EPA-650/2-74-078-0
FIELD TESTING:
APPLICATION
OF COMBUSTION MODIFICATIONS
TO CONTROL
POLLUTANT EMISSIONS
FROM INDUSTRIAL BOILERS - PHASE I
by
G. A. Cato; H. J. Buening; C. C. DeVivo; B. G. Morton; J. M. Robinson
KVB Engineering, Inc.
17332 Irvme Boulevard
Tustin, California 92680
Contract No. 68-02-1074
ROAP No. 21BCC-046
Program Element No. 1AB014
EPA Project Officer: R.E.Hall
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 1974
-------
This report ha:s been reviewed by the Environmental Protection Agency
and approved for publication. Approval does n&t signify that the
I
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.
I
I
I
I
I
i
I
I
I
I
11
-------
I
I
I
I
ACKNOWLEDGMENTS
The authors wish to acknowledge the assistance of Mr. Robert E. Hall,
the EPA Project Officer, whose direction and evaluations were of great benefit.
Acknowledgment is also made to the active cooperation and advice of
Mr. W. H. Axtman of the American Boiler Manufacturers' Association and to
the ABMA members who offered a forum for discussion of the program and con-
structive criticism. Also of assistance were the American Petroleum
Institute, the American Gas Association, and the Naval Civil Engineering
Laboratory.
Special thanks is due the following organizations who volunteered
their boilers as test units:
Baltimore Gas and Electric, Baltimore, MD
Commonwealth of Kentucky, Frankfort, KY
E. I. du Pont de Nemours & Co., Wilmington, DE
The Firestone Tire and Rubber Co., South Gate, CA
Peabody Gordon-Piatt, Winfield, KN
Great Northern Paper Co., Cedar Springs, GA
Industrial Combustion, Inc., Winton, WI
Keeler Co., Williamsport, PA
Kewanee Boiler Corporation, Kewanee, IL
Eastman Kodak Company, Rochester, NY
Lever Brothers Co., Los Angeles, CA
North American Rockwell, Los Angeles Div., Los Angeles, CA
Department of the Air Force, Norton Air Force Base, CA
Texaco, Inc., Wilmington, CA
U. S. Navy Base, Charleston, SC
U. S. Naval Air Station, Patuxent River, MD
University of California, Irvine, CA
University of California, Los Angeles, CA
The assistance of the KVB staff was invaluable. In particular, the
direction and advice of Dr. B. P. Breen and Mr. N. Bayard de Volo was of great
benefit.
111
-------
I
I
I
I
CONTENTS
1.0 SUMMARY 1
1.1 Objective and Scope 1
1.2 Phase I Results 3
1.2.1 Nitrogen Oxides Emissions 8
1.2.2 Particulate Emissions 13
1.2.3 Sulfur Oxides Emissions 16
2.0 TEST BOILER SELECTION 19
3.0 INSTRUMENTATION AND TEST PROCEDURES . 25
3.1 Gas Sampling and Conditioning System 25
3.2 Instrumentation 28
3.2.1 Total Nitrogen Oxides 30
3.2.2 Carbon Monoxide and Dioxide 33
3.2.3 Oxygen 35
3.2.4 Total Hydrocarbons 35
3.2.5 Total Sulfur Oxides 36
3.2.6 Particulates 38
3.2.7 Smoke 39
3.3 Calibration 40
3.4 Test Procedures 40
4.0 DISCUSSION OF TEST RESULTS 45
4.1 Coal Fuel 60
4.1.1 Nitrogen Oxide Emissions 60
4.1.2 Particulate Emissions 62
4.2 Oil Fuel 64
4.2.1 Nitrogen Oxide Emissions 64
4.2.2 Particulate Emissions ' 67
4.3 Natural Gas Fuel 69
4.3.1 Nitrogen Oxides Emissions 69
4.3.2 Particulate Emissions 71
4.4 Mixed Fuel 71
4.5 Ratio of NO- Concentration to Total NOx Concentration 73
4.6 Carbon Monoxide Emissions 73
4.7 Hydrocarbon Emissions 75
4.8 Particle Size 76
4.9 Bacharach Smoke Spots 83
4.10 Boiler Efficiency 83
v
-------
CONTENTS
5.0 EFFECTS OF OPERATIONAL CHANGES 85
5.1 Mixture Ratio Control 85
5.1.1 Excess Oxygen/Air 85
5.1.2 Off-Stoichiometric Firing 89
5.1.3 Air Register Adjustments 96
5.2 Air Preheat Temperature 101
5.3 Fuel Oil Temperature 104
5.4 Firing Rate (Percent Load) 105
6.0 FUEL PROPERTIES 109
6.1 Fuel Nitrogen Content 114
6.2 API Gravity 118
6.3 Carbon Residue 118
6.4 Sulfur Content 120
7.0 BOILER DESIGN CHARACTERISTICS 125
7.1 Burner Design 125
7.1.1 Oil Atomization 125
7.1.2 Coal Burner 134
7.1.3 Natural Gas Burners 134
7.1.4 Burner Heat Release Rate 135
7.2 Furnace Design 138
7.2.1 Firetube Versus Watertube Boilers 138
7.2.2 Furnace Volume and Area 140
7.3 Boiler Efficiency 145
8.0 STATISTICAL ANALYSIS OF DATA 149
8.1 Analysis Strategy 149
8.2 Regression Analysis 153
8.3 Discussion of Statistical Results 157
9.0 REFERENCES 161
10.0 GLOSSARY OF TERMS 163
11.0 CONVERSION FACTORS 169
APPENDIX 177
VI
-------
I
I
I
I
LIST OF FIGURES
Figure
No. Caption Page
1-1 Total Oxides of Nitrogen Concentration at Baseload 9
1-2 Total Particulate Emissions at Baseload for Natural
Gas, Oil and Coal Fired Boilers 15
1-3 Total Sulfur Oxides Emissions at Baseload for Oil and
Coal Fired Boilers 17
2-1 Field Measurement Route and Test Site Locations 24
3-1 Laboratory Trailer Floor Plan and Side Wall Elevation 26
3-2 Flue Gas Sampling and Analyzing System 27
3-3 Exterior and Interior View of Mobile Air Pollution
Reduction Laboratory 29
3-4 Flue Installation of Sulfur Oxides Analyzer 37
3-5 Sulfur Oxides Sample Collection Apparatus 37
4-1 Total Nitrogen Oxide Emissions at Baseload for Coal-
Fired Boilers 61
4-2 Total Particulate Emissions at Baseload for Coal-
Fired Boilers 63
4-3 Baseline Total Nitrogen Oxide Emissions for Oil-Fired
Watertube Boilers 65
4-4 Total Particulate Emissions at Baseload for Natural
Gas and Oil-Fired Boilers 68
4-5 Baseline Total Nitrogen Oxides Emissions for Natural
Gas Fired Boilers 70
4-6 Percentage of Nitrogen Dioxide Concentration in Total
Nitrogen Oxides Concentration 74
4-7 Distribution by Percentage of Catch of the Particulate
Optical Diameter. Natural Gas Fuel. 77
4-8 Distribution by Percentage of Catch of the Particulate
Optical Diameter. No. 6 Oil Fuel. 78
4-9 Distribution by Percentage of Catch of the Particulate
Optical Diameter. Coal Fuel with Pulverizer and Cyclone
Burners. 79
4-10 Distribution by Percentage of Catch of the Particulate
Optical Diameter. Coal Fuel with Underfed and Spreader
Stoker Burners. 80
4-11 Electron Microscope Photographs of Fly Ash from Coal
Fuel and from Oil Fuel 82
VI1
-------
LIST OF FIGURES (cont.)
Figure Caption Page
No.
5-1 Effect of Excess Oxygen on Nitrogen Oxides Emissions
for Firetube Boilers at Baseline Operating Conditions,
Natural Gas and Oil Fuels 86
5-2 Effect of Excess Oxygen in Nitrogen Oxides Emissions
for Watert.ube Boilers at Baseline Operating Conditions,
Natural Gc.s, Oil and Coal Fuels 88
5-3 Effect of Burner Theoretical Air Level on Nitrogen
Oxides Formation 93
5-4 Typical Arrangement of Corner Burner Showing Secondary
Air Distribution to Coal, Oil/Gas and Air Compartments 99
5-5 The Effect of Combustion Air Temperature on Baseline
Nitrogen Oxides Emissions for Natural Gas, Oil and Coal
Fuels 102
5-6 Effect of Fuel Oil Temperature on Total Nitrogen Oxides
Emissions 104
5-7 Effect of Firing Rate on Total Nitrogen Oxides Emissions,
Firetube Boilers ] 06
5-8 Effect of Firing Rate on Total Nitrogen Oxides Emissions,
Gas-Fueled Watertube Boilers 107
6-1 Effect of Fuel Nitrogen Content 115
6-2 Effect of API Gravity on Baseload Nitrogen Oxides
Emissions and Particulates 119
6-3 Effect of Fuel Oil Carbon Residue on Baseload Particulate
Emissions 12i
6-4 Total Sul::ur Oxides Emissions at Baseload for Oil and
Coal Fired Boilers 122
6-5 Ratio of Sulfur Trioxides to Total Sulfur Oxides at
Baseload
-------
I
I
I
I
LIST OF FIGURES (cont.)
Figure
No. Caption Page
7-6 Effect of Furnace Heat Release Volume on Total Nitrogen 142
Oxides Emissions for Oil-Fired Watertube Furnaces 142
7-7 Effect of Furnace Heat Release Volume on Total Nitrogen
Oxides Emissions. Coal Fuel. 144
7-8 Variation of Boiler Efficiency with Excess Oxygen Firing
Level 147
7-9 Effect of Low NOx Operation on Boiler Efficiency 148
11-1 Flue Gas Composition as a Function of Excess Air for
Natural Gas Fuel 171
11-2 Flue Gas Composition as a Function of Excess Air for
Oil Fuels 172
11-3 Flue Gas Composition as a Function of Excess Air for
Coal Fuels 173
A-l Regional Boundaries used in Table A-3 185
A-2 Combined Emissions (1967) from Intermediate Boilers
by Type and Fuel 185
A-3 Federal Power Commission Regions 196
IX
-------
I
I
I
I
LIST OF TABLES
Table
No.
1-1
1-2
2-1
2-2
2-3
3-1
3-2
4-1
4-2
4-3
5-1
5-2
6-1
6-2
7-1
7-2
8-1
8-2
8-3
Caption
Summary of the Effect of Variations in Parameters
and Operating Conditions on NOx Emissions
Data Summary
Distribution of Test Boilers and Test Number Assignments
Number of Test Sets Based on Population and Total
Capacity
Distribution of Fifty Test Boilers by Capacity and Fuel
Emission Measurement Instrumentation
Statistical Evaluation of NO vs. NO
HCXi COLD
Field Test Measurements
Mixed Fuels
Optical Size Distribution of Gas, Oil and Coal Fly Ash
Effect of Burners Out of Service
Effect of Air-Fuel Mixing by Changing the Air Register
Setting
Fuel Analysis Summary
Effect of Fuel Oil Grade on Total Nitrogen Oxides
Emissions and Conversion of Fuel Nitrogen to Total
Nitrogen Oxides Emissions
Test Boiler Design Characteristics
Effect of Oil Atomization Method on Total Nitrogen
Oxides, Particulate Emissions and Boiler Efficiency
Number of Samples for Various Categories
Number of Samples Used in Regression Analysis
Summary of Regression Analysis
Page
4
5
20
22
22
25
34
46
72
81
90
97
111
117
126
129
154
155
156
XI
-------
LIST OF TABLES (cent.)
Table
No. Caption Page
11-1 Emissions Units Conversion Factors 170
11-2 Emissions Units Conversion Factors for Typical
Natural Gc.s Fuel (HV = 23,440 Btu/lb) 174
11-3 Emissions Units Conversion Factors for Typical
Fuel Oil IHV = 19,100 Btu/lb) 175
11-4 Emissions Units Conversion Factors for Typical
Coal Fuel (HV = 12,280 Btu/lb) 176
A-l Distribution of Boilers in Service in the United
States, Circa 1972 177
A-2 Population Breakdown by Fuel Capability (Percentage
Basis) All Industrial Boilers Now in Service 181
A-3 Summary of Capacity, Fuel and Emissions by Boiler
Type and 1-ocation in 1967 182
A-4 Number of Industrial-Size Watertube Boiler Sales
1965 to May 1973, and the Fuel Burned 183
A-5 Capacity of Boilers by Type and User 188
A-6 Estimated Trends of Boiler Types (Percentage Basis)
All Boilers Installed in Years Noted 189
A-7 Estimate Trends by Burner Type (Percentage Basis)
All Burners Installed in Years Noted, Including
Conversions 191
A-8 Estimated Trends by Fuel Capability (Percentage Basis)
All Commercial Industrial Boilers Installed in Years
Noted, Including Conversions 198
A-9 Distribution of Fifty Test Boilers by Capacity, Fuel
and Burner 199
Xll
-------
I
I
I
I
SECTION 1.0
SUMMARY
1.1 OBJECTIVE AND SCOPE
The objective of this program is to determine the effectiveness
of combustion modification techniques to control emissions of nitrogen
oxides (NOx) from industrial boilers of 10,000 to 500,000 pounds of steam
per hour capacity. The results are to be published in final report form
and also as guidelines to assist boiler manufacturers and users in design
and operation of boilers for reduced NOx emissions. In addition to NOx
concentrations, other measurements (particulate loading, sulfur oxides,
hydrocarbons, carbon monoxide, carbon dioxide, and smoke spots) were
made under both normal and reduced NOx emission firing conditions to
establish the magnitude and trend of outputs at different conditions.
The program consists of two phases and this is the final report
from Phase I. The first phase was one year in duration and involved the
selection of representative industrial boilers for testing, assembly of
a mobile analytical laboratory, and field measurement of emissions from
boilers operating normally and in low nitrogen oxides emission modes that
could be obtained without boiler hardware modifications.
The initial task was to make a survey of industrial boilers to
establish; (1) major manufacturers, (2) basic boiler and burner designs,
(3) general trends in boiler type, age, size, and use, (4) total number
and geographical distribution of industrial boilers in use in the United
States, and limitations in boiler uses and fuels, and (5) quantity and
geographical distribution of fuels being used in the United States. The
program scope allowed testing of approximately 50 boilers in Phase I.
It was apparent from the number (about 75,000) and diversity (variation
in furnace design, burner design and fuel characteristics) of industrial
boilers that testing only 50 units would limit the number of tests that
could be made on any one boiler/burner/fuel combination. As a result,
-------
the test plan was focused on measuring the magnitude of emissions from
many such combina.tions and the changes due to combustion modification.
The effects of other particular parameters, e.g., furnace volume, were
to be deduced as carefully as possible from engineering evaluation and
statistical analysis of the data while realizing that many complex inter-
actions of effects would make quantification difficult.
To provide the capability to measure the component emissions
discussed above a.t sites throughout the U.S., a mobile analytical labora-
tory was assembled. Instrumentation and equipment were selected, pur-
chased, and integrated into a system including sampling, sample condition-
ing, sample analysis, and data recording. The sampling and analysis
requirements for each individual component to be measured were considered
separately. Four different sampling systems were needed to provide
appropriate imputs for analysis methods ranging from continuous instru-
mental devices to complex gravimetric and wet chemistry procedures.
The field test portion of the program was then conducted on
industrial boilers in the size and type categories determined in the
selection task. The units tested were those selected from candidate
boilers volunteered by industrial companies and governmental organizations
without whose cooperation this program could not have been conducted. Forty-
seven individual boilers were tested; and some were tested with more than
one fuel and/or ourner; so that a total of 75 sets of test data on different
boiler/burner/fusl combinations were obtained. The data have been analyzed
to establish emission levels and the effect on emission levels of combustion
modifications, operational parameters, fuel characteristics, and design
variations.
Phase II of the program will be directed toward more intensive,
detailed testing of fewer units and may involve boiler modifications to
allow testing of NOx reduction techniques not possible with existing
boilers. Additional sampling and analysis requirements are also part of
Phase II. Partioulate size distribution will be measured before and after
control devices, if present, on thirteen boilers burning oil or coal.
-------
I
I
I
In addition, on ten of these thirteen boilers the concentrations of toxic
gaseous and solid substances will be measured in all fire-side input and
output streams. The distribution of the toxic elements among the differ-
ent particle sizes also will be determined.
1.2 PHASE I RESULTS
It was found that typical, existing industrial boilers have
limited flexibility to allow combustion modifications. This is due to
the small size and simple construction which frequently results in
boilers with single burners, fixed air swirl, unsophisticated control
systems, etc. However, in most cases, NOx reductions could be achieved
by off-stoichiometric combustion, for multiple burner units, by
changes in excess air level, burner adjustments, or other operational
parameters. The effect on NOx emissions of changes in fuel characteris-
tics and boiler design variations were also evaluated to the extent
possible with the available sample. Table 1-1 summarizes the effect of
various parameters on NOx emissions. The importance of the parameters
differed for natural gas, oil and coal fuels. For natural gas fired boilers,
the NOx emissions were independent of furnace dimensions. However-,
the furnace dimensions for oil and coal fired boilers were found
to affect the NOx emissions. For all three fuels the burner size in terms
of heat release rate was found to affect the NOx emissions. Correlations
between fuel properties and NOx emissions were obtained for oil fuel.
Off-stoichiometric firing, where it could be implemented, was effective
in reducing emissions for all three fuels.
The measured data for tests at baseline conditions and at reduced
NOx emissions levels are summarized in Table 1-2. Baseline conditions
were defined as normal boiler settings for a load of 80% of the nameplate
capacity. The low NOx condition was defined as being when minimum NOx emis-
sion was measured for that particular test series. The abbreviations
used in the table are defined in the Glossary of Terms (Section 10.0).
-------
s
s
I *:§
S Cl X O
£ 3
£.5 3
^ ~^
X O >_ J
s g 5 ;.;
5 S^-
3 d t
SIS?
V- Q.
o_
i-e
t_i ir\
^ u
* <-
c?1
-J x -- ^- IU
~* " o f* 2 ^
•^ * >- u o ~
.5°
I
I
I
I
I
I
-------
I
I
I
I
ci 9
r:
o
H
nj
n
ncen
o
u
to
•S
u
r:
0
"-'?
*
*-i
0
t V
Vj : r
fi £ «-
5 1
•p-
o I
U jj^
on?
OQ -
d
"S H ?
S S U
O —
K
Ul C ^
S ".-,'
U 0
"2s£
» o, «;
§ I
0 £
o r
O V." ^
U Q —
S "
£ £-
w cT
*« if
H 3 ^
So
?
1 1
*J rH
W ^
H fct
>, E
M *-* 2
41 -H
r-4 U g
3f I
tn o
H '*'
Vj M TJ TJ M
• H H 15 D H
< < s 3 i:
\f 5 3
3O -H -H O
» 1 1C W r5
5
"^ 1 f I 1
1 1 1 1 1
CO 1 1 rH V£>
CM ^T rH
g
t/1) ""31 O Z 00
•T i n o
CN O 3
CN Q
A i-3
CO CO r-1 0 O
"*' M S ""* ™
W
£ R S S S 5
n (N rH CO
ri -J- CM in
CO O rH in rH
,-H r-
CD rH rH "^ rH
co in r^ in r-l
§ 1 1 1 g,
4J 4J 4J 4J -H
to to 10 i/j oi
^ rH H M
O O O O
(N ID vO >J3 O
o o o m o>
^ CO O^ \D CN
rH
H rH
T3 -0 T3
Q Q -rH Q -rH
a 3 a a a
1 1 1 1 t
1 1 t 1 1
1 1 I t t-
X
o
'2. o o o o vr>
O fN
SO rH
3(N
A
Q CO CO O CN "^
^ vD O~i vf CO r^
(Jl
1* co <•*"> co cn m
o co
rH rH CN
O
rH
o^ o> o CN o CN
•H rH rH r-1
rH O *^T VP (N f~^
iH rH (N CN CN r~\
O CN O O CO O
rn in CN o en r-.
CO CO O CO U5 O
v ^r ^r ^r vr rH
tri tr* D> TJ
w cx & 5 5 m
rH rH .H
(0 (0 n]
O O O O O O
Ch O O O O IT|
""*
rH rH rH rH rH I-H
Tt
•^ Vj M LO t3 nJ
H H H O (U Q
S S 3 « 5 »
T Q O *• H Q
O T CO
* ° 1 ** 1 1
n CN ' o ' '
i i I t I I
O CO | | (N |
CN rH fN
O O O O O O
(NO rH fl
i^j i^) m o I""" CO
H ^ rH °^ °^ °^
o r^ fN I-H vo in
f*~i co CN o vin in
n ro CN CM rH rn
CO C' rH in r^i O
1/1 LT, ID o r* ^r
rH
CM rH CG co -J- m
o c-< u*i f^ CM o
00 rH O O O O
r-. co in ro rj
«H CT> o ^J1 o i
(o o in in rH
'H rH
MM ^
M M CN
in O O O rH O
(N CT* M in
O O rH rH O^ ID co n
0 n 0 0 OJ U>
•31 U3 CO f*l CO f*l
G G E
a CM 4-) 4J 4-> H
cn co w to w te
M M
rH M O O Pi
O O ^D >U in o
o in in o O O
in f~- o u? M u)
rH M M M
Cf» O M CN ro *r
M CN fN fN IN CN
0 «0
td M V* M T3 "; nj
(5 rH -i-l r\ H3 Q O
3 < rf; < Q O »J
rH
S *3 S i-5 a: PQ >5
i i i i i i I
1 1 1 I 1 t 0
o m j f -r i o
0)
O | O O O O p O
M M M «-H Q M
O O M O O C^ fN
H m T f) M rH r*
O m cr^ (T> CN O — < O vD l~~
BCi rH *~^ in o fi o^
C?l O fl CO O CM rH
rj O rH
M *T v£> CN fi in
, >ii *r n o i o n
co m 01 in ^D M
rH M M MM
r- <£> -* n ^
M M t
O O O I 1 O O O
rH M rH M
coconr-r-c^oo
oncNCOinrHco ^r
'<-* 1-1 i-t
ui-HOrMOCnoQ
3_ J5_ " S § S 5 M
^ > « « § CT, > ,
a s. s S1 s a s &
,-,
rH t-( r-i O rH ^
(0 nj DpOO
zouu*^uo
COLDOOOOOM
Mwr-tfimmmin
SSrsSSSS"
-------
des Concentrations
Low Nitrogen Oxi
onccntrations
Baseline C
» £
»$
stii
s&>
S I
M (1.
y I
8 1
80*
"s a
SC £
m
o ^
X
w
""' I
1 i
S I
w a
0 £
8 1
QV< *
O Q ~
4)
*•> C n
X -« C
O rH R.
JS *J ^
K """
8 j?
S r,*
U 0
» «£
sats
o*
Burner
Typ«~ 1
S-8
J5S
if!
*h
2s
H
Tl
^ -H 0 -ii
r< < 3 <
S S S 1
lit 1
III 1
VO rH O
fN rH O
vr
8 o
in 20
0 3 m
rH CM rH
^f u>
U> CTi CO CO i^*
M M ^H rM
r*- in ^ VD u">
r» o>
r- VD M in co
tn
M M V4 O, C
««-4 -H -H 3 -H
rf < *C U «
rH
•H -H -H O
0 0 O
(M
M U> in to C5
«: * * ^ 2
O r-- rH C^ O
ro ^j- in vo ^-
•0
i s a s
:, » < S
S 3 3 3
iiii
i i i i
«t m o -o
m i ,H o n,
w S
§ o « «
r^ o ^ in *°
(U
£
in m f- O
r- co ID
•H O^
.H 1 1 1 CO
O in
1 t t i i
00 | O O i
^7 *n o
C4
I o o n o
lf^ fJD t~* i-t
•H M rH »~i
H
r- r* rH rH M
o co r- o\ r^
^ CM
-D n m co -^
*T rH
*r o ȣ> r- co
O> D> &> CTi
c c c q w
•H H -H -H ^4
pi « « pi »
o o o o c
» 2 S5 SI U
o H r- o o
00 (Jt O j— ( O O O
I ^* CO CO i
co o m
t*- co r-
' o i «j jq
° o o o o
CO — • ^->
rH
H i-4 H
O r- rH rH J—
^J r- VD oo tn
<*1 rH rH rH
CO r^ u> n T
rH
vO ro Oi
•0 tJ
SO -r* H
^ «: <
^ ? * s
s a as
it ii
ii ii
in u-> TJ i i
n fN '
n n
O O 3 m CN
rH rH W O>
(tf T
V
£
2 rH O
M
ro r^ ro \D
i/l I\D VD CO
v£) m i£» vD
CO fn (N r-
O^
t i i n t
o
• ' ' S '
o m i i m
oj r«
in
o o o r. o
rH
u> o in m r-*
m r*» r- vo c»
CM
r- IN VD m m
(M in r~*
r- IN vo ^ ^r
g
t" D» 3
C c M a> M
-H -H -H 4J -H
Pi Pi < W <
^( -H -H
O O O
c e> to CN oj
2 » * * «=
co o r* co co
CO Ol ^H CM ro
'a- 2000 240
•H 2: 2 2
o m
CO i/> m tn in
t)
H O Q
< 0 4
» tn >
S S S
2 0 90 -
t Meas ured
t Measured
0 0 50 476 .0590
t Measured
4 9 48 -
fl O O rH O VD
-1 2 S *-» 2
rH VD 0
•^T rH fNl
rH in rH
r- tn •*?
CM in >JD
in o r-
T rH CO
rH i i m I o
o o o
(N . , e> |
m ' * m <*
•-H : m o>
S S ' S ' "•
o o oo m o o
- g >o n . ..
r-j
A
O O CO r*l o (~-
rH rH rH rH
r*) rH ro CTi r~ r-)
CTv O O rH M 0>
H -^ 4J -P -H 4-1
< o; tn w < ui
-H -r4 H -H H
o o o o o
EN O fN m CM M
«: S: * * *= *=
o o eft o"i r*- co
rH
vo ^ u>
-------
I
I
(N V?
' £:
g
CJ J
-H tn
,Q
*J C
Q
w
5
j
^
j
u
5
u
W
o
*o
o
c
CJ
Oi
O
*J
S
c:
0
•H
2
^
o
o
c:
0
r:
-H
u
o
o
£ £'
ii|
s £
w _c\
L) ^
x j:
„
8 I
s'l'i
ft)
O «H Cij
o —
Fl >,
V, U
V 0 ~
o ;.-
ui o
<"*
5 S.
S I
U> ^Dj
CJ V
CJ [}
*-* el
"
O VI «f
t>
1 2 1-
s&_
X M""*
w o
S g §
H - 1 i
y So
w 3
^s
St *
-a|
*> .
W O
^ rt < <
3 S ? >
5 S S S
III 1
111 1
1 1 1 ^ (
O
r- o o 5 ,
(U
o n cji ?^ o
n O (N 4J rH
rH «H *H 0 ""'
O CQ m (N
CO r^ t^- ir>
oj rxj rn ^r
\£> ^j
^T t~~
O I ' I O
rH ^f in CN
•~t co i r-
rH "31 -«J
1 1 1 1 1
o o o co i
CO
co CN co ^r -c
m o *^ co r-
tN O> O 0' t^
CP> in co co v.*
in T^ in n o
ro v O en o
r*i r-4 ft LI o
rH M VO UJ -H
X X *^
003
= x s ° •
g s s
•3 iJ SI vD o
O O 4J V£J Q
*a "« s
to vi
3 * rH -^
0 CJ
c c
H -H
^ rH
O CJ
U W n ^j,
(3 CQ i-H ,^
•-*
c^ o
CO v-
i ^ . i O
1 «? t 1 0
1 1 1 1 1
i i i
0 O f-H O o
« M ^ » »
U O I~l (U N
>, >, -H 4-> O
U U < W 2;
rH r-H
O O Di
u u
< X pj in H- 0
0 O =*t *fc 2;
m fi n o in
—t rH rH ^H f\J
m in rH 01
•H CN n xj> m
r^ r^ t- r- r-
-o
«
o
^
o
I
i
1
X X
0 0
o *"
.1 ?
^ rH )3
<«' 8 S
rH ^1
rt (0
U) tO
£ * S
C> CJ
c t:
H -H
CJ O
cn w w
v' CQ CM
m
•H O
rH
( rsj t
CO
•-H
in co i
0 O O
r-l oj CD
o •c- rr
O4 CO pi
m co rH
^ m co
o o ""i
CO CO rH
CJ
r-H
N > tP
N rH C
O p -H
rH
O O O
K U K
m m o
(N (N rH
r- co o
r- r- co
•o
§
c:
O
H
(3
V*
CJ
D<
O
X
o
!
c
o
*J
^
o
0,
o
-^
—i
0
w
<3
flL
—•
fM
o
'C
Oj
i
d?
fN
O
^,
1
X
0
M
o
p
[14
f*i co o> ^r n
^o rj ^r *n in
t.i tn ^ o in
en rH in fi o
n rH in in o m
-------
The balance of this section summarizes the emissions results and
briefly discusses the trends and conclusions that can be drawn at the end
of the first phase of the program. Section 4.0 discusses the results
in detail; and Sections 5.0/ 6.0 and 7.0 consider the effects of opera-
tional parameters, fuel properties, and boiler and burner design character-
istics on NOx formation. Section 8.0 discusses the statistical analysis.
1.2.1 Nitrogen Oxides Emissions
The pollutants of primary interest in the first phase of the
program were the oxides of nitrogen [principally nitric oxide, (NO), and
nitrogen dioxide (N02), together called "NOx"]. The NOx measurements
at base load over the capacity range of industrial boilers are summarized
in Figure 1-1. The following paragraphs discuss the NOx emissions from
combustion of coal, oil, and gas fuels.
Coal Fuel -
Base load NOx emissions for coal fired boilers varied from 224
to 800 ppm for a variety of underfed and spreader stokers and pulverized
coal burners^ including one cyclone burner boiler. Only two firetube
units were tested, since there are so few coal fired firetube boilers in
industrial use.
The operating excess oxygen level was found to affect the NOx
emissions level for watertube boilers. The NOx emissions decreased with
decreasing excess ©2 for each coal fuel test conducted and averaged
approximately 50 ppm for each one percent change in 02 (Section 5.1).
The larger the furnace heat release volume, defined as the furnace volume
(ft ) divided by the firing rate (Btu/hr) , for coal-fired watertube boilers,,
the lower were the NOx emissions (Section 7.2). However, other parameters
may be influencing this relationship also, such as the design of the coal
burning equipment.
Boilers with underfed stokers require large furnace volumes for
burnout of the large coal particles and have the lowest NOx emissions.
-------
I
I
I
I
CO
§
o
CO
£
LJ
O
OO
UJ
ca
x
o
' (XON) ' UOT^^USOUOO
go sspfxo
-------
The spreader stokers and pulverized coal burners utilize smaller furnace
volume because of smaller coal particles and have overfire air and/or
steam injection for added turbulence within the furnace. These boilers
have higher NOx emissions than the underfed stoker-fired boilers. The cyclone
coal combustor utilizes the smallest furnace volume because of the high air
injection velocities and turbulent mixing, and it produces the highest NOx
emissions of all types of boilers tested during this program.
The coal fuel data discussed in Section 7.1 show a strong dependence
of NOx emissions en burner heat release rate; although coal burners
cannot be defined completely simply by the number of coal injectors
because grate design is important too. However, pulverized coal burners
and cyclone furnace coal combustors are similar to oil and natural gas
burners in that the fuel and air enter the furnace through a single or
multiple burner port(s). Burners in the 8 to 30 MBtu/hr size range had
NOx emissions between 200 and 400 ppm. Burners in the 30 to 100 MBtu/hr
range had NOx emissions between 370 and 600 ppm. The highest NOx emissions
were 800 ppm with a 255 MBtu/hr cyclone coal combustor.
Combustion air temperature and coal properties did not strongly
affect the NOx emissions for coal-fired boilers.
Oil Fuel -
Base load NOx emissions for oil-fired boilers varied from about 100 to
200 ppm with No. 2 oil, and 150 to 619 ppm with No. 5 and No. 6 oils. The
most important parameter influencing NOx emissions from oil-fired boilers
was found to be fuel nitrogen content. The base load KOx emissions varied
from approximately 105 ppm for fuel oils with less than 0.01% nitrogen by
weight, to approximately 400 ppm for fuel oils with 0.5% nitrogen. This
corresponds to an average conversion rate of the fuel nitrogen to gaseous
nitrogen oxide emissions in the flue gas of 46% (Section 6.1} for normal
operation. The percent conversion can be reduced to 20 to 30% using
"off-stoichiometric" or "staged" combustion.
10
-------
I
I
I
I
Design and operating parameters which were found to affect the
NOx emissions of oil-fired boilers are excess oxygen, burner heat release
rate, furnace heat release volume, and fuel oil temperature at the burner.
These are discussed in Sections 5 and 7. The oil-fired firetube boilers
tested during this program with No. 2 and No. 5 oils showed little depend-
ence of NOx emissions on operating excess 0 level and boiler load. All
firetube boilers used ambient temperature combustion air. The oil-fired
watertube boilers with and without preheated combustion air showed decreas-
ing NOx emissions with decreasing excess O level. The No. 2 oil tests
were not as sensitive to excess 0 level as were the No. 5 and No. 6 oil
tests. The effect of burner heat release rate and furnace heat release
volume on NOx emissins was found to be the same for No. 2, No. 5, and No.
6 oils. Oil burners with larger heat release rates tended to produce more
NOx emissions than burners with smaller heat release rates. The furnaces
with larger heat release volumes had lower NOx emissions than the furnaces
with samller heat release volumes. A test series was conducted with No. 6
oil which showed that lowering the fuel oil temperature at the burner from
its design level increased the NOx emissions.
Design and operating parameters which had little or no effect on
the NOx emissions for oil-fired boilers are combustion air temperature and
method of fuel oil atomization. The No. 2 oil tests were evenly divided
between steam and air-atomized oil guns. A single, high-pressure atomized
burner (Test #54), where atomization is achieved by pressurizing the oil
to at least 100 psig, was tested because sales data indicated that high-
pressure burners are increasing in popularity. The NOx emissions for
these tests did not appear to depend on which atomization scheme was used,
as long as the oil was properly atomized for good combustion. Varying the
differential atomization and fuel pressure during some of the tests did
show some potential for low NOx air and steam atomizer designs. The larger
oil burners tested (greater than 25 MBtu/hr) were all steam atomized.
Natural Gas Fuel -
Baseload NOx emissions for natural-gas-fired boilers varied from
50 to 375 ppm. The combustion air temperature was found to strongly affect
the NOx emissions (Section 5.2). The baseload NOx emissions for boilers
11
-------
using ambient temperature combustion air varied from 55 to 116 ppm. The
watertube boiler NOx emissions were from 70 to 116 ppm, and the firetube
boilers were from 55 to 107 ppm. The baseload NOx emissions for natural
gas fuel watertube boilers with air preheaters (firetube boilers do not
have air preheaters) varied from 90 to 374 ppm. Burner heat release rate
and operating excess O level also affected NOx emissions for natural gas-
fired boilers, and these are discussed in Sections 7.1 and 5.1, respectively.
The magnitudes of these effects depend on combustion air temperature. For
ambient temperature combustion air tests, both the burner heat release rate
and the operating excess O level had only a minor effect on NOx emissions;
*£
however, with preheated combustion air they greatly affected the NOx emissions.
Furnace heat release volume did not affect the HOx emissions for
natural-gas-fired ooilers.
NOx Reduction
The base load NOx emissions were successfully reduced for each fuel
tested during this program. For coal fuel, one of the most successful tests
utilized oil fuel burner ports as air injection ports above a spreader stoker
traveling grate. Reducing the air up through the grates and diverting it to
these burner ports resulted in lower NOx emissions, but higher grate tempera-
tures, such that a compromise between the amount of NOx reduction and an
acceptable operating grate temperature was required. The resulting NOx
emissions were reduced by about 25% over the boiler's operating load range
(Section 5.1.2).
Multiburner oil-fired boilers were successfully operated off-
stoichiometrically by terminating fuel flow to individual burners and using
the burner port as: an air injection port. Using the upper level burners in
a two-level burner bank as air injection ports was most successful in reduc-
ing NOx emissions as discussed in Section 5.1.2. Test 63 resulted in a 17%
reduction due to removing the center top burner out of two rows of three.
Test No. 6 was a boiler with a single row of four burners, for which the NOx
emissions were reduced by 49% by removing a center burner. Test No. 9 was
a boiler with three burners arranged in a triangular pattern. The burner at
the apex was removed from service and the NOx emissions were reduced by 29%.
Some of the change in NOx emissions possibly was due to a change in the test
loads that are listed in Table 5-1.
12
-------
I
I
I
Multiburner natural gas-fired boilers were also successfully operated
off-stoichiometrically by terminating gas flow to a burner and using the
burner port as an air injection port. Test 15 resulted in about 12% NOx
reduction in a boiler with four burners arranged in a square by removing one
of the upper corner burners from service. Test 30 resulted in a 40% reduction
in NOx emissions with a boiler having a similar burner pattern but of larger
capacity. A significant test (77) was conducted with a natural gas fuel
corner-fired boiler by adjusting the air distribution to the different
burner elevations. A 24% reduction in NOx emissions was obtained without
removing any burners from service (Section 5.1.3). In this case, the fuel/
air mixture ratio at the burners was adjusted using air registers.
The low NOx emissions tests conducted during this program demon-
strated the potential for reducing the NOx emissions levels from industrial-
size boilers through off-stoichiometric operation. Greater NOx emission
reductions could have been achieved if more time could have been spent
optimizing the tradeoffs between emissions and operation. Additional data
from this program demonstrate that boilers can be designed for low-NOx
operation, and that different parameters control the amount of NOx formation
for the different fuels evaluated.
1.2.2 Particulate Emissions
Particulate emissions were measured using the EPA sampling train
described in Section 3, and the total particulate emissions (solids plus
condensibles) are listed in Table 1-2 for the baseload and low-NOx
test conditions. It should be noted that EPA stationary source regulations
are based on only the solid or filterable portion of the total particulate
emissions. Table 4-1 lists the total particulates and also the solid or
filterable portion.
The particulate concentrations reported in this Phase I Final Report
were calculated using the revision published on page 32855 of volume 39,
number 117 of the Federal Register, September 11, 1974: "The revision incor-
porates a simplified technique for converting pollutant concentration to
mass emission rate. The technique, which is based on estimating factors
13
-------
and mass balance principles, eliminates the need to measure flue gas flow
rates and also fuel flow rates except when combinations of fuels are simul-
taneously fired."
Figure 1-2 shows the particulate emissions as a function of test
load for coal, oil and natural gas fired boilers. For natural gas fired
boilers, the total particulate emissions were typically 0.004 to 0.007
Ibs/MBtu of fuel input with a few tests above and one below this range.
