[American Institute of Aeronautics and Astronautics 42nd AIAA Aerospace Sciences Meeting and Exhibit...

AIAA-2004-146

A STUDY OF THE EFFECTS OF AIR AND FUEL DILUTION ON NITRIC OXIDE

EMISSIONS FROM GAS JET DIFFUSION FLAMES

Ala R. Qubbaj*

Mechanical Engineering Department, The University of Texas Pan American 1201 W. University Dr., Edinburg, Texas 78539

Alvaro Martinez+

Department of Environmental Engineering, Texas A&M University Kingsville MSC 213, Kingsville, Texas 78363

* Assistant Professor, AIAA Member, E-mail: [email protected] + Assistant Professor, E-mail: [email protected]

ABSTRACT

In this study, a co-flow methane/air diffusion flame at Reynolds number of 6000 was numerically simulated. The co-flow air and fuel streams were diluted with Nitrogen in the range of 0% to 20%. The thermal and composition fields in the far-burner reaction zone (close to the exhaust) were computed, and the effects of diluent’s addition to the air stream (simulating FGR) and to the fuel stream (simulating FIR) were investigated.

The results show that air-side dilution is very effective up to 5% diluent’s addition. For which, 95% and 65% drops in NO and CO emissions, respectively, along with a 16% increase in temperature, are predicted compared to the baseline case (0% dilution). However, beyond 5% dilution, no effect (reaction) has been predicted. On the other hand, the fuel-side dilution has shown an effect for all simulated diluent’s addition (i.e. 0%-20%). However, that effect is not systematic neither on temperature, CO or NO concentrations. For a similar 5% dilution to the fuel-side, a 14% increase in NO and a 97% decrease in CO are predicted, along with a 5.6% increase in temperature.

The simulated results revealed that air-side dilution (simulating FGR) has a dramatic greater

effectiveness in NO reduction, whereas, fuel-side dilution (simulating FIR) has a greater effectiveness in CO reduction. Besides, the results suggest an important role for Prompt-NO Fenimore mechanism.

INTRODUCTION The goal of the present study is to investigate the

effects of air/fuel dilution on the combustion and emission characteristics of jet diffusion flames. Diffusion flames have been selected since they are employed in most of the practical combustion systems such as gas turbines, diesel engines, and utility boiler furnaces. Their advantages in absence of flash back, controlled flammability, heat release rate, as well as, safety make them more attractive and desirable.

Flue gas recirculation(FGR) is a well known technique that is used to control oxides of nitrogen (NOx). The recircualted flue gas is introduced to the co-flow air stream; it reduces flame temperatures, which results in decreased thermal NO production rates. The recirculated flue gas could also be introduced to the fuel stream rather than to the air, in that case the technique is called fuel injection recirculation (FIR). Previous experimental studies reported different effects on NOx reduction for FGR versus FIR (Hopkins et al.1, Reese et al.2) Any

American Institute of Aeronautics and Astronautics

1

42nd AIAA Aerospace Sciences Meeting and Exhibit5 - 8 January 2004, Reno, Nevada

AIAA 2004-146

Copyright © 2004 by Ala R. Qubbaj. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.

improvement in the combustion performance relative to pollutant formation, stability, and overall efficiency requires a careful study of the flow field and mixing processes, particularly in such a highly reactive flow. The fluid dynamic analysis is very useful to provide the necessary information. Therefore, in this study, CFD simulations are used to study the thermal and composition fields in the far-burner reaction zone (close to the exhaust) produced by either air-side or fuel-side dilution. The effects of diluent’s addition to the air stream (simulating FGR) and to the fuel stream (simulating FIR), on the structure and pollutant emissions of co-flow methane/air laminar diffusion flames, are investigated. The primary diluent that was used is Nitrogen (N2). The co-flow air and fuel streams were diluted with Nitrogen in the range of 0% to 20%. The heat capacity of pure N2 is somewhat less than recirculated combustion production gases, however, the use of N2 does not alter the chemical phenomena in any substantial way since N2 is the primary constituents in the products. Besides, experimental runs using a mixture of CO2, H2O, and N2 with typical product-gas composition as the diluent showed the same qualitative behavior as for pure N2 (Feese and Turns3).

