United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-83-045 Nov. 1983
Project Summary
Evaluation of Combustion
Variable Effects on NOX
Emissions from Mineral Kilns
R.J. Tidona, W.A. Carter, H.J. Buening, S.S. Cherry, and M.N. Mansour
Results of tests performed on a lime
kiln, precalciner cement kiln, and
conventional wet process cement kiln
are presented and discussed. Where
applicable, the effectiveness of excess
air variations on pollutant emissions is
quantified and compared to previous
results. Mass balances were also
calculated for the two cement kilns.
A subscale cement kiln simulator was
designed, fabricated and operated to
determine the effect of burner operating
variables on near-flame NOX production.
The effects of combustion air preheat,
carrier air dilution and fuel injection
velocity were the primary variables
assessed for both natural gas and coal.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction and Summary
The activities reported here include
tests on a rotary lime kiln (Location 6),
precalciner cement kiln (Location 8), and
conventional wet process cement kiln
(Location 9). Fuel oil was used in the lime
kiln, and coal in both cement kilns.
Variation in excess air was the N0«
control implemented on the lime kiln and
wet kiln. Only as-found tests were
performed on the precalciner kiln because
it had just recently been started upand its
operation may not have been fully
optimized.
For the lime kiln (Location 6), a
reduction in excess air reduced NOX
emissions by 23 percent. Further reductions
in excess air produced poor quality clink-
er. A new oil tip (with fewer orifices)
caused oil to impinge on the kiln wall, an
unacceptable operating condition.
As-found testing on the precalciner
cement kiln (Location 8) resulted in
emissions higher than the conventional
wet process kiln tested at Location 9. This
result may be due to kiln operation not
being fully optimized at the time of the
test program. Mass balances were
performed for sulfur, sodium, and potassium
by a contractor retained by the plant
operator. Closure on these balances was
good.
Testing at the conventional wet process
cement kiln (Location 9) encompassed
as-found, baseline, and variations in
oxygen level. Linear regression analysis
of the N0« data predicted a 38 percent
reduction in NOX when the oxygen level
was lowered from baseline conditions.
However, a simultaneous increase in
gaseous S02 of 47 percent was also
predicted. Mass balances were made on
seven kiln constituents: the largest single
difference was 29.5 percent. The difference
for all seven constituents was 3.8
percent.
A subscale cement kiln simulator was
constructed and tested to determine the
effect of burner parameters on near-
flame NOX levels for both natural gas and
coal fuels. It was determined that
combustion air preheat, fuel injection
velocity, and oxygen content of the
primary combustion air stream have first-
order effects on NO, levels. This subscale
program will be used to select advanced
combustion modifications for implementa-
tion at the pilot scale.
-------
Comparison of Previous and
Present Cement Kiln Programs
KVB, during a previous EPA contract
and a contract from the California Air
Resources Board, performed emission
measurements on two conventional
kilns. Table 1 summarizes the results
obtained on all four kilns tested.
Of note for Location 3 are the higher
emissions with natural gas fuel as
opposed to either oil or a combination of
coke and natural gas. A similar situation
has been measured in glass furnaces
where NO emissions on natural gas are
higher than oil fuel despite the fuel-
bound nitrogen content of oil fuel. This
difference exists, but it might be due to
higher radiant flame cooling for an oil or
coal flame because of higher emissivity.
Of equal note is the very low NO value
measured during the present EPA contract
for the coal-fired wet process kiln. A
review of the testing procedures indicated
that all measurements were made
properly. There is no explanation why this
particular kiln is a low NO emission
source.
Emissions Test Instrumentation
All emission measurement instrumenta-
tion for the full-scale testing was carried
in an 8 x 42 ft (2.4 x 12.8 m) mobile
laboratory trailer. The gaseous species
measurements were made with analyzers
located in the trailer. The emission
measurement instrumentation used is
listed in Table 2.
The instrumentation used during the
subscale cement kiln testing is listed in
Table 3.
Table 3.
Laboratory Instrumentation Employed
Emission
Species
Oxygen
Carbon Dioxide
Carbon Monoxide
Nitrogen Oxides
Sulfur Dioxide
Manufacturer
Teledyne
Horiba
Horiba
Thermo Electron
Du Pont
Measurement
Method
Fuel Cell
NDIR
NDIR
Chemiluminescent
UV Spectrometer
Model
No.
720P4
PIR200(
PIR200C
10A
411
Results
Location 6 Lime Kiln
The significant results on the lime kiln
are shown in Figure 1 and Table 4. As
noted, a reduction in oxygen from 4.4 to
2.8 percent (test 6/1-1 vs. test 6/1-2)
produced a 23 percent decrease in NO
emissions, while a 85.6 percent decrease
in NO was measured when the oxygen
was reduced to 1.5 percent (test 6/1 -3).