The total particulates for the oil-fired boilers were typically 0.02 to
0.12 Ibs/MBtu of fuel input. The tests with No. 2 oils (1, 52, 54, 59,
65, and 66) ranged from 0.02 to 0.04; No. 5 oils (Tests 3, 33, 35, 44, 45,
46, 63, and 70) remged from 0.04 to 0.12; and No. 6 oils ranged from 0.045
to 0.11 Ibs/MBtu of fuel input, except for Tests 29 and 34, which were
0.35 and 0.50 Ibs/MBtu of fuel input, respectively. For coal-fired boilers,
the total particulate emissions were typically 0.5 to 3.0 Ibs/MBtu of fuel
input. Test No. ill was conducted with a boiler which often burns tree bark
in addition to oil and coal fuel and had exceptionally high particulate
emissions of 10.1 Ibs/MBtu of fuel input. Most of the coal-fired boiler
particulate data presented in Figure 1-2 was measured after the flue gas
had gone through a dust collector; although Test No. 32 was before a dust
collector, and the particulate concentration was not unusually high, only
1.2 Ibs/MBtu. This boiler used two cyclone coal combustors, which by high
combustion temperatures convert most of the potential particulate to
molten slag ; the slag is removed at the furnace hopper.
14
-------
I
I
I
10.0
0 Particulate Concentration, Ibs/MBtu
b ? o i-
oo- '
t-1 M (-• O
/ft
A ^
/<^
A
^6
05
A
@5 ^
_/ — ^P r / — \ r
^^••k J CB)
V^ r T
^5^4-' (20
f\
A8\
&2±
/\
£&
5 _ >-s5 (22)6
2
jSf
1^7! J— J
pM ^'5 i§9
5^1— 1 t|J4
on
&
^
[24j ._,
05J
(s]
D Natural Can — -
Qoii
/\Coal
A
The numerals withi
-t-od-1- nnmHovc ciiip
fuel grades.
[77]
r^j
C_)
A
i the symbols are
OT-cmvi n-t-c: a-ro oil
100
Test Load,
200
k Ibs/hr of Steam
300
400
Figure 1-2.
Total Particulate Emissions at Baseload for Natural Gas, Oil and
Coal Fired Boilers. All Measurements Were Made Downstream of a
Dust Collector Except for Coal Test No. 32, For Which Measurements
Were Made Upstream of a Dust Collector.
6000-28
15
-------
Particulate emissions as measured during this program were found
to depend on the fuel being consumed. Natural gas-fired boilers had very
little particulate emissions. No. 2 oil-fired boilers were much lower
particulate emitters than were No. 5 and 6 oil-fired boilers. Coal-fired
boilers had the greatest particulate emissions, as would be expected based
on the ash content of the coal.
1.2.3 Sulfur Oxides Emissions
Total sulfur oxides (SOx) emissions were found to depend almost
entirely on the sulfur content of the fuel and were not affected by differ-
ent size boilers, burner designs or fuel being fired (Section 6.4). Natural
gas, oil and coal fuels all produced SOx emissions directly related to the
sulfur content of the fuel. Figure 1-3 presents these data for oil and
coal fuels. The total sulfur oxides concentrations from natural gas were
so small, e.g., 0.017 g/MCal, that the measurement was discontinued after
a few tests. The small amount of sulfur that was present was from an odorous
mercaptan that was added in minute quantities by the gas distribution com-
pany so a gas leak could be detected by the odor of escaping gas. The oil
data agree quite well with the calculated values for a typical oil composi-
tion. The coal data show much larger deviation from the; calculated values,
which may be due to the difficulty in obtaining a representative coal sample.
The amount of sulfur trioxide (SO ) in the flue gas appeared to vary
inversely with the amount of total SOx emissions as discussed in Section 6.4.
However, this is not consistent with the knoch chemical kinetics for sulfur
oxide formation. It is felt that the large concentration of SO when the
concentration of SO was small (as shown in Figure 6-5) was due to a syste-
matic experimental error in the measurement method caused by over-filtration,
and that the error increases as the SOx concentration decreases. Therefore,
it appears that 1 to 3% of the SOx is SO .
16
-------
I
I
I
2400
g 2000
a
q
o
•H
-P
03
0)
o
o
w
0)
•o
•H
X
o
Iq
3
-P
O'
1600
1200
800
400
A
Typical Type of Coal
100% Conversion
'
'
'
Typical Type of Oil
100% Conversion
Numerals within Symbols
are Test Numbers.
1—
FUEL TYPE
/\ Coal
O oii
I
0.0
Figure 1-3".
1.0 2.0 3.0 4.0
Fuel Sulfur Content, Dry, %
5.0
Total Sulfur Oxides Emissions at Baseload For Oil and
Coal Fired Boilers
17
6000-28
-------
I
I
I
SECTION 2.0
TEST BOILER SELECTION
The initial task was to select representative industrial boilers
according to boiler and burner design, and present and future fuel to provide
a cross section of the industrial boiler population.
During Phase I of the contract, approximately 50 tests were planned
for specific combinations of fuels and boilers. Additional tests were made
as the opportunities presented themselves. These additional tests typically
were made on another fuel or boiler at a site selected for the basic fifty
boiler tests. No tests were scheduled originally with the No. 2 type of
distillate fuel oil fired in Southern California, but opportunities to test
distillate oil as an alternative fuel were accepted when no significant
delay in the basic program was involved.
Table 2-1 shows the selected distribution of the test boilers among
furnace types, capacity, fuel type and burner type. The numbers in the boxes
are the number of individual boilers to be measured in that category and
burner type. The numbers in parentheses in the boxes are the test series
numbers assigned to that particular category and burner type.
This distribution is a composite of several criteria, such as boiler
population, boiler emissions, burner population, the new United States
energy policies, and present and predicted sales. The data were obtained
primarily from pertinent literature and conversations with the American
Boiler Manufacturers Association. Some of the more useful literature is
listed as References 2 through 9. The boiler selection task is discussed
in more detail and background data from other sources are included in the
Appendix.
Tests No. 11 and 50 eventually were deleted. No. 11 was deleted to
provide time for special tests No. 52, 53 and 54 at Location 19 to investi-
gate the effect on nitrogen oxides emissions of the oil atomization method,
i.e., steam atomization, air atomization and mechanical atomization. Test
19
-------
Tabla 2-'I. DISTRIBUTION OF TEST BOILERS AND TEST NUMBER ASSIGNMENT*
Category
1
2
3
4
5
6
Type
WT
WT
WT
WT
FT
FT
Capacity
k»/hr
10-30
30-100
100-250
250-500
10-16
16- 30
OIL FUEL
Distillate
b Learn
6
5') 62, 63
6,68,74
2
65,70)
8
Air
1
(33)
2
(44,59)
3
Res idual
SttMm
1
(1)
6
(6-11)
3
(21-23)
1
(29)
1
(45)
12
Air
J
(2)
2
(34-35)
1
(46)
4
Hotary
1
(3)
j
(J6)
j
29
GAS FUEL
Ring
Fired
2
(4,5)
7
(12-15
60,67,
f.9)
3
(24-25,
75)
2
130,77)
6
(37-41,
38)
2
(47,48)
22
Other
1
(40)
1
23
COAL FUEL TOTAL
Grate
2
(15,17)
2
(4;'-43)
(50)
5
Spreader
3
as, 19, 20)
2
(27-28)
s
OF UNITS
Putveir- TESTED
5
1
25
1 11
(26)
i
i
3 6
(31,32,
1
12
8
4
14 66
*Number of tests in each category/fuel combination are indicated.
parentheses are the test series numbers.
Numbers in
6000-28
20
-------
I
I
I
No. 50 was deleted because a firetube boiler of capacity greater than
16,000 pounds/hr of steam that burned coal could not be found, and the
time rfas spent at Location No. 12 doing additional testing of corner-
fired boilers (Test No's. 75, 77 and 78).
The initial step was to distribute the 50 test sets according to the
population of boilers of a given category. The results are shown in Table
2-2, Column A. Another distribution was made on the basis of total capacity,
which is a measure of total emissions of the boilers of a given capacity
range (Column B). The total capacity of each size category is important,
because a few large capacity boilers could have more impact on air quality
than many small capacity boilers.
A compromise between the importance of number of units and capacity/
emissions in each category was made by weighting these factors equally.
For example, in Category 3, weighting of the three units that would be
selected by population and the thirteen by capacity resulted in eight
tests planned. The distribution by fuel was with respect to population only.
The distribution shown in Table 2-2 was further modified by consid-
erations of boiler and fuel trends and geographic distribution. Cast iron
boilers were eliminated because they typically are well below the minimum
capacity of 10,000 Ibs/hr. The number of tests with coal fuel was increased
to reflect the recent interest in coal. Discussions were held with the EPA
Project Officer, the American Boiler Manufacturers Association, the American
Petroleum Institute, boiler and burner manufacturers, other EPA contractors,
etc.; and the distribution of the 50 test sets shown in Table 2-3 evolved.
Specific boilers were then sought as candidates for testing. About
one third of the candidate boilers came from owners who were contacted by
KVB or the EPA Project Officer and offered their boilers. Another third
came from referrals from the American Boiler Manufacturers Association.
The remainder were obtained by securing names of owners from trade journals,
21
-------
Table 2-2. NUMBER OF TEST SETS BASED ON POPULATION AND TOTAL CAPACITY
Cate-
gory
1
2
3
4
5
6
7
TOTAL
Furnace
Design
WT
WT
WT
WT
FT
FT
CI
Capacity
MBtu/hr
10-16
16-100
100-250
250-5CO
10-16
16-10C
1-10
.
No. of Test
Sets by
Population
(A)
5
16
3
2
17
5
2
50
No. of Test
Sets by To ;al
Capacity
(B)
3
16
13
5
6
5
2
50
Selected
No. of
Test Sets
(O
4
16
8
4
11
5
2
50
Fuel Type
Oil
1
6
3
1
4
2
1
Oil &
Gas
1
2
1
0
1
1
0
18 1 6
Gas
2
6
2
1
5
2
1
19
Coal
0
2
2
2
1
0
0
7
1 i :
Table 2-3. DISTRIBUTION OF FIFTY TEST BOILERS BY CAPACITY AND FUEL
Category
1
A
3
4
s
6
Furnace
Type
WT
WT
WT
WT
FT
FT
Capacity
M/hr
10-16
16-100
100-250
250-500
10-16
16-100
Oil Fuel
Pistillate
K«ch.
1
(33)
1
Air
1
(44)
1
Residual
Steam
1
(1)
4
(6-9)
3
21-23)
1
(29)
1
(45)
10
Air
1
(2)
2
(10,11)
2
(34,35)
1
(46)
6
Rot
1
(3)
1
(36)
2
20
Gas Fuel
Ring
Fired
1
(4)
3
(12-14)
2
(24-26)
1
(30)
5
[37-41)
2
(47,48)
14
Center
Fired
1
(S)
2
(15-16)
1
(49)
4
IB
Coal Fuel
Grate
2
(17-18)
1
(26)
2
(41-43)
1
(50)
6
Spreader
2
(19-20)
2
(27-78)
4
Pulver-
ized
2
(31-32)
2
12
Total
Number
of Units
Tested
5
15
4
12
6
50
22
6000-28
-------
I
I
reports, word-of-mouth, etc., and soliciting their participation. About
65 individual boilers were offered as candidates, and from these 47 were
selected for testing. In 28 instances, a different fuel and/or burner
type was tested in a given boiler, making a total of 75 test series during
Phase I of the contract.
Eight tests were run on boilers in Southern California that burned
distillate fuel and employed steam atomization. However, the distribution
tabulated in Table 2-3 lists no steam-atomized distillate oil fuel burners.
Steam-atomized distillate burners are found in Southern California, but are
rate in the rest of the country. These tests often were run while awaiting
the resumption of natural gas burning by the boiler being tested. The fall
and winter of 1973-74 was a time of great uncertainty in fuel availability,
and often a fuel switch had to be accepted by the field test crew because
the boiler owner deemed it wise to switch fuels at the time.
The route followed during the testing and the test location numbers
are shown in Figure 2-1.
An attempt was made to include a representative cross section of
brands of boilers and burners. Less flexibility was available with brands,
because 50 boilers did not allow enough freedom of choice for a strict
distribution by brand, as well as by size, furnace type, fuel type and
burner type. The major manufacturers of industrial boilers and/or burners
in the United States are listed in the Appendix. A description of each
boiler tested is presented in Section?, Table 7-1.
23
-------
Figure 2-1. Field measurement route and test site locations.
6000-28
24
-------
I
I
I
SECTION 3.0
INSTRUMENTATION AND TEST PROCEDURES
The emission measurements are made with instrumentation contained
in an 8 by 30 ft Laboratory Trailer. A plan view of this trailer is shown
in Figure 3-1, and exterior and interior views are shown in Figure 3-3. The
gaseous species measurements, except sulfur oxides, are made with analyzers
located in the trailer, while the particulate, smoke spot and sulfur oxides
measurements are made with analyzers taken to the sample port, and the weigh-
ing and titration are done in or near the trailer.
The emission measurement instrumentation used on the project is the
following:
Table 3-1. EMISSION MEASUREMENT INSTRUMENTATION
Species
Manufacturer
Measurement
Method
Model No.
Hydrocarbon
Carbon Monoxide
Oxygen
Carbon Dioxide
Nitrogen Oxides
Particulates
Sulfur Oxides
Smoke Spot
Particulate Sizing
Beckman Instruments
Beckman Instruments
Teledyne
Beckman Instruments
Thermo Electron Co.
Joy Manufacturing Co.
KVB Equipment Co.
Research Appliance Corp.
Millipore Corp.
Flame lonization
IR Spectrometer
Polarographic
IR Spectrometer
Chemiluminescent
EPA Train
Titration
Reflectance
Visual Counting
402
865
32 5A
864
10A
EPA
G2R-100
XX 50
6000-28
3.1
GAS SAMPLING AND CONDITIONING SYSTEM
A flow schematic of the flue gas sampling and analyzing system is
shown in Figure 3-2. The sampling system uses three pumps to continuously
draw flue gas from the boiler into the laboratory. A high capacity (15
CFM) Nash pump is used to draw a high volume of flue gas into the unheated
portion of the system to provide adequate system response. The Nash pump
25
-------
!*
<
K
ID
<
S
o
§
4
°s
Q
G
*|5
s
Ul
01
>-
K
a
o
::r-q
n
^ !
I
S
^
5
<
S
K
Si
•> •<
i|
W
! ^
•H
I
TS
C
al
P.!
O
-P
ai
O
X!
nj
F-!
d)
3
•H
fr.
-------
I
I
I
CO
CN
I
O
O
s
-M
W
tn
c
•H
N
q
•H
CO
01
(0
o
-------
pulls from a manifold connected to 24 unheated sample lines. Selector
valves allow composites of up to 12 points to be sampled at one time. The
probes are connected to the sample manifold with 3/8" nylon line. Stainless
steel quick-disconnect couplings are provided to facilitate the connection
between the sample lines and the instrumentation laboratory. The sample
from each line the:.i passes into individual water traps. The water traps
consist of glass bubblers used to collect water condensed from the sample.
Drain valves for emptying the traps are provided. A positive displacement
diaphragm sample pump draws unheated sample gas from the high volume line
through a refrigerated condenser(to reduce the dew point to 35°F), a rota-
meter with flow control valve, the sample pump, a 1 micron filter, and to
the O , NO, CO and CO instrumentation. Flow to the individual analyzers
is measured and controlled with rotameters and flow control valves. Excess
sample is vented to the atmosphere.
To obtain a representative sample for the analysis of NO and
hydrocarbons, the sample must be kept above its dew point, since heavy
hydrocarbons may be condensible and NO is quite soluble in water. For
this reason, a separate, electrically-heated, Dekoran, sample line is
used to bring the sample into the laboratory for analysis. The Dekoran
line is 3/8 inch Teflon line, electrically traced and thermally insulated.
Metal bellows pumps provide sample to both the hydrocarbon and NOx analyzers.
3.2 INSTBDMENTATION
The laboratory trailer is equipped with analytical instruments to
continuously measure concentrations of NO, NO , CO, CO , 0 and hydrocarbons.
Figure 3-3 presents pictures of the exterior of the laboratory and of the
control panel. The sample gas is delivered to the analyzers at the proper
condition and flew rate through the sampling and conditioning system des-
cribed in the previous sections. This section describes the analytical
instrumentation.
28
-------
I
1
I
t Figure 3-3. Exterior and Interior View of
Mobile Air Pollution Reduction
Laboratory.
6000-28
29
-------
3.2.1 Total Nitrogen Oxides (NOx)
The oxides of nitrogen monitoring instrument is a Thermo-
Electron brand chemiluminescent nitric oxide analyzer. The operational
basis of the instrument is the chemiluminescent reaction of NO and O
to form NO in an excited state. Light emission results when excited NO
^ 2.
molecules revert to their ground state. The resulting chemiluminescence
is monitored through an optical filter by a high sensitivity photomulti-
plier tube, the output of which is electronically processed so it is
linearly proportional to the NO concentration.
Air for the ozonator is drawn from ambient through an air dryer
and a 10 micron filter element. Flow control for the instrument is accom-
plished by means of a small bellows pump mounted on the vent of the instru-
ment downstream of a separator which insures that no water collects in the
pump.
The analyzer is sensitive only to NO molecules. To measure
NOx (i.e., NO+NO ), the NO is first converted to NO. The conversion occurs
£
-------
I
Linearity j^ 1% of full scale
Vacuum detector operation
Range: 2.5, 10, 25, 100, 250, 1000, 2500, 10,000 ppm full scale
Both the total nitrogen oxides (NOx) and nitric oxide (NO) concentra-
tions are measured using a sample line heated to about 120°C (250°F)
and called the "Hot Line" to conduct the gas sample to the analyzer in the
trailer. In addition, the nitric oxide concentration is measured sequentially
using an unheated sample line called the "Cold Line" connected to the same
analyzer in the trailer. Here, the water first is removed from the sample
gas by a drop-out bottle and a refrigerator. Both hot and cold line measure-
ments are listed in the Summary Table 4-2 in Section 4.
When the heated sample line was inoperative, the NOx concentration
listed in the summary table was calculated from the cold line NO concen-
tration measurement by multiplying the NO concentration by 105%, assuming
the NO concentration to be about 5% of the NOx concentration. This
assumption was consistent with the measured NO /NOx ratios for gas fuel
and was conservative for coal and oil fuels where the measured ratios were
1.015 and 1.027%. The N02/N0x ratio will be investigated further in Phase II.
In the cold line system, the sample is in contact with water during the
transfer and drying process for a period of 5 to 50 seconds, depending on
sample rate and line length. Since NO is soluble in water, it is expected
to be lost in the cold sample system, and no measured NOx values from the
cold system are presented.
In addition to losing the NO in the sample, reactions may be postu-
lated in which as much as one mole of NO per mole of NO is lost in the
water. KVB's experience with many different sample configurations, different
amounts of water present, and variable contact times does not indicate that
a significant quantity of NO is actually lost. This is supported by compar-
ison with simultaneous measurements made by the PDS analysis method on
grab samples taken at the stack. The loss of a minor quantity of NO may
be due to an initial NO reaction with water which produces an acid condition
31
-------
that inhibits further reaction.
In the hot line system, water condensation is prevented by electric-
ally heating and insulating the sample line, and NO and NO loss into
condensed water is prevented. However, to present data on a consistent
basis, all results are reported dry at 3% excess O , although the sample
stream on which the measurement was made contained a significant amount of
water vapor. The measured concentration was changed to a dry condition
by a correction factor based on flue gas water content as calculated from
the combustion equations assuming typical oil, coal and natural gas fuel
chemical compositions. The factor was assumed to be constant for all fuels
of a given type. For all natural gas fuels a flue gas water content of
15% was used, for all oils 8% and for all coals 5%.
The moisture content of the flue gas was measured as part of the
particulate measurement using EPA Method 5. The measured concentration
for coal ranged from 9% to 11% with an average of 10%, for oil from 5% to
15% with an average of 10%, and for natural gas from 10% to 19% with an
average of 15%. The moisture contents for natural gas and oil were about
equal to the theoretical, but the moisture content for coal wcis very much
higher. For coal fuel, the effect of water in the coal and in the combus-
tion air on flue gas humidity was investigated. It way found that typical
values of fuel moisture could increase the flue gas moisture by about 0.5%,
and atmospheric humidity could increase it by 1.0% to 1.8%. These effects
are not enough to account for the measured flue gas moisture content being
about double the theoretical amount, however, and the cause is still under
study.
Errors and uncertainties of various kinds in the data, have been
postulated. It is assumed that these errors are random. Rather than
estimate individual error components and their combined effects, it was
felt that the best indication of the measurement error could be derived
from comparison of the NO values measured using the hot and cold line sam-
ple systems. This comparison allows an experimental measure of the various
32
-------
I
possible errors in each sample collection and processing system. Table 3-2
presents the results of a statistical evaluation of all 168 comparisons
possible from Table 4-1. it is shown that the hot line and cold line measure-
ments agree very closely. The correlation indicates about 2% lower cold
line reading than hot line reading with an error of +_ 3 ppm at 99% confidence
level.
3.2.2 Carbon Monoxide and Dioxide (CO and C0_)
2—
Carbon monoxide and carbon dioxide concentrations are measured
by Beckman Model 864 and 865 short path-length nondispersive infrared
analyzers. These instruments measure the differential in infrared energy
absorbed from energy beams passed through a reference cell (containing a
gas selected to have minimal absorption of infrared energy in the wave-
length absorbed by the gas component of interest) and a sample cell through
which the sample gas flows continuously. The differential absorption
appears as a reading on a scale of 0 to 100% and is then related to the
concentration of the specie of interest by calibration curves supplied with
the instrument. A linearizer is supplied with each analyzer to provide a
linear output over the range of interest. The operating ranges for the CO
analyzer are 0-100 and 0-2000 ppm, while the ranges for the CO analyzer
are 0-5% and 0-20%.
Specifications -
Span stability +_ 1% of full scale in 24 hours
Zero stability +_ 1% of full scale in 24 hours
Ambient temperature range 32°F to 120°F
Line voltage 115 +_ 15V rms
Response: 90% of full scale in 0.5 sec.
Linearity: Linearizer board installed for one range
Precision: +_ 1% of full scale
Output: 4-20 ma
33
-------
TABLE 3-2
STATISTICAL EVALUATION OF NO VS NO
HOT COLD
Sample size = 168
N°COLD = -9213 + -978° N°HOT
coef. of determination = .993
correlation coef. = .997
standard error of estimate = 13.551
standard dev. of slope = .0062
standard dev. of intercept = 1.794
=N°COLD' X = N°HOT
Y = A + mX + E
= A + mX + t S /(I - r2)
— a r
= A + mX +_ 2.576 (13.551) /(I - .993) , a = 99%
= A + mX + 3 ppm (at mean)
6000-28
34
-------
I
3.2.3 Oxygen (OJ
A Teledyne Model 326A Oxygen Analyzer is used to automatically
and continuously measure the oxygen content of the flue gas sample. Oxygen
in the flue gas diffuses through a Teflon membrane and is reduced on the
surface of the cathode. A corresponding oxidation occurs at the anode
internally and an electric current is produced that is proportional to the
concentration of oxygen. This current is measured and conditioned by the
instrument's electronic circuitry to give an-output in percent 0_
by volume for operating ranges of 0% to 5%, 0% to 10%, or 0% to 25%.
Specifications -
Precision: +_ 1% of full scale
Response: 90% in less than 40 sec
Sensitivity: 1% of low range
Linearity: +_ 1% of full scale
Ambient temperature range: 32-125°F
Fuel cell life expectancy: 40,000 -hrs
Power requirement: 115 VAC, 50-60 Hz, 100 watts
Output: 4-20 ma
3.2.4 Total Hydrocarbons (HC)
Hydrocarbons are measured using a Beckman Model 402 hydrocarbon
analyzer which utilizes the flame ionization method of detection. The
sample is filtered and supplied to the burner by means of a pump and flow
control system. The sensor, which is the burner, flame is sustained by
regulated flows of hydrogen fuel and air. In the flame, the hydrocarbon
components of the sample undergo a complete ionization that produces
electrons and positive ions. Polarized electrodes collect these ions,
causing a small current to flow through an electronic measuring circuit.
This ionization current is proportional to the hydrocarbon concentration
entering the burner. The instrument is available with range selection
from 6 ppm to 1000 ppm full scale as CH . A summary of the instrument
specifications is presented below:
35
-------
Specifications -
Full scale sensitivity: adjustable from 5 ppm CH to 10% CH
Ranges: Range multiplier switch has 8 positions: Xl, X5, XlO, X50,
X100, X500, X1000, and X5COO. In addition, span control provides
continouously variable adjustment within a dynamic range of 10:1.
Response time: 90% full scale in 0.5 sec
Precision: +_ 1% of full scale
Electronic stability: +_ 1% of full scale per 24 hours with ambient
temperature change of less than 10°F
Reproducitility: +_ 1% of full scale for successive identical samples
Ambient temperature: 32°F to 110°F
Output: ^—20 ma
Air requirements: 250 to 400 cc/min of clean, hydrocarbon-free air,
supplied at 30 to 200 psig
Fuel gas requirements: 75 to 80 cc/min of fuel consisting of 100%
hydrogen supplied at 30 to 200 psig
Electric power requirements: 120v, 60 Hz
Automatic flame-out indication and fuel shutoff valve
Difficulty was experienced in maintaining the h/drocarbon instrument
in the field and cften hydrocarbon measurements were not made. After loca-
tion 22, little hydrocarbon data were obcained because ?£ water in the sample
gas condensing within the instrument. Since then additional insulation has
been added to the plumbing within the analyzer, and the temperature of the
sample collection line has been increased. These changas are expected to
eliminate the water condensation problem.
3.2.5 Total sulJrur Oxides (SOx)
SO concentrations were measured by wet chemical analysis using the
"Shell-Emeryville" method. The gas sample was drawn from the stack through
a heated glass probe (Figure 3-4), containing a quartz wool filter to remove
particulate matter, into a system of thiee sintered glfiss plate absorbers
(Figure 3-5). The first two absorbers contained aqueous isopropyl alcohol
-------
Fluo V/ull
L n d of Opening
15 mm ID
Pyro motor
one!
Thermocouple
Figure 3-4. Flue installation of sulfur oxides analyzer
Sintered
Glass
Absorbers
Sample in
.. Dioph ra a m
Vapor Trap _ a
Pump
Figure 3-5. Sulfur oxides sample collection apparatus
6000-28
37
-------
and removed the sulfur trioxide; the third contained aqueous hydrogen peroxide
solution to absorb the sulfur dioxide. Some of the sulfur trioxide was
removed by the first absorber, while the remainder, which passes through as
a sulfuric acid mist, was completely removed by the secondary absorber
mounted above the first. After the gas sample passed through the absorbers,
the gas train was purged with nitrogen to transfer sulfur dioxide, which
dissolved in the first two absorbers, to the third absorber to complete the
separation of the two components. The isopropyl alcohol was used to inhibit
the oxidation of sulfur dioxide to sulfur trioxide before it got to the
third absorber.
The isoprcpyl alcohol absorber solutions were combined and the
sulfate, resulting from the sulfur trioxide absorption, was titrated with
standard lead perchlorate solution, using Sulfonazo III indicator. In a
similar manner, the hydrogen peroxide solution was titrated for the sulfate
resulting from the sulfur dioxide absorption.
The gas sample was drawn from the flue by a sir.gle probe made cf
5mm ID Vycor glass inserted into the duct approximately one~third to one~
half way. The inlet end of the probe had a section 50mrri long by ISmir OD
which holds a quartz wool filter to remove particulate matter. It is
important that the entire probe temperature be kept above the dew point of
the flue gas during sampling (minimum temperature of 260°C). This was
accomplished by wrapping the probe with heating tape.
3.2.6 Particulates (Part.)
Particulate samples were taken .i\: the same samp? e port as the gas
sample using a Joy Manufacturing Company Portable Effluent Sampler. This
system, which meets the EPA. design specifications for Test Method 5, Deter-
mination of Particulate Emissions from Stationary Sources (Federal Register,
Volume 36, No. 21, page 24888, December 23, 1971) was used to perform both
the initial velocity traverse and the particulate sample collection. Dry
particulates were; collected in a heated case that contained, first, a cyclone
to separate particles larger than 5 microns and, second, a 125 mm glass-
fiber filter for retention of particles down to 0.3 microns. Condensible
particulates were collected in a train of 4 Greenburg-Smith impingers in
a chilled water bath. This study was performed as a research project so
both solid and condensible particulates were measured and the sum is
reported as total particulates. It should be noted that the EPA source
38
-------
I
standards are based on solid particulate only. Therefore, care must be
taken to consider only solid particulate if these data are compared with
EPA standards.
Another point of interest involves the method chosen to calculate
particulate emissions in g/MCal or Ib/MBtu from the experimental data. The
particulate sampling train, properly operated, yields particulate mass per
unit flue gas volume; although some uncertainties exist related to the dis-
tribution of SO and its effect on the solid vs. condensible particulate.
3 3
Having measured Ib/ft , it is necessary to establish the flue gas volume
per million Btu heat input if emissions in Ib/MBtu are desired. The ori-
ginal Method 5 involved a velocity traverse of the stack, the cross sec-
tional area, the flue flow rate, and fuel heating value. KVB experience
is that the measured average gas velocity and the fuel flow rate are subject
to significant errors. A revised and more accurate method has been promul-
gated by the Environmental Protection Agency in Reference 17 that utilizes
a fuel analysis (carbon content, hydrogen content, high heating value, etc.)
and the measured excess O in the exhaust to calculate the gas volume gene-
rated in liberating a million Btus, and it includes excess air dilution.
The velocity traverse approach generally results in a 20 to 30% higher value
and is deemed to be less accurate.
Samples for particulate size determination were obtained by placing
a stainless steel probe in the flue and drawing a sample of the flue gas through
a filter for about one minute using a vacuum pump located downstream. The
filter was contained in a Millipore brand filter holder which, in turn, was
housed in a box heated to 120°C to prevent condensation from the flue gas
sample. The filter itself has a one to two micrometer pore size. The
particulate size distribution was determined by enlarging the image of a
portion of the filter with an electron microscope and visually counting
the particulates.
3.2.7 Smoke
Bacharach Smoke Spots usually were obtained using a Research Appliance
Company Transmittance Particulate Monitor that was modified to measure
reflectance. The instrument measured the amount of light reflected from
a spot on a paper tape that was soiled by passing flue gas through it for
a fixed period of time. The percent reflectance reading was converted into
39
-------
Bacharach Smoke Numbers. In operation, a vacuum pump drew the stack gas
into the instrument and through a sampling nozzle that 01 reeled it ont.>
the filter paper tape. Pollutants in the sample were deposited on the tape
in a 1" diameter spot. After each sample was taken, the tape was automatically
advanced to position a clean section under the nozzle. A simple, straight-
through condenser was mounted upstream of the instrument to remove water
from the flue gas so it would not wet the paper tape. The condenser was
built to specifications supplied by EPA, but it did not remove enough water
vapor during hot weather or when operated in a hot boiler room. When
operated in the field, the paper tape was wetted by the moist flue gas and
frequently tore apart when advanced by the automatic mechanism. To remove
sufficient water from the flue gas, the condenser coil would have to be so
long that a significant quantity of smoke would be scrubbed out, and the
smoke spot measurement would not be useful. A problem that was encountered
during coal firing was excessive particulate buildup that clogged the
mechanism. Most cf the reported Bacharach Smoke Spot data were taken with
a standard hand pump device.
3.3 CALIBRATION
The necessary span and zero gases required for calibrating the
various instruments were carried in the trailer. NOx calibration (500
ppm NO, 50 ppm NO ) , CO, CO calibration (100 ppm CO, 15% CO0), hydrocarbon
calibration (250 ppm CH ) gases, zero gas (pure nitrogen), and hydro-
carbon analyzer hydrogen fuel and air were contained ir. A-size cylinders.
The calibration and zero gases were supplied to the instruments through
2-stage regulators and hand valves to the sample/calibration manifold
for each of the analyzers. Hydrogen free air and hydrogen fuel were supplied
to the HC analyzec from an A-size cylinder independently of the sample/
calibration system.
3.4 TEST PROCEDURES
All measurement equipment was carried from site to site in the
Instrumentation Trailer. The trailer was parked within two hundred feet of
the sample point. The six-man crew was divided into two shifts corresponding
to the hours of ':he day shift and the swing shift of the boiler house
personnel.
40
-------
I
I
Concentrations of the following species were measured:
Species name
1. Total Nitrogen Oxides
2. Nitric Oxide
3. Carbon Dioxide
4. Carbon Monoxide
5. Total Hydrocarbon
6. Sulfur Trioxide
7. Sulfur Dioxide
8. Solid Particulates
9. Condensible Particulates
10. Smoke
Symbol used
NOx
NO
co2
CO
HC
so3
so2
Sid. Part.
Con. Part.
Smoke
The boiler efficiency was calculated and reported using the ASME
Test Form for Abbreviated Efficiency Test, revised September 1965, Power
Test Code 4.1b (1964). Total sulfur oxides and particulate concentrations
were obtained by adding the individual concentrations of Species Number 6
and 7, and 8 and 9, respectively.
All species except sulfur oxides, smoke, and particulates were mea-
sured and displayed continuously by analyzers and strip chart recorders
located in the instrumentation trailer. The sulfur oxides, smoke, and par-
ticulates were measured at the sampling port one time during most baseline
and some low-NOx tests.
Existing sampling ports were used whenever practical to reduce the
expense of the measurements. If the existing ports were too small or poorly
located, the boiler owner was asked to install new ports. If the boiler
had a plate type air preheater, the sample ports were placed upstream of
the preheater to avoid air leakage. If the preheater was tubular, ports
located downstream of the air preheater were acceptable.
41
-------
Prior to starting the measurements, the gas velocity pattern across
the flue was measured by making a series of gas velocity traverses with
the EPA particulate train pitot tube. If the flow pattern was non-uniform,
a profile of the excess oxygen concentration was also obtained. A single
point within the flue was typically selected where the gas velocity and
excess oxygen were' representative, and all gas samples were withdrawn
there. Most sample points on these industrial boilers were in small
ducts far downstream from the furnace where a single sample point should
be sufficient. In cases where more severe concentration gradients existed,
multiple point particulate sampling was used.
During testing, two sets of data were recorded: (1) control room
data which indicated the operating condition of the boiler and (2) mobile
laboratory data that were the readouts of the individual analyzers. Copies
of each of these data forms are included in this section.
While the measurements were being made, the gas console operator
filled in the mobile laboratory data sheet and plotted the total nitrogen
oxides measurements. The plot was used to visualize the trend of the
measurements and to catch any anomalous measurements. Normally the tests
were conducted with the boiler control in manual in order to stabilize
operating conditions and accelerate the test program.
42
-------
I
I
K V B, INC.
CONTROL ROOM DATA
Test No.
Engr,
Test Number
Unit Number
Fuel
Capacity(K#/hr)
Furnace Typo
Date
Owner
Loca tion
Burner Type
1 . Tost Nun, her
2. Load (kil/hr)
3. Control :;c-'_liod An', o/i'a'id
4. Burner-Out-Of -Service
5.
6. Oxygen/Air Level (".,)
7. Drum Pressure (psig)
8. Final Ster.n Pross/Tc : p (nsig/T)
9. Fofdvater r]ow (Mb/hr)
10. Feedwat-rr Prcss/Tenp(psig/°F)
11. Air Flow Prinary/Socondary ( )
12. Air Tcnp Pr inary/Sccondary (°r)
13 Fan Sett ing FD/ID
14.
15. Furl Flow ( Ih/hr ) *
16. Fuel Prt-'S^/Tcrr.p (ps'a/T)
17. Fuel Atomization Press (psig)
18. Pressure Furr.nce/'.'j i.cbox (iwo) |
19. Register Settinq
?0.
21. Smoke Motor
1
-
* Fuel Flow in Ib/hr needed for efficiency calculation.
60-2
43
-------
K VB. INC.
MOBILE LABORATORY DATA
Test No.
Test Engr,
to
Test Number_
Unit Number
Fuel
Date
Capacity (kif/hr)
Furnace Type
Owner
Location
Identification_
Burner Type
1. Test number
2. Load (k#/hr)
3. Flue Diameter (ft)
4. Probe Position
5.
I
7. Water Content (5. vol.)
8. Oxygen (%)
<
9. KOxfhot line) re.''dihq/?3'i Ojl'ppin)'
10. K0(hot line) reading/C^l 02 (PF")
11. NO->(hot line) rrv.O in9/93'i O2(pprs)i
12. NOx dry @ 3* O^Cnot linolppm
13. NO dry @ 3"i Oj (not lino) (ppw)
Id . NO? dry @ 3^ 02 (hot lino) (ppm)
15. Carbon Dioxide (?«)
16. Carbon Monoxide (ppm) uncor./cor.
17 • Hydrocarbon (ppn)
18. Sulfur Trioxicle (ppm)
19. Sulfur Dioxide (ppti)
20. Total Particulate (g/Mcai)
21. Total Particulate (lb/;'.btu)
22 . Smoke Spot (Bacharach)
23. N0(cold line)reading/dty @3"s (ppm)
24 .
25. Atmos. Tcmn. (F°/C°)
26. Dew Point Tenp. (F°/C°)
27. Atmos. Pressure (in. Hg)
I
1
I
1
t
1
;
i
I
i
,
f
i
;
•
•
.
\
?
i
i
i
j
60-3
44
-------
I
I
SECTION 4.0
DISCUSSION OF TEST RESULTS
The field test program resulted in approximately 1250 measurements
These data are discussed in this and the following four sections. This
section presents a summary table of all normal and low NOx emissions data
followed by discussion of the data organized according to fuel fired and
pollutant or property measured. Sections 5, 6, and 7 discuss the data
in terms of the effects of boiler operational methods, fuel properties,
and boiler and burner design characteristics on NOx formation. Section 8
presents the results of a statistical analysis of the baseline operating
data. Unfortunately, some of this material is redundant; but this organi2ation
allows a reader who is only interested in coal firing or design effects for
example, to focus his attention on those topics.
All of the measurements made on the test boilers when operated at
normal settings and at low total nitrogen oxides emissions settings are
summarized in Table 4-1. The data are tabulated in order of Test Run Numbers.
The Test Run Number consists of two parts: the basic test designation which
corresponds to a particular boiler-fuel combination to the left of the dash
and the run number within the given test to the right of the dash. A
typical test consisted of six to ten individual measurement runs made with
different settings of the boiler controls.
The Location Number in the second column positions the test site
geographically on Figure 2-1. Locations distributed throughout the conti-
nental United States were chosen to insure that a variety of fuels would be
tested. The Region Number is the Federal Power Commission Region shown in
Figure A-3 of the Appendix. Since the greatest total consumption of energy
is in Regions 2,3,4 and 5, the majority of the testing was done in these
regions.
The columns from Boiler Number through Capacity indicate where the
particular test falls among the principal variables developed during the
initial test planning as shown on Table 2-1. The distribution shown in
Table 2-1 was developed to provide a cross section of the current boiler/
burner/fuel population.
45
-------
a-;ss
2iU&
s . f
dsl-
a M o
3--U3
Ipl
3311
*r*t
s"fl
sp
,u
all
.0-$.
