Adiabatic Flame Temperature for Controlling the ... - ACS Publications

Subscriber access provided by University of Sussex Library

Combustion

On the adiabatic flame temperature for controlling the macrostructures and stabilization modes of premixed methane flames in a model gas-turbine combustor Ahmed A. Abdelhafez, Medhat Ahmed Nemitallah, Sherif S. Rashwan, and Mohamed A. M. Habib Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.8b01133 • Publication Date (Web): 17 Jun 2018 Downloaded from http://pubs.acs.org on June 18, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

On the adiabatic flame temperature for controlling the macrostructures and stabilization modes of premixed methane flames in a model gas-turbine combustor Ahmed Abdelhafez*, Medhat A. Nemitallah, Sherif S. Rashwan, and Mohamed A. Habib TIC on CCS and Mechanical Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia KEYWORDS: Oxy-combustion; oxygen-enriched air-combustion; premixed flames; stability map; adiabatic flame temperature; gas-turbine combustor.

Premixed oxygen-enriched air-methane flames (CH4/O2/N2) are compared to their oxymethane counterparts (CH4/O2/CO2) in the same model gas-turbine combustor and under identical conditions of oxygen fraction (OF=21–70%vol.) and equivalence ratio (φ=0.2–1.0). The flow rates of non-preheated reactant gases were adjusted for each tested flame to sustain a common bulk velocity at burner throat throughout the whole study, in order to maintain similar flow conditions and turbulence intensities for all isothermal flow fields. The flashback and blowout limits were quantified to identify the combustor stability maps within the OF-φ space. The adiabatic flame temperature (Tad) was also mapped over the same test ranges for both N2 and CO2 flames. The effect of Tad on flame macrostructure and stabilization mode was studied in detail by imaging selected flames. The following novel findings were found to apply to both N2 and CO2 flames at common inlet bulk velocity: Their stable combustion * Corresponding author. Email: [email protected], Tel: +966-13-860-7869. KFUPM Box 2095, Dhahran 31261, KSA.

1 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

zones can both be characterized by Tad only, although they have different Tad maps. Combustion is thus governed mainly by the reaction kinetics (especially near the flashback limits) under similar cold flow conditions. Both N2 and CO2 flames undergo the same changes in macrostructure and stabilization mode as Tad is increased from the blowout limits to the flashback ones. Stable flames of different φ and OF but the same Tad have identical shapes, which shows the direct dependence of flame macrostructure and stabilization mode on Tad under similar cold flow conditions. Both N2 and CO2 stability maps can be subdivided into sub-zones based on Tad only, where each zone has a single prevailing flame macrostructure irrespective of φ and OF. This is yet another proof that Tad is an excellent tool for predicting flame macrostructure at constant inlet bulk velocity. Based on these findings, this study recommends to design and operate future oxy-fuel gas-turbine combustors based on Tad (and not OF or φ), particularly at high and medium loads away from blowout, following the existing common practice among manufacturers of conventional lean-premixed air-fuel gas turbines to quantify combustor performance in terms of Tad.

Nomenclature A

Cross-sectional area of burner throat [m2]

CH4

Methane

CO2

Carbon dioxide

D

Burner throat diameter [m]

DACRS

Dual Annular Counter-Rotating Swirlers

DLE

Dry Low Emissions

H2O

Water vapor

LPM

Lean premixed

Mass flow rate [kg/s]

M

Molecular weight [kg/kmol]

N2

Nitrogen

O2

Oxygen


Page 2 of 26

Page 3 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

OF

Oxygen fraction, volumetric percentage of O2 in the O2/X diluted oxidizer

p

Combustor pressure [kPa]

PD

Combustor power density [MW/m3/bar]

Ru

Universal gas constant [kJ/kmol/K]

Re

Throat Reynolds number

T

Absolute reactant temperature [K]

Tad

Adiabatic flame temperature [K]

v

Bulk throat velocity [m/s]

va

Axial component of bulk throat velocity [m/s]

X

Diluent gas, either N2 or CO2

y

Mole fraction

Greek symbols φ

Equivalence ratio

ρ

Density [kg/m3]

µ

Dynamic viscosity [kg/m/s]

Subscripts mix

Reactant mixture (CH4+O2+X)

i

Each mixture constituent

1.

Introduction

Almost all of today’s combustion applications use air as oxidizer. The resulting exhaust consists primarily of CO2, H2O, O2, and N2. CO2 is a greenhouse gas, and its emissions are a global concern because of the escalating climate-change problem. Fossil fuels are the main source of energy in many countries and will remain so for several decades [1], but the resulting CO2 emissions are inescapable. To prevent global temperature from rising more than 2 °C, international conventions, like the Paris Agreement, have strongly recommended carbon capture and sequestration as an irreplaceable solution [2]. The separation of CO2 from the exhaust stream of air-combustion is, however, complicated and costly [3]. Oxy-fuel combustion, on the other hand, uses pure oxygen as oxidizer, so the resulting products are


