ARTICLE pubs.acs.org/EF
Experimental Investigation of Kerosene Spray Flames in Inert Porous Media near Lean Extinction Chendhil Periasamy*,† and Subramanyam R. Gollahalli Combustion and Flame Dynamics Laboratory, School of Aerospace and Mechanical Engineering, The University of Oklahoma, 865 Asp Avenue, Felgar Hall, Room 212, Norman, Oklahoma 73019, United States ABSTRACT: This paper presents a study of kerosene spray flames in inert porous media near lean extinction to understand stable operating regimes. The porous media are composed of silicon-carbide-coated carboncarbon foam type with 825 pores per centimeter and 87% porosity. Two porous media, one in the evaporation section and the other in the combustion section located downstream, are used in the experiments. Aviation-grade kerosene was injected upstream of the evaporation porous medium with an air-blast atomizer. Stable flames were established both inside and on the downstream surface of the combustion porous medium. Results show that, in interior combustion mode, the flame was completely contained within the porous medium until extinction. In the surface combustion mode, a decrease of the fuel flow rate resulted in partial flame lift and subsequent extinction. The equivalence ratio near lean extinction in each mode was determined. A Damk€ohler number (tres/tchem)-based analysis was developed to study the combustion mode and flame extinction behavior. A nominal value of the Damk€ohler number of 5.0 was required to initiate the interior combustion mode. As the Damk€ohler number was increased, the extinction equivalence ratio decreased, thus extending the range of fuel-lean operation. Axial temperature profiles in evaporation and combustion porous media were measured. Surface temperature uniformity in porous media was also examined near extinction conditions. Also measured were the radiative heat release from porous medium downstream of the exit surface and emissions of pollutants, carbon monoxide and nitric oxide. Results demonstrate the benefits of the porous medium in making NO emission somewhat insensitive to operating parameters, such as the equivalence ratio and location of the injector.
’ INTRODUCTION Reliability of burners in industrial combustion processes depends upon stable firing over the entire range of operating conditions, uniform heat release to load, and low pollutant emissions. One potential solution to improve burner operation is the use of porous media in the combustion section. In this approach, the heat from the combustion zone is fed back to preheat reactants without actually recirculating combustion products.14 Typically, a lowporosity medium, which functions as a flashback arrestor, is placed upstream of another higher porosity medium, which facilitates combustion. This method of combustion could potentially extend the limits of fuel-lean operation, improve stable burner operation over a wide range of loads, enhance reactant mixing, reduce pollutant emissions, and burn fuels with low-calorific values. Flame stabilization in porous media burners employing gaseous fuels has been studied by several investigators.58 Sathe et al.5 numerically studied flame stabilization and multimode heat transfer in porous radiant burners. They concluded that the flame could be stabilized in the upstream half or near the downstream edge of the porous medium, where the velocity profile exhibited positive slopes. Lammers and de Goey6 have conducted a numerical study on the flashback of the premixed methane flames stabilized on the surface and inside of a ceramic burner. Stability diagrams and flashback regimes were presented. Barra et al.7 showed that materials with low conductivity, small heat-transfer coefficients, and large radiative extinction coefficients were desired for the upstream section. For the downstream section of the burner, high conductivity and large heat-transfer coefficients were necessary to enhance the heat transfer. Mathis and Ellzey8 conducted an experimental study to r 2011 American Chemical Society
measure the flame stabilization, operating range, and CO/NOx emissions for two different methane-fueled porous burners. Stable flames were established at or near the interface between large- and small-pore sections. When liquid fuel is burned in porous media, the upstream heat feedback could be beneficially exploited to enhance fuel evaporation and subsequent mixing of the fuel vapor with air, leading to more complete combustion. Kaplan and Hall9 studied four different designs of heptane-fueled radiant burners. Stable operating ranges and emission characteristics were reported. Stable combustion was achieved over the equivalence ratio (ϕ) range of 0.570.67. Tseng and Howell10 investigated liquid fuel combustion in porous media numerically and experimentally. Flame stabilization was achieved at a low equivalence ratio of 0.3. Jugjai and Polmart11 developed a down-flow atomizer-free porous burner. Stable combustion was achieved at an equivalence ratio of 0.2. Jugjai and Pongsai12 extended the study to develop an atomizer-free porous burner. Matrix- and surface-stabilized flame configurations were studied. Matrix-stabilized flames were found to be more stable in a wider operating regime. Mujeebu et al.13 presents a review of investigations on liquid fuel combustion in porous media. Previously, Periasamy et al.14 experimentally studied the evaporation enhancement of liquid spray fuels in porous media. They have also presented a numerical modeling of evaporation Received: November 29, 2010 Revised: July 19, 2011 Published: July 21, 2011 3428
dx.doi.org/10.1021/ef1016076 | Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
ARTICLE
Figure 1. Schematic diagram of the experimental setup.