311
8s!1
8
5-j-
ji^»
V&
go—
M
lll<
M
II.
Hj
1*
Hi
lit
«s&
33
SI
*I
*u
Ijl*
Si*
1 1 I 1 1
m to m in
m i co co co
wi rT
t . i I g 5 ' ' II
rt r-t
m mCDC ^Ttf>tH>H
n\onoinc r-no^
n ^ ^ .-. -»
^> *-» ^ ^^ — -
omcocNi£>r r-mnr^
o
to • r» ^r co
• *H « •
• ^ . M) • f-4 • tH • «H
m^m^in^in^Ln^
^r^^r-^i-^r-^r^
p-' d r-" ci r-" :i ^' ci r-* i
r i
« u -o o «
(0 c i rt J M « C S M
K(4.lQOQ"j('J-HO-H
(Qi-Hi4<3M'3<
O O O O O
§ 1 g
O 0> 0) M M
ui en w < <
S ^ S S i
*» « *» ^r *
r-4 ^ rH M H
m m m m m
r>4 n 04 r« n
23233
O4 rH
H 00 -4 *0 •»
1 1 1 1 1
^i ^i H r4 n
lit t 1 1 1
V 00 TP CD
t-. r- t*. r^
(N aj oo o Mm
r- H
iH rH M
^GDtHO^'CN'^iNO^-^
\OCOVDOO^O*i3^T^T
^, ^, _ ^
SS^SS^S*^*
^-, *— -- ^->
COO^vDOCNncNC-mcN
T
r* <£ m . o^
. . . tr •
. ^| . r-t • M .CO • (N
m*nr>--,ro^-.n^.^r^,
cn.^vi)crv^ocrvio
(111
C* 4 \o i-i
A A A A i
1 1 1 1 t
oo o o r- P»
t-- r- aj r* r*
i i I I i t < I I 1
I i I I i I I 1 1 I
I I i i i i I i t I
i I I I 1 I I I i I
o **-» *n ^*.
m t-4 ^r oo i i i i i i
O T O fN
1
r- H .-. ^-.
A A .*-.«.*-
.-. O ^D (N O
C T CTi (T. O
r~ •-(
ui — o^^u5^^r-—fo^~
DfMcnr^iTr^incNino
r-> tN r^ CO
^q-uir^oinovom^T
ut m
r* ^* o r-
(N . . .
r * ?A m ^
ifJ-*-V^-.(N^-,CNx-,{N — .
• in -in «cri •£?> *cr»
.-ICMrHfMfnrjmtNfniNt
r4-~^i~| — PH — - iH — <-( —
•O *O 4J W T)
OS 3 fl W q n] S i-i
-,-t Q Q Q <0 -H i-t Q Q -H
,.cJHJJrc-tXtJhi<
O O O O O
!7* O* 0* O1 tT>
C C C C C
<3 3 "S ^2 "2
S g g S £
1*1 CO
1 1
.H >-t
ii «n
••^ t-< en m o»
-s \ S. V. X
C4 f* *-4 ft i-t
ff» eft c?> eft 01
M m M M ft
1 1 1 1 1
^ <» m m m
i i i i i
so **! r- ^c •-*
CO CD f^ CO CD
I I I I I I I 1 I I
1
I I I i I I I I i j
I I I i i 1 I I i i
i i 1 i I I i i i i
m ^. m *-, (N . — ,^J-^-.^D^^
OrsiomrMOrHv£>.-t«3>
^li-<»HfHO«HO'~tO*H
. ^ . ~ . . j
CN (N fN O 10 i
O (N vD OJ O* *
r4 ^H ^-* l
_ ' « 1
•^oCcOCS^J'OO^ICD'o^l
C-s-CDtTcr.!!)^^^^ j
1
« « « ^, ^- !
C •tf*l'r-4CT»COLnu")'«T
ccfir-mcnminojuieN
. — — , — <-».-. |
u-'C^-^rN^Dr-vD'^'vO'fl' j
i
MJ p» . m o
«? . j
r- v ^H ^r co
t
S
vi),-M>,-.\o.-.y3^-.«i;«
• LTI • in >LD >Ln *tn
OJODCDCOCOCOCDCOCDCO
f> ~-» <*•) -— • t*i •^- f** •— * ro *— •
W * tJ T3
01 r: id > TJ S v< «** vi
rJ-H-H QQ 6 Q~^i»iQ
Pi^Hcc^j^J^aGO
r.^ rH ^ rH rH
O O O O O
ui m in in in
*fc *fc ^c «; «:
!§!§!§
•u c « « a>
(.1 CO OT CA U)
E2 S • g S g
i
f
t- p- r- r- c*-
X X \ X. >x
tj* en en en ve vO U> «O
I
I
I
I
I
I
-------
I
I
8*8
I'll
s . a
i :i ' '
ml
ulS
-sisa
a a oii
'" (W —
45
«u
.4!
B|!
8*'
all
K"
g3||
O tf
X
*
c -«
X °" **
V MO
*" -g"* £
«'$!
J'i
:£
*fr
il
I?
i, S|
HE
il
3^
SP
its
h • "
S31
In
t i I I i
(N CO
3 g 1 1 1 1 1 1 1 1
<-) .H
°. ° i i i i i i i i
O
CN O 1 1 1 1 t 1 1 1
00
in o j 1 J J 1 1 * I
t 1 1 1 1 1 1 1 > 1
oo*oooSoooo
>£> ty O 4) V <0
S £ g £ £
°* (N tN o>
11 ^ 0 t 1 i t 1 1
(N 0?
in co
CO
It in VD It i i I I
•H '^
V O
t I U1 CO ) 1 till
II II II till
oo* oo oo oooo
r-H •-^ O CM
r*. iH f-t *H iH
Tf rH VDCM t^-ri \£)CNvOOJ
s5 ss s;s ssss
mcN r^fn (*-«*i vpiNiCM
O
. r- X> CO
*§i«| «| -o
co co
i i i i i
(N (N
2 g i i i i i » i i
•«» r-t
(Jl O
"1 "^ 1 1 1 1 1 1 1 1
Ch to
•W vt> | 1 1 1 t 1 1 1
to oo ) r ( i i i i (
i i i i i i i i i i
o o *c
— . CO O -* «-i — * —
OOrHiHOOOCrtOO
r* co r- oi
o o • o o
sISlSlsJlJ
s5wKSSSSo?
. _ . — . — . ~ . —
*IS|?IsIs§
"J1 CO \f> O (N
• iH "CTi • i— t *OH "O
• o" • o * o^ * o" • o*
t
33sS3333S8
•H -H -H -H -H
O O O O O
§ I § § i
0) 4) 01 41 (1)
g g g g g
Ot (N (S
-------
2 8*
*.
a «
S
c
lilt
f < U
9 5 § *
AMU
,J^3
m
*•-*?"
332?
as$
** a, *-
45
«?
fi£*
sit
aO
. s
8|!
•rf*-
M
a-
•HI
s »
•
f%
its
: &
!o~°
& *
*» « "S ?
ilss
—
» .
H«-
a a
«£
H (•
4* r*
!i;
j&
ill
HI
>« a
0
».
if «
88 i
s!<
22i
i i I i i
r- ci o^ r- u>
m in
1 I 1 1 g 0 | | | 1
en in
CO D
.H H
1 1 1 1 0 0 1 1 1 1
in o
i t I I £ • I I i I
0 /i
r- [n
i i i i * • i i i i
o m
o>«o<*-rt.-xr-i--»cM.— •
OCMO^FOrOOO4O»vD<*>onor»tN
O4 »-( fl t-H »CN '^rHtH
.*- -^-r4
OiCO^TCDtn^tirt^HU?O4
HcO*-t^OHr- id 3 u w c «3
•HQQQ^1^*^^ C
XiJJJt-i^m^Kt-
§ g s: g i
S1 S1 I? ? |
3 2 2! 2 3
££-.;££
M CM n CM f*
H >-l p-l C4 04
O> 00
X \ ^v X. X
(N cr> ua
>H
or>r- ^^^.oScr.— .
Ch'y>CNO'HCD'^l'Hr^r>4
(NMOifHCMCTiOJi-li-tCO
vo ? r- in
m in T .-4 i i i i i i
f*l •-( CM iH
StHEN^r^^i-tmCO— ^
^OOfMtNOi/lrHCOvO
rOiHOli-(Oti-4CMiHi-1CD
O
CM . rH o o
CQ^-CO « 01 ^-. 01 — M x^
-O 0) ffl T3 -O
>«3k4tnc sj s O\ \O *0 *O
O O O
o
1 1 1 1 1
•H --* "* ^ *H
1 1 1 1 1
cn o* *n »D r^
in in
1 ' S3 ' ' ' ' ' '
0 0
n us
§s
I I 0 0 1 | t | 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
,-> m — . co --,
co O **i -— M — moTO
r-oir-toino -o -o
t*lfMO*HO^T(NOr*O
. ^. . ^- . ^^ A N A | M (N
^ s- ij ^ s?
« ,0 » ,0 „
0
1 t 1 1 1
•^ ^ ~« -1 •-*
II II 1
P- vD ^ -V F'
1 1 1 1 II II II
i
I
1311 It II IE <
Q\ CO
I I r- r~i II II II
• in
CO O
1 1 CO T II II 11
. LTi
r-\ —*
I 1 1 1 II 11 II
in
iH O — O ~~~ O — . O ^
o o o o c> o
m r- co o o
>y\ CN rN oo T
•-I r-4 rH
jyr^inii) f-i-^p mr< fi-q- i
|-^.^-(*3it/i C^M O tT> v£>CQ
•rr-i^rj r~m *DCN ^?i-t
'"'" '" ' " ' ".
_ -. ~ -
iNCTi uD(N f-f~ -(JCD !
i l ^ in ncr> a\ n r-co 5
\O M r^ (N IT, r-* ^r r-i
COCDO\D cor- 041^1 vco
fficMt--^ cr\«H «H^ r-co j
^ (N v^ tN r- <^i \o f^i T f-i ;
m
^ <& ,-* . as
• iH • CO «O 1/lrH • I/I
eD^^H^1 P- ^D .CN OT ,
O*^»O)^N. (>l*-^ CN — . C4 — * 1
I
oj a* *a 'tJ
^r t/> w c -4 <
•-I r-4 ^H f-t
id « * "S
i 8 8 3 S
DI V XV ^O Ji! O JC
C TJTJO'O'OO'OTJOTJtJO
; -H ca)4-ic«4Jca>4J^iU4J
a DbutJiOtMO)3U4WDtn^
S £ S £ B
r>t n n d oi
6 o o c
ii i i t
U CN| M 01 0*
,-.
O4 M f*) f> **>
Sf > ^ i >
UJ ^ ^f ^ "»
>n J5 3 15
(M 04 vO O
rt ^ rH rt rt
-------
I
I
I
m
1
g
4 S*
"* « 3
« £ M
«-( 4J
« no
fr» D O
•O
V
C
H*
*s«&
sj,r
Solid ic
»»rtlc. •_'
9/XcH -.:
(l/KStu)
• -H 3
« S ? ?
gSsS
rJ* 1.
O X _&
d!
tf
n
g^s
all
s~
8*-
1st
o-'sl:
*S^~
_ 8
*
§3 3 1
0 O*
•
c -*
5.1 S«
B*^I
s«r
u o
to
nil'
*a JiS
£ _
I<*
Is!
If-
i«
|l.
I§
sc
III
:ir
.••a?
si
*l
il
Hi
*3,«
JS .•
si*
lu
i i i i i
IN *r co *ri «N
r. in •* ^
«Jp r~* C* fl
m n O 00
«£ ' ' ' ' ' ' ^M
r, ? ^ ~
CO CN u) r-
(V) rt iH a?
O
n
ii ti it * i «r GO
• r-t
vp ^*-
«
<& *r
II II II II ^ CD
• «H
^ -k^
^ ^
O M
o a & ~
o ^ o-^ a^- ^-rj moo
O D O ^D n O (M
~
•* on o o
• • M * •
o PJ • n rj
•-I M 03 «H ^-1
^. ^ » ^
r-fN» rim MO Of
UltN ii>(N inp) \O(N
cno) OfM O^T .-tin ^-M
•-i\D
• — • - • - •" • -
, - r~SJ * o voS M « c
«-H -HO 00 5-H *B-H
cq<~4 K KJ i-5 t-5 i-5 < « rH
flj « n3 flj (0
5O O O O
O U U O
M M ^ h 'd
M au ID V4 ait4 D id
v -X a> xu xa) x 4)
c*vc:4*-tJciu+Jc:a
pCMWDCucnscnto^c^uaw
S? S £ S £
(N n r>» M CM
rt o m ^ ^ ^ Jr
H r-t <-4 i^
u) co rH ^ m
i i i i i
fx r* r* f* a>
H «-4 rH «H -H
I 1 1 1 J
•H O* O M (T*
o"
^r (N
v o
V CO
1 1 1 1 1 1^ .J. 1 1
in a\
O \£>
if) r*i
Oi
o cn
i i i i i t CN r- i i
• i-i
U> .—
r-«
T r-t
l i i t l l m co 1 l
. f-t
tc _*
r- T3
3 id S )^ rc « q id
5355S3S=ISS
5 8 3 8 S
15 T3 tJ T3 -O
rtj rtj o] nj fl
4t Q) (u n m
rn n *n in in
•H *H r-H (N C+
r-4 U5
1 1 > 1 1
CO 03 CO O* W
I 1 1 1 1
CT\ (N vO f '^
co — r- LT>
IN U3 T M
o v ^r cr»
( , m in ri ^ , , , ,
*r ^- o^ ^
vD CO CO 00
CO V rv n
, , m »n •» ^ ( , ( ,
m
01 \o
1 1 1 1 CO VO 1 1 1 1
- tH
m "- •
00
sj en
i i i i m m j i i i
- ^H
«n —
(M ^t^-.fn^^^«m^.
OCHOCN o*Ho-~(Of-<
co ^. r- — , .-H — » t o«*-^
i/itnno chooiHtn^-i
^ r^ ^> en
CO • « * .
«N 0 C* go
m «H >-H »H
CO-, — . U> — ^^^H^.
^vcofo » ^r
tfl-»U3J--i£)^in^-.O--.
•O OJ 01 TJ T>
S H3 ? M « C <0>(Q
aOO-'-fnJ''-*-^QQQ
•3>3-3
o o o o o
U 0 U U U
•O O* Q*
«n m CN n 04
m in in in tn
M CN f— t tO
^ M ^ °. . 1 1 1 l l
u~i m
l l in co l t i l f l
• •*
^ —
r^rH'i) (Nf>»
. ~- . ^ . — . ^ . —
*"'". S^d^Cl^Ci
en *^ o f- -~4
in \o in co vo
vD.-.fn-^ — eo^^^J^.
n~n~*-~,~m~
1
« 0) *O -T3
> M « C « S nJ J M
ss5sss3aaa
^ iH ^H iH
«J O O O O
U VD 10 \O U>
•4t: * * *
| | | 1 1
ft -p 4J -p 4J
CO « « W W
B S? S £ S
CN
-------
lui
8 , fr
3S S.
88-S
« u"3l
JH§
0. —
. ri 'a
m
41
ill
js
8?.-
eil
&
s__
•51*
_, «
S_-
i^t
£ •
*
33S3
z- ID *O «
CD CO CO CO CO
M ^ m in fi in
04 o en co r^ co
*D Ch vD m CD *f
rH O O O 1 | | ( O O
\O --, O M •* m
O UD fO \D *» ID
fr O CO «3* r- Cfi
rH fH O O 1 | | | .-* O
S>
I I r- (N i i i i i i
• m
TP ri
I i r* *r i i i i i i
• in
xO ^
o o m
• • \D • O
•H «H • fH *
•H iH Cfi r-( OOvoou)ou>o
• O • \O ><^> M(Nto
4tO «J -H Q Q 5 -H * Q
n OAH >H rH iH
O 0 0 0 0
§ § 1 S 1
V 9 9 V B
V) W W 01 M
g B S S S
fi n <*) <*i *n
n \o \o to co
> 5f > i? ^
d 04 CM C4 r«
3 3 2 S S
O vD
M H V CH H
i 1 1 1 1
fH 04 M (N (N
C4 d 01 M fM
1 1 1 1 1
f- \O « ' CD *
r* r- co r* r-
r-
r- -^ f-4 o
O O | | | | | ) O O
^ 1 1 1 1 1 1 1 1
**! S
<*1 o ' ' ' ' ' ' ' '
" •"*
cn^n^co*-^^^-.
oJ^OOl OiHO(N
^^
*-.vD— .rH— .
CN
n r- o -QO
• • • fH
O^ ff> m ^H CTi
^nTpco^H^rxooir^uD
r*-cor4vDr*»HiNCTi
COMICO^OrHCOrHr^ 1 t
. ~ . — . w, . «
iO(Nn\i>r**^f
t 1
«t a> -o na ® a?
n tn n tf
(T» O
co co co r- co
^ ro
M O
o o
1 1 i 1 1 1 O O t
rO M
(N f~4
O O
1 1 1 1 1 1 O O | 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
\o — r* —
O ft O CN
r- .,-, (*i — , o cr>
O vo
CO • ^0 CO
0 - 0
•H a\ M *r ^r
fHincocomtNujrHom
o\ CD tn m
I I i i I i CN co r- in
« CNI- en
cN^jinmcoomcoCTi^
coor^inor-rjcoc^ ^o
in o in
^ o \o • •
^Sr-Pm^aiS-,S
vo — in«vo*— i/iv*^^ •—
^oovoovDOr^cor* a)
• vO • vO *\D • tn — »
I
*O TJ 4> 8J 13
«J > fl * V* W C iQ
^HQOQQ-HnJ-H-HO
xl-5iJi.5t.5l 01 \ X.
\ \ \ M CM
04 CN M H H
vp « «o cn cn
Ci ON O"i vO tO
d <* r- n «r
i i t i i
•^ -v IT in in
. «N * M M
1 1 1 1 1
rH C4 VD in VO
CO ^4
CO fl
r- c^
iiti • a* t i i i
tn £"
vO PI
CO O
i i i i ^ ^ i i i i
vO tn
co in
1 1 1 1 9 CO 1 1 1 t
• M
tfj ^-«
r* -nT
O \£>
I ii i m co i it i
• <-t
\o ~
m^-*co — VD--.CO co
O«HCD«HOt-lO CCiO CO
m « *D — s
in vo f> m o ^- i ii i
v£> l£> (N
CO (N • - •
. . m m f*t
<*i m -H M *H
o .-. rH -q-o O
. ^ . — . ^,-i «^, v-
O m in m
• in .inCMtNtNCMCiCM
r^^^t^-^rH*—,^ . — ,H —
•O W W T1 -O
S (O J M CO C (t. S nj
Q Q 0 -H «J -H -H C Q Q
^4>H t v i N > N > N
C C rH^M-Hr-4-H
ili Q, O fH »H >H
VD ij» -H t^ Cft
t 1 1 1 (
tn tn 10 \o \o
(M IM (M d rj
-------
I
I
I
£
g
I m 5
« £
S«M H
M O
5 «"33
SS>§
P. —
u3 3
3v>uf!
*s*=
rSI
8^5
s
.11
H-
5!
*>~
all
O1
Jr,
8!
- «?
83^
a"
«
•5 3 "e
U?i3
„ „
•
O » -S
*
W
IPI
*3HS
k*
~* *• ^
s<<
a z
u •
M
s«
J!
b
sl
»*•
ill
tit
sl
is
s^
k*i
« o«»
I1*
2Ia
i i i i i
^ m rH o *£>
cr> v"o
(N .-H
m o
! ! ! ! ™ ! ! ! !
1 1 1 1 — 1 1 1 I
r- *n
ID .-H
»H m
I 1 1 1 ••lilt
1 1 1 1 ^ (N ) 1 1 1
1 1 1 1 "lilt
o* r7
in o
1 i i i m n i l it
1 1 ! 1 < rH 1 1 1 1
1 1 1 1 *T *-' I 1 I t
0* ?
0> *T
I 1 1 1 *D CO 1 1 1 I
L 1 1 1 • iH 1 t 1 1
L 1 1 1 "* — 1 i i 1
lO VD
O »n O VO t 1 1 1 1 1
t t 1 II
1 1 1 I O 3" O S O O
fill -^ *- —
T ^r 03
CO • <*>
ij- fO O
tH ^1 m r-4 l/l
O "• O ^ vD «
ID vD f*1 CO (N \D
1 i i i M rn n 'ri >in »m -J3
t -
IN CO
ifl H • 0 •
. «o .m
^. m »H 01 i-^
^?if^3ino^^or---.
.co -co • tN . iri .0
mH-HM«j*»HcO'-teNin
go— -co^in— <£> — CN —
m i/irHO«-iOi-io
OOlOfNCD.HOO'-tCO'-H
^H>— ^H'-^UJ^^^D^-y^ —
1 1
•r4 M N > M 0) -— i i i i i i
m ^
r- VD
I 1 (N f> 1 I I I 1 1
t 1 • CTi 1 1 1 1 t t
I I M — 1 I | I I I
1111111(11
I I i I I i
OO I OOC^OCJOS
CM ro
tO CD O
•H cr» cn <£ *H
i~* — (N^^-V — .CD^.
int^cDcr>O(Tiin(Ticoin
HincNOi/it^-ijr^-^rf^
• ^i1 «i/i *m •i/icoTO
vo-^co— T-^^T^. ^
.•v .in .(£> ->i>crifn
oo oa m
en * • * ot
• o o m .
00 kD •^
*H O O O O O
COiHOtNO
""> -H r- at
en m T m
m {*>*£> «n i i i t i i
* * • • 1 t 1 1 1 t
~ — 1 1 1 t 1 1
^H iTt CO CTi
O TT ^T TT
(N «sr r- P-
\o m 10 m i j i i i i
• • • • i i t i i i
~ — i i i i i l
•V ^~. fH CO
O (N O Cfi
co o co ** i i ; i i i
• CO * rH 1 t 1 1 1 1
CM — «T ^ 1 1 1 1 1 1
r- — T c*
in r- M o
co f-t co m t > i t i t
• CO • rH 1 t 1 1 1 I
1 1
<-' 1 1 0 f 1 1 1 1
CN vC
m (N .04 •
•H • O
CO O> *H H VD 1 | 1 1 1 1 1 1
o> <** t i i t i i \ i
• *- i i i i i i i i
OJ^VOM^rHvDM^M
CT'f*l\DfM'aifHU3(N»aiCN
U3
• m o* in \o
rH JJ -JJ 4J
m tn in m w
£ £ £ g £
X. \ X. X \
in
rH »n ^ m CD
CO cH Q\ O\ Ch
01 CN fM (N M
1 1 1 t 1
1 9 n m ^
1 CO CO CD CO
en r~
•«T -H
fTCT^COvOaQ^flO
1 |iHCOinM)C^CO
I i^rH^f-tmrHmiH
! | . ^- . _- . — • . _^
m^r^^ GOMCO^
VOfN^TrHiJ1(Nf*lrH' n ^- o ff» **i c*
mojcotnr-co *^rcoiri
CTtsrrHCNO<(N^rHi-HCN
r-*^*rH— *rH-*--tP—- rH —
o" o S 'S S
(Nincifincinnron
CNJ— rH^-rH— rH^--M—-
rn v « *a H
3 (0-H-*HOQQQ--H
b< 03'HX^4^J<
*0 O O O O
* Z Z S 2
S ? 2* S* H1
tn 2 2 2 2
£ £ £ £ G
rH r-t iH «-i
" ^ ^ § $
**>
«» rH (*» in
i-H rH rH »H «H
1 1 1 1 1
eh o o o o
(N m tn n PI
-------
«•
?
I'll
N f
ill
_. o « ^
IW
41
il»
uSl
.* a-1
•}I
i«-
S *4 «
8S>~
B3»
S "
C ~4
^-11
a;
L^
6 '
HII
i*l
u
II
II
p
£11
It!
•\
.ii
is
ft
i!i
y*
, , , . i
v v rH oo r-
V C* ** O\
• • rH rH
f 1 1 1 CO 0 • * 1 1
1 1 1 t rH rH (N rH 1 1
1 1 1 1 •- *— 1 1
^" (M O fN
• . CN fN
1 1 1 1 CO 0 • • 1 |
1 1 1 1 rH .H CN rH | |
till '" — 1 I
CO '0 fN (N
i i i i T m a\ rH i i
1 1 1 1 • iH * rH 1 |
t i i | in •- m ^ I i
m n en in
1 t 1 1 VD ID Cn rH | |
1 1 1 1 • iH • rH | |
1 1 1 1 in •- fl — • 1 1
rH «C
O .-. 1 1 O ,-x 1 t 1 1
O rH 1 1 O '» 1 1 1 1
• — 1 1 -'-111!
•H <£ <-*
OCOrHCD 1 O O O O O
O O O CO
o **? o in v
»H Cn rH rH rH
^f ^* CN rH CO CO f> CO
^r rH n rH i i • t" • r-
m ,~. m ^ in -—
1 l l l m -r O en o> co
1 1 1 1 rH — fN — rH --
rH O CD ^ (N
1 1
•rt n v <\t y lH'rIC/l(fiiJrtC it)
•H rH rH
« * <8
§ i 8 S S
tr & > n i
a-H 3 J4 >i O >t O
S fl. (J O -H U H
S £ S £ £
rH rH (M CN
$ $
Ol OS CO CN CN
H rH (N fN CN
X \ X X X
« « „ „ «
*n o o
C7> 0> rH CN CN
O CO
1 1 1 1 1
O O •-( CN (N
i j |
CO CO V V \O
n CN
1 | 1 I O O I 1 1 I
t 1 l l • • l l 1 l
l l l 1 ~- l l 1 l
ft O
r- "f
I I 1 I O O 1 1 1 1
i i i i • • iiit
l i i i — till
CN -^
1 1 1 1 CO CO 1 1 1 1
1 1 1 1 CM CO 1 1 1 1
fill*-' 1 1 1 1
m .-.
1 1 1 1 CO C* till
1 1 1 1 CN CO till
1 1 1 1 • — lilt
O) -- rH .-. CN — .
I | I | . ~ . — . «
OOOOl 1 O O O O
• • m •
•* -^ 0 fN
rH rH Cn rH •-*
SP^SCoP ^^
^ S H ^ N - ! !^-
CO r-, rH ^^ ^. ^,
SSSSSS SS5S
CN rH CN O UJ
ocN^n • r* • ci »in
0 O-ff^. -t-*Tf«
TJ « 0» * -O
rH rH
« m
8 S £ £ £
1 « 1 4)
&-3S.33 3 3
Mfc H H H
s Sk E E
n M
'•
5. 5. ^ ^ s
«««•««
° fN n n «
o
III II
(N CN P» PI m
m o o o
. ... i
* til r-
CO J 1 1 CC
^-s fN CN
n r* CN rH
r- o cr» in
m o* i i i i i i o O
• ' l l i l i l • •
*-" 1 l i 1 l i "
o crT v m
CO CO rH ID
CO -V 1 1 1 > 1 1 »H O
• • 1 1 1 1 1 1 • •
"-* 1 1 1 1 1 1
in rH 1 I I 1 t 1 U> rH
• rH 1 1 1 1 1 1 > 03
M 1 1 1 1 t 1 \V C4
* rH 1 1 1 1 I t >CO
«n •— i i i i i t CM —
ii) CN — m .-* r»
o — . f — , r-. — o --,
OrH OrHOtNOrH CO
oo o co a
o • •
rH CO rH .H r-l
ocK m^cN^o^ cr. o
^Ci ^.C1^^^^ "", d
SS 5,?S5rHr? £S
Sc? m^r^SSS r^S
^ • en so f«
O • •
-^. ^. «. — . en
O • rH . O • O • ^ rH
CO CO CO CD CTi --,
flC ^ ^ M
3333 3
E S E E E
m m in in 1/1
»H !*• r* o*
XXV. X
<0
« « 0
fN CM fN CN fN
rH O
1 111 1
*** *** a — m m
i i i i i
r-* r* M MI l
co at r- r* (
i i i i i i i : i i
1 1 f 1 1 1 t S 1 1
i i i i l i i i i i |
s
i
s i i i l l i i i i
l l i i t l i : t i
1 i 1 1 1 1 i l f i
1 1 1 1 1 1 t 1 i i 1
i 1 i I 1 1 ! i | 1 >
i i i i i i t ii i i !
i
1 ! I f 1 1 1 1! 1 ! !
i I \ I I I i ii I I
1 1 1 1 1 1 1 II 1 1
1 1 1 1 1 1 1 II 1 1
1 1 1 1 1 1 1 1 ' i
t 1 1 1 1 1 i II 1 1
i
— -xioroo ,-.
OOOmrHCTiOC-JQO
CN \D 00 "J
O M O O j
rH rH rH rH Ch
•oS-ifT^S.n^oP
"^"d^CiTi^"
SScjK3ss:;s5
PS'SSSS-'SS
• — • •- • " • -' • ~ 1
t^ r^ (^- ^D o
r~ <*l 10 r^ en
a* •^r^ooin(rif7'T(N
• -«r . M ' rn . rj . ^H
I !
Q O O -H <0 -H -H O O O i
r3r5r5
-------
I
I
4>
H f
|!
2
£
isif
«•*< -t
(4 0
„„•*?
is!§
• <-* 3
~51
41
«u
s'jS
8*'
8 *
Om
"P^S
£"
5 t
So'*3
a -
nil
fr _
PI
w •
fl
*• r4
]i
ill
lit
!|
«
^
Mi
I « •
33 *
ill
It'll
t 1 I 1 0
1 1 1 1 CO
M *t *f G\
vC m MM
O O (MM
* « 1 1 1 1 1 1 * -J.
1 1 1 1 1 1
mm r- m
MO m M
"" I I I I ! i ""
I I i I I i i I i t
I t i i i i i i i :
i i i i i i i i i i
i i i i i i i i i i
t i i i i i t I i i
i i t i i i i i i i
Ss"pSsSSSs
•v s
in o
*-, -* .— • o — .
OOOOOO«M(MOO
CM
"^ M CO « \D
o
1 1 . v . w- .*_..—
SSSSSPSSSS
pjSrtSsSoSiNS
2 n' ] 4 2
". S ". S " 3 " S °. S
1 1
UC ftf i <0 S M 0) C
«iH*H qQQQ'HiO-H
g i g i g
I 1 I ! I
c c s e e
in in in in to
iiii
d d 0* CM
r* r- r^ r*
X X X X \
d d d d d
ch en ch en at
co *•*
1 1 i CO CO
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 t 1 1 1 1 1 1 II
1 1 1 t 1 1 1 1 t 1
I 1 1 1 1 1 1 1 II
1 1 1 1 f 1 1 1 II
1 1 1 1 1 1 1 1 II
IIII))}) 11
O .-. M — . VO ^ -•
in CN r- 03 m *«T i i i i
o m o •* o m i i i i
ll *-* m o o in -^
1 IOOCOr-(NM OO
r- co • • •"*
o o •
S£S§2£2S 23
3sas3pSs is
CD
OOOOOOt^M ChM
, 40 -a
O O C3 O O
z z z z z
1 1 1 1 I
c E e E E
* vo * *
X X X d d
d d CM X X
MM vp
9> 01 cn m m
vo to
M T CO M <*1
IIII 1
co co co ch ch
O t' O
1 O t> 0 1
n i iici
co I i I co
till i i i i i I
lit i i i i i i
iiii i i i i i i
iiii i i i i i I
I i I i i i i i i i
iiii i I i I i I
iiii i i i i i i
iiii i i i I I i
1 I t 1 i i i i I 1
iiii i i i i i I
I i 1 1 i I I I I I
iiii i i I I I i
*n *-+ M --. in -» o o
i iom Mr-OMfSd
2SSS ?§gsSS
vO
* vo O co ^f
M co m r* r-
BsSS BsSsss
2S 2S Sp ^!S SS
OJcoMf-. Mr^-MvO*"*^
in •
Ch o d o
d in M f*- co
d • t^ H O r^ M a)
iS« -0 t .
* JM S c 9 * -a ^wc
O^( m^'-'q Q-H-H fl-H
O tjj O O O
I I III
t* t* B t* fr*
in m in m m
"XX X X d
vO m <•> /n
^ M vO d *«
1 1 1 I 1
|H CO «7 t 1
eo r- i i i i
M (N M CD M M 1 | t I
o ~ m o co o
vo M ^ co o
_ „
25«^l iSrsffri
SlsSHssSsI
o ch o
O • vO • •
• M •* CO
•^ ^. vO r- rt *T>
momoinomomo
alj!saii|J
"3 «
O O O O Q
z a z o o
c c c en ui
2 a 2 g fe
E E E E E
m in m m m
• ,
d d d \ x
d d
o
M in M M M
till!
M M M EM m
-------
I
•
H 3 31
• 11
5
i
3 = H
^!-
8 W u
Iflf
IlII
4!
ifl
Jl
8|l
r
R*'
"f--
a *
•
a *
«5ll
||S
a '
• ill
"ft
i .
in
Ft
y.
n
1*
in
sit
*3*
ii
||
f-
A|i
il«
23s
0 0
1 O 1 1 «H
s s s s s
O ^r oo u>
S 0 1 , , , , , H 0
* • 1 1 1 1 1 1 '
,111111
r- in o r~
P 0 , , , , , , H 10
^ 1 1 1 1 t 1 -J.
1 1 1 1 t 1
r* —. r- ~~.
CO VO vO -1
•*r r^ i i i i i t m o
* r* i i t i i i • co
« -" 1 1 1 1 1 1 M —
*H T CTl CO
in co i i i i i t *n o
• r- i i i i i i >oo
(M — 1 | 1 1 1 1 C4 ^
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
~ -So
OOOOOOOOOrH
» „ r, « 0,
f^SSSfSSKS
s?sssSss^5
slsaslslsl
n o ci so r.
• M .01 . r*- • •-( • F-t
i i
SO) « X) « O
•5-H?'a5-H!3'S
CQ-HK^i-S 4J « ifi
rH H »H H »H
^ " " X X
in in in in in
rt M (N M in m in
10 tf \O ^ ^
« ^" o --.
1 ! r-
;l5i sl33?i
0??o_ o-,o-r.»
> ""a •
o « 5 t) « «
S'H B(-H" O O 5"^ iTl*H
i-5»'((QtH^I»jM»-5'^CQM
tn O O O O
.<* 4? -H -H -H
2 2 222
Ef* f* 1^ f*
64 fr* b K4
^ B fc .
•» m m n TT
XX XXX
VO
« « « « »
n M o* c* «
1 * I 1 1
IP **• r* p* GO
0 O O
O O O 1 1
to oo fl» oo oo
1 1 1 1 f 1 1 1 II
1 1 1 1 1 I 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 i 1 t 1 1 1 II
1 1 1 1 1 1 1 1 II
1 1 1 1 1 1 1 1 II
! ! ! 1 I ! i ! ! !
1 1 1 1 1 I t 1 II
aindt.— <.in.r-i.rMtn >Hin
|o§o§g£f So
<* O
sISIBIs! si
31
r-t m
co ob oo flH cK
-------
I
I
1
Measur
ntinued
n o
H
•a
iiil
&L
33jl
° « *
Total
Partic.
S/>:=al
(«/X8tu)
si!
4-
41
8|!
8" *
^ 0 O,
"I"
*
C -4
11 8 11
:c«;
O
3 *
H?I
fa!
** f
ill
ft
h
P
HI
Jll
M M
31
II
*4«
ii *
n»
O
1 • III
IN
rH i in in in
00 1 CO CO O)
o m
rH f>
III! ^ ! i ! 1
(III ""'lilt
0 CD
co r-
10 m
l l t i °, °. i i t i
l i i i w i i i i
i i i i i i i i
i t > i r- *? i i i i
1 1 1 I rH in 1 1 t 1
till . •— i i i i
1 1 1 1 Cn CM I 1 1 1
1 1 1 1 rH *D t 1 1 1
1 1 1 1 •—'1114
co in c*
\o r- t i i i l i o —
<* m i t i i i i o n
. — i i i t i i « —
S in „ £ P £ £ «
j - — • - * j
M • ...
• rH M tn «n
Bpsi SsssaS
2§SK SSSSSS
(M U3 O
s^sp s?s?s?
co ^-. co m • m in tn •
^sllss^ss
M3fflrH^«rH3<{nrH
O m CM (N CM
B, 1 i
J -.-( *J 4J -H
oi « « "i <
fp ** £ fc! S*
&. b s a s
VO in r-| ,H ^
o o *o tc m
i S in" S S
Tt n N 04 N
1Q S 222
*O iH in fM rH
II t t 1
Ch rH oi CM n
^v in in m in
i i t i l
^" \0 u*> VD 00
CO CO 00 CO 00
n n £ r^ in
OI rH CN rH
0.0.\t°.0.lltt
w 1 1 v 1 1 1 1
II till
Oi rH M rH
' « 1 t ^1 1 1 1 1
II (111
m CM i i i i i i i i
rH V0 1 1 1 1 1 1 1 1
• — 1 1 1 1 1 1 1 1
rH — .
o* m i i i i ( i i i
CM \D t 1 1 1 1 1 I 1
• — 1 1 1 1 1 1 1 t
o -". r-
rH CM O ^r
o o
i i i i °. ° l i i i
iiii « l i i i
i i i i till
i i i i i i i i i i
i i i i i i i i i i
i t i i i i i i t i
i i i i i i i i i i
i i i i i i i t l i
t i i i i i i i i i
in ,~ o — r- o
1 1 1 | (N CO (N *£) CO r-
1 t 1 | O rH O «H (*) rH
111! . ~ . ~ • ~
ocSocjooo^oo
o ^ CN in oj «
<* O ^ O nd
rtHpaHrtrHxSSS
§ o. o> o.
< in 2 QC K
c & s £ e
vo *o m »n m
^H rH H
1 1 1
CM CM CM
r* r- r*
tn in rH rH rH
CM fM X X X,
X X. CM CM CM
U> ^3 rH rH rH
n « « . o>
CM CM m m m
H i-t (N rH in
VD 1-^ CO CO CO
in in m m m
i i i l l
t in i i l
1 GO 1 1 i
rH m
1 1 ° °. 1 1 1 1 1 I
II * * 1 1 1 1 1 1
11 ^ t 1 I 1 1 1
VD rH
O CO
in CM
i l ° ° l l 1 l l i
it ' J. i l l 1 l l
it i i i i i i
^r co
i i r- •* i i i i i i
1 1 ^T rH 1 1 1 t 1 1
ii • *-- i i i i i i
i i co m i i i i i i
1 1 «? rH 1 1 1 1 t 1
II * — 1 I 1 1 1 1
f-HCNOCNOOjOeNOCTi
CO O
°? 0 —.11 ~ ~
(NfMOOl IOOOO
A A^ I'
Oi ^ M CM
. o O O n
|552sis|s5
ritnc * ? (0 S U
S'SmrnseSSSSS
O CM CM CM CM
w
a 3 s 3 s
e e e s e
A * 0 * >0
rH
0^ CF>
in •» ^ •* ^
m in m m in
-------
+J
j
t 8?
