655R80001
EFFECT OF FUEL NITROGEN
ON
INDUSTRIAL BOILER NOX EMISSIONS
KVB11 34204-1244
CONTRACT NO.
68-02-3175
WORK ASSIGNMENT
PREPARED FOR'
U.S. ENVIRONMENTAL PROTECTION AGENCY
INDUSTRIAL ENVIRONMENTAL RESEARCH LAB.
RESEARCH TRIANGLE PARK. NORTH CAROLINA
PREPARED BY:
S.S. CHERRY
RESEARCH & ANALYSES DIV.
KVB, INC.
AUGUST 1980
18006 SKYPARK BLVD., IRVINE, CALIFORNIA 92714 • (714) 641-6200
MINNEAPOLIS, MN (612) 545-2142 • HARTSDALE, NY (914) 949-6ZOO
-------
ABSTRACT
Industrial boiler data were analyzed to quantify the NOX measurements
in terms of a single independent variable—fuel nitrogen content. The
analyses showed that, in general, there was a high degree of correlation
between NO and fuel nitrogen.
X
Specific subgroups (pulverized coal) of the data base showed no corre-
lation primarily because of the narrow range in fuel nitrogen tested, or that
the NO was potentially a function of more than one independent variable
(spreader stokers).
Insufficient data were available to characterize firetube boilers,
cyclone burners, vibrating grates, overfed stokers and underfed stokers.
Recommendations are made to expand the data base, reformulate the
correlation function and include more than one independent variable.
KVB11-34204-1244
-------
CONTENTS
Section Page
ABSTRACT ii
1.0 INTRODUCTION AND SUMMARY 1-1
2.0 DISCUSSION 2-1
2.1 Data Sources 2-1
2.2 Data Base 2-2
2.3 Data Analysis 2-7
3.0 CONCLUSIONS AND RECOMMENDATIONS 3-1
3.1 Conclusions 3-1
3.2 Recommendations 3-2
4.0 REFERENCES 4-1
KVB11-34204-1244
-------
TABLES
Number Page
2-1 Data Base for Baseline Operation 2-4
2-2 Data Base Separation by Boiler Type and Fuel 2-8
2-3 Data Base for Low NO Operation 2-23
iv KVB11-34204-1244
-------
FIGURES
Number Page
2-1 NOX emissions as a function of fuel nitrogen 2-9
content-baseline operation. Watertube industrial
steam boilers
2-2 NO emissions as a function of fuel nitrogen 2-12
content for ambient watertube boilers-baseline
operation
2-3 Correlation of ambient combustion air spreader 2-14
stoker data-baseline operation
2-4 NOX emissions as a function of fuel nitrogen 2-16
content for preheated watertube boilers-baseline
operation
2-5 Fuel nitrogen conversion-linearized form 2-19
2-6 Fuel nitrogen conversion-power form 2-20
2-7 NO,, as a function of excess oxygen 2-22
Jt
2-8 Ambient combustion air watertube industrial 2-27
boilers-low NO operation
X
2-9 Preheated combustion air watertube industrial 2-28
boilers-low NO operation
X
KVB11-34204-1244
-------
SECTION 1.0
INTRODUCTION AND SUMMARY
KVB, during the course of several EPA sponsored field testing
programs, has developed an extensive data base on industrial steam boiler NOX
emissions. The basic purpose of the study reported herein was to analyze the
data base to quantify the effect of fuel nitrogen content on industrial boiler
NOX emissions.
It was the intent to resolve these effects in terms of boiler type,
fuel type and firing mode for both baseline and low NO operation. It was
also the intent of the study to identify those categories which could not be
resolved because of insufficient data.
The study was successful in quantifying the fuel nitrogen effect on
NO,, for the major boiler types:
X \
Watertube - ambient temperature combustion air
Watertube - preheated combustion air
Insufficient data were available on firetube boilers and several coal
firing modes. The study could not resolve the fuel nitrogen effect on NOV for
X
pulverized coal firing primarily because of the relatively narrow range in
fuel nitrogen tested.
Baseline data obtained on spreader stokers suggest that excess oxygen
must be considered as an important independent variable since it was equally
capable of correlating the measured NOX- This tentative conclusion is
supported by spreader stoker data obtained at low NO operation (reduced
excess oxygen) for which there was no correlation between NO., and fuel
X
nitrogen content.
Data obtained when several boilers were operated with more than one
fuel were analyzed to infer the conversion efficiency of fuel nitrogen to
NOX.
1-1 KVB11-34204-1244
-------
SECTION 2.0
DISCUSSION
Data sources, data bases and data analyses employed to quantify the
effect of fuel nitrogen content on industrial steam boiler NO emissions are
discussed in this section.
2.1 DATA SOURCES
Six KVB contracts with the EPA were reviewed to obtain the industrial
steam boiler information required for the present fuel nitrogen study.
2.1.1 Contract 68-02-1074
This program involved field testing of a large number of industrial
steam boilers located throughout the United States. Included were firetube
and watertube types using ambient or preheated combustion air and burning gas,
oil, and coal fuels.
The emissions were characterized over the boiler's normal load range
and excess air levels. Other NOx controls were implemented as available,
e.g., overfire air flow was varied on those boilers so equipped.
The results of this effort, reported in References 2-1 through 2-3,
formed the major data source for the present study.
2.1.2 Contract 68-02-1863
Ten small to intermediate size coal-fired steam boilers were fully
characterized with respect to their ability to utilize both western and
eastern coals.* The characterization included both emission measurements and
operational considerations. The program results (Reference 2-4) were included
in the present study.
*Several of these boilers were in electric utility service. However, they
were retained for this study because of their small size (<29 kg/s,
<230,000 Ib/hr).
2-1 KVB11-34204-1244
-------
2.1.3 Contract 68-02-2144
Two industrial steam boilers were extensively modified to include
advanced NO control technologies, i.e., flue gas recirculation, staged
combustion air and variable air preheat (References 2-5 and 2-6)- The
emissions from these modified steam boilers were determined for natural gas
and #6 and/or #2 fuel oils as a function of the degree of NOX control imple-
mentation .
2.1.4 Contract 68-02-2645
This, the present KVB contract with the EPA, addresses advanced
combustion modification technology applied to a wide variety of process equip-
ment. Test results obtained on a wood-bark/coal-fired boiler (Reference 2-7)
were incorporated into the present study.
2.1.5 Contract 68-02-2645 Task No. 4
The objective of this task is to develop 30-day emission measurements
on industrial boilers. These boilers had been placed in a low NO mode and
were then continuously monitored to determine the effect of a combustion
modification (References 2-8 through 2-11).
2.1.6 Contract Nos. IAG-D7-E681 (EPA), EF-77-C-01-2609 (DOE)
This contract is being performed for the American Boiler Manufacturers
Association (ABMA) and is jointly sponsored by EPA and the Department of
Energy. The objective of this program is to produce information which will
increase manufacturer's ability to design and fabricate economical and
environmental satisfactory coal-fired industrial stoker boilers (References
2-12 through 2-19) .
2.2 DATA BASE
The test results obtained from the data sources were carefully
reviewed for applicability and completeness. The data were retained only if
they satisfied the following constraints:
2-2 KVB11-34204-1244
-------
1. Full fuel analysis reported.
2. Load (steam rate) between 70 percent and 90 percent of
rating.
3. Carbon monoxide level below approximately 400 ppm.
The first constraint is self-evident since if the fuel nitrogen con-
tent was not reported, the data point could not be used.
It has been firmly established that steam boiler NO emissions vary
X
with load. It is not the intent of the present study to analytically and/or
empirically correct the measured NOX levels to a common load range. Rather,
only those data obtained between 70 percent and 90 percent of rated steam flow
were retained. The center of this range (80 percent) was an objective of
contract 68-02-1074, i.e., the bulk of the testing to be performed at a signi-
ficant percentage of the boiler's rated steam flow.
The constraint on carbon monoxide level is somewhat arbitrary but was
selected to acknowledge that there is a limitation on the CO which can be
tolerated.
