EPA-650/2-73-033a
October 1973 Environmental Protection Technology Series
fill >-a$sii. m
-------
DISCLAIMER
This project has been funded at least in part with Federal funds from
the Environmental Protection Agency under contract number 68-02-0216.
The content of this publication does not necessarily reflect the views or
policies of the U.S. Environmental Protection Agency, nor does mention
of trade names, commercial products, or organizations imply endorsement
by the U.S. Government.
-------
EPA-650/2-73-033a
AERODYNAMIC CONTROL OF NITROGEN
OXIDES AND OTHER POLLUTANTS
FROM FOSSIL FUEL COMBUSTION
VOLUME I. DATA ANALYSIS AND SUMMARY OF CONCLUSIONS
by
D.R. Shoffstall and D.H. Larson
Institute of Gas Technology
IIT Center, 3424 South State Street
Chicago, Illinois 60616
Contract No. 68-02-0216
Program Element No. 1A2014
ROAP No. 21ADG47
EPA Project Officer: David W. Pershing
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
October 1973
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names Or commercial products constitute endorsement
or recommendation for use.
-------
TABLE OF CONTENTS
Page
INTRODUCTION 1
SUMMARY OF CONCLUSIONS 2
Input-Output Test Results 2
In-the-Flame Mapping 2
RESEARCH INSTALLATION 3
GENERAL CHARACTERIZATION OF THE FLAME 4
DATA ANALYSIS AND DISCUSSION 8
Baffle Burner 8
"Intermediate" Flame Baffle Burner 8
"Short" Flame Baffle Burner 18
Movable-Block Swirl Burner (I. F. R.F. Design) 27
Flat-Flame Burner 38
Boiler Burner 46
GENERAL OBSERVATIONS 57
11
-------
LIST OF FIGURES
Figure No. Page
1-1 Toroidal Recirculating Cell for a Typical Type I
Flame 6
I-Z Steady Pear-Shaped Recirculating Cell for a Typical
Type II Flame 7
1-3 Assembly Drawing of Axial-Flow Burner With
Ported Swirl Baffles 9
1-4 Modified Gas Nozzle Construction 10
1-5 Normalized NO Concentration as a Function of
Excess Air 12
1-6 Radial Profile for NO at the 5-cm Axial Position 13
1-7 Normalized NO Concentration as a Function of
Excess Air 15
1-8 Radial Profile for NO at the 5-cm Axial Position 16
1-9 Temperature Profile Across Furnace With Gas Input
of 2546 CF/hr and 5. 0-cm Axial Probe Position 17
I-10 Normalized NO Concentration as a Function of
Excess Air 19
1-11 Normalized NO Concentrations as a Function of
Excess Air 21
1-12 Radial Profile for NO at the 7. 6-cm Axial Position 22
1-13 Radial Temperature Profile at the 7. 6-cm Axial
Position 23
1-14 Radial Profile for NO Concentration for the Short-
Flame Baffle Burner Using the Axial Nozzle at a
7. 6-cm Axial Position 26
1-15 Diagram of a Movable-Block Swirl Burner 28
1-16 Arrangement of Swirl Generating Blocks in the
Movable-Block Swirl Burner 29
1-17 Normalized NO Concentration as a Function of
Excess Air 31
1-18 Normalized NO Concentration as a Function of
Excess Air 32
111
-------
LIST OF FIGURES, Cont.
Figure No. Page
1-19 Normalized NO Concentration as a Function of
Excess Air 33
1-20 Normalized NO Concentration Profiles as a
Function of Swirl Intensity and Excess Air on the
Movable-Block Swirl Burner With Gas Nozzle in
the Exit Position 34
1-21 Normalized NO Concentration Profiles as a Function
of Swirl Intensity and Excess Air on the Movable -
Block Swirl Burner With Gas Nozzle in the Throat
Position 35
1-22 Radial Profile of NO Concentration at the 12. 7-cm
Axial Position for the Movable-Block Swirl Burner
With Intermediate Intensity 36
1-23 Radial NO Concentration Profile for the Movalbe-
Block Burner at an Axial Probe Position of 12.7
cm and Set for Intermediate Swirl Intensity, Gas
Nozzle in Throat Position, and a Gas Input of
2008 CF/hr ' 37
1-24 Radial Temperature Profile at the 12. 7-cm Axial
Position 39
1-25 Cross-Sectional View of the Flat-Flame High-
Intensity Burner 41
1-26 Normalized NO Concentration as a Function of Excess
Air for the Flat-Flame Burner 43
1-27 Radial NO Concentration Profile for the Flat-Flame
Burner at a 12. 7-cm Axial Position 44
1-28 Radial Temperature Profiles for the Flat-Flame
Burner at Axial Positions of 12.7, 69, and 130 cm 45
1-29 Guide-Vane Boiler Burner 47
1-30 Method for Measuring Guide-Vane Angle for Boiler
Burner 48
1-31 Normalized NO Concentration as a Function of Excess
Air (Boiler Burner With a 30-deg Angle Vane
Setting; Gas Input, 3020 CF/hr) 49
IV
-------
LIST OF FIGURES, Cont.
Figure No. Page
1-32 Normalized NO Concentration as a Function of
Excess Air (Boiler Burner With a 40-deg Angle
Vane Setting; Gas Input, 3040 CF/hr) 50
1-33 Normalized NO Concentration as a Function of
Excess Air (Boiler Burner With a 60-deg Angle
Vane Setting; Gas Input, 3040 CF/hr) 51
1-34 NO Concentration Versus Swirl Number 53
1-35 Radial Profile for NO at a 60-deg Vane Angle
Setting 54
1-36 Radial Temperature Profile 55
1-37 Radial NO Concentration Profile for the Movable-
Block Burner at an Axial Probe Position of 12. 7 cm
and Set for Intermediate Swirl Intensity, Gas Nozzle
in Throat Position, 3. 6% Excess Oxygen, and a.
Gas Input of 2008 CF/hr 58
1-38 Radial CH4 Concentration Profile for the Movable-
Block Burner at an Axial Probe Position of 12. 7 cm
and Set for Intermediate Swirl Intensity, Gas Nozzle
in Throat Position, 3. 6% Excess Oxygen, and a Gas
Input of 2008 CF/hr 59
1-39 Radial O2 Concentration Profile for the Movable-
Block Burner at an Axial Probe Position of 12. 7 cm
and Set for Intermediate Swirl Intensity, Gas Nozzle
in Throat Position, 3. 6% Excess Oxygen, and a Gas
Input of 2008 CF/hr 60
1-40 Radial Temperature Profile at the 12. 7-cm Axial
Position 61
1-41 Radial Profile of Axial Velocity at the 12. 7-cm
Axial Position 62
1-42 Infrared Spectra Band Locations 63
1-43 NO Concentration Correction Factors for Gas
Samples Containing CaH4 and/or C3H6 65
1-44 Unsteady Pear-Shaped Recirculating Cell 69
1-45 Kidney-Shaped Recirculating Cell 70
-------
LIST OF TABLES
Table No. Page
1-1 Flame Characteristics 5
1-2 Mass Spectrometer Laboratory Analytical Report —
Natural Gas Sample 24
1-3 Mass Spectrometer Laboratory Analytical Report —
Gas Sample for Short-Flame Baffle Burner Axial
Gas Nozzle 25
1-4 Mass Spectrometer Laboratory Analytical Report —
Gas Sample for Movable-Block Swirl Burner 40
1-5 Vane-Angle Setting Versus Swirl Number for
Boiler Burner 46
1-6 Mass Spectrometer Laboratory Analytical Report —
Gas Sample for Boiler Burner 56
1-7 Time-Averaged Directional Flow Data Obtained at
the 12. 7-cm Axial Position 67
1-8 Time-Averaged Directional Flow Data at the
30. 5-cm Axial Position and Obtained Using a
Hubbard Probe 68
-------
INTRODUCTION
This volume of the final report for EPA Contract No. 68-02-0216
presents a synopsis of data collected from investigating the relationship
between combustion, aerodynamics, and pollution emission characteristics
of industrial burners. Five types of industrial burners were studied.
These were a) a utility power boiler burner, b) a high-intensity or flat-
flame burner, c) a movable vane burner design developed by the Inter-
national Flame Research Foundation, d) an axial flow burner, and e) a
partial baffle burner of a design widely used in steel reheat furnaces.
The utility boiler burner was scaled down to fit our experimental system.
The aerodynamic characteristics of each burner were investigated in con-
nection with their influence on the combustion process and on pollution
emissions. The research was divided into two categories:
1. Input-output tests in which the nitric oxide (NO) concentration in the
flue (as well as other flue gas components) was measured as a
function of changes in combustion conditions, such as percent excess
oxygen (air), preheat temperatures of the combustion air, and firing
rate of natural gas.
2. Point-by-point in-the-flame tests were made to measure gas species
concentrations, temperatures, and velocities (magnitude and direction),
in an attempt to determine where and how the NO was being formed.
A companion publication (Volume II) presents all of the raw data and
data plots collected during the program. Volume II also gives a detailed
description of the experimental facility, including dimensional descriptions
of the hot-modeling furnace, the cold-modeling furnace simulator, samp-
ling probes, instrumentation, and the test burners.
-------
SUMMARY OF CONCLUSIONS
From our analysis of the data obtained in this study (Volumes I and
II), we were able to reach the following broad conclusions. A compre-
hensive listing of the more specific conclusions reached for each burner
tested is given at the beginning of the section discussing that burner.
