Energy & Fuels 2009, 23, 1379–1389 1379

Global Combustion Mechanisms for Use in CFD Modeling under

Oxy-Fuel Conditions
Jimmy Andersen, Christian Lund Rasmussen, Trine Giselsson, and Peter Glarborg*
Department of Chemical and Biochemical Engineering Technical UniVersity of Denmark, DK-2800 Kgs.
Lyngby, Denmark

ReceiVed May 16, 2008. ReVised Manuscript ReceiVed December 16, 2008

Two global multistep schemes, the two-step mechanism of Westbrook and Dryer (WD) and the four-step
mechanism of Jones and Lindstedt (JL), have been refined for oxy-fuel conditions. Reference calculations
were conducted with a detailed chemical kinetic mechanism, validated for oxy-fuel combustion conditions. In
the modification approach, the initiating reactions involving hydrocarbon and oxygen were retained, while
modifying the H2-CO-CO2 reactions in order to improve prediction of major species concentrations. The
main attention has been to capture the trend and level of CO predicted by the detailed mechanism as well as
the correct equilibrium concentration. A CFD analysis of a propane oxy-fuel flame has been performed using
both the original and modified mechanisms. Compared to the original schemes, the modified WD mechanism
improved the prediction of the temperature field and of CO in the post flame zone, while the modified JL
mechanism provided a slightly better prediction of CO in the flame zone.

Introduction showed to be superior to the mixed-is-burned model. The

temperature and species concentrations predictions were better,
Oxy-fuel combustion is a promising technique for separating since the effect of molecular dissociation has been accounted
gaseous CO2 with the intention of storing it, for instance in for. An alternative to these approaches is to employ finite-rate
geological reservoirs. The separation of CO2 is facilitated by chemistry, using a scheme consisting of one or several global
removing the atmospheric nitrogen from air before combustion. reactions. A number of simplified methane oxidation mecha-
Recirculated combustion products are used for diluting a pure nisms has been proposed in literature.4-9 Brink et al.10 tested
O2 stream. Due to the higher heat capacity and radiative this approach under perfectly stirred reactor conditions. An
properties of CO2 compared to N2, an increased oxygen equilibrium approach was compared with a three-step irrevers-
concentration is required to obtain the same thermal conditions ible mechanism,10 the four-step mechanism suggested by Jones
as combustion in air. For natural gas, a mixture of 28% O2 and and Lindstedt6 and a detailed mechanism suggested by Glarborg
72% CO2 will result in a temperature field similar to combustion et al.11 Their conclusions were that thermodynamic equilibrium
in air.1 Oxy-fuel combustion will eventually result in a flue gas provides a poor description under conditions with a strong
consisting mainly of CO2 and steam. The flue gas can then coupling between turbulent mixing and chemical reactions. The
undergo a condensation process to remove the H2O, before a global mechanisms resembled the detailed model well at fuel-
part of this flue gas consisting almost entirely of CO2 is lean conditions. At fuel-rich conditions, the accuracy of the
recirculated, while the remaining part is ready for compression three-step irreversible model was less satisfactory, whereas the
and storage.2 four-step mechanism performed better and correctly approached
Computational fluid dynamics (CFD) is becoming an impor- the equilibrium composition at long residence times. The four-
tant industrial tool for trouble-shooting and optimization. step mechanism was reported to be a good compromise between
However, CFD modeling of industrial combustion applications accuracy and computational effort.10
is a computationally demanding task. For this reason, it is often Examples of CFD modeling using the global mechanisms can
necessary to apply simplified reaction mechanisms to reduce be found in the literature.10,12-14 Also, in industrial CFD
the computing effort. Chemistry is often represented by a mixed- modeling, the global mechanisms are used frequently, presum-
is-burned assumption or an assumption of chemical equilibrium.
Breussin et al.3 performed a CFD analysis of a pure natural (4) Dryer, F. L.; Glassman, I. Proc. Combust. Inst. 1972, 14, 987–1003.
gas/oxygen flame, and found good predictions for the fluid (5) Westbrook, C. K.; Dryer, F. L. Combust. Sci. Technol. 1981, 27,
dynamics, temperature and main chemical species concentration 31–44.
(6) Jones, W. P.; Lindstedt, R. P. Combust. Flame 1988, 73, 233–249.
fields (O2, CO2) using both an Eddy Dissipation/mixed-is-burned (7) Dupont, V.; Pourkashanian, M.; Williams, A.; J. Inst. Energy 1993,
approach and an EDC/chemical equilibrium approach. The 66, 20–28.
mixed-is-burned model did, however, fail in predicting CO (8) Nicol, D. G.; Malte, P. C.; Hamer, A. J.; Roby, R. J.; Steele, R. C.
J. Eng. Gas Turbines Power 1999, 121, 272–280.
properly. In all the flames predicted, the equilibrium model (9) Meredith, K. V.; Black, D. L. 44th AIAA Aerospace Sciences Meeting
2006, 19, 14161–14167.
* To whom correspondence should be addressed. Tel: +45 4525 2840. (10) Brink, A.; Kilpinen, P.; Hupa, M.; Kjaldman, L. Combust. Sci.
Fax: +45 4588 2258. E-mail: pgl@kt.dtu.dk. Technol. 1999, 141, 59–81.
(1) Andersson, K.; Johnsson, F. Fuel 2007, 86, 656–688. (11) Glarborg, P.; Alzueta, M. U.; Dam-Johansen, K.; Miller, J. A.
(2) Wall, T. F. Proc. Combust. Inst. 2007, 31, 31–47. Combust. Flame 1998, 115, 1–27.
(3) Breussin, F.; Lallemant, N.; Weber, R. Combust. Sci. Technol. 2000, (12) Brink, A.; Mueller, C.; Kilpinen, P.; Hupa, M. Combust. Flame
160, 369–397. 2000, 123, 275–279.