* S g
'5 *3
a
3
b
lii t
i u 5
ill!
Total
Pirtic.
9/Kc«i
(l/MBtu)
41
• P
g|l.
8 "?I
"
8*'
a1I
S
Mh
0 »-
V I t i i
i i i i in i i i i
I i i i ,-1 — i i i i
CM *-»
at \c i
i i i i r m i i i i
I i i i • in i. i i i
I i i i -i — t i i t
O— . «) .-^ ^ .-* (N —.
rscoi i ui o in o m r-
o*
«*> o H .*•> o .— . o -^
^/J^SoSoooM
VO C' 0
O • **' M •
»H CO f-l H f»-
r,^««s?aa=?Ps;
o m t i i i i t i i
(N O» ( 1 1 1 1 t 1 I
. — i i
sSssHssSs
CD
m o. en •
,4 en CM -w 3
w^cs^m^-oo^co—.
i
s 1 5 s J H 33 $ 3
, 8 § 8
^i f> fi
S 2 S 2 £
» 1 i § S
S It 41 4) 4)
K UI A C/l (/I
E S sS S S
»O CN N W M
3 < S S S
a * 3 H s
*"*
„ o, » . *
<•> H (N r>* C4
H m
1 1 1 1 1
MJ (0 U3 VD SO
O
1 ' H ' '
oo en I ^f m
r- r- 1 co co
*2 £ " t* P
88 £g
•HO oo
..*'.• • • t i
t ! *" ! 1 *— i i
CM VD in CN
r-4 O O O
II ' -^ 1 J * v i i
It II t
I 1 I 1 11
CD m VD x-.
i I ^ vo 1 i r- \o i (
It •Til Of CD 1 1
CN *C H -x
i i in r> i i o ^ i i
li • *t i \ M en t i
1 1 H ^ t 1 • — 1 1
•^- \D 00 CO —.
m 3
"£i0S0£ °S°£
'J1 in
i I i i co CM in en n r-
1 1 1 1 (N >H ^ rH vO in xf
co «->> CD *--. co r* co r* co
>*™S£*SS ?
IssSo^sS^KS
§ 8
M o> 2 £2
0. 0. * "• "
§ § § 1
41 « M « V
ui to < en ui
£ S C S S
CM M ,« n ci
C* *
U3 ^ \D *
1 1 1 t |
GO co co co r^
1 O O 1 1 1 ) 1 1
1 . . 1 1 1 1 1 1
1 *- 1 1 1 1 1 1
o cr» o m
r- m i~< o
o o o o
1 • • 1 1 1 1 • •
i •— i i i e •— -
I i i t t
m ** CM o
1 1 <*1 0 1 1 ! 1 rH •
II • — 1 1 1 1 • <—
VO fH <*> T
1 1 in fH 1 t 1 1 rH •
1 1 m fH I t 1 I O *!•
II • ^ 1 1 1 1 • *-•
r- .-s
«* CO 1 1 1 1 t 1 1 i
o < •
VO fH rH rH C^
S5oPg5Ss,cos
CO CN 1 1 t 1 1 1 1 1
CN __ — , 1
T co co m
K> m •*
.,« o, m „ *
v£ > >H rH iH fH
in ' ' ' '
vo y> ya >o vo
( i t i t i
CO O* CD O 1? ]
r* r^- t^ co r- '
I I t I i I I I i
I i I I > t I i i i
i i i i t i I i t i
t
i i t i i i i t i f '
t I i I i i i l i *
i i I i i i 1 i i i
i i i t in r> t i f i '
i I i i • *r i i i i
j i i t .-t *-" i i i i
*n in
1 1 1 1 UI CO 1 1 1 I
i t i i < ^ i i i i
1 1 1 1 "H — 1 1 1 1
J J 1 J I t i f J i
1 1 t 1 1 1 t 1 I 1
i : i i i i i i i i
1
OOOOIOOOiCJ ^
1
O o M ^
!
t 1 1 1 1 t 1 1 1 i
i i i i i i i i f r
i i i j i i i i i i j
.H _,- „ S s S^ 3 'c;
*0
-------
I
I
I
1
8~
M "O
H P 4)
I n P
« « c
O -H
* £ *J
3 **
2 sj
*
•H
jfl*s
•isH
•H
J! 3 3
•0 J33
a -j s a
"a *i
• .n IT
mi
f\i
*H
Jks
f>
B?l
o*
P 88.
°<£
8*'
— *
C •"• — -
^.-s
1-37
2uII
Q »~"
a:
V
C -1
lm
5 &
8o~°
u *
_ ~
*u*
».
pi
3 s
P
U.
M
"
i b •
es?
Si&
_S3*
M H
• •
n
i*
°j-.
its
ll«
i -
•is
iii ii
rt o >A» » i
o> 05 r- t »
l 1 I 1 1 1 1 1 1 1
I I l l 1 l 1 t l l
1 1 l l 1 l 1 l i l
tl 1 1 I 1 1 1 | l
II l l i l l I i I
(I till I t l t
1 l 1 l l l 1 l l I
1 1 1 1 1 1 III!
1 1 1 1 1 1 lilt
1 1 1 1 1 1 1 1 1 1
t 1 1 1 1 1 | 1 1 1
1 1 1 1 1 t 1 1 1 1
in — .
1 1 1 1 tf> O (lit
1 l i l o r- i i i l
1 i i l • *~ i i i l
n .H -^ o u">
_<-4x-. -^ nco^r^
ooocooo ^fcocr>^
CTI ^r (N
(N •M f-t .
cocotNr-tH-^ r~^» ^
^•^DtJiCDroO O>^> C^
conosnffi^r (NO co
l l t i l l i 1 t i
1 l i i I l i i l i
i)iiii itii
— . r^ x-, m ^.
n,Hor-oy3 TT--N.CO--.
o^nonofn ^COCO\D
o^y .^r -TT (NO^rMoo
^j- o »o CD o
»*> m irt m n
. (N • (N > rH «r>.f*>
m ^, ID ^, m *— Lrt<-.tn^.
* m • m * t/t . in • m
(J^VO^vOO^^D (J> VD O> v£)
V
•U |
» V «
t ^^ui^tn-H m c > n
Q -H * Q * Q &-t
^nCfflOfflo^in-H^ri:
O O O
000
fi n f^
Ul W UJ O O
0- D- 0* Z Z
§s e
« 3 t?i 01
« 4) 0) C C
W W 01 K «
K S K £ 12
o o o r» r-
XXX XX
o* o* o* ^
rt M (N M f*
^T -H
GO €0 00 ^ ft
3 vo is us <»
i i i i t
r- ot «* ">
/-. A) r* i -
n ?
r- t-
vo m i l t i i i l i
O O 1 1 1 1 1 1 1 1
• • 1 1 t t 1 1 t 1
«0 M
r- M
r^ ^j i i i 1 i i i t
°^l i ! ! ! ! i !
€0 M
1- \0 1 1 1 ! 1 1 I 1
• «T 1 1 1 1 1 1 I i
f« •— 1 1 1 1 1 1 1 I
tn «
H (N
m r~ l l l i i i i i
• T 1 1 l 1 1 I t 1
•H «~ t 1 1 1 1 1 t 1
1 1 1 1 1 1 1 1 1 1
1 t 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 I I t ) 1 1 1 1 J
t 1 1 1 1 1 1 1 1 t
** \D o co *r
O r-* O <-H O
o<-tr^MOTr-r^o>«i>
^.(M.HrHt^mQOCQCD
rK(vi^-ror~-cnrs-ncD'Y^
l t l l t 1 l 1 l l
i i i 1 i i i i i i
I I i 1 i i 1 i I i
vDr-f*>r^cri(HCNr^'yiaOf*l(~--ri*»CTiT
•q1 fH (N \D ^D
vo m r- »» u?
• O »O «00 *O «O
coin eoinajtncoincoiri
. c* • (N . n . (N • v
w c/i in w co
s s g s e
tf> m 1/1 m »/i
X X X X X
_| ,H H -H •-(
0, « 0. 0. ' «
01 M (N « (N
1 o
A 4 i i i
i i i i i
I ! 1 1 1
O O^
N 00
i i i i r- (N i i i i
t i i t r- T t i i i
i i i i • j. ' i ' i
o m
r*- n
i t I i ^H m i i i t
1 1 1 1 CD "* 1 1 1 1
lilt • ^ i i i i
t i i i i i i i i i
i i i i t i i i i i
ill iiititi
i i i i i i i i i t
lit i t i i i i l
i i i i i i i i l i
i i i i i l i i l i
i i i i i i i i i t
t i i i i i i i i i
ooo 00000*00
•H O * «« t-
ro^,o— .in^-o-^
tn^ocnr^i^f^mr-.— *
m^cO'q'^Dr-^or-t^
• \r> -r* •* • co i-i r-
OD^-v(N— .in,-**^-
r*ooffi^r-r-»Hco<^^s
*D UD *r- *vo *co>---^t-^(N— • .—
^(^.vD--.^--!^--,
ojyj.-ir-oo'JDOT'--.
r-rHC^Oir^tHO^DOi'^1
. r^ • t- • r- . oo rH co
tn r- T •* M
*n n PO m n
O O O CO CTi «"^
mrsifsomfNinoc^'-i
o o o o o — .
(NOtNOCNOMOffim
CO*TOO'<1'CD't QA1-H-H O «) -H
CQ^tXiJCQ'-IXHOQ'H
oouoooou
S* *> ^ «f *p *• *
OOOOOOOtN
mininmmr'fnr-^fc
1 O 1 0) 1 « 1 O
UCOCOCOCM
65 o5 &5 iJtH^
S S £ & t
f« ri C4 N >
«o CD ^r
^-1
O M m in i i i ( i t
m CN m (N i i i i i i
, ^ * ^ i i i i i i
f-O vOi-'~4^'3>f*lO
O(N QO»-tnO'J'O(T>00
inc^ ^TMTCNxroir'liH
inm ^DfHnf^^inoor*
C^r-1 COfH^OiH^OtHOOO
*TCN ^TtN^-f^frNii^
*n
m « o m oo
^ i-* vo m -- in
_ j_^
• O •<*! •COf«4O4««*l
OO> m^fHt-IOCNC^fH
^ O O in »n m
»rH ••-tINMMNfMr'*
(J»fH Ort-HOOIOeNO(N
i »a "2 t
0) „ 2 T3«« TJ *O
"eS*^1"^ «>*
"J^ijQQ*8^-^ QQ Q
fflME^^*1^31^^^
<4 (4
m u m u O O O
*= D£ 4t 35 Z 2 Z
V A v
S 6 «H «H rH
13 (3 M N M
0
-------
!
I
I
I
t
I
i
i
i
!
\
•I1
« "£
O X 0.
«""
~&
6" -
;^_
3 SI
g « ?, s
•3 "
u
«
C H «
0^16.
p~
g.1 S 6
"^1
: t
S oT^
s -
•III
f 4 "5
pi
+* •
ss
*> V-I
1!
u
i|
Ig
HI
Sx5
S«&
U H
• f
H
s?
i>
ils
•3-3.S
_.
3 6 0
«is
i
in
CO
! 1
! !
t i
i i
> 1
1 i
t l
1 1
1 i
[ i
t 1
r-
o ,-,
0 00
1 I
1 1
o
o
rH
: i
' "
O PJ
n H
35
n. ~
^
-T
0 "1
. CO
CO —
in
M (N
Ss
§
«
r-*
H
! o
! Z
i
^
r*
in
*
i
m
r*-
•
a
I t
i !
! 1
1 E
1
i
i
1
f
1
n
r(
O
(0
(0*
i
to
I")
>D
j
<~*
'D
3
3
I
H
D»
S
3
z
V
-1
N
N
0
Z
$
*T
O4
m
>
o*
U)
&
r
t
in
CO
i 'i
i i
i i
tH \£>
CTi O
rH rH
O O
i t i
i 1 J
1 1 t
1 1 i
1 1 1
1 1 1
-. •»?
1 ' 1
1 i 1
1 J 1
1j [-V r-
r-» VO n
^ Ch o
as .4 t-t J
O O
Z S;
t) O
N 7l
N M
O O
3 z;
£ £
O O
>
CS l"4
0
rl r^
l^ r*
i t i
in ,0 in
CO CO 00
O «£»
IO »H
P*l m
II • -It
I 1 (N .-1 I i
II " t i
CO CO
T f)
11 - • i .
1 ! (N *-\ t \
1 1 — 1 1
*?
Tj- rH
1 1 n CO 1 1
It • rH 1 i
1 t U3 — ' 1 i
CO "T
r- N > N
O 3 * 3 V<
j z c- 4) o. a)
i S S £
I
' O O O
j 04 M (N
M O c-,
£ * £
{N rM CN
C'l
! 1 <
r- co co
, r- r- r^
I I
Mi
CC)
I t i i
! ! ! i
ill!
Ill)
rill
1 i > t
1 1 1 1
.._..
! » '. J
i i i *
n «.,
O — C' -—
O ""I O
o o u o
CO C i
i— r I
f ! 1
t i 1
ill
^J „-, 0 -^
r- o . -» M-
. L/l 0 ^T-
rH — . . f .-^
TJ* CTi ( -J CTi
^a- 10
r~- -ar
mo •"*
. en O ^£'
0 0
in (N iD CN
>*.
333^
n l-f
«i *
0 O
vj '-»
i N r* N
3 V< " lH
o. -r
1 33 rH
1
£
, en
! '2
i£
1
i
!-
i
*
i ^
' Ot
1
1 a
1 o
1 "~*
H
i
J at
i i
*n CD
"• r^
1 ! I '
1 1 1
1 1 1 ;
lit1
till
t 1 i !
1 i 1 !
Ill'
111.
i 1 ? I
lilt
i t : J
! 1 f »
1 1 1 1
o c> o o"
V CN
co' co
co co -n CN
m .H f\j o
O* rH Oi -1
t 1 » (
1-, r>J n-' O
m
cr\ o a* o
« r-i . rH
T3 T3
fl > («
S 3 S 3
CJ O
*s s:
51 IT*
"a «
5 S
\t> LI
Vf) tC
IN" cv
\t> -o
o o
f)
i i
0 Cj
CD 0)
1
^>
i 1
1 1
1 i
1 1
t 1
t i
j i
; i
i i
f~ CTl
•a- o
o
rH
UJ 5"
m m
i-1 rH
: t
IT, -^
o
r^i
T ^~.
0 ffi
Ch O
• rH
CT; r-s
a M
33
1
cn
"2
§
*/l
!-O
\
CM
«
O
O^
1
o
OD
]
j
i
j
:
j
i
S
s
i
!
I
1
i
i
1
1
i
i
j
!
'
|
1
i
1
j
>,
i
i
!
j
!
|
1
-------
I
I
The columns to the right of the one labeled "Test Load" are data
taken during the corresponding Test Run.
For almost all boilers four basic types of measurements were made:
1. Baseline: ~80% of rated capacity and normal control settings.
2. High Load: Highest load obtainable at the time on the unit under
test.
3. Low Load: Minimum load at which unit normally is operated.
4. Low Air: Minimum excess air level at baseline load at which the
boiler could be operated without smoke, excessive car-bon monoxide,
or hydrocarbon emissions.
When a boiler had two or more burners, often a test was run with
the fuel to one of the burners turned off and only air passing through
the burner and into the furnace. The air-only burner then was acting
like an overfire air port. This type of test was designated by "BOOS" for
burners-out-of-service. The test type designation "Register" indicates a
test that investigated the effect on the emissions of increasing or decreas-
ing the air swirl by changing the register setting.
The column titled Test Fuel indicates the fuel being fired at the
time of the test run. When more than one fuel type was being burned, e.g.,
Test Run 23, the entry so indicates, and the reader is referred to Section
4.5 for details.
In the balance of this section, the nitrogen oxides and particulate
emissions measured for each of the three fuels are discussed in detail.
In addition, mixed fuels, emissions other than NOx, and particle size
distribution are discussed.
59
-------
4.1 COAL FUEL
4.1.1 Nitrogen Oxide Emissions
The analyses of the coals tested during this prccrara art listed on
page 107. The moisture contents of the coals varied substantially between
the d-^ferent samples, as did the heatinc values as fired. However , the
variations in the nitrogen content, of the coals analyzed were jniail, varying
from 1.29 to 1.80% by weight. The baseline nitrogen ox;.ce emissions for
these coals are presented in Figure 4-1 as a function of boiler test lead.
Although the data, as shown, indicate that the baseline MGx emissions
increase with increasing boiler size, other parameters uxscuss^d below
probably are contributing to this effect.
The lowest nitrogen oxide emissions were measured with boilers using
•underfed stoker coal burning equipment. These boiler designs also happen
to be of small capacity, less than 50 k Ibs/hr steair. flow, and have a
large furnace volume per unit heat release, S4 ft~VMBtu/hr. The iniodafc-
size boilers, 50 to 400 k Ibs/hr steam f.ow, all nad a lower furnace heat
release volume, 28 to 64 ft; /MBtu/hr, and larger NOx em.LSsians, These boilers
used either pulverized coal burners or spreader stoker coal burning equip-
ment. Two of the cunts with pulverized coal burners, T>;kts 26 ar.d
"8,
corner-fired Combustion Engineering brand boilers. The largest l\Ox er,,.L&si;/r. t,
(Test 32} were measured with & uniLt using two cyclone-type coal coi:.bu::., t^r;.,,,
which have a reputation for being large NOx producers. This a.i so happened
to be one of the largest coal-fired noilers tested, and it nad ^r. extieraely
low furnace heat release volume of only 2 ft /XBtu/hr. Here r.h.-i furnace
volurrie is defined as the volume of the cyclone ccnbustors, since the
combustion reactions are mostly completed before the not gases enter the
boiler furnace for steam generation. Sections 7.1 and 7.2 discuss further
the effects of these bciler design characteristi.es on NO.'. 3rai.s,aacns levels.
The NOx eru.ssi.cns for coal-firec watertube boiZerc cecreased with
decreasing excess oxygen by approximately 50 ppm per on3 percent reduction in
oxygen for each coal test conducted. The NCx ernjuu-'iom for natural gas ana
oil fuel tests were not as sensitive to O0 level ac fc.: coal. Test 1^0., 28
was conducted with a spreader stoker-fired boiler, wnich also had pores for
60
-------
I
I
800
700
a
a,
o
dC
600
•O 500
c
o
•rH
-P
-P
C
0)
O
a
o
u
0)
•a
•H
x
o
c
OJ
400
300
200
4-1
O
100
Cyclone
Pulverized and
Spreader Stokers /
100 200 300 400
Test Load, k Ib/hr of Steam
—I 1 1
50 100 150
Test Load, kkg/hr of Steam
200
500
250
Figure 4-1. Total nitrogen oxide emissions at baseload for coal-
fired boilers.
6000-28
61
-------
oil burners. The oil burners had not been installed, but the air control
mechanisms were operational. These burner ports, when used as overfire
air ports, lowered the NOx emission by 46%. Diverting combustion air
from the grates increeised the grate temperature beyond its allowable
temperature range, and a compromise was necessary between low nitrogen oxides
emissions and acceptable grate temperature. The most acceptable boiler
operating conditions 3:educed the NOx emissions by 20 to 25% over the entire
load range with this mode of firing.
Air register tests were not too effective in lowering NOx emissions
with coal-fired boilers. Combustion air preheat did not strongly affect
the NOx emissions when comparisons are made with other parameters constant.
The solid/gas reaction mechanisms for coal fuel combustion may be providing
sufficient time for furnace gases to penetrate the flame zone, thus eliminating
the effects of air preheat temperature. The effects of operating parameters
on NOx emissions for coal-fired boilers are discussed in detail in Section 5.
4.1.2 Particulate Emissions
Particulate emissions from coal-fired boilers were measured for units
with different coal burning equipment and with coals of varying amounts of
ash. For most units the particulates were measured downstream of dust col-
lectors; however, in some of the tests particulate emissions were measured
upstream of dust collectors. Figure 4-2 presents the total particulate
concentrations at base load for each coal fuel test.
Spreader stokars, underfed stokers and pulverized coal burners all
had similar particulate emission levels. Typically the total particulates
were between 0.5 and 3.0 Ibs/MBtu. Test No. 32 was a cyclone-fired unit
for which the particulates were measured before the dust collectors, rather
than after as with all other coal fuel particulate measurements. The level
of 1.2 Ibs/MBtu was no higher than that of the other units that were mea-
sured after a dust collector. Test No. 31 was a pulverized coal-fired unit
which often burns tree bark in addition to oil or coal fuels. The particu-
late levels for this unit were extremely high (10.1 Ibs/MBtu), perhaps due
to residual bark particles in the flue gas.
62
-------
I
en
O
•rH
4-)
to
u
c
O
u
(1)
4-1
u
•H
4J
(tf
4-1
O
5.0
2.0
L
1.0
0.5
0.2
,A
/9\
r
A
A
A
M,
Numerals
are test
A
• LM
within the
lumbers .
A
tU
symbols
100 200 300
Test Load, k Ib/hr
400
500
Figure 4-2.
Total Particulate Emissions at Base Load for
Coal-Fired Boilers. All Measurements Were Made
Downstream of a Dust Collector Except for Test
No. 32, for Which Measurements Were Made Upstream
of a Dust Collector.
6000-28
63
-------
4.2 OIL FUEL
The oil fuels tested during this program included No's. 2, 5 and 6
type oils, the properties of which are summarized in Table 6-1. The
nitrogen content of the oils varied from essentially zero to greater than
0.5%. Other properties, such as API gravity, viscosity, heating value,
ash content, volatility and Conradson carbon also varied over the normal
ranges for oil fuels. The burners included steam, air, rotary cup and
pressure atomization. Over three fourths of the tests conducted with No.
6 oils were steam atomized, with the remainder being air atomized. One
half of the No. 5 oil tests were air atomized, with the remainder being
evenly divided between steam and rotary cup atomizers. The No. 2 oil tests
were evenly divided between steam and air atomizers, with one test using a
pressure atomizer.
4.2.1 Nitrogen Oxide Emissions
The baseline nitrogen oxides emissions from oil fuel are plotted
in Figure 4-3 as a function of boiler test load. The firetube boiler
data, shown as a cross-hatched area due to large number (14) of data points,
were insensitive to boiler load, combustion air temperature and excess
oxygen level and were between 100 and 300 ppm. The watertube boiler NOx
emission data shown in Figure 4-3 did not vary significantly with test
load but were dependent upon fuel nitrogen content, furnace heat release
volume, burner siue and excess oxygen level.
No. 2 oil total nitrogen oxides emissions data were found to be
insensitive to excess oxygen level for both preheated and ambient-tempera-
ture combustion air. The nitrogen oxides emissions for watertube boilers
burning No. 5 and 6 oils decreased with decreasing excess oxygen for all
64
-------
-
ri!
•H ft
cO 3
0 U
-H
C >i
fti H
V M
-P -H
CO
0).
rH
-H
0
m
- ^.
SN^S:^^N
to
fsj
I
0
O fo
00
-------
tests except one (Test No. 1), for which a peak in the nitrogen oxides
emissions occurred at about 5% excess oxygen. Below this 5% excess oxygen
level, the nitrogen oxides emissions decreased with decreasing excess oxygen.
Burner heat release rate and furnace heat release volume both were found
to influence nitrogen oxides emissions level. The larger burners, 80 to
125 MBtu/hr, had nitrogen oxides emissions greater than 300 ppm. The smaller
burners, 10 to 50 MBtu/hr, had nitrogen oxides emissions varying between
60 and 350 ppm and were dependent upon the nitrogen content in the fuel
(see Section 7.1.4 for further discussion). Furnace heat release volume
defined as the furnace volume divided by the boiler capacity (ft /MBtu/hr),
affected nitrogen oxides emissions for watertube boilers with No. 2, 5, and
6 oils. The larger furnaces, in terms of ft /MBtu/hr, had the lower nitrogen
oxides emissions (see Section 7.2.2 for further discussions).
Off-stoichiometric or staged combustion tests , done by turning off
the burner fuel while leaving the air registers ppen, were successfully
conducted with No's. 5 and 6 oils. The nitrogen oxides emissions were
reduced 6 to 25% for No. 5 oils and 12 to 29% for No. 6 oils. The boiler
firing rate, defined as the percent of boiler capacity, had little effect
on nitrogen oxides emissions for oil-fired boilers. Air registers or dampers
were found to have: a minor effect on nitrogen oxides emissions, but were
most helpful when testing with a burner out of service in reducing the
excess oxygen level at which the boiler could be operated without smoking.
The lower operating 0_ levels resulted in lower NOx emissions.
The temperature of the fuel oil at the burner had a direct effect
on atomizer performance. All No. 2 oils tested during this program were
fired at ambient temperature. The No. 5 oils were nearly all fired at an
elevated temperature between 160 and 180°F. The No. 6 oils were fired at
a temperature between 180 and 250°F with most tests conducted at approximately
200°F. One test with a rotary cup atomizer was fired at 127°F. Two tests
that had the highest NOx emissions were conducted with boilers burning
No. 5 fuel oils at only 130°F temperature at the burner (Tests 63 and 68)
and as large as 40 psi differential pressure between the steam and oil
66
-------
for atomization. Another test (Test No. 70) was conducted with ambient
temperature No. 5 oil and steam pressure 45 psi greater than the oil
pressure at the burner. These three tests conducted with lower than normal
oil temperatures required high (40 to 45 psi) differential steam to oil
pressures at the burners for satisfactory atomization and all three had
high nitrogen oxides emissions.
A test series (Test No. 34) was conducted during which the fuel
oil temperature was decreased from its normal operating temperature. The
nitrogen oxides emissions increased from 300 to 316 ppm as the oil tempera-
ture at the burner was reduced from 250 to 200°F. NOx emission increases
have been observed in some other KVB field test programs when fuel oil temperature
was reduced below normal levels. This effect has been attributed to
changes in viscosity producing larger droplets.
4.2.2 Particulate Emissions
Particulate emissions for oil-fired boilers were measured for No's.
2, 5 and 6 grades of oil and for air, mechanical and steam atomization.
As Figure 4-4 shows, higher particulate emissions result from burning
the heavier oils, and the effect of the method of atomization used depended
upon the type of oil being fired. Particulate emissions are less influenced
by boiler size than thay are by the fuel and burner characteristics.
The particulate concentrations presented in Figure 4-4 show air
atomization of No. 6 oil (Test 34) and No. 5 oil (Tests 35, 44 and 46)
to have higher particulate emissions than for steam atomization of No. 6
and 5 oils. Test number 34 was a special test conducted for this program.
This boiler does not normally burn No. 6 fuel oil. Apparently, steam
atomized the No. 5 and 6 oils better than did air atomization, and a
smaller weight of particulate was formed with steam atomization.
Atomization method effects on particulate emissions are discussed
in detail in subsection 7.1 on burner design.
67
-------
1.00
S. 0.100
(0
o
•H
-P
10
C
0)
u
0
u
0)
-p
3
o
•H
4J
S. o.oio
o
EH
0.001
^
34
-S.
"5ivi
-$c
^^
59
IT-
•<— r— I
^
68
D
12
45 '
>
*
5)
66
«
1
©
H— /'"V-
v§/~
8
V5J" '
63
^3
©
9
(£)
2
1
c
©.-
7
D..
80
5)
'0
©
65
1 ! I
1 1
O
O
O
Numl
ref
Num.
ref
<§>
I
c
Natural Gas
Oil-Steam Atomized
Oil-Air Atomized
> Oil-Mechanical Ato
Oil -Rotary Cup Ate
aers within symbols
sr to oil grade.
bers below symbol
er to test number.
u
mized
mized
I
I
I
I
I
40 80 120 160
Test Load, k Ib/hr
200
Figure 4-4.
Total particulate emissions at base load for
natural gas and oil-fired boilers.
6000-28
68
-------
4.3 NATURAL GAS FUEL
4.3.1 Nitrogen Oxides Emissions
The natural gas burners tested during this program were nearly all
ring burners. The only exceptions were the two corner-fired boilers used
for Tests 75 and 77, that had natural gas nozzles which could be tilted for
steam temperature control. The natural gas ring burners operated at varying
pressure levels depending upon gas pressure available and burner manufacturers.
The NOx emissions for natural gas-fired boilers were found to be
dependent in varying degrees upon furnace type, excess oxygen level, combustion
air preheat temperature, burner size and firing rate. The baseline NOx
emissions are presented in Figure 4-5. A large number of small firetube
boilers, 7 to 20 k Ibs/hr steam flow, were tested, and each individual test
point could not be shown on Figure 4-5, since many boilers had practically
the same concentration. These 10 tests are represented by the Crosshatch.
The natural gas -fired firetube boilers all had baseline NOx emissions
between 50 and 100 ppm and showed very little dependence of nitrogen oxides
on excess oxygen level (Section 5.1). The natural gas fired watertube boilers
had varying amounts of dependence of NOx on excess oxygen level. The boilers
with preheated combustion air typically showed more of a decrease in NOx emissions
with decreasing excess oxygen than did boilers using ambient temperature
combustion air. On two boilers with single large burners of 110 and 160
MBtu/hr heat release rate, the NOx emissions decreased with increasing excess
oxygen.
A comparison of test results from boilers using unheated combustion
air with boilers using preheated air as high as 650°F will find that nitro-
gen oxides emissions usually increase with increasing combustion air tempera-
ture. The amount of increase in nitrogen oxides emissions with an increase in
combustion air temperature appeared to be dependent on burner heat release rate.
With small burners, less than 30 MBtu/hr, an increase of approximately 15 ppm
69
-------
300
200
I
-------
in nitrogen oxides emissions occurred for each 100°F increase in combus-
tion air temperature . For the middle-size burners, 40 to 60 MBtu/hr, this
value averaged 56 ppm/100°F and the large burners, 110 to 160 MBtu/hr,
resulted in approximately 125 ppm/100°F (Section 5.2).
Burner-out-of-service tests conducted with natural gas fuel resulted in
nitrogen oxides reductions from 16 to 42% of baseline. Firing rate affected
nitrogen oxides emissions for boilers using preheated air, but did not have
much effect for boilers with ambient temperature combustion air. The effect
of air register settings on NOx emissions for gas-fired watertube boilers
was found to be minimal for the conventional ring burner. The air registers
for these burners change the amount of swirl in the air flow. The air
dampers on the corner-fired boilers were successfully used to lower the
nitrogen oxides emissions by 25% (Section 5.1.3). These air dampers proportion
the air to different burner compartments rather than add swirl to the air
flow.
4.3.2 Particulate Emissions
Natural gas fired boilers very seldom operate with luminous flames
where the combustion of elemental carbon is occurring and soot or coke
particles are formed by incomplete combustion. The particulate emissions
data taken during this program for natural gas fired boilers and shown oh
Figure 1-2 were low, typically between 0.004 and .007 Ibs/MBtu, as would
be expected for conventional ring burners.
4-4 MIXED FUEL
An additional objective of the Phase I field measurements was to
collect data on the level of nitrogen oxides emitted when a mixture of
fuels was burned. This may materially affect NOx emissions, especially
since waste materials can sometimes be high in organic nitrogen content.
Three sets of measurements were made where a secondary fuel was burned
and the results are listed in Table 4-2.
In Test No. 23 the unit tested burned a mixture of No. 5 fuel oil
and refinery gas. The composition of the refinery gas varied,
71
-------
but on the average it was deemed by the boiler owner to contain about 25%
hydrogen and 30% methane. Even though the excess oxygen level was lower
for Run No. 74-1, the nitrogen oxides concentration was larger by about
45 ppm. Apparently, the "50/50" mixture of oil and refinery gas was a
greater producer cf NOx, This might be explainable if the composition of
the fuels were available as a function of time, since some waste gases
have a high nitrogen content. However, it proved to be impractical to
obtain for analysis an adequate sample of natural gas and refinery gas
fuels that was not diluted by the leakage of air.
When a quantity of wet tree bark, about 20 tons per hour, was fired
with No. 6 oil in Test No. 29-1, the nitrogen oxides concentration increased
by 25 ppm, even though the excess oxygen level had decreased.
Adding 30 to 50% of No. 5 fuel oil to the coal in Test No. 32-2
resulted in a 9% increase of NOx when 30% of oil was fired along with the
coal (Run 72-3). But when the excess oxygen was returned to the baseline
level of 3.4% and the load was reduced to baseline level in Run No. 72-4,
the NOx decreased by 11%. Apparently, at this mixture the nitrogen oxides
Table 4-2. MIXED FUELS
Test
Run
Number
23-1
74-1
29-5
29-1
32-4
32-2
72-3
72-4
71-1
Test
Load
(k#/hr)
88
88
400
400
Burner
Atomization
Method
Steam
Steam
Steam
Steam
32Q Cyclone-
402 Cyclone
409 Cyclone-
320
400
Steam
Cyclone-
Steam
Cyclone-
Steam
Mixed Fuel Type
92% #5 & 8% Refinery Gas
50% #5 & 50% Refinery Gas
#6 Oil
#6 Oil & Wet Bark
Coal
Coal
70% Coal & 30% Oil
70% Coal & 30% Oil
50% Coal & 50% Oil
NOx
(ppm)
172
217
400
425
800
790
860
710
797
X02
(%)
8.0
6.5
9.5
9.0
3.4
3.2
3.6
3.4
3.7
6000-28
72
-------
formation was very sensitive to the amount of excess air being fired.
A 50-50 mixture of coal and oil showed no change in NOx; however,
there was insufficient time available to investigate completely whether
or not it was possible in this latter case to lower the excess oxygen
and thereby lower the NOx, as had been done in Run No. 72-4.
This limited testing of mixed fuels does not provide a good
basis for generalization. However, it appears that the emissions of
total nitrogen oxides may be increased due to the properties of the fuel,
especially if a waste fuel is being burned.
4-5 RATIO OF NO CONCENTRATION TO TOTAL NOx CONCENTRATION
Fifty-seven measurements of the ratio of NO to NOx at base load
were made and are plotted in Figure 4-6. NO values for oil fuel were
typically about 1% to 3% of NOx with a median value of 1.5%. Typical
NO values for coal were about 1% to 6% with a median value of 2.7%. For
gas fuel the typical values were 3% to 13%, and the median of 5.8% was
the highest. A commonly accepted ratio heretofore has been 5%, but this
value would appear to be too high for coal and oil fuels. The variation
in the NO percentage is not unreasonable since the measurement is made
by differences between NOx and NO. For gas testing where total NOx values
are frequently less than 100 ppm, this is especially true.
4.6
CARBON MONOXIDE EMISSIONS
The carbon monoxide (CO) emissions for industrial boilers are
normally near zero, although in a few instances the emissions exceeded
100 ppm. The measured concentrations are listed in Summary Table 4-1.
The presence of over 100 ppm carbon monoxide in the flue gas indicates
either low overall excess O , air/fuel maldistribution, or burner problems.
Oil-fueled boilers typically had no carbon monoxide emissions,
because oil fuels generally are fired with higher excess air/oxygen to
avoid smoke emissions. One exception was the rotary cup atomized firetube
boiler of Test No. 36. It emitted 90 ppm of carbon monoxide with 6.7%
73
-------
I
-------
excess oxygen. A rotary cup-atomized watertube at the same location, Test
No. 3, was fired with more air, 7.6% excess oxygen, and it had no carbon
monoxide emissions.
Spreader stoker-fired boilers tended to emit carbon monoxide
at base load, while pulverized, underfed and cyclone-fired boilers did not.
When the excess oxygen with spreader stokers was below 10%, as in Test Run
No. 20-6, carbon monoxide was present; when the excess oxygen was above 10%,
as in Test Run No. 27-1, no carbon monoxide was measured.
Pulverized-fired units were operating with excess oxygen as low as
5.3% (Test Run No. 26-1) with no carbon monoxide emissions. The cyclone-
fired boiler was run as low as 3.4% excess oxygen (Test Run No. 32-4) with
zero carbon monoxide being measured.
4.7 HYDROCARBON EMISSIONS
Hydrocarbon (HC) emissions measured as methane (CH ) at baseline
conditions with both natural gas and oil fuels were generally in the 0 to
75 ppm range. The two highest baseline values measured were 200 and 575
ppm, and both of these were natural gas-fired firetube units. Ideally, the
hydrocarbon emissions should be near zero, indicating that no unburned fuel
is being lost up the smoke stack.
While the natural gas-fired firetubes were consistently higher in
hydrocarbons, this was not universally true of the oil-fired firetubes. The
single highest baseline measurement with oil fuel was from a firetube boiler,
75 ppm on Test No. 33; but Test No. 59 found about 20 ppm of hydrocarbons,
which was in the same range as the watertubes burning ail. The measured
concentrations for coal fuels were lower, in general, than concentrations
for gas and oil.
There was an indication that natural-gas-fired firetube boilers
tended to emit a greater concentration of hydrocarbon than did watertube
75
-------
furnace type boilers burning natural gas, oil, or coal. Tins higher con-
centration may be caused by the rapid quenching of the products of com-
bustion by the relatively cool walls of the furnace tube.
4.8 PARTICLE SIZE
A limited amount of optical particle size classification was per-
formed and the results are shown in Figures 4-7 through 4-10 and Table 4-3.
The method was to catch the particulate on a heated, Gelman Type A filter,
enlarge and view a portion of the filter with an electron microscope arid
visually count the particles in each size group. Some difficulty was
experienced with the particles smaller than two micrometers (or microns, y)
embedding themselves in the filter matrix where they could not be counted
when using the Millipore MF-AA filter. Therefore, data were not always
available for the 1-2 y size range. The tests for which Table 4-3 shows
an entry in the 1-2 y size range utilized the Millipore MF-DA type filter
with a nominal pore' size of 0.65 y. For the MF-DA filter tests, percentages
are shown in parenthesis for sizes greater than 2 y, and they indicate what
the percentage distribution would have been if the 1-2 y had not been cap-
tured, thus allowing a direct comparison with the MF-AA type filter per-
centages.
The particulates emitted by coal and oil fuels are significantly
different in appearance, as well as in number. This difference is illus-
trated in Figure 4-11, which is a reproduction of an electron microscope
photograph of coal fuel particulate and oil fuel particulate. The coal fly
ash consists mostly of flakes and irregular chunks of material, while the
oil fly ash is mostly spherical. Many of the oil fly ash particles have
holes in them or are hollow.
With natural gas fuel, there were two different types of particulate
size distributions shown in Figure 4-7. For Tests No. 15, 30, and 77, most
of the particulatess were smaller than 6 y; while for Tests No. 14, 24, and
80, the majority of the particles were 6 y or greater in optical size. There
is no ready explanation of the two size groups. All were burning natural gas
from the same area (all but Test No. 77 were within 100 miles of each other),
and there were no anomalies or trends in the corresponding total particulate
catch with the EPA particulate train.
76
-------
Rincr Burner
LOO
40
20
)
1C
) 2C
) 3C
Test
4C
No.
) 5C
15
A
l_y,_...
10
No.
24
•4-1
0 10 20 30 40 50 100
<*° 100
1
0)
°' on
C "0
3 60
0)
•H
13 40
c
•H
I 20
13
U
° 0
§ c
H
TJ
0 100
p
f'l
80
60
40
20
0
)
™^
-
1C
) 2C
) 3C
Test
) 4C
Test 1
No.
