Applied Energy 178 (2016) 19–28

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Effects of oxidizer flexibility and bluff-body blockage ratio on

flammability limits of diffusion flames
Islam A. Ramadan a, Abdelmaged H. Ibrahim a, Tharwat W. Abou-Arab a, Sherif S. Rashwan a,b,
Medhat A. Nemitallah b,c,⇑, Mohamed A. Habib b
a
Mechanical Power Department, Faculty of Engineering, Cairo University, Cairo 12613, Egypt
b
KACST TIC on CCS and Mechanical Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia
c
Mechanical Engineering Department, Faculty of Engineering, Alexandria University, Alexandria 21544, Egypt

h i g h l i g h t s

Experiments on air, O2-enriched-air and oxy-flames anchored over bluff body.

Stability limits, visual appearance and extinction limits of these flames are quantified.

Effect of blockage ratio on flammability limits are presented for the three oxidizers.

Emissions are characterized for the three different flames.

Experiments were performed considering wide ranges of operating parameters.

a r t i c l e i n f o a b s t r a c t

Article history: Concerns about global warming have encouraged the interest in hydrocarbon combustion techniques
Received 14 March 2016 that allow easy capture of carbon dioxide. One technique for achieving this objective is through the
Received in revised form 13 April 2016 use of pure oxygen instead of air for combustion or what is called oxy-combustion carbon capture tech-
Accepted 8 June 2016
nology. The main goal of the manuscript is to study flammability limits, visual flame appearance and
exhaust emissions of diffusion flame stabilized over a bluff body over ranges of operating and design con-
ditions. The operating conditions include flow Reynolds number, equivalence ratio and oxidizer compo-
Keywords:
sition. The design parameter considers the change of blockage ratio (BR) of the bluff body namely,
Air combustion
Oxygen-enriched combustion
BR = 0.36, 0.5, 0.67 and 0.82. Based on this, three sets of experiments were performed utilizing com-
Oxy-combustion pressed natural gas (CNG) as a fuel to be burned with three different oxidizers including air, oxygen-
Flammability limits enriched-air and oxy-fuel mixtures (O2 plus CO2 with a controlled oxygen fractions, OF). The three sets
Diffusion flames of experiments were performed to identify ranges for stable flame operation considering different oxidiz-
Bluff-body ers under different operating conditions. Stability limits, visual flame appearance and extinction limits of
these flames are quantified and analyzed. Furthermore, three different regions were observed; precisely,
jet flames, central-jet dominated flames and recirculation zone flames, depending on the ratio between
oxidizer and fuel momentum. The flame color changed from yellow for air combustion, to bright white
for oxygen-enriched-air combustion and finally to blue with yellow tips for oxy-combustion. The flame
length was the highest for air combustion, then lower for oxy-combustion and the lowest for oxygen-
enriched-air combustion. This was attributed to the effect of oxygen-enrichment which results in
increase in the flame speed making flame length shorter. For the sake of comparison, the flammability
limits of the three sets were reported and the results revealed that oxygen-enriched-air-flames have
higher stability than air-flames and oxy-flames, respectively.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction

⇑ Corresponding author at: KACST TIC on CCS and Mechanical Engineering The use of oxy-combustion technology for CO2 capture is a
Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, promising technique to reduce CO2 emissions. In this technique,
Saudi Arabia.
pure oxygen can be used as an oxidizer rather than air to avoid
E-mail addresses: [email protected], [email protected] (M.A. Nemi-
tallah). nitrogen in the combustion chamber [1]. This technique enables

http://dx.doi.org/10.1016/j.apenergy.2016.06.009
0306-2619/Ó 2016 Elsevier Ltd. All rights reserved.
20 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28

Nomenclature

CH4 methane BR blockage ratio, the ration between the outer diameter of
CNG compressed natural gas the bluff body and the outer diameter of the burner
CO carbon monoxide body
CO2 carbon dioxide D4 confinement diameter
HC hydrocarbons O2 oxygen
D1 burner inner diameter OF oxygen fraction, percentage of oxygen in the oxidizer
D2 burner outer diameter mixture of O2 and CO2
D3 bluff body outer diameter Ro confinement outer radius