10.1021/ef8003619 CCC: $40.75  2009 American Chemical Society

Published on Web 01/28/2009
1380 Energy & Fuels, Vol. 23, 2009 Andersen et al.

Table 1. Westbrook and Dryer Global Multi Step Methane Combustion Mechanism with Kinetic Data (units in cm, s, cal, and mol)
mechanism reactions A β Ea reaction orders ref
WD1 CH4 + 1.5 O2 f CO + 2 H2O 1.59 × 1013 0 47.8 × 103 [CH4]0.7[O2]0.8 4
WD2 CO + 0.5 O2 f CO2 3.98 × 1014 0 40.7 × 103 [CO][O2]0.25[H2O]0.5 4
WD2r CO2 f CO + 0.5 O2 5.0 × 108 0 40.7 × 103 [CO2] 5

ably because the models are simple, cheap, and readily available. detailed mechanisms. Detailed model predictions have been
Furthermore, computationally simple mechanisms do provide obtained from plug flow simulations using the Senkin code from
adequate results if only the main species concentrations and the Chemkin-II library.20 The global multistep model PFR
the temperature picture is of interest. The simplified schemes calculations have been performed in Matlab. Even though the
cannot be expected, however, to work as well under oxy-fuel Chemkin 4.0 package21 can handle the noninteger reaction orders
combustion conditions, as they do for conventional combustion. that are often applied in global mechanisms, the Matlab code was
The exchange of the largely inert N2 with a chemically reactive preferred because it facilitated comparison of several parameters
compound, CO2; at least at high temperatures, has been shown for both the detailed and global models. It should be noted,
to change the importance of some of the elementary reactions however, that convergence can be problematic when using non-
governing the combustion,15 thereby requiring a modification integer reaction coefficients and because of this, the stiff numerical
of the global multistep reaction mechanisms to make them valid solver ode15s was preferred over ordinary ODE solvers.
under oxy-fuel conditions. Detailed Chemical Kinetic Model (DCKM). In this work,
The objective of the present work is to modify two simple the global mechanisms are compared to the detailed chemical
multistep mechanisms, used frequently for describing conven- kinetic mechanism (DCKM) presented by Glarborg and Bentz-
tional combustion, to handle the increased CO2 concentration en.15 To the authors’ knowledge, this mechanism is the only
under oxy-fuel conditions. The two-step hydrocarbon oxidation one that has been validated against oxy-fuel experiments. The
mechanism by Westbrook and Dryer (WD)5 is selected, since mechanism satisfactorily predicts CO, CO2 and O2 concentra-
this scheme is directly available as default in the commercial tions from oxidation of CH4 by an O2/CO2 mixture in a flow
CFD code Fluent.16 The second scheme selected is the four- reactor with residence times of approximately 1 s.15
step mechanism of Jones and Lindstedt (JL).6 The JL scheme The Westbrook and Dryer Two-Step Mechanism (WD).
is more complex but also has a higher accuracy and similar to The WD model consists of two reactions, where the last step,
the WD scheme, it is used regularly in CFD modeling of oxidation of CO to CO2, is reversible. The mechanism is listed
industrial applications. in the form of three irreversible steps,
The selected simplified mechanisms are compared with
reference calculations, conducted with the detailed combustion CH4 + 1.5O2 f CO + 2H2O (WD1)
mechanism proposed by Glarborg and Bentzen.15 Their kinetic
CO + 0.5O2 f CO2 (WD2)
model provides good agreement with oxy-fuel combustion
experiments. The work presented herein bears some resemblance CO2 f CO + 0.5O2 (WD3)
to the work of Brink et al.10 They also tested the Jones and
Lindstedt four-step mechanism with reference calculations with Table 1 displays the reaction rate data for the WD mechanism.
a detailed mechanism, but only for conventional combustion The rate constants for (WD1) and (WD2) originate from Dryer
conditions. and Glassman4 who studied high temperature oxidation reactions
Modeling Approach. The Eddy Dissipation Concept (EDC) of carbon monoxide and methane under fuel lean conditions (λ
is a popular turbulence chemistry interaction model for CFD > 2) in a turbulent flow reactor. Later, Westbrook and Dryer5
analysis of combustion applications. The Eddy Dissipation included the reverse reaction step for CO2 decomposition (WD3)
Concept is an extension of the Eddy Dissipation model17 based in order to reproduce the proper heat of reaction and pressure
on the work by Gran and Magnussen.18,19 In the EDC model, dependence of the [CO]/[CO2] equilibrium.
chemical reactions are assumed to occur in the fine structures The Jones and Lindstedt Four-Step Mechanism (JL).
of the computational cells. These small scale structures can be Jones and Lindstedt6 developed four-step global mechanisms
pictured as a part of the cell, where Kolmogorov-sized eddies for several hydrocarbon fuels. For methane, it involves the
containing combustion species are situated so close together that following steps:
mixing on the molecular level is taking place. The EDC model CH4 + 0.5O2 f CO + 2H2 (JL1)
evaluates the volume of each cell, where mixing on a molecular
scale is occurring, and treats this part of the cell as a Perfectly CH4 + H2O f CO + 3H2 (JL2)
Stirred Reactor (PSR/CSTR). The cell volume fraction and
reactor residence time is depending on turbulence parameters H2 + 0.5O2aH2O (JL3)
for the specific computational cell.18,19
CO + H2OaCO2 + H2 (JL4)
Since the turbulence/chemistry interaction description in CFD
may involve ideal reactor modeling, a comparison of the perfor- The mechanism consists of two irreversible reactions, (JL1)
mance of different global multistep reaction mechanisms should and (JL2), describing the initial oxidation steps of a hydrocarbon.
be conducted under similar conditions. True to the concept of Gran The two reversible reactions, (JL3) and (JL4), control the rate
and Magnussen,18,19 Brink et al.10 tested the simplified models under of reaction for CO and H2. The rate coefficients for methane
perfectly stirred reactor conditions. However, it is questionable combustion are displayed in Table 2. Jones and Lindstedt
whether commercial CFD codes actually employ a PSR solver for validated the model against data for premixed methane and
the chemistry, since the resulting algebraic equations may yield propane flames along with diffusion flame data for a methane-
convergence problems. More likely, codes like Fluent employ a air flame. The mechanism was reported to perform well for both
numerical solver that performs an integration in time. Consequently, fuel-lean and moderately fuel-rich stoichiometries.6 Two sets
in the present work, isothermal plug flow reactor (PFR) modeling, of rate parameters for reaction (JL3) were presented. Set (JL3a)
rather than PSR calculations, is used to compare the global and was the preferred expression, but since it involved negative
Global Combustion Mechanisms Energy & Fuels, Vol. 23, 2009 1381

