PB83-117069
Mechanisms of NOx Formation and Control
Alternative and Petroleum-Derived Liquid Fuels
Energy and Environmental Research Corp.
Santa Ana, CA
Prepared for
Industrial Environmental Research Lab.
Research Triangle Park, NC
1981
U.S. D«?ertiRKit of Commerce
Nations! Technical Information Service
-------
TECHNICAL REPORT DATA
(Pleetr rend lasuuctiunt on the went btfort compte'
1. REPORT NO.
EPA-600/D-82-344
PUBLISHED PAPER
PB83-117069
4. TITLE AND SUBTITLE
Mechanisms of NOx Formation and Control:
Alternative and Petroleum-de rived Liquid Fuels
5. REPORT DATE
1981
6. PERFORMING ORGANIZATION CODE
?. AUTHOR(S)
G.England, M. Heap, D. Pershing, and R.Nihart
(EERC); and G.B.Martin (EPA)
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Energy and Environmental Research Corporation
8001 Irvine Boulevard
Santa Ana, California 92705
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3125
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Symposium paper; 1981
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY MOTES ERL-RTP author Martin's mail drop is 62; his phone is 919/541-
7504. Paper was published in Eighteenth Symposium (International) on Combustion,
The Combustion Institute. 1981. DP 163-174.
s. ABSTRACT
paper gives results of burning petroleum-, coal-, and shale-derived
liquid fuels in a downf ired tunnel furnace to assess the impact of fuel properties on
the formation and control of NOx emissions. A nitrogen-free oxidant mixture (Ar,
CO2, O2) was used to isolate fuel NOx formation. Under excess air conditions, fuel
NOx correlated well with total fuel nitrogen content for both the petroleum and alter-
nate fuels. Under staged combustion conditions, the influence of fuel nitrogen content
was much less pronounced but equally highly correlated except for coal-derived
liquid. Exhaust NOx emissions were directly related to the amount of oxidizable
nitrogen species leaving the first stage. NO, HCN, and NH3 concentrations were
measured in the fuel-rich zone of the staged combustor as a function of stoichiometry
for seven liquid fuels and one CH4/NH3 mixture. Similar characteristics were obser-
ved for all liquid fuels. As the first stage stoichiometry (SRI) was reduced, NO con-
centrations at the first stage exit decreased; however, below SRI = 0. 8, HCN and
NH3 concentrations increased. Thus, the total fixed nitrogen (TFN = NO + HCN +
NH3) concentration passed through a minimum. Experimental data also indicated that
increasing the temperature of the fuel-rich zone decreased TFN concentration which
resulted in reduced exhaust NOx emissions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
C. COSATI 1-ield/GlOUp
Pollution
Fossil Fuels
Combustion
Nitrogen Oxides
Stoichiometry
Pollution Control
Stationary Sources
13B
21D
21B
07B
07D
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS I This Report)
Unclassified
ii.NO. OF PAGES.
13.
JO. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (»-73)
-------
EPA-600/D-82-344
PUBLISHED PAPER
Eighteenth Symposium (International) on Combustion The Combustion Institute. 1981
pgs. 163-174
MECHANISMS OF NO, FORMATION AND CONTROL: ALTERNATIVE
AND PETROLEUM-DERIVED LIQUID FUELS
G. C. ENGLAND,' M. P. HEAP, D. W. H3RSHING, R. K. NIHART
Energy and Environmental Research Corporation. 80O1 Irvine Boulevard. Santa Ana. California 92705
G. B. MARTIN
l). S. Environmental Protection Agency, Industrial Environmental Research Laboratory,
Research Triangle Park, North Carolina 27711
Petroleum-, coal- and shale-derived liquid fuels were burned in a downfircd tunnel furnace
to assess the impact of fuel properties on the formation and control of NO, emissions. A
nitrogen-free oxidant mixture (Ar. CO2, O2) was used to isolate fuel NO. formation. Under
excess air conditions fuel NO, correlated well with total fuel nitrogen content for both the
petroleum and alternate fuels. Under staged combustion conditions the influence of fuel
nitrogen content was much less pronounced but equally highly correlated except in the
case of a coal-derived liquid.
Exhaust NO, emissions were found to be directly related to the amount of oxidizable
nitrogen species leaving the first stage. NO, HCN and NH3 concentrations were measured
in the fuel-rich zcnc of the staged combustor as a function of stoichiometry for seven liquid
fuels and one CH4/NH3 mixture. Similar characteristics were observed for all liquid fuels.
As the first stage stoichiometry (SR,) was reduced, NO concentrations at the first stage
exit decreased, but below SR, = 0.8 HCN and Nil, concentrations increased. Thus, the
total fixed nitrogen (TFN = NO + HCN + NH3) concentration passed through a minimum
Experimental data also indicated that increasing the temperature of the fuel-rich zone decreased
TFN concentration which resulted in reduced exhaust NO. emissions.
