United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/S7-86/003 May 1986
v>EPA Project Summary
Environmental Assessment of
an Enhanced Oil Recovery
Steam Generator Equipped
with a Low-NOx Burner
C. Castaldini, L. R. Waterland, and H. I. Lips
The report discusses emission results
obtained from sampling the flue gas from
an enhanced oil recovery steam generator
equipped with a Mitsubishi Heavy In-
dustries (MHI) PM Iow-N0x crude oil
burner. The test program performed in-
cluded burner performance/emission map-
ping tests, comparative performance
testing of an identical steamer equipped
with a conventional burner, and com-
prehensive testing of the low-NOx-bumer-
equipped steamer at a nominal low-NOx
setting. Emission measurements for the
comprehensive tests included continuous
monitoring of flue gas emissions; source
assessment sampling system (SASS)
testing with subsequent laboratory
analysis of samples to give total flue gas
organics in two boiling point ranges and
specific quantitation of the semivolatile
organic priority pollutant species and other
major semivolatile organics; C, to C6
hydrocarbon sampling; Method 5 par-
ticulate sampling; Method 8 sampling for
SO2 and S03 emissions; emitted particle
size distribution measurements using
Andersen impactors; and N2O emission
sampling.
Full load NOX emissions of 110 ppm
(corrected to 3 percent 02) could be
maintained from the low-NOx burner at
acceptable CO and smoke emissions. This
compares to emissions of about 300 ppm
(3 percent O2) measured from the con-
ventional burner equipped steamer, again
at acceptable CO and smoke emissions.
Comprehensive tests were performed at
a burner operating condition giving NOX
and CO emissions of 106 and 93 ppm,
respectively, with flue gas O2 of 3.0 per-
cent. Under these conditions, SO2 and
SO3 emissions were 594 and 3.1 ppm,
respectively. Particulate emissions were
39 mg/dscm with a mean particle diameter
of 3 to 4 pin (two impactor runs). Total
organic emissions were 11.1 mg/dscm and
almost exclusively volatile (C, to C6)
organics.
Of the polynuclear aromatic hydrocar-
bons analyzed for, only naphthalene (1,4
^g/dscm), phenanthrene (0.3 ^g/dscm),
and pyrene (0.11 m/dscm) ware detected.
Other semivolatile ketones and ox-
ygenated aromatics were measured at
levels ranging from 0.1 to 34 ^g/dscm.
This Project Summary was developed
by EPA's Air and Energy Engineering
Research Laboratory. Research Triangle
Park, NC, to announce key findings of the
research project that is fully documented
in two separate volumes (see Project
Report ordering information at back).
Introduction
The petroleum reserves which can be
recovered through primary production
methods have been essentially exhausted
in many of the older oil fields in the U.S.
Remaining reserves in these regions are in-
creasingly being produced through what
have been termed enhanced oil recovery
(EOR) methods. A popular EOR technique
involves injecting steam into a field to
lower the viscosity of the remaining crude
so that it can be recovered. This steam for
injection is raised by crude-oil-fired steam
generators. Since the aggregate NOX
emissions from many steamers in a region
can be significant, they have received
-------
close regulatory attention in some regions,
notably Kern County, California.
In an effort to reduce NOX emissions
from EOR steamers, burner manufacturers
have experimented with low-NOx burner
designs. One such burner has been
developed in Japan by MHI and is currently
offered in the U.S. by CE-Natco, a steamer
manufacturer.
A steamer equipped with an MHI low-
NOx burner was subjected to burner
performance tests and comprehensive
emission testing. In addition, another
steamer equipped with a conventional
burner was subjected to abbreviated emis-
sion testing for emission comparisons.
Results of these tests are summarized in
this report.
Summary and Conclusions
Source Description
Tests were performed on two CE-Natco
model STOF steam generators rated at 15
MW (50 x 106 Btu/hr) heat output. One
unit was equipped with a conventional
North American burner; the other had
been retrofitted with an MHI PM low-NOx
burner.
The MHI PM burner, shown schemat-
ically in Figure 1, uses a split flame ar-
rangement, whereby an inner air-deficient
diffusion flame is separated from an outer
air-rich premix flame by a blanket of recir-
culated flue gas. This arrangement pro-
duces NOX levels significantly lower than
those from conventional burners. A certain
amount of staged overfire air (OFA),
typically 10 percent, is injected about
half-way down the length of the cylindrical
furnace through three sets of three ports
each, located at about the 4, 8, and 12
o'clock positions. This OFA ensures that
sufficient excess air and mixing are
achieved prior to the combustion gas's
leaving the furnace.