Table 2. Jones Lindstedt Global Multi-Step Methane Combustion Mechanism with the Kinetic Rate Data (units in cm, s, cal, and mol)
no. reactions A β Ea reaction orders ref
JL1 CH4 + 0.5 O2 f CO + 2 H2 7.82 × 1013 0 30.0 × 103 [CH4]0.5[O2]1.25 6
JL2 CH4 + H2O f CO + 3 H2 0.30 × 1012 0 30.0 × 103 [CH4][H2O] 6
JL3a H2 + 0.5 O2 h H2O 4.45 × 1018 -1 40.0 × 103 [H2]0.5[O2]2.25[H2O]-1 6
JL3b H2 + 0.5 O2 h H2O 1.21 × 1018 -1 40.0 × 103 [H2]0.25[O2]1.5 6
JL4 CO + H2O h CO2 + H2 2.75 × 1012 0 20.0 × 103 [CO][H2O] 6

Table 3. Chemkin-Code for the JL Mechanism (3b H2-O2 Reaction) (units: cm, s, cal, mol, and K)
ELEMENTS C O H N END
SPECIES CH4 O2 H2O N2 CO CO2 H2 END
REACTIONS
CH4 + 0.5 O2 ) > CO+ 2H2 7.82 × 1013 0 30.0 × 103 !Jones Lindstedt 88
FORD /CH4 0.5/
FORD /O2 1.25/
CH4 + H2O ) > CO + 3H2 0.30 × 1012 0 30.0 × 103 !Jones Lindstedt 88
H2 + 0.5 O2 ) > H2O 1.209 × 1018 -1 40.0 × 103 !Jones Lindstedt 88
FORD /H2 0.25/
FORD /O2 1.5/
H2O + 0 O2 +0 H2 ) > H2 + 0.5 O2 7.06 × 1017 -0.877 97.9 × 103 !Calculated
FORD /H2 -0.75/
FORD /O2 1/
FORD /H2O 1/
CO + H2O ) CO2+H2 0.275 × 1013 0 20.0 × 103 !Jones Lindstedt 88
END