Introduction of the controlling mechanisms and the trade-off with
smoke emissions. This paper summarizes a series
Increased coal utilization provides a partial solu- of experimental investigations carried out at atmo-
tion to current energy problems; however, the need spheric pressure to define the influence of fuel
for a balanced fuel economy and realities of retrofit properties and combustor design parameters on NO,
capability necessitate that in the future many indus- formation. Twenty-six petroleum-, shale- and coal-
trial users will be required to burn heavy liquid derived liquid fuels were burned in a refractory-
fuels, both petroleum- and coal- or shale-derived. lined tunnel furnace. In-flame measurements were
Since these fuels, particularly the alternativeliquids, made to quantify the influence of temperature and
have relatively high fuel nitrogen contents, there stoichiometry on the fate of fuel nitrogen species.
is potential for increased nitrogen oxides (NO.) The ultimate objective of the study was to provide
emissions unless preventative measures are taken.'" design criteria for full-scale lo .v emission oil burners
Staged combustion has proved to be an effective suitable for a wide range of fuels. The approach
NO, control technique for residual oil-fired was to divide the overall heat release process into
systems."1"'" However, optimization has proven a series of zones: 1) a fuel vaporization region; 2)
to be difficult because of a limited understanding a fuel-rich hold-up zone; 3) a second stage air
addition zone; and 4) a burnout zone. The results
"Author to whom correspondence should be presented in this paper aie mainly concerned with
directed. the optimization of the fuel-rich hold-up zone.
103
-------
164
COMBUSTION GENERATED POLLUTION
Experimental Systems
Tunnel Furnaces
The fuel comparison tests in this study were done
in a downfired, tunnel furnace which has been
described in detail previously.7 The vertical com-
bustion chamber was 2.1 m long and 20 cm in inside
diameter. The overall outer diameter was approxi-
mately 81 cm and the walls consisted of insulating
and high temperature castable refractories. All air-
streams were metered with precision rotameters and
the artificial atmospheres were supplied from high
pressure cylinders.
The staged combustion tests were conducted in
a separate but similar tunnel furnace illustrated in
Fig. 1. In this system the vertical combustion
chamber was 2.3 m long and 15.2 cm in inside
diameter. The walls contained slots which allowed
insertion of stainless steel cooling coils, thus, the
temperature profile within the furnace could be
significantly changed by varying the location and
number of these coils. The slots were also used
for inserting a refractory choke/secondary air injec-
tion system. A 5 cm inside diameter choke was used
to prevent recirculation of second stage air into the
first stage. The secondary air injector was a 10.2
FUEL »
ATOMIZING AIR
PRIMARY AIR
RESIDENCE
TIME
METERS SECONDS
0-1 O-i
0.5-
1.0-
1.5-
2.0-
I.O-J
i_ INSULATINQ
^//f REFRACTORIES
Fic. 1. Furnace configuration showing approxi-
mate residence time for SR, = 0.78.
cm diameter stainless steel ring with 24 6.3 mm
diameter holes drilled parallel to the axis of the
furnace. The injector was designed to provide grad-
ual mixing of the secondary air with the primary
combustion products. The velocity ratio of the sec-
ondary air to the incoming combustion products
was approximately 2-to-l at a first-stage stoichiome-
try of 0.8. In some tests a second refractory choke
was also inserted near the top of the furnace to
shield the initial fuel vaporization zone from cooling
coils in the first stage. This "radiation choke" made
it possible to vary temperatures in the fuel-rich
hold-up zone without influencing the rate of droplet
vaporization and initial combustion.
The full-load firing rate in all experiments (both
furnaces) was 0.53 cc/sec. In both systems the fuel
injector consisted of a 19 mm diameter stainless
steel tube containing an atomizing air supply tube,
fuel supply tube, a cartridge heater for final oil
temperature control, and a chromel/altimel thermo-
couple for accurate measurement of oil temperature.
A commercial, ultrasonic oil atomizer was used
because it provided excellent atomization of the
heavy fuel oils at relatively low flow rates even
though the atomizer is designed for lighter oils. The
spray from this atomizer has been characterized in
a separate study -using a laser diffraction drop size
analyzer and produces a very narrow distr'bution
of drop sizes centered at 20 u,m under the operating
condition of these experiments. .
Analytical System
Exhaust concentrations were monitored contin-
uously using a chemiluminescent analyzer for NO
and NO,, an NDIR analyzer for CO and CO2, a
paramagnetic analyzer for O2, and an FID analyzer
for total hydrocarbons (THC). The flue gas was
withdrawn from the exhaust stack'through a water-
cooled, stainless steel probe1 using a stainless
steel/Teflon sampling pump. Sample conditioning
prior to the instrumentation consisted of an ice bath
water condenser and glass wool and Teflon fiber
filters. All sample lines were 6.3 mm Teflon and
all fittings Type 316 stainless steel.