Table 2-1 presents the baseline data developed from the six KVB data
sources. The sources of these data are:
Lines Contract
1-54 68-02-1074
55-59 68-02-2144
60-69 68-02-1863
70 68-02-2645
71-74 68-02-2645 Task No. 4
75-84 IAG-D7-E681 (EPA)/EF-77-C-01-2609 (DOE)
An examination of Table 2-1 indicates the following number of tests
(n), mean fuel nitrogen content (% N), standard deviation (s) and ranges in
fuel nitrogen for the various fuel types:
Parameter #2 #5 #6 Coal/Solid
n
% N
s
Range
7
0.022
0.017
0.006/0.045
1
0.32
—
—
13
0.31
0.08
0.24/0.44
35
1.22
0.23
0.77/1.80
2-3 KVB11-34204-1244
-------
TABLE 2-1. DATA BASE FOR BASELINE OPERATION
Rated
Steam Flow
Line Location
NJ
«
W
1
W
10
o
i
H
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
(28)
Notes at
1
1
1
1
2
2
3
4
5
S
6
7
9
9
9
10
10
11
12
12
13
14
14
15
«D
end of
Unit
1
2
3
1
2
4
2
4
716-3
248-3
3
3
BC-1
BC-6
VA-1
4
S
1
24
20
2
1
4
123-1
32-10
table
103 Ib/hr
29.0
29.0
30.0
29.0
59.2
65.0
10.0
20.0
25.0
10.0
158.0
85.0
60.0
160.0
300.0
60.0
110.0
135.0
225.0
325.0
500.0
150.0
200.0
19.2
60.0
kg/9
3.7
3.7
3.8
3.7
7.5
8.2
1.3
2.5
3.2
1.3
19.9
10.7
7.6
20.2
37.9
7.6
13.9
17.0
28.4
41.0
63.1
18.9
25.3
2.4
7.6
Bailer
Type*
vt
vt
vt
vt
vt
vt
ft
ft
vt
ft
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
vt
Burner
No./Typet
I/Ring
1/Ring
1/Stn.Atn.
I/Ring
1/Stm.Atn.
6/Ring
6/Ring
1/Ring
I/Ring
I/Ring
I/Ring
4/Stm.Atm.
4/Stm.Atn.
4/Ring
I/Triple
4/Ring
2/Rlng
I/Double
3/SS
8/-
8/PC
8/PC
8/-
6/PC
5/SS
6/SS
1 /Rot. Cup
7/US
Fuel}
NG
NG
t2
NG
t2
NG
NG
NG
NG
NG
NG
•2
15
NG
NG
NG
NG
NG
Coal
NG
Coal
Coal
NG
Coal
Coal
Coal
NSF
Coal
%N
By
Wt. Test No.
0.0
0.0
0.045
0.0
0.045
0.0
0.0
0.0
0.0
0.0
0.0
0.018
0.32
0.0
0.0
0.0
0.0
0.0
1.49
0.0
1.40
1.34
0.0
1.35
1.33
1.80
0.26
1.40
12-4
5-2
66-1
106-1
107-1
13-3
69-1
41-3
38-2
4-1
37-8
65-1
6-1
15-1
24-3
30-1
14-1
80-11
18-3
75-7
26-1
78-1
77-11
31-1
27-1
28-2
3-2
16-1
TAir
•f K
Amb
Amb
350 450
Amb
Amb
Amb
Amb
Amb
Amb
Amb
Amb
300 422
240 389
400 478
330 439
401 478
Amb
Amb
Amb
640 611
645 614
630 606
655 619
445 503
350 450
Amb
Amb
Amb
Load
103 Ib/hr kg/B %
21.0
22.5
23.0
24.0
23.5
47.5
53.0
7.0
14.0
20.0
8.0
115.0
62.5
46.0
136.0
246.0
48.5
85.0
106.0
180.0
181. 0
260.0
260.0
400.0
120.0
162.0
14.5
47.0
2.7
2.8
2.9
3.0
3.0
6.0
6.7
0.9
1.8
2.5
1.0
14.5
7.9
5.8
17.2
31.1
6.1
10.7
13.4
22.7
22.9
32.8
32.8
50.5
15.2
20.5
1.8
5.9
72.4
77.6
76.7
82.8
81.0
80.2
81.5
70.0
70.0
80.0
80.0
72.8
73.5
76.7
85.0
82.0
80.8
77.3
78.5
80.0
80.4
80.0
80.0
80.0
80.0
81.0
75.5
78.3
NOX pp.
02 dry S
% 3% O2
2.8
4.0
5.9
2.6
3.1
4.0
3.8
7.5
6.8
2.9
5.1
5.2
6.5
2.6
3.8
3.2
5.2
8.1
7.0
6.1
5.3
5.8
4.5
9.8
10.3
10.8
5.3
7.5
70
76
123
82
79
132
101
92
108
76
56
185
297
241
374
181
104
94
373
200
383
485
327
578
550
542
142
331
CO
ppn
10
56
0
0
120
133
102
55
—
180
0
11
50
20
63
0
0
0
52
~
0
0
0
0
0
0
282
0
Kff.
—
77
81
80
—
76
—
—
80
78
84
—
79
76
—
80
70
82
84
86
86
85
81
81
80
78
—
(continued)
its.
-------
TABLE 2-1 (CONTINUED)
to
Ul
W
to
o
I
H
10
Rated
Steam Flow
Line
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Location
©
15
16
17
18
18
18
19
20
20
21
21
23
26
27
29
37
38
39
19
38
Unit
32-13
2-1
2
T-8
2
3
4
1
4
42
2
3
1
1
1
5
2
2
BIOS
ID3 Ib/hr
60.0
17.0
65.0
110.0
90.0
105.0
160.0
17.5
80.0
400.0
50.0
75.0
7.0
18.0
100.0
150.0
40.0
45.0
200.0
17.5
45.0
kg/.
7.6
2.1
8.2
13.9
11.4
13.3
20.2
2.2
10.1
50.5
6.3
9.5
0.9
2.3
12.6
18.9
5.1
5.7
25.3
2.2
5.7
Boiler
Type'
wt
ft
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
ft
ft
wt
wt
wt
wt
wt
wt
wt
Burner
No./Typet
7/US
I/Rot. Cup
2/Stm.Atm.
2/Stm.Atm.
3/Stm.Atm.
4/Stm.Atm.
4/Stm.Atm.
1/Stn.Atn.
1/Stm.Atm.
I/Ring
I/Double
2/Cyclonea
2/SS
3/SS
I/Ring
I/Ring
1/Stm.Atffl.
I/Ring
1/Stm.Atm.
2/Rlng
2/Stm.Atm.
2/Stm.Atm.
I/Ring
1/Stm.Atm.
I/Spud
VStm.Atm.
VStm.Atm.
VStm.Atm.
I/Ring
1/Stm.Atn.
Fuel}
Coal
NSF
16
»2
16
16
16
»6
*2
NG
16
16
Coal
Coal
Coal
NG
NG
*2
NG
PS300
NG
16
16
NG
16
NG +
RG
12
»6
NG
NG
16
%N
By
wt.
1.40
0.28
0.29
0.01
0.26
0.26
0.26
0.44
0.006
0.0
0.37
0.30
1.55
1.45
1.43
0.0
0.0
0.020
0.0
0.77
0.0
0.31
0..30
0.0
0.49
0.0
0.008
0.25
0.0
0.0
0.31
Teat No.
