-------
The western coal differed significantly from the other coals not only
in nitrogen content, but in oxygen content. This western coal contained
12.5% oxygen while the other coals averaged about 7%. It is theorized
that the high oxygen content in intimate contact with the fuel nitrogen
enhanced the low temperature conversion of fuel nitrogen to nitrogen
oxides and contributed significantly to the overall high nitrogen
oxides level.
For Test 165 the nitrogen oxides emissions were the lowest of
any coal-fired boiler; 100 ng/J (164 ppm). The fuel averaged about
0.94% nitrogen and the fuel oxygen was 9.9%. It is believed that the
low nitrogen oxides emissions are related to the furnace geometry and
the nature of the combustion process. The boiler was equipped with a
traveling stoker chain grate burner which combusts large coal particles
at a relatively slow rate. The combustion equipment was in poor
condition. Visual examination of the furnace during the tests revealed
low intensity combustion flames of a very lazy and random nature. The
addition of overfire air actually improved the mixing of fuel and air
and resulted in an increase of nitrogen oxides. The excess air was
extremely high and the heat release rate per unit volume was comparatively
low, 0.496 [GJ-hr -m (0.013x10 Btu-hr -ft )], considering the
rated boiler capacity.
5.3.1.2 Temperature -
The effect of oil temperature, or viscosity, on nitrogen oxides
emissions was investigated at five locations during the course of the
program. In all cases the tests were conducted with steam atomized
No. 6 fuel oils over a temperature range of 69°C to 121°C (157°F to
250°F). As seen in Figure 5-37 no consistent trend was observed,
although in all cases the changes in nitrogen oxides emissions were
less than 10%.
169
-------
200
150..
w
a
w
100
50..
e_L
350
300
250
200
o
PO
-o
I
150
100
50
178
The numbers beside
curves are test numbers.
The normal oil
temperature
Test!
Test Fuel Capacity Load!
Series Oil GJ/hr GJ/hr
173
129
120
178
34
#6
#6
#6
#6
#6
84
74
158
42
7.4
150
200
250
I
340
350
360
370
I
380
I
390
OIL TEMPERATURE
Figure 5-37. Effect of fuel oil temperature on total nitrogen oxides
emissions
6001-43
170
-------
The main property change due to increasing the oil temperature
is the reduction of the viscosity; for a typical No. 6 oil the viscosity
will drop from 400 SSU to 110 SSU as the temperatures is increased from
240 K to 365 K (150°F to 200°F). Number 5 and 6 oils are normally
atomized in the viscosity range of 150 to 300 SSU. Fundamentally, as
the temperature decreases and the viscosity increases the energy required
to overcome viscous effects increases, and this detracts from the energy
available for droplet breakup resulting in coarser atomization. This is
minimized somewhat in steam or air atomizers which produce much smaller
drop sizes than their mechanical counterparts since the energy contained
in the atomizing gas stream can be independent of the quantity of liquid
being atomized. Thus, one would not expect nitrogen oxides emissions to
be greatly dependent on oil temperature or viscosity for air or steam
atomized systems.
Another field test crew from KVB, Inc. tested a twin boiler at
Location 38 for the effect of fuel oil temperature on particulate
emissions. They found that the particulate emissions as indicated
by the mass monitor showed a 57% decrease with increasing oil
temperature as shown in Figure 5-38 and a further decrease with increase
oil atomization pressure.
5.3.2 Burner Characteristics
5.3.2.1 Burner Tune-up -
The effect on total nitrogen oxides emissions of tuning the
burner was determined by first measuring the emissions from a boiler
that had not been tuned for a year or so. The local serviceman for
the burner manufacturer then was brought in and he tuned the boiler
to the manufacturer's specifications. Tuning involved examining the
nozzle for worn tips, adjusting the spray angle to make sure unburned
fuel did not strike the side or rear walls of the furnace and adjusting
the flame length so it did not wash the side or rear walls. Much of
this is done by means of adjusting the amount and swirl of the combustion
air.
171
-------
^ICULATE EMISSIONS
ng/J
to cj *». t
o o o c
1 i i i
10.
0
IP
0.2
mm
0.15
m
3
-M
m
*° 0.1
X.
(0
H
0.05
m*
0
\
V
N
Atorslzation
Pressure
kPa
(psig)
k585
(70)
722
O (90)
360
370 380 390
OIL TEMPERATURE^ K
400
Figure 5-38.
Effect of fuel oil temperature and of atomization pressure
on solid particulate emissions, No. 6 oil fuel. "•' '
6001-43
172
-------
Oil Fuel; The chief effect of burner tune-up was a reduction
in carbon monoxide emissions rather than a significant reduction of
nitrogen oxides emissions. During Test 108 at Location 1, the carbon
monoxide from oil fuel was reduced from 139 to 38 ng/J (407 to 110 ppm)
and during Test 112 at Location 27, from 40 to zero ng/J (116 to zero
ppm). During Test 108 this was accomplished by raising the excess
oxygen from 2.7 to 3.8%. The increase in excess air and stack tempera-
ture compensated for the decrease in carbon monoxide in the stack gases
and the heat loss efficiency did not change. After the tune-up during
Test 112, it was possible to operate at a lower level of excess oxygen
than originally without any carbon monoxide in the stack gases and the
efficiency increased by 1% from 81 to 82%.
There was a 13% reduction in the nitrogen oxides emissions from
oil fuel during Test 112 at Location 27 after tune-up, but no reduction
during Test 108 at Location 1.
At Location 27 the particulates were relatively unaffected/
increasing by about 5%. At Location 1, however, the particulate
emissions increased substantially, i.e., doubled, after tune-up. This
increase may have been due to the impingement of the flame on the water
walls of this particular boiler. Even after tune-up there was a
substantial amount of impingement and the relatively cold water walls
may have quenched the flame and increased the creation of unburned
carbon particles. The spray angle was very large and there was not
time during the test for the burner manufacturer to secure and install
a smaller angle burner tip for test purposes.
However, at both locations the total particulate emission was
well below the Environmental Protection Agency limit for new units
of 43 ng/J (0.1 Ib/MBtu) for solid particulate alone.
173
-------
Natural Gas Fuel; With natural gas fuel, tuning the burner
resulted in an increase in nitrogen oxides at both locations. When
the excess oxygen was increased sufficiently during Test 106 at
Location 1 to eliminate the carbon monoxide, the efficiency decreased
due to the larger amount of excess air. For Test 110 at Location 27,
however, it was possible to decrease the excess oxygen and not incur
an increase in the carbon monoxide above ng/J (10 ppm) and the
efficiency increased slightly.
Summary! Thus for both oil and gas fuels, if the burner was
tuned to reduce the carbon monoxide to near zero and/or to improve
the flame texture and color, the nitrogen oxides emissions either
were unchanged or increased. Tune-up universally was successful in
reducing carbon monoxide, however. Reducing carbon monoxide to near
zero generally increased the heat input efficiency: e.g., 0.6 to 1.0%,
because the decrease in combustibles was slightly greater than the
corresponding increase in heated air exhausted up the stack.
In both instances when the fuel was oil and the particulates
were measured before and after tune-up, the particulate increased
rather than decreased when the burner was tuned.
It appears that the most effective way to reduce nitrogen
oxides emission by burner tuning is simply to reduce the excess oxygen
and accept some carbon monoxide, perhaps up to 35 ng/J (100 ppm). The
remaining combustibles in the exhaust gases are offset by the decrease
in excess air exhausted up the chimney and the heat loss efficiency is
not affected significantly.
174
-------
5.3.2.2 Coal Burners -
The data shown in Figure 4-3 have been analyzed to determine
if certain types of coal burners as a class, such as underfed stokers,
spreader stokers and pulverizers, emit less nitrogen oxides than other
types. It was found that boilers equipped with spreader stokers and
pulverizer type burners had the highest nitrogen oxides emissions.
Chain grate and underfed stokers had the lowest emissions. The chain
grate and underfed stokers had less intense flames and larger furnaces
than the others, and this combination of less intense combustion and
large furnace produced a lower level of nitrogen oxides emissions.
The chain grate burner of Test 165 produced the lowest emission
levels, as Figure 4-3 indicates. Nitrogen oxides were 100 ng/J (164
ppm) and particulates were 175 ng/J (0.41 Ib/MBtu). The emissions from
underfed stokers were the next lowest, 134 to 208 ng/J (220 to 340 ppm)
nitrogen oxides, but all boilers with this type of firing were small,
less than 63 GJ/hr capacity. Spreader stokers produced nitrogen
oxides emissions of 220 to 336 ng/J (360 to 550 ppm) . Particulates
from spreader stokers ranged from 103 to 1300 ng/J (0.24 to 3.05
Ib/MBtu), depending on whether the samples were taken before or after
the dust collector.
The cyclone burner of Test 32 was a high emitter of nitrogen
and a low emitter of particulate. These emissions were what one would
expect from the very small volume furnace and a very intense flame of
this type of burner.
The highest nitrogen oxides emissions were from the pulverizer
at Location 31, Tests 131 and 169. The reason for these high emissions
is not known. Originally measurement error was suspected, and the
field crew returned two months later and repeated the test. The results
of Test 169 duplicated those of Test 131, so the high emissions appear
to be real.
175
-------
Particulate emissions from coal burning boilers were slightly
dependent on burner type. Pulverized coal burners generally produced
more particulates than stoker equipped boilers, as is discussed in
Subsections 4.2 and 5.4.
5.3.2.3 Oil Burners -
The types of oil atomizers evaluated during the program were
steam, air, pressure - mechanical, and rotary cup. 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-atomized. The majority 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
temperature 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 0.446 MPa (50 psig), and oil pressure was about 0.377 MPa
(40 psig) at top load. The No. 5 oils were normally fired at from
545 to 355 K (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 365 K (200°F) at the burner, and the
steam/air and oil pressures at the burner were similar to No. 2 and
5 oil atomizers.
A special series of tests, Tests 1, 2, 52, 53, 54, 195, 196,
200, and 201 were run at Location 19 to investigate the effect of the
oil atomization method and oil grade on the total nitrogen oxides and
particulate concentrations. The boiler used was a Keeler Company
packaged steam generator rated at 18.5 GJ/hr (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 1.14 MPa (150 psig).
During this test series, both No. 6 and No. 2 fuel oils were tested with
176
-------
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 5-7 and Figure 5-39. It should be noted that
the No. 2 and No. 6 oils (Tests 1 and 2) used for these tests were the
extremes in API gravity, carbon residue, ash, nitrogen, and sulfur
(see Table 6-1). As a result, relatively high nitrogen oxides and
particulate values were measured for Tests 1 and 2 with No. 6 oil
and low values were measured for No. 2 oil.
The field test measurements from Tests 1, 2, 44, 45, 52, 53,
54, 56, and 57, which were done during Phase I are summarized in
Table 5-8. This table is an excerpt of Table 4-1 of the Phase I
Final Report, Reference 4.
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 0.62 MPa (75 psig) and 93°C
(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. These tests were
repeated during Tests 200 and 201 with a different shipment of No. 6
fuel oil.
As shown in Figure 5-39, the nitrogen oxides emissions increased
with increasing excess oxygen up to about 5% excess oxygen where a
maximum nitrogen oxides value of 213 ng/J (380 ppm) was reached and
beyond this oxygen level the nitrogen oxides emissions decreased with
increasing excess oxygen. The minimum excess oxygen 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 65.5 ng/J (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.
177
-------
Table 5-7. EFFECT OF OIL ATOMIZATION METHOD ON TOTAL NITROGEN OXIDES,
PARTICULATE EMISSIONS AND BOILER EFFICIENCY
Test
NO.