The major pollutants produced by combustion are unburned and partially burned hydrocarbons, nitrogen oxides (NO and NO2), carbon monoxide, and particulate matter in various forms. Carbon monoxide is extremely important to the oxidation of hydrocarbons. The combustion of hydrocarbons can be characterized as a two-step process: the first step involves the breakdown of the fuel to carbon monoxide and radicals, with the second step being the final oxidation of carbon monoxide to carbon dioxide (Turns4).

In hydrocarbon-air flames, different mechanisms account for the production of NO. The most significant, usually named thermal-NO, is the Zeldovich chain of reactions involving O, N, O2, N2, NO, H and OH. This mechanism dominates at high temperatures, and the chemical kinetics is fairly well-understood. This understanding often permits control of thermal–NO via staged combustion or other temperature or residence time modification schemes. Another mechanism of NO production is the prompt-NO route, especially significant in fuel-rich zones. Even though the overall fuel/air ratio may be on the lean side, there are always locally fuel-rich zones in non-premixed flames used in most practical burners. The prompt-NO reaction mechanism, postulated by Fenimore5, has been extensively reviewed and discussed by Miller and Bowman6 and Bowman7. The primary initiation reaction is that of the radical CH with nitrogen to produce HCN and nitrogen atoms. The HCN undergoes several oxidation steps to produce

more N atoms; these atoms then react with OH and O2 to form NO. Some of the NO can be back-converted to HCN through a series of reactions, also involving CH radicals.

Natural gas is used as the fuel due to its availability and low cost. It is assumed that natural gas is composed of pure methane (CH4); this is a highly valid assumption since natural gas is made of approximate 95% methane, with some ethane, propane, butane, pentane, nitrogen, carbon dioxide, and small traces of other gases.

PHYSICAL MODEL

Figure 1(a) shows the actual physical model, which consists mainly of a combustion chamber equipped with fuel/air supply and flue gas recycling facilities. The chamber is made from steel, 31.75 cm x 31.75 cm cross-section and 69.85 cm height. The chamber is provided with air-cooled Pyrex windows of dimensions 10.05 cm x 57.15 cm on all four sidewalls. The top of the chamber is connected to the atmosphere through an exhaust duct (Fig. 1a). The fuel and air are introduced to the combustion chamber through separate streams in a non-premixed or diffusion combustion process. The fuel is introduced through a stainless steel burner, of internal diameter 6.35 mm, inserted in the centerline of the chamber, and the air is introduced through an annular inlet of diameter 0.1 m, surrounding the fuel burner as depicted in Fig. 1a. The fuel/air supply train is designed to support a dual system operation for either FGR or FIR. Before entering the chamber, the air (in case of FGR) or the fuel (in case of FIR) enters through an ejector in order to aspirate the fuel gas directly from the exhaust duct as illustrated in Fig. 1a. The simplified physical model used in the computations, assuming axisymmetric flow conditions, is provided in Fig 1(b). The operating and boundary conditions are given in Table 1. Table 1: Operating and Boundary Conditions

Fuel Natural Gas (95%+) Burner diameter (d) 6.35 mm Jet-exit Reynolds number 6000 Jet-exit/ Fuel axial velocity ux 15.55 m/s Axial air velocity ux 0.3 m/s Far-burner axial location: x/d 30 Ambient temperature 295 K Ambient pressure 100 kPa

NUMERICAL COMPUTATIONS

Computational Model The numerical computations were conducted using the CFDRC-ACE+, advanced computational environment


2

software package 2002, in which CFD-GEOM (Interactive Geometric Modeling and Grid Generation software) and CFD-VIEW (3-D Computer Graphics and Animation Software) are incorporated.