17-1
36-2
10-1
7-1
9-1
21-1
22-1
1-1
52-1
190-3
8-2
170-1
32-1
19-1
20-1
40-1
48-2
57-1
109-1
111-1
113-1
116-1
176-2
180-2
186-1
207-1
19-5
19-97
19-147
200G-2
200-24
TAir
•F K
Amb
Amb
Amb
Amb
Amb
420 489
548 560
Amb
Amb
Amb
Amb
Amb
547 559
Amb
Amb
Amb
Amb
Amb
Amb
Amb
375 464
395 475
227 382
350 450
320 433
Amb
Amb
Amb
Amb
283 413
290 417
Load
103 Ib/hr kg/a »
47.0
15.0
50.0
88.0
71.0
80.0
130.0
14.5
14.0
14.0
60.0
65.0
320.0
40.0
60.0
6.1
15.3
15.7
75.0
85.0
120.0
119.0
32.0
40.0
36.0
160.0
14.5
13.8
14.5
40.0
38.0
5.9
1.9
6.3
11.1
9.0
10.1
16.4
1.8
1.8
1.8
7.6
8.2
40.4
5.1
7.6
0.8
1.9
2.0
9.5
10.7
15.2
15.0
4.0
5.1
4.5
20.2
1.8
1.7
1.8
5.1
4.8
78.3
88.2
76.9
80.0
78.9
76.2
81.3
82.9
80.0
80.0
75.0
81.3
80.0
80.0
80.0
87.1
85.0
87.2
75.0
85.0
80.0
79.3
80.0
88.9
80.0
80.0
82.8
78.8
83.0
88.9
84.4
K
°2
10.6
6.7
3.7
5.3
7.4
7.0
6.8
4.4
3.6
3.2
5.2
3.5
3.0
9.4
7.6
5.0
8.3
8.0
6.6
9.3
5.4
5.0
4.3
1.9
3.0
3.7
3.2
3.0
3.0
1.6
2.9
d?y e
3%0Z
296
185
180
164
246
291
242
423
71
59
328
259
793
476
506
76
53
118
113
458
155
294
195
220
326
192
120
214
95
171»
291*
CO
ppm
0
126
0
0
0
0
0
0
36
15
0
11
0
62
80
203
14
86
0
193
0
0
0
0
0
26
4
4
4
140
22
Eff.
72
—
—
82
—
86
—
—
—
80
—
—
—
—
—
—
80
82
81
—
87
85
85
87
83
82
83
78
81
85
Notes at end of table
(continued)
-------
TABLE 2-1 (CONTINUED)
10
u>
*>.
10
o
10
Line Location
60 Alna
61 (a. Viacf)
62
63 fUr Wisc^1
> ^
64 U. Misc.
65
66 Willmar
67 Fairmont
68 Fremont
69
70 3
71 1
72 2
73 3
74 4
75 A
76 8
77 C
78 (IT)
79
80 F
81
82 G
83
84 (H)
LEGEND
•Boiler Type:
t Burner Type:
Rated
Steam Flow
Boiler
Unit 103 Ib/hr kg/s Type*
13 230.0 29.0
Stout »2 45.0 5.7
1 Eau Claire 60.0 7.6
11
Madison 100.0 12.6
12
13 160.0 20.2
»3 80.0 10.1
16 160.0 20.2
100.0 12.6
100.0 12.6
— 90.0 11.4
260.0 32.8
130.0 16.4
300.0 37.9
200.0 25.3
182.5 23.0
90.0 11.4
—
80.0 10.1
—
— 75.0 9.47
—
— 45.0 5.68
tit « Hater tube
ft » Firetube
Triple • Triple air register
Double ' Double air register
SS * Spreader stoker
PC * Pulverized coal
US ' Underfed stoker
VG - Vibrating grate
OS • Overfed stoker
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
Burner
No./Typet
4/PC
VG
US
3/SS
6/SS
4/SS
4/PC
SS +
Pneumatic
3/SS
3/Stn.Atm.
4/PC
6/SS
7/SS
5/SS
7/SS
VG
3/SS
3/SS
OS
{Fuell NG
RG
NSF
PS300
»N
By
Fuel{ Wt. Test No.
Coal 1.09
Coal 0.91
Coal 1.28
Coal 1.24
Coal 0.78
Coal 1.30
Coal 1.37
Coal 1.06
Coal 1.25
Coal 0.93
Coal-f 1.22
Wood
Coal 1 . 27
16 0.24
Coal 1.25
Coal 0.77
Coal 0.83
Coal 1.44
Coal 1.04
Coal 1.24
Coal 1.23
Coal 1.23
Coal 1.12
Coal 0.84
Coal 1.04
Coal 1.04
- Natural gas
» Refinery gas
- Naval Standard
47
11
25
30
9
15
27
6+7
5
9
1
1/16
4/9
6/1
—
30
13A
9
5
15
1
31
2
23
1A
Fuel
- Pacific Standard 300
ITotal NOx not measured but taken as
two percent
for oili and six percent
TAir
Load
•F K 103 Ib/hr
Preheat
Amb
Amb
Amb
Amb
Amb
Preheat
Amb
Preheat
Preheat
Preheat
Amb
Preheat
Preheat
Preheat
Preheat
Amb
Preheat
Amb
Amb
Amb
Amb
Amb
Amb
Amb
(Similar to a *5)
three percent more
for gas (Ref.2-1).
200.0
33.0
40.0
49.0
80.0
90.0
124.0
60.0
134.0
140.0
82.0
80.0
72.5
198.0
97.0
240.0
144.0
163.0
77.4
79.2
60.0
60.0
63.8
57.0
39.6
than NO
kg/a »
25.3 87.0
4.2 73.3
5.1 88.9
6.2 81.7
10.1 80.0
11.4 90.0
15.7 77.5
7.6 75.0
16.9 83.4
17.7 87.4
10.4 82.0
10.1 80.0
9.15 81.0
25.0 76.0
12.2 75.0
30.3 80.0
18.2 72.0
20.6 89.0
9.77 86.0
10.0 88.0
7.58 75.0
7.58 75.0
8.05 85.0
7.20 76.0
5.00 88.0
for coal i
NOK pp.
02 dry S
* 3« O2
3.7 6541
4.5 165*
5.4 228*
6.6 218*
6.2 303*
9. 1 449*
6.6 428*
8.0 363*
5.4 679*
5.5 502*
9.3 238*
9.5 420*
6.6 234*
6.4 678*
8.9 440*
4.0 483*
7.5 280*
8.9 302*
7.9 233*
7.8 228*
8.9 344
9.9 333
8.9 337*
8.0 444*
9.2 320*
CO
ppm
20
—
—
102
380
215
289
47
20
14
362
99
19
34
—
383
53
66
89
39
146
139
—
—
153
Eff.
%
—
—
—
__
74
76
—
80
87
87
82
—
~
—
86
—
—
82
83
85
—
—
74
76
—
-------
The narrow ranges in fuel nitrogen content for the three fuel oil types
precludes separately specifying their effect on NOX.
Table 2-2 separates the baseline data base into categories based on
boiler type and fuel. The entries correspond to the line numbers of the test
summary in Table 2-1. In addition, the oil type is noted as well as the
firing mode for coal utilization.
It is evident that too few data (2) are available to characterize the
effect of fuel nitrogen on firetube boiler NO emissions. It is also evident
that too few data are available for coal firing by underfed stoker (3),
vibrating grate (4), cyclone burners (1) and overfed stoker (1) in watertube
boilers.
Because of the significant difference between firetube and watertube
boilers, the firetube data were deleted (with one exception) and not combined
with the watertube results in the subsequent discussions. The only exception
was in retaining lines 45 and 46 in subsection 2.3.3.
2.3 DATA ANALYSIS
Figure 2-1 is a scatter plot of NOX in terras of fuel nitrogen content
for the ambient and preheated watertube boilers firing oil and coal. Not
shown are the results obtained on gas fuel which will be used as a pseudo data
point corresponding to a zero fuel nitrogen content. Specifically, the gas
fuel data base indicates:
Ambient Preheated
Combustion Air Combustion Air
Mean NOX, ppm 100.0 233.0
Standard Deviation, ppm 35.4 78.3
Further analysis of the gas fuel (%N=O) results showed that the mean values
were statistically different, i.e., the null hypothesis was rejected at the
95 percent confidence level.
