United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-83-042 Oct. 1983
<&ER& Project Summary
Evaluation of Combustion
Modification Effects on Emissions
and Efficiency of Wood-Fired
Industrial Boilers
R. J. Tidona, W. A. Carter, H. J. Buening, and S. S. Cherry
Results of full-scale tests to evaluate
combustion modifications for emission
control and efficiency enhancement on
two wood-fired industrial boilers are
reported. These modifications consisted
of lower excess air and variations in the
overfire air system operation.
The boiler at Location 3 was fueled
with a combination of wood bark and
coal. The implementation of lower
excess air reduced NO> emissions by
37.2 percent and improved thermal
efficiency by 1.2 percent. Variations in
the overfire air system reduced NO> by
20.7 percent and improved efficiency
by 1.63 percent.
The boiler at Location 6 was fired with
hogged wood as the primary fuel and oil
as the supplemental fuel. The effective-
ness of lower excess air in reducing NO*
was 12.5 percent with a slight improve-
ment in efficiency (0.6 percent). Adjust-
ment of the auxliary air dampers producd
a 17.2 percent NO. reduction and a 1.7
percent improvement in efficiency.
Polycyclic organic matter (POM) was
sampled at both baseline and optimum
low-NO» conditions. On a ftg/m3 basis.
the POM for low-NO> conditions ex-
ceeded the baseline results by a factor
of two to three. The results obtained are
compared to previous POM sampling
on industrial steam boilers.
This Project Summary was developed
by EPA's Industrial Environmental
Rasearch Laboratory, Research Triangle
Park. NC. to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering
information at back).
Introduction
The activities reported here include
tests performed on a wood bark/coal-
fired boiler (Location 3) and a hogged
wood fuel boiler (Location 5). Oil was the
supplemental fuel at Location 5. Varia-
tions in load, excess air and overfire air
system adjustments were the combustion
modifications common to both boilers. In
addition, lower combustion air preheat
and supplemental fuel oil air damper
positioning were implemented at Location
5. Polycyclic organic matter (POM) was
also sampled at Location 5 at both baseline
and optimum low-NO* conditions.
Table 1 summarizes the reductions in
NO and changes in efficiency measured
at Location 3 for each combustion
control. The overfire air system modifica-
tion consisted of increasing the overfire
airports from 1 to 1.5 in. (2.54 to 3.81 cm)
diameter. As noted, the lowest NO level
obtained resulted from implementing
lower excess air before the modification
of overfire air ports. This arrangement
also produced an increase in boiler
efficiency of 1.2 percent.
Table 2 summarizes the NO reductions
achieved at Location 5 and the change in
efficiency for all modifications except
reduced combustion air preheat. This
modification could not be fully imple-
mented since the combustion air tem-
perature could be reduced by only 16-22
K. Also noted in this table is the NO mass
emission factor measured after each
-------
"**"
modification had been implemented.
Increased load (18 percent) actually
reduced the NO concentration and mass
emission factor. This NO characteristic is
somewhat unusual; i.e., peak NO occurs
in the mid-load range. This occurrence
may be due to the boiler's O2 vs. load
characteristic which could produce a
maximum NO concentration at less than
full load.
Polycyclic organic matter (POM) samples
were collected and analyzed for boiler
operation at both baseline and low-NO.
(auxiliary air damper adjustment) condi-
tions. The significant finding was that the
total POM at the low-NO, condition was
two to three times higher than that
measured under baseline conditions.
This large difference could be due more to
fuel property variations than to combustion
modification, although the trend of higher
POM with lower NO, has been observed
previously.
Emissions Test
Instrumentation
All emission measurement instru-
mentation for the full-scale testing was
carried out in an 8 x 42 ft (2.4 x 1 2.8 m) mo-
bile laboratory trailer. The gaseous species
measurements were made with analyzers
located in the trailer. The emission
measurement instrumentation used is
listed in Table 3.
Results
Location 3, Wood Bark/ Coal
Table 1. Summary of Combustion Modifications At Location 3
Combustion modifications implemented
consisted of lower excess air, variation in
overfire air damper positioning, and load
changes. In addition, the overfire air
system nozzle size was increased, and
the effect of lower excess air was re-
evaluated.
Figure 1 depicts the effect of oxygen on
NO emissions before the increase in
overfire air nozzle size. Reducing the
oxygen from 9.3 percent (as found. Test
1) to 7.8 percent (Test 4) lowered the NO
emissions by 37.2 percent while in-
creasing the efficiency by 1.2 percent. As
noted in Figure 2, the effectiveness of
lower excess air after the overfire air
modification was less pronounced.