However, at 1.5 percent oxygen, lime
clinker quality was poor.
Test 6/2-2 and 6/2-3 are of uncertain
accuracy; they were performed during a
±50 percent change in fuel flow rate. The
possibility exists that insufficient time
was allowed after the fuel flow rate
change to permit the kiln to stabilize
thermally. The lower kiln front-end
temperature for test 6/2-2 with respect
to 6/2-1 is most likely a consequence of
the reduced fuel flow rate. However, the
still lower temperature for test 6/2-3
probably reflects an unstabilized kiln
operating condition.
The spread in baseline results (tests
6/1-1, 6/2-1, 6/3-1, and 6/3-4) is not
considered unusual for an industrial
combustion device with direct contact
between the combustion products and
Table 1. Comparison of Cement Kiln NO Emissions
NO
Location
Process
Fuel
Condition
ppm, dry
@ 3% O2 kg/Mg Clinker (Ib/ton)
3 Dry Coke + Nat. Gas
3 Dry Nat. Gas
3 Dry Oil
9 (2144 f Wet Nat. Gas
8 Precalciner Coal
9 (2645J° Wet Coal
^Location 9, EPA Contract 68-02-2144.
"'Location 9. EPA Contract 68-02-2645.
Baseline
Baseline
Baseline
Baseline
As-found
Baseline
1014
1460
640
2474
1264
183
4.0
7.5
3.3
9.1
3.7
0.88
(8.01
(14.9)
(6.6)
(18.2)
(7.5)
(1.8)
Table 2. Emission Measurement Instrumentation
Species
Hydrocarbon
Carbon Monoxide
Oxygen
Carbon Dioxide
Nitrogen Oxides
Particulates
Sulfur Dioxide
Manufacturer
Beckman Instruments
Beckman Instruments
Teledyne
Beckman Instruments
Thermo Electron
Joy Manufacturing
Du Pont Instruments
Measurement Method
Flame lonization
IR Spectrometer
Polarographic
IR Spectrometer
Chemiluminescent
EPA Method 5 Train
UV Spectrometer
Model
No
402
865
326A
864
10A
EPA
400
process material. Variations in the
process material composition and process
rate will require adjustment of the firing
rate and combustion air flow in order to
produce an acceptable product. This lime
kiln's firing rate was manually controlled
by the operators to compensate for
variations in process material composition
and process rate. Therefore, it was not
always possible to reproduce the exact
firing conditions obtained during the
baseline tests.
The total kiln ambient air flow is in two
parts (neither of which could be measured):
a primary supply coaxial with the oil gun
and a secondary circuit distributed
around the kiln interior. Each circuit is
supplied by its own fan. Test 6/3-2 was
conducted by increasing the secondary
flow, decreasing the primary flow, and
maintaining the overall Oz at 3.9 percent.
With respect to the nearest baseline in
time (test 6/3-1), this reduced the NO
emissions by 18.6 percent. Reducing the
total air flow by closing down on both the
primary and secondary air dampers (test
6/3-3) reduced the NO emissions by 31.9
percent; however, clinker quality was
slightly degraded, but still usable.
The original oil tip had seven holes,
each 5.79 mm (0.228 in.) in diameter,
located on a 34.9 mm (1.375 in.)diameter
circle and inclined at 35° from the oil gun
axis. A new tip was designed, fabricated,
and installed with the same oil flow area
but with only four holes. The new tip was
designed to delay mixing between the oil
spray and combustion air and thereby
reduce the NO emissions.
However, the initial testing of the four-
hole tip resulted in the oil spray's
impinging on the kiln insulation in four
locations. In addition, the NO emissions
increased by 22.8 percent. Testing with
the new oil tip had to be suspended
because of the concern for potential
insulation degradation. The tests that
were conducted had to be made with a
lower fuel oil input.
In summary, the tests performed on the
rotary lime kiln showed that lower excess
air had a practical limit in reducing NO;
i.e., to the point where lime quality was
affected. This limitation corresponded to
an NO reduction on the order of 23
percent.
-------
200
175
150
125
.- too
CO
I
* 75
50
25
(6/2-2)
(6/1-4)
O
(6/2-3)
(6/3-1)
(6/3-4)
Process Rate = 49 1 mV/v
Test No.
Baseline
(J Poor Lime Quality
O Modified Operating Conditions
figure 1.
123456
Oxygen, % dry
Location 6 lime kiln-effect of Oz on NO.