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

primarily CO2 and H2O. The former can thus be captured at significantly lower cost after condensing the latter. The inherent absence of air-based N2 also offers two additional advantages: First, nitrogen oxides (NOx) emissions are drastically reduced, and second, the exhaust volume is considerably smaller, which allows for using compact and more efficient treatment equipment. Switching from air to oxy-combustion, however, entails different degrees of freedom for combustor operability. Air is an ample low-cost oxidizer, so lean operation is typical in airfuel gas turbines to control the combustor exit temperature. The separation of air to generate oxygen, on the other hand, poses an energy penalty [4], so oxy-fuel applications need to be operated near stoichiometry to utilize both fuel and oxygen efficiently. However, flames of fuel and pure O2 have extreme temperatures beyond the metallurgical limits of most materials. Dilution is thus implemented to control flame temperature by recirculating part of the exhaust flow back into the combustor. The reactants are thus fuel plus an O2/CO2 diluted oxidizer. The replacement of N2 with CO2 as diluent also affects the flame characteristics, primarily because N2 and CO2 have different properties, such as heat capacity, thermal conductivity, mass diffusivity, and dynamic viscosity [5–7]. CO2 is known to have an adverse effect on the reaction rates, flame speed, and combustion efficiency [8,9]. Consequently, oxy-fuel flames have narrower stability windows compared to air-fuel flames [10,11]. Kutne et al. [12] attributed the reduced stability of oxy-fuel flames to the larger heat capacity of CO2 compared to N2, which reduces the flame temperature and burning velocity. This conclusion was substantiated by the observation that the inner recirculation zones of oxy-fuel flames exist further downstream compared to those of air-fuel ones. Amato et al. [13] examined the extinction mechanisms of premixed flames and showed that CO2 adversely affects the reaction kinetics. This finding was also affirmed by Jerzak and Kuznia [14] who investigated the upper and lower flammability limits of natural-gas combustion in different oxidizer environments, i.e., O2/CO2 and O2/N2. The retarded kinetics and lower burning velocities of 4 Environment ACS Paragon Plus

Page 4 of 26

Page 5 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

the CO2 flames allowed them to resist flashback better than the N2 ones by being stable at greater oxygen fractions before the onset of flashback. The minimum oxygen fraction of O2/CO2 diluted oxidizer needed to replicate the stability of air-fuel flames (21% O2) was found in the literature to be in the range of 30% [15] to 40% [10]. Song et al. [16] reported that 36% is needed to match the temperature profiles. Most of the literature studies on oxy-fuel combustion reported their findings of flame characteristics, stability, and reaction kinetics based on the effect of oxygen fraction in the O2/CO2 diluted oxidizer. The authors of this present work, however, recently published a study of CH4/O2/CO2 flames [17], where they took a different approach in analyzing flame stability and shape based on the adiabatic flame temperature (Tad). It was shown that Tad is the parameter that best represents the stability map of the examined premixed swirl-stabilized model gas-turbine combustor at constant inlet bulk velocity. The flashback limit was observed to coincide with a contour of constant Tad within the investigated two-dimensional space of oxygen fraction (OF) and equivalence ratio (φ). Flames of the same Tad were found to have almost identical sizes and shapes, despite having different OF and φ. This clearly demonstrated that both combustor operability and flame characteristics should be analyzed based on Tad, instead of φ or OF separately, because Tad is an umbrella parameter that combines the individual effects of φ and OF and best describes the prevailing effect of reaction kinetics under similar cold flow conditions. Similar analyses of oxy-fuel flames based on Tad were rarely performed in the literature. Shroll et al. [18] investigated stoichiometric swirl-stabilized premixed CH4/O2/CO2 and CH4/air flames with focus on their dynamic stability characteristics. Tad was controlled through OF. It was found that Tad is the parameter that governs the transition from one stability mode to another in both CH4/O2/CO2 and CH4/air flames, although they have different flame speeds at a given Tad. High-speed imaging was used to confirm this finding by showing comparable vortex breakdown modes indicative of analogous turbulent flame 5 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

geometries that change with Tad. Jourdaine et al. [19] performed an experimental study of the stabilization mechanisms of premixed CH4/O2/CO2 and CH4/air flames. They revealed that, at the same swirl number, CO2 and N2 flames of the same φ have similar shapes while maintaining common Tad and incoming jet velocity. They also reported that no combustor design modifications are needed to switch from air to oxy-fuel operation. It is worth noting here that manufacturers of conventional LPM DLE air-fuel gas turbines commonly quantify combustor performance in terms of Tad (and not φ), whether on the scale of the entire combustor, of each ring/can separately, or even of each lit fuel-air premixer. Another common practice among gas-turbine manufacturers is to consider constant pressure drop across the combustor headend, which maintains constant bulk velocity of reactant mixture at the premixer throat and, consequently, creates a safe margin to flashback. This practice has been implemented by the present authors for CH4/O2/CO2 flames [17] and is employed here for CH4/O2/N2 flames as well. The same bulk throat velocity of 5.2 m/s is considered again here to have direct comparisons of combustor stability and flame macrostructure. Both studies were performed on the same model gas-turbine combustor, which is based on the LPM DLE DACRS technology of operational air-fuel gas turbines [20,21]. The choice of this bulk throat velocity is also based on the desired power density of the present combustor, which is within the range of typical operation of industrial gas turbines, i.e., 3.5–20 MW/m3/bar [22]. In addition to investigating how the combustor stability limits correlate with Tad, a detailed analysis of flame macrostructure is performed here to highlight the effect of Tad. The macrostructures of premixed CH4/air flames were examined in recent literature studies [14,23–25] using dump combustor geometries similar to the one considered here. The flame was found to assume one of four distinct shapes based on φ. The same classification is implemented here for both the CH4/O2/N2 and CH4/O2/CO2 flame sets; however, the present analysis is based on Tad as an umbrella parameter that combines both effects of φ and OF. 6 Environment ACS Paragon Plus

Page 6 of 26

Page 7 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

2. 2.1.