In-flame temperature measurements were made
using a standard 19 mm diameter suction pyrometer
containing a platinum-platinum/rhodium thermo-
couple. In-flame gas samples were withdrawn with
a long, stainless steel water-quench probe. HCN
and NH, were absorbed in a series of impingers
ana concentrations determined using specific ion
electrodes. Sulfide ion interference was minimized
by. the addition of iead carbonate."
Fuels
An independent laboratory determined the
characteristics of the liquid fuels used in this study
which are given in Table I.
-------
ALTERNATE AND PETROLEUM-DERIVED LIQUID FUELS
165
! Fuel Effects-Excess Air
Figure 2 is a composite plot of total and fuel
NO, emissions fcr all of the fuels tested at 5 percent
exce:B O2, 405°K air preheat and 0.53 cc/sec firing
rate. All data were obtained with the ultrasonic
nozzle and an air atomization pressure of 103 kPa
(gage). Fuel viscosity was maintained at 12 ± 1.5
cs by appropriate selection of fuel temperature. Total
NO. refers to emissions produced with air as the
oxidunt. Fuel NO, is defined as the emissions mea-
sured when molecular nitrogen was excluded and
the fuel burned in an atmosphere composed of argon,
carbon dioxide and oxygen.7 Results obtained with
petroleum-derived fuels are presented in Figure 2a.
Both, total and fuel NO, emissions increase with
increasing fuel nitrogen content. Distillate oils
doped with pyridine and thiophene (slashed sym-
bols) correlate well with the actual heavy petroleum
liquids under the conditions of these tests (fine oil
atomization, rapid mixing burner). The difference
between total and fuel NO, data is defined to be
thermal NO, and is approximately constant for all
the heavy residual oils. However, thermal NO,
produced from distillate oils is consistently higher
than that produced from residual oils.
A comparison of total and fuel NO, produced with
alternative and petroleum-derived fuels is presented
in Figure 2b. In general, alternative fuels have higher
nitrogen contents than petroleum-derived fuels and,
therefore, produce higher fuel NO, emissions. With
few exceptions, fuel NO, emissions from both al-
ternative and petroleum-derived fuels appear to
correlate well on the basis of fuel nitrogen content.
Total NO, emissions from alternative fuels (dashed
line) are higher than from pure pedoleum-derived
fuels, suggesting a greater production of thermal
NO,. This may be because the alternative fuels
contain significantly larger fractions of light hydro-
carbons which cause higher peak flame temperatures
since more fuel is burned prior to vitiation of the
combustion air with recirculated combustion prod-
ucts, or because the relatively high vaporization
rates produce higher local combustion intensities.
Staged Combustion
First Stage Stoichiometry
During fuel-rich combustion the original fuel
nitrogen species are ultimately converted to N2,
1000
BOO
geoo
g
5"
: 400
200
2400
2000
1600
1200
BOO
40O
0.2 0.4 O.6 0.8 6 0.3
FUEL NITROGEN (WTX)
1.0
1.3
2.0
2.5
(b)
Fie. 2. Total and fuel NO, produced by (a) petroleum oils (/ denotes doped distillate or residual
oils) and (b) all liquid fuels.
-------
TABLE I
Detailed fuel analyses
Symbol
Ultimate Analysts:
Orlxin. %
IlytlroKrn, %
Nitrogen. %
SlllftlT. %
Ash. %
Oxyjcrn. t
Ojitfjdion Car-
bun Residue, %
Asph-iltene. %
FL«h Point. »F
Pour Point. *F
API Gravity >t
60°F
Viscosity. SSU, at
140-F
at 210-F
Heat of Combustion:
Cross Rtu/lb
Net Btu/lb
Calcium, ppro
Iron, ppm
Mtiijjnese. ppm
MtiK'wsium, ppm
Nickel, ppm
Sodium, ppm
Vanadium, ppm
m
Alaskan
Diesel
*
86.90
12.07
0.008
0.31
<.001
0.62
33.1
33.0
29.5
D2 m
Cali-
VI; - ti>rni»
Trial' No. 2
Diesel . Oil
A 9
mm HUH
9.7-1 12.52
0.028 0.053
1.88 0.27
<.OOI <.001
0.24 038
18.3 32.6
32.0 30.8
28.8 29.5
19.330.
Rl
,
Kast
Coast
o
H6.54
12.31
0.18
036
0023
0(11
2.1
0.34 :
205
50
24.9
131.2
45
• 19.260
18.140
7.1
16
0.09
3.7
6.7
37
14
R2.