1'
2
195"
200
198
203
44
45
52
53s
54
56
57
Oil
Grade
No. 6
No. 6
No. 6
No. 6
No. 6
No. 6
No. 5
No. 5
No. 2
No. 2
No. 2
No. 2
No. 2
Fuel
Nitrogen
(*)
0.44
0.44
0.14
0.14
0.14
0.14
0.10
0.10
0.006
0.006
0.006
0.02
0.02
Atomiza-
tion
Method
Steam
Air
Steam
Air
Steam
Air
Air
Steam
Steam
Air
Mech.
Air
Steam
Test
Load
GJ/hr
(103 Ib/hr)
15
(14)
16
(15)
15
(14)
15
(14)
15.1
(14.3)
14.8
(14.0)
18.6
(17.6)
18.3
(17.3)
15
(14)
15
(14)
13
(12)
16.8
(15.9)
16.6
(15.7)
Normal1
Excess
Oxygen
(%)
3.6
4.4
3.1
2.9
3.1
2.9
7.2
6.7
3.6
3.0
4.3
8.0
8.0
N0x
ng/J
(ppm)
196
(350)
187
(334)
95
(169)
91
(162)
75
(133)
73
(131)
99
(177)
90
(161)
36
(65)
54
(97)
45
(80)
65
(116)
66
(118)
Solid
Particulars
ng/J.
(lb/10 Btu)
62.1
(0.1447)
125
(0.2818)
8.60
(0.020)
10.8
(0.025)
9.90
(0.023)
18.1
(0.042)
17.5
(0.0448)
32.0
(0.0779)
14.6
(0.0339)
5.01
(0.0163)
4.96
(0.0151)
"~
"™
Boiler
Efficiency
(t)
85
85
84
83
82
83
86
86
85
85
85
85
96
Normal operating 02 level defined by burner manufacturer.
ppm is measured value corrected to 3% excess 0 dry.
Particulate data for Test No. 1 were taken for low air run
(2.3% oxygen).
A different shipment of No. 6 oil was used for Test 195, 200
than for Tests 1, 2.
Particulate data for Test 53 were taken for high air run (4.3% oxygen)
6001-43
178
-------
400
200. .
150
w
8
H
X C-
°s
§
- 100
t-i ^
.2 O<
Z C
Normal Excess Oxygen Level
Test No. 1
Steam
Atomization
No. 6 bil
Test No. 2
Air Atomization
No. 6 Oil
Normal Excess Oxygen Level
Test No. 195,196
Steam Atpmizatio
No. 6 Oil
Tests 200i 201
Air Atomized No
Test No. 54
Test No. 53
FivLzation
1 Atomization
—-Test No. 52
Steam Atomization
No. 2 Oil
50..
3.0 4.0 5.0
EXCESS OXYGEN, DRY, %
Figure 5-39.
Effect of oil atomization method and excess oxygen level
on the total nitrogen oxides emissions.
179
6001-43
-------
Table 5-8. PHASE I FIELD TEST MEASUREMENTS
00
o
Mt
•». '
1-13
i-a
1-11
2-5
2-4
2-6
44-4
44-1
44-3
44-6
45-7
4S-1
45-J
4S-S
51-1
52-S
52-2
53-1
53-6
53-2
54-S
55-1
3£~1
turn ir
i»e
St««a
St««M
Steu
Air
Air
Air
Mr
Air
Air
Air
Steaa
Stem
Steui
stem
Air
Steaa
Stea*
Air
Air
Air
Mech
Air
Air
IMC
T»el
16 Oil
16 Oil
16 Oil
16 Oil
16 Oil
16 Oil
15
15
IS
|5
IS
IS
IS
IS
15
12
12
12
12
12
12
12
Twt
*»•
Ba*«-
llm
Low
Load
Low
Air
Base-
line
Low
Air
LoAt*
Press.
Base-
line
Hi
Load
Low
Load
Air
Base-
line
Hi
Low
Low
Air
lines
Hi Loac
Base-
line
Low
Air
Base-
line
Hi
Air
Low
Air
Base-
line
Air
Base-
Hi Load
line
cwxitr
vw
ttf/Vrl
7.95
(17.5)
7.95
(17.5)
7.95
(17.5)
7.95
(17.5)
7.95
(17.5)
7.95
(17.5)
8.17
(18)
8 17
(18)
fi 17
(18)
B 17
(18)
8.17
(18)
8.17
nut
8.17
8.17
(16)
(7)
7.95
(17.5)
7.95
(175)
7.95
(17.5)
7.95
(17.5)
7.95
(175)
7.95
(175)
(17.5)
4.99
81?
BM
(18)
«Mt
LM
It/he)
«M/hrl
6.36
(14)
2.72
(6)
6.36
(14)
6.81
(15)
6.36
(14)
6.36
(14)
7.99
(17.6)
(22)
341
(7.5)
{17.61
7.85
(17.3)
9.99
3.22
7.72
(17.01
(6.7)
6.36
(14)
6.36
(14)
£.36
(14)
6.36
(14)
6.36
(14)
S.4S
(12)
(12)
5.13
"i 1 *
(15.7)
^.^
%v
3.C
11.0
2.3
4.4
2.8
4.7
7.3
6.7
4.7
6.7
3.8
6.3
3.6
2.6
3.0
4.3
1.6
4.3
4.7
•OB
•* LiM
*/*»!
to*
-sot
(350)
.986
(428)
.763
(331)
.770
(33D
.638
(277)
.687
ijqa)
.408
(177)
(188)
(154)
(1831
.371
(161)
.394
.355
.350
(152)
(275)
.ISO
(65)
.145
(63)
.224
(97)
.235
(102)
.198
(86)
.184
(80)
(BO)
.295
(118)
w>
*>X lift*
9/Meal
CH»>
.797
(346)
.961
(417)
.737
(320)
.760
(330)
.611
(265)
.675
(293)
.403
(175)
(185)
350
(152)
(182)
.371
(161)
.385
MR71
.346
.346
(151)
(270)
.147
(64)
.143
(62)
.224
(97)
.230
(100)
.196
(85)
.184
(80)
(78)
.290
(117)
"° J
-oU LtoJ
9/Ncal
in-i 1
.834
(362)
.933
(405)
.758
(329)
.774
(336)
.601
(261)
.664
(288)
.371
(161)
387
(US)
323
(140)
(164)
.357
(155)
.369
f 1Crt\
.327
.320
(139)
(260)
.147
(64)
.145
(63)
.212
(92)
.194
(84)
.184
(80)
(79)
.272
(105)
«,
efy
%
13.4
8.2
14.8
12.9
12.6
10.4
12 4
10 2
10.9
12.6
10.8
13.2
12.6
13.1
13.1
12.5
14.2
12. 0
11.8
CO
•/•C41
lK»>
0
(0)
0
(0)
.062
(45)
0
(0)
0
(0)
0
(0)
0
(0)
(0)
(0)
(0)
.018
(10)
.031
(2n\
0
.030
(20)
(0)
.069
(47)
.675
(485)
0
(0)
0
(0)
.272
206
0
(0)
(0)
0
<")
(0)
•c
t/Mcal
lit-)
.026
<32)
.022
(28)
.053
(60)
.002
(3)
.010
(12)
.OO 7
(9)
tern
«AK»1
(H-)
4.64
(1448)
4.42
(1378)
2.513
(784)
—
2.590
(808)
—
.199
(62)
.221
(69)
•O.
«/N&a
iw->
4.S6
(1424)
4.36
(1359)
2.487
(776)
2.567
(801)
.173
(54)
.199
(62)
Total
>*rtle.
a/Heal
U/taul
.2743
(.1524
. S238
(.2910)
.1175
(.0653)
.1402
(.0779)
.0680
(.0378)
.0295
(.0164)
.0349
(.0194)
— .
(0114
9«tU.
«/XC«l
(l/iatel
.2605
(.1447)
.5072
(.2818)
.0806
(.0448)
.1188
(.0660)
.0610
(.0339)
.0293
(.0163)
.0272
(.0151)
keilac
tm-
clwcr
%
-
81
86
87
87
88
83
83
63
83
84
BO
8(5
85
8S
•Kk*-
»cm
l*et •»•
—
;
1.0
-
-
-
*
-
6001-43
-------
Test No. 2: Air-Atomized No. 6 Fuel Oil. At the baseline
load of 15.0 GJ/hr (14,200 Ibs/hr) the oil pressure and temperature
at the burner were 0.36 MPa (37 psig) and 101°C (214°F) and the atomizing
air pressure at the burner was 0.31 MPa (30 psig). The nitrogen oxides
emissions increased by 6.6% with increasing excess oxygen over the range
investigated. The flame appearance changed with excess oxygen, and the
best flame characteristics occurred at the lower oxygen levels.
Particulate emissions of 125 ng/J (0.2910 lbs/10 Btu) were measured
for Test Run No. 2-6, which was substantially greater than the values
obtained with steam atomization 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 14.8 GJ/hr (14,000
Ibs/hr) operated with 0.55 MPa (65 psig) pressure, ambient temperature
oil and the steam pressure at the burner of 0.60 MPa (73 psig). The
nitrogen oxides emissions increased with increasing excess oxygen up
to about 4%, and between excess oxygen levels of 4 and 5%, the nitrogen
oxides emissions appear to reach a maximum value. A visible haze from
the smoke stack occurred at the lowest level of excess oxygen of 2.6%.
The baseline total particulate emissions of 16.25 ng/J (0.0378 lbs/10
Btu) were measured for this test at an excess oxygen level of 3.6%,
which is about average for steam-atomized No. 2 fuel oil.
Tests 195, 196, 200, 201. The tests were repeated with another
shipment of No. 6 oil and during tests 195, 196, 200, and 201 and the
results are tabulated in Table 5-7 and trends in nitrogen oxides with
excess oxygen are shown in Figure 5-41. The trends obtained with this
series of tests were similar to those obtained during Tests 1 and 2 with
steam atomization producing about 10% more nitrogen oxides than air
atomization. The difference in the absolute levels of the nitrogen
oxides emission is attributable to the nitrogen content of the two
different No. 6 oils. The oil of Tests 1 and 2 had a nitrogen content of
0.44% and for Tests 195, 196, 200, and 201 it was 0.14%.
181
-------
Test No. 53: Air-Atomized No. 2 Fuel Oil. At the
baseline steam flow of 14.8 GJ/hr the oil burner operated with 0.29
MPa (27 psig) oil pressure, ambient oil temperature and 0.26 MPa
(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 56.7 ng/j (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 2.03 MPa
(280 psig) for a boiler load of 12.1 GJ/hr (11,500 Ibs/hr). The
NOx values did not vary significantly over the excess O range
investigated of 3.7 to 6.6%. Particulate emissions of 8.34 ng/J
(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 in Figure 5-40 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 0 , 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 [56 ng/J (100 ppm)] with steam atomization
being the lowest NOx producer [39 ng/J (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.
182
-------
The boiler efficiency did not vary measurably due 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 nitrogen oxides emissions. For
No. 6 fuel oil, atomization resulted in the lowest nitrogen oxides
emissions at the normal operation oxygen level, but yielded substantially
greater particulate emissions than did steam atomization. For the No. 2
fuel oil tests, steam atomization resulted in the lowest nitrogen oxides
emissions and yielded the greatest particulate emissions. The air
atomization test had the greatest nitrogen oxides emissions and yielded
lower particulate emissions than the steam atomized test with No. 2 oil.
Mechanically-atomized No. 2 fuel oil nitrogen oxides and particulate
emissions were in between the air and steam results.