reaction orders that might cause numerical problems, the set These expressions make it possible to implement the forward
for (JL3b) was proposed as an alternative.6 and reverse reactions as irreversible steps, facilitating the use
If (JL3) had been an elementary reaction, the reverse rate in commercial CFD software such as Fluent.16
constant could be determined from expression 1, In the following, only reaction 3b of the JL mechanism is
used. Attempts to use 3a resulted in problems with convergence
ṘJL3,f kJL3,f[H2][O2]0.5 which may limit its use in CFD. Another concern related to 3a
) KJL3 ) (1)
ṘJL3,r kJL3,r[H2O] is that the reverse rate for (JL3a) is independent of the H2O
concentration. This makes sense when the purpose of the
Here, KJL3 is the equilibrium constant, kJL3 is the rate constant, reaction is to dampen the forward reaction, but under conditions
and the f and r subscribts refer to the forward and reverse with large H2O concentrations or even H2O in the oxidizer
reactions, respectively. However, (JL3) is a global reaction and stream a zero reaction order may be inappropriate.
the forward reaction orders do not follow the stoichiometry. A way to implement the JL mechanism in Fluent16 is to
For this reason, the reverse reaction order for reaction (JL3a) import it as a Chemkin input file. In the newer versions of
or (JL3b) cannot be derived according to eq 1. In order for an Chemkin, the FORD keyword is used to overwrite the forward
equilibrium approach to be maintained for a global reaction, reaction order, allowing global expressions to be implemented.
the expression 1 must still hold at equilibrium, here exemplified Table 3 shows an example of a Chemkin input file, with reaction
with reaction (JL3b): (3b) implemented. Note that in the reverse H2-O2 reaction, H2
* and O2 have been added as reactants with 0 as stoichiometry
ṘJL3,f ṘJL3b,f *
kJL3,f[H2][O2]0.5 kJL3b,f [H2]0.25[O2]1.5
) ) ) (2) coefficients. This is required to allow the program to overwrite
ṘJL3,r *
ṘJL3b,r kJL3,r[H2O] k* [H O] the forward reaction orders.
JL3b,r 2

The superscript * refers to the global rate expressions. Since

the forward rate constant for the global and the stoichiometric Results and Discussion
expression, respectively, must be identical (kJL3, f ) kJL3b,
*
f), the In this work, we evaluate the performance of the WD and JL
concentration dependence of the reverse reactions can be found schemes under conditions of conventional combustion and oxy-
as follows: fuel combustion, respectively, assuming plug-flow. Then the two
schemes are revised for use under oxy-fuel conditions and tested
*
kJL3b,r ) kJL3,r[H2]-0.75[O2] w ṘJL3b,r
*
) kJL3,r[H2]-0.75[O2][H2O] again against reference calculations with the detailed mechanism.
(3) Finally, the original and modified schemes are implemented and
compared for CFD calculations of a turbulent diffusion propane/
In the present work, the derivation of the backward rates is done
O2/CO2 flame under conditions similar to those reported recently
by evaluating the forward rates divided by the equilibrium
by Andersson and Johnsson.1
constant at a series of temperatures (from 500 to 2500 K with
Performance of the WD and JL Mechanisms. Both the JL
100 K increments) and then fitting an Arrhenius expression to
and the WD global mechanisms have been used extensively in
the results, to obtain an individual expression for the reverse
CFD models for conventional combustion in air. Before we
reaction. For reaction (JL3a) we obtain the expression,
investigate how the schemes function under oxy-fuel conditions,
49260 we test them under normal combustion conditions by comparing
*
kJL3a,r (T) ) 2 . 6 × 1018 · T-0.877 · exp - ( T ) (4)
with reference calculations with the detailed reaction mechanism.
and for reaction (JL3b), Since the mechanisms were optimized for fuel-lean conditions,
we evaluate them at conditions with excess air.
49260 Figure 1 compares CO, O2 and CO2 concentrations in an
*
kJL3b,r(T) ) 7.1 × 1017 · T-0.877 · exp - ( T ) (5)
isothermal plug flow reactor at 1600 K and an excess air ratio
1382 Energy & Fuels, Vol. 23, 2009 Andersen et al.

Figure 1. Major species concentrations in plug flow calculations. Comparison between the DCKM, the WD and the JL mechanisms at λ ) 1.2 and
1600 K under “air” firing conditions (21% O2 and 79% N2).

Figure 2. Major species concentrations in plug flow calculations. Comparison between the DCKM, the WD, and the JL mechanisms at λ ) 1.2 and
1200 K (top), 1600 K (middle), and 2000 K (bottom) under “oxy”-firing conditions (28% O2 and 72% CO2).