Test Program
A limited set of flue gas emission tests
were performed on the conventional-
burner-equipped steamer. In these tests,
flue gas emissions were measured at two
steamer loads (full and 75 percent) while
varying the overall excess air level. Next,
the low-NOx-burner-equipped steamer
was subjected to relatively detailed burner
performance testing. In these tests, flue
gas emissions were measured while vary-
ing the burner operating parameter set-
tings at full load. The burner parameters
varied were the flue gas recirculation
(FGR) rate; the relative distribution of com-
bustion air to the premixed flame nozzles.
Premix
Air Nozzle
Flue Gas
Recycle
Nozzle
Diffusion
Nozzle
Flue Gas
Recycle
Nozzle
Premix
Air Nozzle
Figure 1. The MHI PM burner nozzle.
the diffusion flame nozzle, and the OFA
ports; and the overall excess air level.
Finally, comprehensive emission testing
was performed on the low-NOx-burner-
equipped steamer with the burner set at
a nominal low-NOx condition. The sam-
pling and analysis procedures for these
comprehensive tests conformed to a
modified EPA Level 1 protocol. The mea-
surements included:
Continuous monitoring for NOX, CO,
CO2, O2, and total unburned
hydrocarbon (TUHC)
Source assessment sampling system
(SASS) training sampling
EPA Method 5 sampling for solid
paniculate
EPA Method 8 sampling for sulfur ox-
ides (S02 and S03)
Particle size distribution measure-
ments using Andersen impactors
Grab sampling for onsite gas
chromatographic analysis of volatile
organics with boiling points in the
G! to C6 range
Grab sampling for laboratory gas
chromatographic analysis of N2O
The analysis protocol for SASS train
samples included:
Analyzing methylene chloride ex-
tracts of particulate and XAD-2 resin
for total organic content in two boil-
ing point ranges: semivolatile
organics with boiling points between
100 and 300°C (nominally C7 to C16
organics) by total chromatographable
organic (TCO) analysis, and non-
volatile organics with boiling points
-------
greater than 300 °C (nominally C16+
organics) by gravimetry
Obtaining infrared (IR) spectra of the
gravimetric residue of all extract
samples
Analyzing all extract samples for 58
semivolatile organic priority
pollutants, including 16 polynuclear
aromatic hydrocarbon (PAH) species
by gas chromatography/mass spec-
trometry (GC/MS) according to EPA
Method 625, with further identifica-
tion and quantitation of major peaks
in the GC chromatogram
Analyzing selected extracts for
general compound category com-
position by direct insertion probe low
resolution mass spectrometry
(LRMS)
Performance/Emission Tests
Results
Results of the limited emission tests on
the conventional-burner-equipped steamer
are summarized in Figure 2 which shows
stack NOX emissions versus flue gas 02
at the two steamer loads. The figure in-
dicates a steady decrease in NOX emis-
sions as excess air is reduced until flue gas
O2 falls below about 3 percent. Below
this 02 level, the rate of NOX emissions
decrease increases. However, smoke emis-
sions (Bacharach smoke number) also in-
creased rapidly at flue gas O2 below
about 3.5 percent. For practical (accept-
able CO and smoke emissions) operation,
NOX emissions of 300 ppm (corrected to
3 percent 02) at flue gas 02 of 3.7 per-
cent were attainable at full load. At 75
percent load, NOX emissions are reduced
to about 250 ppm (3 percent 02) at flue
gas 02 of 4.0 percent.
Results of the low-NOx burner detailed
performance evaluation tests are sum-
marized in Table 1. Emission levels
measured at the furnace outlet (by host
site personnel) and at the stack
downstream of the economizer section are
shown. As noted, NOX emissions from
the unit varied from 95 to 180 ppm (cor-
rected to 3 percent O2) with changes in
the parameters investigated. Certain con-
ditions resulted in NOX emissions at the
stack below 100 ppm (3 percent O2, dry)
but these were, in general, accompanied
by high CO emissions and high smoke
spot. Conditions which resulted in NOX of
about 110 ppm with moderate CO and
smoke might be considered as reflecting
burner capabilities for realistic NOX
reductions.
Figures 2 and 3 illustrate the
dependence of NOX and CO at the stack
on flue gas O2. Both Iow-N0x and con-
ventional burner test data are shown.