Input-Output Test Results
1. Type II* flames with radial gas injection as produced by the axial
burner and the boiler burner with radial gas nozzle will have peak
NO concentrations at an excess air level of 11% or less.
2. Type I* or Type II flames with axial gas injection, axial burner,
swirl burner, or flat-flame burner with axial gas nozzle will produce
their maximum NO emissions at an excess air level greater than 22%.
3. Although for all the burners tested the concentration of NO increases
with increasing preheat temperature of the combustion air, the mag-
nitude of the incremental change is dependent on the type of burner
and the method of gas injection.
4. Changing from axial to radial gas injection consistently resulted in
an increase in the NO concentration, which could be as large as
600 ppm.
In-the-Flame Mapping
1. All NO radial profiles measured for flames with Type I flow patterns,
typified by axial gas injection, resembled an upside down script "m"
(LA/U) in shape, while Type II flow patterns, typified by radial gas
injection, resembled a "W" in shape.
2. For burners with a Type II flow pattern, more than 50% of the flue
NO concentration is formed in the burner block.
3. Because only 50 to 80% of the flue NO concentration was measured
in the flame for both Type I and Type II flow patterns, we postulate
that secondary NO formation occurs either through a continuation of
the nitrogen-oxygen type reactions or because of thermal decomposi-
tion of
•K
A complete description of Type I and Type II flames is given in the
section, "General Characterization of the Flame, " beginning on page 4.
-------
RESEARCH INSTALLATION
All burners investigated were end-wall mounted on a 5x5x15 foot
cast refractory furnace. The combustion air was heated (100° to 600°F)
in an insulated electric preheater. Flue-gas samples were withdrawn
through a water-cooled stainless-steel probe and approximately 50 feet
of Teflon tubing into the various instrumentation remotely located from
the furnace. Concentrations of CO, CO2f CH4, and NO were measured
by multirange nondispersive infrared analyzers (NDIR). An electrolytic
analyzer was used to measure oxygen. The output signals from the NDIR
analyzers were time-averaged electrically and displayed on digital volt-
meters. Flow direction and magnitude in the furnace were measured
with both a five-hole pitot probe and a Hubbard probe. The pressure
differentials between probe tip holes used for determining flow direction
and velocity are measured with an electrical transducer. The signal from
the transducer is analyzed and displayed on a multirange electronic
manometer.
Temperatures are measured on a high-velocity suction pyrometer
containing a Ft — Pt 13% Rd thermocouple which has been calibrated
against flow rate through the tip. Output voltages from the thermocouple
are read directly by a millivolt electrometer. Volume II of this report
presents more detailed information on the installation.
-------
GENERAL CHARACTERIZATION OF THE FLAME
Table 1-1 gives a synopsis of the work being reported. The first
two columns of the table give the burner type and the kind of gas nozzle
being investigated. The characteristic flame type of the burner-nozzle
combination is listed in column 3. Columns 4 through 8 give the condi-
tions at which the furnace was set during the in-flame studies, with the
final two columns giving NO radial-profile data.
A general aerodynamic characterization of the burners can be made
by defining types of flow patterns observed in the flame. A flame may
be divided into four zones of interest: 1) the methane-rich central forward
flow zone, 2) the combustion air forward flow zones, 3) the internal
reverse flow zones, and 4) the secondary recirculation zones. For the
Type I flame shown in Figure 1-1, the methane-rich jet flows along the
axis of the burner and has a velocity large enough to penetrate the in-
ternal reverse-flow regions, which are represented by the lobes on each
side of the burner's axis. The combustion air circulates around the out-
side edge of each of the reversal lobes. The (— • — • —) lines in
Figure 1-1 represent the boundary layers between the forward flow of the
combustion air and the secondary recirculation patterns of the combustion
products. Characteristics of Type I flames found during our investiga-
tion are 1) a long, lazy, and luminous flame; 2) a high methane concen-
tration, not only on the axis of the burner but also in the internal re-
circulation lobes; and 3) in general, the stoichiometric ratio of fuel and
air accompanied by peak temperatures occurring on the outside edge of
the internal recirculation lobes.
The Type II flame is characterized by inability of the central methane
jet to penetrate the internal recirculation zone, as is shown in Figure
II-2. This flame type can be generated either by introducing the gas
radially with respect to the axis of the burner or by having an air flow
with such a large tangential-velocity component that it attaches to the wall
of the burner block and encloses a large toroidal vortex reverse flow
zone situated on the axis of the burner. This pattern enables a matching
of highly turbulent intensity zones with zones of high fuel concentration,
resulting in a large combustion intensity. Characteristics of the Type II
flame are 1) a short, translucent flame and 2) approximately 95% of the
combustion's occurring in the burner block.
-------
Table I-1. FLAME CHARACTERISTICS
Flame Characterization
Burner Type
Axial, Intermediate Baffle
Axial, Short Baffle
Swirl. Movable Block
Flat Flame
Boiler
Gas Nozzle
Axial
Radial
Axial
Axial
Axial
ntermediate
A'irl Intensity)
Axial
Radial
Flame Gas Input.
Type CF/hr
I 2147
II 2547
I 2190
I 2190
I 2008
II 2010
II 3049
Wall
Temp, °F
2570
2680
2570
2524
2453
2480
2534
NO in Flue Oz in
(Normalized), ppm Flue, %
255 2.6
(286)
741 1.7
(800)
189 2.9
(215)
325 3.0
(369)
111 3. 6
(130)
89 4.4
(109)
283 1.9
(308)
NO
Preheat Position,
Temp, °F cm
0
570 18
30
0
510 16
45
0
315 21
54
0
515 21
54
0
12
45
0
24
54
0
270 27
34
Profile
Concentration.
ppm
( 160)
54
142
502
344
666
142
6°
174
202
121
276
5
13
40
79
89
92
242
136
265
B-83-1308
-------
t
9
»
•
MEAN FLOW
DIRECTION
STREAM LINES
REVERSED AXIAL FLOW REGION
BOUNDARY OF RECIRCULATING REGION
STAGNATION POINT
A-83-II97
Figure 1-1. TOROIDAL RECIRCULATING CELL FOR
A TYPICAL TYPE I FLAME
-------
MEAN FLOW
DIRECTION
STREAM LINES
REVERSED AXIAL FLOW REGION
BOUNDARY OF RECIRCULATING REGION
STAGNATION POINT
A-83-II94
Figure 1-2. STEADY PEAR-SHAPED RECIRCULATING
CELL FOR A TYPICAL TYPE II FLAME
-------
DATA ANALYSIS AND DISCUSSION
Baffle Burner
Figure 1-3 illustrates the design of the baffle burner. The gas nozzle
lies parallel to and along the axis of the burner. It is inserted into the
ceramic baffle, thus ensuring that the gas will enter parallel to the axis
of the baffle burner. Combustion air enters perpendicular to the axis
and passes through the six ports in the baffle, which impart a swirl to
the air in some designs. Two of these baffle ports can be seen in
Figure 1-3 with their axes parallel to the axis of the burner. Air exiting
from the ports as illustrated would have an axial velocity component
only, resulting in a "long" flame. To shorten the flame length, a radial-
and tangential-flow component must be added to the combustion air vel-
ocity. This is accomplished by using a baffle where the combustion-air
ports are rotated relative to the axis of the burner. For the "intermediate"
flame length baffle the rotation orientation of the ports is 12 degrees and
for the "short" flame length, 24 degrees. Figure 1-4 shows the gas
nozzle modifications needed to achieve radial gas injection for the baffle
burner. Basically the nozzle end is capped and a series of holes are
drilled radially in the tube wall.
^Intermediate" Flame Baffle Burner
Based on an analysis of the input-output data for the intermediate-
flame baffle, we determined the following:
1. As expected, both increased gas input and increased preheat temper-
ature significantly increased NO emissions. The change in NO
emissions was more strongly affected by changes in preheat temper-
ature. On the average, NO emissions increased 300% for a 600°F
preheated air temperature increase and 25% for a 800 CF/hr increase
in gas input.
2. The emissions of NO as a function of the percent of excess air
(oxygen) showed the characteristic "bell-shaped" relationship. For
the radial gas nozzle, the NO emissions peaked at about 2% oxygen
(10% excess air). This peak shifted to higher excess-air levels
when the gas-air mixing rate was decreased (longer flames).
3. We found that the slow mixing gases of a long flame generally pro-
duce one-third as much NO as the fast-mixed gases of a very short
flame. When the burner nozzle was modified so that the combustion
mixing was slowed down, the flame length increased from about 2
feet to over 13 feet, and the NO concentration decreased from an
average of 700 ppm to 200 ppm.
-------
"A" PIPE SIZE
GAS NOZZLE
ll-l/4-in.-OD
BAFFLE
18 in.-
2 In.
in.
ll-l/2-in. DIAM PORT
I
SEAL BETWEEN
GAS NOZZLE a BAFFLE
AND BAFFLE 8 BODY WITH
R a I 3000 OR EQUAL
'D" DIAM
A-I03-I462
NO SCALE
BAFFLE
a. LONG FLAME
b. SHORT FLAME
c. INTER. FLAME
AIR PRESSURE FOR
40,000 SCF/hr at 850 °F
3.25 in. we
18 in. we
14 in. we
ii.ii
A
1-1/4 in.
3/4 in.
1 in.
"B"
6 in.
5 in.
8 in.
"C"
2-3/8 in.
2-3/8 in.
3-7/8 in.
"D"
13 in.
16-1/2 in.
13 in.