2-7 KVB11-34204-1244
-------
TABLE 2-2. DATA BASE SEPARATION BY BOILER TYPE AND FUEL
Boiler Type
N)
I
00
I
W
10
o
i
H
to
Firetube
Gas Oil Coal
8 30(NSF)
9 46(#2)
11
44
45
Total Tests: 520
Hatertube-Ambient
Gas
1
2
4
6
7
10
17
IS
38
47
54
57
12
Oil
5(*2)
27(NSF)
31(*6)
32(*2)
33(#6)
36(#6)
37(#2)
39(#6)
40(*6)
48(PS300)
55(»2)
56(«6)
12
Coal
19(SS)
26(SS)
28(US)
29(OS)
42(SS)
43(SS)
6KVG)
62 (VG)
63 (US)
64(SS)
65 (SS)
67(SS)
71(SS)
76(SS)
78(VG)
79(VG)
80 (SS)
81(SS)
82(SS)
83(SS)
84 (OS)
21
Watertube-Preheat
Gas Oil
14 3(#2)
15 12(*2)
16 13(#5)
20 34(*6)
23 35 (*6)
49 50(16)
52 51(16)
58 53(*6)
59(#6)
72(*6)
8 10
Coal
21(PC)
22 (PC)
24 (PC)
25(SS)
41 (Cyclone)
60 (PC)
66(SS)
68(PC)
69(PC)
70(SS)*
73 (PC)
74(SS)
75 ( SS )
77(SS)
14
•Wood/coal boiler: pneumatic injection for wood; spreader stoker for coal.
Note: Consult Table 2-1 for symbol explanation.
-------
w w
700
600
(N
O
*
n
4J 500
(0
>,
M
t!
~ 400
&
a
0* 300
z
200
1 1 1 1 1 1 1 w
Q Ambient Combustion Air
— • Preheated Combustion Air A "~
• *
~ ^^ "™
* *
A O
0» rf?>
0 * ° 0«
o ^ *o
^P* 0° 8° °
~ «n ° • . o
XV (9n
L** o * -
100 KT —
0
1 1 1 1 1 1 1 1 1
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.
FUEL NITROGEN CONTENT, % BY WEIGHT
Figure 2-1. NO emissions as a function of fuel nitrogen content - baseline
operation. Watertube industrial steam boilers.
2-9
KVB11-34204-1244
-------
2.3.1 Correlation of Ambient Combustion Mr Data - Baseline Operation
Correlations of NO,, with fuel nitrogen content (%N) were performed
X
using 34 data points (12 oil, 21 coal, 1 pseudo gas) in order to determine the
degree to which a function could explain the data scatter. The functions
investigated were:
Power: NOX = a (%N) b
Quadratic: NOv = a + b (%N) + c (%N)2
X
Logarithmic: NOV = a + b £n (%N + 1)
X.
The power and logarithmic functions were first linearized:
Power: Y = A + B • 2
where Y = Jtn (NO )
A = In a X
B = b
Z = An (%N)
Logarithmic: Y = A + B • Z
where Y = NO
A = a
B = b
Z = SLn (%N + 1)
The coefficients (a,b,c) for quadratic correlation were taken from
Reference 2-20. (This reference or other publications on statistics should be
consulted for a full explanation of the assumptions on which correlation is
based.)
The power function does not adequately reflect the data for small
values of %N since it predicts that NOX approaches zero as %N approaches
zero. Further, the slope of NOX with %N is infinite with b<1. The quadratic
function, when fitted to the data, showed a peak value of NO within the %N
X
range; i.e., it initially increased and then decreased. This was caused by
the negative value of c which eventually caused the function to decrease.
Thus, the quadratic function was also rejected.
2-10 KVB11-34204-1244
-------
The logarithmic function, when fitted to the data, yielded a mean
value of NO of:
NO (ppm, dry, @ 3%0 ) = 140.2 + 271.3 £n (%N + 1) (1)
X 0<%N £1.80
with a correlation coefficient, r, of 0.702. Figure 2-2 is a scatter plot of
the ambient combustion air data and the correlation equation. Also shown are
the 95 percent confidence limits of the mean NOX»
A basic assumption inherent in correlation analysis is that, for a
given value of the independent variable (%N), the dependent variable (NOX) is
normally distributed. The correlation equation then represents the mean value
of this normal distribution. Further, the 95 percent confidence limits cor-
respond to approximately ±2 to 2.5 standard deviations about the mean
(depending on the sample size) .
The significance of the correlation coefficient, r, is that its
2
square, r (coefficient of determination), represents the degree to which the
correlation equation explains the data scatter. In a perfect correlation,
2
with r =1, the correlation function will pass through all of the data
points. Thus, with r2 = 0.493 (0.7022), 49.3 percent of the NOX data scatter
is explained by fuel nitrogen content, with the remainder (50.7 percent) due
to other variables. The high degree of correlation obtained between NOV and
X
fuel nitrogen (r = 0.702) does not imply that there are no significant dif-
ferences in the data for a given nitrogen content. As noted in Figure 2-2,
measured NO values centered about a nitrogen content of 0.3% (#6 fuel oil)
A
differ by a factor of approximately 3.
One possible other variable was noted in Reference 2-4, which reported
on the emission comparison between eastern coals and the lower nitrogen
content western coals. The high moisture content of western coal was deemed
to be responsible for a portion of its lower NOx emissions since the moisture
would reduce the combustion temperature and affect the NO formed by fixation
of atmospheric nitrogen (thermal NO ) .
X
A separate correlation, performed for the 21 coal data points,
resulted in a significantly lower (18.0 percent) coefficient of
2-11 KVB11-34204-1244
-------
800
700
* 600
4J
rO
•o 500
I
&(
- 400
x
300
200
100
I I I I
Ambient Combustion Air
O
O
O
0.4 0.6 0.8 1.0 1.2 1.4
FUEL NITROGEN CONTENT, % BY WEIGHT
1.6
1.8
Figure 2-2. NO emissions as a function of fuel nitrogen content for ambient
watertube boilers-baseline operation.
2-12
KVB11-34204-1244
-------
determination. Furthermore, the correlation of the coal data points was not
significantly different than zero which indicated that fuel nitrogen content,
by itself, could not statistically "explain" the measured NOX. Other
parameters, e.g., coal moisture content, firing mode, excess oxygen, etc.,
could be the important variables for coal NO emissions.
The 13 spreader stoker data points were analyzed to determine if this
coal firing mode was a potentially significant subgroup. The resulting
correlation equation for the mean NOX was:
NO (ppm, dry @ 3% O ) = 103.8 + 367.9 Jin (%N + 1) (2)
X 0.78 <%N <1.80
This is shown in Figure 2-3 together with the data points and 95 percent
confidence limits on the mean. In this instance the confidence limits corres-
pond to ±2.2 standard deviations. The correlation coefficient was 0.578,
which indicated that 33.4 percent of the data scatter can be attributed to
fuel nitrogen.
Also noted in Figure 2-3 are the measured excess &2 levels for the
spreader stoker baseline tests. In general, the excess O_ levels increase
with increasing fuel nitrogen content, which may mask some of the variation of
NOX with nitrogen content. A correlation in the form:
NO = a + b in (% O ) (3)
X £
resulted in a coefficient of determination of 32.9 percent, i.e., comparable
to that calculated with fuel nitrogen as the independent variable. From this
comparison it appears that the measured NOV for the specific spreader stokers
X
tested is probably a function of both fuel nitrogen content and excess 0, .
It is not clear if the correlation for the spreader stoker subgroup is
statistically different from that for the entire ambient combustion air boiler
category. This uncertainty is because of the difference in the range of %N
applicability for each correlation; i.e., the null hypothesis would have to be
true or false for every value of fuel nitrogen.
2-13 KVB11-34204-1244
-------
600
( ) = % Excess O
0.4 0.6 0.8 1.0 1.2 1.4
FUEL NITROGEN CONTENT, % BY WEIGHT
1.6
1.8
Figure 2-3. Correlation of ambient combustion air spreader stoker data -
baseline operation.
KVB11-34204-1244
2-14
-------
2.3.2 Correlation of Preheated Combustion Air Data - Baseline Operation
The 25 data points (10 oil, 14 coal, 1 pseudo gas) were analyzed
following the same procedure described for the ambient combustion air
category.