the exhaust stream to be separated into concentrated CO2 and applications impose limitations on CO and O2 levels [11]. In the
water vapor. The water vapor can be removed by a simple conden- context of CO emissions, Oh et al. [12] studied the effect of CO
sation process leaving a high concentration CO2 stream that can be addition on the flame characteristics of a non-premixed (diffusion)
easily captured. However, burning pure oxygen is unpractical due oxy-methane flames in a lab scale furnace. This experimental study
to the resulting extremely high flame temperature [2]. For this rea- showed that, as the CO concentration decreased, the more the
son, the oxygen is normally diluted with a steam of CO2 in order to flame is stable, the flame length becomes higher as CO is diluted.
control the flame temperature. One of the advantages of this com- Yewn et al. [13] compared the structure of three diffusion flames
bustion technique is the expected significant reduction in thermal utilizing natural gas (NG) as a fuel with three different oxidizers
NOx emissions. Fuel NOx, however, may not decrease if the fuel including oxidizer A (air, 21 vol.% O2 and 79 vol.% N2), oxidizer B
has significant nitrogen content [3]. This technique has been pro- (28 vol.% O2 and 72 vol.% CO2) and oxidizer C (enriched air,
posed for slight modifications of existing gas turbine combined 28 vol.% O2 and 72 vol.% N2). The study showed the variation of
cycles [4] or completely new concepts [5,6]. A review by Yin and centerline temperature with distance from the burner tip for all
Yan [7] was conducted on the fundamentals and modelling of the oxidizers. The results revealed that the temperature of enriched
oxy-fuel combustion of pulverized fuels. The review illustrates air flame is higher than other flames. Also, it was indicated that
the promising importance of the pulverized fuels for CO2 capture the concentrations of NO and CO in O2/CO2 flame are lower than
from power plants and discussing the concerns, developments other flames. Nemitallah et al. [14–16] studied experimentally
and recent studies in the past years. Tan et al. [8] examined an and numerically diffusion oxy-combustion flames in different
oxy-fuel circulating fluidized bed combustion as an alternative combustors. They reported that the stability of oxy-fuel combus-
technology to oxy-fuel pulverized combustion that has many tion flames is adversely affected when the operating oxygen frac-
advantages that can exceed the prediction of the normal oxy-fuel tion reduced below 25%. Ramadan et al. [17] studied the stability
pulverized technology for CO2 capture. They reported results of of CNG/CO2/O2 flames and compared the results with CNG/air
stable oxy-fuel combustion up to 90% of CO2 introduced in the and CNG/air/O2 flames anchored over a bluff body under atmo-
combustion process on a dry basis. Hu and Yan [9] investigated spheric conditions. They observed no lower limits of the flamma-
the characterization of flue gas in oxy-coal combustion process bility due to the use of non-premixed flames. They quantified
for CO2 capture. The economic effective and technical viability and explained the dependence of the upper flammability limit on
for CO2 capture process by using characterization of the flue gas oxygen fraction, fuel and oxidizer momenta and bluff body block-
recirculation, FGR, was conducted for an oxy-coal combustion pro- age ratio. The characteristics of oxy-oil combustion in an existing
cess. The study showed that increasing the O2 (oxidizer) from 20% furnace were experimentally investigated by Chi and Lin [18]. This
to 35% by Volume results in reducing the FGR rate by approxi- experimental study examined the operability of an existing
mately 58%. Furthermore, the FGR rate changes linearly with the 300 kW air–oil furnace to be adapted with oxy-oil combustion.
oxygen concentration. The results revealed some useful references They investigated the effect of oxygen enrichment in the range
for existing power plants efficient optimization and for future from 21% (air) up to 100% while using the FGR system in the retro-
designed power plants. In a further numerical study conducted fitted furnace. They claimed that the existence of air leakage to the
by Hu et al. [10], the heat transfer characteristics in utility boiler systems would decrease the CO2 concentration in the flue gases.
of oxy-coal combustion was investigated. They highlighted that The study reported that the CO2 concentration increases as the
the O2 concentration of 33% in oxy-coal combustion acts in a sim- oxygen fraction increases from 13% to 34.4% at the atmospheric
ilar manner to the air–coal combustion cases in terms of achieving pressure and from 14.7% to 61.1% for the pressurized combustion.
the same highest flame temperature and total heat transfer rate The shifting from air to oxy-fuel did not bring adverse effect to the
through boiler walls. In addition, it was shown that the increase flame stability.
of the moisture content in flue gas has a small impact on the flame A recent experimental investigation by Rashwan et al. [19–21]
temperature. studied the combined effect of flame premixing and oxy-
The current work investigates the combustion characteristics of combustion on flame stability and compared the results with the
confined diffusion CNG flames under different oxy-combustion cases of normal air combustion. They reported a range of oxygen
conditions (different mixtures of CO2 and O2) and compares the fraction, from 29% to 40%, over which the flames are stable. Also,
results with the cases of combustion using other oxidizers includ- they reported that the air–fuel combustion has wider range of
ing pure air and oxygen-enriched-air. The work covers flame flame stability than oxy-fuel combustion due to the adverse effect
appearance and stability limits. It should be noted that CH4/O2/ of introducing the CO2 into the combustion process. Furthermore,
CO2 (oxy-combustion) flames are characterized by slower chemical they claimed that increasing the oxygen fraction compensates
kinetics than methane–air flames. Equilibrium CO emissions in CO2 the negative effect of CO2 until the burner could work similar to
diluted systems are much higher than in corresponding air sys- air–fuel combustion at oxygen fraction of 42%. Ditaranto and Hals
tems. In addition, pipeline specifications in carbon sequestration [22] revealed from their experiment that oxy-combustion requires
 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28 21