Figure 1-3. ASSEMBLY DRAWING OF AXIAL-FLOW BURNER
WITH PORTED SWIRL BAFFLES
-------
SCH 40-NOMINAL PIPE SIZE. A
C DIMENSION
PIPE THREAD
SI2E."B"
BAFFLE
LONG FLAME
SHORT FLAME
INTER FLAME
S
'A"
i
1/2
3/4
— j \
I J — •-) r*— 1/2 in.
~l 1
I SILVER SOLDER
4 '*-. ' END CAP
r i
I '
"o"
i
5ft Oin.
x-PORT Dl/
/ SIX ECU/
/<£-- ~O^ SPACED (
"B" "C" "D" "E" ff ^<^
1 I.Z5!g;gf 5 7/32 U U
1/2 0.75 !g go 4 7/32 U //
3/4 1.0 +°-°° 5-1/2 7/32 DO FIRST \\ //
~ ^^^•'S^ -^^V
fl.93-860
Figure 1-4. MODIFIED GAS NOZZLE CONSTRUCTION
-------
4. We also found that the change in NO emissions as a result of chang-
ing gas input is larger at higher preheat temperatures. Increasing
the gas input from 2300 CF/hr to 3100 CF/hr at a 100°F preheat
temperature generally resulted in a 75-ppm increase in NO. How-
ever, increasing the gas the same amount while using 450°F preheated
air resulted in an average NO emission increase of 150 ppm.
The flame from the axial burner with the intermediate baffle and the
axial gas nozzle exhibited a Type I directional flow pattern. The burner
block had a 2. 7-degree divergent angle with a 16. 5-cm radius at the in-
side edge of the furnace wall.
Typical measured NO concentrations in the flue as a function of
percent excess oxygen (air) using the axial burner, intermediate-flame
baffle, and axial gas nozzle are shown in Figure 1-5. These data were
gathered at a 2147 CF/hr gas input for combustion air preheat temper-
atures of 100°, 270°, and 570°F. The NO concentrations presented in
this paper as input-output test results were normalized by dividing the
weight of the flue products at the stoichiometric mixture of fuel and air
into the measured concentration of NO, and multiplying this ratio by the
weight of the flue products for the input conditions under which the mea-
surements were taken.
Each curve in Figure 1-5 displays a different positive slope for the
change in normalized NO concentration vs. percent of excess oxygen.
Note that no point of inflection is shown by these curves. Measured
concentrations of carbon monoxide greater than 100 ppm are denoted by
listing their numerical value in parts per million next to the appropriate
data point.
Figure 1-6 represents the radial profile for NO concentration in the
flame and the secondary recirculation zone at an axial position of 5 cm
from the front wall of the furnace. The data were collected at a gas
input of 2147 CF/hr with 2. 6% excess oxygen in the flue and a 570°F
preheat temperature of the combustion air. The NO concentration has a
maximum value of 162 ppm on the axis of the burner. The concentration
decreases to a minimum of 56 ppm at an 18-cm radial position and then
increases in value to 144 ppm in the secondary recirculation zone.
Assuming a symmetrical NO concentration profile close to the axis of
the burner, a shape of the profile would resemble a "W. " The central
11
-------
400
PREHEAT
TEMPERATURE,
»F
O 570
A 270
O 100
300
o.
O.
6
o
UJ
N
OC 200
o
100
NOTE.DATA OBTAINED USING AXIAL
BURNER, INTERMEDIATE BAFFLE,
AND AXIAL NOZZLE. GAS INPUT,
2147 CF/hr
60
3 4
02 IN FLUE,%
1 Figure 1-5. NORMALIZED NO CONCENTRATION
AS A FUNCTION OF EXCESS AIR
A-53-742
12
-------
w
Q
n
X
o
(J
» VS NO,
161.18
159.05
154.80
152.6B
150.56
146.tl
144.19
142.C6
rnrsr
I 35.69
1>3.57
111.44
127. 19
125.07
122.95
120.82
118.70
116.58
114.45
112.33
110.2C
10U.CB
105.96
103.83
101.71
19.56
13.21
11.C9
ob.S6
P4.72
12.59
30.47
If. 35
Ft.2?
71 .97
•-9.br>
(.7.73
c,5.f>C
61 .35
55.23
AXIAL BURNEf^lMERf'iCUIE BAFFLE CUARU PROBE . JEPT . 29.197?
5."CC
3.CCC
18.0CC 21.CCC 24.CCT
RADIAL POSITION, cm
Figure 1-6. RADIAL PROFILE FOR NO AT THE 5-cm AXIAL POSITION
(Intermediate-Flame Baffle — Axial Gas Nozzle — Gas Input, 2147 CF/hr —
Preheat Temperature, 570°F — 10% Excess Air)
-------
point of inflection of the "W" shape would be on the axis of the burner
and would contain a higher concentration of NO than the two end points
of the "W" corresponding to NO concentrations in the secondary recircu-
lation zone.
Figure 1-7 gives the normalized NO concentrations as a function of
percent excess oxygen in the flue for the "intermediate" flame baffle
burner with the radial gas nozzle at a gas input of 2335 CF/hr and com-
bustion air preheat temperatures of 100°, 295°, and 570°F. Because of
the improved mixing conditions with radial gas injection, the minimum
amount of excess air needed for proper combustion (defined as a flue
concentration of carbon monoxide below 100 ppm) decreased by approxi-
mately 5% from the amount needed with axial gas injection. Each curve
in Figure 1-7 shows that the point of maximum NO concentration shifts
toward higher percentages of excess oxygen as the air preheat temperature
is increased. In addition, the values of the measured NO concentrations
have increased by as much as 650 ppm from the levels measured with
axial gas injection.
A NO concentration profile for the axial burner with an intermediate-
flame baffle and radial gas nozzle is plotted in Figure 1-8. These data
were derived at a higher gas input of 2547 CF/hr with 10% excess air
at a 510°F preheat air temperature. On the axis of the burner the mea-
sured NO concentration was 502 ppm. The concentration value decreases
to a minimum of 346 ppm at a 16-cm radial position and then increases
to a rather constant value of 655 ppm in the secondary recirculation zone.
A symmetrical distribution of the NO profile about the axis of the burner
again displays a "W"-shaped curve. However, unlike the profile produced
by axial gas injection (Figure 1-6), the maximum NO concentrations oc-
curred in the secondary recirculation zones (at the ends of the W) rather
than on the axis of the burner (corresponding to the central peak of the
W).
A radial temperature profile for the furnace conditions listed in
Table 1-1 is shown in Figure 1-9. The temperature is relatively con-
stant in front of the burner block at about 3080°F. At a 28-cm radial
position, the temperature begins to decrease rapidly until the secondary
recirculation zone temperature of 2600°F is reached at a radial position
14
-------
900
800 —
PREHEAT
TEMPERATURE.
°F
300
NOTE: DATA OBTAINED USING AXIAL
BURNER, INTERMEDIATE BAFFLE,
AND RADIAL NOZZLE. GAS INPUT,
2335 CF/hr
,4000
200 —
100
0
2 3
02 IN FLUE,%
A-53-740
Figure 1-7. NORMALIZED NO CONCENTRATION
AS A FUNCTION OF EXCESS AIR
15
-------
PIFFLE - mri»L NCZZLE - SKINLESS -•<. ?c ic1 2.'?
cr-
E
P.
w
Q
X
O
u
t—I
H
M
2
13.JCO
so.scr
39.6CC
RADIAL POSITION, cm
Figure 1-8. RADIAL PROFILE FOR NO AT THE 5-cm AXIAL POSITION
(Intermediate Flame Baffle — Radial Gas Nozzle — Gas Input,
2547 CF/hr — Preheat Temperatures, 510°F - 10"^ Excess Air)
-------
3100
3000
2900
< 2800
cr
ui
Q.
5
UJ
2700
2600
2500
NOTE: DATA OBTAINED USING AXIAL
BURNER. INTERMEDIATE BAFFLE
AND RADIAL NOZZLE.
8
12
16
20
40
44
48
52
24 28 32 36
RADIAL POSITION, cm
Figure 1-9. TEMPERATURE PROFILE ACROSS FURNACE WITH GAS
INPUT OF 2546 CF/hr AND 5. 0-cm AXIAL PROBE POSITION
56
6O
A-II2-IO68
-------
of 40 cm. The slight decrease in temperature between the axis of the
burner and the 12-cm radial position may be caused by the central in-
ternal recirculation zone.
"Short" Flame Baffle Burner
Based on an analysis of the input-output data for the short-flame
baffle, we determined the following:
1. As expected, both increased gas input and increased preheat temper-
ature significantly increased NO emissions. The change in NO emis-
sions was more strongly affected by changes in preheat temperature.
On the average, NO emissions increased 200% for a 550°F preheat
air temperature increase and 20% for a 500 CF/hr increase in gas
input.
2. For the radial gas nozzle, the emissions of NO as a function of the
percent of excess air (oxygen) showed the characteristic "bell-shaped"
relationship. The NO emission peaked at about 2% oxygen (10% ex-
cess air). This peak shifted to higher excess air levels when the
preheat air temperature was increased.
3. We found that the slow mixing gases of a long flame generally pro-
duce one-half as much NO as the fast mixing gases of a short flame.
When the burner nozzle was modified so that the combustion mixing
was slowed down, the flame length was visually observed to increase
and the NO decreased from an average of 730 to 400 ppm.