of λ ) 1.2. These results, along with other simulations, show conversion of CH4 to CO2 to be of the order of 10-3 s, which
that both the WD and JL mechanisms adequately predict O2 is satisfactory.
and CO2 levels at fuel lean conditions. Thereby, they would The performances of the WD and JL mechanisms under fuel-
also yield a satisfactory prediction of the heat release for lean oxy-firing conditions are displayed in Figures 2 and 3. At
nonisothermal conditions at these stoichiometries. Only the JL low temperatures (<1300 K) consumption of CO2 is practically
mechanism predicts CO correctly at longer times under fuel negligible,15 and CO2 is expected to have little chemical effect
lean conditions. The WD mechanism tends to overpredict the under these conditions. At higher temperatures, such as in the
CO exit concentration and may not be sufficiently accurate for figures, atomic hydrogen may start to convert CO2 to CO,
CO emission modeling. resulting in a change in both the CO/CO2 ratio and the
As expected, the results show differences in the predicted composition of the O/H radical pool compared to conventional
ignition time between the three models. Only the detailed model combustion. In general, the level of agreement between the
can describe the slow build-up of the radical pool that eventually global mechanisms and the reference calculations is similar to
lead to ignition in a plug-flow calculation. The global schemes that observed under conventional combustion conditions. As
cannot account for an ignition delay. The WD mechanism was expected, the WD mechanism cannot predict the CO exit
developed to describe the postignition fuel-consumption rate concentration accurately; differences are seen at all temperatures.
in PFR experiments, whereas the JL mechanism was optimized With the exception of the ignition timing, predictions of the JL
for flame conditions, where a radical pool is available through mechanism generally compare well with those of the detailed
diffusional processes. It is difficult to assess how inaccuracies mechanism. The ability of the JL scheme to compensate for
in the description of ignition affect a CFD calculation. In the the water-gas shift reaction (JL4) allows it to capture most
EDC approach, PSR/PFR reactor residence times may be in the changes caused by the elevated CO2 concentration. Conse-
range of 10-4 to 10-3 s, i.e., time scales where there are quently, it predicts correct levels for all major species at longer
considerable differences between modeling predictions. How- reaction times, even under fuel-rich conditions. However, the
ever, all three models predict the time scale for complete CO peak levels are overpredicted in most cases.
Global Combustion Mechanisms Energy & Fuels, Vol. 23, 2009 1383

Figure 3. Major species concentrations in plug flow calculations. Comparison between the DCKM, the WD and the JL mechanisms at 1600 K and
λ ) 0.8 (top), λ ) 1.0 (middle), and λ ) 1.2 (bottom) under “oxy”-firing conditions (28% O2 and 72% CO2).

Table 4. Modified Multi Step Methane Combustion Mechanisms with Kinetic Rate Data - units in cm, s, cal, mol
reaction number reactions A β Ea reaction orders
WD1 CH4 + 1.5 O2 f CO + 2 H2O 1.59 × 1013 0 47.8 × 103 [CH4]0.7[O2]0.8
WD2(modified) CO + 0.5 O2 f CO2 3.98 × 108 0 10.0 × 103 [CO][O2]0.25[H2O]0.5
WD3(modified) CO2 f CO + 0.5 O2 6.16 × 1013 -0.97 78.4 × 103 [CO2][H2O]0.5[O2]-0.25
JL1 CH4 + 0.5 O2 f CO + 2 H2 7.82 × 1013 0 30.0 × 103 [CH4]0.5[O2]1.25
JL2 CH4 + H2O f CO + 3 H2 3.00 × 1011 0 30.0 × 103 [CH4][H2O]
JL3(modified) H2 + 0.5 O2 f H2O 5.0 × 1020 -1 30.0 × 103 [H2]0.25[O2]1.5
JL3 reverse H2O f H2 + 0.5 O2 2.93 × 1020 -0.877 97.9 × 103 [H2]-0.75[O2][H2O]
JL4 CO + H2O a CO2 + H2 2.75 × 1012 0 20.0 × 103 [CO][H2O]

Refined Schemes for Oxy-Fuel Combustion. It is apparent energy for (WD2) was lowered to match better the detailed
that the WD mechanism needs improvement to become ap- model predictions.
plicable for oxy-fuel conditions due to the poor CO prediction. When comparing the CO levels predicted by the JL mech-
The JL mechanism works satisfactorily under oxy-fuel condi- anism with the detailed mechanism, it was observed that besides
tions, even though the prediction of the peak CO levels could the difference in ignition timing, the peak CO levels deviated,
be improved. In the present work, both global models are most pronounced under oxy-fuel conditions (Figures 2 and 3).
modified to improve the prediction of the steady state concentra- The predicted CO oxidation rate in the JL scheme is governed
tions and the CO trends compared to the reference calculations. by the reaction with water vapor (JL4). Examination of the
The global mechanisms, refined for oxy-fuel conditions, are mechanism showed that it was the availability of H2O (formed
summarized in Table 4. The following criteria were employed in reaction JL3), as well as the rate constant for (JL4), that
in the modification procedure: limited the CO oxidation rate. Attempts to modify the JL4 rate
• In time, the concentrations should approach correctly the constant proved unsuccesful; a decrease in kJL4 resulted in
chemical equilibrium values. undesired changes at some conditions (prolonged burnout
• The peak CO concentrations should approximate the values period), whereas an increase had little effect, since the formation
predicted by the detailed model.
The modification window involved temperatures of 1200-2000 (13) Mueller, C.; Brink, A.; Kilpinen, P.; Hupa, M.; Kremer, H. Clean
K and stoichiometries in the range 0.8 < λ < 1.5. Air 2002, 3, 145–163.
In the WD scheme, the major issue was the rate coefficients (14) Saario, A.; Oksanen, A. Energy Fuels 2008, 22, 297–305.
(15) Glarborg, P.; Bentzen, L. L. B. Energy Fuels 2007, 22, 291–296.
for reaction (WD3), which did not secure an approach to (16) Fluent 6.2 users guide, Fluent Inc., Centerra Resource Park, 10
equilibrium values for CO and CO2. These were consequently Cavendish Court, Lebanon, NH 03766, 2005.
modified by applying the global mechanism equilibrium ap- (17) Magnussen, B. F.; Hjertager, B. H. Proc. Combust. Inst. 1971, 13,
649–657.
proach for the CO-CO2 reaction, following the procedure (18) Gran, I.; Magnussen, B. F. Combust. Sci. Technol. 1996, 119, 171–
discussed above (see eqs 2-5). This change caused the WD 190.
scheme to predict the approach to equilibrium correctly, but (19) Gran, I.; Magnussen, B. F. Combust. Sci. Technol. 1996, 119, 191–
resulted in an underprediction of CO at higher temperatures. 217.
(20) Lutz, A. E.; Kee, R. J.; Miller, J. A.;Senkin: A Fortran Program
To improve the CO prediction, it was necessary also to modify for Predicting Homogeneous Gas Phase Chemical Kinetics With SensitiVity
(WD2). Consequently, both the A-factor and the activation Analysis, Report No. SAND87-8248 Laboratories, 1990.
1384 Energy & Fuels, Vol. 23, 2009 Andersen et al.