Table 1. Combustion Chamber Dimensions and Operating Conditions parameter
value Test Chamber cross-section, 76 76; height, 163
test chamber dimensions (cm) Pyrex glass plate window (cm)
25 135
porous medium housing (cm)
cross-section, 5 5; height, 25 Porous Media (Manufacturer, Ultramet)
pores per centimeter
31, 25, 18, 12, and 8
cross-section (cm) thickness (cm)
44 2.5
porosity (%)
87 Operating Conditions
fuel
kerosene
fuel tank pressure (atm)
0.34 (gauge)
injector type (Delavan model 3060-1)
solid cone
spray angle (deg)
40
atomizing air flow rate (L/min)
48
co-flow air flow rate (L/min) co-flow air velocity (m/s)
75195 13
temperature of the secondary co-flow (K)
cold450
Reynolds number (on the basis of test section
5600
dimensions and co-flow air velocity at 450 K)
enhancement of liquid spray with porous media.15 This paper shifts the attention to the near-extinction regime of spray flames
in porous media. The specific purpose of the current work is to study extinction behavior and determine a stable operating 3429
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
ARTICLE
Figure 3. Experimental arrangement of liquid spray combustion in porous media.
Table 2. Typical Properties of Porous Medium Figure 2. Photographs of the porous media used in experiments and microscopic view of a typical open-cell porous medium.
property
regime of spray flames in porous media. The data would be useful for adapting porous media burners in the development of lean combustors for gas turbines and furnaces. Both surface and interior flames were studied. A Damk€ohler-number-based approach is presented to understand lean extinction characteristics. Axial temperature profiles, surface temperature uniformity, and pollutant emissions are also examined near extinction conditions.
’ EXPERIMENTAL SECTION Laboratory Combustion Chamber. The experimental setup consisted of a housing for evaporation and combustion porous media, rectangular glass test sections, air and fuel supply system, and co-flow air preheater. The setup was housed in a vertical steel test chamber of 76 76 cm cross-section and 163 cm height. A schematic of the test chamber and experimental setup is shown in Figure 1. A base plate with a square opening side of 11 cm was fitted at the bottom of the test chamber. The porous medium housing sections were mounted on the base plate, and the air-settling chamber was located immediately below the base plate. The top of the test chamber was open to the atmosphere through an exhaust duct. The ambient pressure inside the laboratory was maintained slightly above atmospheric pressure to ensure a positive draft inside the test chamber. The dimensions of the test chamber and nominal operating conditions are listed in Table 1. Test Porous Media. Open-cell silicon-carbide-coated carbon carbon matrix porous media of cross-section 4.3 4.3 cm and a height of 2.5 cm were used. Two porous media with linear pore densities [defined as the number of pores per centimeter (PPCM)] of 8 and 25 were used. Both porous media had a porosity of 87%. Figure 2 shows photographs of 8 (Figure 2a) and 25 (Figure 2b) PPCM porous media used in the experiments. The 25 PPCM porous media were used in the evaporation section [referred to as the evaporation porous medium (EPM)], and the 8 PPCM porous media were used in the combustion section [referred to as combustion porous medium (CPM)]. Figure 3 shows a typical arrangement of the porous media in the test section. EPM also functioned as a flame arrestor, preventing flashback. Typical properties of porous media are listed in Table 2. In making engineered porous media, reticulated vitreous carbon foams are typically created out of polyurethane foams from a pyrolysis process. This foam is then infiltrated with structural ceramic or metal
value
units
bulk density
0.11.45
g/cm3
ligament density
3.2
g/cm3
surface area
0.08
m2/cm3
specific heat maximum use temperature
1.424 1700
J g1 °C1 °C
thermal conductivity
13
W m1 K1
porosity
87
%
number of pores per centimeter
831
material to form open cell foam. Figure 2c shows a microscopic view of a typical open-cell porous medium. The focus of this paper is to study the influence of porous media on evaporation enhancement, lean extinction, and combustion processes compared to those without porous medium, as in the conventional spray burners. Hence, only the bulk properties of porous media, such as porosity and pores per centimeter, are considered rather than the properties that depend upon the internal structure of the porous medium. While these bulk properties depend upon internal structures, characterization details regarding the internal flow path were not considered. Further, pressure drop measurements have been made across the two porous media used in the experiment. The results are reported by Periasamy et al.14 and Periasamy16 and compared to the Ergun equation.1 According to the Ergun equation,1 the pressure drop across a packed column depends upon the fluid flow rate, fluid properties, closeness and orientation of packing, and size, shape, and surface conditions of the particles. The