Experimental Setup and Test Conditions Combustor and instrument specifications

This study compares the combustion characteristics, stability limits, and macrostructures of confined swirl-stabilized premixed CH4/O2/N2 and CH4/O2/CO2 flames experimentally at atmospheric pressure. Both flame sets were examined in the same model gas-turbine combustor for a one-to-one comparison within the same geometry, which is based on the LPM DLE DACRS technology of operational air-fuel gas turbines [20,21]. A schematic of this combustor is shown in Figure 1 to show how the O2/N2 and O2/CO2 diluted oxidizers were furnished and introduced. More details are available in Abdelhafez et al. [17]

Figure 1. Schematic drawing of the experimental setup. 7 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 26

The O2/N2 diluted oxidizer was furnished by supplying non-preheated flows of air and enrichment oxygen to the combustor after metering them to the desired oxygen fraction. A common bulk throat velocity of 5.2 m/s was maintained for all N2 and CO2 flames. All gases were supplied from compressed cylinders, except for air, which was supplied from a remote two-stage reciprocating Ingersoll-Rand compressor with inter- and after-cooling and a big buffer tank. This cooling, together with the large line length and the inline pressure/flow controls, assured that air reached the rig at ambient temperature. More information about the equipment can be found in past publications by the authors [17,22]. Flame macrostructure was analyzed by capturing images of visual flame appearance using a 10-Mega-pixel NIKON-D3100 camera with 1/60-s shutter speed, 5.6 f-stop, and 1600 ISO. 2.2.

Test conditions

The present study compares premixed CH4/O2/N2 and CH4/O2/CO2 flames at fixed flame-inlet bulk velocity. Since the use of different diluents (N2 vs. CO2) is expected to induce significant differences in the combustion characteristics, operability limits, and flame macrostructure, both N2 and CO2 are treated here as reactants. In other words, the terms “bulk velocity” and “bulk Re” refer to these quantities of the CH4/O2/N2 and CH4/O2/CO2 reactant mixtures as a whole and not of CH4+O2 only. Abdelhafez et al. [17] provided a detailed explanation of the approach taken to fix the bulk throat velocity at 5.2 m/s for all examined CH4/O2/CO2 flames. If this approach is generalized for a diluent X ≡ either CO2 or N2, the following species flow rates are obtained.

= 16

= 64

= 2

1 +

+

(1)

(2)

Thus, = 88

(3)


(3a)

Page 9 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

and ! = 56

(3b)

where OF and φ were examined over the ranges 21–70% and 0.2–1.0, respectively. An important remark to make from Equation 1 is that does not depend on the diluent (X) properties, i.e., the combustor consumes the same amount of fuel in both test sets for common φ and OF. is used here to define the combustor power density (PD) as

#$ =

%&'( × &

× *+,-./ +0 1-2345 6+7087/./74

(4)

where HC is the standard tabulated enthalpy of combustion of methane (55.5 MJ/kg) [26]. It is worth noting here that the choice of 5.2-m/s bulk throat velocity is based on the desired power density of the present combustor, which is within the range of typical operation of industrial gas turbines, i.e., 3.5–20 MW/m3/bar [22]. As mentioned earlier, airflow was enriched with pure oxygen to generate the O2/N2 diluted oxidizer of desired oxygen fraction. Using ! from Equation 3b, combined with the fact that air is 76.7% by mass nitrogen, the necessary flow rate of air was calculated from

9: = ! ⁄0.767

(5)

The flow rate of enrichment O2 was thus determined from Equations 2, 3b, and 5 as

?@ = + ! − 9:

(6)

The bulk dynamic viscosity of reactant mixture was calculated using the formula [27]

B%9C =

∑ EF GF HIF

(7)

∑ E F H IF

The bulk density of reactant mixture was determined from

J%9C =

∑ %F

(8)

∑K%F⁄IF L

The Reynolds number at the burner throat was thus calculated from

MN = O $ J%9C ⁄B%9C

(9)

The adiabatic flame temperature (Tad) was also mapped for both the CH4/O2/N2 and CH4/O2/CO2 flame sets over the same examined ranges of φ and OF. Abdelhafez et al. [17] 9 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

estimated the Tad of CO2 flames using a simplified approach that is extended here to N2 flames as well, since the associated error is expected to be even smaller with the lower concentration of CO2 in the exhaust of N2 flames.

3. 3.1.