) 5C
^Io. 14
30
L
) ^ 1C
-n
)0 C
)
••
—
i
0 2
0 3
Test
0 4
I'est He
No.
0 5
T. 80
77
i
0 ' 1
00
0 10 20 30
40 50 100 0 10
Optical Diameter, pm
20 30 40 50 100
Figure 4-7. Distribution by Percentage of Catch of the Particulate Optical
Diameter. iJatural Gas Fuel.
77
-------
100
80
60
40
20
Steam Atorniiicd
Test
No.
0 10 20 30 40 50
-
-1
)
No.
I
1 )
~^m.
1
1
A
10 20 30 40 50 v
<*> iuu •
1
0)
§ 8°
K
M
(U f,n
§
fC
•rH
Q zin
c
•H
fi
U on
4J ^U
fl
U
M-l
§ o
•H
-P
M i no
o -LUU
&
n
&i
on
/r p,
An
90
n
1C
b=
2C
3C
Test
40
Test
No.
50
No.
22-1
-Ji—
1(
29
,-X-,.
0 0
•
Burr
Test
er-Out
No.
-Of-Se
22-]'
rvic
>l
r
10 20 30 40 50
0 10 20 30 40 50 100
Optical Diameter, ym
Figure 4-8. Distribution by Percentage of Catch of the Particulate Optical
Diameter. No. 6 Oil Fuel.
78
-------
100
80
bO
40
20 .
Cyclone and Pulverizer
•
'
-
•MMH
Tout
Cycl
No.
3n e
$2
1
lil_ , y
10 20 30 40 50 100
100
anae
00
O
M
-------
J.UU
80
60 •
40 .
20 .
I
-i
i
i
[
Test
Sprea>
i
i
i
No.
ler
16
1
|
0 10 20 30 40 50 100
e\°
1 00
0)
c
"5 80
K
M
0)
4-)
3 20
M-<
0
0 °
Tl
••••
Test
Spre.
Ho.
ider
20
;jj 0 10 20 30 40 50 f 100
IH
o
°i i no
o
M
P^
80
60
40
20
n
mm
mm
mtfum
Test
Unde
No.
rfed
17-2
., L
j
Coal
Unr'erfed and Spreader--Stoke r
I
-
Test
I
"
No.
dor
~1
'" 1
•
1
10
10 20 30 40 50 100 0 10
Optical Diameter, ym
20
30
40
50
n
}\_
Test
Unde:
No.
fed
17 -6
J
\
20
30
40
Figure 4-10. Distribution by Percentage of Catch of the Particulate Optical
Diameter. Coal Fuel with Underfed and Spreader-Stoker Burners.
80
-------
Table 4-3
OPTICAL SIZE DISTRIBUTION "OF"CAS, OIL AND COAL FLY ASH
Test Run
Number
Gas Fuel
14
15
24
30
77
80
Burner
Type
Ring
Ring
Ring
Ring
Ring
Ring
No. 6 Oil
8
9
22-1
22-16
29
Coal
17-2
17-6
18
26
31
78
16
19
20
32
71
Steam
Steam
Steam
Steam
Steam
Underfired
Under fired
Pulverizer
Pulverizer
Pulverizer
Pulverizer
Spreader
Spreader
Spreader
Cyclone
Cyclone
Relative Number, %
1-2 2-4
0
77
0
20 45
63
0
64
60
56
66
78 19
52
71 15
96 4
65 32
62 23
23 65
34 31
51
42
70
92
Particle Di
4-6
0
21
24
(56) 16(20)
18
0
27
23
37
23
(86) 2(9)
33
(52) 10(35)
(100) 0
(91) 2(6)
(62) 10(27)
(84) 9(12)
(36) 24(14)
31
40
25
7
ameter,
6-10
0
0
66
11
13
0
7
15
6
9
1
11
3
0
1
3
3
9
15
13
5
1
pm
10-50
94
2
10
(14) 8
5
95
2
3
1
3
(5) 0
1
(10) 1
0
(3) 0
(8) 1
(4) 0
(14) 2
3
5
0
0
•
>50
6
0
0
(10) 0
1 I
5
0
0
0
0
0
0
(3) 0
0
0
(3) 0
0
(3) 0
0
0
0
0
i
6000-28
81
-------
Test 20-1
Coal Fuel
2000X
Spreader Stoker
Test 8-8
#6 Oil Fuel
2000X
Figure 4-11. Electron microscope photographs of fly ash from coal
fuel and from oil fuel caught on filter paper.
Magnification is 2000X.
6000-28
82
-------
The optical particle size distribution for the tests with No. 6
oil fuel depicted in Figure 4-8 all have practically the same distribution
and uniformly favor the largest number in the "fine particulate" size range
of 4 y and smaller. Test No. 22-16 was run with a burner out of service to
investigate the effect on particle size of this type of low-nitrogen oxides
firing. Apparently, burner-out-of-service firing had no major effect on
the size distribution of the particulate material.
The pulverized and cyclone-fired coal fueled boilers also have a
preponderance of particles in the "fine particulate" sizes, 2-4 y (Figure
4-9). The spreader-stoker and underfired stoker units have relatively
greater numbers of particles larger than 4 p (Figure 4-10).
Test No. 71, shown on Figure 4-9, was run to investigate the effects
of a 50% oil-50% coal fuel mixture. Apparently, utilizing oil fuel increased
the proportion of the smallest size particles, 2-4 p.
During Phase II the aerodynamic diameter of particulates also will
be measured using a'cascade impactor.
4.9 BACHARACH SMOKE SPOTS
The Bacharach Smoke Spot measurements in general followed the
carbon monoxide emissions. For example, in Test No. 3 when the test load
was raised from the baseload of 12 k Ib/hr to the high load of 14 k Ib/hr,
the carbon monoxide emissions rose from 0 to 234 ppm and the smoke spot
rose from 5 to 7. As discussed in Section 3.2, this measurement proved
to be difficult to make in the field on a routine basis, and the significance
of the data is questionable.
4.10 BOILER EFFICIENCY
Boiler efficiencies were calculated using the measured fuel and
emission data, and at the base load the efficiencies ranged from 77% to 87%,
depending on the fuel type, excess air level and age of the boiler.
The major factor affecting efficiency was found to be excess O
level. A reduction in excess 0 improved efficiency by about 0.5% per 1%
change in 0 .
83
-------
I
I
The effect on efficiency of combustion modifications to reduce NO
emissions was primarily determined by the excess 0 effect. If 0 could •
^ ^ ^^
be lowered along with changing air registers or taking burners out of
service, efficiency was improved. If an operational change resulted in a I
higher 0 requirement, efficiency was degraded. In either case, the
magnitude of the change was typically in the 1 to 3% range. fl
With coal fuel, cyclone-type burners had the highest efficiencies.
The lowest efficiencies were with underfed stokers due,, in part, to their
being older than the other coal-fired boilers. Older boilers tended
to have lower efficiency because of higher stack temperatures and lack of
efficiency-enhancing design features, such as economizers and/or air
preheaters. The larger capacity boilers were more efficient than smaller
boilers probably for the same reason. The median baseline boiler efficiency
for units of 50 k#/hr output was about 79%, while the median for units
around 400 k#/hr output was about 83%.
Boiler efficiency is discussed in more detail in Section 7.3.
84
-------
SECTION 5.0
EFFECT OF OPERATIONAL CHANGES
Changes in boiler operational methods and firing practices were
evaluated to establish the effectiveness of these changes in controlling NOx
formation. Effective parameters were found to be the air/fuel mixture
ratio control, air preheat temperature, fuel oil temperature, and firing
rate. Mixture ratio control, which proved most effective, could be
achieved by changing excess oxygen, by taking burners out of service, and
by air register adjustments. Some of these control methods may not be
applicable to a given industrial boiler since many industrial units have only
one burner, fixed air registers, no air preheat, no preheat control, limited
oil temperature control, or severe load constraints. Excess oxygen is
controllable on all units and one or more other methods should be applicable
to a typical unit.
5.1 MIXTURE RATIO CONTROL
The air/fuel mixture ratio at the burner is one of the most important
variables to be considered due to its effect on both flame temperature and
the concentration of oxygen atoms available for NOx formation. Theoretically,
there is a peak in the flame temperature when the fuel/air mixture ratio
is slightly air rich. Measured NOx concentrations have a similar trend for
premixed or well-mixed flames. This peak rarely has been observed in large
utility boilers, apparently due to slower mixing, and for utility boilers
reducing excess oxygen reduces nitrogen oxides formation. However, peak NOx
values have sometimes been observed by KVB while testing smaller utility
boilers.
5.1.1 Excess Oxygen/Air
The test results showed a definite difference between the effect
of excess oxygen level on NOx emissions for firetube and watertube boilers.
Figure 5-1 presents all of the firetube boiler data. NOx emissions were
85
-------
I
O
0)
c
to
1X3
rt)
en
rH
•H
o;
3
+j
•H
4-1
o
•H
CO •
CO 0)
•H r-\
&
Cn en
O nJ
•H rH
C (0
C C
0)
Cn -
>i en
§§
to -H
Q) -a
u C
X O
0) O
M-I en
O C
•H
-P 4->
u aJ
d) H
IM a)
>w 04
W O
in
-------
found to be relatively insensitive to excess oxygen level for natural gas,
No. 2 oil and No. 5 oil fuels. Test No. 34 conducted with No. 6 oil fuel
did show some dependency of NOx emissions on excess oxygen. However, the
test results are not typical, because the No. 6 oil was run as a special
test fuel for this program using a boiler and atomizer designed for other
fuels.
The effects of excess oxygen on NOx emissions for watertube boilers
are presented in Figure 5-2 . The data for coal and oil fuels show the
typical reduction of NOx emissions with decreasing excess oxygen. The coal
data show relatively large effects of O level on NOx emissions with the
average being approximately 50 ppm change for each one percent change of
excess oxygen.
The oil data for both heated combustion air and ambient temperature
combustion air indicate that the NOx emissions for No. 2 oil are only
slightly affected by excess 0 level and average about 10 ppm change for
each one percent change of excess 0 . The data for the No's. 5 and 6
fuel oils with both preheated and ambient air show larger effects of excess
oxygen on NOx emissions than with No. 2 oil. An average of about 20 ppm for
each one percent change of excess oxygen level was observed. Test No. 1
with No. 6 fuel oil showed NOx emissions peaking at about 5% excess oxygen
and a slight reduction occurred as the excess oxygen was further increased
to about 6%.
The data for natural gas fuel with ambient temperature air show that
the NOx emissions were only slightly affected by excess O level except for
Test No. 80. The preheated air data show a rather significant effect of
excess O level on NOx emissions. The change in NOx concentration with
excess O varied from about 5 to 40 ppm decrease for each one percent decrease
in excess O level (see, for example, Tests No. 15 and 77). The one excep-
tion with preheated air was Test No. 24, for which the NOx emissions decreased
with increasing excess 0 level. Tests No. 24 and 80 both showed decreasing
NOx emissions with increasing excess O level. The burners had heat release
rates of 160 and 110 MBtu/hr, respectively, which are much larger heat
87
-------
I
&
a
3
s
800
700
600
500
400
300
200
600
500
400
300
200-
100
»
78
19
16
COAL TOEL
10 12 14
OIL FUEL
The Nuncrals at the End
olE the Curves arc Test
Ntji~l-ers, and the Nucierals
in Brackets are the Fuel
Oil Grade.
Combustion Air Temperature
Ambient
Preheated
400
300.
200
100 -
NATURAL GAS FUEL
4 6 8 10
Flu* G*« Excess Oxygen, Dry, %
12
Figure 5-2. Effect of excess oxygen in nitrogen oxides emissions
for watertube boilers at baseline operating conditions,
natural gas, oil and coal fuels.
88
6000-28
-------
release rates than the burners used in the other boilers. Burner di'siqn
parameters and their effects on NOx emissions are discussed in Section 7.1.
5.1.2 Off-Stoichiometric Firing
Section 5.1.1 discussed the effect of the air/fuel mixture ratio
on nitrogen oxides formation and pointed out the benefits of firing at minimum
excess oxygen. However, a limit is reached where the overall air level
cannot be further reduced without causing incomplete combustion and carbon
monoxide and/or smoke. A technique to further change the local mixture
ratio, termed "off-stoichiometric (O/S) firing " or "staged combustion "
involves the development of a more fuel-rich flame zone than normal close
to the burner and the addition of air at an appropriate location to give
good secondary mixing and complete combustion. This secondary mixing may be
accomplished either by diverting part of the air to points out of the primary
flame zone, or by terminating only the fuel flow to one burner while
maintaining the total fuel and air flows constant. In each case the fuel/
air mixing is affected such that the primary flame zone is more fuel-rich
and the balance of the required combustion air is provided further down-
stream to complete combustion and prevent smoke, carbon monoxide or unburned
hydrocarbon formation.
Few existing industrial-size boilers are amenable to this NOx
reduction approach, because they have only one or two burners and no
capability to introduce additional combustion air through overfire or
NOx ports. Table 5-1 presents the results of the tests on those boilers
which did have multiple burners and where O/S firing could be achieved by
terminating the fuel flow to one of the burners. NOx ports are more frequent-
ly being included in the newer boiler designs that will be operational start-
ing in 1975.
Test Series 63 results are typical of O/S firing of light oil fuel
in a small boiler. An empirical evaluation of taking burners out
of service one at a time was performed to determine which burner-out
provided the greatest NOx reduction and still allowed operation at low
89
-------
TABLE 5-1
EFFECT OF BURNERS OUT OF SERVICE
Test Test Fuel
No . Load Type
k#/hr
63-6 46 Oil
63-16 47 PS 300*
63-17 43
63-18 43
63-19 45
63-20 47
68-2 50 Oil
68-12 51 PS 300*'
68-14 52
6-16 58 Oil
6-19 51 No. 5
6-26 50
6-27 50
6-28 50
6-32 45
6-36 69
6-23 40
6-37 42
6-41 41
22-2 112 Oil
22-13 118 No. 6
22-16 105
21-1 80 Oil
21-11 77-5 No. 6
21-12 76
21-13 76
21-14 76
21-15 75
21-16 76
21-20 76
Excess Burner Out
Oxygen of Service
% Number i
2 . 9 None
4 . 0 None
6.1 1
8.1 5
7.5 3
5.5 3
5.8 None
5.0 2
5.0 4
7 . 3 None
8.1 4
8.4 3
8.4 2
8.2 1
8.0 3
6.0 3
8.3 4
7.6 3
8.1 1 _,
7 . 8 None
8.3 2
10.5 2
7.0 None
6.2 3
6.0 3
5.95 3
6.15 3
6.30 4
6.55 4
6.6 4
[NOx] Burner
Theoret-
ppm ical Air
%
619 117
652 124
647 118
680 135
610 129
516 113
466 139
458 109
437 109
338 152
220 122
246 124
273 124
286 123
244 121
173 105
214 123
210 117
243 122
281 157
169 123
201 148
289 149
240 105
227 105
215 105
229 105
220 106
214 108
222 108
Comments
Normal excess oxygen
Higher excess oxygen
Burner Numbers
135
0 ° 0 ° 0 °
246
Minimum excess oxygen
Burner Numbers
135
0 ° 0 ° 0 °
246
Burner Numbers
0 O 0 0
4321
Lower Air
49% NOx Drop
Lower Load
Normal Excess O^
2 1 2
0 0 Smoke
0 0
Clear Stack J 4
Normal Excess O 4 3
O 0
0 O
Register Open Full
Registers Reset
Registers Reset
Registers Open full
Registers Reset
Registers Reset
* The API Gravity of this oil was typical of No. 5 - see Section 6.3.
90
6000-28
-------
TABI.E 5-1 Continued
EFFECT OF BURNERS OUT OF SERVICE
^est Test Fuel
o. Load Type
k#/hr
3-1 71 Oil
(9-10 60 No. 6
15-1 46 Nat-
15-10 40 Ural
Gas
15-11 41
.15-12 41
15-13 41
15-14 41
30-25 200 Nat-
30-26 199 "ral
Gas
30-27 204
30-28 202
Excess Burner Out
Oxygen of Service
% Number
7 . 4 none
8.2 1
*ta*«
2.6 None
1.8 4
2.6 4
4.4 4
3.0 4
2.4 4
2.7 None
3.4 3
3.8 3
5.0 3
[NOx] Burner
Theoret-
ppm ical Air
%
246 115
175 109
242 114
152 82
203 85
228 89
210 87
190 84
178 114
102 90
104 91
169 97
Comments
Normal. Excess O0 1
* 0 ° 0
2 3
[CO] >2000 ppm
Burner Numbers
3 4
0 0
O 0
1 2
Same burner no's, as Test 15
[CO] = 1300 ppm
Negligible [CO]
6000-28
91
-------
excess oxygen. It should be noted that the excess oxyyen level was rather
low under normal firing conditions, and taking a burner out of service
required an increase in the excess air/oxygen to prevent smoking. Therefore,
the theoretical air at the burner was decreased only slightly from 117% to
113%. The data show the center burner in the top row (No. 3) was the best
choice, and that an 18% NOx concentration reduction was obtained. More
testing time probably would have allowed the test crew to find an air
register setting that would allow even lower excess oxygen, burner theoreti-
cal air and nitrogen oxides.
The successful application of off-stoichiometric firing depends on
the condition and characteristics of an individual boiler. It is not
difficult to terminate fuel flow to a burner. The difficulty lies in
empirically selecting the best burner to turn off and the optimum settings
of excess 0 , air register positions, etc., which will allow satisfactory
operation without smoking, increasing overall 0 level, etc. This program
allowed only a few days per boiler for all tests, so sufficient time was not
available for optimization of 0/S operation. As a result, many of the tests
summarized in Table 5-1 do not show theoretical air below 100% at the burners.
The limitations imposed by this constraint are indicated in Figure 5-3 which
shows NOx concentration as a function of burner theoretical air. Due to
smoke and CO limits, only two tests show operation in the region below 100%
theoretical air where the NOx reduction effect is most significant. More
time will be allocated in Phase II to further define the potential of O/S
firing for industrial boilers.
A unit similar to the Test 63 boiler with the same burner pattern,
etc., was also tested on PS300 oil fuel (Test Series 68). Equipment problems
that arose during the test sequence precluded taking the top burners out
of service. Terminating fuel flow to bottom burners had less effect than
top burners previously tested. The best choice in the bottom row was the
center burner, and a nitrogen oxides concentration reduction of 6% was
measured. These results agree with previous experience, i.e., removing
top row burners from services reduces NOx more than removing bottom row
92
-------
tn
C
o
•H
w
10
•H
X
O
0)
4J
•H
2
C
0
(U
0)
id
O
-H
-P
n
3
en
uidd
93
-------
burners. It also indicates that removing inner, rather than outer, burners
(center burners in this case) results in better mixing of the air from the
fuel-off burner with the outputs of the other burners and allows operation
at lower excess oxygen levels.
A four-burner boiler with all burners in a single horizontal
row was tested with No. 5 oil fuel and each burner was taken out of service
one at a time (Test Series 6). The results in Table 5-1 again show
that removing one of the two inside burners (even though there is only
a single row of burners in this case) was best for minimum excess
oxygen and NOx. The NOx reduction for No. 3 burner out-of-service and
6% excess oxygen was 49% (Test Run No. 6-36) in spite of the fact that
this was a high-load run. Twenty-two runs were made with different com-
binations of load, excess air and burner-out-of-service. The nine entered
in Table 5-2 were selected to illustrate the effect of the burner theoreti-
cal air level on emissions.
In Test No. 21, a reduction from 149% to 105% in the burner
theoretical air and a corresponding decrease in NOx of 21% was achieved
With either No. 3 or No. 4 burner fuel turned off. Adjusting the air
registers resulted in an additional small NOx reduction to yield a total
of 26% reduction in nitrogen oxides concentrations. These test results
illustrate the value of proper setting of the registers in the burner-out-
of-service firing mode. More of this type of 0/S testing will be done
during Phase II of the program.
The burner pattern for Test No. 9 was unusual in that it was
triangular. When the fuel to the upper burner was turned off, it acted
as a large overfire port. A 29% reduction in nitrogen oxides was achieved.
A four-burner boiler with a square burner pattern, two rows and two
columns, was tested in Test Series 15 with natural gas fuel. The baseline
oxygen level was unusually low for normal operation. Taking a top burner
out of service at the same excess oxygen level reduced NOx by about 16%
while the carbon monoxide concentration was maintained below 200
ppm (Run 15-11). The excess oxygen had to be increased to about 4.4%
in order to eliminate completely the carbon monoxide (Run 15-12).
94
-------
Test Series 30 also was on a boiler with the square burner pattern
and burning natural gas. Taking a top corner burner out of service
reduced NOx emissions by 42% at an excess oxygen of 3.8%.
A larger unit with the burners arranged in a square and burning
No. 6 oil needed a substantial increase in excess oxygen from 7.8% to 10.5%
when a top burner was taken out of service in Test No. 22. The stack was
not clear at 8.3% excess oxygen. The burner theoretical air decreased only
9%, but the NOx concentration decreased by 28%.
In the three instances when a boiler was fired with the fuel to
one of the burners turned off, the particulate emissions increased. In
Test No.s 21 and 63, with No. 6 and No. 5 oil, respectively, the particu-
late emissions increased by 65% to 70%. In Test No. 22, also with No. 6
oil, the particulate emissions doubled. It is unclear whether the observed
increases in particulate emissions were an unavoidable result of O/S firing
or just another indication that insufficient time was available to optimize
the modified combustion operation.
Most industrial-size boilers tested had only one burner, so O/S
firing was feasible in only eight instances. On the multiple burner
boilers that were tested, significant NOx reductions of 18 to 49% were
obtained with both oil and gas fuels. Further reduction was not possible
because industrial boilers typically must be fired air rich when a burner
is taken out of service, as pointed out above. This limitation suggests
that redistributing the combustion air using air injection or "NOx ports"
would be an effective way to change materially the air-fuel mixing ratio.
Neither the air injection point, the proportion of total air, nor the
effectiveness in reducing NOx can be defined adequately from the data
collected during Phase /I. Only one boiler (a spreader stoker coal fired unit)
was tested in Phase I, that had the equivalent of overfire air in the form
of auxiliary oil burner throats. When these were used as overfire air ports,
NOx reductions of 20-25% were obtained with satisfactory boiler operation.
95
-------
Experience with utility boilers indicates that the use of NOx or overfire
air ports in reducing the total nitrogen oxides concentrations is effec-
tive, and this technique will be investigated further in Phase II.
5.1.3 Air Register Adjustments
The air-fuel mixture ratio was varied by changing the settings
of the air registers. Air registers on face-fired boilers typically
consist of a group of interconnected vanes oriented so that they all
move simultaneously {similar to a cylindrical Venetian blind) to vary
the area and angle through which the air enters the burner, and thereby
vary the flow rate and degree of swirl. The area and direction are
usually changed simultaneously by a lever mechanism so that decreasing
flow area is accompanied by increased air speed and swirl. Most of the
smaller boilers tested were single burner boilers and had fixed air control
with vanes bolted or tack welded in position. In these cases, swirl
and mixing were not parameters that could be investigated within the
scope of the first phase of the program.
Experience with multiburner boilers has shown the most important
effect of air register adjustments to be in air flow rate to individual
burners to control the air distribution and air/fuel mixture ratio across
the burner front. The swirl effect on the NOx production of an
individual burner usually appeared to be relatively small. At constant
air flow, closing an air register should increase swirl and mixing to
produce a shorter, hotter flame. This would normally result in higher
nitrogen oxides formation, unless the improved mixing allows operation at
lower excess air; but since air-fuel mixture ratio is a major factor,
the mixture ratio might obscure the swirl effect.
Table 5-2 summarizes1 the results of tests run on both face-fired
and corner-fired boilers where the excess air and load were held practically
constant, while the air register settings were changed.
96
-------
TABLE 5-2
EFFECT OF AIR-FUEL MIXING
BY CHANGING THE AIR REGISTER SETTING
Face-Fired Boilers
Test
Rian
Number
30-14
30-19
7-10
7-13
70-11
70-10
10-2
10-12
10-10
Fuel
Type
NG
NG
#2
#2
#2
#2
#6
#6
#6
Test
Load
k#/hr
259
254
88
82
100
100
54
49
51
o2%
3.0
2.8
5.7
6.6
6.6
6.6
4.7
3.9
5.2
Register
Setting
% Open
70
100
100/100
100/70
50
100
65/65
100/45
100/100
NOx
Meas.
ppm
197
204
177
180
383
405
186
174
228
Change
%
— .—
+3.5
—
+1.7
+5.7
—
-6.5
+22.6
Comments
Baseline
All 4 regis-
ters reset
Baseline
Baseline
Baseline
Burner
Pattern
0- 0
0 0
0
0
0
0 0
Corner-Fired Boiler No. 24
Pulverized Coal Fuel
Test
Run No.
26-1
26-10
26-11
26-12
26-13
26-14
26-15
Excess
%
5.3
6.0
6.0
5.8
6.1
5.1
5.4
NOx
Meas.
ppm
378
468
466
427
405
368
376
Chancre
%
-
+23.8
+23.3
+13.0
+ 7.1
- 2.6
- 0.5
Comments
This is baseline run.
Used baseline register positions
and fuel flow, but higher O .
Same O but symmetrical register &
damper positions.
Now fuel to upper burners decreased
& lower increased.
Fuel to upper burners decreased
to minimum.
Return to baseline 0 ; lowest NOx.
Open upper 3 registers wider.
97
6000-28
-------
TABLE 5-2 Continued
EFFECT OF AIR-FUEL MIXING
BY CHANGING THE AIR REGISTER SETTING
Corner-Fired Boiler No. 20
Natural Gas Fuel
I
I
I
I
Test Run
Number
77-11
77-16
77-17
Excess
?'
4.5
4.5
5.8
NOx
Me as.
ppm
320
338
242
Change
%
—
+5 . 6
-24.-4
Auxiliary
Air Register
Position
2
1
5
Coal
Damper
Position
1
1
2
Comments
Baseline
Air Register Closed
Air Register Opened
& Coal Damper Opened
i
Corner-Fired Boiler No. 20
Pulverized Coal Fuel
Test Run
Number
78-9
78-10
78-11
78-12
78-13
KVB
Excess
Oxygen
%
5.6
6.5
6.5
6.5
5.9
Damper Settings
Upper
Auxiliary
Air
(0-5)
2
2
2
2
5
Center
Oil
(0-5)
2
5
5
5
?.
Lower
Auxiliary
Air
(0-5)
">
2
2
2
fj
Upper
Coal
(0-5)
4.5
4.5
4.5
4.5
4.5
Lowe r
Coal
(0-5)
4.5
4.5
4.5
4.5
4.5
NOx
viea .
ppm
494
567
535
508
492
Hhancjo
%
— j
4-14.8
+ 8.3
f 2.8
-0.4
Comments
Baseline
this series.
Opened oil
damper full.
Lowered
excess oxygen
Lowered excess
oxygen more.
Oil damper
returned.
Auxiliary air
opened full.
6000-28
98
-------
For both Test No. 30 and 70 on face-fired boilers, the normal
setting appeared to give the lowest output of nitrogen oxides. Opening
the registers to the fully open position increased the NOx concentration
slightly. Opening the registers in Test No. 10 from the 65/65 position
to the 100/100 position also caused the NOx to increase partly due to an
O increase. Apparently, decreasing the swirl by opening the air registers
increases nitrogen oxides formation. When the lower register was closed
in Test No. 7 to force more air through the upper burner in an attempt to
produce an overfire air effect, the nitrogen oxides measurement was
practically unchanged in spite of a 0.9% excess 0 increase.
In general, with these face-fired units, when the air registers
were opened the nitrogen oxides emissions increased, and when they were
closed the emissions decreased. This result is contrary to what had been
expected. Register settings will be investigated further during Phase II.
A special series of combustion air distribution tests were made
on two corner-fired boilers, No's. 20 and 24, at Location No. 12, burning
either pulverized coal or natural gas fuels. There were five burners arranged
vertically in each of the four corners of the furnace as shown in Figure 5-4.
„ f
COAL
OIL/GAS
COAL
AIR
7 — — — ^"
R
ftn
Czi3
1 I
EE3
COAL FIRING
UPPER AUXILIARY-AIR
COAL FUEL- AIR
COAL AUX. - AIR
COAL FUEL - AIR
LOWER AUX. ADR
OIL/GAS FIRING
UPPjuR AUXILIARY AIR
OIL/GAS AUX. - AIR
OIL/GAS FUEL - AIR
OEL/GAS AUX - AER
LOWER- AUX. - AIR
Figure 5-4. Typical Arrangement of Corner Burner Showing Secondary
Air Distribution to Coal, Oil/Gas and Air Compartments
6000-28
99
-------
There was no standard setting of an individual air register. The position
of an individual register depended upon the desires and judgment of the
boiler operator on duty at the time.
Test Runs No. 26-1 and No. 26-10 were run on Boiler 24 with the
register positions existing at the time measurements began. Test Runs 26-11
through 26-15 were run with all four corner registers of a given level
at the same setting. Then changes were made in all four registers of the
same level and the resulting emissioas measured. Fuel flow also was
varied.
The sequence of air register settings and fuel flow that was followed
is tabulated in Table 5-2 Overall, there was no significant reduction in
NOx between the baseline Run No. 26-1 and the best-adjusted register Run
No. 26-14. However, when going from normal fuel distribution in Run 26-11
to the least possible fuel through the upper four burners in Run 26-13, the
NOx decreased a total of 13% from that measured in Run 26-11.
Reducing the fuel while holding the air constant in the upper level
burners is akin to using overfire air. In utility boilers, changing the
fuel and air mixing by diverting the overfire air has been found to be
effective in reducing NOx emissions without sacrificing generating capacity.
This test indicates a similar effect in industrial boilers.
Tests also were made on Boiler No. 20 at the same location burn-
ing pulverized coal and burning natural gas. The measurements are listed
in Table 5-2. Three runs were made with natural gas fuel and the air
register repositioned to reduce the nitrogen oxides emissions. One
cannot draw sweeping conclusions from only three data points, but it
appears that when the auxiliary air registers were closed one setting from
Position 2 to Posi-ion 1, the nitrogen oxides concentration increased.
When both the air register and the coal damper were opened, the NOx
decreased markedly fror. 320 ppr. to 242 ppn.
When this latter unit was fired on pulverized coal in Test No. 78,
register position adjustments were not as successful in reducing nitrogen
100
-------
oxides concentrations. Run 78-9 was the baseline for this series of tests,
and the test load was 270 k Ib/hr of steam. When the damper on the center
oil burner was opened from Position 2 to the fully open Position 5, the
NOx increased from 494 ppm to 567 ppm. Lowering the overall excess air/
oxygen in Runs 78-11 and 78-12 reduced the NOx to 508 ppm, but this still
was above the series baseline. In Run 78-13 the center oil register was
returned to its normal Position 2, and the auxiliary air registers above and
below the two coal burners were fully opened. The NOx then dropped to
492 ppm, which was about the baseline level. Thus, it was not possible
during this limited series of runs to decrease the NOx emissions by resetting
the air/fuel mixture control system.
5.2 AIR PREHEAT TEMPERATURE
Industrial-size boilers normally do not use combustion air preheaters.
Of the 47 boilers tested, 14 operated with preheated combustion air ranging
in temperature from 250 to 650°F. These boilers were watertube boilers
which did not have ducting to bypass the preheater and vary the windbox
air temperature. This limitation precluded a systematic study of the effect
of preheat temperature on NOx emissions for a single unit. Experience with
utility boilers has shown that preheat temperature influences NOx emissions.
The magnitude of this influence varies between units, fuels, and operating
conditions with an increase in combustion air temperature usually resulting
in an increase in NOx emissions.
The baseline NOx emissions as a function of combustion air tempera-
ture are presented in Figure 5-5 for natural gas, oil and coal fuels.
The natural gas data indicate that the effect of combustion air temperature
on NOx emissions depends on burner heat release rate. The lower the heat
release rate per burner the less sensitive the NOx emissions are to combus-
tion air temperatures. More data are required to substantiate this conclu-
sion , but if true it could have a significant effect on sizing of a burner
for a particular boiler. Minimum NOx emissions for boilers designed for
101
-------
I
300
700
600
500
•100
300
200
I
I
0-60 ft3/(MBtu/hr)
/
.Spreader and Pulverizers ^~
}(Jnder£ed Stokers 84 ft /(Mhtu/hr)
-A-
I
Cyclone
2 it /(Kbtu/hr)
Heat Release Volume*
COAL
100
10 50
200 3CO 400 500
Combustion Air Temperature, T
700
100 ISO 200 250
Combustion Air Temperature, °C
300
350
Figure 5-5.
The Effect of Combustion Air Temperature on Baseline
Nitrogen Oxides Emissions for Natural Gas, Oil and Coal
Fuels.
6000-28
102
-------
preheated combustion air and gas fuel may be achieved by using a number of
small burners in place of a single, large burner; whereas boilers designed
for ambient temperature combustion air and gas fuel may require only a
single, large burner. The effects of burner design on NOx emissions are
discussed in Section 7-1. The regression equation developed for natural
gas fuel showed a strong correlation of NOx with air preheat as discussed
in Section 8.
The oil fuel data indicate that combustion air temperature does
not strongly affect the NOx emissions for industrial boilers. The two
points shown above 450 ppm for No. 5 oil are for boilers whose emissions
consistently were atypical. The cause is still under study.
The coal fuel data presented in Figure 5-5" at first appears to
show some effect of combustion air preheat; however, the slight rise with
combustion air temperature may actually be due to the volume and heat
loading of the furnace. The NOx emission level was shown to be correlatable
as a function of parameters other than air preheat temperature. The pre-
heated air NOx emissions are the same as the unheated air NOx emissions for
furnaces that have the same furnace volume per unit heat release. This
similarity indicates that combustion air temperature does not strongly affect
the NOx emissions for a coal-fired unit, but that furnace volume and
burner heat loading do have an effect. The two boilers with a large
furnace volume per unit heat release rate (84 ft /MBtu/hr) are both
underfed stokers, and underfed stokers currently are installed only in
boilers of 30 k Ibs/hr output or less. The boiler data in the midrange
of furnace volume/unit heat release rate (30-60 ft MBtu/hr) include
mostly spreader stokers with a traveling grate and a few pulverized coal
burning units. The boiler with the smallest furnace volume per unit heat
release rate (2 ft /MBtu/hr) and the highest burner heat release rate
(256 MBtu/hr/burner) is a cyclone furnace. Section 7.1 discusses the
effects of burner design on NOx emissions, and Section 7.2 discusses the
effects of furnace design.
103
-------
I
I
5.3 FUEL OIL TEMPERATURE
Test results showed that NOx increased as oil temperature was
decreased from the normal operating range. The fuel oil temperature was
varied during Runs 9 and 10 of Test No. 34 to determine its effect on NOx
emissions. Figure 5-6 shows the results from these tests. The lowest
temperature tested was 95 °C (200°F) on Run No. 10, which increased the NOx
emissions to 316 ppm as compared to 298 ppm measured on Run 11 with the
oil temperature at 120*C (250°F). Run 9 was with the oil temperature at
102«C (218«F) and 305 ppm NOx were measured. The dashed circle at 250°F
in Figure 5-6 is for Run 11 adjusted from 5.4% 0 to the 5.6% 0 level of
Runs 9 and 10. Similar effects of oil temperature on NOx emissions have
occurred in some other KVB field test programs and have been attributed to the
atomizing pressure being insufficient to atomize properly the more viscous
oil.
320
I
§
u
M
0) O
*O
•H <#>
"
91 >i
!S
•H
310
300
290
Numerals within Symbols
are Run Numbers for
Test No. 34.
Air Atomization
No. 6 Fuel Oil
Normal Operating Range
200 210 220 230
Fuel oil Temperature at Burner, °F
240
250
Figure S-g. Effect of Fuel Oil Temperature on Total Nitrogen Oxides
Emissions.
6000-28
104
-------
5.4 FIRING RATE (PERCENT LOAD)
Although reduction of total steam load to control NOx emissions
would not be an acceptable control strategy in any but the most drastic of
circumstances, the change in NOx emissions with unit firing rate might be
utilized where the required steam load could be produced using all boilers
at part load, or a few boilers at full load, etc. A minimum NOx firing
strategy might be possible without limiting steam production.
The effect of firing rate on the level of nitrogen oxides emissions
was investigated by raising and lowering the boiler load from the base load
point of 80% of nameplate capacity. The boiler control settings, including
excess 0 , were normal for each load. In general, changing the firing rate
did not have a strong effect on nitrogen oxides emissions. Usually the NOx
reduction effect of lowering the load was compensated for by the increase in
excess air at reduced load that was called for by the boiler firing procedure
used by the boiler owner, and the net result was that the NOx either did not
change or even increased at the lower firing rates.
Figure 5-7 illustrates the results for firetube boilers. The oil
and the gas data each fell in well-defined bands, and both bands are
relatively insensitive to load changes.
Watertube gas-fired boilers also were relatively insensitive to
load changes unless they had air preheaters. The measurements from Tests
No's. 15, 25 and 77 that are plotted on Figure 5-8 are the data collected
from boilers with preheated combustion air. The NOx concentration
dropped sharply from about 275 ppm to 200 ppm as the firing rate dropped
from 85% of capacity to 60% of capacity. A combination of lower air
preheat temperatures and poorer fuel-air mixing and the resulting lower
temperature combustion products probably caused this decrease in NOx
production. Reduction in air preheat temperatures alone would not have
caused a decrease in NOx concentration of 150 ppm, since the combustion
air temperature drop is on the average only 40°F from the high to low
firing rate.
Watertube oil-fired boilers also showed little or no relationship
between the NOx emission and the firing rate.
105
-------
•H
O
in
(0
o
o
uidd 'Z
CD
to
(N
in
03
rH
•H
3
-P
OJ
M
•H
4-1
in
c
O
-r-l
!/]
en
•H
I
Ifl
0)
c
0)
JJ
•H
C
O
-P
C
O
-------
M
(P
0)
.a
3
-P
0)
r-l
(U
in
C
O
tn
•H
§
U)
0)
T3
•rH
X
O
c
0)
-P
-H
C
(0
-M
O
4J
a
o
0)
-p
c
•H
+J
O
cu
«4-t
M-l
W
00
I
in
Q)
w
3
uidd
107
-------
I
I
Generally, coal units showed an increase in NOx emissions when
firing below 60% boiler capacity. This increase coincided with a signifi-
cant increase in the excess air level.
108
-------
SECTION 6.0
FUEL PROPERTIES
The physical form and chemical composition of the fuel have a
strong effect on pollutant emissions and emission levels can be reduced
readily by shifting to a different fuel. For example, oil-fueled boilers
generally have lower nitrogen oxides emissions than do coal-fired boilers.
A shift from residual oil to distillate oil would result in lower nitro-
gen oxides emissions because the fuel-bound nitrogen content of the
lighter oil is less.