Middle
K*st
b.
80.78
1 1.1)5
0.18
O.B7
0.012
041
6.0
3.24
350
48
19.S
490
I3l.fi
19,070
17.9SO
1.2
2.6
0.02
008
13
0.98
25
R3
Law
Sulfur
No. 8
Oil
*
Hfi.57
12.52
0.22
0.21
' 0.02
0.46
•
44
0.94
325
105
25.1
222.4
69.6
19.110
17.970
9.52
123.6
0.46
2.23
14.10
3.74
3.11
R4
Indo/
Malay-
sian
A
86 .5.1
11.93
0.24
0.22 '
0.036
1.04
3.98
0.74
210
61
21.8. •
199
65
.
• 19.070
17.980 •..
14
16
0.13
3.6 ,
19
15
• 101
R5
Venezu-
elan
Desul. -'
fur-
feed
0
S5.92
12.05
0.24
093
0033
O.K3
' 5.1
2.59
176
48
• 23.3
113.2
50.5
18.400 .
17.300 -•
S.7
6.5 :
0.09
3.6
-
'
R6
Pennsyl-
vania
(Antarada
Mess)
b
14.82
11.21
0.34
2. 26
0.067
1.3
12.4
4.04
275
68
' 15.4
1049
240
18.520
. , 17.500
9.2
13.2
0.10
3.3
32.7 .
-". 64.5
81.5
R7
Gulf
Coast
O
84.62 '
10.77
0.36
2.44
0027
1.78
14.8 :
7.02 •
• 155
40
13.2
835
. 181
18.240
" 17.260
4.4 '
. 19
; 0.13
0.4
. 29
: 3.6 '
45
R8 .
Vene- •
Kuelan
0
8.5.24
10.96
0.40
2.22
0.081
1.10
6.8
8.4 •
210
58
14.1
742
'.96.7
18.240
17.400
9.1
11 -
019
3.8
52
32
226
R9
- . _ •.
•' Alaskan
D
86.04
11.18
0.51
1.63
0.034 "
0.61
12.9
5.6 '
!15
38
15.6
1.071
" 194
18.470 .
. 17.580
6.9
24
0.06
1.4
50
37
67
RIO
-•
Calif-
Fomia
V
85.73
11.83
0.62
1.05
0.038
0.71
19.5
246.1
70.00..
•"
:
2
ec.
i
o
0
[=1
D
r
c
o
z
-------
TABLK
I
(continued) ,
Ultima!? Ana!v
Calcium, ppni
Iron, ppni
Manijalifse. ppni
Magnesium, ppni
Ni. If 1. ppni
Sodium, ppni
Va.tatltum. ppni
Rll
Call-
furi.ia
0
S54
11.44
0.77
1.69
0.043
0.71
fl.72
5. IS
Ck(
3S ''.
15.4
K54
129
IS.470
17.430
21
7.1
O.S
5.1
65
21
44
RI2
Call.
fornia
0
S5.33
11.23
0.79
1.BO
0.0.12
1.02
9.22
S.IS
150
.10
15.1
74S.O
131.0
IS.4SO
17.440
14
53
0.1
3.S
S2
2.6
53
RI3
Cali-
fornia
C2
SB.OK
10.44
O.S«
0.99
0.20
O.h5
15.2
S.62
ISO
42
12.6
720
200
IS.2,10
I7.2HO
U0.6
77.2
O.S7
3 1. 4
ss.o
22.3
66.2
RI4
Cali-
fornia
(Kern
Counts-}
0
S6.61
10.93
O.S3
1.16
0.0.10
0.44
Mi
3.98
225 '
65
1 2.3
46.10
3.52
IS.430
17.4.10
4.4
15
0.15
1.1
6S
.1.4 •
.19
Al
Shall--
l)i-rivt-«t
UKM
f
SK.IS
13(10
0.24
0.51
OIKll
1.07
4.1
0.0.16
IS2
40
33.1
3B.I
30.7
IH.430
1S.240
0.13
6,1
0.06
0.43
0.09
1.1
A2
Synthuil
.
SB.30
7.44
1,16
O.SO
I.56
2,54
2.1.9
IB.55
210
SO
S'l.l-l
IO.S«0
f •**?
IH.-ISO
15.SOO
1 .670
109
«.2
170
2.6
I4S
(15
A3
Crinlr
Shall-
.
S4.6
11,1
2.0S
0.63
.026
1,16
2.»