Results and Discussion Combustor stability maps

Nitrogen and oxygen have very similar properties, i.e., comparable densities and heat capacities in particular. Changing the oxygen fraction of O2/N2 diluted oxidizer is thus expected not to affect the apparent oxidizer properties much. Carbon dioxide, on the other hand, has significantly different properties; the density and heat capacity of CO2 are both about 1.5 times their O2 and N2 counterparts at the same pressure and temperature. Oxygen fraction is thus expected to play a more significant role in oxy-fuel flames. These facts affect key combustion parameters like the adiabatic flame temperature (Tad), which is of pivotal importance in the analyses of premixed flames as will be shown here. Figure 2 shows the stability maps (flashback and blowout limits) of the examined model gasturbine combustor geometry for both the CH4/O2/N2 and CH4/O2/CO2 flames at fixed flameinlet bulk velocity (similar cold flow conditions). The color backgrounds are the respective Tad maps plotted against the same color scale for a one-to-one comparison. Flashback data points are indicated by triangular symbols in Figure 2, while the blowout points are depicted using circular symbols. The stable combustion zone on each map is the area between the limits, whereas the outside areas are the flashback and blowout zones on the top and bottom, respectively. The reason behind the scattered nature of blowout data points is that intermittent flame lifting and reattachment occurred prior to blowout, which made it difficult to quantify this limit accurately. The flashback limit was thus more distinctly identifiable than the blowout one during experiments.


Page 10 of 26

Page 11 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 2. Combustor stability maps plotted against the respective contours of constant Tad for CH4/O2/N2 flames (top) and CH4/O2/CO2 flames (middle). Both background maps of Tad share a common color scale. The bottom plot compares the stability limits directly. The bulk throat velocity is fixed at 5.2 m/s throughout.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Two preliminary observations can be made from Figure 2. Firstly, CO2 flames always have lower Tad than their N2 counterparts at the same φ and OF, which is attributed to the higher heat capacity of CO2 compared to N2. Secondly, the slopes of constant Tad contours [P⁄PQR L S ] are steeper in CO2 flames, indicating a greater influence of OF on Tad in oxyfuel flames as mentioned earlier, again because of the higher heat capacity of CO2 compared to N2. Figure 2 also provides one of the primary findings of this study, namely that flame flashback occurs consistently at constant Tad for both the CH4/O2/N2 and CH4/O2/CO2 flame sets at constant flame-inlet bulk velocity. Although the contours of constant Tad have different slopes in the N2 and CO2 maps as discussed earlier, both N2 and CO2 flashback limits occur consistently at constant Tad. This was reported recently by the authors for CH4/O2/CO2 flames only [17], but the novelty of the present work is proving that this finding is universal and applies to CH4/O2/N2 flames as well. This behavior can be attributed to the fact that the flashback limits are controlled mainly by reaction kinetics under similar cold flow conditions (constant inlet bulk velocity). A delicate balance between flame speed and flow velocity exists at the flashback limit. The constant bulk velocity examined here allows for inferring that each flashback limit occurs at constant flame speed, which further proves the dominant role of reaction kinetics. The flame speed was shown to depend directly on Tad [17,28–30], which completes the picture and explains why both N2 and CO2 flashback limits occur at constant Tad. It can also be observed in Figure 2 that Tad is generally the optimum parameter to describe the stability maps of premixed natural-gas combustors operating at constant inlet bulk velocity. Some deviation is observed near the blowout limit, which indicates that the blowout mechanism is not function of Tad only but depends on other flow parameters as well. The aforementioned intermittent lifting and reattachment of the flame prior to blowout is another proof of this.


Page 12 of 26

Page 13 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Based on this analysis, it can be recommended to design and operate premixed oxy-fuel gasturbine combustors based on Tad, particularly at high and medium loads away from blowout, similar to the existing state-of-the-art LPM air-fuel gas-turbine combustors. These are commonly designed and operated based on Tad, whether on the scale of the entire combustor, of each ring/can separately, or even of each lit fuel-air premixer. The flame temperature of an air-fuel combustor is controlled through the equivalence ratio, of course, but gas-turbine manufacturers design, operate, and evaluate their combustors based on flame temperature and not equivalence ratio. The present study thus recommends the same practice for oxy-fuel gas turbines as well, especially that maintaining constant flame-inlet bulk velocity is another common practice in LPM air-fuel gas turbines. The combustor is typically operated at constant pressure drop across its headend, which maintains constant bulk velocity of reactant mixture at the premixer throat and, consequently, creates a safe margin to flashback. Multiple other important observations can also be made from Figure 2 regarding the reaction kinetics of CH4/O2/N2 vs. CH4/O2/CO2 flames:

•

The 2320-K Tad of a stoichiometric CH4/air flame is matched at OF=30% in a stoichiometric CH4/O2/CO2 flame, which agrees with the findings of Ditaranto et al. [15], who reported this 30% OF value but did not explain it in terms of Tad

•

Stable CH4/air flames (OF=21%) were attainable for φ > 0.55, whereas the same OF was beyond the blowout limit of oxy-flames. This agrees directly with the findings of all past studies that investigated the minimum OF needed to match the stability of airfuel flames [10,15,16]

•

CO2 flames blowout at higher φ than N2 ones for any given OF, i.e., CO2 flames generally exhibit inferior resistance to blowout compared to N2 ones, which narrows the stable combustion zone of oxy-flames significantly and shows the negative impact