With respect to boiler operation at 80
percent of rated load (Test 31) only a 7.9
percent reduction in NO was measured
when the load was reduced by 51 percent
(Test 29) as shown in Figure 3. This
control caused the boiler efficiency to
decrease by 4 percent.
Variations in the overfire air damper
positioning, at constant overall oxygen
Control
Lower Excess Air" (146?
Lower Excess Air" (184)
Overfire Air Dampers" (1 74)
Load Reduction <51%f* (140)
NO Reduction.
%
37.2
18.5
20.7
7.9
Efficiency
Change, %
+1.2
+0.9
+ 1.6
-4.0
NO After Control
ng/J*
92
150
138
129
'NO as NO*
^Before overfire air system modification.
c Value in parentheses is baseline NO (ng/J) before combustion modification.
"After overfire air system modification.
'Load reduction referenced to nominal operation at 80% of rating.
Modification
Lower Excess Air (4Of
Increase Overfire Air (46)
Auxiliary Air Damper (36)
Load Change"
+18% (40)
-30% (40)
NO Reduction,
%
12.5
21.7
17.2
27.5
3O.O
Efficiency NO After Modification
Change, % ng/J*
+O.6
-1.3
+1.7
+0.9
+1.8
35
36
3O
29
28
'NO as NOz.
''Value in parentheses is baseline NO (ng/J) before combustion modification.
"Load change referenced to nominal operation at 76.5% of rating.
Table3. Emissions Measurement Instrumentation
Species
Hydrocarbon
Carbon Monoxide
Oxygen
Carbon Dioxide
Nitrogen Oxides
Particulates
Sulfur Dioxide
Manufacturer
Beckman Instruments
Beckman Instruments
Teledyne
Beckman Instruments
Thermo Electron Co.
Joy Manufacturing Co.
DuPont Instruments
Measurement Method
Flame lonization
IR Spectrometer
Polargraphic
IR Spectrometer
Chemiluminescent
EPA Method 5 Train
UV Spectrometer
Model
No.
402
865
326A
864
10A
EPA
400
(9.8-9.9 percent), were shown to reduce
NO emissions by 20.7 percent and
increase efficiency by 1.6 percent. This
result was obtained by partly closing the
lower row of overfire air ports and fly-ash
reinjection ports.
Total and solid particulate emissions
were measured downstream of the
multiclone at the low-NO, condition.
Total particulate was 138 ng/J (0.320
lb/106 Btu) and the solid particulate was
118 ng/J (0.274 lb/106 Btu) with the
unit operating at 8.2 percent 62.
The low-NO, cascade impactor test is
shown in Figure 4. Particulate diameter
as a function of cumulative proportion of
impactor catch is plotted. About 27
percent of the particles are below 3 //m
aerodynamic diameter. The geometric
mean and geometric dispersion are 6 and
1.099 //m, respectively. A comparison of
the baseline and low-NO, results indicates
that the geometric mean particle size for
baseline operation is approximately 50
percent of that measured during low-NO,
operation (3.2 fjm vs. 6 (im). Closing the
dampers for the overfire air and fly-ash
reinjection (low-NO, configuration) re-
sulted in the production of larger partic-
ipates, but at a reduced mass rate (118
vs. 155 ng/J).
Location 5, Hogged Wood
Boiler
NO emissions from this boiler were
very low: only one measurement exceeded
100 ppm. Combustion modifications
implemented were lower excess air, load
variations, increased overfire air flow,
and auxiliary air damper positioning. The
effectiveness of reduced combustion air
preheat could not be established since
only a modest (16-22 K) reduction was
possible when the steam coil portion of
the air heater was bypassed.
The effect on NO of excess air variations
is shown in Figure 5 for three different
loads. With respect to nominal operation
at 19.3 kg/s (153,000 Ib/hr) steam flow
and 7.2 percent O2 (Test 5/2-1 A), a load
-------
reduction of 30 percent produced a NO
reduction of 30 percent and an efficiency
increase of 1.8 percent.
The effectiveness of lower excess air at
constant load is also shown in Figure 5.
At a load of 19.3 kg/s (153,000 Ib/hr) nd
7.2 percent 02, a 12.5 percent reduction
in NO was computed when the Oz was
lowered to 5.7 percent (Test 5/2-4).
The overfire air flow, as a percentage of
total combustion air, was increased from
a baseline value of 5.7 percent to 9.7
percent. This had the effect of reducing
the primary combustion air admitted
under the grate. This modification
reduced the NO emissions by 21.7
percent; however, the boiler efficiency
was also lowered by 1.3 percent.
Each auxiliary oil burner has an
independent air supply. Opening this air
supply (with the burner off) implements
another form of staged combustion. Tests
conducted in this configuration, when
compared to all baseline tests, produced a
NO reduction of 17.2 percent and a 1.7
percent increase in efficiency.