Gas fuel presents the simplest situation, since only gas-gas mixing
is involved. Natural gas fuel is mostly methane with minor amounts of C
and heavier constituents. Natural gas is relatively consistent and already
in a state allowing easy mixing and combustion. The properties do not
materially affect the emissions. An exception to this generalization may
exist for process waste from chemical plants or refineries where gas streams
high in organic nitrogen may be burned, or with future fuels, such as low
Btu gas derived from coal.
Combustion of oil fuel is significantly more complex. It must be
atomized and vaporized to burn properly; so fuel properties such as viscosity,
specific gravity, volatility, ash, Conradson carbon, and heating value become
important parameters. Atomization can be accomplished in different ways and
can significantly affect emissions. The design aspect of this problem is
considered in the following section on design parameters.
In evaluating the effects of oil parameters on emissions, the degree
of sameness and difference from one oil to another should be considered. Oil
was formed by the same basic mechanism, so crude oils have a great deal of
similarity. At the same time, location-to-location differences in temperature,
pressure, and raw material cause variations in chemical composition and
characteristics. Typically, crude oil is further processed and segregated
into fractions, defined for commercial purposes as Number 2, Number 6,
etc. where each oil designation has a specified allowable range of properties.
The result is that a given grade of oil from two sources will typically be
109
-------
I
very similar in chemical and physical properties and in NOx emission
characteristics. Variations will exist due to location differences, and
these variations may sometimes be magnified by blenajung procedures which
can result in unusual characteristics. One effect of this situation is
that correlations of emissions with a particular oil property become some-
what questionable. It is not clear whether emissions versus API gravity
has a causal relationship or that gravity indicates a Number 5 oil which
has a certain typical fuel nitrogen content. Fuel nitrogen content is
known to be very important and is discussed in detail, and other proper-
ties (API gravity, carbon residue, and sulfur content) are briefly dis-
cussed in spite of this uncertainty.
j
Coal presents even more problems, since it is mined as solid material,
contains more impurities, is highly variable, and must be crushed or pulverized
for burning on grates or in air suspension. The difficulties of coal handling,
grinding, feeding, slagging, and flyash collection can easily become the
predominant design and operating problems.
Table 6-1 lists the properties of the fuels that were used for the
coal and oil fuel tests. This quantity of fuel property information, for
a variety of fuels from throughout the country, collected and presented on
a consistent basis probably represents the most extensive published
information available. It proved to be impractical to collect and ship
a natural gas fuel sample back to the laboratory for analysis, so typical
analyses of natural gas for those parts of the country where natural gas
fuel testing was done are listed.
110
-------
o
o
o
'
£
n
§
rH
CO 0)
H 3
tO h
H! rH
f6 -rl
5 o
i-i
M
(uidd) jaATTS
(uidd) snjoifdsoqd
(wdd) mrvrava
{mdd} omTUiojti{3
(radd) 4T»qoo
(uidd) umTuv^ti
(mdd) OUT 2
(uidd) umfpos
(mdd) j^ddoo
(mdd) wrvptv?
(mdd) asauc&uvM
(mdd) UODTTTS
(mdd) UTJ,
(uidd) uojci
(uidd) t^ipTN
(add, umTp*u*A
(cff /n^g)
-UKJO JO }E3H
(%) JnjTns
(%) M»V
IdV A^TABJO
adAj, -[TQ
MqmvN t»S3i
i i i
^ * 9
rH rH g
Q Q Q
22 2
CO
•* *» •*
(N Ol O
22 2
(M (N
00 g
(NO* O
CD CO Q
O
O O vO
00 S
CD CO
00 g
* * s
rH rH •
ON
m r" o
0* (M «T
rH rH
CO CO O
(N r>4 rsi
o* (N r-
m
rH rH O
m m
in
r* r- in
in
m m rH
r* r-
o o o o
CO CD m -q-
CD CO CO 0*1
O O ON
O O O O
; r,
rH rH . . O
m m ON O P^
kD vO in in (N
r< c. n -0 P.
4> S
° ° g 1 i
ft)
O 4 Q P O
Z £ 2 2 2
• Q Q Q Q
-42222
m •* o o o
^ CO V0 Ot oj
V 0)
co m r^ o u
r^ (N m o> o*
O (M ^- O O
m ^ ^0 in in
rH m O
• rH oi 2 2
r*.
O rH O rH rH
o* r* r-
rH O rH O O
o
• m fM 0 0
0** rH Ol rH rH
-H r* r- r*
ON 0 UN ^ ^
01 • Ol Ol
co m vO co co
rH ui m m n
m
CO rH rH rH rH
PJ \D v in in
Ol - •
. m IN 01 OJ
o o o o o
r-
o
rH in in in m
. -. . .
CO ON O rH Oi
g
M
in
ON
o
CO
o
s
00
t-.
rH
0)
O( i£ vD O
O CO O O O
O O O O Ol
ON CO ON Cn vD
rs m oi 04 *
in Ui <-H in in in
z b
« * T U1 »
g
r*
rH
ON
m
1
tn
o
01
rH
0
(Q tfl in tfl
Q _, rH ,H rH
2 2 2 S S
*. 0 U 0 U
jc — jz — j=— j:
CM D, O. Ot
Ol W 14 kl kl
CP CT> C7* 0*
o o o c
O *-> *J 4J *-»
I- O 41 O V
a, a. £, a
01 4 « « «
o o o o
OJ
ui u tn u
01 c c c e
-r< -H -iH -H
m o U u O
03333
• w> t/> in n
tn m tn «
r-
1) C V w to
rH rH rH
2s 2
M W k*
*> *> 4-*
S. i S.
4 (Q «
O O O
c c. jr
0) U (ft
« < 4
S S S
U U U
33 3
tn w tn
tn m n
XX x
0 0
rH rH
CO CO O (N m
rH
O O O r-t
o o o o
V V
rH rH O
t • < m
o
fN 01 rs oj ui O
* r- m « -,
go o
.
« 01
« «
o co m
« 0 r-
H
2
• . o
rH ON Ol
a
o
co m oi m rH
m co rH
O O O r-( O)
O O O O O
>O rH ^ kO
m O O
• • .mm
o o
O< Ol O) U5 Ot/) O
^ m « co o
1
to
rH
u
a
*j
&
«
0
X
q j
U
»
C
m
j
4t
1
ON ^
S *
O rH
o o
V
r- vo
o
01 m
m „
111
-------
D
CO
10
rH >H
rQ J
(8 rtj
EH 2
w
D
Pu
rH
0)
3
fc
O
U
^ juj i L j-tuu
(uidd) uiriT&st?4O^
(uidd) uirij^uoj^s
(uifTd) umr ;io-i TT'/
(uidd) }"[CC1O3
(uidd) 'im-cpos
(uidd) I3AT.TS
(uidd) uirupeuEA
(uid3) Ufi
(indd) T.3XDTN
(uidd) pE3T
(mdd) 3S3UP&UPW
(uidd) ucaoa
(uidd) umTT.tA'.iag
(uidd) amfjca
(uidd) aaddoo
(uidd) umTUi^H'Tj,
(uidd) umTS3u5t?w
(odd OOOT) irntofBO
(gxld OOOT)
uoaj
(uKld OOOT)
(inJd OOOT
UOOf ITS
(%) ^sv
(%) xnjtns
us&o- A
(%) uoqj^o
ssojg uoT^snc
-U1OO JO HM33H
jraqinrm 3S3J,
I
]
I
i
rH rH r-
m i/i [--
o o o
CO CO CT»
rH rH *3"
in u"i ro
••J ^f lA
CO CO fN
r* r- o
o o o
rH rH O1
. * r-
O O
VD r~- co
co m o
CN fN ro
O
a p ti i
X V-. rJ
H
or- o
rH fN
t f*l (N
O O O
U3 CO M
0)
CO O
Q in nj |
CO VD rH
0 0
i-J i-l CN
tn H
r-
ro in o
fN rH ^«
O ^r r-
O4 fN n
in m n
O
o
to Q m
O vD O
or- T
rH rH
CT> \D fN
O O O
m co vc
m n »^)
o o o
CO CO ^O
v£) v£> O^
rH fN fN
CO fN
in un r-
r-t o> n
rH *H
rH rH CN
rH 0 1-
CN r-l O
co m *H
in m \o
i in *3* vD CT\
*£ vO CTi f"»
in ro O ro
\o in O ao
t*- in CN in
rH rf rH
O o vo r-
ro rs r-t -H
m fN CN rj
o c.) ;
-j- TJ r i n '
r- H & „•
o o o ;
rg in a) a)
vo ^--i in to
CO O'1' r^t n">
vD (N [*- r*>
O O O O
VO O O O
vo ^ in in
rH
Q Q Q Q
O fN CN
rH ITl CTt CT* '
41
O
ro ^ oj CN
CO
^ • rH rH
r- o o o
o m n m
^* PI fN fN
m (N -q* rH ^
0
CO ^T CN fN
rH ^t CN CN
O O O O
CO fN CN fN
O O O
O O CN CN
T O X CO
O r-i
(N fN 0^ CT\
rH CO rH rH
• • • 1
fO CO ^* •*
CO CO 00
, CN rH rH r-i
f> & t~~ t*~
CN CN O O
m m vo \o
rn n co co co
• r- in m
, o in m r- r--
rH rH H rH rH
CO fN
• rH fN rH rH
O o o
n ro vD r^ r^-
o o o r- r^
fN tn cr> m ro
M OT rH O O
rn CO rn 0"* 0*i
^ rH fN CN m
CN
a
c
c.
i
a
cc
0
O-.
ir
C,
2,
. 1
r-l
1
r-l j
C| |
1-i
CCi |
C'
r-l
O
CXI
r-l
r-l
O
O
O
r-l
O
O
••3'
r-l i
r-l
r-l
f)
IT)
cn
r-l
O
IT)
r-l
r-l
ai
CO 10
in in
rH r-l
rH H
? s
c^ '
^3* LTi
CO r-|
CN Ifi
fN C'
VO I-*
o o
^O LTi
rH tt'
CT. VC1
rH t-
I
I
112
-------
Table 6-1
FUEL ANALYSIS SUMMARY
Gas Fuel
Test Number
4
5
12
13
14
15
24
25
30
37
38
39
40
41
47
48
49
58
60
67
69
75
77
80
f
CM
O
O
.9
.9
.9
.9
0
0
0
.9
0
.9
.9
.8
.5
.9
.5
.8
.8
.9
.9
.9
.9
2.5
2.5
0
CM
O
0
0
0
0
.3
.3
.3
0
.3
0
0
.6
0
0
0
.6
0
0
0
0
0
0
0
.3
dp
CM
Z
.1
.1
.1
.1
4.6
4.6
4.6
.1
4.6
.1
.1
6.3
3.9
.1
3.9
6.3
8.4
.1
.1
.1
.1
1.3
1.3
4.6
dp
5*
85.8
85.8
85.8
85.8
74.0
74.0
74.0
85.8
74.0
85.8
85.8
88.9
95.6
85.8
95.6
88.9
84.1
85.8
85.8
85.8
85.8
23.6
23.6
74.0
d?
X
OJ
U
13.2
13.2
13.2
13.2
20.6
20.6
20.6
13.2
20.6
13.2
13.2
3.4
0
13.2
0
3.4
6.7
13.2
13.2
13.2
13.2
69.7
69.7
20.6
High Heating
Value (Btu/ft
1108
1108
1108
1108
1129
1129
1129
1108
1129
1108
1108
964
966
1108
966
964
967
1108
1108
1108
1108
1548
1548
1129
Density - 0.046 #/ft3
6000-28
113
-------
I
I
6.1 FUEL NITROGEN CONTENT
There are two important mechanisms for the formation of NOx. One I
is thermal fixation of atmospheric nitrogen, and the other is conversion
of nitrogen compounds in the fuel. The magnitude of the potential fuel
nitrogen effect is about 1300 ppm of nitrogen oxides for complete conversion
of 1% nitrogen in a typical oil and about 1900 ppm for a typical coal.
Partial conversion of the fuel nitrogen occurs and the percent conversion
depends on the fuel nitrogen content and the availability of oxygen. The
percentage conversion is high for low nitrogen oil and decreases with in-
(10)
creasing nitrogen content.
The fuel nitrogen content of residual oils used in industrial and
utility boilers ranges from 0.1 to 1.0% by weight. Distillate oils are
generally 0.2% or lower in nitrogen content. Crude oils, which contain
distillate and residual fractions, are intermediate. Shale oils have
nitrogen contents as high as 2.5%, and pyrolytic oils made from waste materials
could conceivably contain 5% or more of nitrogen. The oils tested during
this program varied in nitrogen content from .006 to 0.52% by weight.
Table 6-1 presents the nitrogen content for each oil and coal fuel tested.
The nitrogen contents for the No. 2 oils were from .006 to .031%, the No. 5
oils were from .10 to .52%, and the No. 6 oils were from .26 to .46%. The
nitrogen content for the coal fuels tested during this program varied from
1.29 to 1.80% by weight as fired.
The baseline nitrogen oxides emissions as a function of fuel
nitrogen content are presented in Figure 6-1 for the oil and coal fuel
tests. Not all data points were included since a lot of the data were
nearly identical and would lie on the top of the points shown. The oil
fuel tests No's. 63 and 68, which are inconsistent with the remaining
data, are the PS300 oil tests conducted with nearly ambient temperature
fuel oil at the burner instead of the 160 to 180°F typical for No. 5 oils.
The fuel nitrogen content of the coals did not vary over a large enough
range to show any dependence of baseline nitrogen oxides emissions on fuel
114
-------
O Z
M C rH
rH G
0)
-P
G
O
O
0 G
4J
0) c
tT> d)
O -p
•H
2
0)
fa
0
u
C
0)
-p
•H
-P
U
0)
w
0)
rH
3
en
-H
CM
CO
C-i
t
o
o
o
UD
uidd
'0
'uoTq\eaq.uaouoo sap-pxo
115
-------
nitrogen content. This does not indicate that no such dependence exists.
It does indicate very similar nitrogen content for many coals.
During th;.s program fuel oils of varying nitrogen contents were
burned in the same boiler at four test locations. Table 6-2 summarizes
these data. At Location 19 changing from No. 2 oil with .006% nitrogen
to No. 6 oil with .44% nitrogen resulted in a 43% conversion of the fuel
nitrogen to nitrogen oxides for air-atomized tests and 51% conversion for
steam-atomized tests. Tests conducted at Location 23 with air-atomized
No. 5 and 6 oils with fuel nitrogen contents of 0.28 and 0.27%, respec-
tively, resulted in 44% conversion of the fuel nitrogen to nitrogen oxides
for the No. 5 oil and 52% conversion for the No. 6 oil. Similar air atom-
ized tests conducted at Location 24 on No. 5 oil with 0.20% fuel nitrogen
resulted in 41% conversion of the fuel nitrogen to nitrogen oxides. The
test series conducted at Location 26 when No. 2 oil with 0.02% fuel nitro-
gen and No. 5 oil with 0.1% fuel nitrogen were burned both with air and
steam atomizers resulted in 60% and 56% fuel nitrogen conversion to nitro-
gen oxides, respectively. The average for these tests is 50% fuel nitro-
gen conversion which agrees quite well with the average of 46% for all the
data.
The curve to fit empirical data from an in-house KVB, Inc. lab-
oratory investigation of the influence of fuel nitrogen on NO emission is
also presented in Figure 6-1. The KVB laboratory curve is nitric oxide
concentration measurements versus fuel nitrogen content for 130% of theore-
tical air at the burner. The percent theoretical air for the measurements
of this study are written beside each data point. The Phase I data are
slightly above the KVB laboratory curve. The intercept at zero fuel
nitrogen content is the thermal NO contribution, and the slope of the
curve is the contribution of converted fuel nitrogen. This approach leads
to the conclusion that for normal operation conditions the thermal NO for
the tests shown vras in the 60 to 200 ppm range and that the fuel nitrogen
conversion averaged 46%. The thermal NO and fuel nitrogen conversion in
116
-------
TABLE 6-2
EFFECT OF FUEL OIL GRADE ON TOTAL NITROGEN OXIDES EMISSIONS
AND CONVERSION OF FUEL NITROGEN TO TOTAL NITROGEN OXIDES EMISSIONS
Location
Number
19
19
19
19
19
23
23
23
24
24
26
26
26
26
Test
No.
1
2
52
53
54
64
51
34
73
46
56
57
44
45
Fuel
#6 oil
#6 oil
#2 oil
#2 oil
#2 oil
#2 oil
#5 oil
#6 oil
#2 oil
#5 oil
#2 oil
#2 oil
#5 oil
#5 oil
Burner
Type
Steam
Air
Steam
Air
Pressure
Air
Air
Air
Air
Air
Air
Steam
Air
Steam
NOx dry
@ 3% 02
ppm
350
334
65
97
80
127
275
298
84
186
116
118
173
161
Excess O
dry, %2
3.6
4.4
3.6
3.0
4.3
6.8
6.3
5.4
3.1
3.2
8.0
8.0
7.3
6.7
Fuel Nitrogen
Content ,
Wt. %
0.44
0.44
.006
.006
.006
.015
0.28
0.27
.014
0.20
.020
.020
0.10
0.10
Conversion ,
%
51
43
*
*
*
*
44
52
*
41
*
*
60
56
, ,1
6000-28
*Fuel nitrogen content was too low to determine a realistic conversion
percentage. The conversion was near 100%.
117
-------
the field-tested boilers were similar to the laboratory burner used for
the subscale study. This further indicates a lack of nitrogen oxides
variation with unit size for oil fuel. Other investigators have reported
similar values of fuel nitrogen conversion. ' ' ' Sufficient data
were not collected to allow evaluation of fuel nitrogen conversion under
off-stoichiometric conditions; however, the KVB laboratory tests discussed
above showed a reduction in fuel nitrogen conversion to about 20% for
fuel rich combustion.
6.2 API GRAVITY
The API gravity of the fuel oil burned was measured at 20°C. The
nitrogen oxides and total particulates are shown as a function of API
gravity in Figure 6-2. The data points marked "B" are from the Battelle-
Columbus field investigation.
The measured NOx fell into two groups: 1) where the fuel oil gravity
matched the API gravity specification for diesel or No. 2 oil the NOx was
between 100 and 200 ppm, and 2) where the fuel oil gravity matched No. 5 or
6 oil and the NOx was between 170 and 620 ppm. The specific grade of oil
being fired is listed on Tables 6-1 and 7-1. The fuel burned for Test
Nos. 63, 68, and 70 was designated as PS300 which when analyzed was found
to have properties much like No. 5 oil.
It should be noted that the data might be correlated as well by
fuel grade number, as indicated at the top of the figure. While fuel grade
number could in no way be considered a natural property, it does reflect a
grouping of properties and reflects the similarity between different oils
as previously discussed.
6.3 CARBON RESIDUE
An oil fuel property that appears to correlate with the particulate
emissions is the carbon residue. The measurements from the current field
118
-------
OUl'
*7 r^r\
/oo
CN
O £on
OUU
n
>i
. 500
U)
C
5
<0
4-1
S 400
Qj ~ */ V
u
G
O
U
co
(U
•o
•^ 300
G
0)
en
0
4J
vd ?oo
^^ ** " ^*
H
«5
O
, EH
-| *-* *~i
100
0
Fuel r...
Grade P
N|
V
O 44
C§/ 5"v
Vzy
#4
II .1^
^| |<- #5 -^j |4
©
— ^
-------
tests with oil fuel are plotted in Figure 6-3, along with data points
labeled with the letter "B" from a Battelle-Columbus Laboratories report .
With coal fuels, however, the carbon residue-particulate relationship is
not clear-cut (see Table 6-1).
6.4 SULFUR CONTENT
The results of the measurements of total sulfur oxides in the flue
gas are shown in Figure 6-4. The curve shows sulfur oxides concentration
emitted as a function of the sulfur content of the fuel and compares it
with calculated values assuming 100% conversion of fuel sulfur to sulfur
oxides (SOx). The measurements of which these data are a part indicate
that the sulfur emissions were dependent almost solely upon the sulfur con-
tent of the fuel.
It is apparent that for oil fuel, practically all of the sulfur is
emitted as gaseous products of combustion and an insignificant amount is
contained in the fly ash or other particulates. The coal fuel data are not
as consistent as the oil data, and this may indicate that the higher sulfur
coals (greater than 3%, dry) have inorganic sulfate which does not convert
to gaseous sulfur oxides but, rather, contributes to the particulate emiss-
ions.
Figure 6-5 shows the ratio of sulfur trioxide (SO ) to total sulfur
oxides (SOx) is the typical 1% to 2% conversion, except when the SOx con-
centration dropped below about 500 ppm. This increase at low total sulfur
concentrations has been further investigated, because there appears to be
no theoretical mechanism that would increase the proportion of sulfur trioxide
at low total sulfur concentrations. A possible cause of the higher values alt-
low concentrations of sulfur oxides is experimental error due to overtitration.
This experimental error in titration endpoint of the standard Shell-
Emeryville method is always positive, and the effect is greater at low SOx
levels. Therefore, experimental tolerance may be responsible for the
apparent trend. A modified titration procedure will be developed for use
during Phase II.
There appears to be no strong effect of fuel type other than its
sulfur content. For example, No. 6 oil data are shown between 500 ppm
120
-------
o
•H
-P
-U
o
EPA Regulation
For Utility
Sized Boilers
0 2 46 8 10 12 14
Fuel Oil Carbon Residue, %
Figure 6-3. Effect of Fuel Oil Carbon Residue on Base Load Particulate
Emissions.
6000-28
121
-------
2400
e 2000
a
a.
ro
(Si
c
o
•rH
-P
to
M
-p
c
0)
o
8
in
o
-o
•H
X
o
1600
1200
800
400
./9\
0.0
/
i l
Typical Type of Coal
100% Conversion
/
Typical Type of Oil
100% Conversion
Numerals within Symbols
are Test Numbers.
FUEL TYPE
A
O
Coal
1.0 2.0 3.0 4.0
Fuel Sulfur Content, Dry, %
5.0
Figure 6-4. Total Sulfur Oxides Emissions at Baseload For Oil and
Coal Fired Boilers
6000-28
122
-------
W
OMIZATION TY
H
A
a
J
U)
»a
Oj (N
M =tt=
-p tH
cd *H
2 0
ne
O (
6 C
N
IO vO
H r-l
•H -H
0 0
0®
D C
D il
H
jreader
ilverizer
>-M ^
W pj
H r-l
n3 nj
8 8
<^<1
3 C
J ^
•\
iderfed
^clone (
P u
0} «»
8 8
<3«3
o
vv
D C
J1 r
H
C-
V_
3 C
*
xl C
-i T~
\ e
^
3 C
3 0
-(
|
3 C
0 •£
^
c
<5
:
.]
•
rm
^
e
o
CN
^
?
ll
^
^
e
o
0 32C
1 Sulfur
o m
00 4J
CN 0
H
4-1
O
O
0 -H
0 E 4J
•^ ft U
CN ft C
fa
CN
O n3
T3
O rd
>00 2G
itration, dry
ies at Basel c
DO ie
)xides Concer
L Sulfur Oxic
CN 03
r-l M 4J
3 O
M-j EH
'S O
U) 4J
rH U3
(0 0)
0 4-1 T)
0 8(
T<
Sulfur Tjriox
asured.
O 0)
"31 'W S
0
in
O 0!
•H T3
4-> -H
03 X
« O
O
lf>
1
vD
s jo
123
-------
I
and 1500 ppm and the SO /SOx decreases with total sulfur oxides just as with
the other fuels. For coal the type of coal burner has no significant effect
on the SO /SOx ratio in the exhaust gas.
124
-------
SECTION 7.0
BOILER DESIGN CHARACTERISTICS
Although the design of existing boilers cannot be adjusted day to day,
the influence of boiler and burner design on emissions is of interest in
terms of new unit design and potential modification of existing units.
The major influences are expected to be in burner design (degree of mixing,
ingestion of recirculated gases, atomization, etc.) and the rate of heat
loss from the flame (burner face cooling, burner spacing, furnace area,
furnace volume, etc.). The specifications of the boilers tested are listed
in Table 7-1.
The Phase I data presented in this report have been evaluated in
terms of determining boiler design parameter effects on pollutant emission
levels, especially NOx emissions. These results are limited by the data
sample size but do provide interesting trends.
The boiler design characteristics discussed in this Section are
considered on an individual basis only. The regression analysis dis-
cussed in Section 8 considers simultaneous interactions of these parameters.
The results are presented to indicate trends in the data and define the more
important parameters influencing NOx and other pollutant formation. The
individual relationship presented will vary slightly if the data were
normalized for each of the other important parameters.
7.1 BURNER DESIGN
7.1.1 Oil Atomization
Oil atomizers evaluated during the program consisted of steam, air,
pressure - mechanical, and rotary cup - mechanical atomizers. The No. 2
oil burners were evenly divided between steam and air atomized, with one
test conducted using a pressure-mechanical atomizer. The No. 5 oil burners
were divided into about one-fourth steam atomized, one-half air atomized,
and the remainder rotary cup-mechanically atomized. Over three fourths
125
-------
tn
u
H
B
10
a
r^ g
Q) Z
rS O
H
O
m
'
t
I
s
3
C
1
M
\
i
X
8
I
u
•
2
It?
3JE
i!s
c
Hi
M *J •
M M C
M • —
13 S
« Q -
IU
0
£
9
£
** ••*
.
-. .< s
,>.
Jii
Is
4 «*«
£ £
* -
!i*
**
2 i
» u
1
i1 M
«
3
X
t
1'
I*
i issssSsill
SaaaaSSaSa
oor-,Q1Qr''i,£ir- 200
OOfN"Gj|r-EjOOCO
inmoiOOm O o O m
.H rH <-l (N Pi (N in CO O f*l
1 1 1 1 1 O CD 1 \D O
1 I i I I ^r i M
l i i i i .-H o i r- o
1 i t t i r> i f> n
•a
aJ o .0 C T3
^3Q>- jrK'OfttX'O
OOO OOOOO
m ^ ° in M
000 0 0 .A ^ in 0
OOO O O ^O *•••—*(""
cvcr>>/> i coooC'Cov
04^0 in o o o o o
asiiSlsS^p
mi i ao H o -i m
1 1 1 1 1 -H 1 -H
co co o i i *r r- i <-« i
EB-SBEBSSi:
EtEKfiKSifiSB
in in oi
*H
Illiililsi
o o i t o i i D itiiiiii i i i i i > i i i i
llOO^lllvOtl \O O CD 1 CO I \C *£) ^ 1 (js 1 ( 1 t 1 1 1 t Illlllill)
livmilr-lt •» v£. -CT l •* I fo ?n CO imiillitii li II
1 _,
II
a>>>, oo w>i "^"""^'S^RS ^I~l2^Et^f'^£m
" * " 5GS '**§ * ^ oit^»tj> Itrtn
1 u
,^gggS23S3 ,o£«gicoo>og S8;is£feg§zS ^8S2^jSzz
0 0 0 0 tn 0; 1_ 0! « j> en o 0 0 <0 ^ 0 CD o CO r; to ^ « t- Chr-^r- ooiooococNcomo mcorjtTvOr-co'^^'J' rM "^in in
moooooooo ooooooounoo ooc:-l*^j3Jlt*"j2ijJ ® mS S '
^4 ,-4 tv pH O4
i 1 t •-« 1 1 1 1 1 1 1 1 1 O^ t 1 I I » I 'O _O I ! m ' ' '
ll^tOinin^HfvjM) r4tNvDl^r-r-o-H04 ^i ^ -M ^^ S CS t 1 S ^ ' ' S 1
li'7'*i3'i*'i<'|i't*'l4V iiimiiiiii |O'IQQI'Q^ Qlloi''!:!1
SS^S^^S P^feS^S^ SEfcSEEEEEfc £££££££££_
!gggsse5se esssssssee EBecccccet tccctEctt
ICN.OvO^^vor.ltnr- SS^SSSSSSS SS" "»-"•<« rt-< -1* -1
•^OOCD^CICOIJI^ §S5™^SSSS? SSSS°SS§^S S55f^SSSiP!S
O^ Mfl >>
UNCN "OIN rf •» CNCO
CO n n n > ^
j^eMomiQiq.^jH^ toeDcom^c.iw^^cn H°'"cNCMlQin'rcMr^ rJcN
-------
co
u
M
EH
CO
2
w
EH
tj
i a o>
0 3
a -H
O 4J
H C
W O
§ Si
PS
w
hH
H
O
m
ti
w
w
EH
r
fr
2
U
L
*
j
e)
?
a
£
s
£
>
„
fc
y
S
n, j~
* •*
«. tJ" —
•^ v- _>
* •-•
~
U^
> a --
S .: ^
3
5
« S
H IH
j
j V.
.?";
? ~
> ^
r:
•c «
?i 31 s
u J
? ,°
t
5,
•9 ~
v r
o
*j
3
D>
•••••IB"-
3
-
!55SSSsSSi SSISSSKIS SS! sIlSSS
IIIIIIIIIIIIIS^III HI1§i§ll
SISIISIHIiSSIISSIS |||s|il!i
mmi/ioooooo ooooooooo oooo*-* VD^D
^ (N fN r-t
O O O OOOO
itiiiiiiii icoioiieocoi ooivocoi
t i I i I l ) I i I i r-t i l l CN rs i l rn oo uu+isiotio
i/i(N(NfN(NcNfMS*. r^25 (N v" (N M "N K ^ 11 wi "^ "^ f1^ i/l S2 1 5K O S5
^T"Tij>^T^HlrH^c*\^j'^ OOOOOOOOO (NfN OCDIQOO
<*SS"'*><^*n'1r-i^'>0'*'T'»ino'io^^roo-» "'"i mo loom
„ „ mr,M rtrt ,n « OOM
sSSs-li^ssSs^pSss? S£ sliSp
r-J fl -H<-H(N (MCO'COCD
1 f O5 CO * * *-» OO IIJ(l
-------
of the No. 6 oil tests were with steam atomized oil guns, the remainder
being air atomized. All No. 2 oil atomizers operated with ambient tempera-
ture oil at the burner. The oil and steam/air pressures at the burner
varied from unit to unit; but typically, steam/air pressure was about
50 psig, and oil pressure was about 40 psig at top load. The No. 5 oils
were normally fired at from 160 to 180°F at the burner with steam/air and
oil pressures similar to the No. 2 oil atomizers. The No. 6 oils were
normally fired at approximately 200°F at the burner, and the steam/air and
oil pressures at the burner were similar to No. 2 and 5 oil atomizers.
The baseline NOx emissions for the No. 2, 5 and 6 oil tests were
not dependent upon atomization techniques when the oil atomizers were
operated near their design conditions. As discussed in Section 5.6,
lowering the fuel oil temperature at the burner from its normal value
increased NOx emissions. A single test (Test No. 2), where the atomizing
air pressure at the burner was reduced while the oil flow rate and pressure
were held constant, resulted in about 50 ppm reduction in nitrogen oxides
emissions. The atomizer may have been operating with too high an air
pressure and lowering the air pressure produced the proper momentum ratio
of the two streams for best atomization.
A special series of tests, Tests. No's. 1, 2, 52, 53 and 54, were
run at Location 19 to investigate the effect of the oil atomization method
and oil grade on the total nitrogen oxides and particulette concentrations.
The boiler used was a Keeler Company packaged steam generator rated at
17,500 Ibs/hr steam flow and was installed in 1970. The furnace ceiling
and side walls consisted of tangent-wall tubes with a tile floor and burner
wall. This saturated steam boiler operated at a nominal steam pressure of
150 psig. During this test series, both No. 6 and No. 2 fuel oils were
tested with steam and air atomizing oil guns and No. 2 fuel oil was also
tested with a mechanical-pressure atomizing oil gun. Ambient temperature
combustion air was used in all tests. The measurements are summarized in
Table 7-2 and Figure 7-1. It should be noted that the No. 2 and No. 6 oils
used for these tests were the extremes in API gravity, carbon residue, ash,
128
-------
TABLE 7-2
EFFECT OF OIL ATOMIZATION METHOD ON TOTAL NITROGEN OXIDES,
PARTICULATE EMISSIONS AND BOILER EFFICIENCY
Test
No.
I3
2
44
45
52
534
54
56
57
Fuel
#6
#6
#5
#5
#2
#2
#2
#2
#2
Atomiza-
tion
Method
Steam
Air
Air
Steam
Steam
Air
Mech.
Air
Steam
Steam
Flow
(klb/hr)
14
15
17.6
17.3
14
14
12
15.9
15.7
Normal
Excess Oxy-
gen (%)
3.6
4.4
7.2
6.7
3.6
3.0
4.3
8.0
8.0
2
NOx
(ppm)
350
334
177
161
65
97
80
116
118
Partic.
Ib/MBtu
0.1524
0.2910
0.0653
0.0779
0.0378
0.0164
0.0194
Boiler
Efficiency
(%)
85
85
86
86
85
85
85
85
86
1. Normal operating O level defined by burner manufacturer.
2. ppm is measured value corrected to 3% excess 0_ dry.
3. Particulate data for Test-No. 1 were taken far low air run
(2.3% oxygen).
4. Particulate data for Test No. 53 were taken for high air
run (4.3% oxygen).
6000-28
129
-------
I
04
o
oo
O
•rH
-P
-P
C
(1)
O
C
O
U
w
0)
-0
O
C
a)
en
o
40^
300
250
200
150
100
50
0
1
Figi,
Test No. 1
Steam Aton
No. 6 Oil
X
Test No.
Air Atom
No. 2 Oi
— --
Test 11
Steam
No. 2
lization
'
53
ization
1
-—- <
—
0.52
Atomization
Oil
-------
nitrogen, and sulfur (see Table 6-1). As a result, relatively high NOx
and particulate values were measured for the tests with No. 6 oil and low
values were measured for No. 2 oil.
Test No. 1: Steam-Atomized No. 6 Fuel Oil. The steam-atomized
oil burner used for this test operated at the baseline load with oil
pressure and temperature at the burner of 75 psig and 200°F. The oil was
atomized by steam impingement within the atomizing tip and injected into
the furnace through burner tip orifices, which were similar to the common
B&W Y-jet atomizer design.
As shown in Figure 7-1, the NOx emissions increased with increasing
excess 0 up to about 5% excess 0 where a maximum NOx value of 380 ppm
4!~ ^-
was reached and beyond this 0 level the NOx emissions decreased with
increasing excess O . The minimum excess 0 level, below which incomplete
combustion occurred, as evidenced by excessive CO emissions and a visible
smoke plume, for this test was 1.6%. Particulate emissions of 0.1524
lbs/10 Btu were measured for the low air Test Run No. 1-11, which is one
of the higher emission levels recorded for steam-atomized No. 6 fuel oil.
Test No. 2: Air-Atomized No. 6 Fuel Oil. At the baseline steam
flow of 14,200 Ibs/hr the oil pressure and temperature at the burner were
37 psig and 214°F and the atomizing air pressure at the burner was 30 psig.
The NOx emissions increased with increasing excess 0 over the range inves-
tigated from 217 to 6.6%. The flame appearance changed with excess O ,
and the best flame characteristics occurred at the lower O levels. Parti-
,- ^
culate emissions of 0.2910 lbs/10 Btu were measured for Test Run No. 2-6,
which was substantially greater than the values obtained with steam atomi-
zation on Test No. 1.
Test No. 52: Steam-Atomized No. 2 Fuel Oil. The steam-atomized
oil burner used for this test at a steam flow of 14,000 Ibs/hr operated
with 65 psig pressure, ambient temperature oil and the steam pressure
at the burner of 73 psig. The NOx emissions increased with increasing
excess 0 up to about 4%, and between excess O levels of 4 and 5% of
131
-------
the NOx emissions appear to reach a maximum value. A visible haze from
the smoke stack occurred at the lowest level of excess oxygen. Parti-
culate emissions of 0.0378 lbs/10 Btu were measured for this test, which
is about average for steam-atomized No. 2 fuel oil.
Test No. 53: Air-Atomized No. 2 Fuel Oil. At the baseline steam
flow of 14,000 Ibs/hr the oil burner operated with 27 psig oil pressure,
ambient oil temperature and 23 psig atomizing air pressure. The NOx
emissions increased with increasing excess 0 up to about 4.0% O beyond
which the NOx was relatively constant at 101 ppm. Particulate emissions
were 0.0164 lbs/10 Btu, which is one of the lower values for air-atomized
No. 2 fuel oil.
Test No. 54: Mechanically-Atomized No. 2 Fuel Oil. The mechanically-
atomized oil burner used for this test operated with ambient temperature fuel
oil at a burner pressure of 280 psig for a boiler steam flow of 11,500 Ibs/hr.
The NOx values did not vary significantly over the excess O range investi-
6
gated of 3.7 to 6.6%. Particulate emissions of 0.0194 lbs/10 Btu were
measured, which is one of the lower values measured for No. 2 fuel oil.
The No. 6 oil data presented on the upper part of Figure 7-1 show
steam atomized fuel oil burners to have slightly higher NOx emissions than
air atomized burners for normal operating excess oxygen levels. As the
excess O level is increased, both of the NOx emissions increase until, at
5% excess O , the NOx emissions for steam atomization are less than for air
atomization.
The NOx emissions with No. 2 fuel oil were not very sensitive to
excess oxygen. Air atomization resulted in the highest NOx emissions (100
ppm) with steam atomization being the lowest NOx producer (70 ppm). The
mechanically atomized No. 2 fuel oil tests were conducted at a reduced
load and yielded NOx emissions greater than the steam, but less than the
air-atomized data.
132
-------
The boiler efficiency did not vary measurably duo to use of different
oil and atomizers.
The particulate emissions for both the No. 6 and No. 2 fuel oil
tests were inversely related to the NOx emissions. For No. 6 fuel oil,
air atomization resulted in the lowest NOx emissions at the normal
operating 0_ level, but yielded substantially greater particulate emissions
than did steam atomization. For the No. 2 fuel oil tests, steam atomiza-
tion resulted in the lowest NOx emissions and yielded the greatest parti-
culate emissions. The air atomization test on No. 2 fuel oil had the
greatest NOx emissions and yielded lower particulate emissions than the
steam atomized test. Mechanically-atomized No. 2 fuel oil NOx and par-
ticulate emissions were in between the air and steam results.
A second special series of tests, Tests No. 44, 45, 48, 56 and 47,
was run at Location 26 with No. 2 and No. 5 oils with both steam and air
atomization. In Tests No. 56 and 57 with No. 2 oil, the NOx emissions
listed in Table 7-2 for air and steam atomization were the same, whereas
for Test 52, steam atomization produced significantly less nitrogen oxides
emissions. With No. 5 oil in Tests No. 44 and 45, the emissions with air
atomization were greater than with steam, rather than less, as for Tests
No. 1 and 2 with No. 6 oil. It appears the NOx emissions depend on atom-
ization and mixing characteristics of a given burner. The mechanism for
atomization (steam, air, or mechanical) may be less important than the
degree of atomization achieved. Evaluation will continue in Phase II.