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

of CO2 on reaction kinetics and blowout. This agrees with the findings of Rashwan et al. [10], Kutne et al. [12], and Amato et al. [13] •

Flashback of CO2 flames occurs at higher OF compared to N2 flames of the same φ. This is further evidence of the retarded chemical kinetics of CO2 flames, as more oxygen is needed to trigger flashback, which is in direct agreement with the findings of Jerzak and Kuznia [14]

•

It is only at high oxygen fractions (70%) where the CO2 and N2 flames exhibit almost identical flashback and blowout equivalence ratios, because the concentrations of CO2 and N2 in their diluted oxidizers have dropped enough to undermine the effect of their different properties

To further prove that Tad is the parameter that best describes the stability maps of premixed natural-gas combustors (CH4/O2/CO2 and CH4/O2/N2) operating at constant flame-inlet bulk velocity, the stability maps of Figure 2 were replotted against the contours of constant power density (PD, Figure 3) and throat Reynolds number (Re, Figure 4). Recall from the analysis of governing equations that the N2 and CO2 flame sets have identical PD maps, as seen in Figure 3, because the combustor consumes the same amount of fuel in both sets for a given common combination of φ and OF.

Figure 3. Combustor stability maps plotted against the contours of constant PD for CH4/O2/N2 flames (left) and CH4/O2/CO2 flames (right). The bulk throat velocity is fixed at 5.2 m/s throughout. Both flame sets have identical PD maps in the background. 14 Environment ACS Paragon Plus

Page 14 of 26

Page 15 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 4. Combustor stability maps plotted against the respective contours of constant bulk throat Re for CH4/O2/N2 flames (left) and CH4/O2/CO2 flames (right). The bulk throat velocity is fixed at 5.2 m/s throughout. Each background map of Re has its own color scale.

It can be seen in Figure 3 that the flashback limit of CH4/O2/N2 flame set follows a contour of constant PD, but the same does not apply to CH4/O2/CO2 flames, which proves that, unlike Tad, PD is not the optimum parameter for describing the stability maps. Interestingly, the observed trend of N2 flashback limit is merely an artifact of the physical properties of O2 and N2. As discussed earlier, O2 and N2 have very similar densities and heat capacities. So, changing the OF in O2/N2 diluted oxidizer at constant φ changes Tad because the concentrations of exhaust O2 and N2 change and not because of the minute difference in heat capacities of O2 and N2. The PD map of CH4/O2/N2 set is also primarily affected by species concentrations because of the comparable densities of O2 and N2. This explains why the Tad and PD maps look so similar. To demonstrate this, Figure 5 replots the stability map of CH4/O2/N2 flames against the contours of the ratio of Tad to PD. It can be clearly seen that the map of this ratio is almost identical to those of Tad and PD; consequently, the N2 flashback limit follows a contour of constant Tad-to-PD ratio.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5. Combustor stability map of CH4/O2/N2 flames plotted against the contours of constant ratio of Tad to PD.

By comparison, O2 and CO2 have different properties, which results in dissimilar Tad and PD maps for the CH4/O2/CO2 set. Therefore, the observed dependence of flame flashback on Tad (Figure 2) is not observed with PD (Figure 3) as well. This study thus agrees with the findings of Abdelhafez et al. [17], namely that combustor flashback does not always occur at constant fuel flow rate. The effect of O2 vs. CO2 or N2 densities is also evidenced in the Re maps of Figure 4. The slopes of constant Re contours [P⁄PQR L? ] are significantly steeper in CO2 flames, indicating that the Re map is dominated by OF in oxy-fuel flames, similar to what was observed earlier with Tad. This is attributed to the higher density of CO2. It can also be seen in Figure 4 that none of the stability limits in either map follows a contour of constant Re, which further proves that Tad is indeed the optimum parameter to describe the stability maps of premixed natural-gas combustors operating at constant flame-inlet bulk velocity. The effect of Re was examined here because a previous study by the authors [31] revealed that the stability and structure of coaxial non-premixed CH4/O2/CO2 and CH4/O2/N2 flames were affected by the Reynolds number of diluted oxidizer.


Page 16 of 26

Page 17 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

3.2.

Flame-macrostructure analysis based on Tad

3.2.1. Flame classification according to stabilization mode Past studies have identified four distinct stabilization modes of premixed CH4/air flames, which are described in Table 1. Figure 6 shows the images of selected CH4/O2/N2 flames at OF=40%, which were captured to visualize the mode classification of Table 1. Flame A (Tad = 2094 K) is a tall unstable flame that assumes shape I (single cone) near the blowout limit with intermittent lifting and reattachment, because this limit is not function of Tad only, as discussed earlier. Increasing Tad induced a transition to shape II (double cone, flame B, Tad = 2241 K). This metastable shape was found to exist within a narrow range of Tad with intermittent transition back to shape I. Further increase of Tad stabilized the flame within the corner recirculation zone and boundary layer at the quartz walls (shape III, flame C, Tad = 2382 K). This cup shape was observed to enjoy consistent stability and to span a wider range of Tad. Flame D (Tad = 2516 K) was carefully selected as it visualizes the upper limit of zone III, where the flame is considerably shorter, adopting a plate shape that is stabilized within the corner recirculation zone only. Intermittent transition to shape IV was observed here. Further increase of Tad made the flame acquire shape IV (flame E, Tad = 2644 K) consistently, which is the typical V-shape expected for swirl-stabilized flames. Approaching the flashback limit (flame F, Tad = 2767 K), the flame shape did not change much, but the audible noise increased, indicative of combustion dynamics prior to flashback. This mode classification of flames based on their Tad will be used to subdivide the stable combustion zone into four sub-zones I – IV, bounded by contours of constant Tad, as seen on the stability map in Figure 6. Each zone is named after the flame shape that prevails throughout it. The upcoming analysis will then answer the question: Do all flames within each zone have a common shape irrespective of φ and OF? In other words, is the knowledge of Tad only enough to accurately predict and characterize the macrostructures and stabilization modes of stable flames? 17 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 26