A second special series of tests, Tests 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 56 and 57 with No. 2 oil, the nitrogen oxides
emissions listed in Table 5-7 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 44 and 45, the
emissions with air atomization were greater than with steam, rather than
less, as for Tests 1 and 2 with No. 6 oil.
Tests 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.
183
-------
5.3.2.4 Oil Atomization Pressure -
During three test series data were collected to determine the
effect on nitrogen oxides emissions of changes in the pressure of the
atomizing fluid. The results were that when the fuel and/or atomization
pressure was increased the nitrogen oxides increased too. In the one
instance where the effect on particulate emissions was measured, they
decreased.
At Location 36 the pressure of the atomizing steam was varied
to determine the effect of atomization pressure on the nitrogen oxide
emissions. These tests were carried out in a steam-atomized watertube
boiler firing No. 2 fuel oil. At a steam rate of 55 GJ/hr (52,000 Ib/hr)
and an excess oxygen level of 5.9%, the steam atomization pressure was
varied from 340 kPa to 670 kPa (35 psig to 83 psig). The normal pressure
setting at this load was 590 kPa (59 psig). The effect of nitrogen
oxide emissions and smoke are shown in Figure 5-40. As the steam
atomization pressure was increased over the pressure range, the nitrogen
oxides emissions increased by 6% and the smoke levels decreased by two
Bacharach smoke numbers.
Although the changes in the total nitrogen oxides emissions
were small in these tests, the trend was consistent with that obtained
(4)
previously- The results of Test 2 of Phase I were that when the
pressure of the atomizing air was reduced the nitrogen oxides emissions
decreased.
The ASME heat loss boiler efficiency was not significantly
affected by this combustion modifications, remaining at 85% throughout
the tests at Location 36.
The effect of atomization pressure on nitrogen oxides and
particulate emissions was investigated at length by Laurendeau, et al.
They tested a boiler at Location 38 that was a twin to the one tested
under this program. One set of runs consisted of raising the fuel
184
-------
w
w
a
o
W Q 50-
o 2
§ CO
EH It)
H
2 h)
, X 40-
rij C
g
«Ninn _
dC
PI
*
& 80
•a
'&
<;n
163-2
163- 3Q
j
ii J
—
-O
163-1
Load: 55 GJ/hr of equivalent
Excess 0 : 5.9% saturated steam
1 i 1
^ 30 40 50 60 70 80 90
psig
ill1
400 500 600 700
kPa
H
CD O
O
sz
e
3J
O
DA
30
40
70
80
90
PSi
I-
400
500
kPa
600
700
Figure 5-40.
Effect of steam atomization pressure on total nitrogen
oxides emissions and smoke level.
6001-43
185
-------
and steam atomizing pressure while maintaining a 140 kPa (20 psig)
spread between them. Raising the pressures caused the particulate
emissions to decrease as is shown in Figure 5-38 when the atomizing
pressure was raised from 585 to 722 kPa (70 to 90 psig). However,
raising the pressure also caused the nitrogen oxides emissions
to increase. At fuel and atomizing steam pressures of 515 kPa and
650 kPa (60 and 80 psig) the nitrogen oxide emissions were about 140
ng/J (250 ppm). When the pressures were raised to 700 kPa and 825 kPa
(87 and 105 ppm) the NO emissions rose to about 163 ng/J (290 ppm).
5.3.2.5 Natural Gas Burners -
The majority of the industrial-sized boilers tested in Phases
I and II were equipped with multijet ring type natural gas burners.
This type of burner injects the gas jets radially inward (toward the
burner center axis) into a swirling air stream. The ring burner produces
good fuel and air mixing and the combustion starts in the fuel-rich
combustion zone near the injection orifices and continues downstream
of the burner throat. Ring burners have generally been found to be
low nitrogen oxides producers and have the capability of operating
fuel-rich over a large range of fuel flow rates with a stable flame.
Two boilers, used for Tests 75 and 77 had corner-fired furnaces
which use multijet gas nozzles where the gas and air streams are
injected into the boiler in parallel directions.
The boiler used for Tests 153-155 utilized a single burner
comprised of three multi-orifice gas nozzles. Combustion air is
supplied through primary and secondary air registers. The gas guns
186
-------
are located within the primary (inner) air passage and inject fuel
outward into the swirling air stream at an angle of approximatley 45°
from the center axis. The boilers used for Tests 149-152 and 207-212
were also fitted with gun type burners, however refinery gas was the
fuel.
Because of the lack of variation in gas burner designs, no
concrete conclusions could be drawn on the effect of burner design on
emissions. Emissions from the boilers equipped with nozzle type
burners were similar to those from boilers fitted with ring burners.
Generally, nitrogen oxide emissions from natural gas fired boilers
were found to be more dependent upon firing parameters, such as
burner heat release rate, excess air, and combustion air temperature.
5.3.2.6 Burner Size -
The total nitrogen oxides emissions measured during the program
were found to be larger when the burner size in terms of heat release
level in joules per hour was large. The relationship differed for each
of the three fuels, but in general it was found that the larger the
burner the larger the nitrogen oxides emissions. This relationship
suggests that an effective form of combustion modification would be to
use two smaller burners rather than one larger one. It is recognized
that coal fuel burning equipment sometimes can not be defined simply
in terms of individual burners size; 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 enters the furnace through a
burner port.
The relationship between the nitrogen oxides emissions and the
burner heat release rate or size for the natural gas and coal-fired
boilers is depicted in Figure 5-41. The coal fuel data on Figure 5-41
187
-------
w
w
Q
M
X
o
z
w
EH
H
2
600
500
400
id
0
o
^300
O
en
C
200
100
- 500
400
300
"N
O
2
tfl
(0
S1
200
100
1000
900
Symbol Fuel Type
/\ Coal
D
Natural Gas
Numerals within symbols
are test numbers.
50 100 150 200 6 250
RATED BURNER HEAT RELEASE,(10 Btu/hr)/burner
50
250
100 150 200
(GJ/hr)/burner
Figure 5-41. Effect of burner heat release rate on total nitrogen oxides
emissions for coal and natural gas fuels.
60C1-43
188
-------
show a strong dependence of nitrogen oxides emissions on burner heat
release level. The natural gas burner data, however, show a somewhat
lower dependence of nitrogen oxides emissions on burner size than
does the coal burner data.
(4)
The Phase I data had been interpreted as indicating that
the dependence of emissions on burner size was stronger when the
combustion air was preheated than when it was not. However, during
Phase II the measurements made for Tests 140 and 165 on boilers that had
preheated combustion air were the same level as those taken previously
for unheated combustion air. It now appears that the degree of
sensitivity of nitrogen oxides emissions to burner size depends more
upon the characteristics of the individual boiler than upon whether or
not the combustion air is preheated.
Figure 5-42 presents the effect of burner heat release level
on the total nitrogen oxides emissions for all of the oil-fired boilers
tested. The two data points for No. 5 oils which have nitrogen oxides
emission levels greater than 225 ng/J (400 ppm) are from tests where
the fuel oil 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 from oil fuels was not as great as previously discussed for
coal fuel, but was greater than for natural gas burners with or without
preheated combustion air. The type of oil atomizer did not seem to
affect this relationship. The No. 2 oil burners were smaller, all
being below 53 GJ/hr (50x10 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 131 GJ/hr
(125xl06 Btu/hr).
189
-------
800
Z
W
£H
H
a
g
8
iti
O
O
OJ
O
O
en
w
Q
H o
X g
NJ
O
O
ng/J as
M
O
O
700
600
500
400
dP
n
©
300
200
100
0«-
250
RATED BURNER HEAT RELEASE, (10 Btu/hr)/burner
I I | | I
0
50 100 150
(GJ/hr)/burner
200
250
Figure 5-42,
Effect of burner heat release rate on total nitrogen
oxides emissions, oil fuel.
6001-43
190
-------
5.3.3 Boiler Furnace Characteristics
5.3.3.1 Firetube Boilers -
A large number of firetube furnace boilers in addition to
watertube furnace boilers was tested during Phase I, and the results
are discussed in Subsection 5.1.1.4. Comparison of the test results
showed that the emissions of nitrogen oxides from firetube boilers
was less sensitive to changes in the excess air level than they were
(4)
from watertube boilers when burning the same fuel. The Phase II
testing concentrated on watertube boilers only.
5.3.3.2 Furnace Volume and Area -
Nitrogen oxides are formed at high temperature by the combina-
tion of oxygen and nitrogen, and the length of time that the products
remain at high temperature is critical to the formation of nitrogen
oxides. Consequently, the furnace heat absorption volume and area
were evaluated as design parameters which could influence the time/
temperature history. The furnace heat absorption volume parameter
was defined as the products of the furnace heat release per hour
divided by the furnace volume from the burner face to the end of the
furnace. The furnace heat absorption area parameter was defined as the
ratio of the furnace heat release per hour to the projected wall, floor
and ceiling areas of the furnace. The furnace heat absorption parameters
are listed in Table 7-1 of Section 7, Test Boiler Design Characteristics,
for each boiler tested in Phase II.
The heat absorption parameter data are discussed in Subsection
(4)
7.2.2 of the Phase I report, and the conclusion drawn from these
data was that the nitrogen oxides emissions were not dependent upon
furnace heat absorption area or volume. Phase II test results support
this conclusion.
191
-------
5.4 PARTICULATE EMISSIONS
5.4.1 Particulate Concentration
Figure 5-43 compares the change from the baseline value in
solid or filterable particulate emissions and the corresponding change
in the total nitrogen oxides emissions when the different combustion
modification techniques were applied. Data from both Phase I and Phase
II are included. The figure is divided into quadrants. One is labeled
"Best Quadrant" and a second "Worst Quadrant." The criterion for the
Best Quadrant with solid particulate emissions is that the effect of
the modification was to reduce the emissions of both the total nitrogen
oxides and the particulates. The Worst Quadrant is when the effect was
to increase both emissions.
The effect of the various combustion modification methods was
as follows:
1. Reduced excess air: This was the best method because the
particulate emissions decreased by as much as 30% in four out of the
six tests.
2. staged combustion air: The change in particulates was measured
in three of the six staged combustion air tests. In all three instances
it increased by 20 to 48% of the baseline level.
3. Burners-out-of-service: This method had the advantage that the
nitrogen oxides emissions always decreased and the boiler efficiency was
maintained. However, the particulate emissions always increased by from
25 to 95% of the baseline level.
4. Burner register adjustment: Readjusting the burner registers
had no significant effect on the particulate emissions.
5. Reduced combustion air temperature: Only one test was run and
the air temperature was increased by 11 K. The particulate emissions
decreased by 53%. However, a measurement of the particulate emissions
change with an air temperature reduction of 100 K was made on the boiler
192
-------
COMBUSTION MODIFICATION METHOD
ftir Temp. Reduction O Staged Air
Reduced Firing Rate O Burner Tuneup
Gas Recirc .
Excess Air
O Burner-Out-Of-Service
CO
w
Q
H
X
O
2
W
g
H
g
J3
H
w
O
u
4-
I
O - -+20
D - ^+1°
Best Quadrant ::xxX::x:::::::x:x:
Worst Quadrant
+ 40
+ 100
_-io
A O
o
O
_-30
-40
O
CHANGE IN PARTICULATES, %
Figure 5-43.
Effect of combustion modification methods on solid
particulate emissions.
193
6001-43
-------
at Location 38 as part of the work reported in Reference 17. The
results were reported in a private communication and were that no
change in particulate emissions occurred.
6. Flue gas recirculation: Recirculating 25% of the flue gas
resulted in a nitrogen oxides reduction of about 12% and a particulate
emission increase of about 15% of the baseline levels.
