERCIALIZATION
Combustion
Modif icatlons
UNITED STATES
FOREIGN
PILOT COMMERCIAL
PLANT USAGE
PILOT COMMERCIAL
PLANT USAGE
1. Low Excess
Air
Available now
2. Overfire
Air
Available now
3. New
Techniques
Thermal DeNO
+
+
0
0
3-5
3-5
Dry Catalytic
Postcombustion
Treatments
3-5
+Has Been Demonstrated
^Has Not Been Demonstrated
-------
(4)
TABLE 2. SUMMARY OF COtWERCIAL APPLCATIONS OF EXXON'S THERMAL DeNO^ PROCESS
Source
Fuel
Burned
Location
Initial Nitric
Oxide Emissions
(ppm as measured)
DeNO Rate
( Percent)
Additive*1
Comments
Steam boiler
45 MM heat input
Gas/oil
Japan
120-150
60
Yes
No reduction obtained at full load
Incinerator
7 ton/hr
Haste
Japan
100-180
20-70
NAC
Difficult source to retrofit because
of constant change in fuel
Crude heater
150 x 10 bbl/day
Gas/oil
Japan
150
35-65
Yes
Best reductions achieved at high
load
Steam boiler
76 m heat input
Gas/oil
Japan
95-145
35-50
NA
No details of retrofit system are
available
Utility boiler
275 MW heat Input
Gas/oil
Japan
80-120
60
NA
No details of retrofit system are
available
Utility boiler
275 heat input
Gas/oil
Japan
80-120
50-60
NA
No details of retrofit system are
available
Crude heater
150 x 10 bbl/day
Gas/oil
Japan
80-85
40-65
NA
Best reductions achieved at high
loads
Thermal recovery
heater
Oil
USA-
California
260
50-70
NA
First commercial U.S.
installation
Utility boiler
375 W
Residual
oil
Japan
NA
40
No
Does not use Exxqn NH^ Injection
technology — NH^ emission limited
to 10 ppm
Exptl. fire tube
boiler, 0.9
Bitum. &
subbitum.
coal
USA-
California
500-800
65d
No/Yes
Experimental study sponsored by
EPRI; conducted by KVB, Inc.
aExcess oxygen varied between 3-5 percent for all sources
"Yes" inldcates that hydrogen was injected together with NH- to obtain reported NO reduction performances
* A
"N0m Indicates that no additive was used
c
NA - no data are available
^Pealt reduction
-------
Table 3
HITACHI ZOSEN PILOT PLANT EXPERIENCE 6i)
Number of Plants
Heavy fuel oil-fired boilers 21
LN6- and LPG-fired furnaces 6
Iron ore sintering 3
Heavy oil-fired cement kilns 2
Heavy oil-fired glass smelting
furnaces 2
Coke ovens 3
Coal-fired power plants 3
LPG-fired simulation gas 1
(for gas turbine)
Total 41
3
Total raw gas .flow through test plants = 35,130 Nm /h
-------
Table If
COMMERCIAL PLANTS USING OR INSTALLING HITACHI
DRY CATALYTIC NOx REMOVAL PROCESS ((,)
No.
Plant3
Gas Flowrate
(Nm3/h)
Fuel
DeNOx
Eff
<*)
Operating
Date
Remarks
1
A-l
300,000
Crude oil
80
June
1977
25* x 450 MW
2
A-2
2,000,000
LNG
80
March
1978
700 MW
3
A-3
2,000,000
LNG
80
March
1978
700 MM
4
A-4b
483,000
Heavy oil
50
July
1978
175 MM
5
A-5b
466,000
Heavy oil
30
July
1978
156 MM
6
A-6
280,000
Coal
90
Sept.
1980
25% x 350 MM
7
A-7
980,000
Kerosene
80
March
1981
90 MW (Stag)
8
B-l
15,000
Heavy oil
70
Sfept.
1976
9
B-2
20,000
Heavy oil
90
Sept.
1977
10
B-3
16,000
Kerosene
90
Oct.
1977
11
B-4
20,000
Heavy oil
90
Apri 1
1978
12
B-5
30,000
Kerosene
90
Sept.
1978
13
C-l
500,000
COG/BFG
95
Oct.
1976
14
C-2
3,600
Electric
furnace
84
Jan.
1977
15
C-3
50,000
Heavy oil
80
Oct.
1979
aA = utility boiler; B = industrial boiler; C = others.
bIn A-4 and A-5, plate-type catalysts are installed in the restricted space of
the flue gas duct between economizer and air preheater.