Figure 2-4 is a scatter diagram of the baseline NO vs. fuel nitrogen
content. Also shown is the correlation function:
NO (ppm, dry @ 3% 0 ) = 158.4 + 456.5 An (%N + 1) (4)
X 0<%N <1.55
with a correlation coefficient (r) of 0.786. This implies that the fuel
nitrogen function, Equation (4), explains 61.8 percent of the NOV data
X
scatter. Also shown in Figure 2-4 are the 95 percent confidence limits of the
mean NOx, which correspond to ±2.1 standard deviations.
The seven pulverized coal data points were separately analyzed to
determine if they were a statistically significant subgroup. The result was a
correlation coefficient of -0.246, indicating that larger values of %N were
associated with smaller values of NO . Further, it was shown that r was not
statistically different from zero, so that the arithmetic means of NO and %N
X
were just as likely estimators since NOV and %N were not correlated; i.e.,
X
NO = 570.7 ppm and %N = 1.22 would be appropriate to describe the seven
X
pulverized coal data points. The failure to achieve a statistically
significant correlation may be due to the narrow range in fuel nitrogen
content for the pulverized coal boilers (0.93 to 1.40 %N).
2.3.3 Fuel Nitrogen Conversion Efficiency
The test results shown in Table 2-1 contain data on nine boilers which
were tested on gas fuel and either oil or coal, with two of the boilers tested
with two types of oil. These data offer the opportunity to infer the conver-
sion efficiency of fuel-bound nitrogen to NO .
X
If the total NOX can be expressed as the sum of a "thermal" and a
"fuel" component:
NOX (%N) = N0x (thermal) + N0x (fuel) (5)
2-15 KVB11-34204-1244
-------
800
700 —
600
1 I
Preheated Combustion Air
I I
I
I I
I
0.2 0.4 0.6 0.8 1.0 1.2 1.4
FUEL NITROGEN CONTENT, % BY WEIGHT
1.6 1.8
Figure 2-4. NO emissions as a function of fuel nitrogen content for preheated
watertube boilers-baseline operation.
2-16
KVB11-34204-1244
-------
where NO (thermal) is independent of the fuel NO component, then the data
x x
can be used to assess the efficiency of conversion of fuel nitrogen to N0x.
Define a parameter, A, by:
NO (%N) - NO (thermal)
—* *100 (6)
NO (max)
x
Where NOX (max) is the fuel NOX component if all the fuel nitrogen were
converted to NOX (100 percent conversion efficiency). For oil:
NO (max) = 2.508 • 10? (%N)/(Btu/lb) , ppm, dry @ 3% 02
For coal:
NO (max) = 2.314 • 10? (%N)/(Btu/lb) , ppm, dry @ 3% O2
The parameter A compares the inferred fuel nitrogen component with that which
would occur if all the fuel nitrogen were converted. It is to be noted that
the validity of this comparison is based on the assumption that the thermal
NOX is independent of the fuels burned in the boiler.
The parameter A was correlated with fuel nitrogen content, %N, using a
function of the form:
A = c(%N)d (7)
which was linearized to:
Y = C + D • Z (8)
with: Y = Jin A
C = An c
D = d
Z = Jin (%N)
2-17 KVB11-34204-1244
-------
The resulting correlation yielded:
A = 14.59 (%N)"°'59° (9)
0.006 _<%N _<1 .40
with a correlation coefficient (r) of -0.901 and a coefficent of determination
(r ) of 0.812. The negative value of r indicates that the conversion
2
efficiency (A) decreases with increasing %N . The value of r shows that the
fuel nitrogen conversion is highly correlated with fuel nitrogen content .
The use of a power function for the conversion efficiency resulted in
the fuel NO , NO (%N) - NOv (thermal), monotonically increasing with
A A X
increasing fuel nitrogen content, i.e.:
NO (%N) - NO (thermal) ~ A • NO (max)
XX X
~ (%N)d + 1
where d + 1 >0 . An exponential function for the conversion efficiency, A =
2
exp (c + d • %N), also correlated the data well (r = 78.9%). However, the
exponential damping (with d <1 ) more than offset the linear increase in
NO (max) with fuel nitrogen content so that the fuel NO first increased and
then decreased with increasing %N .
Figures 2-5 and 2-6 are graphic representations of the data points,
the correlation equation of the mean, and the 95 percent confidence limits of
the mean . The former figure is in terms of the linearized equation and the
latter for the power form.
The high degree of correlation obtained between conversion efficiency
and fuel nitrogen content does not necessarily confirm the assumption that the
thermal NOV in a given boiler is independent of the fuel fired. However, it
X
does not necessarily negate this assumption .
2.3.4 Fuel Nitrogen Effects At Low NOV Operation
As previously mentioned most boilers were tested over a range of
excess air levels . Other NO controls were also implemented if the boiler
2-18 KVB11-34204-1244
-------
FUEL NITROGEN CONTENT, % BY WEIGHT
0.006 0.02 0.1 0.2 0.5 1.0 1.
< 4
c
=>?
I
+95%CL
Mean
I
-6 -5
0
1 I
Watertube
Watertube
Firetube
I I
ambient
preheat
-4
1
I
-3 -2
£n(%N)
o
-1
300
300
200
150
100
80
60
40
30
10
5
3
+1
H
H
H
U
H
§
w
20 8
H
H
z
Figure 2-5. Fuel nitrogen conversion - linearized form.
2-19
KVB11-34204-1244
-------
260
Watertube
Watertube
Firetube
+95% CL
Mean
-95% CL
0.2 0.4 0.6 0.8 1.0 1.2
FUEL NITROGEN CONTENT, PERCENT BY WEIGHT
1.4
Figure 2-6. Fuel nitrogen conversion-power form.
2-20
KVB11-34204-1244
-------
could be so operated. For example, a boiler fitted with overfire air ports
would be tested with variations in the overfire air flow. (This was the
primary NO control for contract IAG-D7-E681.)
The data sources were reviewed to obtain NOV levels achieved by
3\
lowering total excess air (LEA) (as opposed to maintaining total air flow but
biasing the air flow to selected burners). This technique has been an effec-
tive and widely applied NOX control.
Before discussing the data it is appropriate to expand on the LEA
applicability consideration. Figure 2-7 is a sketch of NO as a function of
excess oxygen (a measure of excess air). The NO reduction depends on:
X
. The baseline operating point.
. The amount by which excess oxygen can be reduced.
Consider Point A as the baseline operating condition and that a margin
above the CO/smoke limit is to be maintained (to accommodate rapid load
changes and/or fuel property variability). Then, it is apparent that LEA
would not be an applicable NO control for this particular boiler. With Point
B as the baseline operating condition, some amount of NO reduction could be
achieved. Implementing LEA with Point C as baseline could actually increase
NO emissions as the excess oxygen was reduced toward Point D.
X
In summary, although LEA has been shown to be an effective and widely
applicable NOV control, there are very definite instances where its implemen-
X
tation will produce little or no NOV reduction or even increase NO,, emission,
X X
i.e., the applicability of LEA is less than 100 percent and must be determined
on a boiler-by-boiler basis.
The data sources were reviewed to establish a data base describing low
NO operation resulting from the implementation of LEA. Table 2-3 summarizes
X
the data base with each point satisfying the three conditions discussed in
subsection 2.2. The baseline results (from Table 2-1) are repeated for those
boilers for which LEA did not produce a N0x reduction since these data were
considered to correspond to a controlled condition. Also shown in Table 2-3
are the NOX reductions and efficiency changes resulting from LEA
implementation.
2-21 KVB11-34204-1244
-------
to
I
to
to
i
to
to
O
o
*
ro
TJ
EXCESS OXYGEN, PERCENT
Figure 2-7. NO as a function of excess oxygen (not to scale)
I
I-1
SJ
-------
TABLE 2-3.