Figure 6 presents all paniculate meas-
urements made at the boiler outlet as a
function of the corresponding NO levels.
A trend of lower particulate emissions
with lower NO emissions is noted. The
reason for this behavior is not known at
present.
Polycyclic organic matter (POM) samples
were obtained at the boiler outlet under
both baseline and low-NO. (auxiliary oil
burner air dampers open) conditions.
Table 4 presents the speciated analyses
from which it is noted that the total POM
under Iow-N0x conditions is two to three
times greater (on a /jg/m3 basis) than for
baseline operation. This trend (i.e., higher
POM with lower NO*) was observed
previously,* and is attributed to more
fuel-rich conditions in the burning zone.
•Carter, W. A., and Buening, H. J., "Thirty-Day Field
Tests of Industrial Boilers, Site 1—Coal-Fired
Spreader Stoker," EPA-600/7-80-085a (NTIS PB
80-211386), April 1980.
400
300
200
100
Fuel: ^ 70% Coal
s 30% Wood Bark
( ) Test Number
Boiler Load; 37.2-37.7 Mg/h
(4)
Smoke Limit'
@ 7.8% Oz
10
11
Oi %. dry
Figure 1. Location 3—NO emissions as a function ofOt before overfire air nozzle modification.
-------
400
300
200
100
(
0 Baseline before modification
H Oz Variation before modification
& Overfire air O
~ O Baseline after modification |
Q Oz Variation after modification
O Coal only
OH
8H-
^ 0
•
Fuel. ^80% Coal
a 20% Wood Bark
Boiler Load 34.9-38. 1 Mg/h
1 I t I I 1 i 1 I I
32465/0
JUU
200
X
Q
C3
I
O'
* 700
Stack Oxygen. %. dry
ure 2. Location 3 — NO emissions as a function of stack O»
1 1 1 1 1 1 1 1 1 1
(32) (?)
7
rj^^
\ -ST^T^" -
-
f ) Test Number
I 1 i I I 1 1 I 1 1
0 20 40 60 80 100
Load. 103 Ib steam/hr
Figure 3. Location 3—NO emissions as a function of load.
-------
10
s
Q
a 0.5
Q
i
O
0.1
Location 3 Wood Bark Boiler
Test No. 3-24
21% Wood Bark. 79% Coal
Load: 34.9 Mg/h
O2 = S.S%
Brink Impactor
Downstream of Multiclone
0.01 2 5 10 20 50 95 9899
Cumulative Proportion of Impactor Catch. % by mass
Figure 4. Aerodynamic particle diameter— Iow-N0t conditions.
140
120
^ 100
80
i'
20
i i i i i i i i i i i r
Steam Flow:
kg/s(103lb/hr)
O 19.31153)
n 22.7(180)
A 13.5(107)
Fuel: Hogged Wood (100%)
O
5/2-2
5/3-2 5/3~3A
1
1
024 6 8 tO 12 14
Stack Oxygen, %. dry
Figure 5. NO emissions as a function of stack oxygen for three
loads in a hogged fuel boiler.
-------
i
ti
I
3870
(9.0)
3440
(8.0)
3010
(7.0)
2580
(6.0)
2150
(5.0)
7720
(4.0)
^ '290
.o (3.0)
860
(2.0)
430
- 5/2- IE
5/7-2
5/2-4A
Q 5/4-2A
40 50 60 70
NO. ppm. dry @ 3%. O2
80
90
100
Figure 6. Location 5—Boiler total paniculate emissions as a function of NO emissions.
-------
Table 4. Summary of POM Analyses For Location 5—Wood-Fired Spreader Stoker
Low NO, Test
XAD-2
POM
Phenanthrene
Anthracene
Methyl Anthracenes/
Phenanthrenes
Fluoranthene
Pyrene
Methyl Pyrene/ Fluoranthene
Benzoldphenanthrene
Benzfa/anthracene
Chrysene
Methyl Chrysenes
Dimethylbenz anthracenes
Benzofluoranthenes
Benzofajpyrene
Benzofe/pyrene
Perylene
Methylcholanthrenes
lndeno(1,2. d-cdjpyrene
Benzofg.h. i/perylene
Dibenz anthrancenes
Dibenzpyrenes
Coronene
TOTAL"
Sample volume, rrf
H9
0.7
A/0*
0.7
<0.5
NO
3.2
<0.4
3.4
0.8<1.3
1.0
0.7<1.2
ND
ND
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
ft U.S. GOVERNMENT PRINTING OFFICE: 1983-759-102/0783
------- |