at least 30% of oxygen to perform in a comparable matter with flow rates. The gas pressure and temperature are measured
respect to air–fuel combustion. Liu et al. [23] studied numerically downstream of each rotameter using calibrated pressure gauges
the characteristics of oxy-methane combustion in a gas turbine and thermocouples. The O2 and CO2 streams are fed into a mixer
model combustor. They concluded that higher oxygen fraction in having L/d ratio of 200 and designed to ensure fully mixing
the oxidizer mixture results in higher flammability limits. Joo between the O2 and CO2 streams. Then, the mixed O2/CO2 and
et al. [24] compared structures and soot concentrations of laminar CNG streams are fed into the burner.
methane–oxygen and methane–air diffusion flames over a range of The setup allows for the use of different oxidizers including air,
operating pressure up to 60 bar. It was shown that the soot oxygen-enriched-air, and oxy-mixture with different O2/CO2 ratios.
concentrations of air–methane flames are higher than those of oxy- The burner used is a concentric-flow, bluff-body stabilized, gas
gen–methane flames over the whole operating pressure range. Fur- fired burner as shown in Fig. 1 in the right side. The burner consists
thermore, Guo-neng et al. [25] quantified and analyzed the of two concentric tubes, the inner-thin tube supplies the fuel and is
structure of different diffusion flames utilizing CH4 as a fuel and made of stainless steel with an inner diameter, D1, of 5 mm and a
two different oxidizers (O2/CO2 oxidizer and O2/N2 oxidizer) with thickness of 0.5 mm and ended with a nozzle of 2.5 mm in diame-
different oxygen concentration (from 25% to 45% by volume). The ter. Oxidizer is supplied between the inner and the outer tubes. The
results of their experiments showed that the temperatures of O2/ outer tube is made of stainless steel with an inner diameter (bur-
CO2 flames are higher than those of O2/N2 flames and the nitric ner outer diameter), D2, of 30 mm and a thickness of 1 mm. A per-
oxide concentrations in the O2/CO2 flames are too small comparing forated plate is mounted between the two concentric tubes to
to O2/N2 flames. Nevertheless, Kutne et al. [26] studied the effects ensure the concentricity of the two tubes and the uniformity of
of oxygen fraction on oxy-fuel combustion characteristics. They the oxidizers flow. A bluff body is used to create a re-circulation
reported that operation with oxygen fractions lower than 22% zone to enhance the flame stability. In this work, the bluff body
was not possible even at stoichiometric conditions because of the is made of a circular disk mounted on the outer surface of the fuel
adverse effect of adding CO2 to the combustion chamber. A com- nozzle. Different bluff bodies were used with the same inner diam-
parative study between air–coal and oxy-coal combustion is con- eter (6 mm) and different bluff-body outer diameters, D3, which is
ducted by Hu et al. [27], studying the effect of flue gas corresponding to each blockage ratios (BR), (D3/D2)2 of 0.36, 0.5,
recirculation and examined several configuration methods for 0.66 and 0.82 respectively.
emission removal. The study showed that the oxy-coal combustion In order to operate in a confined environment, a steel cylinder
could replace the air–coal combustion and enhance the radiative is installed around the burner, with an inner diameter, D4, of
heat transfer. In oxy-coal combustion, the amount of flue gases 150 mm and a length of 500 mm. The confinement ratio (D4/
that needs to be treated is reduced by 40% as compared to air–coal D2)2 is 25, which eliminates the wall effects [28]. A rectangular
combustion, which will reduce the flue gas treatment equipment. sight quartz glass (50 mm  500 mm) was mounted and sealed
The emission results shows a reduction in NO and SO2 emissions. on the outer surface of the confinement to allow capturing digital
Also they indicated that the dew point of flue gas in oxy-coal com- images of the flames. An exhaust section with a conical end was
bustion is higher than its corresponding value in air–coal combus- installed above the confinement to minimize the effects of free-
tion due to the higher moisture content. air entrainment on the flame characteristics. A pilot LPG flame
As per the open-relevant literature and the above review on the burner was used for initiating the flame. The pilot flame consists
oxy-fuel combustion technique, it is concluded that among the of two concentric tubes, where the LPG fuel passes through the
several technologies used for carbon capture and storage (CCS), inner tube and air passes through the outer tube. The pilot flame
oxy-combustion technique is one of the recent technologies in car- was turned off after a stable CNG diffusion flame was achieved
bon capture that eliminates the CO2 and NOx emissions but it faces and before measuring the data. Before start up, the combustor
many challenges like flame stability. As a matter of fact, CNG/O2/ is purged with air from the compressor and, then, the pilot flame
CO2 flames are known to be characterized by slower chemical is lighted up. The CNG fuel is, then, fed into the burner and the
kinetics than CNG/air flames. As a result, flame stability and flame oxidizer is introduced. After a stable flame is achieved, the pilot
length are more problematic and require quantification and com- flame is shut down. The measurements were taken after a steady
parison with the conventional cases that use air as an oxidizer. This state period of approximately five minutes. The visual flame
work focuses on the oxy-fuel combustion technology, where the appearance under different conditions is documented using a dig-
oxidizer is a mixture of O2 and CO2 and comparison is made while ital camera (Canon EOS D1100, 14 Mega pixel resolution, 1/8-s
using different oxidizers. The work studies the effects of blockage exposure time). The exhaust emission concentrations were mea-
ratio, oxygen fraction with three different oxidizers (air, oxygen- sured using QUINTOX (KM9106) commercial gas analyzer using
enriched-air and oxy-mixture of CO2/O2). Flame appearance, stabil- electrochemical emission sensors. The probe is made of stainless
ity, and emissions are reported over wide ranges of equivalence steel with a length of 300 mm and a diameter of 8 mm. The sam-
ratio, oxygen fractions and flow Reynolds number considering dif- pling process is continuous where the flue gas sample is taken by
ferent oxidizers. a built-in pump through the probe and passes it over different
sensors that provide electric output signals which vary according
to the values of the measured concentrations. The results are dis-
2. Experimental setup played on a LCD screen and, then, it can be printed out.

2.1. Specifications of the combustor 2.2. Emission measurements

The experimental setup used to conduct the present work is The emissions were measured using a gas analyzer which mea-
shown on the left side in Fig. 1. It consists of an air compressor, sures the carbon dioxide concentration (CO2) with infrared cells
CO2 bottles (25 kg at 70 bar), O2 bottles (7 m3 at 120 bar) and com- and also measures oxygen concentration (O2) with an electro-
mercial CNG bottles (85% CH4, 9% C2H6, 3% C3H8, 2.4% CO2 and 0.6% chemical cell. The emissions were represented as emissions index
N2 by volume at 200 bar). Each cylinder is equipped with a pres- that defined as the mass flow rate of measured emission divided by
sure regulator and a calibrated rotameter. A needle valve is the mass flow rate of fuel. The emission index is calculated from
inserted at the exit of each cylinder to precisely control the gas the following formula:
22 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28

Fig. 1. Schematic of the experimental setup (left) and a schematic of the bluff-body and the burner used (right).

!
gemission _
M W  P V exhaust
emission index ¼ C ppm  ð1Þ 0.06
kgfuel RT m_fuel
0.05 Extinct flame zone
Air momentum, N.m