DATA BASE FOR LOW NO OPERATION
x
to
I
to
w
I
U)
to
o
Rated
Steam Flow Boiler Burner
Line Location Unit
1
2
3
4
5
6
7
a
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1
1
1
1
2
2
3
4
5
5
6
7
9
9
9
10
10
11
12
12
13
14
14
15
15
IS
1
2
3
1
2
4
2
4
716-3
248-3
3
3
BC-1
BC-1
VA-1
4
5
1
24
20
2
1
4
123-1
32-10
32-13
103 Ib/hr
29.0
29.0
30.0
29.0
59.2
65.0
10.0
20.0
25 .0
10.0
158.0
85.0
60.0
160.0
300.0
60.0
110.0
135.0
225.0
325.0
500.0
150.0
200.0
19.2
60.0
60.0
kg/s
3.7
3.7
3.8
3.7
7.5
8.2
1.3
2.5
3.2
1.3
19.9
10.7
7.6
20.2
37.9
7.6
13.9
17.0
28.4
41.0
63.1
18.9
25.3
2.4
7.6
7.6
Type
wt
wt
wt
wt
wt
wt
ft
ft
wt
ft
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
No ./Type
I/Ring
I/Ring
1/Stn.Ata.
I/Ring
1/Stn.Atn.
6/Rlng
6/Rlng
I/Ring
1/Rlng
I/Ring
I/Ring
4/Stm.Ata
4/Stm.Ata.
4/Ring
I/Triple
4/Ring
2/Ring
I/Double
3/SS
8/-
8/PC
8/PC
8/1-
6/PC
5/SS
6/SS
I/Rot .Cup
7/US
7/US
Fuel
NG
NG
•2
NG
12
NG
NG
NG
NG
NG
NG
12
»5
NG
NG
NG
NG
NG
Coal
NG
Coal
Coal
NG
Coal
Coal
Coal
NSF
Coal
Coal
IN
By
Wt. Teat No.
0
0
0.04S
0
0.045
0
0
0
0
0
0
0.018
0.32
0
0
0
0
0
1.49
0
1.40
1.34
0
1.35
1.33
1.80
0.26
1.40
1.40
12-8
5-1
66-5
106-1
107-1
13-3
69-1
41-3
38-7
4-1
37-8
65-1
6-1
15-1
24-7
30-5
14-1
80-11
18-6
75-7
26-2
78-1
77-13
31-1
27-4
28-6
3-2
16-2
12-7
*lr Load
•F
Amb
Amb
350
Amb
Amb
Amb
Amb
Amb
Amb
Amb
Amb
300
240
400
322
401
Amb
Amb
Amb
640
653
630
640
445
350
Amb
Amb
Amb
Amb
K 103 Ib/hr kg/8 %
22.5
22.0
450 24 .0
24.0
23.5
47.5
53.0
7.0
14.0
20.0
8.0
422 115.0
389 62 .5
478 46.0
434 135.0
478 248.0
48.5
85.0
114.0
611 180.0
618 183.0
606 260 .0
611 260.0
503 400.0
450 123.0
150.0
14.5
47.0
54.0
2.9
2.8
3.0
3.0
3.0
6.0
6.7
0.9
1.8
2.5
1.0
14.5
7.9
S.8
17.0
31.3
6.1
10.7
14.4
22.7
23.1
32.8
32.8
50.5
15.5
19.0
1.8
6.0
6.8
77.6
75.9
80.0
82.8
81.0
80.2
81.5
70.0
70.0
80.0
80.0
72.8
73.5
76.7
84.4
82.7
80.8
77.3
84.4
80.0
81.3
80.0
80.0
80.0
82.0
75.0
75.5
78.3
90.0
NOX pp*
O2 dry 8 CO
% 3»02 ppm
1.9
3.4
2.8
2.6
3.1
4.0
3.8
7.5
4.8
2.9
S.1
5.2
6.5
2.6
2.6
2.5
5.2
8.1
4.9
6.1
4.5
5.8
3.5
9.8
8.9
8.9
5.3
6.0
7.9
65
70
104
82
79
132
101
92
91
76
56
185
297
241
339
171
104
94
338
200
364
485
287
578
470
358
142
297
221
114
159
17
0
120
133
102
55
—
180
0
11
50
20
39
24
0
0
157
—
0
0
0
0
0
0
282
0
0
Eff.
t
—
80
83
80
—
76
~
~
—
78
—
84
—
79
83
—
80
70
82
84
86
86
85
81
82
83
78
—
™
Reduc .
7.1
9.1
15.5
NA
NA
NA
NA
NA
15.7
NA
NA
NA
NA
NA
9.4
5.5
NA
NA
9.4
NA
5.0
NA
12.2
NA
14 .S
34.0
NA
10.3
25.3
Eff.
i Change*
—
+3.9
+2.5
NA
NA
NA
NA
NA
—
NA
NA
NA
NA
NA
+9.2
—
NA
NA
0
NA
0
NA
0
NA
+ 1.2
+3.8
NA
~
—
(continued)
-------
TABLE 2-3 (CONTINUED)
to
I
to
I
U)
*>
(O
o
*>.
i
M
10
i&
•C.
Rated
Stean Flow _ Boiler Burner
Line Location Unit
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
IS
16
17
18
18
18
19
20
20
21
21
23
26
27
29
37
38
39
19
38
2-1
2
T-8
2
3
4
1
4
42
2
3
1
1
1
5
2
2
B10B
—
—
103 Ib/hr
17.0
65.0
110.0
90.0
105.0
160.0
17.5
80.0
400.0
50.0
75.0
7.0
18.0
100.0
150.0
40.0
45.0
200.0
17.5
45.0
kg/8
2.1
8.2
13.9
11.4
13.3
20.2
2.2
10.1
50.5
6.3
9.5
0.9
2.3
12.6
18.9
5.1
5.7
25.3
2.2
5.7
Type
ft
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
ft
ft
wt
wt
wt
wt
wt
wt
wt
No J Type
I/Rot .Cup
2/Stn.Atn.
2/Stn.Atn.
3/Stn.Atn.
4/Stn.Atn.
4/Stn.Atn.
1/Stn.Atn.
I/Ring
I/Double
2/Cyclones
2/SS
3/SS
I/Ring
I/Ring
1/Stn.Atn.
I/Ring
1/Stn.Atn.
2/Ring
2/Stn.Atn.
2/Stn.Atn.
1/Rlng
1/Stn .Atn .
1/Spud
1/Stn.Atn.
I/Ring
1/Ring
1/Stn .Atn .
Fuel
NSF
te
»2
16
16
«6
16
12
NG
16
16
Coal
Coal
Coal
NG
NG
12
NG
PS300
NG
16
•6
NG
16
NG+RG
12
•6
NG
NG
•6
%N
By
wt.
0.28
0.29
0.01
0.026
0.26
0.26
0.44
0.006
0
0.37
0.30
1.5S
1.45
1.43
0
0
0.020
0
0.77
0
0.31
0.30
0
0.49
0
0.008
0.25
0
0
0.31
TAir
Test No.
36-2
10-1
7-1
9-6
21-8
22-9
1-4
52-1
191-3
8-2
172-2
32-1
19-9
20-4
40-1
48-2
57-1
109-1
111-8
113-1
117-2
179-1
180-2
187-5
207-3
19-76
19-132
19-147
200G-2
201-12
•F
Anb
Anb
Amb
Anb
415
542
Anb
Anb
Anb
Anb
Anb
547
Anb
Anb
Anb
Anb
Anb
Anb
Anb
375
388
231
350
320
Anb
Anb
Anb
Anb
283
284
Load
K 103 Ib/hr kg/s %
15.0
50.0
88.0
72.0
486 79.5
557 120.0
14.5
14.0
14.0
60.0
63.5
559 320.0
41.0
62.0
6.1
15.3
15.7
75.0
82.0
464 120.0
471 122.0
384 32 .5
450 40.0
433 37.0
163.0
14.5
14.2
14.5
413 40.0
413 38.8
1.9
6.3
11.1
9.1
10.1
15.2
1.8
1.8
1.8
7.6
8.0
40.4
5.2
7.9
0.8
1.9
2.0
9.5
10.3
15.2
15.4
4.1
5.1
4.7
20.6
1.8
1.8
1.8
5.1
4.9
88.2
76.9
80.0
80.0
75.7
75.0
82.9
80.0
80.0
75.0
79.4
80.0
82.0
82.7
87.1
85.0
87.2
75.0
82.0
80.0
81.3
81.3
88.9
82.2
81.5
83.0
81.1
83.0
88.9
86.2
N'
- °2
6.7
3.7
5.3
7.0
6.1
6.0
2.7
3.6
2.0
5.2
2.7
3.0
5.8
5.5
S.O
8.3
8.0
6.6
5.9
5.4
3.1
3.8
1.9
1.6
3.1
1.1
0.98
3.0
1.6
1.6
Ox ppn
dry 9 CO
3%O2 ppn
185
180
164
216
225
233
338
71
55
328
255
793
330
359
76
53
118
113
401
155
246
179
220
243
181
97
150
95
171»
230»
126
0
0
26
0
96
0
36
76
0
201
0
24
104
203
14
86
0
0
0
0
0
0
120
50
181
183
4
140
65
Eff.
t
72
—
—
83
85
86
—
—
ao
80
83
—
82
—
—
—
80
82
—
—
88
—
85
87
—
83
84
78
81
86
NOX Eff .