modified to describe higher hydrocarbons by adopting the

appropriate rate data from the original mechanisms.4,6
However, the inability of the global schemes to predict
ignition timing has not been addressed.
PFR Tests of the Refined Schemes. The impact of the
modifications on the CO prediction is illustrated in Figure 4
for lean conditions and 1600 K. The modification of the WD
mechanism results in an improved prediction of the exit CO
concentration, while the peak value is slightly lowered. The
modified JL mechanism predicts a reduced peak CO level,
due to the increased value of kJL3. The major species
concentrations predicted by the modified mechanisms are
displayed in Figures 5 and 6. The results confirm that the
modified schemes constitute an improvement over the WD
and JL mechanisms for oxy-fuel conditions. Compared to
the original schemes, the prediction in particular of the CO
concentration has been improved. The improvement is most
Figure 4. CO concentrations in plug flow calculations. Comparison
pronounced for the WD mechanism, which now predicts both
between the DCKM, WD, JL, and modified mechanisms at 1600 K the trend in CO and the approach to equilibrium concentra-
and λ ) 1.2 under “oxy”-firing conditions (28% O2 and 72% CO2). tions reasonably well. For the modified JL mechanism, the
predictions of the CO trend and the peak CO levels have
of H2O (through JL3) became rate limiting. Consequently, been improved. Under fuel-rich conditions, both modified
predictions with the JL scheme were improved by modifying mechanisms approach steady-state concentrations reasonably
the rate coefficients for JL3, increasing the pre-exponential factor well, as shown in Figure 6. This indicates that the schemes
and decreasing the activation energy. may be applied also for modeling staged or slightly under-
The parameters for the fuel-specific reactions were not part
stoichiometric combustion applications.
of the modification procedure; the kinetic data for the
Figure 7 shows the performance of the modified global
initiating reactions, i.e., (JL1), (JL2) and (WD1), were all
retained. This is consistent with the flow reactor results of mechanisms when simulating a recirculated flue gas containing
Glarborg and Bentzen15 that indicate that the temperature for H2O as well as CO2; an option considered for oxy-fuel-fired
the initiation of reaction are very similar for reactive flows power plants. Calculations were conducted for an oxidizer
with and without CO2. This implies that the main chemical stream consisting of 28% O2, 40% CO2 and 32% H2O instead
difference between combustion in O2/N2 and O2/CO2 relates of the 28% O2 and 72% CO2. The results confirm that the
to the CO/H2 reaction subset. Since the fuel-specific steps modified schemes predict the major species concentrations fairly
have been retained, the proposed schemes can easily be well also under these conditions.

Figure 5. Major species concentrations in plug flow calculations. Comparison between the DCKM, the modified WD and the modified JL mechanisms
at λ ) 1.2 and 1200 K (top), 1600 K (middle), and 2000 K (bottom) under “oxy”-firing conditions (28% O2 and 72% CO2).
Global Combustion Mechanisms Energy & Fuels, Vol. 23, 2009 1385

Figure 6. Major species concentrations in plug flow calculations. Comparison between the DCKM, the modified WD and the modified JL mechanisms
at 1600 K and λ ) 0.8 (top), λ ) 1.0 (middle), and λ ) 1.2 (bottom) under “oxy”-firing conditions (28% O2 and 72% CO2).