Typical normalized NO concentrations in the flue as a function of
percent excess oxygen and combustion air preheat temperature for the
"short" flame baffle burner with a radial gas nozzle are shown in Figure
1-10. This is a Type II flame. The gas input during these meas-
urements was 2593 CF/hr. In comparison with the flue analysis data
for the "intermediate" flame baffle burner (Figure 1-7), there is an increase
in the measured NO concentrations from the "short" flame baffle burner
for the 100° and 300°F air preheat temperatures. Characteristically, the
maximum NO concentrations occur at higher percentages of excess air
as the preheat temperature of the air is increased.
The flame from the axial burner with the short baffle and an axial
gas nozzle exhibited a Type I directional flow pattern. The burner block
had an 8. 9-degree divergent angle with a 21-cm radius at the hot face.
18
-------
800
700
600
E
O.
O
UJ
N
cc
O
400
300
200
100
4499,
9780
TEMPERATURE PREHEAT
°F
NOTE- DATA OBTAINED USING AXIAL
BURNER, SHORT BAFFLE,
AND RADIAL NOZZLE.
GAS INPUT, 2593 CF/hr
0
2 3
02 IN FLUE, %
A-53-737
Figure I-10. NORMALIZED NO CONCENTRATION
AS A FUNCTION OF EXCESS AIR
19
-------
Figure I-11 illustrates normalized NO concentrations for a 2109 CF/hr
gas input as a function of excess air and preheat temperature for the
"short" flame baffle burner with axial gas injection. These graphs show
greater curvature and higher concentrations of NO than those obtained
from the intermediate baffle and axial gas nozzle (Figure 1-5). As was
the case for the intermediate baffle, the short-flame baffle with axial gas
injection consistently produced lower NO concentrations than with radial
gas injection for comparable levels of excess air.
A radial profile of the NO concentration at a 7. 6-cm axial position
with 15% excess air at a 3l5°F preheat temperature and a 2190 CF/hr
gas input injected axially is illustrated in Figure 1-12. The maximum
"measured" NO concentration occurs along the axis of the burner with a
215 ppm peak concentration. At the 21-cm radial position a minimum
value of 70 ppm is reached; the concentration then increases until it
reaches a rather constant value of 172 ppm in the secondary recircula-
tion zone. This profile resembles a "W" with the central inflection point
higher than the two end points.
Figure 1-13 shows the radial temperature profile for the same set of
furnace conditions used when measuring the NO profile in Figure 1-12.
It should be noted that a "cold" temperature of 2070°F occurs at the 3-cm
radial position, which corresponds to the location of the maximum
"measured" NO concentration. The "hot" spots (2680°F and 2870°F) occur
at +15 and —15 cm. (The stoichiometric ratio between oxygen and meth-
ane occurs between the radial positions of 15 and 18 cm and the radial
positions of —12 and —15 cm.) The time-averaged NO concentrations at
these points are 142 and 129 ppm, respectively. The central "cold" spot
corresponds to the maximum methane concentration, whereas the "cold"
spots at —18 and +21 cm are areas of relatively high oxygen concentra-
tions (8. 3% and 10. 6%).
To investigate the possible formation of hydrocarbons in the flame,
a grab sample was taken at the 0-cm radial position and at a 7. 6-cm axial
position. This sample was analyzed using a mass spectrometer; the re-
sults are listed in Table 1-2. The natural gas used to fire the furnace
was also analyzed and is shown in Table 1-3. A comparison of these tables
reveals that 0. 4 mol % of ethylene (CzFLi) and 0. 5 mol % of acetylene
were formed in the flame.
20
-------
500
400
£
Q.
Q.
O
LJ
M
2
IT
O
300
200
100
50
1,175
PREHEAT
TEMPERATURE,
°F
O 515
A 315
D 90
NOTE: DATA OBTAINED USING AXIAL
BURNER,SHORT BAFFLE,AND
AXIAL NOZZLE. GAS INPUT,
2109 CF/hr
02 IN FLUE,%
Figure 1-11. NORMALIZED NO CONCENTRATION
AS A FUNCTION OF EXCESS AIR
6
A-53-739
21
-------
SXIAL BU-INEK SHIIRT Sf»IM.ESi SHEPHEIO'S PRU6E. NOV.3, 1972
7.60
-17.200 -2S.800
-3.000
8.4UO
19.800
3l.2!'0
RADIAL POSITION, cm
Figure 1-12. RADIAL PROFILE FOR NO AT THE 7. 6-cm AXIAL POSITION
(Short Flame Baffle — Axial Gas Nozzle — Gas Input, 2190 CF/hr — Preheat
Temperature, 315°F - 3% Excess Air)
-------
29
28
27
26
CNJ
o
UJ
cr:
25
24
a.
5
LJ
23
22
21 —
20
I
I
I I
65 55 35 15 5 0-5 -15
RADIAL POSITION,cm
-35
A-I2Z-I233
Figure 1-13. RADIAL TEMPERATURE PROFILE AT THE 7.6 cm
AXIAL POSITION (Short Flame Baffle - Axial Gas Nozzle -
Gas Input, 2190 CF/hr — Preheat
Temperature, 310°F - 3.3% Excess Oxygen)
23
-------
Table 1-2
MASS SPECTROMETER LABORATORY
ANALYTICAL REPORT
M.iie,,.il 8933 Sample # 2 11/16/7Z Oale 11/17/7Z
3289
by - M. S.
Uol X Mol '
0.4
C.l.hon Un,in.:ilP . 4'°
C.vhoii Diomde 2- ?
• 2
Ethylene
Ethyl
Slyrenc
Indene
Napthalene
C.ilc. H. V., Bin SCF Air i
C.ilc. sp U'(Alf 1.000) ^^_^__^^^___ Approved by
Argon "• 3
W.ilcr Vnpor ___^^______
Hplinin _^_^___^_^__ Cyclopenladiene _^_—_
\o (. 05
Mctli.ine JV- b Acetylene LL
TOTAL 10°-°
24
-------
Table 1-3
MASS SPECTROMETER LABORATORY
ANALYTICAL REPORT
M.-Mrri.il 8933 Natural Gas from Pilot Plant 0a|
I by - M. '
11/8/72
3238
C.lrhnii Mnnni:H»
C.lrhou I
Argon
Wnlcr Vapor
Helium
Elhniie
IsolniUiiir
Pcnl.ines
Heiancs
C.ilc. H. V., Bin SCF
Cnlc. sp ijr (Air 1.000
Uol X
'
91.74
3-93
0-
°-07
0. 06
0.03
0-02
Ethylenr
P'opylene
Butenes
Pentenes
Heienes
1,3-Buladiene
Ppiiladienes
Cyclopentadienp
Acetylene
Methyl Acetylene
Vinyl
Ethyl
Styrene
Indene
TOTAL
Air I
Approved by
Mol %
100. 0
25
-------
Because the central peak of the radial NO concentration profile
closely resembled the methane profile and because the maximum NO con-
centration occurred in one of the "coldest" regions in the flame, the NDIR
analyzer was investigated for possible optical interference between NO
and another molecule in the gas sample.
We discovered that ethylene (CzH4) and propylene (CjHf,) would optically
interfere with the optical analysis of NO concentrations by the NDIR tech-
nique. The intensity of the interference was measured as 100 ppm of
ethylene (CzH^, indicating a 2. 1 ppm concentration of NO, and 100 ppm
of propylene (CaHt) being analyzed as 1. 1 ppm of NO. This analysis
indicated that a correction for the NO data must be applied to measure-
ments made in regions of high hydrocarbon concentrations. (A complete
discussion of the correction method is given later in this report. ) The
measured profile, when corrected (Figure 1-14), is very similar in shape
to that of the corresponding temperature profile, Figure 1-13.
170
E 150
Q.
Q.
tr
t-
z
UJ
o
o
o
130
110
90
70
I
I
I
40 30 20 10 0 -10
RADIAL POSITION, cm
-20
-30 -40
A-83-M90
Figure 1-14. RADIAL PROFILE FOR NO CONCENTRATION
FOR THE SHORT-FLAME BAFFLE BURNER USING THE AXIAL
NOZZLE AT A 7. 6-cm AXIAL POSITION. GAS INPUT, 2190 CF/hr;
EXCESS OXYGEN, 3.0%; PREHEATED AIR, 315°F
26
-------
The "hot" regions (2680°F and 2870°F) of the flame occur at +15 and
—15 cm while the radial position of -1-15 and —6 cm correspond to the two
maximum internal inflection points of the NO profile.
Movable-Block Swirl Burner (I. F. R. F. Design)
Figure 1-15 shows the design of the movable-block swirl burner.
The fuel gas is introduced through a 3/4-inch pipe along the axis of the
burner. The combustion air enters perpendicular to the axis of the bur-
ner, passes through an array of swirl generating blocks, and exits into
a 0-degree angle burner block with a 7. 6-cm radius. The swirl gener-
ating blocks are arranged as shown in Figure 1-16. There is a total of
sixteen blocks; eight are fixed to the burner, and the remaining eight
are mounted on a movable plate. If the plate is positioned in its maxi-
mum clockwise rotation, the tangential entry channels are completely
blocked. Thus, the combustion air has only an axial velocity component.
However, if the plate is positioned at its maximum counterclockwise
rotation, the air can enter only through the tangential channels. If the
movable blocks are adjusted to some intermediate position, the combus-
tion air enters with some combination of axial and tangential velocity
components.
The swirl burner was operated at three different swirl intensities
for the input-output tests, with two gas nozzle positions for each swirl
intensity. For the first gas nozzle position, the nozzle tip was located
even with the inside edge of the burner wall (hot face), while in the second
position the nozzle tip was withdrawn into the burner block, 6 inches
from the hot face wall. (For the remainder of this report, these posi-
tions will be referred to as the "exit position" and "throat position, "
respectively. ) The input-output tests were conducted at gas inputs of
1578, 1976, and 2382 CF/hr, with between 10 and 80% of excess air.