, Reduc. Change,
% t
NA
NA
NA
12.2
22.7
3.7
20.1
NA
6.8
NA
1.5
NA
30.7
29.1
NA
NA
NA
NA
12.5
NA
16.3
8.2
NA
25.5
5.7
19.2
29.9
NA
NA
21.0
NA
NA
NA
+ 1.2
—
0
—
NA
—
NA
—
NA
—
—
NA
NA
NA
NA
NA
NA
+ 1.1
—
NA
0
—
+ 1.1
+ 1.2
NA
NA
+0.94
(continued)
-------
TABLE 2-3 (CONTINUED)
to
to
ui
Rated
Stean Flow Boiler Burner
Line
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
Location Unit
Alma *3
U.Hiac./ «2
Stout
U.Hiac./ (1
Eau Claire
U.Hiac./ »2
Kadi eon
Hillnar *3
Fairmont *3
Fremont *6
3
1
2 —
3
4 —
A •"
B —
C
D —
—
F
—
G —
—
H
103 Ib/hr
230.0
45.0
60.0
100.0
160.0
80.0
160.0
100.0
100.0
90.0
260.0
130.0
300.0
200.0
182.5
90.0
80.0
75.0
45.0
kg/8
29.0
5.7
7.6
12.6
20.2
10.1
20.2
12.6
12.6
11.4
32.8
16.4
37.9
25.3
23.0
11.4
10.1
9.47
5.68
Type
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
wt
No./ Type
4/PC
VB
US
3/SS
6/SS
4/SS
4/PC
SS +
Pneumatic
3/SS
3/Stm.Atn.
4/PC
6/SS
7/SS
5/SS
7/SS
VG
3/SS
3/SS
OS
Fuel
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal+
Hood
Coal
*6
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
%N
By
Wt. Teat No.
1.09
0.91
1.28
1.24
0.78
1.30
1.37
1.06
1.25
0.93
1.22
1.27
0.24
1.25
0.77
0.83
1.44
1.04
1.24
1.23
1.23
1.12
0.84
1.04
1.04
42
11
25
30
9
15
34
9
5
11
3
1/26
4/9
6/1
—
30
13C
9
5
15
4
34
2
23
7A
TAir
•F K
Preheat
Amb
Amb
Amb
Amb
Amb
Preheat
Amb
Preheat
Preheat
Preheat
Amb
Preheat
Preheat
Preheat
Preheat
Amb
Preheat
Amb
Amb
Amb
Amb
Amb
Amb
Amb
Load
103 Ib/hr kg/a
200.0
33.0
40.0
49.0
80.0
90.0
133.0
62.6
134.0
115.0
82.0
77.0
72.5
198.0
96.0
240.0
144.0
163.0
77.4
79.2
60.0
60.0
63.8
57.0
34.2
25.2
4.2
5.1
6.2
10.1
11.4
16.8
7.9
16.9
14.5
10.3
9.72
9.15
25.0
12.1
30.3
18.2
20.6
9.77
10.0
7.58
7.58
8.05
7.20
4.32
%
87.0
73.3
88.9
81.7
80.0
90.0
83.4
78.3
83.4
71.8
82.0
77.0
81.0
76.0
74.0
80.0
72.0
89.0
86.0
88.0
75.0
75.0
85.0
76.0
76.0
- °2
2.9
4.5
5.4
6.6
6.2
9.1
5.9
6.5
5.4
4.3
8.2
6.8
6.6
6.4
8.4
4.0
4.7
8.9
7.9
7.8
7.8
6.2
8.9
8.0
8.2
NOX ppm
dry 9 CO
3%02 ppm
596*
165*
228*
218*
303*
449*
346*
290*
6791
402*
189*
281*
234*
678*
319*
483*
2041
302*
233*
228*
325
229
337*
444*
217*
7
—
—
102
380
215
216
155
20
12
387
103
19
34
—
383
28
66
89
39
137
96
—
—
56
Bff.
—
—
—
—
74
76
—
80
87
88
83
—
—
—
—
—
—
82
83
85
—
—
74
76
— —
Reduc
1
8.8
NA
NA
NA
NA
NA
19.0
19.9
NA
20.1
20.4
33.0
NA
NA
27.5
NA
27.3
NA
NA
NA
5.5
31.2
NA
NA
32.1
Eff .
. Change,
__
NA
NA
NA
NA
NA
—
+0.9
NA
+1.0
+ 1.5
—
NA
NA
—
NA
—
NA
NA
NA
—
—
NA
NA
--
i
to
10
O
Mote 1 NA - Not applicable
Note 2 See Table 2-1 for symbol definition
•Total NOX not measured but taken aa three percent more than NO for coal) two percent for oili and six percent for gas (Ref. 2-1).
(O
-------
The analysis of the low NOX data base followed that performed for the
baseline data base. Indeed, the separation of the data by boiler and fuel
type (Table 2-2) is directly applicable to the low NOX results. The conclu-
sion regarding the insufficiency of firetube boiler results is still valid.
The mean NO levels achieved by the ambient and preheated boilers with
gas fuels (%N = 0) were 97 ppm and 223 ppm, respectively. These did not
differ significantly from the corresponding baseline values, i.e., the overall
effect of implementing LEA on the entire boiler sample firing gas fuel did not
produce a significant decrease in NO. This is not to be interpreted to mean
that LEA was ineffective on all the boilers (a maximum NO reduction of 15.7%
X
was achieved on gas fuel) but rather that the boiler sample, as a whole,
showed only a minor NO reduction (2 ppm for ambient units and 11 ppm for
preheated units).
Figures 2-8 and 2-9 are scatter plots of the low NO data for the
ambient and preheated boilers, respectively. Also shown in these figures are
the correlation equations (in the form NO = a + b in (%N + 1)) and the 95%
confidence limits on the mean values. The parameters from the correlations
were:
Ambient
Preheat
% N Range
0-1.80
0-1.55
a
142.1
131.0
b
195.5
445.8
r
0.637
0.755
r2(%)
40.5
57.0
The coefficients of determination (r ) for low NOX operation compare favorably
with those calculated for baseline conditions, i.e.:
Ambient: 40.5% vs. 49.3%
Preheat: 57.0% vs. 61.8%
A comparison of the data for both ambient and preheated boilers
indicated that there was no statistical difference between the baseline and
low NO operating modes. This implies that the functional dependency of NO
X X
with fuel nitrogen content is not statistically dependent on whether the
boiler was operated at baseline or low NO (lower excess air) conditions.
This may be surmised by examining Table 2-3 and noting the number of boilers
for which lower excess air was ineffective.
2-26 KVB11-34204-1244
-------
600
0.2 0.4 0.6 0.8 1.0 1.2
FUEL NITROGEN CONTENT, % BY WEIGHT
1.4 1.6
1.8
Figure 2-8. Ambient combustion air watertube industrial boilers - low NO
operation. }
2-27
KVB11-34204-1244
-------
800
0.2 0.4 0.6 0.8 1.0 1.2
FUEL NITROGEN CONTENT, % BY WEIGHT
+95% CL
Mean
-95% CL
1.4 1.6
Figure 2-9. Preheated combustion air watertube industrial boilers - low
NO operation.