where C ppm is the concentration of measured emission (ppm), M W is

the molecular weight of measured gas, P is the pressure of exhaust 0.04

at measurement point (kPa), R is the universal gas constant

0.03
(8314.4 J/kg K), T is the temperature of exhaust at measurement
_
point (K), V exhaust is the volumetric mass flow rate of exhaust (m3/ 0.02
s) and m_fuel is the mass flow rate of fuel (kg/s). The flame emissions Stable flame zone
were measured using the gas analyzer at a cross section far from 0.01
burner tip by a vertical distance of 500 mm. This section was
selected to minimize the stratification in the cross section and to 0
0.0E+00 1.0E-04 2.0E-04 3.0E-04 4.0E-04 5.0E-04
ensure that the reactions were ended at that vertical location. The
Fuel momentum, N.m
emissions were measured at different radial locations on the same
cross section to determine if there was any stratification over a BR=0.36 BR=0.5 Br=0.667 BR=0.82
range of equivalence ratio. The emission readings were taken when
Fig. 2. The effect of the blockage ratio on flammability limits in air–fuel combustion
the fluctuations of the measured parameters are disappeared. The mode.
sample was draught using an internal pump, where the sample
passes through a water trap to prevent water from entering to ana- disc diameters were 18 mm, 21.2 mm, 24.5 mm and 27.2 mm
lyzer, then passes through particulate filter to prevent soot from which were corresponding to blockage ratios, BR, of 0.36, 0.5,
entering to analyzer. Finally the sample enters the analyzer to be 0.66 and 0.82 respectively. The flame appearance studied in this
analyzed for providing the value of emission concentration. work was defined as the flame length and flame color. A digital
camera was used to capture photos for flames through the glass
window in the combustor. While running the experiments, in
2.3. Operating conditions
order to determine the flame length, a white mark with defined
length was fixed on the glass window to be the scale of captured
The objective of this study is to investigate the flammability
photos. The present study experimentally investigates flammabil-
limits, visual flame appearance and exhaust emissions of CNG
ity limits, flame appearance, stability, and emissions over a wide
flames with three different oxidizers in a concentric flow, bluff-
range of equivalence ratio (from 0.25 to 1.75), oxygen fractions
body burner. In order to accomplish this objective, the following
(from 29% to 36%) and flow Reynolds number (from 1100 to
experimental program was performed. The flammability limit
1900) in order to provide explanation and quantification of the sta-
studied in this work is defined as the fuel to oxidizer momentum
bility and visual flame lengths issues under different combustion
ratio at which flame extinction occurs, when the oxidizer flow rate
conditions for all oxidizers.
is gradually increased at a fixed fuel flow rate, blockage ratio, and
oxygen fraction. The flow momentum was calculated from the
mass flow rate multiplied by the mean flow velocity at the exit 3. Results and discussion
of the flow stream. The effect of changing blockage ratio, as the
main parameter representing the burner geometry, was investi- The results of this work are presented in three separate sections
gated through four replaceable discs of different diameters. The representing different combustion operating conditions. The first
 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28 23

section investigates air–fuel combustion (CNG/Air), the second

exhibits the oxygen-enriched air–fuel combustion (CNG/Air/O2)
and the third represents the oxy-fuel combustion (CNG/O2/CO2)
mode. The results of each section are illustrated in terms of
flammability limits, visual flame appearance and exhaust emis-
sions over ranges of operating conditions.

3.1. Air–fuel combustion (CNG/air)

This section presents the results of air–fuel (CNG/Air) fuel com-

bustion at different oxidizer (air) and fuel momentum values to
obtain the flammability limits, visual flame appearance and
exhaust emissions under different operating blockage ratios.

3.1.1. Flammability limits of CNG/air flames

In this set of experiments, the flammability limits experiments
were performed at four different bluff body-blockage ratios
(namely, 0.36, 0.50, 0.67 and 0.82). Fig. 2 shows the relation
between the air momentum values that are required to achieve
stable flame operation at different fuel-momentum values.
Flammability limits were identified by the following procedure;
the fuel flow rate was kept constant at a certain value and the air
Fig. 3. CNG/air flames at BR of 0.36 and fuel momentum of 1.15E5 N m for
flow rate was increased gradually until the upper flammability
different air momentum values, 0.00047, 0.00121 and 0.00423 N m, from left to limit occurs. While the lower flammability limit was obtained by
right. fixing the fuel flow rate and decreasing the air flow rate step by
step until the lower flammability limit occurs. The results revealed
for the same blockage ratio that increasing the fuel momentum

1800 200 1800 200

O2 O2
1600 BR=0.36 1600 BR= 0.5
emission index {g (O2 or CO2)/kgfuel}

emission index {g (O2 or CO2)/kgfuel}

CO2 CO2
emission index (gHC /kgfuel)

emission index (gHC /kgfuel)

1400 1400
HC 150 HC 150
1200 1200

1000 1000
100 100
800 800

600 600

400 50 50
400

200 200

0 0 0 0
0.6 0.7 0.8 0.9 1 1.1 1.2 0.6 0.7 0.8 0.9 1 1.1 1.2
Equivalence ratio ( ) Equivalence ratio ( )

1800 1800
200 BR=0.82 200
O2
1600 BR=0.67 1600
emission index {g (O2 or CO2)/kgfuel}

CO2
emission index {g (O2 or CO2)/kgfuel}

1400
emission index (gHC /kgfuel)

1400
emission index (gHC /kgfuel)

HC 150 150
1200 1200

1000 1000
100 100
800 800

600 600 O2
50 400 CO2 50
400
HC
200 200

0 0 0 0
0.6 0.7 0.8 0.9 1 1.1 1.2 0.6 0.7 0.8 0.9 1 1.1 1.2

Equivalence ratio ( ) Equivalence ratio ( )

Fig. 4. Emission distributions of CNG/air flames at a distance of 500 mm from burner tip and at the center of the combustor for different blockage ratios.
24 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28

0.03 0.03
BR = 0.36 BR = 0.5

Oxidizer moomentum (N.m)

Exnct flame zone

Oxidizer momentum (N.m)

0.025 Exnct flame zone 0.025

0.02 0.02 Unstable

Unstable
zone
zone

0.015 0.015
Stable flame zone Stable flame zone

0.01 0.01
20 21 22 23 24 25 20 21 22 23 24 25
Oxygen fraction (%vol.) Oxygen fraction (%vol.)