Tests No. 3 and 36 were run on a rotary cup type atomizer firing
No. 5 and NSF oil, respectively. Although rotary cup oil burners once
were commonplace, now they are becoming rare. The total nitrogen oxides
concentrations were somewhat high for oil-fueled boilers of this small
size, but not seriously so. The particulate emissions were slightly less
than those of boilers burning No. 6 fuel oil.
133
-------
800
o
ro
a
o
•H
4-1
0)
o
o
u
w
01
0)
-P
•H
(0
4-1
O
700
600
500
400
300
200
100 H
Coal Fuel
Numerals within Symbols
are Test Numbers.
\Gas Fuel
Preheated Air
\ \
Figure 7-2 .
100 150 200 250
Rated Burner Heat Release, MBtu/hr/Burner
Effect of burner heat release level on total nitrogen oxides
emissions for coal and natural gas fuels.
6000-28
136
-------
upon burner heat release rate, the magnitude of which depends on combustion
air temperature as discussed in Section 5.2.
7.1.4 Burner Heat Release Rate
The nitrogen oxide emissions as measured during this test program
were generally found to increase with increasing burner heat release rate.
This dependence of nitrogen oxides emissions on burner heat release rate is
different for each of the fuels tested. Coal fuel burning equipment sometimes
can not be defined in terms of individual burners; however, pulverized coal
burners and cyclone furnaces are similar to oil and natural gas burners in
that a certain portion of the fuel and air enter the furnace through a
burner port.
The nitrogen oxides emissions versus burner heat release rate for
the natural gas and coal-fired boilers are presented in Figure 7-2. The
natural gas burner data show a much lower dependence of nitrogen oxides
emissions on burner heat release rate than the coal burners. The natural
gas fuel data for ambient temperature combustion air show less dependence
of nitrogen oxides emissions on burner heat release rate than do the pre-
heated combustion air data. The coal fuel data show a strong dependence of
nitrogen oxides emissions on burner heat release rate. Figure 7-3 presents
the effect of burner heat release rate on nitrogen oxides emissions for all
of the oil-fired boilers tested. The two data points for No. 5 oils which
have NOx emission levels greater than 400 ppm are from tests where the
fuel ©il was not heated, but was near outside air temperature. Atomization
was poor, and they are not considered to be representative data points.
The effect of burner heat release rate on nitrogen oxides emissions is not
as great as previously discussed for coal fuel, but is greater than for
natural gas burners with or without preheated combustion air. The type of
atomizer did not seem to affect this relationship. The No. 2 oil burners were
smaller, all being below 50 x 10 Btu/hr, and defined the lower region of
the oil data. The No. 5 and No. 6 oil burners included the complete range
of burner size investigated from the smallest up to 125 x 10 Btu/hr.
135
-------
800
o
c
0
•H
-P
(0
n
-p
c
0)
o
c
o
u
w
0)
•o
•H
c
OJ
Dl
o
M
-p
•H
2
700
600
500
400
300
200
100 -i
Coal Fuel
Numerals within Symbols
are Test Numbers.
\Gas Fuel
Preheated Air
\ \ /
WX
Gas Fuel
Ambient Air
\
Figure 7-2 .
'50 100 150 200 250
Rated Burner Heat Release, MBtu/hr/Burner
Effect of burner heat release level on total nitrogen oxides
emissions for coal and natural gas fuels.
6000-28
136
-------
800
700
I
ca
o
•H
-P
(0
M
-P
§
O
c
8
0)
T3
•H
X
O
c
0)
tjl
0
M
-P
•H
2
•U
0
EH
600
500
400
300
200
100
Atomizer
Symbol
O
• o
<0
V
Numbers within symbols
• refer to oil grade.
Steam
Air
Mechanical
Rotary Cup
50 100 150
Rated Burner Heat Release, MBtu/hr/Burner
200
Figure 7-3. Effect of burner heat release level on total nitrogen
oxides emissions, oil fuel.
6000-28
137
-------
7.2 FURNACE DESIGN
7.2.1 Firetube Versus Watertube Boilers
The firetube boilers tested during the program had furnaces of
varying diameters and lengths. Some of the boilers had furnace tubes
partially submerged in the boiler water, and some of them were of the
"wet back" design, where the furnace backwall is water cooled for added
heat transfer surface. The largest capacity firetube boilers tested were
20,000 Ibs/hr steam flow. All firetube furnaces were either saturated
steam boilers or pressurized hot water heaters. The watertube boilers of
the same size ranee as the firetube boilers varied widely in design, but
all had a single burner with a refractory burner face.
Figure 7-4 presents the nitrogen oxides emissions data for all of
the firetube units tested, as well as the smaller watertube units (below
30,000 Ibs/hr steam flow). The nitrogen oxides emissions are practically
the same for both types of furnaces. Test No's. 1, 2 and 34 were with No.
6 fuel oil and the nitrogen oxides emissions were between 300 and 350 ppm.
Tests 35, 36, 44, 45, 46 and 51 were with firetube furnaces and No. 5 fuel
oil and the nitrogen oxides emissions varied from 160 to 275 ppm. Test
No. 3 was with a watertube furnace and No. 5 fuel oil and the nitrogen oxides
emissions were 200 ppm. The remaining oil fuel tests were all with No. 2
fuel oil and the nitrogen oxides emissions varied from 65 to 195 ppm and
were independent of furnace type. The nitrogen oxides emissions for natural
gas fired firetube furnaces varied from 55 to 105 ppm and for the watertube
furnaces varied from 70 to 100 ppm. Two small underfed stoker coal-fired
firetube furnaces were tested (Tests 42 and 43). The nitrogen oxides emis-
sions were 275 and 345 ppm for these units. Coal-fired watertube furnaces
of this small capacity are rare and none were tested during the program.
There does not appear to be a significant difference between NOx
emissions from firetube and watertube boilers when burning the same fuel.
138
-------
Oil
10 20
Test Load, k Ibs/hr
Figure 7-4. Baseline nitrogen oxide emissions of firetube and small
watertube boilers.
6000-28
139
-------
7.2.2 Furnace Volume and Area
Nitrogen oxides are formed at high temperature by the combination
of oxygen and nitrogen, and the length of time that the products remain at
high temperature is critical to the formation of nitrogen oxides. The fur-
nace volume and area were evaluated as design parameters which could influ-
ence the time/temperature history.
The furnace heat release volume was defined as the furnace volume
from the burner to the end of the furnace divided by the combustion
heat release rate, i.e., heating value of the fuel times the fuel flow rate.
This parameter, in terms of ft /MBtu/hr, has been calculated for each baseline
test conducted during this program whenever sufficient furnace geometry infor-
mation was available and is presented in Table 7-1.
Figure 7-5 presents the nitrogen oxides emissions versus furnace
heat release volume for all natural gas-fired furnace tests. The nitrogen
oxides emissions were not dependent upon furnace heat release volume for
these tests. The combustion air temperatures are included in Figure 7-5
for all tests with preheated combustion air. The other tests were all with
ambient temperature combustion air. The NOx emissions for the ambient temp-
erature combustion air tests were all between 50 and 100 ppm independent of
furnace heat release volume. The preheated combustion air tests were all
between 200 and 375 ppm of nitrogen oxides, and like the ambient data were
independent of furnace heat release volume.
The nitrogen oxides emissions versus furnace heat release volume for
all oil-fired watertube furnaces are presented in Figure 7-6. The firetube
furnace volumes (iot shown in Figure 7-6) were all less than 10 ft /MBtu/hr
and the nitrogen oxides emissions varied from 125 to 300 ppm depending on
the fuel nitrogen content and did not depend on furnace volume. The differ-
ence in the No. 2 oils as compared to the No. 5 and 6 oils is due to the
lower fuel nitrogen content characteristic of No. 2 oils. The two No. 5 oil
fuel tests which showed greater than 400 ppm NOx emissions were the tests
conducted with nearly ambient fuel temperature at the burner instead of the
160 to 180°F typical for No. 5 fuel oils.
140
-------
a.
t)
d
•S
+J
nj
a)
u
c
o
u
in
0)
TJ
•H
X
O
d
a;
o
E-i
suu
700
600
son
A r\r\
300
POO
100
39-
0
03
& ^
a®
22° F
1
i
i
Numerals within Symbols are
Test Numbers. Temperatures
Adjacent to Symbols are Air
Preheat Temperatures.
1
, 400 °F
J
1] 400° F
3
._, 640° F
13
-,650°F
3
20 40 60
Heat Release Volume, ft / MBtu/hr
80
100
Figure 7- 5 . Effect of furnace heat release volume on total nitrogen
oxides emissions for natural gas,.
141
6000-28
-------
800
O
c;
o
u
to
\\ Met
V Roi
Number wi
""refers to <
®e
(1
amizer
?air>
tr
^hanical
tary Cup
thin symbol
ail grade. ~
8
9
©
(Q\
»
d
©
©
©
20 40 60 80
Heat Release Volume, ft3/MBtu/hr
100
Figure 7-6.
Effect of furnace heat release volume on total nitrogen
oxides emissions for oil-fired watertube furnaces.
6000-28
142
-------
Figure 7-7 presents the nitrogen oxides emissions versus furnace
heat release volume for the coal-fired watertube furnaces. As discussed
in Section 7.1.2, coal burning equipment design varies greatly. The
larger furnace heat release volumes are for older boilers with underfed
stoker coal burning equipment and had the lowest nitrogen oxides
emissions. The underfed stokers burn very large coal particles when compared
to spreader stokers, pulverizers and cyclone furnaces. The larger particles,
because of the surface area to volume ratio, burn much slower, providing more
time for furnace gas recirculation into the flame zones and require large
furnaces for complete combustion.
The majority of the coal-fired data presented in Figure 7-7 are for
spreader stoker and pulverized coal burners. These coal burning equipment
designs require less furnace volume for complete combustion of the smaller
coal particles and sometimes include overfire air and/or steam injection
for added turbulence within the furnace. They produce more intense combustion
zones and have higher nitrogen oxides emissions. The smallest furnace heat
release volume and most intense combustion zone was the cyclone coal combustor.
This design utilizes high air velocities and a swirling flow pattern within
the small combustor to achieve complete combustion of the coal fuel. The
molten slag produced by the high bulk gas temperature acts as insulation,
helping to produce more of an adiabatic combustion zone. The nitrogen oxides
emissions for the cyclone furnace combustor were the highest values measured
for all tests conducted during this program.
Another furnace geometry parameter related to nitrogen oxides emissions
is the furnace heat release area, which is the wall area for heat release from
the flame zone. The data from this program were plotted by defining this area
as the surface area corresponding to the heat release volume. The correlation
was similar to but not as well defined as that obtained with the furnace heat
release volume. Table 7-1 presents the furnace heat release area parameter
for each baseline test for which the required furnace geometry information
was available.
143
-------
Total Nitrogen Oxides Concentration, dry @ 3% Q^, ppm
|— 'to U) ib. U1
-------
The furnaces tested during this program had many different types of
wall construction, from tangent tubes on over two-thirds of the furnace walls
to widely-spaced tubes with large areas of refractory surface. The older
boilers had the greatest portion of refractory furnace walls but also had the
larger furnace heat release volumes. Table 7-1 lists the wall construction
for each furnace tested. Unfortunately, the wall construction is much more
complicated than simply tangent tube (TT), welded fin (WF), refractory and
tube (RT), or refractory (R), because each wall of the furnace may be of dif-
ferent tube spacing or material. During Phase II of this program, more emphasis
will be placed on wall construction and heat transfer effects.
7.3 BOILER EFFICIENCY
Gas-fired units in the watertube category tended to have a larger range
of efficiency than gas-fired firetube units. The latter maintained an efficiency
value between 80% and 83% throughout their size range for baseline conditions.
Watertube units fired at baseline conditions showed an efficiency range
between 77% and 85%, with larger units having the higher values.
Both firetube and watertube units firing number 2, 5, and 6 oils
had higher efficiency values than similar units firing natural gas due to
lower water losses. The efficiency ranged from 84% to 88% for firetube and
80% to 87% for watertube furnaces. The higher efficiencies coincided with
significantly lower excess oxygen level in the flue gas.
Coal-fired boilers showed no major efficiency differences based on
the type of burner employed, although pulverized and spreader stoker units
had slightly higher efficiencies than did the under-feed stokers. The cyclone
boiler efficiency was the highest measured due to the use of an air preheater
and low excess air firing.
High exhaust gas temperature results in lower efficiency. High stack
losses can result due to boiler design, improper equipment maintenance, or
high excess air operation. Older boilers typically are about 3% less
efficient than are newer boilers. In older firetube units, high stack
145
-------
temperatures are caused by obstructions in the firetubes that interfere
with the transfer of heat from the gases of combustion or by an insufficient
number of passes within the boiler shell before the gas is exhausted. Older
watertube units usually do not have an air preheater or feedwater economizer,
and this decreases their efficiencies.
One benefit of reducing the excess oxygen/air to lower the nitrogen
oxides emissions was that the boiler efficiency was increased. The increase
was due to less heated air being exhausted up the stack. This effect is
illustrated in Figure 7-8 that plots the change of boiler efficiency for
natural gas fuel in over a dozen instances where the excess oxygen was changed.
The increase in efficiency varied from boiler to boiler, but in general,
the efficiency increases about 0.5% for each 1% decrease in excess oxygen.
A plot for oil fuel is similar in appearance and the increase in efficiency
is slightly larger, about 0.6% for each 1% decrease in excess oxygen.
Reducing boiler load generally caused the efficiency to decrease.
See, for examples, Test Runs No. 32-4 and 32-2 in Table 4-1. The decrease
mainly was due to the higher excess air utilized at lower loads and the
resulting larger volume of heated air going up the stack.
Low NOx operation achieved by taking burners out of service or
resetting the air registers had various effects on boiler efficiency. As
Figure 7-9 illustrates, the efficiency decreased in seven out of twelve
instances. The points for burner-out-of-service runs are marked "BOOS" and
for register adjustment are marked "Reg." The decrease in efficiency for
BOOS runs was caused by the increase in excess air that was required. In
those BOOS cases where efficiency increased, no increase in excess air was
necessary. Efficiency increases, while firing oil fuel, were more frequently
encountered on un.xts using steam atomization.
146
-------
0)
c
0)
Cn
>.
X
o
to
a)
O
X .
w
•H
Q)
•H
O
•H
<*-!
in
W
c
o
•H
-U
(0
-H
CO
I
r-
0)
I
•H
fc.
O
0>
O
CD
147
-------
u
C
Q)
-H
U
•H
44
W
O
CQ
d
o
•H
p
rrf
(U
a
o
X
o
-p
o
0)
w
^
0)
en
XON
'AousTDtjgpg
148
-------
SECTION 8.0
STATISTICAL ANALYSIS OF DATA
The data analysis was confined to a statistical evaluation of data
collected from 47 boilers at 26 different locations in the conterminous U.S.
Although numerous data were collected, the analysis reported herein was
confined to a study of NOx emissions at baseload for 66 "complete" sets of
data; i.e., more than one set of data was collected for several of the 47
boilers studied by KVB. A set of data consists of measured boiler, fuel,
and operating parameters assumed to be related to NOx emission levels.
The objective of the statistical analyses was to investigate the
extent that a general unifying theory could be developed from the data by
constructing a functional relationship between NOx and a set of boiler
operation and design parameters. For instance, could a general quantitative
relationship be developed for all boiler data or must the data be separated
into strata based on fuel type, etc.? The emphasis of this section is to
explore the degree that boiler data can be lumped. Individual boiler properties
are discussed in other sections of the report. Another objective was to
isolate significant causative parameters and recommend improved data
collection procedures for subsequent phases of the overall project.
8.1 ANALYSIS STRATEGY
The basic view taken here is that a deterministic model quantifying
NOx emissions can be constructed by conceptualizing the appropriate variables
entering into the processes producing NOx and constructing a relationship
by a statistical evaluation of observed parameters. A deterministic approach
recognizes at the outset that NOx emissions are governed by nonrandom
factors. A purely probabalistic approach assumed beforehand that the
process of interest, NOx formation, is governed by random processes.
There are several standard statistical methods for evaluating
random data. Although all statistical methods proceed from the assumption
of randomness, there are certain fundamental differences between the various
149
-------
statistical techniques, depending on the analysis objectives and the nature
of the data.
A widely used technique for analyzing statistical data is analysis
of variance which consists of testing various hypothesis of causality and
dependence or interaction between the data. Analysis of variance techniques
consists of latin squares analysis, incomplete block design, and so forth.
However, a key precept of such procedures is that the data be collected in
an unbiased or random manner; e.g. selection of population samples by
chance.
In the case of the data analyzed herein, NOx emissions from industrial
boilers, it was impossible to select boilers for sampling by a random
selection process. Instead, candidate types and sizes were selected followed
by a test-as-available selection criterion for practicality and economy.
Consequently, the data may be biased or skewed since "clean" or "dirty"
boilers may inadvertently have been selected. Generally, with a small
data sampling it is not possible to determine if bias exists. The only
valid method for proving the goodness or badness of the data set and hence
the analysis is to verify the data by additional emperical evidence.
The method of analysis appropriate to the data analyzed here consists
of regression analysis, an alternative statistical analysis method to analysis
of variance. Re9ression analysis is a powerful statistical tool that not
only permits a quantitative relationship to be derived but also permits an
evaluation of the degree of data interdependence or interaction. Moreover,
it is not generally necessary to assume randomness in the data. Multiple
regression analysis also permits the evaluation of several effects at the
same time; i.e., multidimensional models can be examined.
The form of regression analysis used here consists of multiple
linear regression which has been widely used in other disciplines. Multiple
linear regression consists of fitting a set of data to equations of the
form
Y - ao + Vl + V2 + •'• +Vn (1)
150
-------
where Y is the dependent variable and X , X ,.. X are a set of variables
selected as the independent variables. The regression coefficients of best
fit, a , a ,...a are determined by a set, N, of observed Y's and X.'s.
0 1 n i
For strict statistical purposes, N should be infinite and the Y's and X.'s
should be random. However, in practice and particularly for the case con-
sidered here, N is finite and the Y's and X."s are nonrandom. This latter
problem is handled by writing
Y = a + ax, + a^X. + ... + a X + £ (2)
0 11 22 n n
where e is an error due to measurement of the Y's and X.'s and the
i
incompleteness of X.'s; i.e., all X.'s that determine Y were not observed
since some are unknown. It is reasonable to expect e to be a random variable
containing only random errors.
It is common to transform the Y's and X.'s by various transformations,
usually common logrithms. The advantage of doing this is to account for a
curvalinear fit of the data since oftentimes a simple log-linear relationship
is obtained between the variables, the marginal distributions of the trans-
formed variables more closely approximates a normal distribution (e.g., bias
is minimized), and the variance of points along the regression plant (N+l
dimensional) is more homogeneous (i.e., the standard deviation is stabilized).
Taking logrithmic transformations, (2) becomes for purposes of linear
regression analysis
log Y = Bn + Bnlog Xn + Bnlog X_ + ... + B log X + 6 (3)
01122 n n
= log a(
6 = log e
where B = log a
or
B, B B0 B
Y = 10 °. x. X. X, 2. . . X n « 106
12 n
o
where 10 is a measure of the error associated with the individual data.
There are a number of statistical parameters of fit that can be
used to make judgements of the quality of the fit (i.e. the quality of
the model). These will be discussed below.
151
-------
The multiple correlation coefficient is an mdic.it jon ol t lu- ijooilno:..!*
of fit between the independent variables and the dependent variable. A
coefficient of one indicates a perfect fit, and a coefficient of zero
indicates no fit.
The coefficient of determination is a measure of the amount of
variance accounted for by the regression. For example, a coefficient of
determination of 0.923 indicates that 92.3% of the data variance in the Y's
and X.'s is accounted for by the regression.
One of the powers of regression analysis is the ability to estimate
the error of the model. Since no model is exactly the same as the prototype
(note that all engineering equations are models), there is always some error
or uncertainty associated with the model. This error is indicated by £
or 6 in the above equations. This error can be approximately estimated by
using the standard error of estimate; i.e.
£ ^ Sy /(I -
where S is the standard error of estimate for the regression using Y as the
dependent variable and R is the coefficient of determination (R is the
multiple correlation coefficient). The variable t is the value of the
Students t distribution at an a confidence level at a certain degree of
freedom.
The level of confidence in the multiple correlation coefficient can
be judged by Fisher's F-ratio. This statistic is the ratio of the variance
of estimate of the regression equation to a chance variance, and indicates
the probability of an actual relationship.
The significance of the individual fit parameters (i.e., the
independent variables X.) can be measured by Student's t statistic. The
t-test is a measure of the level of importance of the individual X. in the
regression equation. Parameters with a low probability of significance are
eliminated from consideration and only those with a high probability of
significance level are retained.
152
-------
8.2 REGRESSION ANALYSIS
Initially, tests were conducted to determine the quality of regression
fits using simple linear multiple regressions and logrithmic transformed
multiple linear regression. The results of preliminary studies suggested
that in some cases lograthmic transformations improved results while in
other cases such transformation did not make a significant difference.
Consequently, all regression analysis were based on logrithmic transformation
of the variables; i.e., equations of the form
log Y = B + B log X + B log X + ... + B log X
were used.
Numerous independent variables were tried in various combinations
to affect the best possible fit from the available data. The following
parameters were investigated:
X ; Baseline load (80% of capacity), in klb/hr
X ; Number of burners (or stokers where applicable)
X ; Excess 0 , in %
X ; API gravity (for oil only)
X ; Fuel nitrogen, in %
X ; Air temperature, °F
6 2
X ; Furnace area, in ft
7 3
X ; Furnace volume, in ft
o
X ; Preheat temperature, in °F
X ; Furnace length, ft
X11<- Mean distance between burners, (multiple burners only) , .in inches
These parameters were selected based on indications of potential
importance in the NOx formation mechanisms. The above independent variable
sets were correlated to NOx (in ppm, corrected to 3% excess o ).
153
-------
The selected parameters may be divided into those that characterize
the physical design of the boiler: X , X , X , X , and X.. ; that characterize
f. I o 10 11
the operating conditions: X , X , X , and X ; and that characterize the
fuel: X and X . Also furnace type, watertube or fireitube, and expecially
fossil fuel type; gas, oil, or coal; can be expected to influence the level
of NOx emissions.
Should a strong correlation exist between the X.'s, called
autocorrelation when included in a regression, the interpretation of the fit
statistics is subject to uncertainty. Moreover, there is little use in
including two or more variables strongly correlated since only one is sufficien
to predict the effects associated with a given process. In general, it was
found that volume was highly correlated with area? therefore, only one
parameter X or X was used in the regression analysis or the ratio Xg/X^,
was used. API gravity was moderately correlative with fuel nitrogen, and con-
sequently only fuel nitrogen was used as a predictive parameter. The remain-
ing parameters indicated only weak or no correlation with each other.
The available data were initially divided into broad categories
depending on fuel and boiler type as shown in Table 8-1
TABLE 8-1
NUMBER OF SAMPLES FOR VARIOUS CATEGORIES
Boiler Type
Watertube
Firetube
Total
Fuel
Gas
14
10
24
Oil
24
14
38
Coal
11
2
13
Total
49
26
75
6000-28
154
-------
Although considerable data were obtained for various loadings and excess
O , only baseline loading data are included in the above table. In several
instances, not all of the samples shown contain complete data sets, which
are required for regression analysis. Consequently, for purposes of
regression analysis, Table 8-2 indicates the number of data sets available
in the various categories.
TABLE 8-2
NUMBER OF SAMPLES USED IN REGRESSION ANALYSIS
Boiler Type
Watertube
Firetube
Total
Gas
13
8
21
Fuel
Oil
22
12
34
Coal
11
0
11
Total
46
20
66
6000-28
Initially, regression correlations were attempted within each category;
however, the variations accounted for by regression were not significantly
different from regressions that lumped boiler type. Significant differences
exist between fuel types and correlation attempts were confined to three
classes: gas with 21 samples, oil with 34 samples and coal with 11 samples.
The results of the regression analysis are summarized in Table 8-3.
155
-------
TABLE 8-3
SUMMARY OF REGRESSION ANALYSIS
Fuel
NOx Data, ppm
Mean
Standard Deviation
i Range
Best Regression
Equation
Parameters
B
0
P
1
g
2
•D
3
; B
i 4
B5
Regression Statistics
j
i Sample Size
1
Multiple Correlation
Coefficient
Coefficient of
i Determination
Standard Error
of Estimate
Fisher's F and
(significance level, %)
Confidence about
mean at 90% level
Student's t and
(significance level, %)
t
t
B2
t
B3
t
B4
^5
Gas
140
93
57 - 374
f ^i\ / a\ B-3
1 A \ [ \ v J
NOX = B-. V TT" j I v / Q
o\x2; \x7/ 9
12.4
0.108
0.160
0.387
21
0.860
0.740
0.138
16.15 (99)
+_39 ppm
1.06 (70)
1.24 (76)
3.36 (99)
Oal
220
114
65 - 619
Y Bl v B3 y B4
X2 5 8
NOx = BQ x B2 x B5
701.5
0.304
0.751
0.247
0.109
0.455
34
0.702
0.493
0.186
5.45 (99)
+94 ppm
1.77 (90)
2.61 (98)
3.59 (99)
1.42 (84)
1.43 (84)
Coal
467
160
224 - 800
fxAV
NOx = B°vv V
90.4
0..332
0.320
11
0.923
0.852
0.068
22.94 (99)
+_52 ppm
6.57 (99)
1.86 (90)
Xn = Baseline load, K Ib/hr; X = Number of burners; X = Excess O ,%; X = Fuel Nitroc
1 2. 2 J 3 ^ 5
content,%; X = Furnace wall area, ft ; X = Furnace volume, ft. ; X = Primary air
temperature, °F; x
10
Furnace length, ft.
156
6000
-------
8.3 DISCUSSION OF STATISTICAL RESULTS
The statistical results shown are quite helpful in discerning
which variables were indicated to influence NOx formation, and, equally
of help, which were not. However, the limitations of the results due to
methodology and input data must be understood. An attempt was made to group
the input data to avoid "comparing apples and oranges." An attempt was made
to collect the data into like groups so that statistical analysis could
appropriately be applied. Unfortunately, although this probably represents
the largest number of boilers ever tested in a single, coherent effort, such
groups were each very small. As a result, the only useful groupings were
baseline condition data for gas, oil, and coal fuels. For instance, when
grouping all data together and including all fuel types in the same regression
attempt, it was impossible to develop a meaningful relationship. Similarly,
there are great differences between boilers using a similar type of fossil
fuel.
Experience has demonstrated that all boilers are not members of the
same population with random variations about some nominal characteristic.
For instance, the NOx formation characteristics of a B&W opposed fired cell
burner unit, a Riley Turbofired unit, a CE tangential fired unit, a cyclone
burner unit, and a Scotch-Marine faretube unit can not reasonably be considered
as members of the same population with some standard deviation in NOx
formation due to random variations about a mean value. Although regression
attempts that separated watertube and firetube boilers did not show a
significant difference between the two basically different boiler types,
separation of the data into small more homogeneous strata markedly influenced
the quality of the regression. This was particularly true for oil when
multiburner units were analysed together. The regression equation showed a
much higher multiple correlation coefficient than the case shown in Table 8-3
which includes all oil fired types.
The effect of changing the position of an air register or taking
a particular burner out of service has a significant effect on NOx emissions.
Registers and burners are unique in that each one is specially located in a
157
-------
different place and thereby have different nonrandom effects on fuel/air
distribution. Analysis of limited tests of these effects suggests that
burners in or out of service may influence NOx from 10 to 30 percent. Some
anomalies result from the limitations discussed above; e.g., no significant
effect of excess 0 at baseline conditions is shown for the gas and coal
equations and the oil equation indicates that NOx increases as excess 0
decreases. This does not conflict with the data describing the effect of
excess 0 on an individual unit (Section 5.1.1) which typically shows NOx
to decrease as excess O decreases. Rather, it indicates that at baseline
conditions, including baseline 0 , the NOx emissions for gas and coal units
are primarily determined by factors other than excess O .
It is also difficult to interpret the significance of a variable or
group of variables without an understanding of the physical process occurring.
For instance, NOx emissions for oil fuel correlated well with API gravity;
but API gravity correlated well with fuel nitrogen content which is
obviously the variable of real influence. A conclusion that API gravity was
of prime importance would erroneously devert attention from the variable of
real significance. For gas and coal fuels, the furnace volume divided by
area is indicated to be an important parameter; but either volume or area
alone gave almost as good a correlation. Volume over area could be considered
a characteristic furnace length. Length could also be related to residence
time. Clearly, many possibilities exist.
The significance levels obtained along with the Student's t values
indicate the relative importance of each variable in explaining the NOx
variation. Variables of significance level below 70% were not included.
Those of primary importance (significance level over 90%) were few. For
coal fuel where the variables were burner heat loading and volume to area
ratio, the former was most important (99%) and the latter was relatively
important (90%). For gas fuel, the same two variables plus air preheat
temperature were included but air preheat was the most significant variable
(99%). For oil fuel, several variables were included but only two, fuel
nitrogen content (99%) and excess 0 (98%) were found to be highly significant.
158
-------
The regression equation for the oil fuel data contains five variables.
It was found that the selection of different variables to be included resulted
in quite different calculated coefficients. In some cases, a coefficient
would change from positive to negative. This would seem to indicate that the
exact form of the equation and values (or even signs) of the coefficients are
rather uncertain. However, the overall equation in the form shown predicts
the measured NOx values with the uncertainty specified.
Variable selection and data interpretation to provide meaningful
results and assist in formalizing and quantifying an appropriate physical/
chemical model will require additional effort. In the meantime, the regression
equations in Table 8-3 provide a simple method to predict NOx emissions from
industrial boilers with uncertainties of about +_ 11%, 28%, and 43%, for coal,
gas, and oil, respectively.
159
-------
SECTION 9.0
REFERENCES
1. Barrett, R.E. and S.E. Miller. Field Investigation of Emissions from
Combustion Equipment for Space Heating, Final Report. Battelle-Columbus
Laboratories, Columbus, Ohio. Prepared for the U.S. Environmental
Protection Agency and the American Petroleum Institute. EPA Report No.
EPA-R2~73-084a. NTIS No. PB 223-148, or API Publ. 4180, June 1973.
2. Locklin, D.W., et al. Design Trends and Operating Problems in Combustion
Modifications of Industrial Boilers. Battelle-Columbus Laboratories,
Columbus, Ohio. EPA Report No. EPA-650/2-74-032. Prepared for
the U.S. Environmental Protection Agency, April 1974.
3. Ehrenfield, J.R., et al. Systematic Study of Air Pollution from
Intermediate-Size Fossil-Fuel Combustion Equipment. Walden Research
Corporation, Cambridge, Massachusetts. Prepared for U.S. Environmental
Protection Agency. NTIS No. PB 207110, July 1971.
4. Bartok, W. et al. Systematic Study of NOx Emission Control Methods for
Utility Boilers. Esso Research and Engineering Company, Linden, New Jersey.
Prepared for the U.S. Environmental Protection Agency. NTIS Report No.
PB 210739. December 1971
5. Smith, S., Emissions from Fuel Oil Combustion - An Inventory Guide.
U.S. Department of Health, Education and Welfare, Public Health Service,
Cincinnati, Ohio, November 1962.
6. U.S. Energy Outlook, An Initial Appraisal, 1971-1985. Interim Report,
Committee on U.S. Energy Outlook, National Petroleum Council, Vol. No. 1
and 2, November, 1971.
7. Hall, R.E. and D.W. Pershing. Proceedings, Coal Combustion Seminar,
June 19-20, 1973, Research Triangle Park, NC. Prepared for U.S.
Environmental Protection Agency. EPA-650/2-73-021. September 1973.
8. Stationary Watertube Steam and Hot Water Generator Sales, 1972. American
Boiler Manufacturers Association, Arlington, Va.
9. Guide for Installation and Operation of Oil Burning Units. American
Boiler Manufacturers Association, Arlington, VA, 1971.
10. Setter, J.G., "The Effect of Fuel Nitrogen on NOx Production in Oil-Fired
Utility Boilers," Final Report, KVB, Inc. prepared for WEST Associates,
Report No. 51-137, December 17, 1973.
11. Pershing, D.W., J.W. Brown, and E.E. Berkau. Relationship of Burner
Design to the Control of NOx Emissions Through Combustion Modification.
Presented at EPA Pulverized Coal Combustion Seminar, Research Triangle
Park, NC. June 1973.
161
-------
12. Turner, D.W. and C.W. Siegmund. Staged Combustion and Flue Gas Recycle:
Potential for Minimizing NOx from Fuel Oil Combustion. Presented at the
American Flame Research Committee Flame Days, Chicago, 111., Sept. 1972.
13. Kato, K., C. Kurata, K. More, and K. Fujii. Formation and Control of
Fuel Nitric Oxide. Published at a conference of thermo-engineering
held by the Japanese Society of Mechanical Engineers, Nov. 1973.
14. Blakeslee, C.L., and H.J. Burbach. Controlling NOx Emissions from Steam
Generators. Presented at the Air Pollution Control Association's 65th
Annual Meeting.
15. Zerban, A.H. and E.P. Nye. Power Plants, Second Edition. International
Text Book Company, Scranton, NY, 1957.
16. Hilt, M.B.f and D.V. Giovanni. Particulate Matter Emission Measurements
from Stationary Gas Turbines, ASME, 73-Pwr-17.
17. Federal Register, Vol. 39, No. 177, Part II, paragraph 60.46, September
11, 1974.
'162
-------
SECTION 10.0
GLOSSARY OF TERMS
Air
Amb
Atm
API
B#XOOS
B&W
Brn
Burnh
CO
co
C2H6
CE
CI
CL
Cl Brk
Coen
Coppus
c
Cup
Con Part
cor.
Usually referring to air-atomized fuel oil burner
Ambient temperature
Atomization
American Petroleum Institute
Burner Number X out of service
The Babcock and Wilcox Company
Burner
Burnham/Golden Scotch
Carbon monoxide
Carbon dioxide
Multiple carbon atom hydrocarbons
Methane
Ethane
Combustion Engineering, Incorporated
Cast iron furnace walls
Unheated sample line (cold line)
Cleaver-Brooks Division
The Coen Company
Coppus Engineering Corporation
Coal
rotary cup fuel oil atomizer
Condensible particulates
Data corrected to standard conditions
163
-------
cyclone or eye.
cm
D
Det Stk
°C
°F
EPA
FD
FT
FW
Faber
ft
g/MCal
H
HC
HL
hrs
Hz
IBW
ID
IR
Ind. Comb.
in Hg
Cyclone furnace coal combustor
Centimeters
Diameter
Detroit Stoker Company
Temperature in degrees centigrade
Temperature in degrees Fahrenheit
Environmental Protection Agency
Forced draft
Furnace tube furnace
Foster Wheeler Corporation
Faber Engineering Company Incorporated
Feet
Grams of constituent per million calories of fuel
input computed at 3% excess oxygen dry in the flue gas
Height
Unburned hydrocarbons measured as methane
Heated sample line (hot line)
Hours
Hertz or cycles per second
International Boiler Works Company
Inside diameter or induced draft
Infrared
Industrial Combustion, Incorporated
Pressure in inches of mercury, usually gage
16'4
-------
iwg
kg/hr
KPH or k Ib/hr
Kewan
Keeler
L
Ibs or #
MBH or MBtu/hr
MCH or M Cal/hr
MR
Mfg
ma
iJm or \i
min
mm
N
N2
NG or G
NO
N02
NOx
NO.
No. Am,
Nebr
NSP - oil
Pressure in inches of water column gage
Kilograms per hour
Mass flow rate in thousands of pounds of steam
per hour
Kewanee Boiler Corporation
E. Keeler Company
Length
Pounds
One million British thermal units per hour
One million calories per hour
Mixture ratio in terms of air flow rate divided by
the fuel flow rate
Manufacturer
Milliamps
Micrometer or "micrron" (10 meters)
Minutes
Millimeters
Molecular nitrogen content in fuel percent by weight
Nitrogen gas
Natural gas fuel
Nitric Oxide
Nitrogen dioxide
Total nitrogen oxides (NO+NO-)
Number
North American Co., Cleveland, Ohio
Nebraska Boiler Company
Navy Standard Fuel - oil (similar to No. 5 oil)
165
-------
o
°2
OD
O/S
P
Pulv.
PS-300
Peabody
#/MBtu or Ib/MBtu
ppm
psi
psia
psig
R
RG
RT
Ray
Reg
Riley
rms
S
Oil
Oxygen gas
Outside diameter
Off-stoichiometric
Preheated combustion air when outside of data symbol
and pulverized coal burner when inside of symbol
Pulverized coal burning equipment
Pacific Standard Fuel Oil No. 300 (similar to No. 5 oil)
Peabody Engineering Company
Pounds mass of constituent per million British
thermal units of fuel input
Parts of constituent per million parts of total volume
Pressure in pounds per square inch
Pressure absolute in pounds per square inch
Pressure gauge in pounds per square inch
Refractory
Refinery gas fuel
Water wall tubes spaced such that refractory tile is
exposed to flame
Ray Burner Company
Air registers
Riley Stoker Corporation
Root mean square
Sulfur content in fuel percent by weight, or when
inside coal data symbol refers to spreader coal
burning equipment
166
-------
Sprd.
so2
so3
SOx
Sec.
Sid. Part.
Steam
Supr.
TIW
TT
Temp.
Todd
Trane
t
U or UFS
Union
uncor.
V
Vol
W
WF
WT
Wall Const.
Wtgh
Winkler
Spreader stoker coal burning equipment
Sulfur dioxide
Sulfur trioxide
Total sulfur oxides (SO +SC>3)
Seconds
Solid particulates
Usually referring to steam-atomized oil burners under
burner type
Superior Combustion Industries
Titusville Iron Works
Furnace walls where the watertubes are tangent
Temperature
Todd - CEA Incorporated
The Trane Company
Metric ton (1000 kg)
Underfed stoker coal burning equipment
Union Iron Works
Data presented as measured and not corrected to a
standard condition
Voltage in volts
Volume
Width
Furnace wall constructed with welded fin design
Watertube furnace
Furnace wall construction
Westinghouse
Winkler burner manufacturer
167
-------
SECTION 11.0
EMISSIONS UNITS CONVERSION FACTORS
Pollutant emission regulations are written in various sets of units.
Converting data from commonly measured engineering parameters into these
various sets of units requires basic knowledge as to the chemical composition
of the fuel and the chemical processes involved.
Table 11-1 presents equations for converting emissions data from
one set of units to another. The values of N used for computing these
conversion factors depends on (1) conditions of flue gas for quoting data,
i.e. 3% O dry, 12% CO., etc. and (2) the chemical composition of the fuel.
-------
SI
s
H
u
en
EH
H
§
H
CO
W
co
C
•rH
ft)
M
Oi
•
H
O
£>
*-^
g
O<
*>—
m
6
(Q
M
Oi
(0
a
4J
A
CN
O
2
vo
T OS
o
II 0
n
5 in
en
rd
a
3
rH
Pw
MH
O
B
CO
(0
•H
to
CCJ
0)
^
H
a
rH
id
o
•D
o
rH
3
•M
lD
O
rH
tn cy
-H
•*
dp
^i
CO
^1
rH
ft
-H
•M
rH
I/
/
I
3
C
H
j
/
/
/
/
/
/
/
f
•S
1
Q
o
*-sj
H
x^>
^^
z
<
^^
^.^^
:e
in
o
ws,
5
-— s
rfj
W
o
o
0
^
o
H
.»— it
t>
5c
X-t*
*""*•
rf
V-*
^*.