2-28
KVB11-34204-1244
-------
A separate analysis was performed for the 13 ambient temperature
combustion air spreader stokers operating in a low NOX mode. Of these seven
boilers, nine were responsive to LEA, achieving NO reductions ranging from
X
5.5% to 34.0% with the mean NO reduction achieved by all 13 boilers being
16.9%. The analysis indicated a correlation coefficient (r) of 0.024, and
that r was not statistically different from zero, i.e., the NOX measured for
the 13 spreader stokers was not dependent on the fuel nitrogen content. This
result allows the NOX levels to be specified in terms of the mean measured
NO —328.3 ppm and the 95% confidence limits—+ 43.2 ppm.
The seven preheated pulverized coal results obtained during low NO
operation were analyzed for a dependency of NO on fuel nitrogen content
(0.93 < %N < 1.40). Pour of the seven boilers achieved a NO reduction when
i^_ ^BV J£
LEA was implemented (reductions ranged from 5.0% to 20.4%); however, the
analysis showed no correlation between NOX and %N. As previously discussed,
this may be due to the relatively narrow range in fuel nitrogen content.
Thus, the seven preheated pulverized coal boilers, operating in a low NO
mode, could be characterized in terms of the mean NO,.—540.3 ppm and the 95%
A
confidence limits--+ 117.0 ppm.
2-29 KVB11-34204-1244
-------
SECTION 3.0
CONCLUSIONS AND RECOMMENDATIONS
This study has addressed the dependency of industrial boiler NOV
A.
emissions on a single variable—fuel nitrogen content. It is concluded that
such a dependency exists and has been quantified by correlating the measured
data in the form:
NO = a + b Jin (%N + 1)
x
The degree of correlation, as measured by the coefficient of
determination, is such that fuel nitrogen can explain a large percentage of
the data scatter. It is recognized that other variables, e.g., boiler
cleanliness, fuel oil atomization details, heat release/furnace volume, burner
spacing, etc., have an effect on boiler NOX emissions. An extensive effort
would be required to quantify these variables.
3.1 CONCLUSIONS
The main conclusions established by this study are:
1. Insufficient data is available from KVB tests to
characterize NO vs. %N for the following boiler types:
firetubes; cyclone burners; vibrating grates; overfed
stokers; and underfed stokers.
2. The data base precludes establishing NO vs. %N for
pulverized coal boilers primarily because of the limited
range in fuel nitrogen tested.
3. Fuel nitrogen, by itself, is responsible for 49.3% of all
the NO data scatter during baseline operation of ambient
temperature combustion air watertube boilers. This was
established for the full range of fuel nitrogen tested
(0-1.80%).
4. For low NOX operation (lower excess air implemented) of
ambient watertube boilers, correlation with fuel nitrogen
explained 40.5% of the NOX data scatter.
3-1 KVB11-34204-1244
-------
5. The functional dependency of NO on fuel nitrogen for
ambient and preheated watertube boilers was not
statistically different for operation at baseline or low NO
conditions* This conclusion must be viewed with caution
since it was based on a rather restricted data base.
6. Fuel nitrogen, by itself, can explain 61.8% of the NOy data
scatter for baseline operation of preheated combustion air
watertube boilers for the full range of fuel nitrogen tested
(0-1.55%).
7. Low NO operation of preheated watertube boilers resulted in
a fuel nitrogen correlation which explained 57.0% of the NO
data scatter.
8. NO emissions from ambient spreader stoker boilers during
baseline operation is probably dependent on both fuel
nitrogen (0.78 _£ %N <_ 1.8) and excess oxygen. Low NO
operation of spreader stokers resulted in the NO being
independent of fuel nitrogen content.
3 . 2 RECOMMENDATI ONS
1. The data sources utilized in this study were limited to those
prepared by KVB. It is recommended that other publically available data be
obtained and reviewed for inclusion in the data base. It is especially impor-
tant to secure data for firetube boilers operating with high nitrogen fuel
oils.
2. The logarithmic correlation function:
NO = a + b fcn (%N + 1)
x
was selected since it was more physically reasonable than either a power
function :
or a quadratic:
NO = a (%N)
NO = a + b (%N) + c
x
It is recommended that other functions be investigated to determine if they
can better correlate the NOX measurements with fuel nitrogen content.
3-2 KVB11-34204-1244
-------
3. It was noted that the spreader stoker baseline NOX data appeared
to be a function of both fuel nitrogen and excess oxygen. It is recommended
that these, and other, data be correlated in terms of both variables, i.e.:
NO = f (%N, %0 )
X ^
It is anticipated that correlations with more than one independent variable
will require a high-speed digital computer.
4. Industrial boiler NO emissions, as has been previously discussed,
are most likely a function of an extensive array of independent variables:
N0x= f (x,, x2, .", xn )
and the present study has addressed only one of these variables—fuel nitrogen
content. An extensive effort would be required just to obtain data of suffi-
cient quantity and quality to perform the multi-dimensional correlation. It
is recommended that this matter be considered in the context of a long-term
effort.
3-3 KVB11-34204-1244
-------
SECTION 4.0
REFERENCES
2-1. Cato, G. A. et al., "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial Boilers -
Phase I," EPA-650/2-74-078a, October, 1974.
2-2. Cato, G. A. et al., "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial Boilers -
Phase II," EPA-600/2-76-086a, April, 1976.
2-3. Hunter, S. C., and Buening, H. J., "Field Testing: Application of
Combustion Modifications to Control Pollutant Emissions from
Industrial Boilers - Phases I and II (Data Supplement),"
EPA-600/2-77-122, June, 1977.
2-4. Maloney, K. L. et al., "Low-Sulfur Western Coal Use in Existing Small
and Intermediate Size Boilers," EPA-600/7-78-153a, July, 1978.
2-5. Carter, W. A. et al., "Emission Reduction on Two Industrial Boilers
with Major Combustion Modifications," EPA-600/7-78-099a, June, 1978.
2-6. Fisher, K. T., "Emission Reduction on Two Industrial Boilers with
Major Combustion Modifications (Data Supplement)," EPA-600/7-78-099b,
December, 1978.
2-7. Carter, W. A. et al., "Application of Advanced Combustion Modification
Technology to Industrial Process Equipment," Progress Report No. 27,
November 15, 1979.
2-8. Carter, W. A. et al., "Thirty-Day Field Tests of Industrial Boilers:
Site 1—Coal-Fired Spreader Stoker," EPA-600/7-80-085a, April 1980.
2-9. Carter, W. A. et al., "Thirty-Day Field Tests of Industrial Boilers:
Site 2~Residual-Oil-Fired Boiler," EPA-600/7-80-085b, April 1980.
2-10. Carter, W. A. et al., "Thirty-Day Field Tests of Industrial Boilers:
Site 3—Pulverized-Coal-Fired Boiler," EPA-600/7-80-085c, April 1980.
2-11. Carter, W. A. et al., "Thirty-Day Field Tests of Industrial Boilers:
Site 4—Coal-Fired Spreader Stoker," EPA-600/7-80-085d, April 1980.
2-12. Gabrielson, J. E. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site A,"
EPA-600/7-78-136a, July 1978.
4-1 KVB11-34204-1244
-------
2-13. Gabrielson, J. E. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site B,"
EPA-600/7-79-041a, February 1979.
2-14. Gabrielson, J. E. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site C,"
EPA-600/7-79-130a, May 1979.
2-15. Gabrielson, J. E. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site D,"
EPA-600/7-79-237a, November 1979.
2-16. Langsjoen, P. L. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site E,"
EPA-600/7-80-064a, March 1980.
2-17. Langsjoen, P. L. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site F,"
EPA-600/7-80-065a, March 1980.
2-18. Langsjoen, P. L. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site G,"
EPA-600/7-80-082a, April 1980.
2-19. Langsjoen, P. L. et al., "Field Tests of Industrial Stoker Coal-Fired
Boilers for Emissions Control and Efficiency Improvement - Site H,"
EPA-600/7-80-112a, May 1980.
2-20. Chatfield, C., Statistics for Technology; A Course in Applied
Statistics, Halsted Press Book (1978).
4-2 KVB11-34204-1244
------- |