The input-output data from the movable-block swirl burner showed
the following:
• The maximum measured NO concentration occurred at the lowest
levels of gas input (1578 CF/hr) and swirl density.
• At excess oxygen levels below 6% , generally more NO was formed
when the gas nozzle was in the throat position than when it was in
the exit position. Insufficient data are available to evaluate the
relative effect of burner nozzle position when operating with more
than 6(#- excess oxygen.
27
-------
CONNECTING FLANGE
oo
SWIRL GENERATOR
BLOCKS
-12 in.
A-23-290
Figure 1-15. DIAGRAM OF A MOVABLE-BLOCK SWTRL BURNER
-------
A-23-291
Figure 1-16. ARRANGEMENT OF SWIRL GENERATING
BLOCKS IN THE MOVABLE-BLOCK SWIRL BURNER
29
-------
• Increasing gas input (and consequently gas velocity) always reduced
the normalized concentration of NO independent of swirl intensity
and percent excess air, when the burner was in the throat position.
However, when the nozzle was in the exit position, changing gas
input had little or no effect on the normalized NO concentration.
This was observed for intermediate and high swirl intensity. Insuf-
ficient data were obtainable for the case of low swirl intensity.
Figure 1-17 shows normalized NO concentrations as a function of the
amount of excess air and the gas nozzle position for high swirl "intensity
(axial entry channels closed) with a 1976 CF/hr gas input. Figures 1-18
and 1-19 show normalized NO concentrations as a function of excess air
and gas nozzle position for intermediate and low swirl intensities.
A comparison of Figures 1-17,1-18, and 1-19 at 8-22% excess air
(the region of practical interest) discloses several observations concern-
ing the effect of swirl on NO production. For the throat and exit gas
nozzle positions, the minimum NO concentration (76 and 34 ppm) are
generated by an intermediate swirl intensity with 2% excess oxygen,
Intermediate swirl intensity also produces the maximum NO concentrations
of 146 ppm with the gas nozzle in the throat position and 82 ppm with
the nozzle in the exit position at an excess air level of 22%. Normalized
NO concentration profiles as a function of swirl intensity and excess air
are shown in Figure 1-20 for the nozzle in the exit position and in
Figure 1-21 for the nozzle in the throat position.
Figure 1-22 displays the radial NO concentration profile at a 12. 7-cm
axial position for intermediate swirl intensity with the gas nozzle in the
throat position. The data were collected at a gas input of 2008 CF/hr
with 3.6% oxygen in the flue and no air preheat (100°F).
The NO profile corrected for the optical interferences of ethylene
and propylene is presented in Figure 1-23. The corrected data are rep-
resented on the NO transparency (Figure 1-37) by the dashed line. As
was the case for the "short" flame baffle burner, the corrected NO pro-
file is similar in shape with the temperature profile. The "hot" regions
of the flame (2309°F) occur at +11 and —5 cm while the radial position
of -111 and —4 cm correspond to the internal peaks in the NO concentra-
tion profile.
30
-------
130
no
90
o.
ex
O
LL)
N 70
CE
O
Z
50
30
10
O THROAT
V EXIT
3 4
02 IN FLUE,%
A-23-293
Figure 1-17. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Movable Block Swirl Burner -
High Swirl Intensity - Gas Input, 1976 CF/hr)
31
-------
170
ISO
130
110
E
Q.
Q.
S 90
N
70
50
30
10
O THROAT
V EXIT
02INFLUE,%
Figure 1-18. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Movable-Block Swirl Burner -
Intermediate Swirl Intensity — Gas Input, 1976 CF/hr)
7
A-23-297
32
-------
170
ISO
130
110
o.
O.
S 90
N
70
50
30
10
123456
02 IN FLUE,%
Figure 1-19. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Movable Block Swirl Burner -
Low Swirl Intensity - Gas Input, 1976 CF/hr)
A-23-296
33
-------
90
I. 80
Q.
O 70
S
tr
UJ
o
Q
UJ
N
60
O 50
o
40
30
20
10
% EXCESS AIR
0 8
A 14
D 22
LOW INTERMEDIATE
SWIRL INTENSITY
HIGH
A-53-736
Figure 1-20. NORMALIZED NO CONCENTRATION PROFILES
AS A FUNCTION OF SWIRL INTENSITY AND EXCESS AIR ON
THE MOVABLE-BLOCK SWIRL BURNER WITH GAS
NOZZLE IN THE EXIT POSITION
34
-------
150
140
E
Q.
Q.
- 130
o
UJ
u
2
O
Q
UJ
M
ee
O
z
120
110
100
90
80
70
% EXCESS AIR
O 8%
A 14%
D 22%
LOW INTERMEDIATE
SWIRL INTENSITY
HIGH
A-53-738
Figure 1-21. NORMALIZED NO CONCENTRATION PROFILES
AS A FUNCTION OF SWIRL INTENSITY AND EXCESS AIR ON
THE MOVABLE-BLOCK SWIRL BURNER WITH GAS
NOZZLE IN THE THROAT POSITION
35
-------
30 20 10
RADIAL POSITION, cm
-10
A-83-II89
-20
Figure 1-22. RADIAL PROFILE OF NO CONCENTRATION
AT THE 12.7-cm AXIAL POSITION FOR THE MOVABLE-BLOCK
SWIRL BURNER WITH INTERMEDIATE INTENSITY. GAS INPUT,
2008 CF/hr; EXCESS OXYGEN, 3.6%; NOZZLE IN THROAT POSITION
36
-------
RP
ft
P
(X
P.
w*
p
x
o
u
*"j
\L{
H
^
VS ^0,
131.31
126.73
126.16
123.56
121.00
116.43
113. 28
110. 70
108. 13
105.55
102.96
100.40
17.63
95.25
12.6U
90. 10
HI.':}
04.95
U2. )7
(7.22
14.65
12. 0»
69.50
66.92
04.35
61.77
59.20
56.62
54.04
51.47
46.69
46. 32
38.59
36.02
30.67
28.29
25.72
23.14
20.56
17.99
15.41
12.64
10.26
7.69
5.11
2.5*
-0.03
KOVtlBlE BLOCK SKIRL BURNER - INTtRMEOUIt SWIKL - STAINLESS SHEPHERD'S PROB6
«P« 12.70
N
-15.000
-9.000 -3.000
3.000
\s
15.000
21.000
27.000
33.000
39.000
45.000
RADIAL POSITION, cm
Figure 1-23. RADIAL NO CONCENTRATION PROFILE FOR THE
MOVABLE-BLOCK BURNER AT AN AXIAL PROBE POSITION
OF 12.7-cm AND SET FOR INTERMEDIATE SWIRL INTENSITY,
GAS NOZZLE IN THROAT POSITION, 3.6% EXCESS
OXYGEN, AND A GAS INPUT OF 2008 CF/hr
37
-------
The radial temperature profile at the 12. 7-cm axial position is shown
in Figure 1-24. The methane central "cold" spot occurs at a 3-cm radial
position. The combustion-air "cold" spots correspond to radial positions
of—6 and +13 cm with 15.2 and 15.9% concentrations of oxygen, respectively.
The maximum temperatures are measured at +11 and —1-cm radial posi-
tions, while the stoichiometric ratios of fuel and air occur at +11.5 and
—5. 5 cm.
A visualization of these variations in temperature, velocity, Oa, CH4,
and NO as a function of radial position at the 12. 7-cm axial position can
be aided by the accompanying transparencies.
To investigate the presence of higher hydrocarbons in the flame, a
grab sample was taken at the —1. 0-cm radial position and a 12. 7-cm
axial position. This sample was analyzed using a mass spectrometer
and the results are listed in Table 1-4. In addition to the ethylene and
acetylene which were measured for both the axial burner "short" flame
baffle (Table 1-2) and the movable-block swirl burner (Table 1-4), the
swirl burner also produced 0. 2 mol % of propylene.
Flat-Flame Burner
When penetration of the flame into the furnace is not desired, a flat-
flame burner can be used. Figure 1-25 illustrates a partial cutaway view
of a nozzle mixing burner of this type. The flat-flame or high-intensity
burner actually heats its own refractory burner block tile and the re-
fractory surface of the surrounding furnace wall, primarily by convection,
from the high-velocity combustion gases thrown sideways from the burner.
These hot refractory surfaces then radiate heat to the furnace load.
Because the hot gases have no final velocity component along the burner
axis, an extreme case of Type II flame pattern results.
The flat-flame burner was operated at three different gas inputs, all
over a range of fuel/air ratios expressed as percentage of oxygen in the
flue. No changes were made in the burner nozzle position or the swirl
vanes in the burner housing because these were fixed by design. The
38
-------
14
12
840
RADIAL POSITION,cm
-4
-8
Figure 1-24. RADIAL TEMPERATURE PROFILE AT THE
12. 7-cm AXIAL POSITION (Movable-Block Swirl Burner -
Intermediate Flame Intensity — Nozzle in Throat Position —
Gas Input, 2008 CF/hr - 3. 6% Excess Oxygen)
-12
A-23-301
39
-------
Table 1-4.
MASS SPECTROMETER LABORATORY
ANALYTICAL REPORT
H'JJJ Furnaci- Product CJas
Maten.il Radial Position. -1 cm; Axial Position, Oi11c 1/11/73
..--O- K.