7. Reduced firing rate: In the one instance where the particulates
emission change was determined, the nitrogen oxides emission increased
by 10% and the particulate emission decreased by 45%. This was one of
the largest particulate emission decreases that was encountered.
8. Fuel oil viscosity: Another KVB, Inc. field test crew measured
the particulate emissions change as the oil temperature of a twin boiler
at Location 38 was increased from 351 K to 388 K. The particulate
emissions "showed a pronounced decrease with increasing oil temperature."
9. Burner tune-up: Tuning the burner reduced the nitrogen oxides
emissions and had no effect on the particulates in Test 112. During
Test 108 the emissions rose by 150%, because the tune-up resulted in
increased flame impingement and quenching on the water walls. The
results of Test 112 are deemed to be the more representative, since
tuning the burner resulted primarily in reducing the carbon monoxide for
a given level of excess air. Reducing the carbon monoxide emissions
should reduce, or at the worst not affect, the particulate emissions.
10. Fuel oil atomization method: No generalized conclusions can be
drawn from the five test sets that are listed in Table 5-9. There were
four instances where the atomization method of a given burner was changed
from steam to air. When the oil was No. 6 the particulate emissions
increased by from 26 to 101% of the baseline level. For the one case
when the oil was No. 2 the emissions decreased by 69%.
194
-------
Table 5-9. EFFECT OF ATOMIZATION METHOD ON THE
PARTICULATE EMISSION LEVELS
Test
Run
No.
1-11
2-6
195-1
200-3
198-12
203-7
44-4
45-7
52-5
53-6
54-5
Test Type
Baseline
Changed
Atomization
Baseline
Changed
Atomization
Baseline
Changed
Atomization
Baseline
Changed
Atomization
Baseline
Changed
Atomization
Changed
Atomization
Oil
Grade
No. 6
No. 6
No. 6
NO. 6
No. 6
NO. 6
No. 5
NO. 5
No. 2
No. 2
No. 2
Atom.
Method
Steam
Air
Steam
Air
Steam
Air
Air
Steam
Steam
Air
Mech.
Solid Part.
ng/J
(lb/106 Btu)
62.1
(0.1447)
125
(0.2818)
8.60
(0.020)
10.8
(0.025)
9.90
(0.023)
18.1
(0.042)
17.5
(0.0448)
32.0
(0.0779)
14.6
(0.0339)
5.01
(0.0163)
4.96
(0.0151)
Change in
Particulate
Emission
%
+101
-1-26
+83
+83
-69
-69
6001-43
195
-------
When the change was made from air to steam atomization on
another bu/.ner, rather than from steam to air, the particulate emissions
did not decrease as would have been expected from the -results of Tests
1 and 2, 195 and 200, and 198 and 203. Instead the emissions from
Tests 44 and 45 increased by 83%.
Apparently, the effect of atomization method on the particulate
emissions is unique to each fuel-burner-boiler combination and it cannot
be generalized.
12. Fuel oil atomization pressure: The results of one set of tests
that are discussed in Subsection 5.3.2.3 were that an increase in the
atomization pressure of 23% reduced the particulate emissions by 75%.
Although this was only one test set, extensive data were taken carefully,
and one can conclude that it is possible to reduce the particulate
emissions by increasing the atomization pressure.
5.4.2 Particulate Size
The effect of some of the forms of combustion modification on
the particulate size distribution also was determined and is discussed
in this section. Table 5-10 lists the combustion modification methods
that were investigated and the corresponding size distribution results.
Figure 5-44 shows the effect of the particulate size distribu-
tion of modifying the combustion of oil fuel by reducing the amount of
excess air/oxygen. Test 176 was run with the baseline amount of
excess oxygen of 4.3%, while Test 179 was run with 4.0% excess oxygen.
Reducing the excess oxygen from 4.3% to 4.0% reduced the proportion
of fine particulates from about 58% to about 50%. (The total nitrogen
oxides concentration dropped from 195 ppm to 174 ppm.) Apparently the
modified combustion resulted in a decrease in the proportion of the
smaller and an increase in the proportion of the larger size particulates.
196
-------
Table 5-10. PARTICUIATE SIZE DISTRIBUTION WITH COMBUSTION
MODIFICATIONS
OIL FUEL
Test
111
112
162-36
162-11
162-5
176
179
166-3
166-B
Location
27
36
37
35
Load
GJ/hr
(103lb/hr)
90(85)
90(85)
65(62)
63(60)
93(88}
34(32)
34(32)
116(110)
116(110)
Burner
or Oil
Type
PS 300
PS 300
NO. 2
NO. 2
No. 2
No. 6
No. 6
Chain
Grate
Chain
Grate
Proportion of Total Weight of Catch
Particles
Inhaled
Then
Exhaled
<0.5 \tm
%
60
-
1
3
0.3
31
27
11
18
Particles
In The
"Fine"
Particulate
Size Range
<3 pm
%
81
97
26
40
5
60
50
24
40
articles
Reducing
Visibility
by Hie
Scattering
0.4-0.7 lam
%
10
-
0.8
0.9
0.1
1
1
5
13
Soot
Included
No
NO
NO
No
No
No
No
NO
NO
Combustion
Modification
None (Baseline)
After Tuneup
None (Baseline)
Low Excess Air and
Registers Reset
Registers Reset
None (Baseline)
Low Excess Air
None (Baseline)
Low Excess Air
6001-43
197
-------
100
0.1
0.3
1.0 3.0 10
pm
100
Figure 5-44.
Effect of combustion modifications on particulate size,
Oil Fuel.
6001-43
198
-------
Also shown in Figure 5-44 is the effect on the particulate size
distribution of modifying the fuel and air mixing by resetting the
burner registers. The test fuel here was No. 2 oil, and the testing
was done at a very low firing rate, i.e., 33% of capacity. The upper
curve for Test 162 was drawn from data taken after the registers had
been reset. The most striking effect was that the proportion of fine
particulate rose from a baseline value of about 26% to about 40%.
When the fuel was coal burned on a chain grate the effect of
reducing the excess air was different. The reduction in the percentage
of excess air was 0.4. This is illustrated on Figure 5-45. Reducing
the excess air raised all of the proportions of the total weight of
the catch, rather than reducing them as with oil fuel.
When the firing rate of the boiler used for Test 162 was
raised from a level of 65 GJ/hr that was 33% of capacity to 93 GJ/hr
(47% of capacity) and the registers reset for the lowest nitrogen oxides
emissions the proportion of fine particulate decreased from about 26%
to about 5%.
The effect of modifying combustion by tuning the burner is
illustrated in Figure 5-46. The data for the upper curve were taken
before the oil burner at Location 27 was tuned and those for the lower
curve were taken after. After tuning there was a larger proportion of
the fine particulate. No data were available below on aerodynamic size
of about 0.5 pm because the back-up filter was damaged and could not
be reweighed.
199
-------
100
a
o
CJ
OJ
O
f4
U
<
M
4)
4J
(0
-U
to
X! to
4J 03
flj
J*~4 ^^
4-* X)
B to
O 10
H €)
O C
&< o
O -i-l
ft M -H
O 13
W Qi
> O
H M
O
M
/
1
Y
6
•*
0.1 0,3 3
e
.•
_^-^
x?T^
Test
"~Chai
VH
**-
A
A
****
*a»
ma
am
Ml
-AU_J=/__,
\
:^~
• ==C
A
3 166 arid 1G8, Location
rt Grate Stoker Fired
-0 3.0
No
•
35-
0 30 T^
00
figure 5-45. Effect of low air modification on
particulate size. Coal fuel,
6001-43
200
-------
30
g „ 10
f*°i _C!
IVE OF C
jportion less than
iiameter, % by
H W
0 0
COMULAT]
i.e. prc
stated c
o o
M Ul
0 - -
_Befc
— Tune
»rt>
•up.
f
r
t
§
\
— LL
«M<
'
i
t
«
•*
r?
II
e
a
ter
leup
i flLg
S3
•m
mt
•1 0.3 l.o 3.0
m
K
^^^
: . — .
. — |
Symbol Test
O in
i i 1
10 30
[ML
100
DIAMETER, ]im
Figure 5-46,
Effect of burner tune-up on particulate size,
oil fuel
6001-43
201
-------
5.5 BOILER EFFICIENCY
The effect of the various techniques used to reduce nitrogen
oxide emissions on boiler thermal efficiency have been evaluated and
are discussed in this section. The test data are from Phase II in
which the nitrogen oxides reducing techniques of low excess oxygen,
overfire air, burners out of service, reduced combustion air preheat,
and flue gas recirculation were investigated. The data are presented
in graphs wherein the percentage change of nitrogen oxides (the change
in NOx resulting from a particular combustion modification divided by
the baseline NOx level) is plotted versus the corresponding change
in boiler efficiency.
In general, the nitrogen oxide reduction techniques of low
excess oxygen firing, burners out of service, and flue gas recirculation
resulted in boiler efficiencies equal to or better than baseline
levels. Staged air and reduced combustion air preheat produced a
degradation of efficiency.
5.5.1 Effect of Excess Air
The effect of low excess air firing on boiler efficiency is
illustrated in Figure 5-47. The majority of the data points are located
in the best quadrant, i.e., where a reduction in emissions is accompanied
by an increase in efficiency. Efficiency was bettered by as much as 2.5%
in two cases. On a coal-fired boiler reducing excess oxygen resulted in
a 44% reduction in nitrogen oxide emissions along with a 2.0% increase
of boiler efficiency. In three cases with gas fuel, lowering the excess
oxygen resulted in an increase in emissions at a higher efficiency.
This behavior of increasing nitrogen oxides with decreased oxygen is
unusual and is discussed in Section 5.1.1. In two instances, lowering
excess oxygen resulted in a decrease of efficiency; however, the
magnitude of the changes were insignificant compared to the accuracy of
the procedure used to determine them.
202
-------
t
<#>
u
H
U
4-
I
BEST QUADRANT
<9-
+2
CD
-50
4—1
•30
D
D
+10 +30 +50
-1
-2
-3
WORST QUADRANT
CHANGE IN TOTAL NITROGEN OXIDES, % •*• +
^ Coal Fuel
O Oil Fuel
Natural Gas Fuel
Figure 5-47. Effect on boiler efficiency of reducing the excess
combustion air.
6001-43
203
-------
As a whole, the efficiency of the boilers tested during Phase
II responded as expected to the effects of reduced excess oxygen. In
agreement with Phase I results, the degree of efficiency increase
averaged to be 0.5% for each 1.0% decrease in excess oxygen.
5.5.2 Effect of Staged Air
The effect of staged combustion air on boiler efficiency is
illustrated in Figure 5-48. Generally, the reduction of nitrogen oxide
emissions by using staged air had a negative affect on efficiency.
This behavior was to be expected since staged air normally requires
that the level of excess oxygen be maintained at higher than baseline
levels to assure complete combustion. This greater quantity of heated
air being exhausted through the stack contributes significantly to the
negative influence on efficiency. A few boilers exhibited increases
in efficiency (best quadrant) when staged air was used. These boilers
had staged air ports which were part of the original boiler design and
were therefore more carefully sized and located.
5.5.3 Effect of Burners Out of Service
Figure 5-49 presents the effect of burners out of service on
boiler efficiency. The efficiency changes were generally small, 0.6%
or less, but were mostly in the positive direction. One would expect
the effect of burners out of service to be similar to staged air since
both techniques involve staging combustion. The quantity of test data
from burners out of service is small, making it difficult to draw any
concrete conclusions.