1.33
250
SO
20.3
97
44.1
IS.290
17,260
1.5
47.9
0.17
5.40
5.00
11.71
<,1
A4
site n
Klrnd
A
S9.91
9.27
0.45
0.065
OOO4
0,111
«. IS
4.10
70
<-72
10.0
40.6
' 32.5
I7.!«0
17.1.10
0,13
3.9
<0,5
0.17
<0,5
0.31
<1.0
A5 A6
Shall-
KI-II-
-------
168
COMBUSTION GENERATED POLLUTION
HCN, NO or NH3.010" The intent of staged com-
bustion is to minimize the concentration of total
fixed nitrogen (NH;, + NO + HCN = TFN) prior
to second stage air addition. Figure 3 presents data
obtained from detailed in-flame measurements for
a distillate oil (Dl-Alaskan diesel), a typical residual
oil (R9-Alaskan Bunker C), and an alternate liquid
fuel (A9-SRC-H heavy distillate). These measure-
ments were made on the centerline of the furnace
at a distance of 103 cm (approximately 630 msec)
from the oil nozzle. Detailed radial measurements
indicated that the concentration profile was essen-
tially uniform at this location. These in-flame data
are reported on a dry, as-measured basis. Second
stage air was added at 107 cm subsequent to each
in-flame measurement in these experiments to bring
overall theoretical air to 116 percent, and exhaust
NO, measurements, (NO,)K, are also shown (on a
dry, 0 percent O2 basis).
Decreasing the first stage stoichiometric ratio
reduced NO concentration leaving the first stage.
However, below a stoichiometric ratio of approxi-
mately 0.8 significant amounts of hydrocarbons
(THC) and HCN were measured. Thus, there exists
a minimum in exhaust NO, concentrations because
of a competition between decreased first stage NO
and increased oxidizable nitrogen species such as
HCN. It should be noted that HCN concentration
increased rapidly below SR, = 0.75 and that this
increase was accompanied by an increase in hydro-
carbon content of the partially oxidized combus-
tion products.
Data for the Alaskan diesel oil (Fig. 3a) also show
the presence of much smaller but significant con-
centrations of HCN and NHn, although this fuel
is essentially nitrogen-free. Total conversion of the
• fuel nitrogen would produce 15 ppm TFN at SR,
= 0.7. This confirms previous work'"'" 12 which
demonstrated that reactions involving hydrocarbon
fragments and N2 or NO can produce HCN.
Total fixed nitrogen measured at the exit of the
fuel-rich hold-up zone is presented in Fig. 4 as a
function of stoichiometric ratio for the Alaskan diesel
oil, three petroleum-derived liquid fuels, three alter-
nate liquid fuels, and methane doped with 0.79
weight percent nitrogen as ammonia. All fuels in-
vestigated showed similar trends. TFN reached a
minimum concentration at approximately SR, = 0.8,
240 -
O.6
(o)
(b)
Fir.. 3. NO, HCN, and NHn production in the fuel-rich hold-up zone for (a) Dl-Alaskan diesel oil,
(b) R9-Alaskan bunker C oil, and (c) A9-SUC-II heavy distillate oil (3% overall excess oxygon).
-------
ALTERNATE AND PETROLEUM-DERIVED LIQUID FUELS
169
1000
BOO
g 600
400
200
1800
1600
1400
1200
1000
800
600-
400
200
0.6
0.8
1.0
0
SR,
0.6
0.8
I.O
(0)
(b)
Fie. 4. The effect of fuel type and stoichiometry on the production of TFN (XXN,) for (a) petroleum
oils, and (b) alternate fuels.
and then increased significantly as the primary zone
became more fuel-rich. The TFN for both the am-
monia-doped methane and the fuel oil of similar
nitrogen content (R14) showed similar characteris-
tics under fuel-lean conditions. However, below SR,
= 0.7 the heavy oil produced significantly less TFN
than the gaseous fuel.
Figure 5 shows the partition of TFN as a function
of the first stage stoichiometric ratio. Data for the
six liquid fuels and the 0.79 weight percent nitrogen
(as NH3) in methane gas are presented as the per-
centage of the original fuel nitrogen existing as HCN,
NH3 or NO at the end of the fuel-rich first' stage.
In each case the missing fuel nitrogen has either
been converted to Na or still exists in some partially
oxidized fuel fragment. All of the liquid fuels exhib-
ited the same general trends:
• Above approximately SR, = 0.8, NO was the
dominant TFN species.
• Below SR, = 0.75 the major TFN species was
HCN and under very fuel-rich conditions HCN
usually accounted for more than 50 percent of
the initial fuel nitrogen.
• First stage Nil., levels were always low, generally
less than 20 percent of the fuel N. Axial profiles
indicated that near SR, = 0.8, significant amounts
of NH3 were formed early in the rich zone, but
decayed rapidly.
Alternative liquid fuels produced TFN distribu-
tions very similar to those observed with petroleum-
derived liquids. These results are in marked contrast
to similar data obtained with pulverized coal" which
demonstrated that the TFN yield and the preferred
species were both strong functions of coal composi-
tion.