Uol \
Nitrr)rjpii
C.lrhon Unnnicrtp _,,__.
Nil-,,™*™
Oiyrjpn '
3-8
L.Trlinn l)in«irt»
4 0
HyrlrnijPii . ,
0.6
Arijnn ________
W.llrr V.npnr
Hpliiini _________
12. 5 cm 3,.c.4
U Q 0,,,, Wn
Mol '.
_ Fthylfnt _
M-Bi-'-itliPiir
_ PPIII.II|IPI|P<, __._____^,
Cvrlonpnl^rtipup
31.8 0. 4
Fllnnp ' niarplulpnr
Prnp.T.p ° '
•••Diil.niP ,. __. ,..,
Knlinl.iiir ... ,. _
IV 1 HIP*
_ Uplllyl A/-o|ylono _________________
_ Vinyl Arpfylfnp __________..,_,.
Tnl.ipnp ._.._
Flhyl Rpii7PiiP _ _,
Styrpup
Indenp
r.iir. H. v. Rh, ^rp
T.nlr. -;p ijr (Air 1,000)
_ Naplhalpnp
1000
Air fnnlpnt _
40
-------
I DIA.-8 HOLES EQUALLY SPACED
STRADDLE it's AS SHOWN
f PIPE PLUG WHEN
PILOT IS NOT USED
? BOLT 8 NUT
TACK WELD HEAD TO FCE
Figure I-Z5. CROSS-SECTIONAL VIEW OF THE FLAT FLAME HIGH-INTENSITY
BURNER
41
-------
input-output tests were conducted at gas inputs of 1670, 2010, and 2394
CF/hr, with between 1. 0 and 7. 0% oxygen in the flue by volume.
Figure 1-26 shows the normalized NO concentrations measured in
the flue as a function of excess air and gas input for the flat-flame
burner.
Based on the analysis of the input-output results, we concluded the
following:
1. From 1. 0 to about 3. 75% excess oxygen in the flue, the gas input
rate made little difference in the amount of NO formed so long as
the flame had a visible appearance of being flat. Spot-check runs
for NO in the flue gases at gas inputs below 1670 CF/hr (where the
flame lost its flat appearance) showed differences as gas input was
changed. However, the flame was very lazy and concentration read-
ings erratic. Definite measurements could not be made, only gross
differences observed. Consequently, measurements at gas inputs
below 1670 CF/hr were not pursued further.
2. The NO concentration at all gas inputs and excess air levels tested
was considerably lower for the flat-flame burner than any other
"commercial type" burner tested. As the gas input was increased,
the peak NO concentrations decreased in magnitude and occurred at
lower percentages of excess oxygen. We postulate that this occurs
because at higher inputs the flame is held closer to the burner block
and the wall of the furnace. Because of increased heat transfer,to
these walls, the combustion area is cooled. This cooling effect, in
turn, tends to reduce the energy available to the NO -forming reactions.
X
The NO concentration radial profile at a 12.7-cm axial position is
given in Figure 1-27. We observe very little curvature or noticeable
shape in this profile; however, an increase occurs in the rra gnitude of
the average NO concentration moving across the furnace from left to right.
Hydrocarbon correction of these data was not necessary. The measured
data points can be considered to be within the interval concentration level
of 74 ppm ±10. The average concentration of this profile is 77 ppm.
As compared with the flue concentration of 88 ppm, this would seem to
indicate that most of the NO was formed during the initial combustion
processes inside the burner block.
The radial temperature profiles for axial positions of 12. 7, 69, and
130 cm are displayed in Figure 1-28. At the 12. 7-cm axial position, a
minimum occurs on the axis of the burner where the main contributions
would be caused by recirculating gases. There is only an 80°F difference
between this minimum temperature and the maximum temperature which
42
-------
110
OJ
I
90
E
o.
a
O
r
o
UJ
a:
O
70
50
O 1670 CF/hr
A 2010 CF/hr
O 2394CF/hr
30
3 4
02 IN FLUE, %
A-53-743
Figure 1-26. NORMALIZED NO CONCENTRATION AS A FUNCTION
OF EXCESS AIR FOR THE FLAT-FLAME BURNER
-------
RP
FLAT FLAME BURNER - STAINLESS SHEPHERO.S PROBE
»P- 12.70
12.000
24.000
36.000
48.000
60.000
Figure 1-27. RADIAL NO CONCENTRATION PROFILE FOR
THE FLAT FLAME BURNER AT A 1Z. 7-cm AXIAL POSITION
44
-------
Ul
27
CVI
O
X
LU
^
£ 25
Q.
24
WALL TEMPERATURE: 2510 °F
AXIAL POSITION, cm
I I
55
35 15 5 0 -5 -15
RADIAL POSITION, cm
-35
-55
A-53-741
Figure 1-28. RADIAL TEMPERATURE PROFILES FOR THE FLAT
FLAME BURNER AT AXIAL POSITIONS OF 12. 7, 69, AND 130 cm
-------
occurs at a Z4-cm radial position. The radial temperature profiles at
the 69 and 130-cm axial positions are essentially the same, maintaining
a fairly constant value of 2480° ± 20°F.
Boiler Burner
A guide-vane boiler burner is illustrated in Figure 1-29. The com-
bustion air enters perpendicular to the axis of the burner and then passes
through a register of guide zones which impart a degree of swirl dependent
on their orientation before entering the burner block. Figure 1-30 illus-
trates how the angle of the guide vanes is measured. The burner block
has a 30-degree divergent angle with a 45-cm diameter exit into the fur-
nace. All the tests presented here were conducted using radial gas in-
jection with a gas throughput of 3040 CF/hr.
Normalized NO concentrations in the flue as a function of excess air,
combustion air preheat, and vane-angle setting are given in Figures 1-31,
1-32, and 1-33. In Figure 1-31 at a 30-degree vane-angle setting, the
NO curves have very little curvature and are similar in intensity to the
short-flame baffle burner with the axial gas nozzle (Figure I-11). For a
40-degree vane-angle setting (Figure 1-32), there is a large (52%) increase
in the magnitude of the peak NO concentration for the 530°F preheat
temperature as compared with a 31% increase in peak concentration for
the 270°F preheat. Figure 1-33 displays the normalized NO concentra-
tion in the flue for a 60-degree vane-angle setting. For the 265° and
530°F preheat temperatures, the shape and magnitudes of the NO profiles
remain relatively unchanged from those obtained at a 40-degree vane-
angle setting. However, for ambient air temperature, the nitric oxide
has a positive slope whereas for the40-degree angle vane it had no slope.
The swirl numbers corresponding to the vane-angle settings investi-
gated during this project are listed in Table 1-5.
Table 1-5. VANE-ANGLE SETTING VERSUS
SWIRL NUMBER FOR BOILER BURNER
Vane Angle, deg Swirl Number
30 0.52
40 0. 71
60 1.22
46
-------
A-83-II99
Figure 1-29. GUIDE-VANE BOILER BURNER
47
-------
Figure 1-30. METHOD OF MEASURING GUIDE
VANE ANGLE FOR BOILER BURNER
48
-------
500
E 400
Q.
Q.
o"
Q
iH 300
O
2 200
100
2 3
% 02 IN FLUE
550°F PREHEAT
285°FPREHEAT
5 6
A-63-934
Figure 1-31. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Boiler Burner With a 30-deg
Angle Vane Setting; Gas Input, 3020 CF/hr)
49
-------
750
700
600
E
o.
o.
•»
O
UJ
N
500
400
g 300
200
100
530°F PREHEAT
234
% 02 IN FLUE
265°F PREHEAT
85°F PREHEAT
5 6
A-63-933
Figure 1-32. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Boiler Burner With a 40-deg
Angle Vane Setting; Gas Input, 3040 CF/hr)
50
-------
700
600
I 500
»
o
z
o
M 400
o
z
300
200
150
530°F PREHEAT
265°F PREHEAT
85°F PREHEAT
2 3
% 02 IN FLUE
4 6
A-63-932
Figure 1-33. NORMALIZED NO CONCENTRATION AS A
FUNCTION OF EXCESS AIR (Boiler Burner With a 60-deg
Angle Vane Setting; Gas Input, 3040 CF/hr)
51
-------
Figure 1-34 presents a set of curves plotting normalized NO concen-
trations versus swirl number. All the curves were drawn for 2% excess
oxygen with each curve representing a different air preheat temperature
whose values are labeled. There is no appreciable difference between
the curves shown in Figure 1-34 and the curves which would represent a
3% excess oxygen level. From Figure 1-34 we conclude that the swirl
number producing the maximum concentration of NO, independent of the
preheated air temperature, would be 1.0. The swirl number corresponds
to a 53-deg vane angle.
The radial NO profile is plotted in Figure 1-35 for the boiler burner
with a 60-deg vane angle setting. These data were collected at a 3039
CF/hr gas input with 8% excess air at a Z70°F preheat temperature. On
the axis of the burner the measured NO concentration was 242 ppm.
The concentration value decreased to a minimum of 136 ppm at a 27-cm
radial position and then increased to a rather constant value of 260 ppm
in the secondary recirculation zone. A symmetrical distribution of the
NO profile would produce a "W"-shaped curve.
The radial temperature profile in Figure 1-36 shows a shape char-
acteristic of Type II flames. It has a relatively flat maximum, 2840°F,
across the central portion of the burner block, and then falls to a con-
stant temperature of 2560°F in the secondary recirculation zone. The
large temperature gradients occur in the region of forward flow out of
the burner block, with the maximum temperature plateau corresponding
with the internal recirculation zone.