5.5.4 Effect of Combustion Air Temperature
The effect of varying the combustion air preheat temperature is
shown in Figure 5-50. As expected, lowering the temperature to reduce
emissions resulted in a degradation of boiler efficiency, because a
reduction in air preheat was accompanied by an increase of flue gas
temperature. The five instances where the efficiency increased were
204
-------
+
•K
dP
B
M
U
H
Cn
lz
H
W
0
u
4-
i
BEST QUADRANT
•••
•w
D
All/
_+3
_ +2
udx
•
-so Or30 Df\ 1 +1° +3° +5°
u ^ A
0 ^ -
D
A
•••
D
D D
D
— i
- -2
- -3
WORST QUADRANT
- «- CHANGE IN TOTAL NITROGEN OXIDES, %
Coal Fuel
Natural Gas Fuel
Figure 5-48. Effect on boiler efficiency of staged combustion air.
6001-43
205
-------
+
t
dp
£
U
1
H
H
fe'
W
H
W
a
o
*
i
BEST QUADRANT
•••
WM
D
an
i m n i P
— r 1 1 A 1 r
-50 O £^30 -10
^^
^•^
_+3
+2
^^
+1
d 1 1 1 —
^H 1 1
+10 +30 +50
- -1
2
- -3
WORST QUADRANT
CHANGE IN TOTAL NITROGEN OXIDES, %
Coal Fuel
O Oil Fuel
D Natural Gas Fuel
Figure 5-49.
Effect on boiler efficiency of operating with burners
out of service.
6001-43
206
-------
+
t
rr
LNGE IN EFFICIENi
3
u
4-
i
BEST QUADRANT
•M
«•
Q
0
1 1 n
1 1
-50 -30 -10
DC
«••
o<&
•^
a o
^M
^
+3
••^
-4-2
-1-1
D| Pi 1°
1 1 1 1
+10 Q +30 +50
J
.=•1
-2
iHB>
_-3
WORST QUADRANT
CHANGE IN TOTAL NITROGEN OXIDES, % •+ +
^ Coal Fuel
O Oil Fuel
D Natural Gas Fuel
Figure 5-50.
Effect on boiler efficiency of the combustion air
preheat temperature.
6001-43
207
-------
tests where the air temperature was raised, rather than lowered,
e.g., Runs 130-1, 142-1, 118-1. Efficiency losses were as high as
3.3% with a 32% reduction of nitrogen oxides emissions from a coal-
fired boiler where the air temperature was reduced from 365 K to 355 K
by opening a by-pass duct (Test 138-2). If reduced air preheat is to
be adopted as a permanent nitrogen oxide emissions reduction technique
for a particular boiler, the stack losses can be recouped by redesigning
the steam side of the boiler for more heat absorption. An example of
this wou.ld be the installation or enlargement of an economizer.
5.5.5 Effect of Flue Gas Recirculation
The effects of flue gas recirculation on boiler efficiency
are shown in Figure 5-51. Also illustrated are the effects of flue
gas recirculation combined with staged air. Flue gas recirculation,
per se, had only small effects on efficiency. The changes were 0.6%
or less and varied from positive to negative. However, when sidefire
air was added, efficiency dropped by about 1.5% as would be expected
due to the necessary increase in excess oxygen for complete combustion.
5.6 GENERAL NITROGEN OXIDES EMISSIONS CORRELATION
A general correlation of nitrogen oxide emissions from
industrial boilers was developed using the test data from Phase I and
Phase II. The correlation relates nitrogen oxides emissions to three
boiler operational factors. These factors are (1) the excess air level,
(2) a term describing the rate of heat input, and (3) the nitrogen
content of the fuel. The correlation holds for all combination of
boilers and fuels tested during the program. To the knowledge of the
authors of this report, this is the first of any such general correlation
and is believed to be of major significance.
208
-------
<#>
u
z
w
H
u
H
BEST QUADRANT
--+3
o
w Q
2
Open symbols = flue
gas recirculation
only
Solid symbols = flue
gas recirculation
plus overfire air
+1
I I I I I I
w
O
-50 D - -10
+10 +30 +50
"*•»!
-2
-3
WORST QUADRANT
- •«- CHANGE IN TOTAL NITROGEN OXIDES, %
Q Oil Fuel
D Natural Gas Fuel
Figure 5-51.
Effect on boiler efficiency of flue gas recirculation
and staged combustion air.
6001-43
209
-------
An overall review of nitrogen oxide emissions test data from
Phase I and Phase II revealed that there was no single design or opera-
tion factor that could provide an acceptable correlation for the levels
of emissions from industrial boilers. Nitrogen oxide emissions were
dependent in varying degrees upon fuel properties, excess air, boiler
design, boiler firing rate, combustion air temperature, etc. The
effects of these factors on emissions are discussed individually
earlier in this section of the report, but no one factor correlated
with all of the emission trends that were encountered.
For the majority of test cases it was found that nitrogen oxide
emissions consistently increased along with an increase in two factors:
excess air and a factor describing the rate of heat release. This
second factor was more specifically defined as the ratio of total heat
release per unit furnace-heat-absorbing-area. A third factor was the
fuel nitrogen content. For coal and, especially, oil fuels, nitrogen
oxide emissions increased as the fuel nitrogen content became greater.
It was found that by using the rate of heat release as the
basic parameter and then correcting for the excess air level and fuel
nitrogen, a reasonable correlation of all the data could be achieved.
The correlation parameter was formulated as the produce of the three
individual factors and was as follows:
NO = (1 + 46N)(TA)(Q/A)
X
where N = fuel nitrogen content, % by weight
TA = fraction of theoretical air
-1 -2
Q/A = heat release per unit heat absorption area, joules-hour -meter
The nitrogen factor of (1 + 46N) was developed from the field
test finding that the proportion of conversion of fuel nitrogen to nitrogen
oxides was 46%. This conversion factor is discussed in Sections 5.3 and
6.0. The unity portion of the term was included in the nitrogen factor
to account for nitrogen oxide emissions resulting solely from the thermal
210
-------
fixation of atmospheric nitrogen (for example, when the fuel burned
contained no bound nitrogen, as with natural gas fuel).
The fraction of theoretical air provides for the excess air
effect on nitrogen oxides. One hundred percent of theoretical air is
the stoichiometric air required for complete combustion under perfect
conditions. Anything greater than 100% is excess air. For example,
when a boiler is operated with 55% excess air the "TA" factor would
be 1.55.
As mentioned above, the heat release factor is the ratio of
the heat released by combustion to the furnace heat transfer area.
It is the product of the full load fuel flow, fuel heating value,
and the fraction of the boiler load for the test divided by the area
of the furnace heat transfer surfaces surrounding the flame. This
area factor is very difficult to evaluate, since a boiler furnace
usually has an odd shape and a variety of waterwall tube sizes and
spacings. A significant amount of scatter in the correlation data
is caused by this uncertainty in the actual heat absorbing area.
The results of the correlation are presented in Figures 5-52
and 5-53, wherein the nitrogen oxide emissions level is plotted versus
the correlation parameter. One variable that affects nitrogen oxides,
but was not taken into consideration for the correlation, was the
temperature of the combustion air. For this reason, two plots were
made, one for boilers with ambient temperature combustion air and one
for preheated air. The plot for preheated air has more scatter than the
ambient air plot. This is because the amount of preheat temperature
varied significantly and the variation in ambient temperatures was
quite small. Additional scatter in the data bands may be due to
different burner designs and fuel oil atomization schemes, fuel oil
temperature, and coal particle size.
211
-------
1000
N)
H
to
CO
w
Q
H
X
o
Z
W
0
frf
H
z
a
<
fsj
O 100
*
PI
©
>1
u
•o
1
10
NOx = 1.51[(1+46 N)(TA)(Q/A}]
Q - Natural fuel
O - Oil fuels
^-J - Coal fuels
.362
ft2
Figure 5-52. MOx correlation for boilers with ambient combustion air.
6001-43
-------
1000
to
i-j
u>
NOx = 9.21 [(1+46 N) (TA) (Q/A)]
- Natural fuel
O - Oil fuels
- Coal fuels
Figure 5-53. correlation for boilers with combustion air.
6001-43
-------
Curves have been drawn through the data points by eye and
the curves behave according to the following equations:
and
A = 1.51 B°"36 for ambient air
A = 9.21 B°' for preheated air
where
A = nitrogen oxides emissions in ppm, dry @ 3% 0,
B = (1 + 46N)(TA)(Q/A).
214
-------
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 nitrogen 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 ethane 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.
In evaluating the effects of oil parameters on emissions, the
degree of sameness and difference from one oil to another should be
considered. All oils were 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
215
-------
material cause variations in chemical composition and characteristics.
Typically, crude oil is further processed and segregated into fractions,
classified for commercial purposes as No. 1 through No. 6, where each
oil designation has a specific allowable range of properties. The
result is that a given grade of oil from two sources will typically be
very similar in chemical and physical properties and in nitrogen oxides
emission characteristics. Variation will exist due to location differ-
ences, and these variations may sometimes be magnified by blending
procedures which can result in unusual characteristics. One effect of
this situation is that correlations of emissions with a particular oil
property become somewhat questionable. It is not clear whether emissions
and API gravity have a causal relationship or whether gravity indicates
a particular oil which has a certain typical fuel nitrogen content
and consequently a characteristic level of nitrogen oxides emissions.
Fuel nitrogen content is known to be very important and is discussed
in detail, and other properties such as ash and sulfur content are
discussed because of their effect on particulates and sulfur oxide
emissions.
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.
6.1 NATURAL GAS
Table 6-1 lists the properties of the gaseous fuels which
were tested in Phase II. Natural gas comprised the majority of
the gaseous fuels. Refinery gas was tested on one boiler and a
mixture of natural and refinery gas was tested in another instance.
216
-------
Table 6-1. FUEL ANALYSIS SUMMARY
Gas Fuel
Test No.
101
104-106
109-110
113-115
122-125
140-148
149-152
153-155
180-185
190-194
207-212
Type
of Gas
Nat.
Nat.
Nat.
Nat.
Nat.
Nat.
Ref .
Nat.
Nat.
Nat.
Nat.+
Ref.
CH4
%
94.52
94.52
90.00
88.11
88.11
91.8
76.40
93.91
96.99
97.26
48.43
C2H4
C2H6
%
4.26
4.26
2.90
4.33
4.33
5.84
5.10
4.37
1.98
1.82
10.96
c#
C3HB
0.20
0.20
0.40
0.68
0.68
0.51
5.30
0.71
0.10
0.027
1.68
C4;8
c'>
0.062
0.062
Trace
0.15
0.15
0.08
0.40
-
0.04
0.08
0.15
%""
0.032
0.032
-
0.04
0.04
0.01
0.20
-
0.01
-
0.02
C,H
V4
0.032
0.032
-
0.02
0.02
0.01
-
-
-
-
-
H2
%
-
-
-
-
-
-
8.80
0.25
-
-
35.67
*?
0.35
0.35
0.10
0.67
0.67
-
1.40
0.74
0.60
0.37
-
°l
0.12
0.12
-
-
-
-
-
-
-
0.01
0.02
N2
%
0.37
0.37
6.60
5.96
5.96
-
1.60
-
0.28
0.28
2.68
H20
%
0.05
0.05
-
0.04
0.04
-
-
-
-
-
-
Density
kg/m3
0.725
0.725
0.746
0.765
0.765
0.720
0.765
0.733
0.710
0.705
0.576
Ib/ft3
0.0447
0.0447
0.0460
0.0472
0.0472
0.0444
0.0472
0.0452
0.0438
0.0435
0.0355
Higher
Heating Value
GJ/n>3
0.0388
0.0388
0.0364
0.0373
0.0373
0.0391
0.0391
0.0390
0.0377
0.0381
0.0310
Btu/ft3
1042
1042
976
1000
1000
1050
1050
1047
1011
1023
831
6001-43
-------
The natural gases were composed mostly of methane with small
amounts of ethane and traces of heavier hydrocarbon gases. The
methane contents varied from 88 to 97% and the ethane proportions
were between 1.8 and 5.8%. The nitrogen contents of natural gases
varied significantly, from zero to as high as 6.6%. Nitrogen in
natural gas does not add significantly to the production of nitrogen
oxides as with liquid or solid fuels. The reason is that the
nitrogen is in its molecular form (N2) as in the combustion air.