FM=0.0000115 N.m FM=0.0000471N.m FM=0.0000115 N.m FM=0.0000471N.m

0.03 BR = 0.67 0.03 BR = 0.82

Exnct flame zone

0.025
Oxidizer momentum (N.m)

0.025 Oxidizer momentum (N.m)

Unstable
0.02 Stable flame zone 0.02 Zone

0.015 0.015
Stable flame zone

0.01 0.01
20 21 22 23 24 25 20 21 22 23 24 25
Oxygen fraction (%vol.) Oxygen fraction (%vol.)
FM=.0000115N.m FM=0.0000115 N.m FM=0.0000471N.m

Fig. 5. Flame stability curves for different oxygen fractions, blockage ratios and fuel momentum.

resulted in wider stability limits. This may be attributed to the flame as shown in the left flame in all figures. At intermediate air
increase in the Reynolds number and associated improvement in to fuel momentum ratios, the fuel momentum still able to pene-
turbulence and mixing. The results also indicated that the opti- trate recirculation zone but part of fuel was held behind the recir-
mum value of blockage ratio for stable flame operation is 0.67. culation zone forming a central-jet dominated flame (intermediate
The reason can be explained in terms of the recirculation zone size flame). At high air to fuel momentum ratios, the fuel was held
and strength and fuel–air mixing. When the blockage ratio is too behind the recirculation zone forming a recirculation-zone flame
low, the recirculation zone is small so that the mixing between (right flame); similar results were observed by Kang and Yang [29].
the fuel and the oxidizer is weak, resulting in less stable flames.
When the blockage ratio is too high, the distance between the fuel 3.1.3. Exhaust emissions of CNG/air flames
and the oxidizer streams is increased which weakens the mixing Fig. 4 shows the emission distributions at a distance of 500 mm
between them, resulting in lower flame stability. The lower extinc- from burner tip and at the center of the combustor considering dif-
tion limits occurred at very low flow rate of air. ferent blockage ratios. In all plots, the emissions of CO2, O2 and HC
are presented as function of operating equivalence ratio. As shown
3.1.2. Visual flame length and appearance of CNG/air flames in the figure, the emission index of oxygen is decreased as the
The visual flame length and appearance are obtained by a pro- equivalence is increased. The emission index of unburned hydro-
fessional camera allowing the investigation of its dependence on carbons (HC) is increased as the equivalence is increased and its
the air and fuel momentum values. The shapes of the CNG/air value became zero at equivalence ratio of 0.9 due to excess air used
flames are presented in Fig. 3 at BR of 0.36. The shapes of the in combustion. The emission index of CO2 is almost constant
flames in each figure are recorded at fixed fuel momentum of because the emission index of CO2, theoretically, should be con-
1.15E5 N m for different air momentum values, 0.00047, stant for lean and stoichiometric combustion (U 6 1). The number
0.00121 and 0.00423 N m. As shown in the figure, the CNG/air of moles of CO2 depends on the completeness of combustion which
flames have a yellow color due to the burning of carbon atoms occurs in lean and stoichiometric combustion. To sum up, the
and the increase in the soot concentration associated with the dif- emissions of blockage ratios of 0.36, 0.5, and 0.667 were almost
fusion flames. At low air-to-fuel momentum ratios, the fuel jet the same. However, at blockage ratio of 0.82, the maximum value
penetrates into the inner recirculation zone and forms a like-jet of O2, HC, and CO2 emissions indexes are too high in comparison
 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28 25

with their values at other blockage ratios. This may be attributed to

the large distance between the fuel and air streams at high block-
age ratios resulting in low level of mixing between both streams
and, accordingly, high emissions of O2, HC, and CO2.

3.2. Oxygen-enriched air–fuel combustion (CNG/air/O2)

This section represents the results of the oxygen-enriched air fuel

combustion (CNG/air/O2) considering ranges of oxygen fraction and
blockage ratio. Oxygen fraction (OF) in this case is defined as the per-
centage of total volumetric flow rate of oxygen ðV O2 þ 0:21  V Air Þ
divided by the total volumetric flow rate of oxidizer ðV O2 þ V Air Þ.

3.2.1. Flammability limits of CNG/air/O2 flames

In this set of experiments, the flammability limits were investi-
gated for different values of blockage ratio including 0.36, 0.5, 0.67
and 0.82. Fig. 5 represents a relation between the oxidizer momen-
tum (air plus O2) and the oxygen fraction values for different block-
age ratios. The flammability limits were identified by the same
procedure of the first set of experiments by keeping the fuel flow
rate constant and gradually increasing/decreasing the oxidizer flow
rate until the upper/lower flammability limits are obtained. The
stability curves for four different blockage ratios (0.36, 0.50, 0.67
and 0.82) are presented in Fig. 5. For each disc, different oxygen
fractions with respect to air were tried (21%, 22%, 23% and 24%)
Fig. 6. CNG/air/O2 oxygen enriched air flames at BR of 0.36, fuel momentum of
and two fuel momentum were tried in each case (1.15E5 and
1.15E5 N m and OF of 24% for different oxidizer momentum values including 4.71E5 N m). The study revealed that for the same blockage ratio
0.00285, 0.0024 and 0.00446 N m, from left to right. and oxygen fraction, increasing the fuel momentum increases the
flame stability, as observed in the CNG/air combustion cases. This
can be attributed to the increase in the flow Reynolds number
and, accordingly, the improved mixing and turbulence within the
flame zone. Also, for the same blockage ratio, increasing the oxygen
fraction in the enriched air leads to an increase in flame stability.

1800 O2 200 1800 200

O2
1600 BR=0.36 CO2 1600 BR=0.5 CO2
emission index (gHC /kgfuel)
HC HC
emission index {g (O2 or

emission index {g (O2 or

1400 1400
emission index (gHC /kgfuel)

150 150
1200 1200
CO2)/kgfuel}

CO2)/kgfuel}

1000 1000
100 100
800 800
600 600
400 50 400 50

200 200
0 0 0 0
0.6 0.7 0.8 0.9 1 1.1 1.2 0.6 0.7 0.8 0.9 1 1.1 1.2
Equivalence ratio ( ) Equivalence ratio ( )

1800 200 1800 200

O2 O2
1600 BR=0.67 1600 Br=0.82 CO2
CO2
emission index (gHC /kgfuel)

HC
emission index (gHC /kgfuel)
emission index {g (O2 or

1400 1400
emission index {g (O2 or

HC 150 150
1200 1200
CO2)/kgfuel}
CO2)/kgfuel}

1000 1000
100 100
800 800

600 600
50 400 50
400
200 200

0 0 0 0
0.6 0.7 0.8 0.9 1 1.1 1.2 0.6 0.7 0.8 0.9 1 1.1 1.2
Equivalence ratio ( ) Equivalence ratio ( )