Table 1. Shapes and stabilization modes of premixed CH4/air flames from past studies Flame Shape

Description / Stabilization Mode

I

Single cone

At low φ. Flame stabilized within the inner shear layer [24,25]

II

Double cone

Flame stabilized within both the inner and outer shear layers [24,25]

III Corner-stabilized

Flame stabilized within the corner recirculation zone and the boundary layer at the combustor walls [14]

IV Swirl-stabilized

Typical V-shape of swirl-stabilized flames [23]

V-shape

CH4/O2/N2 Flames at OF=40% A

B

C

D

E

F

Shape I

Shape II

Shape III

Shape III – IV

Shape IV

Shape IV

2094 K

2241 K

2382 K

2516 K

2644 K

2767 K

Figure 6. Top: Visual images of six CH4/O2/N2 flames (A – F) at OF=40%; Tad is listed for each flame. Bottom: Combustor stability map highlighting data points A – F and the four subzones of the stable combustion zone. The bulk throat velocity is fixed at 5.2 m/s throughout. 18 Environment ACS Paragon Plus

Page 19 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

3.2.2. Dependence of flame macrostructure on Tad The effect of Tad on flame macrostructure and stabilization mode will be analyzed from three different aspects: changing Tad by varying φ at constant OF=60%, changing Tad by varying OF at constant φ=0.5, and keeping Tad constant by varying both OF and φ simultaneously. All three approaches will be performed on both CH4/O2/N2 and CH4/O2/CO2 flames. The selected data points for which flame images were captured to execute this analysis are shown on the stability maps of Figure 7. Figure 8 compares the CO2 flame images side by side to their N2 counterparts, and the following observations and conclusions can be made from Figure 8:

•

The same distinct 4 shapes observed with the N2 flames are also observed with the CO2 ones, namely, single cone (shape I), double cone (shape II), corner-stabilized (shape III), and the swirl-stabilized V-flame (shape IV). This agrees with the findings of Jourdaine et al. [19]

•

Flames (2 & 8)CO2 have a common Tad of ~2000 K, and both have shape II. Flames (3 & 7)CO2 at Tad ≈ 2170 K have shape III. Even the transitional plate-shaped flames (4 & 6)CO2 have a common Tad of ~2340 K. The same analogy is observed for flames (4 & 5)N2 at Tad ≈ 2690 K, (3 & 6)N2 at Tad ≈ 2500 K, and (2 & 7)N2 at Tad ≈ 2300 K. The only exception is flame 8N2 (shape II) vs. flame 1N2 (shape I), but even this discrepancy is justifiable because the blowout limit is not function of Tad only

•

It is very important to note here that the two flames within each of these “couples” have different OF and φ but the same Tad. This singles out the direct dependence of flame macrostructure on reaction kinetics and Tad at constant inlet bulk velocity and agrees directly with the findings of Shroll et al. [18]

•

Approaching the flashback limit, flames 5CO2, 4N2, and 5N2 have shape IV, and approaching the blowout limit, flames 1CO2 and 1N2 have shape I


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

All these observations can be combined to answer the aforementioned questions and provide two novel conclusions for stable flames (away from blowout) at constant inlet bulk velocity. First, the knowledge of Tad alone is indeed enough to accurately predict and characterize flame macrostructure and stabilization mode. Second, the stable combustion zone of a premixed natural-gas combustor can be subdivided into Tad-based sub-zones, where each zone has a single prevailing flame macrostructure irrespective of φ and OF, because Tad is an umbrella parameter that combines the individual effects of φ and OF and best describes the prevailing effect of reaction kinetics under similar cold flow conditions. Figure 7 illustrates these sub-zones for the CH4/O2/N2 and CH4/O2/CO2 flame sets. The soundness of these conclusions can be further attested by comparing more than just two flames at the same Tad. Flames (3, 6, 9-13)N2 are all zone-IV flames at Tad ≈ 2500 K, while flames (9-13)CO2 are all III–IV transition flames at Tad ≈ 2380 K, see Figure 7. The images of these flames are portrayed in Figure 9. It is undeniably clear that all seven N2 flames have identical shapes at common Tad, although the OF ranges from 30% to 66% and φ ranges from 0.37 to 0.79. The same applies to the five CO2 flames with OF=50–70% and φ=0.38–0.58. This highlights the pivotal importance of Tad as the sole parameter needed to predict, characterize, and control the macrostructure and stabilization mode of stable premixed natural-gas flames (CH4/O2/N2 and CH4/O2/CO2) at constant inlet bulk velocity. A concluding remark to be made here is that the findings of this study trigger the recommendation to design and operate future oxy-fuel gas-turbine combustors based on Tad and not dilution ratio or percent exhaust recirculation. Tad can be easily estimated for any given set of operating conditions. With the knowledge of Tad only, the shapes and stabilization modes of stable flames can be easily inferred. The heatshields at combustor headend can thus be designed accordingly, and the interactions of adjacent flames (and the resulting instabilities) in multi-premixer combustors can also be predicted.