Nitrogen contained in liquid or solid fuel is released in its atomic
form (N) and reacts at relatively low temperatures with oxygen to
form the pollutant. The heating values of the natural gases varied
from 0.0364 to 0.0391 GJ/m3 (976 to 1050 Btu/ft3).
The refinery gas used for Tests 149 through 152 was composed
of 76% methane, approximately 5% each of ethanes and propanes,
and 8.8% hydrogen. The heating value was comparable to natural
gas at 0.0391 GJ/m (1050 Btu/ft3).
For Tests 207-212 a mixture of natural and refinery gas was
fired. The proportion of methane was comparatively low at 48%.
The ethane of about 11% and 36% of the gas was hydrogen. The gas
had a heating value of 0.0310 GJ/m3 (831 Btu/ft3).
6.2 COAL AND OIL
Coal and oil fuel properties are discussed together in this
section since their characteristics influence emissions similarly.
The properties of the coals and oils tested in Phase II are
summarized in Table 6-2. The effects of the individual fuel
properties are discussed in the following subsections.
218
-------
Table 6-2. FUEL ANALYSIS SUMMARY
Coal and Oil Fuels
Test No.
102-103
107-108
111-112
116-121
126-130
131-133
134-139
156-159
159
160-164
165-168
169
170-175
176-179
186-189
195-206
Type of
Fuel
#2 Oil
#2 Oil
PS 300
(#5 Oil)
#6 Oil
#6 Oil
Coal
Coal
Coal
#6 Oil
#2 Oil
Coal
Coal
#6 Oil
#6 Oil
#6 Oil
#6 Oil
c
i *
o <* S of o *
85.94 12.72 0.89
85.94 12.72 0.89
86.46 10.31 1-07
85.95 10.21 0.22
No Fuel Sample Obtaii
62.50 5.21 12.51
68.30 4.70 11.01
69.89 4.67 8.10
85.44 11.35 0.002
87.14 12.71
63.42 4.84 9.85
62.50 5.12 12.51
86.60 10.94 0.31
86.17 11.55
85.10 11.10 3.12
86.68 12.16 0.65
•>.
C
0)
M IP
3 0
«H ^ »
i-l -P A
3 -H ra
w <* ys * < *
0.40 0.045 0.003
0.40 0.045 0.003
1.29 0.77 0.10
2.74 0.31 0.03
led For Analysis
1.15 0.83 10.5
1.16 1.46 9.71
1.36 1.50 14.48
2.80 0.33 0.078
0.31 0.013 <0.001
3.05 0.94 13.7
1.15 0.83 10.5
1.60 0.30 0.25
1.91 0.30 0.07
0.19 0.49 0.08
0.37 0.14 0.009
g) * «•
H i
3 C 3 4J
4j o 'O --H H
01 XI "H > Oj
•H H W «J rtj
O rt 0> ^
a <*> o os <» o o
0.071 31.1
0.071 31.1
8.50 15.1
_ - -
7.29 -
3.66 -
3.37 -
12.33 15.8
0.027 35.1
4.15 46.2
7.29 -
9.00 15.1
_
8.53 25.7
1.61 29.2
Higher Heating
Value
GJ/kg
0.0452
0.0452
0.0426
0.0423
0.0276
0.0286
0.0286
0.0432
0.0451
0.0276
0.0276
0.0434
0.0436
0.0446
0.0450
Btu/lb
19,440
19,440
18,333
18,213
11,863
12,293
12,300
18,580
19,390
11,873
11,863
18,660
18,773
19,227
19,365
to
O
O
CO
-------
6.3 VUEL SULFUR CONTENT
The results of the measurements of total sulfur oxides in
the flue gas are shown in Figure 6-1. 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 content 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
emissions.
Figure 6-2 shows that the ratio of sulfur trioxide (SO ) to
total sulfur oxides (SO ) is typically 1% to 2%, except when the
X
sulfur oxides concentration dropped below about 400 ng/J (500 ppm).
The steep rise below 400 ng/J is deemed to be due to the measurement
method itself, since the standard Shell-Emeryville method always
yields relatively high sulfur trioxide ratios when the total sulfur
oxides concentrations are below 400 ng/J. The instant of the color
change when titrating is difficult to determine precisely, and only
one drop of titrating solution can have a large effect on the calculated
concentration of SO when the absolute concentration of SO is low.
220
-------
I
§
2000
1800.
1600-
1400-
1200'
ki
O
800-
I
600-
400-
200-
0.
1800 .
1600 '
1400 .
1200 -
M
1000-
3
fc 800 -
»
•.
c 600 -
400 -
200.
0 .
2400
2000
1600
»
•
CM 1200
O
*
CO
0
&
•a
.^ son
- a
04
400
»
0
A
15
£
/
^/W
/3-I02
^°°
f
"TO ^
^\P)
1 I
Typical Type of Coal
/ 100% Conversion
/
/
/ /
/(D
/
176
&
A
, Typical Type of Oil
' 100% Conversion
\
Numerals within
are Test Numbers
Symbols
FUEL TYPE
/\ Coal
O <*
1
1
0.0 1.0 2.0 3.0 4.0 5.
FUEL SULFUR CONTENT, DRY, %
Figure 6-1.
Total sulfur oxides emissions at baseload for oil and
coal fired boilers.
6001-43
221
-------
FUEL AND BURNER TYPE
SULFUR TRIOXIDES TO TOTAL SULFUR OXIDES, PERCENTAGE
f— ' 1— * k_i i_i L_| t
u *. c* pop 10 E £ co !
=> ° o OfTT— 1° b b b b J
&
2 o
0
m
_u
fe
^JCD~
(T
or
^
i/rr >
e<
A
^^
^-C->'
I 1 Nature
(J) Oil #:
Q Oil #£
^ Oil #€
/yy Coal £
/^. Coal F
AA Coal U
^\ Coal C
/^.
— «. t
&>L
ffi
il Gas
>
preader
ulverizer
nderfed
yclone
^/v
^o-
4Sf
ffiSP
| 0 400 800 V*"/ 1200 1600 2000
ppm, dry @ 3% 0
h ' '
0
0
i i "i
400 800 1200
ng/J, for Natural Gas
l | i -
400 800 1200
ng/J, for Oil Fuel
0 400 80*0 12*00 1600
ng/J, for Coal Fuel
TOTAL SULFUR OXIDES CONCENTRATION
Figure IS-HT Ratio of sulfur trioxides to total sulfur oxides at
baseload as a function of total sulfur oxides measured.
222
6001-43
-------
There appears to be no strong effect of fuel type other than
its sulfur content. For example, No. 6 oil data are shown between
400 and 1200 ng/J (500 and 1500 ppm) and the SO,/SOV decreases
J x
with total sulfur oxides just as with the other fuels. For coal
the type of coal burner has no significant effect on the SO /SO
3 x
ratio in the exhaust gas.
6.4 FUEL ASH CONTENT
A coal and oil fuel property that correlates reasonably well
to solid particulate emissions is the fuel ash content. Figure 6-3
illustrates baseline solid particulates as a function of fuel ash
content for coal and Nos. 2, 5, and 6 fuel oil testing from Phase I and
II. Particulates were the lowest for the relatively ash-free No. 2 oils,
then increased as the oils became heavier and higher in ash content,
as with the No. 5 and 6 oils. Particulates were the highest with
coal.
Figure 6-3 also shows a line of equality corresponding to the
mass of solid particulate matter contained in the combustion produce
gases being equal to the mass of fuel ash input. The data do not lie
on this line. For oil fuels more solid particulate matter was emitted
than ash input. For coal the particulates were less than the ash
input. The coal data are easily explained; a significant amount of
the fuel ash drops out in the furnace bottom as dry ash or slag and
does not appear as part of the particulate measurement. For oils the
answer is slightly more complex. When the ash content of a fuel is
determined by an ultimate analysis all combustible materials including
sulfur are eliminated from the fuel sample prior to the determination
of ash content. However, the combustion process occurring in the
boiler is incomplete, resulting in carbon particles being present in
the combustion gases. In addition, a very small amount of sulfur may
combine with other materials to form solid sulfate compounds. The
223
-------
100
10,000
FUEL ASH CONTEKT
Figure 6-3. Effect of fuel ash content on baseline solid particulate emissions for coal and
oil fuels.
6001-43
-------
carbon and sulfate particles are combined with the fuel ash and
the sum can be larger than the amount of ash in the original oil.
A limited number of particulate tests were conducted on
natural gas fired boilers in Phase I. The particulate levels were
very low, nevertheless, they were significant. This suggests that
airborne dust particles may contribute to particulate emissions
since natural gas contains no ash.
The amount of volatile substances, such as unturned
sulfur or carbon, contained in the fly ash from six coal fuels and
one No. 6 oil fuel burned at baseline conditions was determined
during Phase I and Phase II. A quantity of fly ash was placed in a
crucible, weighed, baked in a furnace and then reweighed. The results
are tabulated below:
Fuel Content Ash Volatile
Test No.
19-6
20-6
32-4
134-2
156-2
165-1
Ref. 17
Loc.
No.
21
21
20
30
13
35
38
Test Load Carbon
Burner Type
SpStk
SpStk
Cyclone
SpStk
Pulv.
ChGrt
Steam Atom.
GJ/hr
42
66
338
87
422
110
47
%
76
76
77
68
70
63
85
Sulfur
%
0.
1.
2.
1.
1.
3.
2.
76
6
9
2
4
0
0
Ash
%
6.
6.
7.
9.
14
14
o.
8
9
8
7
05
Content
%
74
29
12
20
89
73
2
The ash from the cyclone burner was the lowest in volatile
content of the coal fuel tests. This low residual volatile content
is not surprising, because the combustion zone temperature of a cyclone
burner is very high and nearly all of the volatile substances in the
coal should be driven off.
The fly ash that had the highest volatile content
was from a pulverizer. This was expected, since other information had
indicated that the carbon content of the particulate from spreader
225
-------
(8)
stokers should have been the highest. However, the sulfur content
of the coal was 1.4%, and the presence of sulfur and sulfates in the
fly ash could account for the high volatile content. An analysis
of No. 6 oil fuel fly ash that is reported in Reference 17 found that
71% of the fly ash was sulfur and sulfates and only 8% was carbon.
The volatile content of the fly ash from a spreader stoker
found in Test 19-6 was relatively high, but the contents from Tests 20-6
and 134-2 using spreader stokers were not. The volatile content of
the chain grate fly ash was high, which is consistent with the
contention that the larger the size of the coal as fired, the larger
is the carbon content of the fly ash. It is likely that sulfates also
are a significant factor, since the sulfur content of the coal was a
high 3.0%. The grouping of variations in volatile content is deemed
to be due to the characteristics of the combustion process as well as
the coal feeding method, because there appears to be no fuel property,
such as sulfur content, that would cause certain test results to
belong to one group or another. This contention can be verified by
referring to Table 6-1 of this report for the properties of the coal
burned during Phase II and to Table 6-1 of Reference 1 for the Phase
I coal properties. Phase I tests were numbers 19-6, 20-6, and 32-4.