Fig. 7. Emission distributions of CNG/air/O2 flames with oxygen fraction of 21% at a distance of 500 mm from burner tip and at the center of the combustor for different
blockage ratios.
26 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28

0.035 0.035

Oxidizer momentum, N.m

Oxidizer momentum, N.m
0.03 0.03

0.025 Exnct flame zone 0.025

0.02 0.02

0.015 0.015

0.01 0.01
Stable flame zone Stable flame zone
0.005 0.005

0 0
0 0.0001 0.0002 0.0003 0.0004 0 5E-05 0.0001 0.00015 0.0002 0.00025
Fuel momentum (N.m) Fuel momentum (N.m)

BR=0.36 BR=0.5 BR=0.667 BR=0.82 BR=0.36 BR=0.5 BR=0.667 BR=0.82

Fig. 8. The stability curves for different blockage ratios and two oxygen fractions including 28% (left side) and 32% (right side).

0.035 0.035
Extinct Flame zone
0.030 Extinct Flame Zone 0.030
Oxidizer Momentum (N)

Oxidizer Momentum (N)

0.025 0.025
BR = 0.5
0.020 BR = 0.36 0.020

0.015 0.015

0.010 0.010
Stable Flame Zone Stable Flame Zone
0.005 0.005

0.000 0.000
0.00000 0.00020 0.00040 0.00060 0.00000 0.00020 0.00040 0.00060
Fuel Momentum (N) Fuel Momentum (N)
Air OF=28% OF=32% Air OF=28% OF=32%

0.050 Extinct Flame zone 0.040 Extinct Flame zone

0.035
Oxidizer Momentum (N)

0.040
BR = 0.67
Oxidizer Momentum (N)

0.030

0.030 0.025 BR =0.82

0.020
0.020
0.015

0.010
0.010 Stable Flame Zone Stable Flame Zone
0.005
0.000
0.000
0.00000 0.00020 0.00040
0.00000 0.00020 0.00040
Fuel Momentum (N) Fuel Momentum (N)
Air OF=28% OF=32% Air OF=28% OF=32%

Fig. 9. Comparison of flammability limits of air–fuel combustion and oxy-fuel combustion under different blockage ratios.

This can be attributed to the enhanced chemical kinetics and the unburned gases beside the flame and, accordingly, more stable
improved rates of reactions in the flame zone while increasing oxy- flame was obtained. Increasing or decreasing the size of the bluff
gen concentrations. For the same oxygen fraction and fuel momen- body around blockage ratio value of 0.67 resulted in modified
tum, the best flame stability was observed at a blockage ratio of recirculation zone size and unstable operation regime.
0.67 similar to what was obtained in the cases of CNG/air combus-
tion. In this particular blockage ratio, it was not possible within the 3.2.2. Visual flame length and appearance of CNG/air/O2 flames
ability of the setup to observe flame extinction for the fuel Fig. 6 represents the flame images captured at steady state con-
momentum of 4.71E5 N m. Also, the lower extinction limits are dition at blockage ratio of 0.36, oxygen fraction of 24% and fuel
represented in Fig. 5 for the case of BR = 0.67 as a line identically momentum of 1.15E5 N m. the comparison of the flame shapes
on x-axis, because the flame was extinct when the oxidizer flow is done for different oxidizer momentum values including
rate is too small at all fuel flow rate. This may be attributed to 0.00285, 0.0024 and 0.00446 N m. As shown in the figure, CNG/
the creation of stable recirculation zone at this given size of the air/O2 flames have bright white color due to presence of excess
bluff body which resulted in better mixing between burned and oxygen that leads to high flame temperature. The flame shape
 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28 27

Fig. 10. CNG/O2/CO2 flames at BR of 0.36, fuel momentum of 1.15E5 N m and OF of: (left side) 28% at oxidizer momenta of 0.00064, 0.00121 and 0.00202 N m, from left to
right, and (right side) 32% at oxidizer momenta of 0.00069, 0.00134 and 0.00275 N m, from left to right.