pressure drop results from our study are found to be in good agreement with the Ergun equation.1 Combustion Air, Fuel Supply, and Pilot Flame. Combustion air was obtained from a laboratory compressed air source. This was filtered, dried, and preheated using a high-throughput variable power supply electrical heater. From the preheater, the hot air was supplied through the insulated aluminum settling chamber located below the base plate in the test chamber (Figure 1) and then admitted through porous media. Aviation-type kerosene (Jet A) was stored in a nitrogenpressurized tank and used as fuel. This was injected onto the upstream surface of the EPM using an air-blast atomizer. The atomizer generated a solid cone fuel spray with a total angle of 40°. A calibrated rotameter was used to meter the fuel flow rate. For detailed injector arrangement, the reader is referred the study by Periasamy et al.14 A non-preheated air stream was supplied through an injector as atomizing air. A propane pilot flame was used to ignite kerosene vapors and was removed after a stable flame was established in the CPM. 3430
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels Evaporation and Combustion Test Sections. Two glass test sections of 5 5 cm inside cross-section and 25 cm height were installed one on each side of the porous medium housing. The upstream section was made of borosilicate glass with 4 mm thickness. The injector protruded into the upstream test section. The downstream glass test section was made up of a high-temperature Vycor glass. The evaporated fuel vapors were flared using a pilot flame at the exit of the downstream test section. Stainless-steel, porous disk flame arrestors were installed to prevent the propagation of the flame into the evaporation test section. Instrumentation. Flame photographs were taken using a Canon digital SLR camera with a shutter speed of 1/100 s. Solid-phase temperatures on the surface and interior of the porous medium were taken using thermocouples (type K and R) with wire and bead diameters of 0.3 mm and 400 μm, respectively. A two-dimensional image of the porous medium surface temperature was also obtained non-intrusively using an infrared camera to check the uniformity of the temperature distribution. Infrared images were recorded using a DVD recorder and exported to a personal computer. Porous medium surface radiation was measured using a pyrheliometer (instrument sensitivity = 28.4 W m2 mV1). Gas samples were collected using a Pyrex cone placed above the upstream test section. An uncooled quartz sampling probe with an internal diameter of 2 mm was placed at the top end of the cone. The sample was filtered and passed through a glass condenser placed in an ice bath. A vacuum pump was used to draw the sample, and the flow rate of the sample gas was monitored with a rotameter. Concentrations of CO and CO2 were measured using a nondispersive infrared analyzer. The NO concentration was measured using a chemiluminescence NONO2NOx analyzer. The oxygen concentration was measured with a polarographic sensor. From the measurements, the global emission indices were calculated. Experimental Procedure, Test Conditions, and Uncertainty Estimates. Co-flow air preheater was first turned on. The flow rate of air was set to produce the desired velocity in the test section using a calibrated rotameter. The liquid fuel tank was pressurized using compressed nitrogen. A propane pilot flame was ignited downstream of the CPM. The flow rate of kerosene was then set at the desired value. Atomizing air flow was adjusted to provide a steady spray. Once the fuel vapors started to burn continuously and a stable flame was established, the propane pilot flame was turned off. The injector distance from the upstream surface of the porous medium was set at four values (36 cm), and the overall equivalence ratio ranged from 0.9 to 0.2. The ranges of other parameters are given in Table 1. Uncertainties in experimental measurements were calculated using Student’s t distribution at 95% confidence interval and are reported in appropriate figures.
’ RESULTS Kerosene fuel spray was injected on the leading surface of the EPM, and stable flames were established both inside and on the surface of the CPM. First, the flame appearance and its general characteristics are described. A Damk€ohler number (Da = tres/tchem)-based model has been proposed to describe lean extinction in porous media. Axial temperature profiles and surface temperature uniformity are also presented near extinction conditions. Further, radiation from the porous medium surface and CO and NO emissions are reported. Flame Appearance. Initially, co-flow air and fuel corresponding to a fuel-rich condition were supplied through the porous medium. The fuel vaporair mixture was ignited by the pilot flame, and a stable flame was established in the porous medium. A stable flame is defined as one that is entirely contained within or on the surface of the porous medium for a given fuel and air
ARTICLE
Figure 4. Top view of porous media during interior and surface combustion modes.