8
10
11
Location 8 Precalciner Kiln
Only as-found tests were performed
since it had recently come on-line and its
operation may not have been fully
optimized. Major system components are
shown in Figure 2. It is claimed that the
cyclone stages and flash furnace can
complete up to 95 percent of the
calcination prior to the feed's entering the
rotary kiln.
The test results are given in Table 5.
The average value of N0« measured in the
stack* was 972 ng/J (2.26 Ib NO2/106
Btu) which was higher than anticipated.
(The high value of stack oxygen is due to
the ambient temperature quench air
introduced upstream of the alkali bypass
baghouse.)
Samples were obtained of the coal, raw
feed, clinker, alkali bypass baghouse
catch, and crusher/dryer baghouse
catch, for the purpose of performing mass
balances for sulfur, sodium, and potassium.
The sample analyses, together with the
process weights, were analyzed by a
contractor retained by the plant owner.
The results of the mass balances are
shown in Table 6. Of note is the high
*A single stack is used for the discharge streams
from the alkali bypass baghouse and crusher/
dryer baghouse. Stack measurements were
made downstream of both streams
Table 4. Summary of Gaseous Emissions from a Lime Kiln - Location 6
Test No.
6/1-1
6/1-2
6/1-3
6/1-4
6/2-1
6/2-2
6/2-3
6/3-1
6/3-2
6/3-3
6/3-4
6/4-1
6/4-2
6/4-3
6/4-4
6/4-5
6/4-6
6/4-1 A
6/4-7
6/5-1
6/6-1
6/6-2
6/6-3
6/6-4
6/6-5
6/6-6
6/6-7
Date
1979
10-25
10-25
10-25
10-25
10-25
10-25
10-25
10-25
10-25
10-25
10-25
10-26
10-26
10-26
10-26
10-26
10-26
10-26
10-26
11-12
11-12
11-12
11-12
11-12
11-12
11-12
11-12
Process
Rate
m3/ha
49.1
49. J
48.6
49 1
49.1
49.1
49.1
49.1
49.1
49.1
49.1
486
48.6
48.6
48.6
48.6
49.1
49.1
49.1
34.1
34.1
34.1
34.1
34.1
34.1
34.1
34.1
Fuel
Flow
m3/h*
1.63
1.44
1.48
1.21
1.21
0.57
1.82
1.21
1.17
1.21
1.19
1.21
1.17
1.21
1.21
1.21
1.17
1.21
1.21
1.32
1.32
1.14
1.14
1.14
0.40
1.14
1.14
Oz
%
4.4
2.8
1.5
6.5
4.7
11.0
3.4
4.0
3.9
3.2
38
4.5
6.0
3.0
1.3
5.0
5.8
4.2
1.6
5.8
5.4
3.5
4.2
4.8
6.7
6.1
4.2
COz
%
19.9
20.5
20.5
18.3
18.8
9.8
20.0
19.2
18.6
20.5
18.9
20.5
18.0
21.0
22.0
20.5
18.8
19.6
20.5
20.5
20.5
20.5
20.5
20.5
16.0
19.2
20.2
NO
ppm
dry at
3% Oz
90
69
13
160
119
192
132
113
92
77
104
158
246
90
60
105
115
154
108
127
156
no
119
138
162
147
124
NO
ppm
wet
-
-
-
-
-
-
-
_
-
-
-
-
-
-
-
-
33
22
18
15
16
20
12
CO
ppm
dry at
3%Oz
33
25
17
20
20
18
15
13
28
33
33
49
54
45
2.165
81
62
54
71
41
43
43
37
34
43
40
36
SOz
ppm
wet
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
50
40
57
38
110
105
95
44
Kiln Front
End Temp.
K
1,402
1,478
1,478
1,478
1,478
1,267
1,200
1,436
1,339
1,353
1,367
1,422
1.300
1,381
1,464
1,436
1.378
1,356
1.467
1.378
1.467
>1,478
1,450
1,461
1,456
1,417
1,444
Comments
Baseline
Minimum primary air
Low 02
High O2
Baseline
Low fire
High fire
Baseline
Increased secondary air flow
Low O2
Baseline repeat
Baseline without secondary air
High Oz without secondary air
Medium Oz without secondary air
Low Oz without secondary air
Baseline with minimum secondary air
High secondary air
Baseline without secondary air
No secondary air - no odor gas
Baseline old oil tip
Baseline new oil tip
No secondary air - new oil tip
No secondary air - no odor gas
No secondary air, high Oz - no odor gas
Minimum secondary air - with odor gas
Min. sec. air, low prim, air -with odor gas
Minimum secondary air - no odor gas
"gal./min - m3/h • 4.40
-------
degree of closure on all three mass
balances.