Table 1-6 presents a chemical species analysis of a grab sample
taken in the furnace at a 12.7-cm axial position and a 30-cm radial
position. This radial position was selected because the flame displayed
a Type II flow pattern with a region of forward flow within the radial
positions of 24 cm and 34 cm. Hydrogen and carbon monoxide were the
only combustibles measured. By comparison, the Type I flames, investi-
gated for combustibles within the furnace (Tables 1-2 and 1-4), not only
contained hydrogen and carbon monoxide but also methane, ethane, pro-
pane, normal-butane, ethylene, propylene, and acetylene.
52
-------
800
700
600
Q.
O.
»
O
z
O
HI
N
-------
BOILER BURNER - RADIAL CAS NOZZLE - BLUNT STAINLESS PROBE
AP= 12.70
-I?.000
-4.800
9.600
16.800
2<.,000
52.800
Figure 1-35. RADIAL PROFILE FOR NO AT A 60-deg
VANE ANGLE SETTING (Boiler Burner - Gas Input, 3039 CF/hr -
Preheat Temperature, 270°F - 8% Excess Air)
-------
29
28
CM
O
27
UJ
CC
5
Sj 26
0.
25
24
WALL TEMPERATURE: 2534 °F
PREHEAT: 270°F
I
55
35 15 5 0 -5
RADIAL POSITION, cm
-15
-35
A-83-II92
Figure 1-36. RADIAL TEMPERATURE PROFILE
55
-------
Table 1-6. MASS SPECTROMETER
LABORATORY ANALYTICAL REPORT
Material: 8933 Furnace Gas Sample Date: 3/16/73
Requested by: M.S. Run No. : 3875
Mol %
Nitrogen 61. 3
Carbon Monoxide 3. 3
Oxygen 0. 15
Carbon Dioxide 6. 3
Hydrogen Z. 3
Argon 0. 74
Water Vapor 25. 9
Helium 0. 01
Total 100.00
56
-------
GENERAL OBSERVATIONS
Although the "measured" radial profiles of NO concentrations for
Type I flames consistently resembled a "W" in shape, with the central
inflection point being higher than the two end points, these profiles gave
rise to two questions: 1) Why does the central peak in the NO profile
closely resemble the methane profile, and 2) why does the maximum NO
concentration occur in one of the "coldest" regions of the flame? Both
of these situations contradict the work of other investigators. (See
Figures 1-37 through 1-41.)
One possible answer was that of an optical interference within the
NDIR analyzer between NO and another molecule in the gas sample being
examined. Water has such a strong optical interference with NO.
However, being aware of this, we very carefully dried each gas sample
before allowing it to enter the analyzer. Before beginning this investi-
gation, dried air, methane, CO, and CO2 were passed through the NO
analyzer without detectable interference. However, several molecular
species were measured in the flame which had not been investigated for
optical interferences. These included ethane, propane, normal-butane,
ethylene, propylene, and acetylene as shown in Tables 1-3 and 1-4.
Figure 1-42 shows the infrared spectra of these species as well as the
infrared spectra of CO, CO2, CH4, NO, and NO2. The lines in Figure
I-4Z indicate only the location of the center of the infrared bands, but
do not represent the width of the bands. The NDIR analyzers used the
infrared spectra bands with a wave number of 1876 cm"1 for NO, 2143
cm"1 for CO, 2350 cm"1 for CO2, and approximately 3020 cm'1 for CH4.
The CO, CO2, and CH4 analyzers have no optical filters and use a quartz
window for transmission. Unlike these analyzers, the NO analyzer has a
filtering system which passes light within the range of wave numbers
1660 cm-1 to 2060 cm"1. Figure 1-42 indicates that possible optical inter-
ferences with the 1876 cm"1 NO band are the 1932 cm"1 band for CO2,
1936 cm"1 band for propane, 1890 cm"1 band for ethylene, and the 1830
cm"1 band for propylene. To test for optical interference of thes.e bands
with the 1876 cm"1 band of NO a sample of each gas was measured by the
NO analyzer. Only the propylene and ethylene samples indicated any
interference. To determine if any of the propylene or ethylene would be
57
-------
8 4 0-4-8
RADIAL POSITION, cm
-12
-16
Figure 1-37. RADIAL NO CONCENTRATION PROFILE FOR THE
MOVABLE-BLOCK BURNER AT AN AXIAL PROBE POSITION OF
12.7 cm AND SET FOR INTERMEDIATE SWIRL INTENSITY, GAS
NOZZLE IN THROAT POSITION, 3.6% EXCESS OXYGEN
AND A GAS INPUT OF ZOOS CF/hr
58
-------
1.4
q
d
ID
UJ
O
z
o
o
UJ
z
UJ
1.2
1.0
0.8
0.6
0.4
0.2
CH4(METHANE)
-BURNER WALL-
I I I
16 12 8 4 0-4 -8
RADIAL POSITION, cm
-12 -16
Figure 1-38. RADIAL CH4 CONCENTRATION PROFILE FOR THE
MOVABLE-BLOCK BURNER AT AN AXIAL PROBE POSITION OF
12. 7 cm AND SET FOR INTERMEDIATE SWIRL INTENSITY, GAS
NOZZLE IN THROAT POSITION, 3.6% EXCESS OXYGEN,
AND GAS INPUT OF 2008 CF/hr
59
-------
1.4
o^
N-'
g
55
UJ
o
z
o
o
UJ
o
X
o
1.2
1.0
0.8
0.6
0.4
0.2
OXYGEN
-BURNER WALL-
16 12 8 4 0-4 -8
RADIAL POSITION, cm
-12 -16
Figure 1-39. RADIAL O2 CONCENTRATION PROFILE FOR THE
MOVABLE-BLOCK BURNER AT AN AXIAL PROBE POSITION
OF 12.7 cm AND SET FOR INTERMEDIATE SWIRL INTENSITY,
GAS NOZZLE IN THROAT POSITION, 3.6% EXCESS OXYGEN,
AND A GAS INPUT OF 2008 CF/hr
60
-------
u.
o
o
ro
<*•
CO
LJ
tr
a:
UJ
a.
S
UJ
1.4
1.2
1.0
0.8
0.6
0.4
0.2
TEMPERATURE
-BURNER WALL-
I I I
16 12 8 4 0-4 -8
RADIAL POSITION, cm
-12
-16
Figure 1-40. RADIAL TEMPERATURE PROFILE AT THE 1Z. 7-cm
AXIAL POSITION (Movable-Block Swirl Burner - Intermediate
Intensity; Gas Input, 2008 CF/hr; Excess Oxygen, 3.6%;
Nozzle in Throat Position)
61
-------
-0.4
8 4 0-4-8
RADIAL POSITION, cm
-12 -16
8-73-1142
Figure 1-41. RADIAL PROFILE OF AXIAL VELOCITY AT THE
12.7-cm AXIAL POSITION (Movable-Block Swirl Burner - Intermediate
Intensity; Gas Input, ZOOS CF/hr; Excess Oxygen, 3.6%;
Nozzle in Throat Position)
62
-------
M.
1
S.
I
W.
1
S.
1
M.
I
S.M.
II
S.
1
M.
(1830)
M. S. M.
II 1
vs.
1
S.
I
(1890)
W.
I
w.
1
S. S.
1 1
vs.
1
S.
1
W.
M.
S.
1
C2H4
C2H2
S.
1
M.
|
M.
|
M.M.
||
S.
|
S.
1
w.
1
w.
1
M.
1
S.
1
C2H6
N02
C02
CH4
CO
NO
S.
S.
1
S.M.
II
S.
1
1
vs.
1
M.
1
vs.
1
M. M. VjS.
1 1 (2350)
S.
1
1
1
1(3020)
(2143)
(\B
I
INTENSITY LEGEND
VS.= VERY STRONG
S.= STRONG
M.= MEDIUM
W. = WEAK
I
500
1000
1500 2000
WAVE NUMBER, cm'1
2500
3000
3500
6-83-1198
Figure 1-42. INFRARED SPECTRA BAND LOCATIONS
-------
removed oy the gas-drying system, the tests were performed again intro-
ducing the gas samples into the drying system before they were analyzed.
The test results were identical to those of the previous investigation. A
graphical representation of these experimental relationships between the
ethylene or propylene concentration and the equivalent NO concentrations
indicated by the NDIR analyzer are shown in Figure 1-43. Other optical
interferences measured were propane with NO and ethylene and propylene
with CO. These interferences were 25 times smaller than the ones re-
ported above and have been disregarded in this report.
To correct the radial position versus NO concentration profiles for
these optical interferences, we assumed that the ethylene and propylene
concentrations measured at the burner center line decreased radially
proportional to the CH4 concentration changes.
The NO concentration radial profile for the movable-block swirl
burner can also be corrected for optical interferences. This profile was
shown in Figure I-ZZ. The corrected data are represented on the NO
transparency (Figure 1-37) dashed line. Beyond the region from —6 to
11 cm (where the dashed and solid lines join), there is no correction.
As for the "short" flame baffle burner, the profile resembles an upside
down script "m" in shape. This profile does not have as much structure
as the "short" flame baffle burners profile because the flame has not
had as long to develop. The distance between the sampling point and
the point of gas injection is only 30 cm for the movable-block burner
while it was 53 cm for the "short" flame baffle burner. The tempera-
ture profile illustrated in Figure 1-24 and the temperature curve in
Figure 1-40 (transparency) also reflects an upside down script "m" shape,
very similar to the temperature profile of the short-flame baffle burner.