Page 20 of 26

Page 21 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

# 1 2 3 4 5 6 7 8 9 10 11 12 13

# 1 2 3 4 5 6 7 8 9 10 11 12 13

OF 60%

54% 48% 42% 36% 66% 54% 42% 36% 30%

OF

60%

55% 50% 45% 70% 65% 60% 55% 50%

φ 0.30 0.35 0.40 0.45 0.50

0.37 0.45 0.57 0.66 0.79

φ 0.30 0.35 0.40 0.45 0.50

0.38 0.42 0.46 0.52 0.58

Tad [K] 2080 2297 2499 2688 2694 2509 2311 2097

2500

Tad [K] 1806 1995 2173 2342 2501 2336 2169 2001

2380

Figure 7. Combustor stability maps highlighting the selected data points for which flame images were captured to analyze flame shape. The bulk throat velocity is fixed at 5.2 m/s throughout.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 26

CH4/O2/N2 Flames

CH4/O2/CO2 Flames

OF = 60%

OF = 60%

1

2

3

4

1

2

3

4

5

Zone I

Zone III

Zone IV

Zone IV

Zone I

Zone II

Zone III

Zone III – IV

Zone IV

2080 K

2297 K

2499 K

2688 K

1806 K

1995 K

2173 K

2342 K

2501 K

φ = 0.5

φ = 0.5

8

7

6

5

8

7

6

5

Zone II

Zone III

Zone IV

Zone IV

Zone II

Zone III

Zone III – IV

Zone IV

2097 K

2311 K

2509 K

2694 K

2001 K

2169 K

2336 K

2501 K

Figure 8. Visual images of CH4/O2/N2 flames (left) and CH4/O2/CO2 flames (right) at OF=60% (top) and φ=0.5 (bottom). The CO2 images were adopted from Abdelhafez et al. [17]. Tad is listed for each flame.


Page 23 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

CH4/O2/N2 Flames Tad ≈ 2500 K, Zone IV

9

3

10

6

11

12

13

CH4/O2/CO2 Flames Tad ≈ 2380 K, Zone III – IV

9

10

11

12

13

Figure 9. Visual images of CH4/O2/N2 flames (top) and CH4/O2/CO2 flames (bottom) at constant Tad. The CO2 images were adopted from Abdelhafez et al. [17]. OF and φ are listed for each flame. Conclusions Premixed oxygen-enriched air-methane flames (CH4/O2/N2) were compared to their oxymethane counterparts (CH4/O2/CO2) in the same model gas-turbine combustor and under identical conditions of oxygen fraction (OF=21–70%) and equivalence ratio (φ=0.2–1.0). The swirl-stabilized combustor mimics the geometry of operational air-fuel gas-turbine combustors implementing the LPM DLE DACRS technology. The flow rates of nonpreheated reactant gases were adjusted for each tested flame to sustain a common bulk velocity of 5.2 m/s at burner throat throughout the whole study, in order to maintain similar flow conditions and turbulence intensities for all isothermal flow fields. The artificial air (O2/N2) was produced by enriching atmospheric air with pure oxygen to the desired level of oxygen fraction. The flashback and blowout limits were quantified to identify the combustor 23 Environment ACS Paragon Plus

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

stability map within the OF-φ space. The adiabatic flame temperature (Tad) was also mapped over the same test ranges for both N2 and CO2 flames. The effect of Tad on flame macrostructure and stabilization mode was studied in detail by imaging selected flames. The following novel conclusions were made for both N2 and CO2 flames at common inlet bulk velocity. a) Their stable combustion zones can both be characterized by Tad only, although they have different Tad maps. The stable combustion zones are thus governed mainly by the reaction kinetics (especially near the flashback limits) under similar cold flow conditions. b) Both N2 and CO2 flames undergo the same changes in macrostructure and stabilization mode as Tad is increased from the blowout limits to the flashback ones. c) Stable flames of different φ and OF but the same Tad have identical shapes, which proves that the knowledge of Tad alone is enough to accurately predict and characterize flame macrostructure and stabilization mode under similar cold flow conditions. d) Both N2 and CO2 stability maps can be subdivided into sub-zones based on Tad only, where each zone has a single prevailing flame macrostructure irrespective of φ and OF. This is yet another proof that Tad is an excellent tool for predicting flame macrostructure at constant inlet bulk velocity. Based on these findings, the present study recommends to design and operate future oxy-fuel gas-turbine combustors based on Tad (and not OF or φ), particularly at high and medium loads away from blowout, following the existing common practice among manufacturers of conventional LPM DLE airfuel gas turbines to quantify combustor performance in terms of Tad.