flow rate and remained steady. The former type is referred to as interior flames, and the latter type is referred to as surface flames. Interior Flames. In interior combustion mode, the flame was stabilized inside the CPM. The flame was completely contained within the porous medium. Figure 4a shows a typical flame during interior combustion mode. The porous medium glowed in bright orange color. With a reduction in the fuel flow rate below a certain value (corresponding to the stable lean limit of interior flames), the flame appeared as a transient weak blue flame on the surface and immediately extinguished. The porous medium was able to withstand the combustion continuously for more than 2 h. The combined operation of one porous medium exceeded more than 100 h, maintaining structural stability. No clogging of pores was observed. Further, interior flames operated quieter than surface flames. Surface Flames. In surface combustion mode, the flame was always stabilized on the downstream exit surface of the porous medium. The flame covered the entire porous medium surface. Figure 4b shows a typical surface flame observed in this study. The flame appeared as a contiguous flat bright blue sheet. The flame zone was located about 13 mm above the porous medium surface. As the fuel flow rate was decreased below a certain equivalence ratio, the flame first lifted from the surface on locations where the gas velocity was higher than the local flame speed. As the mixture was made leaner, the flat flame structure weakened and transformed into a cellular structure. With a further decrease in the fuel flow rate, the flame was completely lifted from the surface and immediately extinguished. A fully lifted flame could not be stabilized. Lean Extinction Limits. From the initial stable fuel-rich condition, the fuel flow rate was decreased at the rate of 0.01 ϕ/min while keeping the co-flow air velocity constant. In interior combustion mode, below a critical fuel flow rate (or equivalence ratio), the flame appeared as a transient weak blue flame on the surface and blew out immediately. In surface combustion mode, below a critical fuel flow rate, the flame lifted from the exit surface of the CPM and blew out immediately. The condition where the flame (interior or surface) is not present in the porous medium is referred to as flame extinction. Such extinction of the visible flame was detected by visual observation and porous medium surface temperature measurements. Because a large part of the radiance from a flame is at infrared wavelength, one effective way to detect the flame extinction is by the use of a radiometer. However, the presence of the solid medium continued to emit significant radiation even after the flame is extinct. Hence, the authors believe one appropriate way to detect flame extinction is by visual means. The fuel flow rate just prior to the extinction of the flame was recorded, and the equivalence ratio at extinction was calculated. This procedure was repeated for a range of co-flow air flow rates and injector locations. 3431
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
ARTICLE
Figure 5. Variation of the extinction equivalence ratio with gas velocity and injector locations in interior and surface combustion modes.
Figure 6. Effect of the Damk€ohler number on the extinction equivalence ratio of kerosene spray flames in porous media.
Figure 5 presents the variation of the extinction equivalence ratio with effective gas velocity (defined as the ratio of the total air flow rate and porous medium surface area normal to the flow direction) for different injector locations. Each data point in the figure denotes the lowest equivalence ratio up to which the flames (interior or surface combustion) could be established in the porous medium. Surface combustion data points are shown as filled symbols in Figure 5. This figure illustrates the operating regimes of interior and surface combustion flames in porous media. The results show that a decrease in the effective gas velocity decreased the extinction equivalence ratio for all injector locations. The effective gas velocity determined the location of the flame in the porous media, and a decrease in the effective gas velocity moved the flame further inside the porous medium. Uncertainty in the measurement of the extinction equivalence ratio was less than (8% of the mean value. For injector locations >3 cm, because the droplets have a longer residence time both upstream of and inside the porous medium, a more uniform fuelair mixture is made available for combustion. Hence, the effective gas velocity does not significantly influence the extinction equivalence ratio until the mode changed to interior combustion. € hler Number Analysis near the Flame Extinction. A Damko characteristic residence time for the fuel spray upstream and inside EPM (tres) can be calculated as follows:
the porous medium, as follows:
tres ¼
dip þ atp u
ð1Þ
where dip is the distance between the injector exit surface and upstream surface of the porous medium, a is the porous medium thickness correction factor (3.4), tp is the thickness of the porous medium, and u is the axial gas velocity. Note that one unit length of porous medium provides more residence time than one unit length of open space. Resistance offered by the porous medium could be calculated using the “hydraulic radius model”.17 This model assumes the presence of imaginary hydraulic tubes in the porous medium. These hydraulic tubes are responsible for the randomness in the porous medium structure, because they do not generally follow a straight path. The resistance offered by the porous medium depends upon the length of hydraulic tubes (LH) in the medium. Tortuosity (τ) is defined as the ratio of the thickness of the porous medium and the length of the longest hydraulic tube in
τ¼
tp LH
ð2Þ
where tp is the thickness of the CPM. Typical tortuosity values are presented by Kaviany.17 In the present study, a tortuosity of 0.3, typically corresponding to packed beds, was chosen. For a porous medium of thickness 2.54 cm, the length of the longest hydraulic tube was calculated to be 8.54 cm. A characteristic chemical time (tchem) for the combustion of kerosene spray can be calculated as follows: tchem ¼
δ SL
ð3Þ
where δ is the laminar flame thickness and SL is the laminar flame speed of kerosene with air at stoichiometric conditions. In eq 3, the laminar flame thickness and flame speed of kerosene with air at stoichiometric conditions were used.18 In the present study, the equivalence ratio employed ranged from 0.9 to 0.2. Hence, the flame thickness and speed were extrapolated to lean conditions and employed in the above expression. Figure 6 presents the effect of the Damk€ohler number on the extinction equivalence ratio for different injector locations. The operation regimes of interior and surface combustion modes are also marked in the figure. The specific mode of combustion is determined by the completeness of vaporization and the quality of the reactant mixture in CPM. Availability of a flammable mixture in CPM favors interior combustion. If a flammable mixture is not prepared until the exit surface of EPM and CPM, surface combustion is most likely to occur. Because the air flow rate was held constant, the average gas velocity through the porous medium was also constant. When the fuel flow rate is reduced, at some fuel flow rate, the local gas velocity exceeded the flame speed and blew out the flame. Temperature Distribution near Extinction Conditions. The porous medium interior temperature and surface temperature uniformity measurements were taken to understand the upstream heat feedback rate and detection of flame extinction. Porous Medium Interior Temperature. Figure 7 presents the measured centerline temperature of EPM and CPM for different equivalence ratios prior to flame extinction. The figure indicates that the axial temperature in the EPM increased only slightly. On the other hand, at the interface between the EPM and CPM, the 3432
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
ARTICLE
Figure 7. Measured centerline temperature of EPM and CPM for different equivalence ratios prior to flame extinction.