Figure 2 shows that the NOX emissions
from the Iow-N0x burner were 45 to 65
percent lower than from the conventional
burner at a given flue gas 02. Figure 3
shows that CO emissions from the low-
NOx burner increased steeply at flue gas
O2 below 2.5 to 3.0 percent. This con-
trasts with conventional burner behavior
where CO emissions were still low at flue
gas O2 down to 2.5 percent. These
higher CO levels from the low-NOx burner,
which were accompanied by high smoke
spots (see Table 1), are attributed to flame
impingement which was observed at vir-
tually all burner settings. Higher CO levels
are attributed to increased flame impinge-
ment and excessively low diffusion zone
stoichiometries during low 02 and high
OFA tests.
The effect of OFA flowrate on both CO
and NOX levels is illustrated in Figure 4.
CO levels decrease sharply at OFA rates
below 10 percent. At 3 percent OFA, CO
levels are nearly those of the conventional
burner (see Figure 3). NOX emissions at
minimum OFA, however, are not
significantly higher than those at high OFA
rates.
The effect of the flue gas recirculation
(FGR) rate on NOX and CO emissions is
shown in Figure 5. FGR had a greater ef-
fect at higher O2 and lower OFA levels
(4.2 percent O2 and 8 percent OFA) than
it did at lower O2 and higher OFA levels
(2.6 percent O2 and 19 percent OFA).
That CO responded in an opposite man-
ner can be explained in part by the great-
er mixing occurring at higher burner
stoichiometries combined with lower FGR
rates. This mixing tended to partly cancel
the low-NOx properties of the split frame.
Conversely, that higher FGR rates com-
bined with lower burner stoichiometry,
while keeping the flames separate, caused
greater impingement of the premix flame,
which increased the CO levels.
Comprehensive Emission Test
Results
Following these performance tests,
operating conditions were selected for
comprehensive emission testing using the
SASS train and other aspects of the
sampling protocol noted above. This test
was conducted at 9.5 percent FGR, 10
percent OFA, about 54 percent premix air,
and 36 percent diffusion air (test No. 21
in Table 1).
Table 2 summarizes the gaseous and
particulate emission levels measured dur-
ing these tests. Continuous monitor
measurements at both the stack and the
furnace outlet (provided by host site per-
sonnel) are shown. As indicated, stack
NOX and CO emissions averaged 106 ppm
and 93 ppm, respectively, with smoke
Q.
I
i"
350
300
250
200
750
700-
50
0§
O Full Load MHI Burner Tests
Full Load Conventional Burner
75% Load Conventional Burner
MHI FD Fan Limit
2345
Oa, percent (Dry!
Figure 2. /V0« emissions versus flue gas Oa.
-------
Table 1. MHI Burner Performance Test Results
Air Distribution Fuel Rate
Heat Input
Stack*
Furnace outlet
Test
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21e
FOR
Rate
8.9
9.4
13.4
9.9
9.9
9.4
8.8
9.1
9.8
9.3
8.8
8.2
8.4
8.8
6.6
2.6
8.4
9.1
9.8
9.8
9.5
Premix3
OFA Flame Air
(%l 1%)
19
18
19
19
13
15
15
11
9
8
8
8
8
8
8
8
7
6
7
3
10
48
52
52
53
52
51
51
56
57
57
57
58
55
55
55
55
59
58
58
62
54
Diffusion8
Flame Air
f%> ll/s)
33
30
29
29
35
34
34
33
34
35
35
34
37
37
37
37
34
36
35
35
36
0.383
0.385
0.381
0.390
0.379
0.377
0.388
0.386
0.386
0.388
0.386
0.396
0.388
0.390
0.390
0.390
0.405
0.377
0.386
0.388
0.386
(BPDj
208
209
207
212
206
205
211
210
210
211
210
215
211
212
212
212
220
205
210
211
210
IMWt
16.1
16.2
16.0
16.4
16.0
15.9
16.3
16.3
16.3
16.3
16.3
16.7
16.3
16.4
16.4
16.4
17.0
15.9
16.3
16.3
16.3
(Million
Btu/hrl
55.0
55.2
54.7
56.0
54.5
54.2
55.8
55.5
55.5
55.8
55.5
56.8
55.8
56.0
56.0
56.0
58.2
54.2
55.5
55.8
55.5
02
(%)
3.5
2.7
2.6
2.5
2.6
3.4
3.8
3.2
2.2
3.1
4.1
4.2
4.2
3.4
4.2
4.2
4.2
3.6
2.8
2.8
3.0
C02
13.4
13.9
13.9
14.0
14.0
13.9
13.3
13.2
14.6
13.3
12.5
12.5
12.4
13.0
12.3
12.3
12.6
12.9
13.5
13.6
13.3
CO
(ppmf1
99
266
215
236
269
60
60
51
79
141
70
64
51
85
66
54
60
62
80
64
93
NOX
(ppmf1
119
1O2
99
95
97
119
140
145
110
111
145
180
126
111
131
152
143
116
106
133
1O6
TUHC
(ppmf1
3.2
4.4
3.2
5.3
3.9
2.2
1.4
1.1
1.0
4.5
8.5
8.6
1.8
1.1
1.4
1.1
1.4
0.5
0
0
0
Smoke
No.