changed from like-jet flame at low ratio of oxidizer to fuel momen- ranges of oxygen fraction (vol.%) and blockage ratio. Oxygen frac-
tum, to central-jet dominated flame at intermediate oxidizer to tion (OF) in this case is defined as the percentage of total volumet-
fuel momentum. At high oxidizer to fuel momentum ratio, a strong ric flow rate of oxygen ðVO2 Þ divided by the total volumetric flow
stable recirculation-zone flame was obtained. Increasing the ratio rate of oxidizer ðVO2 þ VCO2 Þ.
of oxidizer to fuel momentum resulted in shorter flame length
and wider flame size in the traverse direction of the flow. This
may be attributed to the improved mixing between fuel and oxi- 3.3.1. Flammability limits of CNG/CO2/O2 flames
dizer over the bluff body and the creation of strong recirculation In this set of experiments, the flammability limits were studied
zone for flame stabilization. at the same four different blockage ratios like in air–fuel and
oxygen-enriched air combustion cases. In this set of experiments,
3.2.3. Exhaust emissions of CNG/air/O2 flames two different oxygen fraction (OF) have been studied, namely
The emissions characteristics are presented in Fig. 7 for CNG/ 28% and 32% (by vol.). To determine the upper extinction limits
air/O2 flames considering oxygen fraction of 21%. Fig. 7 represents of the flame, the fuel flow rate was kept constant and the oxidizer
the emission measurements at the center of the combustor and flow rate was increased gradually at fixed oxygen fraction till the
close to its exit section for oxygen fractions of 21%. The measure- extinction point. The stability curves for four different blockage
ments were recorded considering blockage ratios of 0.36, 0.5, ratios (0.36, 0.50, 0.67 and 0.82) are presented in Fig. 8. For the
0.67 and 0.82. As it can be observed from the plots in Fig. 7, the same oxygen fraction and blockage ratio, the results show that as
emission index of oxygen was decreased as the equivalence ratio the fuel momentum increases, the flame stability improves, as
is increased. For the same blockage ratio and equivalence ratio, observed in air and oxygen-enriched air combustion cases. For
the emission index of O2 in case of CNG/air/O2 flame is lower than the considered two oxygen fractions, the operation under blockage
that index in case of CNG/air flame (compare the results in Figs. 4 ratio of 0.82 showed the best flame stability conditions. For sake of
and 7). As presented in Fig. 7, the emission index of hydrocarbon comparison, the best flame stability conditions occurred in case of
(HC) is increased as the equivalence is increased and its value air combustion at blockage ratio of 0.67 (see Fig. 2), which is not
became zero at equivalence ratio of 0.9 due to excess oxidizer used the case for oxy-combustion as presented in Fig. 8. Considering
in combustion. Also, the emission index of hydrocarbon in case of fixed blockage ratio and fuel momentum, the results showed that
CNG/air/O2 flame is lower than that index in case of CNG/air flame increasing the oxygen fraction improves the flame stability, as
considering the same operating conditions (compare the results in observed also while operating under oxygen-enriched air combus-
Figs. 4 and 7). The emission index of CO2 is almost constant and tion conditions.
higher than that index when pure air is used as an oxidizer, for For the sake of comparison, Fig. 9 compares the stability lim-
the same blockage ratio and equivalence ratio. The emission its of the oxy-fuel combustion with their correspondents of air
indexes of O2, CO2 and HC at blockage ratios of 0.36, 0.5, and fuel combustion at the considered four blockage ratios. The fig-
0.667 are very close. However, the maximum emissions indexes ure shows that the flammability limits increases as fuel momen-
values of O2, HC, and CO2 at blockage ratio of 0.82 are too high in tum increases and the CNG/CO2/O2 flames with oxygen fraction
comparison with the values at other blockage ratios as shown in of 32% have higher extinction limits than the flames with oxygen
Fig. 7. This may be attributed to the too large distance between fuel fraction of 28%. Also the CNG/air flames have higher extinction
and air streams at blockage ratio of 0.82 which prevents better limits than those of CNG/CO2/O2 flames. This may be attributed
mixing between both streams and, accordingly, emission concen- to the adverse effect of adding CO2 to the oxidizer mixture on
trations are increased. flame stability. As a matter of fact, using CO2 as diluent impacts
the flame in four ways as illustrated by Amato et al. [30,31]
3.3. Oxy-fuel combustion (CNG/O2/CO2) through changes in (1) mixture specific heat and adiabatic flame
temperature; (2) transport properties (thermal conductivity,
This section discusses the flammability limits and visual flame mass diffusivity, viscosity); (3) chemical kinetic rates; and (4)
length and appearance of CNG/O2/CO2 oxy-combustion flames over radiative heat transfer.
28 I.A. Ramadan et al. / Applied Energy 178 (2016) 19–28