Figure 9. Contours of radiance qualitatively representing the difference between the maximum and minimum temperatures on the porous medium surface during interior combustion (conditions: dip, 5 cm; coflow air velocity, 125 cm/s; Damk€ohler number, 5.0).
Figure 8. Decay of the porous medium centerline temperature of interior flames prior to extinction (Da = 5.0).
temperature increased rapidly. This indicates that the combustion began at the upstream surface of the CPM. After the flame zone, the axial temperature profiles exhibited a decreasing trend. Further, the peak temperature attained in the porous medium at each equivalence ratio decreased with a decreasing equivalence ratio. In this case, a stable interior combustion in the CPM was observed until the equivalence ratio was reduced to 0.35. Temperature measurements at the centerline near the porous medium exit surface were taken, and Figure 8 shows such results at different equivalence ratios prior to extinction for a Damk€ohler number of 5.0. The figure shows that the surface temperature decreased as the fuel flow rate was reduced; i.e., the lean extinction was approached. Because thermocouples were embedded just below the downstream surface of the porous medium, they continue to record high values even after flame extinction because the porous medium retained the heat for a considerable time. After the flame was extinguished, one more measurement was taken in each case to determine the surface temperature just below which the extinction occurs for all cases. All of the flames were extinguished at and below a surface temperature of 900 K. Babb et al.19 experimentally measured an extinction temperature of 1500 K for liquid heptane open diffusion flames (no porous medium was employed). In the present study, an extinction temperature of less than 900 K was recorded. This demonstrates that combustion in porous medium
could be stabilized at much lower temperatures than that in open flames. Hence, the insertion of the porous medium in the combustion zone of a spray flame markedly widens the range of stable burner operation. Note that a thermocouple embedded in a porous medium measures an average of gas and solid temperatures because it is in contact with both solid and fluid media. It does not indicate the true porous medium temperature. Other methods, such as infrared imaging and/or modeling, are helpful for understanding thermocouple measurements. Hence, in an earlier study by the authors,15 thermocouple measurements are compared to infrared imaging and two computer models (equilibrium and nonequilibrium models). Surface Temperature Uniformity. The porous medium surface temperature uniformity denotes spacious homogeneity of the fuelair mixture prior to the flame zone. Figure 9 presents the measurements of radiance contours qualitatively representing the porous medium surface temperature uniformity during interior combustion at four different equivalence ratios prior to extinction for a Damk€ohler number of 5.0. The contours qualitatively show that the peak temperature decreased as the equivalence ratio was decreased, as also observed with thermocouple measurements. The results further show that the surface temperature was uniform within (50 °C. Note that the figure shows only the difference between the maximum and minimum temperatures because the emissivity of the porous medium was not known accurately. Semi-quantitative results could, however, be obtained by combining the thermocouple and infrared images. Khanna et al.20 measured the exit plane temperatures of methaneair combustion in the porous medium. They reported a temperature of 1250 K for an equivalence of 0.6. In the present study with kerosene spray flames in the porous medium, the 3433
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
ARTICLE
Figure 10. Radiation from the porous medium surface for different equivalence ratios prior to extinction (co-flow air velocity = 96 cm/s).
surface temperature measured was 1241 K. A comparison of our results to that by Khanna et al.20 reveals that our data are in good agreement with the literature. Radiation from the Porous Medium near Flame Extinction. Radiation from the porous medium during combustion was measured using a radiometer. The radiometer was located directly 35 cm above the exit surface of CPM. A quartz window (spectral transmission of 0.24.5 μm) was used to cover the sensing surface, and hence, only the radiant energy output over a wavelength range of 0.24.5 μm from the porous medium surface was measured. Figure 10 presents the radiation from the CPM at different injector locations during the interior combustion mode. While the co-flow air flow rate was held constant at 90 L/min, the fuel flow rate was decreased. The figure shows that the flame radiation decreased as the extinction was approached. As the fuel flow rate was reduced, the heat input was decreased, and hence, the radiant heat energy from the flames was also decreased. Decreasing porous medium surface temperature profiles during flame extinction also support the decreasing trend of flame radiation. Although there were some non-systematic variations in the radiant energy output with the location of the injector observed, these were within the experimental uncertainties (mean value (100 W/m2). Pollutant Emission near Flame Extinction. Emission indices of carbon monoxide (CO) and nitric oxide (NO) were measured. A Pyrex glass funnel was placed directly 1 cm above the exit plane of the combustion test section. A quartz probe was inserted at the exit of the glass funnel, and the sample was drawn. From the measurements, the global emission index was calculated as follows21: emission index of species i, EIi ¼
Xi xMW i XCO þ XCO2 MW fuel ð4Þ
where Xi is the mole of fraction of the species of interest, XCO is the CO mole fraction, XCO2 is the CO2 mole fraction, x is the number of carbon atoms per mole of fuel, MWi is the molecular weight of the species of interest, and MWfuel is the molecular weight of the fuel. Emission Indices of Carbon Monoxide (EICO). Figure 11 presents global CO emission indices for four equivalence ratios and two injector locations prior to extinction. CO emission indices ranged from 13 to 100 g/kg of fuel. Equilibrium calculations showed a CO emission index of 4.77 g/kg of fuel at an equivalence ratio of 0.7.22 Note that the above equilibrium
Figure 11. Measured emission indices of CO at four equivalence ratios prior to extinction for two injector locations (co-flow air velocity = 96 cm/s).