>10
10
9.5
10
8
6
3.5
3.5
a
10
6
3.5
4
8
6
2.5
4
6
a
6
8
02
4.3
3.3
3.1
3.1
3.O
4.O
4.4
3.3
2.3
2.2
3.2
4.2
4.6
3.6
4.5
4.6
4.1
3.3
2.3
2.5
2.5
CO
(ppmf1
70
80
80
73
76
59
44
45
58
104
54
46
55
67
58
53
50
64
87
55
68
C02
12.9
13.9
13.9
14.0
13.9
13.0
12.9
13.9
14.5
14.7
13.8
12.9
12.5
13.5
12.7
12.6
13.1
13.8
14.5
14.5
14.5
NOX
(ppmf1
124
113
108
109
105
126
140
144
112
102
126
174
125
114
131
149
144
113
98
126
108
S02
tppml"
597
594
582
592
576
573
581
568
585
616
595
570
587
573
572
582
556
558
574
583
586
aPremix and diffusion nozzle combustion air flows were not measured. Values shown here were estimated based on blower discharge
pressure and static pressure readings in the windbox for diffusion and premix zones.
bEmission measurements by Acurex.
cEmission measurements by Getty Oil Company.
dDry at 3 percent O2.
e Using SASS train.
spot of 8 and flue gas O2 of 3.0 percent.
S02 and S03 emissions were 594 and 3.1
ppm, respectively, by Method 8; the S02
result agrees well with the furnace outlet
continuous monitor. Method 5 paniculate
load was 39 mg/dscm. The Andersen
particle-sizing measurements indicated
that mean (50 percent cut point) particle
size was between 3 and 4 /^m (two runs).
Table 3 summarizes measured organic
emissions from the low-NOx burner-
equipped steamer by organic boiling point
range. The organic emissions are
dominated by the volatile (C1 to C6) frac-
tion, which is composed primarily of com-
pounds in the C3 and C4 boiling range. No
semivolatile organics were detected.
However, nonvolatile organics (nominally
C16+) were found in the particulate,
though not in the sorbent module. This
confirms the high smoke emissions for the
tests and suggests that soot was forming.
Table 4 summarizes the PAH and other
semivolatile organic priority pollutant
species identified by GC/MS analysis. Also
noted in the table are levels of other
organic species identified and quantitated
in the GC/MS sample analyses. Of the
PAHs, only naphthalene, phenanthrene,
and pyrene were found, and these only in
280
240
o
£ 200
01
Q.
160
I
o'
o
120
80
40
- O
O Low-NO^ BurnerFull Load
Conventional BurnerFull Load
O» percent (Dry)
Figure 3. CO emissions versus flue gas O2.
4
-------
790
O 150
I 100
O
so
240
. 200
160
120
80
40
CO, ppm (Dry at 3 percent
O NO
A CO
FGR = 9.4 to 9.9 Percent
Diff = 29 to 35 Percent
Prem" - 52 to 62 Percent
02 = 2.5 to 3.0 Percent
a = Diffusion flame air
" = Premixed flame air
J_
_L
-L
14
16
Figure 4.
2 4 6 8 10 12
OF A Rate, percent
Effect of OFA rate on /VOX and CO emissions from the low-NOi burner.
18
20
the paniculate. The other species detected
are generally oxygenated aromatics and
fused-ring aromatics.