The computational domain encompassed half of the flame jet assuming axisymmetric flow conditions (as seen in Fig. 2) and extended to 69.85 cm in the axial direction and 15.875 cm in the radial direction. The grid cells were generated with increasing spacing in the radial and axial directions; this provided an adequate resolution where gradients were large, i.e., near the centerline, and saved CPU time where gradients were small, near the edges.

A cell-centered control volume approach was used, in which the discretized equations or the finite difference equations (FDE) were formulated by evaluating and integrating fluxes across the faces of control volumes in order to satisfy the Favre-averaged continuity, momentum, energy and mixture fractions conservation equations (Eqs. 1, 2, 4 and 9, respectively). The first order upwind scheme was used for evaluating convective fluxes over a control volume. The well-known SIMPLEC algorithm, proposed by Van Doormal and Raithby8, was used for velocity pressure-coupling. SIMPLEC stands for “Semi-Implicit Method for Pressure-Linked Equation Consistent”, in which an equation for pressure correction is derived from the continuity equation. Governing Equations The code CFD-ACE+ employs a conservative finite-volume methodology and accordingly all the governing equations are expressed in a conservative form in which tensor notation is generally employed. The basic governing equations are for the conservation of mass, momentum and energy: Continuity equation:

0)( =+ jj

uxt

ρ∂∂

∂∂ρ ...............………………(1)

where uj is the jth Cartesian component of velocity and ρ is the fluid density. Momentum equations: (j=1, 2, 3)

ji

ij

jji

ij f

xxpuu

xu

tρ

∂∂τ

∂∂ρ

∂∂ρ

∂∂

++−=+ )()( …..(2)

where p is the static pressure, τij is the viscous stress tensor and fj is the body force. For Newtonian fluids τij can be expressed as:

ijk

k

i

j

j

iij x

uxu

xu

δ∂∂µ

∂∂

∂∂

µτ ⎥⎦

⎤⎢⎣

⎡−⎟

⎟⎠

⎞⎜⎜⎝

⎛+=

32 ..….(3)

where µ is the fluid dynamic viscosity and δij is the Kronecker delta.

Energy Equation: The equation for the conservation of energy can

take several forms. The static enthalpy form of the energy equation can be expressed as:

)4.......(..............................).........(

)()(

mhmjJjxjx

iuij

jx

pju

t

p

jxjq

hjujx

ht

∂

∂

∂

∂τ

∂

∂

∂

∂

∂

∂ρ

∂

∂ρ

∂

∂

−+

++−=+

where Jmj is the total (concentration-driven + temperature-driven) diffusive mass flux for species m, hm represents the enthalpy for species m, and qj is the j-component of the heat flux. Jmj, hm and h are given as:

j

mmj X

YDJ∂∂

ρ−= ....................................……….(5)

ofm

hdTTT

oT mPCmh +∫= )( ....................……….(6)

mhn

mmYh ∑

==

1................................……………(7)

where D is the diffusion coefficient, Cp is the constant-pressure specific heat, and hf

o is the enthalpy of formation at standard conditions (Po=1 atm, To=298 K). The Fourier’s law is employed for the heat flux:

jj x

TKq∂∂

−=′′ .................……………...……(8)

where K is the thermal conductivity. Mixture Fractions:

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛=+

jxkf

Djxkfju

jxkft ∂

∂

∂∂ρ

∂∂ρ

∂∂ )()( ....(9)

where D is the diffusion coefficient, fk is the mixture fraction for the kth mixture.