calculations assume prevaporized and premixed combustion of kerosene vapors with air. Also, no porous medium was employed. As the equivalence ratio was decreased or flame extinction was approached, EICO showed a somewhat non-systematic variation. This variation was, however, within the experimental uncertainty of (12 g/kg of fuel. The injector located farther from the leading edge of EPM produced lower EICO. Such emission index measurements are independent of any dilution by air and widely used in evaluating the efficiency of combustion systems. In a methane-fueled porous medium burner, Khanna et al.20 measured the CO emissions as 5120 ppm for equivalence ratios of 0.60.87. In the present study, dependent upon the location of the injector, CO emissions were obtained from 40 to 160 ppm, corrected to 3% oxygen. Note that Khanna et al.20 used gaseous methane as fuel. However, a comparison of our results to the results by Khanna et al.20 reveals that the results are in conformity with the literature. Emission Indices of Nitric Oxide (EINO). Figure 12 presents EINO for four equivalence ratios and two injector locations prior to extinction. Results show that EINO values were less than 2.5 g/kg of fuel. EINO did not vary significantly with the location of the injector or the equivalence ratio. This demonstrates the benefits of the porous medium in making NO emission somewhat insensitive to operating parameters. This is due to uniformity as well as low dependence of the reaction zone temperature to operational parameters because of the large thermal mass of the porous medium and its role in evenly distributing fuel. Measurement uncertainties calculated using Student’s t distribution at 95% confidence interval are (0.46 g/kg of fuel. Kaplan and Hall9 measured NOx emissions from a n-heptanefueled porous media radiant burner. Their results indicated that the NOx concentration varied from 15 to 20 ppm, corrected for 3% oxygen. In the present study with kerosene combustion, the NO concentration had a maximum value of 6.5 ppm, corrected to 3% oxygen. Although there is a difference in fuel, a comparison of our results to the results by Kaplan and Hall9 shows that the results are in excellent qualitative agreement. Puri and Gollahalli23 measured the transverse NO concentration profiles of kerosene spray flames without using a porous medium. Measurements were taken at different axial locations for a fuel flow rate of 0.35 g/s and a secondary air velocity of 0.4 m/s. At the centerline of the spray, the authors reported a NO concentration of 22 ppm (corrected to 3% oxygen). In the present study, the NO concentration (corrected to 3% oxygen) varied only 3434
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels
Figure 12. Measured emission indices of NO at four equivalence ratios prior to extinction for two injector locations (co-flow air velocity = 96 cm/s).
from 2 to 7 ppm over an equivalence ratio range of 0.40.7. This demonstrates that the combustion in porous medium reduces the NO emission from spray flames considerably. This is due to enthalpy augmentation of reactants via heat feedback and shorter residence time in the flame zone.
’ DISCUSSION Operating regimes of the flames depend upon the completeness of evaporation and quality of the fuelair mixture available in CPM. Availability of enough flammable mixture in CPM favors interior combustion. If the flammable mixture is not fully prepared until the exit surface of EPM and CPM, surface combustion is most likely to occur. The transition from exterior surface combustion to interior combustion may also be considered equivalent to the flame base approaching the surface closer than the quenching distance of the premixed portion of the flame. In fact, the flame photographs confirm the quenching distance (of the order of 12 mm) of hydrocarbon fuels.24 As the distance of the injector from the upstream surface of the porous medium increases, the evaporation rate in the free jet spray is enhanced, pushing the equivalence ratio toward stoichiometric (Figure 7). That increases the laminar flame velocity at the flame base. The quenching distance is proportional to the laminar flame thickness,25 which varies inversely as the laminar flame velocity.26 Thus, the flame can exist closer to the downstream surface of the CPM, which enhances the heat transfer to the porous medium. The higher temperature at extinction (Figure 7) and higher gas velocity that can be supported at extinction (Figure 5) accompanying the increase in the injector distance can be attributed to this. The extinction of interior flames can be explained by the extinction limit analysis of premixed flames. Choudhuri et al.27 state that the extinction limit, contrasting with the flammability limit, is the condition when the chemical reaction is not selfsustaining. Although the extinction concentration limits have been analyzed for some pure fuels, such as methane and propane,28 and their mixtures with hydrogen,27 the values for kerosene could not be found. Nonetheless, as the extinction limit fuel concentration is lowered with the addition of the higher reactive component in the mixture (H2/CO mixtures in the study by Choudhuri et al.27), one can see that the higher injector distance tending the equivalence ratio toward stoichiometric necessitates higher gas velocity, as seen in Figure 5.