-------
catalytic technologies in coal-fired facilities. The special cata-
lyst bed design and other process modifications needed for applica-
tion to flue gases from coal-burning operations have now been
developed and tested at the pilot plant level and the technology is
moving toward commercialization. Although no coal-fired facilities
are as yet installed, at least one installation is planned and
demonstration of this dry catalytic NOx removal process on coal-
fired boilers should occur within the next three to five years.
Costs
The costs for pertinent NO* control/removal technologies are
compared in Table 5. These costs, taken from various sources and
modified to a common base, should be taken as ballpark figures only.
They do, however, illustrate the levels of increasing expenditures
needed as the technologies progress from the relatively simple main-
tenance of low excess air levels in the boiler to the more sophisti-
cated combustion modifications and postcombustion treatments.
The cost for maintaining low excess air is virtually zero. In
fact, there are situations where some net benefit in fuel savings
occurs.
Combustion modifications, as represented by overfire air addi-
tion, still have relatively low costs. These costs, are almost
entirely capital expenditures for boiler firing installations.
Retrofit costs are significantly higher than those for new boiler
installations.
6-11
-------
TABLE 5
COMPARISON COSTS FOR NOx CONTROL/REMOVAL
500 MW COAL-FIRED PLANT
CAPITAL COSTS
OPERATING COSTS
TOTAL ANNUAL
OPERATING COSTS
TECHNOLOGY
(lb
A NOx RANGE $
. NOx/10bBtu)
MILLS/
KW-HR
iiYR
MILLS/
KW-HR
$
MILLS/
KW HR
Low Excess Air
0.7 - 0.5 ~0
~0
~0
~ 0
~ 0
~ 0
Overfire Air
0.7 - 0.4 $500,000
.025
~0
~3
90,000
.025
Retrofit
New Installation
$500,000
120,000
.025
.006
~0
~0
~0
~0
90,000
21,600
.025
,006
Thermal DeNox
0.7 - 0.4 4,400,000
0.23
1,300,000
0.37
2,210,000
**
' 0.60
0.4 - 0.2 4,070,000
0.21
966,000
0.28
1,830,000
**
0.49
Catalytic DeNox
0.7 - 0.4>30,000,000*
>1.54
2,800,000
0.80
8,200,000
>2.34
Catalytic DeNo^
0.4 - 0.2>30,000,000*
>1.54
2,460,000
0.70
7,860,000
>2.24
Overfire Air (new)
-L
OFA
0.7 - 0.4 30,390,000
0.17
966,000
0.28
1,576,000
0.45
T
Thermal DeNox
td
0.4 - 0.2
Overfire Air (new)
_L
OFA
0.7 - 0.4>30,000,000*
1.54
2,460,100
0.70
7,880,000
>2.25
T
Catalytic DeNO^
SCR
0.4 - 0.2
*Includes particulate removal with ESP
r*
Includes licensing fees
-------
Costs for removal of N0X from flue gases are at least an order
of magnitude higher than for overfire air addition. Costs for the
Exxon Thermal DeNOx process include both significant capital
investment and operating expenses. Most of the operating expenses
involve purchase of ammonia.
The^ catalytic DeNOx systems as applied to coal-fired opera-
tions include hot flue gas treatment, to remove particulates, and
catalyst beds in addition to ammonia handling facilities. Therefore,
the capital expenditures are several times greater than for the
thermal DeNOx system. Non-reagent operating costs (maintenance,
catalyst replacement, labor, etc.) are also higher than for the
thermal DeNOx process.
Combinations of combustion modifications and flue gas treatment
technologies provide the most cost-effective way to achieve low (0.2
lb. N0X/10^ Btu or less) levels of N0X. Use of overfire air in
a tangentially-fired boiler can reduce the NOx level to approxi-
mately 0.4-0.5 lb. N0x/106 Btu) at much lower cost than possible
with any postcombustion process. Once combustion modification has
reduced the N0X level, removal of the remaining N0X with ammonia
reduction not only becomes more practical but also lower in cost.
Other Influencial Factors
Use of the N0X control/treatment technologies is influenced by
factors other than N0X removal performance and economics. These
factors, including such considerations as boiler corrosion, energy
6-13
-------
and environmental impacts, process problems, and background and
experience, are shown in Table 6.
Combustion modifications such as low excess air and overfire air
have been demonstrated in numerous commercial boilers and the operat-
ing conditions to keep CO, hydrocarbon, and soot emissions within
acceptance limits have been established. Energy impacts, as based on
boiler derat'ing, are very low. Effects such as corrosion or deposits
on the firewall still need further documentation, as was discussed in
the previous combustion modification presentation.