Figure 7. Major species concentrations in plug flow calculations at 1600 K. Comparison between the modified WD, JL, and detailed mechanisms
under “oxy”-firing conditions with an oxidizer stream of 28% O2, 40% CO2 and 32% H2O.
1386 Energy & Fuels, Vol. 23, 2009 Andersen et al.

Figure 8. Temperature [K] Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental data from
Andersson and Johnsson.1 Dashed line: Modified WD mechanism. Solid line: Original WD mechanism.

scales or larger cell areas, the global mechanisms are capable

of predicting satisfactorily the heat release and major species
concentrations, provided the turbulence modeling provides
accurate turbulence level predictions as input for the EDC
turbulence chemistry interaction model.
CFD Modeling with the Refined Schemes. Recently,
Andersson and Johnsson1 reported measurements in a turbulent
diffusion flame operated under oxy-fuel conditions. The ex-
perimental setup consisted of a 100 kW down-fired refractory
lined furnace, with a swirl burner configuration. The fuel used
was propane and the oxy-fuel experiments (27% oxygen in CO2)
were compared with data obtained in air. As part of the
evaluation of the global schemes, we have conducted a CFD
analysis the oxy-fuel flame. The geometry of the setup was
simplified to a 2D case, even though the setup was not entirely
axi-symmetric, due to four cooling tubes near the furnace walls.
The heat loss to the cooling tubes was accounted for by applying
Figure 9. Temperature [K] Axial temperature at the centerline of the a piecewise constant wall temperature to obtain gas temperatures
setup. Dots: Experimental data from Andersson and Johnsson.1 Dashed near the walls in agreement with measurements. The grid was
line: Modified WD mechanism. Solid line: Original WD mechanism. constructed with ∼30.000 cells. A grid-independency check
indicated that this grid size was sufficient.
Table 5. Submodels and settings for the CFD analysis The commercial CFD code Fluent was used for the calcula-
chemistry
tion. The realizable k-ε turbulence model was adopted along
grid turbulence radiation interaction mechanisms with the P1 radiation model. Second-order upwind discretization
2D realizable P1 Eddy Dissipation modified was used for all transported scalars. The CFD settings are
30.000 cells k-ε concept (Table 3) summarized in Table 5. Both the Eddy Dissipation Model
original JL and (EDM) and the Eddy Dissipation Concept (EDC) were applied
WD (Tables 1,2) in the modeling of the oxy-fuel flame. We recommend to use
the EDC turbulence chemistry interaction model, when modeling
The turbulent mixing time scales applied in the individual oxy-fuel flames. The traditional “mixed-is-burned” approach
CFD cells are of the order of 0.1-100 ms.22 As expected, offered by the EDM is not likely to be applicable in reaction
the modified mechanisms do not match the detailed mech- systems where reverse reactions play an important role, as is
anism on the smaller timescales. This implies that the global the case under oxy-fuel conditions due to CO2 decomposition
schemes have a limited accuracy in describing changes at high temperatures. Limitations of the EDM model are
occurring on a small time scale or on an individual cell basis discussed in more detail by Brink et al.12
in a CFD computation. However, in terms of larger time The CFD results presented here were all performed using
the EDC turbulence-chemistry interaction model, applying both
(21) Kee, R. J.; Rupley, F. M.; Miller, J. A.; Coltrin, M. E.; Grcar, J. F.; the original WD and JL mechanisms and the modified versions
Meeks, E.; Moffat, H. K.; Lutz, A. E.; Dixon-Lewis, G.; Smooke, M. D.;
Warnatz, J.; Evans, G. H.; Larson, R. S.; Mitchell, R. E.; Petzold, L. R.; suggested in Table 4. Since the fuel used in the experiments is
Reynolds, W. C.; Caracotsios, M.; Stewart, W. E.; Glarborg, P.; Wang, C.; propane, rate data and stoichiometric relationships for the
Adigun, O.; Houf, W. G.; Chou, C. P.; Miller, S. F.; Ho, P.; Young, D. J.; initiating reactions (JL1, JL2, and WD1, see Table 4) were
CHEMKIN Release 4.0, Reaction Design, Inc., San Diego, CA ( 2004).
(22) Kjaldman, L.; Brink, A.; Hupa, M. Combust. Sci. Technol. 2000, modified according to the original references.4,6 These reactions
154, 207–227. are identical in the original and modified versions.
Global Combustion Mechanisms Energy & Fuels, Vol. 23, 2009 1387

Figure 10. Radial CO concentration (%dry) Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental
data from Andersson and Johnsson.1 Dashed line: Modified WD mechanism. Solid line: Original WD mechanism.

Figure 11. Radial O2 concentration (%dry) Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental
data from Andersson and Johnsson.1 Dashed line: Modified WD mechanism. Solid line: Original WD mechanism.

Figure 12. Temperature [K] Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental data from
Andersson and Johnsson.1 Dashed line: Modified JL mechanism. Solid line: Original JL mechanism.