Location 9 Wet Cement Kiln
NO* results obtained on this kiln are
shown in Figure 3 as a function of oxygen.
Also shown is a linear regression
between NO* and 02 which can explain
39.9 percent of the NOX scatter. Figure 4
presents similar information on the
variation in SOz with oxygen for which
the linear regression can explain 43.6
percent of the data scatter. (Coal sulfur
content was in excess of 3 percent.)
Based on these analyses it is predicted
that a reduction in oxygen from 2.85
(baseline average) to 1.5 percent would
reduce NOX emissions by 37.6 percent.
Kiln ID Fan
Coal Mill -^
Stack
•—^*-feed from Silos
Preheater
Cyclone (1 of 4)
Quench
Air
Figure 2. Precalciner kiln arrangement ( G = gas. S = solid. SP = sample point,
PA = primary air. SA = secondary air, FF = flash furnace air).
Table 5. Emissions Data Summary - Precalciner Cement Kiln Location 8
Test No
8/1-1
8/1-1P
8/1-1P
8/1-1G
8/1-1A
Date
1980
8-6
8-7
8-7
8-8
8-8
Kiln Feed
Rate
kg/s ft/h)
24.1
26.5
265
28.8
28.8
(95.4)
11051
(105)
(114)
(1141
02
%
135
13.4
5.8
13.1
8.5
C02
%
12.8
13.7
20.0
15.0
NO
ppm"
1371
1249
912
1173
NCK
ng/J°
1054
960
545"
902
CO
ppm"
241
355
12
365
SOz
ppm'
51
0
0
24
Solid Total
Paniculate Paniculate Probe
Ib/W'Btu" ng/J° lb/W,Btuc ng/f Location
Stack
0.0886 38.1 01434 61.7 Stack
Kiln Outlet
Stack
Flash Furn.
Outlet
Comments
As found, gaseous emissions
As found, paniculates
As found
As found
As found, excess Oi at
flash furnace outlet
''dry, corrected to 3% 02.
b/VO, as NO,.
"dry, corrected to 3% 0& corrected for COi generation in the kiln and precalciner.
"dry, corrected to 3% Oz corrected for COj generation in the kiln only.
-------
Table 6. Mass Balances for Precalciner Cement Kiln
Element Input Ib/hr (kg/hr)
Sulfur
Sodium
Potassium
Coal
Raw Feed
Coal
Raw Feed
Coal
Raw Feed
324
319
643
0
139
139
0
1081
1081
1147)
(145)
(292)
10)
(63.1)
(63.1)
(0)
(490)
(490)
Clinker'
Alkali Bypass Baghouse"
Crusher/ Dryer Baghousec
SOi in stack
•
Clinker3
Alkali Bypass Baghouse"
Crusher/ Dryer Baghouse*
Clinker*
Alkali Bypass Baghouse"
Crusher/ Dryer Baghousec
Output Ib/hr
484
27
87
0
598
141
2
18
161
859
38
183
1080
(kg/hr)
(220)
(12.3)
(39.5)
(01
(272)
(64.0)
(0.9)
(8.2)
(73)
(390)
(17)
(83.0)
(490)
Output/Input
0.93
1.16
1.00
"Assumed clinker production = 0.65 of raw feed to preheater.
"Based on 0.5 ton/hr (454 kg/hr) of waste dust from alkali bypass baghouse.
cBased on 12 ton/hr (10.886 kg/hr) of return dust from crusher/dryer baghouse.
300
£200
<0
§
*
5: 100
\ I
NO* = 108.8 + 25.33 (% 0-J
Coefficient of Determination (R2) = 0.40
O Baseline
0
Figure 3.
Oxygen, %
Location 9—/VOx vs. oxygen.
However, it \s also predicted that the SOS
emission would increase by 46.6 percent.
The SOz dependence suggests a
reaction between SO2 and feed alkali
components in the presence of oxygen.
Laboratory and full-scale tests have also
shown that water vapor speeds up the
reaction between SOz and alkali. In this
respect the cement feed is performing as
a flue gas desulfurization agent; i.e.,
SCMg) + 0.5 CMg) + CaCOafs) + 2 H20(g) -
CaSO, • 2 H2O(s) + CCMg)
The above global reaction indicates that
both oxygen and water vapor are required
for the reaction between SO2 and
limestone (or lime).
Samples of coal, feed, clinker, and
precipitator catch were obtained and
analyzed in order to perform constituent
mass balances. The measured SOa
volumetric concentration was converted
to the corresponding sulfur mass rate.