The fraction of fuel nitrogen converted to HCN
was somewhat less for the CH4/NH, mixture (JJO
than for most of the liquid fuels tested (Fig.
5a); however, conversion to Nil, (Fig. 5b) was
higher than the liquids fuels. The light (API =
21.8) Indonesian/Malaysian residual oil (A) pro-
duced consistently higher conversion of fuel
nitrogen to HCN, NH., and NO than the much
heavier (API = 12.3) Kern County crude oil (Q)-
Conversely, conversion to HCN and NIf, for the
+850°F. shale residuum (II) was higher than for
the lighter parent crude shale (®); however, conver-
sion to NO was lower. The exact relationship of
-------
COMBUSTION GENERATED POLLUTION
1.0
(b)
(e)
Fie. 5. The conversion of fuel-bound nitrogen to (a) HCN, (b) NH3, and (c) NO.
simple fuel properties to conversion of fuel nitrogen
to intermediate XN species is not apparent at this
time.
TFN Contersion in Second Sfagc
Exhaust NO, emissions in a staged combustor
result from conversion of TFN exiting the first stage
and any thermal NO, production during burnout.
Second stage thermal NO, production is not consid-
ered to be significant in this study because changes
in heat extraction during burnout had almost no
effect on final emissions. Figure 6 shows minimum
exhaust NO, and the associated TFN as a function
of total fuel nitrogen content for the seven liquid
fuels and the CII.,/NH,. In general, the minimum
exhaust NO, concentration occurred at SR, = 0.78
± 0.02. Under these optimum staged conditions NOt
emissions (and TFN) correlate well with total fuel
nitrogen content, but the slope is significantly less
than under excess air conditions (Fig. 2). Only the
SRC-II heavy distillate (ft) exhibited unusually high
emissions, and this was the direct result of a high
TFN yield (Fig. 6b). NO, NH, and HCN were all
substantially higher for this fuel than with compara-
ble fuels at the minimum stok-.iiometry (0.78). The
increased TFN may be the result of basic chemical
bonding differences between the parent coal from
which the SRC-II was derived and petroleum liq-
uids.
Figure 7 shows the percent of the TFN exiting
the first siage which was converted to NO, in the
second stage (based on NO, measured in the furnace
exhaust). Data for all eight fuels and first stage
sloichiomctrics between 0.5 and 0.8 have been in-
cluded. As the TFN concentration increased (due
to decreasing SR,), the percentage conversion de-
creased. This behavior is analogous to that reported
'previously for poorly mixed excess air diffusion
flames." The second stage burnout can be consid-
ered an excess air flame with a gaseous hydrocarbon
furl. The general conversion decrease rngy be the
result of a competition between N2 formation reac-
tions which are second order in XN and first order
NO formation reactions.
The distillate oil results «^) in Fig. 7 are substan-
tially below all of the residual petroleum and al-
ternative liquid data and the CH4/NH, results are
-------
ALTERNATE AND PETROLEUM-DERIVED LIQUID FUELS
171
400
g
8
8 40O
!
no
UINIUUU (N0x)f
I I I I I I I I I I I I
0.6 1.2 I « 2.0
FUEL MTftOOEN (WTXI
2.4
Fic. 6. The effect of fuel-bound nitrogen content
on exhaust NO. and TFN in the primary zone (£XN,)
(SR, = 0.78. 3% overall excess Oz).
slightly high. This suggests that TFN conversion
is also dependent on the nitrogen specialion and/or
Che oxidaliv,:. uf the partial products of combustion
at the exit oi" the fuel-rich zone. At the TFN = 145
ppm point, the first stage exhaust from the distillate
oil contained 10,000 ppm hydrocarbons and the
dominant TFN species was HCN (SR, = 0.6). In
contrast, the comparable heavy oil points had less
l:han 400 ppm hydrocarbons and NO was the major
TFN species (SR, = 0.78).
Fic. 7. Conversion of TFN to exhaust NO, (0.5 £
SH.sO.8).
Time-Temperature History
The TFN concentrations shown in Fig. 6 arc in
excess of equilibrium levels, and Sarofim and co-
workers" have suggested that increasing the tem-
perature of the primary zone would prove beneficial.
The results presented in Figs. 8 and 9 obtained
with the shale crude (A3) demonstrate the impact
of first stage heat loss on the fate of fuel nitrogen.
The lu)t conditions refer to the furnace without
primary zone cooling and with the radiation choke
(as shown in Fig. 1). To cool the furnace cooling
coils were added prior to the second stage choke
and the radiation choke was removed. Figure 8
shows exhaust NO, and primary zone gas tempera-
ture as a function of stoichiometric ratio for the
two conditions. Minimum NO, emissions were re-
duced for the hot case, and Ihe optimum stoichi-
ometry was shifted toward more fuel-rich condi-
tions.