Chemistry/Reaction Model

The reaction model used by CFD-ACE+ was the instantaneous chemistry model in which the reactants are assumed to react completely upon contact. The reaction rate is infinitely rapid and one reaction step is assumed. Two reactants, which are commonly referred to as “fuel” and “oxidizer”, are involved. A surface “flame sheet” separates the two reactants (this assumption can be made only for nonpremixed flames). The mass fractions for this model are computed by first using Eq. 10 to obtain the composition that would occur without the reaction. The “unreacted” composition, denoted by the superscript “u”, is given by


3

..............................................(10) ∑=

=K

kkik

ui fY

1

)( ξ

where ξik is the mass fraction of the ith species in the kth mixture, Yi is the mass fraction of the ith species and fk is the mixture fraction of the kth mixture. The change in composition due to the instantaneous reaction is then added to the unreacted mass fractions, as described below.

A stoichiometrically correct reaction step needs to be specified. The mass of species i produced per unit mass of fuel consumed is

rMMi

i i

f f= −

νν

.................................................(11)

where ν is the stoichiometric coefficient of the species in the overall reaction; positive for product species and negative for fuel and oxidizer. The instantaneous reaction mechanism consumes either all the fuel or all the oxidizer, whichever is limiting. The amount of fuel consumed is

∆YY

rY

rff

u

f

oxu

ox=

− −⎛

⎝⎜

⎞

⎠⎟min

( ),( ) .................(12)

The change in each species due to the reaction is proportional to the change in fuel, with the proportionality constant given by Eq. 11. The mass fraction of each species is then given by

...........…................(13) Y Y r Yi iu

i= +( ) ∆ f

The right-hand side of the above equation is only a function of the K mixture fractions. Therefore, K-1 transport equations were solved for the mixture fractions. These equations have no source terms due to chemical reactions (for more details see Qubbaj9).

RESULTS AND DISCUSSION

Air-Side Dilution Figures 2 (a, b, c, d, e) show the radial

temperature, CO2, NO, CO, and O2 in-flame profiles, respectively, in the far-burner (close to the exhaust) region. Each graph is for 0%, 5%, 10%, 15%, and 20% of N2 diluted in the air-stream. For 5% N2 dilution, dramatic results could be seen. However, beyond that (i.e. for 10%, 15%, and 20% N2 dilution), the simulated results for CO2, NO, and CO were zero, indicating no chemical reaction. This could be explained by the fact that air is mainly composed of N2 (76.8% N2 and 23.2% O2 by mass). Therefore, higher nitrogen dilution means less oxygen, thereby pushing the mixture out of its flammability limits.

Fig 2a shows a peak in-flame temperature of 960K for the baseline case (0% N2 dilution). For 5% N2 dilution, an increase of 16% has been seen. That temperature increase is due to the dominance of the

fuel-lean conditions in such a far-burner region, and therefore, the Nitrogen dilution would push the mixture towards stoichiometry, thereby leading to the increase in temperature. The temperature profiles do not have an off-axis peak due the merge of the stoichiometric contour in the far-burner region.

Fig. 2b shows a peak CO2 concentration of almost 4% for the baseline case. For 5% dilution, an increase of 24% has been seen. The CO2 profiles have the same trend as the former temperature ones since CO2 is a major combustion product driven by temperature.

Fig 2c shows a peak in-flame NO concentration of 105 ppm for the baseline case (0% N2 dilution). For 5% dilution, a dramatic drop of 95% from the baseline concentration is noticed. The NO concentration profiles do not follow the trends of the corresponding temperature profiles. This indicates that the temperature-dependent thermal-NO is not the dominant NO production mechanism and suggests a role for Prompt-NO Fenimore mechanism. This conclusion is substantiated by the temperature profiles that depict peak in-flame temperatures below 1800 K. That temperature (1800 K) has been reported to be the threshold for the Thermal-NO mechanism (Turns2; Bowman5]. Fig. 2d depicts a peak in-flame CO concentration of 0.1204% for the baseline case. At 5% dilution, a 65% decrease is noticed. The CO profiles have a peak value in the fuel rich region and it decreases in the radial direction towards both the fuel-lean and centerline regions.

Fig. 2c shows the O2 in-flame concentrations. The general trend of O2 profiles is that the O2 is minimal at the center of the flame and increases toward the radial direction till it attains the ambient value of 23% (by mass). As the Nitrogen content is increased by 5%, the O2 content decreased by 5 %.