The flue gas treatment technologies avoid any corrosion or
buildup effects in the boiler. The major considerations for these
technologies for coal-fired facilities involves process and environ-
mental effects of flue gas components. When ammonia is added to the
flue gas, care must be exercised to avoid discharging an excess to
the environment. Side reactions are also important influences.
Reaction with S0X in the flue gas, gives liquid and/or solid
sulfates which may plug equipment or otherwise interfere with the
process operation. Particulates may contaminate or plug the
catalysts beds. Also, some energy is expended in handling and
transfer of the ammonia, carrier gases, and the flue gas itself.
Electrostatic precipitators used to clean the flue gas of
particulates and avoid catalyst bed pluggage also consume energy.
The main shortcoming for flue gas N0X treatment technologies,
however, is the lack of demonstrated performance in coal-fired
facilities.
6-14
-------
TABLE 6
COMMENTS ON TECHNOLOGIES
ENERGY AND BOILER
RATING IMPACTS (%)
INFLUENCE ON REDUCTION OF BOILER BOILER OTHER
TECHNOLOGY OTHER EMISSIONS EFFICIENCY CORROSION FACTORS
Low Excess Air
Overfire Air
CO, Unburned hydro-
carbons and carbon loss
increase as excess air
is decreased beyond
practical limits.
-0
CTN
I
Oi
No significant
deleterious
effects
Demons t ra t i on
to date show no
significant in-
creases . Long
term study is
underway
May even improve
boiler efficiency
Effect on furnace
waterwall deposits
being studied
Thermal DeNox
0.2
NH^ not in
contact with
boiler tubes
on wall
General lack of
knowledge on scale-
up and demonstration
factors for coal-
burning boilers
Other Post
Combustion
Treatments
0.3
No. U.S.
experience
-------
Conclusions
To sum up the conclusions regarding new or retrofitted
tangentially-fired boilers, these boilers can usually achieve 0.7 lb.
N0x/10^ Btu emission levels without mechanical modification or
special firing conditions. Use of overfire air as a combustion modi-
fication, should reduce NOx emission levels to 0.4-0.5 lb. NOx/
10^ Btu. Costs to achieve this level will be small.
There is no current U.S. demonstrated technology to reduce N0X
emissions below the 0.4-0.5 lb. N0x/l0^ Btu level and it will be
3 to 5 years before any such technology can be expected to be
commercialized. If N0X emissions below 0.4-0.5 lb N0x/10^ Btu
are required, postcombustion ammonia injection using either thermal
or catalytic processes appears to be the most promising approach.
Costs for NH3 injection technologies will be one or two orders of
magnitude higher than presently used combustion modifications.
6-16
-------
REFERENCES
1. Varga, G.M., Tomsho, M.E., Ruterdories, B.H., Smith, G.J. and
Bartok, W. "Applicability of the Thermal DeNOx Process to
Coal-fired Utility Boilers," EPA-600/7-79-079, March 1979.
2. Muzio, L.J., Arand, J.K., and Malonty, K.L., "Noncatalytic NOx
Removal with Ammonia," EPRI Report FP-735, April 1978.
3. Marshall, J.J. and Selker, A.P., "The Role of Tangential Firing
and Fuel Properties in Attaining Low NOx Operation for
Coal-Fired Steam Generation," Proceedings: Second NOv Control
Technology Seminar, EPRI Report FP-1109-SR, July 1979.
k. Castaldini, C., Salvesan, K.G., and Mason, H.B., "Technical
Assessment of Thermal DeNOx Process," EPA-600/7-79-117, U.S.
Environmental Protection Agency, May 1979.
5. Saleem, A., Galgano, M., and Inaba, S., "Hitachi-Zosen DeNOx
Process for Fossil-Fired Boilers," Proceedings: Second NO^
Control Technology Seminar, EPRI Report FP-1109-SR, July 1979.
6. Kurda, H. and Nakajima, F., "Some Experiences of N0X Removal in
Pilot Plants and Utility Boilers," Proceedings: Second NO^
Control Technology Seminar, EPRI Report FP-1109-SR, July 1979.
7. Burrington, R.L., Cavers, J.D., and Selker, A.P., "Overfire Air
Technology for Tangentially Fired Utility Boilers Burning Western
U.S. Coal," EPA-600/7-77-117 U.S. Environmental Protection
Agency, October 1977.
8. Mobley, J.D., "Status of EPA1s NOx Flue Gas Treatment Program,"
Proceedings: Second NOv Control Technology Seminar, EPRI
Report FP-1109-SR, July 1979.
6-17
------- |