•S
%•-»
o
o
o
•b
00
rH
^
^
5
5
5
O
o
o
o
rH
rH
•P
•§>H
Q) &4
* H
s^
*^s
K
tl
a
|
5c
»— i
00
H
*•-!
**^
Jgj
s—*1
o
o
o
o
^^
^
K
*»«*
*«^
51
rH
co
5
3
0
O
O
rH
^^
fjj
£
a
x^
f>
N*^
CO
I/)
o
o
*~%
a
**-S
S
CO
rH
J>
JC
rH
cn
rH
,-x.
s
o
o
o
CO
rH
.^.^
^>
5s
*»«^
^^
s
P
r,
Cf
3
o
o
rH
33
P4 en
E^ in
*~ o\
Q
O
•
rH
*— s
a
j»-^*
j> 3S
3S **^
03
*
f~|
> S
i
o *-*
o a
O *"
O ri?
rH ^^
.
rH 10
O -H
€
sir
l-l
s*^
*&
•«W
Ch
m
o\
o
0
•
*-•*
QjJ
£-4
%-•*
/»*.
t>
f^
*^_r
ro
(V)
U")
o
O
*— *
ffS
PHI
'*-*'
^"N
a
r— %
s
CT>
m
W
o
o
*
^
pj
Ej,
5
s*^
^r^
JgJ
"*-••
O^
•
in
OS
*--*
1^w%
*2
HM**
^
"^
CO
fi
-H
ro *— * n
O rH
H 00 \
< II ffi 4J
«k 5* •» "^-r-
— H CO CD O
•P — ' 3 O
•H Kj1 *H )25
J3 VD *0 dP
>-^ > CM
^* It i~\
rH rH
S t)
H U
•H
ttt " CU CN
X S 0
» rH «J *>
T3 o» +J in
Q> 3 CO • rH
M i^ H
0) SO)
•O MH o 3
•H O M t-4
W
•r) > O CO
eg rd
<0 Cn <4H en
Xt C O H
•H 0) 0)
<4H
+J R -H
X 0) >i
tn H » M
-H Q) ,
•rj O CO TJ -P
G rH CO CU C C
0 d) rH rH rd 3
•P 3 O Q -P O
< PM S S CO U
N II II II t!
170
-------
UJ
s:
o
CO
UJ
u
a:
UJ
a.
^^
CM
o
CM
O
o
CM
GASy % BY VOLUME
C2H6 C3H8 C^HjQ C5H12
,17 6,0^ 2,49 0,27 0,02
40 50 60 70 80
% EXCESS AIR
90
100
Figure 11-1. Flue Gas Composition as a Function of Excess Air for
Natural Gas Fuel.
171
-------
LU
O
m
LU
u
01
LU
a.
2?
C_>
CNl
O
CNl
O
O
CM
86,4 11.9 0=4 0,45 0,015 0,83 5
40 „ 50 60 70
% EXCESS AIR
80 90
Figure 11-2. Flue Gas Composition as a Function of Excess Air for
Oil Fuels.
172
-------
20
18
16
14
UJ
12
o
PCI
H-
z:
LU
CJ
oc
UJ
O.
10
8
CXI „ ——
o
CXI
0
COAL, % BY WEIGHT
C H N S ASH 0
73,83 5,20 1,80 ,86 10,36 7,95
40 50 60
% EXCESS AIR
Figure 11-3. Flue Gas Composition as a Function of Excess Air for
Coal Fuels.
173
-------
174
-------
175
-------
a
H
4J
[Q
fc U
U
en OP
\ fN
[/) rH
G
H 0;
\
\
a in
a
-------
APPENDIX
BOILER SELECTION
Table A-l shows the final distribution of test boilers
by size, fuel type and burner type. This distribution is a
composite of several criteria, such as boiler population,
boiler emissions, burner population, the new energy policies
of the United States, and present and predicted sales. The
appendix discusses how the final distribution was developed.
Boiler Population
Table A-l shows a breakdown of industrial boiler types
by percentage of the entire industrial boiler population. This
information is from Appendix Reference A-l and is a "best estimate"
of the number of watertube and firetube boilers in service
based on the total capacity of the type in operation about
1972. Similar data, but listed by number of units in service
in 1967, were available from Reference A-2, and these data are
listed below.
Cate-
gory
1
2
3
4
5
6
Furnace
Design
Watertube
Watertube
Watertube
Watertube
Firetube
Firetube
Capacity
k#/hr or
MBtu/hr
10-16
16-100
100-250
250-500
10-16
16-100
Boiler
Population
in 1967
7,550
26,800
4,015
942
27,000
8,000
Total
Capacity
10^Btu/hr
91
833
700
35
350
450
74,307 Total
The selection of 50 test boilers that was based solely
on published data was a compromise between population and total
capacity data shown in these two tables. This selection was
177
-------
Table A-l
DISTRIBUTION OF BOILERS IN SERVICE IN THE UNITED STATES,
CIRCA 1972.
Industrial
Btu/hr or
103 Ib ttm/hr
RATED
CAPACITY.
SIZE RANGE
Boiler Horsepower
Industrial Type>1CT # Steam/hr
Packaged
Field erected
Commercial Type OO'' # Steam/hr
Coil
Firebox
Other
Packaged Scotch
Firebox
Vertical
Horizontal Return Tubular (HRT)
Misc. (Locomotive type, etc.)
MISC. 'Tubetesi. etc.)
From Reference A--1
178
-------
further modified to reflect sales and fuel use as discussed
in the following sections.
Sales
The following table is a recent distribution of boilers
(A-2)
by capacity developed by Battelle-Columbus from water-
tube sales data for the period of 1965 to May 1973, supplied
to them by the American Boiler Manufacturers Association.
CAPACITY
(k
1.
2.
3.
4.
Ib/hr)
10-16
16-100
100-250
250-500
NUMBER
'65
108
800
128
54
'66
99
874
186
41
'67
54
657
130
24
OF BOILERS
'68
56
696
154
0
'69
65
753
207
35
SOLD
'70
35
663
193
34
'71
40
704
226
25
'72
44
709
172
46
'73
45
303
90
39
Total
Sales for
Period
516
6159
1486
298
8459
The significant difference between the population and
total capacity information cited previously and in the table
above is the preponderance of watertube boilers in the 16,000 -
100,000 pounds of steam per hour category. According to these
sales data, 73% of the 32 watertube boilers to be tested, i.e.,
23 units, should be Category 2, rather than the 16 units
that were determined from population and total capacity data.
This large number was discussed with parties concerned with this
project and the consensus was that if 23 watertubes of Category
2 were tested, there would be an insufficient number of test
units remaining in the other watertube categories. Therefore,
the number of Category 2 units to be tested initially was left
at 16. Later, it was further reduced to 15, and the span of
Category 2 was narrowed to 30,000 to 100,000 pounds per hour.
179
-------
The boiler population, sales and total capacity information,
then, were used to develop the initial Selected Number of Test
Sets that is listed below.
Cate-
gory
1
2
3
4
5
6
7
Furnace
Design
Watertube
Watertube
Watsrtube
Watertube
Firetube
Fir=tube
Cast Iron
Capacity
MBtu/hr
10-16
16-100
100-250
250-500
10-16
6-100
1-10
TOTAL :
Selected No.
of Test Sets
4
16
8
4
11
5
2
50
Fuel Type
Oil
1
6
3
X
4
2
1
18
Oil &
Gas
1
2
1
0
1
1
0
6
Gas
2
6
2
1
5
2
1
19
Coal
0
2
2
2
1
0
o
7
i
Fuel Burned
A further distribution of the 50 test sets was made
on the basis of the principal fuel burned, and results by fuel
type are listed above. One basis of this distribution by fuel
types was the data on the percentage breakdown by fuel capability
in 1972 contained in Table 2, taken from Reference A-l. Data
from Reference A-2 on the amount of fuel used in 1967 by inter-
mediate boilers of the three basic types shown in Table A-3 also
were consulted. The proportion of fuel actually used only by
industrial sized boilers is shown on the seventh line of
Table A-3.
An additional basis of the initial selection was the
information in Table A-4 on the principal fuel of industrial-
size watertube boilers sold since 1965, supplied to Battelle-
(A-2)
Columbus Laboratories by the American" Boiler Manufacturers
Association. This compilation shows that the number of gas
180
-------
Table A-2_
POPULATION BREAKDOWN BY FUEL CAPABILITY (PERCENTAGE BASIS)
ALL INDUSTRIAL BOILERS NOW IN SERVICE.
|~ niuuiuim •"— ^^^"1
106 Btu/hr or
RATED 103 |b jtm/hr
SIZE RANGE Bo,|or Horsepower
FUELS
Oil Only
Gas Only
Coal Only
Oil & Gas and Gas & Oil
Oil & Coal and Coal & Oil
Gas & Coal and Coal & Gas
Misc. Fuels
(alone or with alternate fuels)
Total
OIL
Distillate. No. 2
Resid
No. 4 and Light No. 5 (No preheat)
Heavy No. 5 and No. 6 (Preheated)
Total Oil
10-16
301-500
35
45
3
16
1
100%
10
W)
20
70
100%
17-100
35
35
10
18
2
. 100%
2
(98)
2
96
100%
101-250
30
22
18
26
0.5
0.5
3
100%
2
(98)
nil
98
100%
251-500
22
22
22
23
3
3
5
100%
2
OB)
nil
93
100%
From Reference A-l
181
-------
Table A-3
SUMMARY OF CAPACITY, FUEL AND EMISSIONS
BY BOILER TYPE AND LOCATION IN 1967.
Capacity
(106 pph)
Watertube <500,000 pph
Flretube
Cast Iron
INTERMEDIATE BOILERS*
Residential
Comerclal
Industrial
Utilities
ALL BOILERS
INTERMEDIATE BOILERS
BY REGION
Atlantic
Sreat Lakes
Far West and South
Central Urban
South East
Rural Horth
1.833
613
757
3.370
2.117
1.341
1.515
1,800
6.773
913
818
770
399
362
108
Coal
(106 tons)
132
11
11
154
26
64
271
381
40
62
2
17
26
7
Res id.
(106 bbls)
194
93
56
348
...
106
192
156
454
202
42
60
11
17
16
Dist.
(106 bbls)
14
69
37
120
355
58
60
2
475
74
26
5
3
S
7
Gas
(1012 cu ft)
1.72
1.12
1.03
3.87
3.15
1.33
1.99
2.76
9.23
.26
.76
1.73
.72
.32
.03
SO,
(106 tons)
6.93
.88
.73
8.54
.24
1.52
4.27
15.17
21.20
2.47
3.40
.40
.94
1.05
.23
*>.
(10S tons)
1.54
.24
.15
1.93
.23
.36
1.04
3.40
S.03
.56
.60
.24
.21
.24
.08
(•articulates
Controlled Uncontroll
(106 tons) HO6 tons
3.83
.29
.27
4.39
.09
.77
2.39
2.03
5.28
1.23
1.75
.06
.54
.62
.19
r\i
nj>
r,i
7.99
na
n,i
n.i
n.l
25. n
2.08
3.29
.11
1.02
1.11
.38
TotlU do not necMtirlly adil up dut to rounding.
"a • «ot avatlibl* from runs presently completed xltn STRAT.
From Reference A-3
182
-------
TABLE A-4
NUMBER OF INDUSTRIAL-SIZE WATERTUBE BOILER SALES
1965 TO MAY 1973 AND THE FUEL BURNED.
FUEL
BURNED
Bituminous Coal
Oil
Natural Gas
Woodbark
Bagasse
Black Liquor
Other Fuels
Waste Heat
With Auxiliary Firing
65
161
316
594
13
2
19
60
0
0
66
130
346
707
4
2
15
64
0
0
YEAR
67
78
214
598
6
0
4
46
0
0
68
38
204
619
11
3
5
35
0
0
69
49
230
767
20
6
12
45
2
3
70
38
311
618
7
3
9
13
0
5
71
33
386
535
17
10
16
16
40
5
72
37
353
548
27
9
18
15
25
2
5/73
14
162
208
32
7
11
27
12
0
From Battelle-Columbus Laboratories Reference A-2.
183
-------
fueled units remained at 50% of total sales throughout the
period. The proportion of new oil-fueled units increased from
27% in 1965 to 34% in 1973, while coal burners dropped from
14% to 3%. There were no units reported that burned lignite
as the primary fuel.
Regional Distribution
The regional distribution of commercial and industrial
size boilers was deduced in part from the "intermediate boilers
by region" data listed in the lower section of Table A-3. The
regions are pictured in Figure A-l. The significant features
of the information in Table A-3 are: 1) the capacity of inter-
mediate or industrial size boilers is divided nearly equally
between commercial and industrial units with a minor number of
small size utility boilers, 2) the Atlantic, Great Lakes, Far
West and South regions represent approximately 75% of the United
States capacity of intermediate boilers, and the distribution of
boiler capacity in these regions corresponds to the human popu-
lation, age, and industrial characteristics of these areas
relative to the country as a whole, and 3) a little over half
of this 1967 capacity consists of watertube boilers and the
other half is more or less equally distributed between firetube
(A-3)
and cast iron units
Emissions
The right hand columns of Table A-3 summarize the emissio
by boiler furnace type and user for 1967, based on annual averag
operating factors and on uniform, average data for each region
considered. However, this summary may downplay the importance
of emissions from industrial-size boilers relative to all sta-
tionary sources, because:
1. These sources tend to be located centrally in urban and
metropolitan areas and to release stack gases at low
heichts.
184
-------
Figure A-l_. Regional Boundaries used in Table A-3.
(From Reference A-3)
0 = COAL BURNING BOILERS
Q = OIL AND GAS BURNING BOILERS
NOTE: VALUES IN PERCENT OF TOTAL
Figure A-2. Combined Emissions (1967) from Intermediate Boilers
By Type and Fuel. (From Reference A-3)
185
-------
2. Over half of the emissions of all three pollutants is
produced in the Great Lakes and Atlantic regions, where
the above factor is most significant - the areas are
well developed and highly centralized, and located in
the northerly part of the country.
3. The larger utilities, particularly new plants, and
large industrial complexes tend to be located remotely
and utilize tall stacks.
The particulate emissions from intermediate or industria?
sources shown in Table A-3 are about 80% of the total particular
emissions from all stationary sources, considering present
installed controls. The total national emissions of S02 from
intermediate sources represent about 40% of the total emissions
from all stationary sources. Intermediate boilers contribute
only approximately 20% of NOx emissions of all stationary source
Although the largest stationary emission source is utilities,
the effect of urban plant location and lower stack height of
the intermediate-size sources may make them a significant
(A-3)
source
Another factor entering into the selection of the test
units was the relative emissions of the various types of fuel
used. Watertube boilers burning coal have by far the greatest
emissions (about 70% of the total emissions from intermediate
size boilers), according to Figure A-2. This was one of the
reasons we arbitrarily increased the number of coal burning test
units from seven to twelve.
Oil and gas-burning watertube units are the next most
significant emission source. Coal-burning firetube emissions
are nearly as great, i.e., 4.4% , as both oil and gas-fired
186
-------
firetube boilers combined, i.e., 5.3%.
The industrial-size watertube class is the major emission
source, with utilities close behind. The use of coal in
commercial-size watertube boilers is relatively small, e.g., the
number of coal-fired watertube units is only about one fourth
the number of intermediate-size coal-fueled watertube boilers.
As mentioned before, a relatively small fraction of total
boiler capacity makes the major contribution to emissions. For
example, coal-burning watertubes in the Great Lakes and Atlantic
(A-3)
regions contribute nearly half the total burden
Boiler Trends
Tables A-5(A~4) and A-6^'1^ are predictions of boiler
population by capacity and by percentage of boilers in service.
Some of the trends that can be deduced from these and other
tables in this section are as follows:
- A higher portion of packaged boilers in the smaller
sizes and field-erected boilers in the larger sizes.
- More firetube and cast-iron boilers in the smaller
sizes and watertube boilers in the larger sizes.
Since the industrial-sized watertube package boiler
was first introduced in the early 1940's, they have become very
popular. In the period of 1930-1950, almost 95 percent of
the 10,000 to 100,000 pounds of steam per hour watertube
boilers were field erected. However, it is anticipated that
by 1990 99% of this class will be packaged boilers. Until 1950,
all of the watertube boilers in the range of 100,000 to 500,000
pounds of steam per hour were field erected. The forecast
indicates that by 1990 about 90% of the sizes up to 250,000
(A-l)
pounds of steam per hour will be packaged
187
-------
TABLE A-5
CAPACITY OF BOILERS BY TYPE AND USER.
Boiler
By Type
Total Watertube
Size 1 - Under 100 k#/hr
Size 2 - 100-250 k#/hr
Size 3 - 250-500 k#/hr
Size 4 - Over 500 k#/hr
Firetube
Cast Iron
Residential
TOTAL BOILERS
intermediate Size Boilers
By User - Total Boilers
Commercial
Industrial
Utility
Residential
Output in Million Pounds
1967
3,086
921
658
259
1,248
813
757
2,117
6,773
3,370
1,341
1,515
1,800
2,117
1975
4,595
1,045
700
286
2,564
1,365
985
2,877
9,822
4,335
1,857
1,989
3,099
2,877
1980
5,686
1,123
745
282
3,537
1,783
1,098
3,359
11,927
5,030
2,192
2,333
4,043
3,359 '
of Steam
1985
6,950
1,201
810
276
4,663
2,255
1,330
3,844
14,379
5,838
2,578
2,770
5,187
3,844
per Hour
1990
8,379
1,275
898
262
5,944
2,650
1,461
4,344
16,834
6,602
2,964
3,188
6,338
4,344
*Totals do not necessarily add up due to rounding.
From Reference A-4.
188
-------
TABLE A" 6
ESTIMATED TRENDS OF BOILER TYPES (PERCENTAGE BASIS)
ALL BOILERS INSTALLED IN YEARS NOTED.
Industrial"
RATED
CAPACITY,
SIZE RANGE
10G Btu/hr or
103 Ib stm/hr
Boiler Horsepower
10-16
301-500
17-100
101-250
251-500
WATER TUBE
Industrial Type > 104 # Steam/Hr
Packaged
Field erected
Commercial Typa <104 # Sleam/Hr
Coil
Firebox
Other
FIRE -TUBE
Packaged Scotch
Firebox
Vertical
Horizontal Return Tubular (HRT|
Misc. (Locomotive type, etc
30 '50 '70 '90
(25)(I7)(19)(20)
0 2 18 20
30 '50 '70 '90
30 '50 '70 '90
(94) (97) (94) (90) JOO) (100) (100X100)
'20 '50 '70 '90
COOQoo)(ioo)aoo
25 15
0
0 8 30 89
94 89 14 1
0 0 80 90
100 100 20 10
0
1 20
100 100 99 80
nil 35 40 45
20 40 40
CAST IRON
MISC (TUBELESS. ETC)
From Reference A-l.
189
-------
Burner Trends
The types of burners now in service and the trends to
1990 are shown in Table A-7. The table is from Reference A-l,
and indicates for the smallest capacity units that air and
steam atomizing oil burners are replacing the pressure and
rotary burners that formerly predominated. For boilers larger
than 17 thousand pounds per hour, steam atomizing burners
are the most common.
Over the last decade there has been a lively business
in converting small coal-fired units to oil and/or gas firing.
The smaller units no longer were being built with stokers, but
conversations with packaged boiler manufacturers indicate that
the high price of oil and gas fuels has reawakened the interest
in coal-fueled boilers. For larger coal-fired units the pulver-
ized coal burners are supplanting spreader stoker units.
The following table from Reference A-5 lists the age
of existing oil burners in several sections of the United
States. The table lists more new burners and fewer old burners
in the South Atlantic than in other sections. The majority
of burners in use are 11 to 15 years old.
SECTION
New England
Mid-Atlantic
South Atlantic
Midwest
West
All Sections
Under
5 Years
26
24
27
6
17
19
6-10
Years
25
17
39
21
20
23
11-15
Years
21
29
19
34
35
28
16-20
Years
16
17
8
28
17
19
> 20
Years
12
13
7
11
11
11
190
-------
TABLE A-7
ESTIMATE TRENDS BY BURNER TYPE (PERCENTAGE BASIS)
ALL BURNERS INSTALLED IN YEARS NOTED, INCLUDING
CONVERSIONS.
Industrial
RATED
CAPACITY,
SIZE HAN'GE
OIL F'JPN'ERS
1C5 Gtu/hr or
103 Ib stm/hr
EOI!C' Horiopower
Air Atomizing
Steam Atomiiing
Pressure or
Rotary
Mechanical Atomi/ing
Total Oil
COAL BURNER
Sprcac'cr
Underfeed
Overfeed
Pulverized
Other
Total Coal
10-16
301-500
30 '50 '70 '30
10 20 35 40
30 30 35 40
25 20 20 20
35 30 10 nil
100 100 100 100
5 10 5 nil
75 75 85 90
15 10 5 5
17-100
'30 '50 '70 '90
5321
75 80 88 90
15 14 10 9
5 3 nil nil
100 100 100 100
15 50 50 nil
50 35 35 85
30 10 10 10
5555
00 100 100 100
5555
100 100 tOO tOO
101-250
•30 '50 '70 '90
2 1 nil ntl
93 94 95 95
5555
ww/#w?/w/-
4&<'dM/4Ztf't64'
100 100 100 100
15 40 40 nil
50 30 20 20
25 15 15 15
5 10 20 60
5555
100 100 100 100
%
251-500
'30 '30 '70 'SO
2 1 nil nil
93 94 95 95
5555
'/&/%&&&£/%&/.
100 ICO 100 100
15 30 20 nil
40 20 10 10
20 15 10 10
20 30 55 75
5555
100 100 100 100
From Reference A-l.
191
-------
When combined with the data above on the age of oil
burners, the data below on life expectancy indicate the rapid-
ity with which burner replacement will take place: replacing
rotary cup burners with air atomizing burners, for example. For
oil burners, steam atomizing gives the longest life (30 years)
while pressure/mechanical atomizing gives the shortest life
of 15 years. For coal-burners, the life expectancy of the
spreader stoker type is slightly better than the other stoker
types.
Expectancy
Years
OIL BURNER TYPE:
Air Atomizing 20
Steam Atomizing 30
Pressure/Mechanical Atomizing 15
Rotary 20
COAL BURNER TYPE:
Spreader 20
Underfeed 15
Overfeed 15
Pulverized 18
Other 15
Sales data below (which were adapted from Reference
A-6) indicate that while the use of both low pressure atomizing
and rotary cup burners has declined, the rotary cup type has
decreased the most. Sales of rotary burners for small boilers
have decreased steadily from 24 percent in. 1951, and they
presently account for only two percent of annual sales. Con-
versations with representatives of the American Boiler Manufac-
turers Association and with boiler manufacturers confirm that
this trend is continuina.
192
-------
YEAR
1951
1955
1960
1965
1970
Percentage
LOW PRESSURE
ATOMIZING
%
45
35
43
22
77*
of Total Burners Sold
HIGH PRESSURE
ATOMIZER ROTARY
% %
31
36
36
54
24
24
13
9
2
NOT
SPECIFIED
BY TYPE
%
—
5
8
15
21
TOTAL
NUMBER
OF
SALES
40.6
42.9
37.9
36.2
26.5
* Total low pressure atomizing and high pressure atomizing
oil burners.
Another trend in small boiler burners indicated by the
table is the decrease in low pressure atomizing burner sales,
which by 1965 accounted for only 22 percent of annual sales. It
is deemed likely that high pressure atomizing burners will
continue to increase in popularity and will account for an
increasing percentage of total sales.
Mechanical stoker sales data from Reference A-6 for
industrial and commercial use are tabulated below. They show
that the annual shipment of mechanical stokers has declined
steadily for the last 20 years, with the industrial sector
having the highest percentage drop. Sales in 1970 amounted to
only 5 percent of the sales in 1950; however, it is possible that
the recent interest in the use of coal may reverse this downward
trend.
193
-------
YEAR
1950
1955
1960
1965
1970
(Thousands of
COMMERCIAL
6.6
4.2
2.4
.9
.2
Units Shipped)
INDUSTRIAL
.8
.3
.2
.1
.04
TOTAL
7.4
4.5
2.6
1.0
0.24
I
COMMERCIAL: Capacity 61-1200 Ibs/hr of Coal
INDUSTRIAL: Capacity 1201 and over Ibs/hr of Coal
Fuel Trends
Fuel, demand depends upon total energy consumption, and
the table below shows the predictions of two organizations of
what the total energy consumption will be. The prediction of
Reference A-5 is in Btu's and the prediction of Reference A-7
is in the equivalent number of barrels per day of oil required
to generate the required energy.
YEAR
1970
1975
1980
1985
2000
REFERENCE A- 5
Btu
io15
68
83
103
124
-
REFERENCE A-7
Equivalent
Barrels of
Oil Per Day
IO6
30
38
47
55
95
INCREASE
Over 1970
%
23-27
51-57
84-85
- 217
194
-------
The geographical distribution of the total consumption
of energy is indicated in the following table from Reference
A-5. The energy consumed originally was given by Petroleum
Administration for Defense Districts, but it is presented here
by the corresponding Federal Power Commission Regions. The
FPC Regions are used in the subject contract, rather than the
PAD Districts, and are delineated in Figure A-3.
FEDERAL POWER
COMMISSION
RFfiTON
2 & 4
3 & 5
6
8 & 9
1970
4,716
6,336
4,681
1,549
ENERGY
io12
1975
5,190
6,994
5,390
1,864
CONSUMED
Btu's
1980
5,652
7,641
6,188
2,145
1985
6,044
8,387
7,030
2,417
The tabulation above indicates that the eastern half
of the country and Texas are the largest users of energy and
will continue to be the largest in the future. Consequently,
most of the industrial boiler measurements are planned for sites
east of the Mississippi River or in Texas.
The initial selection of test boiler fuels was based
on the trend of consumption of the principal industrial boiler
fuels. Reference A-7 makes the following predictions of boiler
fuel trends (All fuels have been converted to barrels per
day of oil equivalent).
195
-------
Figure A-3. Federal Power Commission Regions.
196
-------
Year
1970
1980
1990
Energy Consumption
Total Industrial
Million BBLS/Day Equivalent Million BBLS/Day Equivalent
Oil
13.9
21.5
29.3
Coal
7.4
10.5
14.0
Gas
10.3
11.9
12.0
Oil
1.6
2.7
7.1
Coal
2.5
3.5
3.3
Gas
4.6
4.7
4.0
other 3.1
By 1990 it is predicted that an additional 3.1 million
barrels of oil per day, equivalent, will come from sources
other than oil, coal, or gas burned on site.
Table A-8 from Reference A-l presents data on the trends
of fuel consumption from 1930 to 1990 in terms of the fuel capa-
bility of boilers of the various industrial sizes. We have
encountered thus far very few boilers that can burn only gas
or oil; most industrial boilers can burn both. Table A-8
indicates that the trend toward dual fuel boilers will continue
through this decade and the next.
The test boiler distribution among the original 50
measurement sets that was developed during the investigation
discussed in this appendix is shown in Table A-9. This distribu-
tion was discussed with many people and organizations in the
power and environmental industries, and it finally evolved into
the distribution shown on Table 2-1.
197
-------
TABLE A-8
ESTIMATED TRENDS BY FUEL CAPABILITY(PERCENTAGE BASIS)
. ALL COMMERCIAL INDUSTRIAL BOILERS INSTALLED
IN YEARS NOTED, INCLUDING CONVERSIONS.
10GBtu/hroc
RATED 103 Ibstm/hr
CAPACITY, — —
SIZE RANGE Eoilor Horicpower
FUEL CAPABILITY
Oil Only
Gas Only
Cojl Only
Oil & Gas and Gas & Oil
Oil & Coal and Coal & Oil
Gas & Coal and Coal & Gal
Misc. luclt
(alona or with ilteinait fueli)
Total
OIL
Distillate. No. 2
Beid
No. 4 & Light No. 5 (No preheat)
Heavy No. S & No. 6 (Preheated)
Total Oil
•« • Industrial — ' — — —
10-16
301-500
30 '50 '70 '00
17 43 30 30
5 20 30 30
75 10 5 nil
nil 25 30 30
17-100
'30 '50 '70 '00
13 30 30 25
10 30 30 25
75 30 5 nil
nil 5 30 35
&W////MW0M, ^W^W'-l
/?%?/!• rsl'S'ty"'/1/.- '•' -y •/••?:•• ///'"'-'-'.'.'-'- 4
/&/%&falj&fSZ&
32 5 10
100 100 100 100
X
5 2 10 30
(95) (98) (90) (70)
20 23 10 nil
75 75 80 70
100 100 JOO 100
H
ys^i^/Zfati&i
2 5 5 15
100 100 100 100
X
ill till 10 20
(100X100) (90) 60
ill 3 nil nil
100 95 90 80
100 100 100 100
X
101-250
30 '50 '70 '90
5 20 24 20
5 20 24 20
90 38 15 nil
nil 10 25 40
nil 5 55
nil 5 55
nil 2 2 10
100 100 100 100
*
nil nil 5 10
aoO)(100)(95)(90
ai nil nil nil
100 100 95 90
100 100 100 100
X
251-500
'30 '60 '70 '90
5 15 20 10
5 15 20 10
90 60 20 10
nil 5 20 20
nil 3 ' 10 20
nil 2 10 20
nil nil nil nil
100 100 100 100
*
nil nil 5 10
;iOOH100)(95)(90j
nil nil nil nil
100 100 95 90
100 100 100 100
x
From Reference A-l
198
-------
TABLE A-9
DISTRIBUTION OF FIFTY TEST BOILERS
BY CAPACITY, FUEL AND BURNER
Category
1
*
3
4
S
6
Furnace
Type
WT
WT
WT
WT
FT
FT
Capacity
k*/hr
10-16
16-100
100-250
250-500
10-16
16-100
Oil Fuel
Distillate
Kcch.
1
(33)
1
Air
1
(44)
1
Residual
Steam
l
1
(1)
4
(6-9)
3
21-23)
1
(29)
1
(45)
10
Air j Rot
1
1
(2)
2
(10,11)
1
(3)
2
(34,35)
1
(46)
6
1
(36)
2
20
Gas Fuel
Ring
Fired
1
(4)
3
(12-14)
2
(24-26)
1
(30)
5
(37-41)
2
(47,48)
14
Center
Fired
1
(5)
2
(15-16)
1
(49)
4
18
Coal Fuel
Crate
2
(17-18)
1
(26)
2
(42-43)
1
(50)
6
Spreader
2
(19-20)
2
(27-28!
Pulver-
ized
2
(31-32!
Total
Number
of Units
Tested
5
15
!
4
i
! 1 1
4
2
12
12
i
6
50
199
-------
An attempt also was made to get a representative cross-
section of brands of boilers and burners. We did not have as
free a hand with brands as we did with size and fuel, because
only 50 test sets did not allow enough degrees of freedom for
a strict distribution by brand, as well as by size, furnace
type, fuel type and burner type. The major manufacturers of
industrial boilers and/or burners in the United States listed
in alphabetical order are the following, according to the
American Boiler Manufacturers Association:
The Air Preheater Corporation
Wellsville, NY 14895
The Babcock & Wilcox Company
Barberton, Ohio 44203
W. N. Best Combustion Equipment Co.
Danbury, Connecticut 06810
The Bigelow Company
New Haven, Connecticut 06503
Bryan Steam Boiler Company
Peru, Indiana 46970
Cleaver-Brooks Division
Milwaukee, Wisconsin 53201
The Coen Company
Burlingame, California 94010
Combustion Engineering, Inc.
Windsor, Connecticut 06095
Continental Boilers, Inc.
Middletown, Pennsylvania 17057
Detroit Stoker Company
Monroe, Michigan 48161
Eclipse Lookout Company
Chattanooga, Tennessee 37405
The Engineer Company
South Plainfield, New Jersey 07080
200
-------
Foster Wheeler Corporation
Livingston, New Jersey 07039
Gordon & Piatt
Winfield, Kansas 67156
Hoffman Combustion Engineering
Lincoln Park, Michigan 48146
A. F. Holman Boiler Works, Inc.
Dallas, Texas 75212
Industrial Boiler Company
Thomasville, Georgia 31792
Industrial Combustion, Inc.
Milwaukee, Wisconsin 53211
International Boiler Works Company
East Stroudsburg, Pennsylvania 18301
Iron Fireman, Dunham-Bush, Harrisonburg
Harrisonburg, Virginia 22801
S. T. Johnson Company
Oakland, California 94608
Johnston Brothers, Inc.
Ferrysburg, Michigan 49409
E. Keeler Company
Williamsport, Pennsylvania 17701
Kewanee Boiler Corporation
Kewanee, Illinois 61443
Lasker Boiler & Engineering Corporation
Chicago, Illinois 60608
James Leffel & Company
Springfield, Ohio 45500
Mid-Continent Metal Products
Chicago, Illinois 60614
Nebraska Boiler Company, Inc.
Lincoln, Nebraska 68501
Orr & Sembower
Middletown, Pennsylvania 17057
201
-------
Oswego Package Boiler Company
Oswego, NY 13126
Sid E. Parker Boiler Mfg. Co.
Los Angeles, CA 90058
Peabody Engineering Corporation
Stamford, Connecticut 06907
Preferred Utilities Mfg. Corporation
Danbury, Connecticut 06810
Ray Burner Company
San Francisco, CA 94112
Raypak, Incorporated
Westlake Village, California 91361
Riley Stoker Corporation
Worcester, Massachusetts 01606
Spencer Boilers
Lancaster, Pennsylvania 17601
Superior Boiler Works, Inc.
Hutchinson, Kansas 67501
Superior Combustion Industries
New York, New York 10017
Thermo-Pak Boilers, Inc.
Memphis, Tennessee 38113
TODD-CEA
New York, New York 10022
Trane Company
La Crosse, Wisconsin 54601
Vapor Corporation
Chicago, Illinois 60648
Henry Vogt Machine Company
Louisville, Kentucky 40201
William & Davis Boiler & Welding Co., Inc.
Hutchins, Texas 75141
202
-------
York-Shipley, Incorporated
York, Pennsylvania 17405
Zurn Industries, Inc., Erie City Energy Division
Erie, Pennsylvania 16512
203
-------
APPENDIX REFERENCES
A-l. Barrett, R. E. and S. E. Miller. Field Investigation of
Emissions from Combustion Equipment for Space Heating,
Final Report. Battelie-Columbus Laboratories, Columbus,
Ohio. Prepared for the U.S. Environmental Protection
Agency and the American Petroleum Institute. EPA Report
No. EPA-R2-73-084a. NTIS No. PB 223-148, API Publ. 4180, 6
A-2. Locklin, D. W., et al. Design Trends and Operating
Problems in Combustion Modifications of Industrial
Boilers. Battelie-Columbus Laboratories, Columbus,
Ohio. EPA Report No. EPA-650/2-74-032. Prepared for
the U. S. Environmental Protection Agency, April 1974.
A-3. Ehrenfield, J. R., et al. Systematic Study of Air Pollu-
tion from Intermediate-Size Fossil Fuel Combustion Equip-
ment. Walden Research Corporation, Cambridge, Massachu-
setts. Prepared for U. S. Environmental Protection Agency
NTIS No. PB 207110, July 1971.
A-4. Bernstein, R. H., et al. Systems Study of Air Pollution
from Industrial Boilers. Walden Research Corporation.
American Power Conference, April 1972.
A-5. "Special Study," Fuel and Oil Heat, January 1971, p. 22,.
A-6. U. S. Department of Commerce, Bureau of Census: Facts
for Industry, Heating and Cooking Equipment, Series
M34N and M51N.
A-7. Joint Committee on Atomic Energy: Understanding the
National Energy Dilemma, 93rd Congress, First Session,
August 17, 1973.
204
-------
;rir
PORT NO.
A-650/2-74-078-Z
-|_t A.'.'O SUBTITLE
Id Testing: Application of Com:.; V
3 Control Pollutant Emissions .' :u ;a;.t
oilers --Phase I _ __
THOR(S)
A. Gate; H. J. Buening; C. C, ;)> ' ,'
G. Morion; J. M. Robinson
~ NATION NAME AN"D ADDRESS""
3 Engineering, Inc.
32 Irvine Boulevard
tin , -LiTornia 92680
'ONSOHING AGENCY NAME AND ADDRESS
\, OlYiee of Research and Deve!c'pr..enl
SC-RTP, Control Systems Laboratory
earch Triangle Park, NC 27711
IPPl EMENTARY NOTES
'". -icCif lENT'S ACCESSION NO.
>'• >>.r PORT OATl
jOrtober 1974
JB )•( HFOHMlNCi OROANI/A1 ION I'UIH
I
]a Fe^To^RMiNG ORGANIZATION"REPORT NO.
|SN-8000-28
(10. f=!OC-HAM ELEMENT NO.
1AB014; ROAP 21BCC-046 j
.TTTONTHACTTG RAN T~NCT ~l
-'68-02-1074 !
'13. TYPE OF REPORT AND PERIOD COVERED ?
:Fmal (Phase I): 6/73-7/74 J
4. SPONSORING AGENCY CODE
ISTRACT The report gives results of field measurements made on 47 representative
istrial boilers (75 boiler/fuel combinations) of 10,000-500,000 Ib/hr of steam
icity throughout the Continental U.S. Poimtants measured were total nitrogen
Jes (NOx), total particulates , total bullir oxides, CO 2 , CO, and hydrocarbon.
.surements were made of emissions 1'rorn cokl. oil, and natural gas fuels and a
.ety of coal stoker and oil atomization methods. The effectiveness of reduced
3ss air, reduced load, air register readjustment, and off-stoichiometric firing
educing NOx emissions is evaluated. T;ie report covers the selection of the test
.ers and discusses the emission measurements during the first phase of the
jram. The second phase will include results of more detailed long-term testing
epresentative current boilers.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
it. iDbNTIFit-RS/OPtN ENDED TERMS jc. COSATI Field/Group
; Air Pollution Control
Pollution; Combustion; Boilers
ners; Emission; Nitrogen Oxides j Stationary Sources ISA;
jr Oxides; Carbon Monoxide; Smoke ; Combustion Modification
1; Fuel Oil; Natural Gas; Hydrocarbons ; Industrial Boilers
j 13B; 21B
07B
:icle Size
21D;
; Pardculate; No, 2 Oil; 14B
1 No. 5 Oil; No. 6 Oil
07C
rm 2220-1 (9-73)
TRIBUTION STATEMENT
mited
;9. SECURITY CLASS (This Report)
Unclassified
;20 SECURITY CLASS {This page)
: Unclassified
21 NO. OF PAGES
214
J
i
1
22. PRICE 5
1
1
-------
------- |