ARTICLE
Interior combustion flames were stabilized at as low of an equivalence ratio as 0.2. At higher effective gas velocities, flames were stabilized only on the surface of the porous medium. Further, a critical effective gas velocity that distinguished interior and surface combustion modes was found. In the present configuration, increasing the effective gas velocity beyond 130 cm/s resulted in surface flames. To develop a general understanding of extinction behavior in porous media, following the approach by Williams25 to analyze the transient phenomena, such as extinction, we adopted a Damk€ohler-number-based approach. This number depends upon a characteristic residence time and chemical time in the porous media. The characteristic preheating residence time of fuel spray was varied by changing the (a) distance between the porous medium and injector and (b) co-flow air flow rate. In the experiments, at a given Damk€ohler number (i.e., injector location and air flow rate), the fuel flow rate was decreased and the extinction limit was determined. Because the air flow rate was held constant, the average gas velocity through the porous medium was also constant. A decrease in the fuel flow rate reduced the flame speed. At some fuel flow rate, referred to as the extinction condition in this paper, the local gas velocity exceeded the flame speed and blew out the flame. A large preheating residence time (or Damk€ohler number) denotes more time for spray evaporation and mixing and, hence, leads to interior combustion. If the residence time is small, evaporation and mixing are not complete and a surface combustion mode is favored. A nominal Damk€ohler number of 5.0 was required to initiate interior combustion mode. As Da was increased, the extinction equivalence ratio decreased. At a given Da, interior flames could be stabilized over a range of equivalence ratios. For instance, at a Da of 6.0, interior flames were stabilized over an equivalence ratio range of 0.20.45. Because the operating conditions are in the fuel-lean regime, the production of more fuel vapors tends to move the mixture toward an equivalence ratio of unity and, thus, increases the reaction rate. Hence, a stable interior combustion mode is established under these conditions, and the heat transfer upstream to EPM became more efficient because of an increased porous medium temperature. Investigations by other researchers5,7 showed that stable flames could be established at the interface between upstream and downstream sections. Similar observations are noted in the present study also. The flame was located near the interface between EPM and CPM. The surface combustion mode is similar to the operation of the flat flame burner but with a thicker porous disk functioning as an evaporator, a mixer, and a flame holder. In interior combustion mode, combustion takes place inside the porous medium and, hence, heat transferred to the solid portion is more efficient. This, in turn, improves upstream heat transfer to EPM and fuel spray. Further, the pollutant emissions of porous media flames are lower than corresponding open flames.
’ CONCLUSION An experimental study of combustion of liquid spray in porous media near lean extinction has been reported in this paper. Stable spray flames can be established both inside and on the downstream exit surface of the CPM. Using a characteristic preheating residence time, a Damk€ohler number approach has been developed to study flame extinction. A higher preheating residence time decreases the extinction equivalence ratio and a stable flame is achieved at as low of an equivalence ratio as 0.2. A minimum 3435
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
Energy & Fuels Damk€ohler number of 5.0 is required to initiate interior combustion. The porous medium surface temperature is uniform within (50 °C. NO emission is insensitive to operating parameters, such as the equivalence ratio and location of the injector. Interior flames are beneficial to effectively transfer combustion heat upstream to enhance evaporation. The porous burner concepts developed in this study have potential applications in gas turbine combustors, air-heating systems, industrial burners, porous chemical reactors, and hybrid burners for biofuels.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses †
Air Liquide R&D, Newark, Delaware 19702, United States.
’ ACKNOWLEDGMENT This research was supported by the U.S. Army Research Laboratory and the U.S. Army Research Office (Grant DAAD 190210082), with Dr. David Mann as the contract monitor. ’ NOMENCLATURE EIi = emission index of species i LH = length of hydraulic tubes in the medium MWi = molecular weight of the species of interest SL = laminar flame speed of kerosene with air at stoichiometric conditions Xi = mole of fraction of the species of interest a = porous medium thickness correction factor dip = distance between the injector exit surface and upstream surface of the porous medium tchem = characteristic chemical time tres = characteristic residence time tp = thickness of the porous medium u = axial gas velocity x = number of carbon atoms per mole of fuel τ = tortuosity δ = laminar flame thickness ’ REFERENCES (1) Howell, J. R.; Hall, M. J.; Ellzey, J. L. Combustion of hydrocarbon fuels within porous inert media. Prog. Energy Combust. Sci. 1996, 22, 121–145. (2) Viskanta, R. Interaction of combustion and heat transfer in porous inert media. In Transport Phenomena in Combustion; Chan, S. H., Ed.; Taylor and Francis Group: London, U.K., 1995; pp 6487. (3) Kamal, M. M.; Mohamad, A. A. Combustion in porous media. Proc. Inst. Mech. Eng., Part A 2006, 220, 487–508. (4) Wood, S.; Harris, A. T. Porous burners for lean-burn applications. Prog. Energy Combust. Sci. 2008, 34, 667–684. (5) Sathe, S. B.; Peck, R. E.; Tong, T. W. Flame stabilization and multimode heat transfer in inert porous media: A numerical study. Combust. Sci. Technol. 1990, 70, 93–109. (6) Lammers, F. A.; de Goey, L. P. H. A numerical study of flash back of laminar premixed flames in ceramic-foam surface burners. Combust. Flame 2003, 133, 47–61. (7) Barra, A. J.; Diepven, G.; Ellzey, J. L.; Henneke, M. R. Numerical study of the effects of material properties on flame stabilization in a porous burner. Combust. Flame 2003, 134, 369–379.