6.5 API GRAVITY
An oil fuel property which correlates with nitrogen oxide
and particulate emissipns is the API gravity measured at 20°C.
This is not a unique correlation, since fuel nitrogen and ash
contents decrease in going from heavy to light oils or as the
gravity increases.
The nitrogen oxides and solid particulates are shown as a
function of API gravity in Figure 6-4. The measured NO fell into
*:wo groups: (1) where the fuel oil gravity matched the API gravity
226
-------
450
400 „
i
IA
n)
51
o>
g
350
300
250
200
Q
H
150
50
10
(ill
\/
Nitrogen
Oxides
Particulates
Numbers within symbols
are test numbers
90
80
70
60
50
40
30
20
10
i
D
U
i-
8
15 20 25 30 35
API GRAVITY OF FUEL OIL AT 293 K
40
Figure 6-4.
Effect of API gravity on baseload nitrogen oxides
and particulates emissions.
6001-43
227
-------
specification for diesel or No. 2 oil the nitrogen oxides was
between 56 and 110 ng/J (100 and 200 ppm) , and (2) where the fuel
oil gravity matched No. 5 or 6 oil and the nitrogen oxides was
between 95 and 350 ng/J (170 and 620 ppm). The fuel burned for Tests
63, 68, and 70 was designated as PS 300 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.
228
-------
SECTION 7.0
BOILER DESIGN CHARACTERISTICS
7.1 FURNACE AND BURNER 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 in Phase II are listed in Table 7-1.
7.2 COST OF MODIFICATION
Four methods of combustion modification that required modifi-
cation of the boiler were investigated during Phase II. These were:
Staged combustion air
Variable combustion air temperature
Flue gas recirculation
An 18.5 GJ/hr (17.5xl03 Ib/hr) D-type boiler at Location 19 was
modified to add a staged air and flue gas recirculation capability.
The installation is discussed and pictured in Subsection 5.2.2. The
cost of modifying this boiler was estimated at about $5,000. The
current cost of a new boiler of this type is about $60,000.
The cost of a similar modification on other modern D-type
boilers could run as high as $7,500 if the existing burner registers
could not be used. At Location 19 the windbox depth was increased
and a second set of registers to control the flue gas being recirculated
was installed within the extension.
229
-------
Table 7-1. TEST BOILER DESIGN CHARACTERISTICS
Test Ho.
Fron-Thru
101 only
102-103
104-106
107-108
109-110
111-112
113-115
116-121
122-125
126-13O
131-133
134-139
140-142
143-148
149-152
153-155
156-159
160-164
165-168
169 only
Loc.
No.
1
1
1
1
27
27
29
29
28
28
31
30
32
32
33
34
13
36
35
31
BOILCR
No.
2
2
1
1
1
1
5
5
1
1
7
8
4
1
32
2
2
6
6
6
Mfg.
BtW
BtW
BtW
BtW
BtW
BtW
Rile)
Rile;
Erie
Erie
Erie
Erie
ClBrl
Erie
BtW
(Con
Cyc)
CE
BSW
BtW
Erie
Erie
Date
1964
1964
1964
1964
1967
1967
1972
1972
1957
1957
1963
1963
1969
1963
1965
1972
1967
1971
1960
1963
Cap
GJ/hr
(10 Ib/hr)
31
(29)
31
(29)
31
(29)
31
(29)
106
(100)
106
(100)
158
(150)
158
(150)
74
(70)
74
(70)
274
(260)
132
(125)
137
(130)
127
(120)
580
(550)
264
(250)
528
(500)
211
(200)
227
(215)
274
(260)
FURNACE
rype
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
WT
Hall
Const
RT
RT
RT
RT
WF
WF
TT
TT
RT
RT
TT
RT
TT
TT
TT
TT
TT
TT
TT
TT
Size
L-H-H
Meter
(feet)
4.0-1.8-2.1
(13-6-7)
4.0-1.8-2.1
(13-6-7)
4.0-1.8-2.1
(13-6-7)
4.0-1.8-2.1
(13-6-7)
6.4-1.8-2.7
(21-6-9)
6.4-1.8-2.7
(21-6-9)
8.5-3.4-5.8
(28-11-19)
8.5-3.4-5.8
(28-11-19)
2.7-4.3-6.1
(9-14-20)
2.7-4.3-6.1
(9-14-20)
6.4-6.4-12
(21-21-39)
5.2-4.3-10
(17-14-33)
1O. 4-2. 1-5.'
(34-7-18)
5.2-4.0-8.6
(17-13-29)
11-10-8
(35-32-27)
14-3.7-2.4
(46-12-8)
26-6.1-6.1
(86-20-20)
11-2.1-3.0
(35-7-10)
4.3-8.2-13
(14-27-43)
6.4-6.4-12
(21-21-39)
Area
7ft2,
49.7
(535)
49.7
(535)
49.7
(535)
49.7
(535)
68.6
(738)
68.6
(738)
167
(1795)
167
(1795)
104
(1115)
104
(1115)
460
(4950)
193
(2080)
181
(1949)
200
(2150)
488
(5250)
158
(1700)
684
(7360)
117
(1260)
398
(4282)
460
4950)
Vol.
m3
21.4
(755)
21.4
(755)
21.4
(755)
21.4
(755)
32.1
(1134)
32.1
(1134)
154
(5450)
154
(5450)
76.5
(2700)
76.5
(2700)
552
(19500
190
(6700)
123
(4340)
177
(6250)
856
(3024Q
HEAT
ABSORPTION
Area
GJ/hr
n2
(S
.631
:.0556)
.631
(.O556)
.631
[.0556)
.631
.0556)
1.53
1.135)
1.53
1.135)
.946
.0833)
.946
.0833)
.693
.0610)
.693
[.0610)
.847
.0746)
.931
1.0820)
1.05
[.09261
.902
(.0794)
1.19
.105)
1
123 1.67
(43S5)|t.l47)
988 1 .772
( 3490U|( . 068O)
j
69.4 j 2.15
( 2450) k. 189)
460 1 .568
(16254TI.0500)
552 | .847
(19500^.0746)
Vol.
GJ/hr
™"in^
eft
1.62
(.0435)
1.62
(.0435)
1.62
(.0435)
1.62
(.J435)
3.30
(.0885)
3.30
(.0885)
1.02
(.0275)
1.02
(.0275)
.965
(.0259)
.965
(.0259)
.70S
(.0190)
.946
(.0254)
1.55
(.0415)
1.01
(.0272)
.678
(.0182)
2.14
(.0575)
.533
(.0143)
3.65
(.0980)
.496
(.0133)
.708
(.0190)
BURNER
Test
Fuel
NG
12
NG
*2
NG
PS
300
NG
16
NG
•6
Coal
Coal
NG
NG
Ref
Gas
NG
Coal
Oil
•2
Coal
Coal
Type
Ring
Stean
Ring
Steal
Ring
Steal
Ring
Steal
Ring
Steai
Pulv.
SpStic
Ring
Ring
Spud
Spud
Pulv.
Stean
Steav
ChGrt
Pulv.
Mfg.
BtW
BtW
BtW
BtW
Coen
Coen
Coen
Coen
Todd
Todd
CE
Oet
Stk
Pea-
body
Erie
B&w
CE
BtW
Pea-
body
BtW
CE
CE
No.
of
1
1
1
1
1
1
2
2
3
3
4
4
1
4
1
1
6
3
1
4
4
BURNER SPACING
Horis.
Dilt.
(in.)
_
_
_
_
^
_
_
_
_
.
_
_
-
.
91.4
(36)
91.4
(36)
198
(78)
150
59
_
183
(72)
_
_
_
183
(72)
_
_
.
_
198
(78)
Vert.
Dist.
cm
(In.)
_
_
.
.
_
.
_
_
_
_
.
168
(66)
168
(66)
.
_
193
(76)
_
_
_
122
(48)
_
..
_
168
(66)
_
_
_
_
193
(76)
Brn.
Load
CJ/hr
Brn.
HBH/Brn.
30.6
(29.0)
30.6
(29.0)
30.6
(29.0)
30.6
(29.0)
106
(100)
106
(100)
79.1
(75.0)
79.1
(75.0)
24.6
(23.3)
24.6
(23.3)
96.1
(93.0)
44.3
(42.0)
190
(180)
45
(43.0)
101
(96.0)
264
(250)
87.9
(83.3)
253
(240)
56
(54)
(93.0)
Fuel
Temp.
« Brn.
•C
(V)
_
Aib
Anb
_
Anto
Anb
_
71.1
(160)
118
(245)
_
93.9
(201)
_
»
_
_
.
_
_
Aato
Amb
_
_
65.6
ISO
„
_
Aib
Anb
-
Prim.
Air
Teep
•c
CF)
Amb
Amb
Anb
tatt
Amb
Anb
Anb
Anb
Anb
Anb
Anb
193
(380)
202
(395)
168
(335)
171
(340)
260
(500)
93.3
(200)
199
(390)
*
443
(830)
110
(230)
216
(420)
Arab
Anb
107
(225)
26O
1500)
Sta<*
Tecip>
•c
(*F)
268
(515)
288
(550)
293
(560)
243
(470)
171
(340)
168
(334)
143
(290)
154
(310)
127
(260)
135
(275)
166
(330)
160
(320)
149
(300)
160
(320)
171
(340)
377
(71O)
216
(420)
199
(390)
143
(290) .
166
(330)
U>
O
•O»it haa air preheat, but it waa not po«aible to
coafcuaUon air temperature.
6001-43
-------
Table 7-1. Continued
Test No.
rr^B»-'n.ru
170-175
176-179
180-105
186-189
190-194
195-199
200-203
207-212
Loc.
Mo.
20
37
38
38
19
19
19
39
HO1LLK
Ho.
4
2
2
2
1
1
1
BIOS
Hfy.
CE
HicXra
CE
CE
Koole
Heeler
feeler
BCN
Date
1966
1955
1951
1951
1970
1970
1970
1974
Cap
GJ/hr
HEAT
ABSORPTION
Area
c.l/lir
(S-)
1.81
(.159)
.OlS
(.0806)
.710
(.0625)
.710
(.0625)
.6075
(.0535)
.6075
(.0535)
.6075
(.0535)
.263
(.0232)
vol.
C.j/hr.
(£)
2.91
(.0781)
1.04
(.0278)
.846
(.0227)
.846
(.0227)
1.66
(.0446)
1.66
(.0446)
1.66
(.0446)
2.93
(.0787)
UUKNUK
Test
Fuel
16
16
NG
16
NG
16
16
Ref.
Gas
Type
Steam
steam
Ring
Steam
Ring
Steam
Air
Spud
Mfg.
Coen
Pea-
body
Pea-
body
Pea-
body
Faber
Faber
Fabar
BtW
No.
of
1
2
1
1
1
1
1
1
BUKNKR St'ACINo
lloriz.
Dist.
cm
(in.)
.
-
112
(44)
-
-
-
-
-
-
-
-
-
-
-
VOL- I.
Dist.
on
(in.)
.
-
-
.
-
-
-
-
-
-
-
-
-
-
-
Brn.
Load
GJ/hr
Brn.
MDll/Di n .
121
(115)
21.1
(20.0)
47.5
(45.0)
47.5
(45.0)
18.5
(17.5)
18.5
(17.5)
18.5
(17.5)
211
(200)
Fuel
TVmi>.
9 Brn
"C
CF)
65.6
(150)
9O.O
(194)
-
-
113
(235)
-
-
93
(200)
93
(200)
-
Prinv
Air
Tempi
•c
CF)
Amb
Arab
141
(285)
177
(350)
160
(320)
Anb
Anb
Amb
Anb
Anb
Amb
Anb
Anb
itaJ<.