Fuel-Side Dilution Figures 3 (a, b, c, d, e) show the radial

temperature, CO2, NO, CO, and O2 in-flame profiles, respectively, in the far-burner (close to the exhaust) region. Each graph is simulated for 0%, 5%, 10%, 15%, and 20% of N2 diluted in the fuel-stream (originally composed of 100% CH4). Unlike the air-side dilution, a reaction occurs for 10%, 15%, and even 20% N2 dilution. Since the fuel is pure CH4, a 20% maximum dilution means 80% CH4, and therefore, the mixture remains within the flammability limits. In Fig. 3a the temperature effects for 0%, 5%, 10%, 15%, and 20 % dilution are computed. The peak temperature for 0% dilution (baseline) is 963K. For, 5%, 10% and 15% dilution, an increase of 5.6%, 0.5%, and 1.5% is noticed, respectively, whereas, for 20% dilution, a decrease of 0.2% is noticed. The in-flame


4

temperature increase as a result of fuel dilution has a similar explanation as that in air-side dilution.

In Fig. 3b, the effects of 0%, 5%, 10%, 15%, and 20% fuel dilution on CO2 concentration are shown. The CO2 concentration for 0% dilution (baseline) is 4.1%. For 5%, 10%, 15%, and 20% fuel dilution, an increase of 9.55%, 0.7%, 2.5%, and .2%, respectively, is noticed from the baseline concentration. The trends of CO2 profiles do qualitatively follow those of temperatures.

In Fig. 3c the effects of 0%, 5%, 10%, 15%, and 20% fuel dilution on NO concentrations are seen. The baseline peak NO concentration is 105 ppm. For 5% dilution, a 14% increase is seen. Whereas for 10%, 15%, and 20% dilution, a decrease of 9%, 3%, and 12.4%, respectively, from the baseline value, is predicted. The NO profiles do not follow the trends of temperature profiles (as seen in the air dilution case). Besides, the in-flame temperatures point towards the dominance of prompt-NO rather than Thermal-NO production mechanism (Fenimore5). In Fig. 4d, the effects of 0%, 5%, 10%, 15%, and 20% fuel dilution on CO concentrations are given. The baseline peak CO concentration is 0.1204%. For 5%, 10%, 15%, and 20% fuel dilution, a 97% decrease, a 72% increase, a 96% decrease, and a 16% decrease is noticed, respectively, from the baseline value. It can be seen that the fuel-side dilution has a greater influence on CO than on NO emissions.

Fig. 3e shows the O2 concentration profiles for 5%, 10%, 15% and 20% fuel dilution, respectively. In general, the O2 concentration is minimum at the center of the flame, and increases toward the radial direction. For 5%, 10%, 15% and 20% dilution, the minimum O2 concentrations vary at the center of the flame but as the radial distance increases the O2 concentrations increases until it reaches the ambient value (23.2%).

CONCLUSIONS

The present study has revealed the followings: • Air-side dilution of 5% nitrogen has resulted in a

dramatic drop in both NO and CO emission concentrations by 95% and 65%, respectively, along with a 16% increase in temperature, from the baseline case.

• Air-side dilution has a substantial effect up to 5% nitrogen, beyond which, no effect (reaction) has been predicted. This is due to the increase in N2 concentrations and the consequent O2 decrease, thereby, pushing the mixture out of the flammability limits.

• The fuel-side dilution has shown an effect for all simulated diluent’s addition (i.e. 0%-20%). However, that effect is not systematic neither on temperature, CO or NO concentrations.

• For a similar 5% dilution to the fuel-side, a 14% increase in NO and a 97% decrease in CO are predicted, along with a 5.6% increase in temperature.

• The air-side dilution (simulating FGR) has a dramatic greater effectiveness in NO reduction, whereas, fuel-side dilution (simulating FIR) has a greater effectiveness in CO reduction.