A satisfactory agreement between major species concentra- was achieved for the air case. Here, only results from the
tions predicted by the CFD model and experimental results oxy-fuel case is presented. It should be noted that deviations
1388 Energy & Fuels, Vol. 23, 2009 Andersen et al.

field is essential; unfortunately, no data for velocity and

velocity fluctuations are available from the oxy-fuel experi-
ments. Radiative properties are also affected when changing
from air to oxy-fuel combustion, and it is likely that
differences in the temperature profiles are due to insufficient
radiative modeling along with the simplifications in geometry
(neglection of the cooling tubes). Furthermore, soot forma-
tion, which is important for the flame radiation, was not
accounted for in the modeling.
Figures 8 and 9 show comparison between experimental
results1 and modeling predictions with the WD schemes for
the radial and centerline temperature profiles, respectively,
while Figures 10 and 11 show results for radial CO and O2
concentrations. The CO and O2 concentrations are shown for
two measurement planes, 21.5 and 38.4 cm downstream of
the burner. Solid lines denote calculations with the original
WD scheme, while dashed lines donte predictions with the
Figure 13. Temperature [K] Axial temperature at the centerline of the modified WD scheme. In general, the modified WD mech-
setup. Dots: Experimental data from Andersson and Johnsson.1 Dashed anism provides an improved prediction of both the temper-
line: Modified JL mechanism. Solid line: Original JL mechanism.
ature and the postflame conditions and emissions, compared
between the CFD predictions and the experimental results to the original scheme. The modified mechanism predicts well
can only partly be attributed to an insufficient combustion the CO levels in the region outside the flame zone (Figure
mechanism. A satisfactory prediction of turbulence and flow 10). The improvement can be attributed to the new equilib-

Figure 14. Radial CO concentration (%dry) Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental
data from Andersson and Johnsson.1 Dashed line: Modified JL mechanism. Solid line: Original JL mechanism.

Figure 15. Radial O2 concentration (%dry) Left: at 21.5 cm downstream of the burner. Right: at 38.4 cm downstream of the burner. Dots: Experimental
data from Andersson and Johnsson.1 Dashed line: Modified JL mechanism. Solid line: Original JL mechanism.
Global Combustion Mechanisms Energy & Fuels, Vol. 23, 2009 1389

rium fit for the WD3 reaction. The modified model also initiation reactions involving the hydrocarbon fuel were unal-
predicts a different CO level in the flame zone, but due to tered. Changes were made in the CO-CO2 reaction subset to
the limited experimental data points for CO in this region, obtain an improved fit for CO levels and steady state emissions
both sets of predictions are considered to be within experi- as predicted by the detailed mechanism.
mental uncertainty. The differences in the predicted CO levels The modified schemes provide improved agreement with the
do, however, emphasize the importance of the chemical detailed mechanism compared to the original mechanisms for
mechanism in the CFD computation. In the very fuel-rich isothermal plug flow reactor modeling under oxy-fuel combus-
parts of the flame, the CO levels are underpredicted in this tion conditions. The improvement was most pronounced for WD
CFD analysis. This indicates that it may be necessary to apply mechanism, where the modified scheme yielded a better
a more complex model than the WD scheme for reliable prediction of the peak and exit concentrations of CO. The
modeling of fuel-rich regions of a combustion system. modified JL mechanism offers a slight improvement in predict-
Comparisons for temperature profiles (Figures 12 and 13) and ing CO trends.
CO/O2 concentrations prediction (Figures 14 and 15) have also A CFD analysis of a propane oxy-fuel flame was performed
been conducted for the JL schemes. Figures 14 and 15 compare using both the original and modified mechanisms. The
CO and O2 concentrations from the original JL mechanism and modified JL mechanism performed slightly better than the
the modified mechanism with the experimental measurements at original, regarding CO predictions in the flame zone. The
two measurement planes 21.5 and 38.4 cm downstream of the modified WD mechanism improved the prediction of CO
burner. From Figure 14, it can be seen that the modified JL especially in the post flame zone, but also a better temperature
mechanism predicts a lower CO concentration in the center of the agreement with experimental data was obtained. In general,
furnace inside the flame, improving the agreement with experiment. it is recommended to use the EDC turbulence chemistry
The prediction of a lower CO level by the modified scheme is interaction model, when modeling oxy-fuel flames. The
consistent with the PFR calculations discussed above. The modi- traditional “mixed is burned” approach offered by the Eddy
fications in the JL mechanism also induce changes in the temper- Dissipation Model is not likely to be applicable in a
ature predictions (Figures 12 and 13) and the O2 prediction (Figure combustion system where the CO2 dissociation can have a
15), but whether these changes constitute improvements over the significant chemical effect.
original mechanism cannot be concluded.
Acknowledgment. This work was funded by Vattenfall Research
Conclusions and Development AB with Karin Eriksson as project manager. The
authors would like to thank Anders Brink and W.P. Jones for helpful
Two global multistep mechanisms, the two-step mechanism discussions.
by Westbrook and Dryer and the four-step mechanism by Jones
and Lindstedt, have been tested and refined for oxy-fuel
conditions, based on comparison with model predictions with
a detailed chemical kinetic mechanism. In the modification, the EF8003619