3.3.2. Visual flame length and appearance of CNG/CO2/O2 flames combustion: technology, fundamentals and numerical simulations. Appl
Energy 2015;154:1020–47.
The shapes of the CNG/O2/CO2 flames at BR of 0.36 and fuel
[2] Charles, Stren R. Oxygen-enhanced combustion: library of congress; 1998, 0-
momentum of 1.15E5 N m are presented in Fig. 10 considering 8493-1695-2.
two values of oxygen fraction including 28% and 32% for different [3] Andersson K. NO emission during oxy-fuel combustion of lignite. Int J Eng
oxidizer momentum values. As shown in the figure, CNG/CO2/O2 Chem 2008;47:1–11.
[4] Gabrielli R, Sing R. Thermodynamic performance analysis of new gas turbine
flames at low oxygen fraction (i.e. 28 vol.%) have blue color. This combined cycles with no emissions of carbon dioxide. ASME; 2002.
may be attributed to high levels of H radicals within the flame core. [5] Sanz W. Thermodynamic and economic investigation of an improved Graz
At high oxygen fraction (i.e. 32 vol.%), the flame color is changed cycle power plant for CO2. Eng Gas Turbines Power 2005;127:765–73.
[6] Mathieu P, Nihart R. Zero-emissions MATIANT cycle. Trans ASME
from white to blue while increasing the oxidizer momentum. This 1999;121:116–20.
can be attributed to the decrease in flame temperature and the [7] Yin C, Yan J. Oxy-fuel combustion of pulverized fuels: combustion
increase in concentrations of H radicals within the flame core. fundamentals and modeling. Appl Energy 2016;162:742–62.
[8] Tan Y, Jia L, Wu Y, Anthony E. Experiences and results on a 0.8 MWth Oxy-fuel
These flames changed from like-jet flame at low ratio of oxidizer operation pilot-scale circulating fluidized bed. Appl Energy 2012;92:343–7.
to fuel momentum to central-jet dominated flame at intermediate [9] Hu Y, Yan J. Characterization of flue gas in oxy-coal combustion processes for
ratio of oxidizer to fuel momentum and, finally, a stable flame was CO2 capture. Appl Energy 2012;90:113–21.
[10] Hu Y, Li H, Yan J. Numerical investigation of heat transfer characteristics in
obtained due to the creation of strong recirculation-zone at high utility boilers of oxy-coal combustion. Appl Energy 2014;130:543–51.
ratio of oxidizer to fuel momentum. [11] Fu Q, Kansha Y, Song C, Liu Y, Ishizuka M, Tsutsumi A. A cryogenic air
separation process based on self-heat recuperation for oxy-combustion plants.
Appl Energy. http://dx.doi.org/10.1016/j.apenergy.2015.03.039.
4. Conclusions [12] Oh Jeongseog, Noh Dongsoon, Lee Eungyeong. The effect of CO addition on
flame behavior of a non-premixed oxy-methane jet in a lab-scale furnace. Appl
Energy 2013;112:350–7.
In this work, an experimental investigation has been performed [13] Tan Yewen, Douglas Mark A, Thambimuthu Kelly V. CO2 capture using oxygen
on different diffusion flames including air–fuel, oxygen-enriched enhanced combustion strategies for natural gas power plants. Fuel
air fuel and oxy-fuel combustion flames anchored over a bluff body 2002;81:1007–16.
[14] Nemitallah MA, Habib MA. Experimental and numberical investigations of an
with different blockage ratios. The flammability limits, visual flame atmospheric diffusion oxy-combustion flame in a gas turbine model
appearance and exhaust emissions have been investigated and combustor. Appl Energy 2014;111:301–415.
compared for all flames. For all flames, the results reveled that, [15] Nemitallah MA, Habib MA. Design of an ion transport membrane reactor for
application in fire tube boilers. Energy 2015;81:787–801.
for the same blockage ratio, increasing the fuel momentum results [16] Nemitallah MA, Habib MA, Pervez A, Mostafa HS, Hassan MB, Inam M, et al.
in an increase in the flame stability. This is attributed to improve- Experimental analysis of oxygen–methane combustion inside a gas turbine
ments in Reynolds number and associated improvements in turbu- reactor under various operating conditions. Energy 2015;86:105–14.
[17] Ramadan I, Abou-Arab TW, Ibrahim AH. Oxy-combustion of CNG flames as
lence and mixing. For air–fuel and oxygen-enriched air fuel flams,
compared to air and enriched air: appearance, stability and emissions. In: 7th
the results indicated that the optimum flame stability was Mediterranean combustion symposium, 2011, Sardinia, Italy. p. 11–5.
obtained at a blockage ratio of 0.67. For oxygen-enriched air fuel [18] Chi Chung C, Lin Ta Hui. Oxy-oil combustion characteristics of an existing
and oxy-fuel flames, increasing oxygen fraction leads to an furnace. Appl Energy 2013;102:923–30.
[19] Rashwan SS, Ibrahim AH, Abou-Arab TW. M.Sc. Thesis. Experimental
increase in flame stability due to the enhanced chemical kinetics investigation of oxy-fuel combustion of CNG flames stabilized over a
in the flame zone. For oxy-fuel flames, operation at blockage ratio perforated plate burner. Mechanical Engineering Department, Cairo
of 0.82 resulted in best flame stability for different oxidizer University; 2014. http://dx.doi.org/10.13140/RG.2.1.1312.4240.
[20] Rashwan SS, Ibrahim AH, Abou-Arab TW. Experimental investigation of oxy-
mixture compositions. The flame shape was mainly dominated fuel combustion of CNG flames stabilized over a perforated plate burner. In:
by the ratio of air to fuel momentum. At low air momentum, the 18th IFRF members’ Conference-Flexible and Clean Fuel Conversion to
fuel jet penetrates the recirculation zone and forms a like-jet flame. Industry, Freising, Germany, 1, 2, 3 June 2015, Paper 25. http://dx.doi.org/10.
13140/RG.2.1.3537.9686.
When the air momentum increases, the fuel is held behind the [21] Rashwan SS, Ibrahim AH, Abou-Arab TW, Nemitallah MA, Habib MA.
recirculation zone and forms a recirculation zone flame. In case Experimental investigation of partially premixed methane-air and methane
of comparable values of fuel and air momentum values, a oxygen flames stabilized over a perforated-plate burner. Appl Energy
2016;169:126–37.
central-jet dominated flame was formed. The flame length was [22] Ditaranto M, Hals Jorgen. Combustion instabilities in sudden expansion oxy-
higher in air combustion than in oxy-combustion cases. The max- fuel flames. Combust Flame 2006;146:493–512.
imum hydrocarbon emission index observed at blockage ratio of [23] Liu CY, Ghen G, Sipocz N, Assadi M, Bai XS. Characteristics of oxy combustion
in gas turbines. Appl Energy 2012;89:387–94.
0.82 due to the increase in distance between oxidizer and fuel
[24] Joo PH, Charest MRJ, Groth CPT, Gulder OL. Comparison of structures of
stream. These results and findings enable the prediction of the laminar methane–oxygen and methane–air diffusion flames from atmospheric
stability and emission characteristics of diffusion flames when ret- to 60 atm. Combust Flame 2013;160:1990–8.
rofitting an existing furnace/gas turbine to be adapted for oxy-fuel [25] Guo-neng LI, Zhou H, Cen K. Emission characteristics and combustion
instabilities in an oxy-fuel swirl-stabilized combustor. J Zhejiang Univ
combustion. 2008;9:1582–9.
[26] Kutne P, Kapadia BK, Meier W, Aigner M. Experimental analysis of combustion
behaviour of oxyfuel flames in a gas turbine model combustor. Proc Combust
Acknowledgement Inst 2011;33:3383–90.
[27] Hu Y, Yan J, Li G. Effects of flue gas recycle on oxy-coal power generation
This present work was funded by the National Plan for Science, systems. Appl Energy 2012;97:255–63.
[28] Aneke M, Wang M. Process analysis of pressurized oxy-coal power cycle for
Technology and Innovation (MAARIFAH)-King Abdulaziz City for carbon capture application integrated with liquid air power generation and
Science and Technology through the Science & Technology Unit binary cycle engines. Appl Energy 2015;154:556–66.
at King Fahd University of Petroleum & Minerals (KFUPM)-the [29] Kang YP, Yang JT. Blockage effect on turbulent flame structure behind a disc
stabilizer with a central fuel jet. In: Proceedings of PSFVIP-4, France; 2003. p.
Kingdom of Saudi Arabia, award number (14-ENE58-04). 1–12.
[30] Amato A, Hudak R, Noble DR. Methane oxy-combustion for low CO2 cycles:
blowoff measurements and modeling. In: Proceedings of ASME turbo expo.
References
[31] Amato A, Hudak R, Noble DR. Methane oxy-combustion for low CO2 cycles:
measurements and modeling of CO and O2 emissions. In: Proceedings of ASME
[1] Taamallah S, Vogiatzaki K, Ghoniem A, Alzahrani F, Mokheimer E, Habib MA. turbo expo.
Fuel flexibility, stability and emissions in premixed hydrogen-rish