The axial profiles presented in Fig. 9 provide an
explanation for this shift in the minimum emission
levels. Heat extraction in the first stage impacts the
rate of decay of TFN. Under cold conditions both
NO and HCN essentially freeze, whereas without
heat extraction the initial rate of decay for all three
species is much, faster, leading to !ow TFN con-
centrations at the exit of the fuel-rich first stage.
It should be noted that heat extraction also affects
the rate of CO oxidation.
Discussion
NO, control via staged combustion involves opti-
mizing the first stage of the combustor with respect
to (1) thermal environment, and (2) sloichiometry.
It appears that heat loss from the first stage should
be minimized for liquid fuels because elevated
temperatures accelerate the rate of XN species decay
toward low equilibrium levels. The data presented
in Fig. 8 clearly indicate that an increase in first
stage temperature results in a decrease in NO,
emissions.
The optimization of first stage stoichiometry is
more complex. Measurements within Ihe fuel-rich
first stage show that at the optimum stoichiomelry
NO, HCN and Nil, co-exist. TFN exists mainly as
HCN and Nil, for more fuel-rich conditions, and
for conditions less fuel-rich than the optimum, NO
is Ihe primary fixed nitrogen species. Three factors
influence Ihe sloichiometry at which Ihe minimum
exhaust NO, emissions will occur: (1) Ihe amount
of TFN species present, (2) the relative conversion
efficiencies of Ihe TFN species to NO in the second
stage flame, and (3) Ihe influence of other partial
combustion products on TFN conversion.
Figure lOa shows Ihe percentage conversion of
Ihe first stage TFN for the •: 850°F shale fraction.
-------
172
COMBUSTION GENERATED POLLUTION
600
TOO
£400
*
3OO
ZOO
100
D HOT PRIMARY
O COOL PRIMARY
I
I
ISOO
1200
1100
IOOO
I
J_
I
0.3 0.6 O.7 0.8 0.9 0.3 O.6 0.7 0.6 0.9
SRl
Fie. 8. The influence of fuel-rich hold-up zone temperature profile on exhaust NO,.
0.2 O.I 0.6 0.80 O.2 0.4 0.6 O.6
SECONDS
Fie. 9, CO oxidation and decay of XN species for two temperature profiles.
-------
IOO
I80
z- 60
a
8 20
ALTERNATE AND PETROLEUM-DERIVED LIQUID FUELS
100
0.80
0.5O
3OO 900 700 IOOO 2000 400O 6000
PPM rXN,(ORY. 0%0Z1
(b)
Fie. 10. Influence of first stage exit composition on second stage NO, formation (A8 — +850°F shale
fraction).
As the first stage sloichiometry is progressively
reduced from SR, = 1.05, the TFN decreases (due
to decreased first stage NO formation) and the
percent conversion also decreases (due to increased
second stage combustion). These two effects result
in a rapid decrease in exhaust NO,. Below SR, = 0.8,
the TFN begins to rise dramatically because of HCN
and NH, formation, but the percent conversion
continues to decrease. Thus, in general, there exist
two first stage stoichiometries with identical TFN
levels (e.g., 1000 ppm) but with vastly different
conversion efficiencies (77 percent versus 26 per-
cent) and, hence, different exhaust emissions.
Figure lOb indicates that there are two major
differences between the upper and lower percent
conversion curves. First, below approximately SR,
= 0.8, the major TFN species shifts from NO to
HCN. Second, at first stage stoichiometries less than
0.8, significant amounts of unburncd hydrocarbons
are carried into the second stage flome. Folsom el
al." have shown that in a simple diffusion flame
the conversion of HCN and NH3 to NO and the
retention of NO arc affected by the hydrocarbon
content of the fuel. NH3 conversion is low when
hydrocarbons are absent, and hydrocarbons are
necessary for the reduction of NO or HCN to N2.
Thus, staged combustion requires the minimization
of TFN, as well as the production of an optimum
TFN/fuel mixture for burnout.
Conclusions
Data have been presented showing the impact of
fuel properties on NO, formation from liquid fuels
under both excess air and staged conditions. Inves-
tigations with a wide range of fuels indicate that:
1. Fuel NO formation from petroleum-, shale- and
coal-derived liquid fuels can be correlated by
the total fuel nitrogen content under excess air
conditions.
2. NO, control for high nitrogen fuels is most
effective when a rich primary zone is held at
an optimum stoichiometry to minimize both the
TFN concentration and the TFN conversion to
NO in the second stage flame. This concentration.
can be minimized by increasing the temperature
of the fuel-rich zone and/or reducing the initial
fuel nitrogen content. The second stage conver-
sion decreases with increasing hydrocarbon con-
tent.