Table 1-6 gives the mass spectrometer's analysis of a gas sample
from the boiler burner taken in the furnace at a 12. 7-cm axial position.
There were no measurable amounts of either ethylene or propylene. Thus,
the NO profile measured for the boiler burner needs no corrections
because of optical interferences. Like the boiler burner, all burners
which had a Type II flame would burn approximately 95% of the com-
bustibles before the gas entered the furnace. Since there were only
trace amounts of methane measured in the furnace, we can conclude that
no optical interferences caused by ethylene and propylene occur in a
Type II flame. 54
-------
220
200
180
E
o.
Q.
z
UJ
I
3
O
UJ
160
140
(T
Z 120
UJ
O
Z
o
o
80
60
40
20
I
I
I
0 O.I 0.3 0.5 0.7 0.9 I.I
CONCENTRATION OF C3H6 OR C2H4 , %
1.3
B-83-II9I
Figure 1-43. NO CONCENTRATION CORRECTION FACTORS
FOR GAS SAMPLES CONTAINING CaH4 AND/OR C3H6
65
-------
As discussed earlier, there are two types of mean flow patterns:
Type I (Figure 1-1) is a toroidal recirculation cell, and Type II is a
steady pear-shaped recirculating cell. Two burners displayed flow pat-
terns that could not be described by either the Type I or the Type II
flow patterns. Tables 1-7 and 1-8 present direction flow measurements
taken for the movable block burner with intermediate swirl intensity, a
2008 CF/hr gas input, and 20% excess air. The probe used in these
measurements is described in Volume II of this report. The sign of the
time-averaged A P (pressure differential) determines the directional flow
of the gases. In Tables 1-7 and 1-8, a plus sign correlates with forward
flow (flow away from the burner) and a negative sign indicates a recircu-
lating flow (flow toward the burner). From Table 1-7, we then can deduce
that at a 12.7-cm axial position the flow pattern resembles that of a
Type I flame. This can be visualized by turning to the axial velocity
transparency (Figure 1-41) present in this report. However, the data
from Table 1-8 indicate that the flow pattern at a 30. 5-cm axial position
resembles that of a Type II flame. Velocity data for the short-flame
baffle burner shows a similar type of flow mixture between Type I and
Type II patterns.
Figures 1-44 and 1-45 present two possible flow patterns which can
explain these data. Figure 1-44 illustrates an unsteady pear-shaped re-
circulating cell. In this case the recirculating region would oscillate
back and forth across the center line. However, because of the frequency
resolution of the measuring systems (10 cycles/s) as compared with the
frequency of recirculating cells (~4 cycles/s) we should have been able
to detect these oscillations. Nonetheless, we detected no oscillations.
Figure 1-45 illustrates a kidney-shaped recirculating cell. From the
figure, we can see that a radial traverse near the downstream edge of
the recirculating cell would indicate a Type I flow pattern while a traverse
near the upstream edge of the cell would indicate a Type II flow pattern.
Although we plotted the x and y velocity components for the 12. 7-cm and
30. 6-cm axial positions, a comparison with the stream lines shown in
Figure 1-45, proved to be inconclusive since we do not know what our
position is relative to the recirculating cell. Thus to verify a kidney-
shaped recirculating cell would require at least one more directional
flow profile.
66
-------
Table 1-7. TIME-AVERAGED DIRECTIONAL FLOW DATA
OBTAINED AT THE 12. 7-cm AXIAL POSITION (Movable
Block Swirl Burner — Intermediate Swirl Intensity).
GAS INPUT, 2008 CF/hr; 3. 6% EXCESS OXYGEN;
NOZZLE IN THROAT POSITION.
Time Time- ., Time
RP* cm Averaged Ap RP,* cm Averaged AP RP,' cm Averaged Ap
20
17
15
14
11
10
9
8
7
-1.31
-1. 11
0. 00
+ 6. 87
+ 202. 0
+204. 8
+40. 6
-0. 1
-5. 76
6
5
4
3
2
1
0
-1
-2
-7. 59
-7. 5
-6. 3
+ 0. 06
+ 7. 22
+ 12. 37
+ 12. 76
-4.51
-19. 31
-3
-^
-5
-6
-7
-8
-9
-10
-13
-26. 37
-20. 93
-7. 39
-29. 87
+ 148. 3
+ 282. 7
+213. 0
+ 91. 56
-2. 28
Radial Position
-------
oo
Table 1-8. TIME-AVERAGED DIRECTIONAL FLOW DATA
AT THE 30. 5-cm AXIAL POSITION AND OBTAINED USING A
HUBBARD PROBE (Movable Block Swirl Baffle - Intermediate
Swirl Intensity). GAS INPUT, 2008 CF/hr; EXCESS OXYGEN,
3.6%; NOZZLE IN THROAT POSITION
Time- .. Time- . Time-
RP,'' cm Averaged AP RP, cm Averaged A P RP," cm Averaged A P
-13
-10
—7
-4
-3
-2
-1 55. 86
+ 59. 82
+31. 52
+4. 77
+ 1. 58
-0. 94
-1
2
5
6
7
8
-2. 09
-3. 38
-1.48
-0. 88
-0. 08
+ 1. 86
11
14
17
20
23
26
29
H7. 21
l 42. 63
+ 33. 37
+8. 59
+ 1. 08
-0. 53
-0.77
.
Radial Position
-------
INSTANTANEOUS
FLOW PATTERN
/UNSTEADY
< AXISYMMETRIC
[FLOW FIELD
STEADY FLOW
RECIRCULATING
REGION
MEAN FLOW
DIRECTION
STREAM LINES
REVERSED AXIAL FLOW REGION
BOUNDARY OF RECIRCULATING REGION
STAGNATION POINT
A-83-II96
Figure 1-44. UNSTEADY PEAR-SHAPED RECIRCULATING CELL
69
-------
MEAN FLOW
DIRECTION
STREAM LINES
REVERSED AXIAL FLOW REGION
BOUNDARY OF RECIRCULATING REGION
STAGNATION POINT
A-83-II95
I
Figure 1-45. KIDNEY-SHAPED RECIRCULATING CELL
70
-------
We can conclude from a survey of the input-output data that the flue
concentrations of NO for Type II flow patterns are approximately 3-1/2
times larger than those for Type I flow. A similar difference in the
magnitude of NO concentrations can be observed by comparing NO radial
profiles. In addition to the flow pattern, a major difference between
Type I and Type II flames is the region of combustion. A Type I flame
will typically consist of 50% unburned combustibles when it leaves the
burner block and enters the furnace while gases of a Type II flame will
have less than 10% combustibles. Since 50% of the combustion of a
Type I flame occurs in the furnace, a lower flame temperature will re-
sult because of the rate of heat release and the diffusion and entrainment
of recirculating combustion products. These recirculating combustion
products will absorb thermal energy, making less energy available for
the kinetics of NO formation.
71
-------
A C K NO W L F. DC! M l:: NT
The Environmental Protection Agency gratefully acknowledges the
assistance of J. D. Nesbitt, senior adviser, Institute of Gas Technology,
for his aid on this project.
72
-------
BIBLIOGRAPHIC DATA
SHEET
1. Ki-port No.
EPA-650/2-73-033a
3. Recipient's Accession N».
4. Til Ir iiinl Sulu if !<•
Aerodynamic Control of Nitrogen Oxides and Other
Pollutants from Fossil Fuel Combustion
VnliimP J--T")ata Apa)ygig anf^ Summary nf f!nnphiaif>ns
5. Hrpiirt Dnii.-
October 197 ^
6.
7. Author(s)
D. R. Shoffstall and D. H. Larson
8- Performing Orjr,iini/.uiion Kepi.
No.
9. Performing Organization N;imr .ind Adiln-s.s
Institute of Gas Technology
IIT Center
3424 South State Street, Chicago, Illinois 60616
10. I'lojciM/Task/Woik Unii N,
11. ( ontruci/(muii No.
68-02-0216
12. Sponsoring Organization Name :>nd Address
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, North Carolina 27711
13. Type of Report A Period
Covered
Final
14.
IS. Supplementary Notes
16. Abstracts.
The report gives a synopsis of data collected from investigating the relation-
ship between combustion, aerodynamics, and pollutant emission characteristics of
industrial burners. Five types of burners were studied: a scaled-down utility power
boiler burner; a high-intensity (flat-flame) burner; a movable-block burner devel-
oped by the International Flame Research Foundation; an axial flow burner; and a
baffle burner used in steel reheat furnaces. Broad conclusions , applicable to all
burners tested, were: burners using radial gas injection produce peak NO at 11% or
less excess air; burners using axial gas injection produce peak NO at 22% or more
excess air; NO concentration increases with increasing air preheat, but the magni-
tude of the change depends on burner design and the gas injection method; and
changing from axial to radial gas injection consistently results in an increase in NO
concentration. Volume n is subtitled Raw Data and Experimental Equipment.
17. Key words and Document Analysis. 17o. Descriptors
Air Pollution
Nitrogen Oxides
Aerodynamics
Natural Gas
Combustion Control
Burners
Flames
17b. Identifirrs/Opcn-Kndcd Terms
Air Pollution Control
Stationary Sources
Axial Injection
Radial Injection
Swirl
Type I Flame
Type n Flame
Radial Profiles
Industrial Burners
17c. C.OSATI FieUi/C.roup
13B, 21B
18. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Security Class (This
Page
UNCLASSIFIED
21- No. of Pages
80
22. Price
FORM NTIS-33 (REV. 3-721
73
USCOMM-DC H95Z-P72
------- |