United States
Environmental Protection
Agency
Industrial Environmental
Research Laboratory
Research Triangle Park NC 2771 1
Research and Development
EPA-600/S2-84-095 July 1984
Project Summary
Evaluation of the Efficiency of
Industrial Flares: Test Results
J.H. Pohl, R. Payne, and J. Lee
The report gives results of Phases 3
and 4 of a four-phase research program
to quantify emissions from, and effi-
ciencies of, industrial flares. Phase 1
involved the experimental design;
Phase 2, the design of the test facilities;
Phase 3, development of the test
facilities; and Phase 4, data collection
and analysis. (NOTE: Report EPA-
600/2-83-070 gives results of Phases
1 and 2.)
The combustion efficiency of large
pilot-scale flares was measured. The
flame structure and combustion effi-
ciencies were correlated with operating
conditions of the flare, size of the flare
head, and properties of the flared gases.
The combustion efficiency was corre-
lated with the ratio of heating value of
the gas flared to the heating value
• required to maintain a stable flame, and
was independent of the flame head size.
In turn, the heating value required to
maintain a stable flame was correlated
with the reciprocal of an estimated
flame temperature based on properties
of the flared gas. The length of the
flame, entrainment into the flame, and
liftoff distances were also correlated,
using combinations of the Richardson
Number, jet theory, and properties of
the flared gas.
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 docu-
mented in a separate report of the same
title (see Project Report ordering
information at back).
Test Facility
EPA's Flare Test Facility (FTF) was
constructed at Energy and Environmental
Research Corporation's (EER'S) El Toro
test site. The FTF (Figure 1) includes a pad
and structure for installing and testing
flare heads, screens to shield the flame
from the wind, parallel delivery systems
to accurately meter the wide range of gas
flows to the. flare, a hood to sample the
entire plume, a movable rake probe to
simultaneously sample five radial posi-
tions, high-speed movie and photographic
equipment to record the structure of the
flare flame, and a room from which to
control the flare and analyze gas samples.
Techniques were developed to operate,
sample, analyze, and reduce the data.
Analysis included visual and photographic
observation of the flare flame structure,
and samples of soot, 02, CO, COz, total
hydrocarbon, and SO2, which was used as
a tracer. The data were corrected for the
measured background of combustion
species and for dilution of the flare plume
by ambient air. Dilution and local Com-
bustion efficiencies were calculated at
each probe position, and the maximum
potential error in the dilution and
combustion efficiencies were estimated
for each data point. The local combustion
efficiencies were integrated, using
velocity profiles estimated by jet theory to
yield a global combustion efficiency for
each flare flame.
The combustion efficiencies were
measured for a wide range of operating
conditions typical of commercial flares:
• Head type—
- 3-in.* EER prototype.
- 6-in. EER prototype.
- 12-in. EER prototype.
- 3-in. EER prototype with conver-
gent ring.
"To convert to metric equivalents, please use the
{actors at the end of this Summary.
-------
Induction
Fan
Analyzers
CO. C02
HC. SO2
(
Figure 1. EPA's flare test facility at EER.
- 3-in. EER prototype with divergent
ring.
- 12-in. commercial (three manu-
facturers—k, B, and C).
• Gas—
- Propane/nitrogen mixtures.
- Natural gas.
• Heating value of the flared gas —
270 - 2350 Btu/ft3.
• Flow rates—
- Velocities, from 0.2 - 428 ft/sec.
- Reynolds Numbers, from 337 to
217,000.
- Richardson Numbers, from 2.9 x
10~5to8x 102.
Results
Table 1 summarizes combustion effi-
ciencies. In addition to combustion
efficiencies, other calculations were
made:
• Combustion intensity was found to
be 90,000 Btu/hr/ft3, independent
of flare or flame conditions.
• The flame length was correlated with
the Richardson number.
• The liftoff distance was correlated
with ratios of velocities and concen-
trations of combustible gas.
Table 1.
Combustion Efficiencies of Various Flare Heads
Purpose
of Test
No Retaining
Ring
700% CsHa
77%C3«e
56% CaM,
5O%CiHe
Stable Flame
Limit
Low-Btu
Effect of
Steam
Steam,
Smokeless
City Gas
3-in.
EER
—
--
--
98.37-
98.95
-
—
90. 19-
99.92
98.94-
99.96
—
-
3-in.
EER
Hi-Vel
95.11-
99.66
99.74-
99.87
99.73-
99.88
97.27-
99.33
99.72-
99.87
99.81-
99.88
—
—
—
--
6-in.
EER
—
--
--
9847-
99.76
—
.-
92.24-
99.36
99.89-
99.92
.-
-
12-in.
EER
—
--
--
98.29-
99.50
-
--
94.89-
99.73
--
99.32
99.91
12-in.
Indus A
—
--
--
99. 12-
99.78
—
--
98.49
--
--
-
12-in.
Indus B
—
—
—
99.48-
99.65
—
—
99.21-
99.72
—
99.84
-
12-in.
Indus C
—
--
—
99.08
99.65
-
—
91.16
99.52
—
—
-
• The flame stability was correlated
with the reciprocal of an estimated
flame temperature.
• The combustion efficiency correlated
with a dimensionless heating value
of the gas fired.
The term "flame stability" means that
a flame is maintained; flame instability
occurs when the jet velocity exceeds the
flame velocity and the flame goes out.
Figure 2 shows the gas heating value
versus the gas exit velocity at the point of
instability (i.e., at the point where the
flame starts to "go out"). This point is
-------
JJUU
1000
900
^. 800
g
*
a 700
1
9>
•S 600
w
S 500
CD
400
300
200
O 3-ln
Flare Head
' 9 3-ln. Flare Head. Divergent Ring
• 3-ln
- A 6-ln.