• The Prompt-NO Fenimore mechanism plays an important role in NO production for both the air-side and fuel-side dilution.

ACKNOWLEDGMENT

This project has been partially supported by the National Science Foundation through partnership in a Center for Research Excellence in Science and Technology at Texas A&M University Kingsville (Award #: 0206259).

REFERENCES 1Hopkins, K. C., Czerniak, D. O., Youssef, C., Radak, L., and Nylander, J. (1991). ASME paper 91-JPGC-EC-3, International Power Generation Conference, San Diego, CA, October 6-10. Reese, J. L., Reddy, V., Lange, H. B., Chang, C., 2Radak, L. J., and Youssef, C. F. (1994). American And Japanese Flame Research Council, Pacific Rim International Conference on Environmental Control of Combustion Processes, Maui, HI, October 16-20. 3Feese, J., and Turns S. (1993). Combustion and Flame 113:66-78. 4Turns, S. R. (2000). An Introduction to Combustion: Concepts and Applications. McGraw-Hill Inc., NY. 5Fenimore, C. P. (1970). Formation of Nitric Oxide in Premixed Hydrocarbon Flames. Thirteenth Symposium (International) on Combustion, Combustion Institute, Pittsburgh, PA, pp. 373-380. 6Miller, J. A. and Bowman, C. T. (1989). Mechanisms and Modeling of Nitrogen Chemistry in Combustion. Progress in Energy and Combustion, Vol. 15, p. 287. 7Bowman, Craig. T. (1992). Control of Combustion-Generated Nitrogen Oxide Emissions: Technology Driven By Regulations. Twenty-Fourth Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, pp 859-878. 8Van doormaal, J. P. and Dryer, F. L. (1984). Enhancements of the SIMPLE Methods for Predicting Incompressible Fluid Flows. Numerical Heat Transfer, Vol. 7, pp. 147-163. 9Qubbaj, A. R. (1998). An Experimental and Numerical Study of Gas Jet Diffusion Flames Enveloped by a Cascade of Venturis. Ph.D. Dissertation, School of Aerospace and Mechanical Engineering, University of Oklahoma, Norman, OK.


5


6

(a)

(b)

Figure1: (a) Actual physical model (b) Simplified Problem geometry

Figure 2 (a-e): Radial temperature, CO2, NO, CO and O2 profiles for 0%-20% air dilution

��

0

200

400

600

800

1000

1200

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

Tem

pera

ture

(K)

(a)

��0

1

2

3

4

5

6

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

CO

2(%)

(b)

��0

20

40

60

80

100

120

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

NO

(ppm

)

(c)

��0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

CO

(%)

(d)

��

��

0

0.05

0.1

0.15

0.2

0.25

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

Radial distance (m)

O2(

mol

e fr

actio

n)

0% 5% 10%��15%

��20%

(e)

7American Institute of Aeronautics and Astronautics

Figure 3 (a-e): Radial temperature, CO2, NO, CO and O2 profiles for 0%-20% fuel dilutio

��

��

��

��

0

200

400

600

800

1000

1200

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

Tem

pera

ture

(K)

(a)��

��

��

��

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

CO

2(%)

(b)

��

��

��

��

��

��

0

20

40

60

80

100

120

140

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

NO

x(ppm

)

(c)

��

0

0.05

0.1

0.15

0.2

0.25

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

CO

(%)

(d)

��

��

��

��

��

0.15

0.16

0.17

0.18

0.19

0.2

0.21

0.22

0.23

0.24

0.25

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

O2(m

ole

fract

ion)

0% 5% 10%��15%

��20%

(e)

8American Institute of Aeronautics and Astronautics

[American Institute of Aeronautics and Astronautics 42nd AIAA Aerospace Sciences Meeting and Exhibit...

Documents

Transcript of [American Institute of Aeronautics and Astronautics 42nd AIAA Aerospace Sciences Meeting and Exhibit...