Acknowledgement The authors would like to acknowledge the support of King Fahd University of Petroleum and Minerals (KFUPM) under DSR project IN161027. The support of SABIC to this work through project ME 2394 is also acknowledged.


Page 24 of 26

Page 25 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

References [1]

[Online], IEA, 20 Years of Carbon Capture and Storage 2016, https://www.iea.org/publications/freepublications/publication/20YearsofCarbonCaptur eandStorage_WEB.pdf.

[2]

[Online], UN, Paris Agreement 2015, https://unfccc.int/files/essential_background/convention/application/pdf/english_paris_ agreement.pdf.

[3]

Li, H.; Yan, J.; Yan, J.; Anheden, M. Appl. Energy 2009, 86, 202–213.

[4]

Cao, Y.; He, B.; Ding, G.; Su, L.; Duan, Z. Energy Fuels 2016, 30, 9953−9961.

[5]

Nemitallah, M. A. J. Nat. Gas Sci. Eng. 2016, 28, 61–73.

[6]

Nemitallah, M. A.; Habib, M. A.; Mezghani, K. Energy 2015, 84, 600–611.

[7]

Habib, M. A.; Salaudeen, S. A.; Nemitallah, M. A.; Ben-Mansour, R.; Mokheimer, E. M. A. Energy 2016, 96, 654–665.

[8]

Williams, T. C.; Shaddix, C. R.; Schefer, R. W. Combust. Sci. Technol. 2007, 180, 64– 88.

[9]

Zhang, J.; Mi, J.; Li, P.; Wang, F.; Dally, B. B. Energy and Fuels 2015, 29, 4576– 4585.

[10] Rashwan, S. S.; Ibrahim, A. H.; Abou-Arab, T. W.; Nemitallah, M. A.; Habib, M. A. Appl. Energy 2016, 169, 126–137. [11] Nemitallah, M. A.; Rashwan, S. S.; Mansir, I. B.; Abdelhafez, A.; Habib, M. A. Energy & Fuels 2018, acs.energyfuels.7b03607. [12] Kutne, P.; Kapadia, B. K.; Meier, W.; Aigner, M. Proc. Combust. Inst. 2011, 33, 3383– 3390. [13] Amato, A.; Hudak, B.; D’Carlo, P.; Noble, D.; Scarborough, D.; Seitzman, J.; Lieuwen T. J. Eng. Gas Turbines Power 2011, 133, 061503. [14] Jerzak, W.; Kuźnia, M. J. Nat. Gas Sci. Eng. 2016, 29, 46–54. [15] Ditaranto, M.; Hals, J.; Combust. Flame 2006, 146, 493–512. [16] Song, Y.; Zou, C.; He, Y.; Zheng, C. Int. J. Heat Mass Transf. 2015, 86, 622–628. [17] Abdelhafez, A.; Rashwan, S. S.; Nemitallah, M. A.; Habib, M. A. Appl. Energy 2018, 215, 63–74. [18] Shroll, A. P.; Shanbhogue, S. J.; Ghoniem, A. F. J. Eng. Gas Turbines Power 2012, 134, 051504. [19] Jourdaine, P.; Mirat, C.; Caudal, J.; Lo, A.; Schuller, T. Fuel 2017, 201, 156–164. [20] Joshi, N. D.; Epstein, M. J.; Durlak, S.; Marakovits, S.; Sabla, P. E. Interational Gas Turbine and Aeroengine Congress and Exposition 1994, 94-GT-253, 1–9. [21] Huang, Y.; Yang, V. Progress in Energy and Combustion Science 2009, 35, 293–364.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[22] Nemitallah, M. A.; Habib, M. A. Appl. Energy 2013, 111, 401–415. [23] Taamallah, S.; Shanbhogue, S. J.; Ghoniem, A. F. Combust. Flame 2016, 166, 19–33. [24] Taamallah, S.; Labry, Z. A.; Shanbhogue, S. J.; Ghoniem, A. F. Proc. Combust. Inst. 2015, 35, 3273–3282. [25] Taamallah, S.; Labry, Z. A.; Shanbhogue, S. J.; Habib, M. A.; Ghoniem, A. F. J. Eng. Gas Turbines Power 2015, 137, 071505. [26] Cengel, Y. A.; Boles, M. A. Thermodynamics: an Engineering Approach 8th Edition, 2015. [27] Carr, N. L.; Kobayashi, R.; Burrows, D. B. J. Pet. Technol. 1954, 6, 47–55. [28] Kaskan, W. E. Symp. Combust. 1957, 6, 134–143. [29] Van Maaren, A.; Thung, D. S.; De Goey, L. P. H. Combustion Science and Technology 1994, 96, 327–344. [30] Galmiche, B.; Halter, F.; Foucher, F.; Dagaut, P. Energy and Fuels 2011, 25, 948–954. [31] Habib, M. A.; Rashwan, S. S.; Nemitallah, M. A.; Abdelhafez, A. Appl. Energy 2017, 189, 177–186.


Page 26 of 26

Adiabatic Flame Temperature for Controlling the ... - ACS Publications

Recommend Documents