The mass balance results are presented
in Table 7, which shows an overall
balance within 4 percent on a total basis.
Subscale Cement Kiln Simulator
The cement kiln simulator (Figure 5)
was designed to investigate four effects
(combustion air preheat, fuel injection
velocity, primary air oxygen content, and
excess air) on near-burner NOX for both
natural gas and coal fuels. Results
obtained at the subscale level are to be
used to select advanced combustion
modifications for implementation at full
scale.
Figure 6 shows the significant effect of
preheat on NOX production in the near-
burner zone. Also of importance is the
effect of fuel injection velocity, especially
at the higher preheats investigated. The
data at high preheat suggest that NOX
decreases at very high fuel injection
velocities. This effect may be due to the
decreased gas residence time within the
combustor which would inhibit NOx
production. Another possible explanation
would be that, at very high fuel injection
velocities, the mixing is so rapid that the
combustion would correspond to a
premixed flame for which the maximum
NOx would occur at 0 percent excess air.
The curve shown at high preheat is a
quadratic regression of NOX as a function
of fuel injection velocity, V,et; i.e.,
NO, = a + b V, + c V,lt
This function is able to account for 56
percent of the data scatter. The effect of
high fuel injection velocity on NOx is less
pronounced at the lower combustion air
temperatures.
Figure 7 quantifies the impact of carrier
oxygen content on NOX with coal fuel. The
implementation on full-scale kilns could
be accomplished by replacing a portion of
the carrier air with inert gas; e.g., flue
gas.
Overall oxygen content effectiveness in
reducing NOX is shown in Figure 8 for
coal fuel with and without air preheat.
-------
2500
2000
1500
en
g.7000
500
SO2 = 2610 • 454 (%
Coefficient of Determination (R2) = 0.44
O" Baseline
2 3
Oxygen. %
Figure 4. Location 9 • SO2 vs. oxygen.
Table 7. Mass Balance Results
In
Constituent Mq/d (tons/day)
Ah03
Si'Oi
Fez03
CaO
MgO
K20
S
40.1
152.1
21.7
468.2
25.4
5.96
5.98
(44.2)
067.7)
(23.9)
(516.1)
(28.01
(6.57)
(6.59)
Out
Mg/d (tons/day)
41.4
143.0
28.1
443.9
25.5
5.03
5.30
(45.6)
(157.6)
(31.0)
(489.3)
(28.1)
(5.55)
(5.84)
% Difference*
3.2
-6.0
29.5
-5.2
0.4
•15.6
-11.4
Total
719.4
(793.1)_
692.2
(763.0)
-3.8
100.
Hot or Cold
Secondary
Combustion Air
Air or
Nitrogen
^
Internal
Coo/ing
Air
. Cooling
\ Air
Cooling
Air
OJM^ —
tooling Air—' Pooling Air •
Refractory
Refractory
12 in.
24 in.
(30.5 cm) I (61 cm.)
\5or8 in. (12.7 or 20.3 cm)
— uiameier
Refractory
Refractory
>
Coal or Natural Gas
Figure 5. Schematic of subscale test furnace.
6
-------
50
700
/«t, m/s
150
200
3000
2500
f,2000
1500
I
7000
500
0
Figure 6.
OHigh Preheat. 1600°F(1144 K). 2.5-2.7% 02
O Medium Preheat. 800°F (700 K), 2.45-2.65% O2
A No Preheat, 100°F(311 K). 2.4-2.6% 02
O
O
250
300
I
7200
SOO
>t
i
400
200
400 600
Viet, ft/sec
800
1000
NO* emissions vs. injection velocity - natural gas fuel.
700
600
I
i
500
400
m
1
0 10
O2 in Carrier Stream, %
Figure 7. Effect of carrier 02 on NO* - coal fuel (Vjet = 31 ft/sec).
20
•fr U. S. GOVERNMENT PRINTING OFFICE 1983/759-102/0802
-------
I"
g.
x-
i
700-
j
600
500
400
350
w/Preheat
800-900°F
(700-756
w/o Preheat
= 52 ft/sec
(15.9 m/s)
R. J. Tidona, W. A. Carter, H. J. Buening. S. S. Cherry. andM. N. Mansourare with
KVB, Inc., Irvine, CA 92714.
Robert E. Hall is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Combustion Variable Effects on NO*
Emissions from Mineral Kilns," (Order No. PB 83-259 655; Cost: $11.50,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
123456
Excess 02, %
Figure 8. NOx vs. 02 - coal-constant V/et.
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
------- |