Tcwpv
•c
CF)
332
(630)
218
(425)
227
(440)
204
(400 )
254
(490)
254
(490)
254
(490)
166
(330)
NJ
U)
•Unit has air preheat, but it was not poMibl* to Baasura combustion air tufxratur*.
6001-43
-------
During discussions with a manufacturer of boilers in the
320 GJ/hr (300x10 Ib/hr) and one million dollar size and cost range
it was estimated that a staged air installation in general would add
two to four percent to the cost of the boiler. For A-type boilers the
additional cost would be about two percent and for D-type boilers about
three percent. If another booster air fan were required the cost would
be increased by about an additional one percent.
232
-------
SECTION 8.0
FUTURE RESEARCH
Staged combustion air, variable combustion air temperature and flue
gas recirculation were very effective in reducing the emissions of nitrogen
oxides. But before the advantages of incorporating these forms of
combustion modification into current boiler designs can be determined,
a large body of parametric data is needed.
In the past it was not practicable to gather good design data
on these combustion modification methods. Laboratory research suffered
from scaling inconsistencies. Research with full-size boilers in the
field was limited, because it was difficult to reproduce prior conditions
exactly, and process needs frequently interrupted testing. But most
serious was the complete lack of flexibility. For example, industrial
boilers with staged combustion air ports are in service, but the
port location is fixed and the effect of port number and location
cannot be investigated.
Two boilers now exist at Locations 19 and 38, where controlled
research over a wide range of combustion parameters can be done readily.
It is recommended that a field research program be initiated to
investigate one or all of the staged combustion air, variable combustion
air temperature and flue gas recirculation combustion modification
methods.
The measurement of emissions and the effect on emissions of
combustion modifications should be extended to industrial combustion
equipment. Industrial combustion equipment includes waste-product-
fueled boilers, kilns, glass melting furnaces, steel furnaces,
incinerators, etc.
Industrial combustion devices contribute a large fraction of
the total air pollution from stationary sources. Recent studies have
shown as much as 40% of the stationary source nitrogen oxides emissions
233
-------
originate from industrial devices. A similar figure was obtained for
oxides of sulfur, while particulate emissions from stationary industrial
sources account for more than 80% of the total. ' ' Combustion
modifications for industrial boilers have been demonstrated in this
report which can reduce emissions of nitrogen oxides, carbon monoxide,
and hydrocarbons while improving boiler efficiency. Application of
these modifications to industrial combustion devices, if successful,
could have a profound impact on air quality and energy conservation.
In order to apply these modifications, baseline emissions and efficiency
from industrial combustion devices must be determined. Then, applica-
tion of combustion modification techniques under controlled conditions
can be performed to determine efficiency and emission trends.
234
-------
SECTION 9.0
REFERENCES
1. Bartz, D. R. , et al., "Control of Oxides of Nitrogen From Stationary
Sources in the South Coast Air Basin," KVB Engineering Report No.
5800-179, State of California Air resources Board, September 1974.
2. Barrett, R. E., et al., "Field Investigation of Emissions From
Combustion Equipment For Space Heating," Battelle-Columbus
Laboratories, EPA-R2-73-084a, June 1973.
3. Ehrenfield, J. R., et al., "Systematic Study of Air Pollution From
Intermediate Size Fossil-Fuel Combustion Equipment," Walden Research
Corp., NTIS No. PB 207110, July 1971.
4. Cato, G. A., et al., "Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions From Industrial
Boilers - Phase I," Report No. EPA-650/2-74-078-a, NTIS No. PB 238
920/AS, October 1974.
5. Cato, G. A., et al., "Field Testing: Toxic Metals and Organic
Emissions From Industrial Boilers," KVB, Inc., Report No. EPA-650/2-
74-078-c, To Be Published Spring 1976.
6. Smith, W. B., et al., "Particulate Sizing Techniques for Control
Device Evaluation," EPA Report No. EPA-650/2-74-102a, August 1975.
7. Burchard, J. K., "The Significance of Particulate Emissions," J. Air
Pollution Control Assoc., V 24, No. 12, December 1974.
8. Dorsey, J. A. and J. O. Burckle, "Particulate Emissions and Process,"
Chem. Engr. Progress, 67, 92, 1971.
9. Cuffe, S. T. and R. W. Gerstle, "Emissions From Coal-Fired Power
Plants: A Comprehensive Summary," U. S. Dept. of Health, Education
and Welfare, Public Health Serivce Publication No. 999-AP-35, 1967.
10. "Power Test Codes for Steam Generating Units," The American Society
of Mechanical Engineers, PTC 4.1-1964, December 1964.
11. Turner, D. W. and C. W. Siegnumd, "Staged Combustion and Flue Gas
Recycle: Potential for Minimizing NOX from Fuel Oil Combustion,"
Presented at the American Flame Research Committee Flame Days,
Chicago, IL, September 1972.
235
-------
12. Muzio, L. J., et al., "Package Boiler Flame Modifications for
Reducing Nitric Oxide Emissions, Phase II of III," EPA Report
R2-73-2925, API Publication 4208, 1974.
13. Bell, A. W., et al., "Combustion Control for Elimination of Nitric
Oxide Emissions from Fossil Fuel Power Plants," Proceeding of the
13th International Symposium on Combustion, University of Utah,
August 23-29, 1970.
14. 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.
15. Kato, K., et al., "Formation and Control of Fuel Nitric Oxide,"
Published at a conference of thermo-engineering held by the
Japanese Society of Mechanical Engineers, November 1973.
16. Blakeslee, C. L. , and H. J. Burbach, "Controlling NOX Emissions
from Steam Generators," Presented at the Air Pollution Control
Association's 65th Annual Meeting.
17. Laurendeau, R., et al./ "The Reduction of Particulate Emission
from Industrial Boilers by Combustion Optimization," ASME Paper
75-WA/APC-3, 1975 Winter Annual Meeting, November 1975.
236
-------
SECTION 10.0
GLOSSARY OF TERMS
SJ
Ul
ABIS
Air
Air Reg
Amb
API
Atm
Atom Press
Base
BOOB
Bt X COS
Brn
BrTune
Btu
Bumh
BtN
c
c
c2
CE
C2H6
ChGrt
CI
CL
Cl Brk
CO
Coen
Comb Cyc
Con Part
Coppus
oor.
Cup
cyclone or eye
D
Damper
•C
•r
Det Stk
EPA
Equivalence Ratio,
All burners in service
Usually referring to air atomized fuel oil
burner
Air Registers
Ambient temperature
American Petroleum Institute
Atomization
Burner atomizing pressure adjustment
Baseline
Burnur(s) out of servlca
Burner number X out of service
Burner
Boiler tune-up
British Thrnul Unit
Burnham/Golden Scotch
The Bibcock and Hilcox Company
Coal
centii one-hundredth
Multiple carbon atom hydrocarbons
Combustion Oiglneering, Incorporated
Methane
Ethane
Chain Grate
Cast iron furnace walls
Unheated sample line (cold line)
Cleaver-Brooks Division
Carbon sonoxide
Carbon dioxide
The Coen Company
Combined cycle
Condensible particulates
Coppus Engineering Corporation
Data corrected to standard conditions
Rotary cup fuel oil atomizer
Cyclone furnace coal conbustor
Diameter
As test typei air danper adjustment
Degrees Centigrade or Celsius
Degrees Fahrenheit
Detroit Stoker Company
Environmental Protection Agency
# The actual fuel to air ratio divided by the
otoichiometric fuel to air ratio.
« > li Fuel rich
• < li Air rich
Faber
FD
FGR
ft
FT
FW
9
G
H
HC
High Air
Mi load
HL
hr(s)
Hz
IBM
ID
in
in. Hg
Ind. Cost).
IR
iwj
J
K
k
Keeler
Kewan
L
Low Air
Ibs or t
H
n
MB o* MBtu
NBH or MBtu/hr
Meal
MCH or Mcal/hr
Mfg
•in
MR
V
Vm or u
Faber Engineering Company Incorporated
Forced draft
Flue gas recirculation
foot
Firetube furnace
Foster Wheeler Corporation
Grams
Giga, one billion
Height
Unburned hydrocarbons measured as methane
High excess air
High load
Heated sanple line (hot line)
Hour(a)
Hertz; cycles per second
International Boiler Works Company
Inside diameter
Inches
Pressure in inches of mercury, usually gage
Industrial Combustion Incorporated
Infrared
Pressure in inches of water column gage
Joule
Kelvin temperature scale
Kilo; one thousand
E. Keeler Company
Kewanee Boiler Corporation
Length
Low excess air
Pounds
Mega; one million
As prefix: milli; one-thousandth
Meter
One million British thermal units
One million British thermal units per hour
One million calories
One million calories per hour
Manufacturer
Minutes
Mixture ratio in terms of the air flow rate
divided by the fuel flow rate
Micro; one-millionth
«ierom»ter or "micron" (1O~* meter)
-------
03
"2
Nebr
HG or G
NO
No.
No. Am.
NO I
Nrml Air
NSF-Oil
OD
OTA
o/s
p
Pa
Pea body
t
ppm
PS 300
psi
psia
psig
Pulv.
R
Ray
Ref Gas or RG
Riley
Ring
rms
RT
SCA
S
Sec
Sid. Put.
Molecular nitrogen content of fuel, percent
by weight
Nitrogen gas
Nebraska Boiler Company
Natural Gas Fuel
Nitric oxide
Number
Nitrogen dioxide
North American Company, Cleveland, Ohio
Total nitrogen oxidea (NO + NO )
Normal excess air
Navy standard fuel-oil (similar to No. 5 oil)
Oxygen gas
Outside diameter
Overfire air
Of f-stoichiometric
Preheated combustion air
Pascals, newtons per square meter
Pcabody Engineering Company
See Equivalence Ratio
Parts of constituent per million parts of
total volume
Pacific standard fuel-oil No. 300
(similar to No. 5 oil)
Pressure in pounds per square inch
Pressure absolute in pounds per square inch
Pressure gauge in pounds per square inch
Pulverized coal burning equipment
Refractory
Ray Burner Company
Refinery gas
Riley Stoker Corporation
Natural gas ring
Root mean square
Water wall tubes spaced such that refractory
tile is exposed to flame
Staged combustion air
Sulfur content of fuel, percent by weight
Seconds
Seconds
Solid Particulates
Sngl Cyc
"2
S03
SOX
Sprd. or SpStk
Spud
Steam
Steam Injac.
Supr*
t
Temp.
TIW
Todd
TP
Trane
TT
U or UFS
uncor.
Union
V
viscosity
Vol
VPH
W
w
Wall Const.
WF
Kinkier
VTT
wtgh
Single cycle
Sulfur dioxide
Sulfur trioxide
Total sulfur oxides (SO + SO )
Spreader stoker coal burning aquipaant
Natural gas gun
As Burner Type: steam atomized oil burners
Steam injection
Superior Combustion Industrial
Metric ton (103 kilograms)
Temperature
Titusville Iron Works
Todd-CEA Incorporated
Toxic Particulate
The Trane Company
Furnace walls wh^-re the watertubes arc tangent
Underfed stoker coal burning equipment
Data presented as measured and not corrected
to a standard condition
Union Iron Works
Voltage in volts
As test type: fuel oil viscosity variation
via temperature change
Volume
Variable combustion air preheat
As unit of power: Watt
As a dimension: Width
Furnace wall construction
Furnace wall constructed with welded fin design
Winklei- burner manufacturer
Matertube furnace
Westinghouse