Flare Head, Convergent Ring
Flare Head
O 12-ln. Flare Head
^S
. D 12-ln. Industrial Head C
tS 12-ln. Industrial Head A
0 12-ln. Industrial Head B
-
-
O
-
!
/
'I
4
I
3
S.
/
,f1
1
M
,
.
/
4
'
S
t
/
/
/
A
IT
i***
1 1
i
4
/
4
m
/
9
i
/
,'
f
1
J
'
S
/
.
^
1 L
*
f
{j
^ r
/ i
r 1
\j
f*
/
-
K»i
lolo-
•^qp" -
, g^
V
"'C/ = Confidence Interval.
'"Based on flares burning propane/ ~
nitrogen mixture with no pflot
flame.
lr v v
1.0
1O.O
Exit Velocity,
i 1
100.0 1000.0
ft/sec
FiguH 2. Region of flame instability. (bl
determined by establishing a propane/
nitrogen flame at a given velocity and
then decreasing the flow of propane until
the flame goes out. The combustion
efficiency is then measured at the
conditions just prior to flameout.
The region shown in Figure 2 indicates
the minimum gas heating value required
to produce a stable flame at the gas exit
velocity within the 95 percent confidence
limits of the mean. For any given velocity,
a gas with a heating value above this
region produces a stable flame; a gas with
a heating value below this region pro-
duces an unstable flame. Velocity/gas
combinations in or below the region tend
to produce flames with lower combustion
efficiency. Thus, for any given test velo-
city, the minimum gas heating value for a
stable flame can be determined. By
dividing the actual gas heating value by
the minimum value required for stability,
a ratio is obtained which is greater than 1
for stable flames, and less than 1 for un-
stable flames. Figure 3 plots combustion
efficiencies versus this ratio and shows
that high combustion efficiencies are
ach ieved when the ratio exceeds 1. When
the ratio is 1 or less, lower combustion
efficiencies are often obtained. Note that
even at a ratio less than 1, high combus-
tion efficiencies are sometimes achieved,
demonstrating the uncertainty associa-
ted with the stability measurements. In
general, however, stable flames are effi-
cient, and unstable flames can be ineffi-
cient. Flames near the stability limit are
very sensitive to perturbations and, when
perturbed, can easily produce high emis-
sions of unburned material.
All conclusions are based on the data of
this study and are limited to head geo-
metries, gases, and variables examined.
Head geometries were limited to:
• Simple pipe flare of 3-, 6-, and 12-in.
in diameter.
• Three commercial 12-in. flare heads
of different design and manufacture.
The gases studied were limited to:
• Propane/nitrogen mixtures with
heating values of 270 - 2350 Btu/ft3.
• One test with natural gas.
The variables examined were:
• Velocities from 0.2 to 420 ft/sec.
• Reynolds numbers from 340-217,000.
• Richardson numbers from 2.9 x 10"5
to 8 x 102.
• Steam flow from 0 to 1 Ib steam/lb
fuel.
The flare flames were shielded from the
wind, and combustion efficiencies were
not measured in winds greater than 5
mph.
The following conclusions are based on
study results:
• Flares operating with unstable flames
can have low combustion efficiency.
• Combustion efficiency does not
depend on flare size or geometry.
• Successful correlations were devel-
oped for flare flames:
- Flame length was correlated with
a modified Richardson number.
/UL/
98
•g
S 96
I
1 94
!§
S
c a-,
.0 yz
1
a 90
88
86
&SSVP O ctt^^wKcxft o ' ' ' '
c~
-
^
-
$%v A) K^(O'| Q
%
£
o
A
L
Q
^
L
-
ow
O 3-ln. EER
A 6-ln. EER
O 72-/n. f£/?
* k 72-//J. Commercial A
0 72-/n. Commercial B
D /2-/n. Commercial C
-
^a' Reason for low combustion efficiency is unknown.
' 'Based on flares burning propane/nitrogen mixture
with no pilot flame.
1 1 1 . 1 1 1 1
Figure 3.
12345678
Heating Value/ Minimum Heating Value for Stability""
Combustion efficiency near the lower limits of flame stability.
-------
- Liftoff distances were correlated
with ratios of velocities and con-
centrations.
- Flame stability was correlated with
a pseudo flame temperature.
- Entrainment was correlated with
ratios of distances and the Richard-
son number or with velocity.
• Combustion efficiency was high for
flares with high velocities, provided
the heating value of the gas was in
the region of stability.
• Steam injection completely suppressed
soot production but did not apprecia-
bly alter combustion efficiency un-
less the flame was oversteamed
(>0.5 Ib steam/lb fuel), and then
combustion efficiency decreased.
Conversion Factors
To convert nonmetric units used in this
Summary to their metric equivalents,
please use the following factors:
Nonmetric Multiplied by Yields Metric
Btu
ft
ft3
in.
Ib
mi
1.055
0.305
0.028
2.54
0.454
1.609
kJ
m
m3
cm
kg
km
J. H. Pohl. R. Payne, and J. Lee are with Energy and Environmental Research
Corporation, Irvine. CA 92714.
Bruce A. Tichenor is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of the Efficiency of Industrial Flares:
Test Results," {Order No. PB 84-199 371; Cost: $ 17.50, 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
-tl U S GOVERNMENT PRINTING OFFICE; 1984 — 759-015/7742
\
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
PS
u
I* O I ID PKU'f"^'*
SJSwaa-
------- |