Acknowledgments
These investigations were carried out under Unit-
ed States Environmental Protection Agency (EPA)
Contract 68-02-3125 and the authors gratefully ac-
knowledge '.he guidance of the Project Officer, \V.
S. Lanier. The authors also wish to acknowledge
the assistance of J. A., Naccarato, J. G., Llanos and
J. H. Tomlinson in conducting the experiments, and
particularly to \V. C. Rovesti of Electric Power
Research Institute, J. E. Haebig of Gulf Research
and Development Company and L. Lvikens of the
United States Department of Energy for their help
in obtaining several alternative fuels.
REFERENCES
1. MANSOUH, M. N. AND M. GKKSTKIN, "Correlation
of Fuel Nitrogen Conversion to NO, During Com-
-------
174
COMBUSTION GENERATED POLLUTION
bustion of Shale Oil Blends in a Utility Boiler,"
Symposium on Combustion of Coal and Syn-
thetic Fuels, American Chemical Society,
(March 1978).
2. HKAP. M. P. "Control of Pollutant Emissions
from Oil-Fired Package Boilers," Proceedings
of the Stationary Source Combustion Sympo-
sium, EPA-600/2-76-1526, NTIS, Springfield,
VA (1976),
3. MAHTIN, G. B. AND E. E. BKRKAU, "An Investiga-
tion of the Conversion of Various Fuel Nitrogen
Compounds to NO in Oil Combustion," AlChE
Symposium Series No. 126, 68 (1972).
4. TURNKR, D. W., R. L. ANDHKWS, ANI> C. W. SIKC.
MUND, "Influence of Combustion Modification
and Fuel Nitrogen Content on Nitrogen Oxide
Emissions from b'uel-Oil Combustion," AlChE
Symposium Series tfo. 126, 68(1972).
5. APPLKTON, J. P. AND J. B. HKVWD, "The Effects
. of Imperfect Fuel-Air Mixing in a Burner on
NO Formation from Nitrogen in the Air and
the Fuel," Fourteenth Symposium (Internation-
al) on Combustion, Combustion Institute, Pitts-
burgh, PA (1973).
6. PKRSHINI:, D. \V., G. B. MARTIN, AND E. E. BKHKAU,
"Influence of Design Variables on the Produc-
tion of Thermal and Fuel NO from Residual
Oil and Coal Combustion," AlChE Symposium
Series No. 148. 71 (1975).
7. PKRSHINC:, D. W., J. E. CICIIANOWICZ, G. C. ENG.
LAND. M. P. HKAP, AND G. B. MAHTIN, "The In-
fluence of Fuel Composition and Flame Tem-
perature on the Formation of Thermal and Fuel
NO, in Residual Oil Flames," Seventeenth
Symposium (International) on Combustion, The
Combustion Institute. Pittsburgh, PA (1979).
8. CHKN, S. L., M. P. HKAP, R. K. NIIIART. D. W.
PKKSHINC AND D. P. RKI:S, "The Influence of Fuel
Composition on the Formation and Control of
NO, in PuKeri'/ed Coal Flarres," paper present-
ed at the 'A'SS Combustion Institute, Irvine,
California .1980).
9. TAKAOI.T.. T. TATSUMI AND M.OC.ASAWARA, "Nitric
Oxide Formation from Fuel Nitrogen in Staged
Combustion: Roles of HCN and NH,," Com-
bustion and Flame, 35, 17 (1979).
10. FKNIMOKK, C. P., "Studies of Fuel Nitrogen Spe-
cies in Rich Flame Cases," Seventeenth Sympo-
sium (International) on Combustion, The Com-
bustion Institute, Pittsburgh, PA (1979).
11. HAVNKS, B. S., "Reactions of NH3 and NO in
the Burnt Cases of Fuel-Rich Hydrocarbon-Air
Flames," Combustion and Flame, 28, 81 (1977).
12. CKRHOLD, B. W.. C. P. FKNIMORK AND P. K. DK.
DKRICK, "Two Stage Combustion of Plain and
N Doped Oil," Seventeenth Symposium (In-
ternational) on Combustion. The Combustion
Institute, Pittsburgh, PA (1979).
13. SAHOFIM, A. F., J. H. POHL AND B. R. TAYLOR,
"Mechanisms and Kinetics of NO, Formation:
Recent Developments," paper presented at 69th
Annual Meeting AlChE, Chicago, November
(1976).
14. FOLSOM, B. A., C. W. COURTNEY AND M. P. HKAP.
"The Effects of LEG Composition and Com-
bustor Characteristics on Fuel NO, Formation,"
paper presented at the ASME Gas Turbine
Conference and Exhibit and Solar Energy Con-
fercnce^San Diego, California (1979).
-------
------- |