ARTICLE
(8) Mathis, W. M., Jr.; Ellzey, J. L. Flame stabilization, operating range, and emissions for a methane/air porous burner. Combust. Sci. Technol. 2003, 175, 825–839. (9) Kaplan, M.; Hall, M. J. The combustion of liquid fuels within a porous media radiant burner. Exp. Therm. Fluid Sci. 1995, 11, 13–20. (10) Tseng, C.-J.; Howell, J. R. Combustion of liquid fuels in a porous radiant burner. Combust. Sci. Technol. 1996, 112, 141–161. (11) Jugjai, S.; Polmart, N. Enhancement of evaporation and combustion of liquid fuels through porous media. Exp. Therm. Fluid Sci. 2003, 27, 901–909. (12) Jugjai, S.; Pongsai, C. Liquid fuels-fired porous burner. Combust. Sci. Technol. 2007, 179, 1823–1840. (13) Mujeebu, M. A.; Abdullah, M. Z.; Abu Bakar, M. Z.; Mohamad, A. A.; Abdullah, M. K. A review of investigations on liquid fuel combustion in porous inert media. Prog. Energy Combust. Sci. 2009, 35, 216–230. (14) Periasamy, C.; Sankara-Chinthamony, S. K.; Gollahalli, S. R. Experimental evaluation of evaporation enhancement with porous media in liquid-fueled burners. J. Porous Media 2007, 10, 137–150. (15) Periasamy, C.; Gollahalli, S. R. Numerical modeling of evaporation enhancement of aviation-grade kerosene spray in porous media combustors. J. Porous Media 2010, 13, 707–723. (16) Periasamy, C. An experimental and numerical study of evaporation enhancement and combustion in porous media. Ph.D. Dissertation, University of Oklahoma, Norman, OK, 2007; UMI Dissertation Reference Number AAT 3261119. (17) Kaviany, M. Principles of Heat Transfer in Porous Media; SpringerVerlag: New York, 1995. (18) Annamalai, K.; Puri, I. K. Combustion Science and Engineering; CRC Press: Boca Raton, FL, 2007; pp 987992. (19) Babb, M.; Gollahalli, S. R.; Sliepcevich, C. M. Extinguishment of liquid heptane and gaseous propane diffusion flames. J. Propul. Power 1999, 15, 260–265. (20) Khanna, V.; Goel, R.; Ellzey, J. L. Measurements of emissions and radiation for methane combustion within a porous medium burner. Combust. Sci. Technol. 1994, 99, 133–142. (21) Turns, S. R. An Introduction to Combustion: Concepts and Applications, 2nd ed.; McGraw-Hill: New York, 2000. (22) Olikara, C.; Borman, G. L. A computer program for calculating properties of equilibrium combustion products with some applications to I.C. engines. SAE [Tech. Pap.] 1975No. 750468. (23) Puri, R.; Gollahalli, S. R. Effects of location and direction of diluent injection on radiation and pollutant emissions of a burning spray. J. Energy Resour. Technol. 1989, 111, 16–21. (24) Kanury, A. M. Introduction to Combustion Phenomena, 7th ed.; Gordon and Beach Science Publishers: New York, 1992. (25) Williams, F. A. Combustion Theory, 2nd ed.; Benjamin/ Cummings Publishing Co.: Menlo Park, CA, 1985. (26) Glassman, I. Combustion, 2nd ed.; Academic Press: New York, 1987. (27) Choudhuri, A. R.; Subramanya, M.; Gollahalli, S. R. Flame extinction limits of H2CO fuel blends. J. Eng. Gas Turbines Power 2008, 130, 03151-1–03151-7. (28) Law, C. K.; Zhu, D. L.; Yu, G. Propagation and extinction of stretched premixed flames. Symp. (Int.) Combust., [Proc.] 1986, 21, 1419–1426.
3436
dx.doi.org/10.1021/ef1016076 |Energy Fuels 2011, 25, 3428–3436