The flue gas N20 levels measured in
these tests, along with those measured in
other tests performed in this EPA project,
are summarized in Figure 6 which shows
N20 plotted versus corresponding NOX
emissions. It appears from Figure 6 that
N2O emissions are roughly proportional to
a unit's NOX emissions. A least squares fit
of the data, which were taken from a
range of combustion sources, suggests
that N20 emissions correspond to 22 per-
cent of NOX emissions, with a correlation
coefficient (r2) of 0.88. The curve in
Figure 6 represents this fit.
Results of several quality assurance
(QA) activities performed in these tests
are discussed in Volume I of the project
report. These results establish that the
data quality was of an acceptable level in
terms of the project's QA objectives.
190
O ISO
c
I
«
'00
i
50
Figure 5.
240
- 200
160
120
80
40
~ CO, ppm (Dry at 3 percent Oi)
OFA = 8 Percent
Diff = 37 Percent
Prerrf - 55 Percent
Oz = 4.2 Percent
OFA = 19 Percent
Diff = 29 Percent
Prerrf = 52 Percent
O2 = 2.6 Percent
O
* = Diffusion flame air
b = Premixed flame air
0
4
12
14
6 8 10
FGR Rate, percent
Effect of FGR rate on /VO, and CO emissions from the /ow-A/0, burner.
16
18
-------
Table 2. Flue Gas Emissions
Stack3
Furnace outlet*1
Pollutant
As measured:
O2, percent dry
CO2, percent dry
NOX, ppm dry
N2O, ppm dry
CO, ppm dry
TUHC, ppm dry
SO2, ppm dry
Continuous monitor
Method 8
SO? ppm
Method 8
Bacharach smoke number
Range
2.7 to 3.3
13.1 to 13.5
108 to 115
12.9 to
20.5°
45 to 135
<1
_d
_e
_e
8
Average
3.0
13.3
106
17.0
93
<1
_d
594
3.1
8
Range
2.4 to 2.7
14.4 to 14.5
110 to 112
_d
68 to 75
_d
550 to 610
_d
_d
_d
Average
2.5
14.4
111
_d
71
_d
600
_d
_d
_d
Corrected to 3% O2
NOX (as NO2)
N2O
CO
TUHC (as CH 4)
SO2
Continuous monitor
Method 8
SO3 (as W2SOV
Method 8
Paniculate
Method 5
SASS
Andersen
ppm
106
17
93
<1
_d
594
3.1
mg/dscm
39
118
573
ng/Jf
73.7
11
39
-------
Table 3. Total Organic Emissions Summary
Organic Category
mg/dscm
ng/J
Volatile organics analyzed in the
field by gas chromatography
C,
Ce
Total C,-C6
Semi volatile organics analyzed
by TOO
Filter
XAD-2
Total C7-Ct6
Nonvolatile organics analyzed
by gravimetry
Filter
XAD-2
Total C,6+
Total organics
0.2
0
8.4
2.2
0
0
10.8
<0.004
<0.004
0.3
0.3
11.1
0.07
0
3.0
0.80
0
0
3.9
-------
I
§
J
140
120
100
80
60
40
20
O
O
V This test
Q Coal/oil-mixture-fired industrial boiler
O Oil/refinery-gas-fired crude oil heater
Coal/water-slurry-fired industrial boiler
Oil/refinery-gas-fired industrial boiler
Coal/'plastic-water-fired commercial boiler
Coal-fired commercial boiler
Coal/water-slurry-fired industrial boiler
O EOR steamer equipped with the EPA low-NO* burner
I 1 I
100 200 300 400
NO* ppm (Dry at 3 percent Oy)
500
600
Figure 6. A/ZO emissions from combustion sources as a function of NO* emissions.
C. Castaldini, L. Water/and, andH. Lipsare withAcurexCorp., Mountain View, CA
94039.
Robert E. Hall is the EPA Project Officer (see below).
The complete report consists of two volumes, entitled "Environmental Assess-
ment of an Enhanced Oil Recovery Steam Generator Equipped with a Low-NO*
Burner:"
"Volume I. Technical Results." (Order No. PB 86-159 837/AS; Cost: $11.95)
"Volumell. Data Supplement,"(Order No. PB86-183 290/AS; Cost: $16.95)
The above documents will be available only from: (cost subject to change)
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
S. GOVERNMENT PRINTING OFFICE: 1986/646-116/20836
-------
5
3>
o
o
oo
o>
8
co
yo
5 3
Q) O
£8
CD
c
01
GO
8
> m
CO 3
------- |