Fuel 201 (2017) 148–155

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Super-adiabatic flame temperatures in premixed methane flames: A

comparison between oxy-fuel and conventional air combustion
Björn Stelzner ⇑, Christof Weis, Peter Habisreuther, Nikolaos Zarzalis, Dimosthenis Trimis
Karlsruhe Institute of Technology, Engler-Bunte-Institute, Combustion Technology, Karlsruhe, Germany

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

Article history: As an alternative to usual combustion processes with air as oxidant, oxy-fuel processes are attractive in
Received 30 May 2016 high temperature thermal or thermochemical processes, novel power plant concepts or in gasification
Received in revised form 4 November 2016 processes. Especially for production of basic chemicals or synthetic fuels, high-purity synthetic gas with-
Accepted 5 January 2017
out nitrogen dilution is required for further industrial processing.
Available online 13 January 2017
In case of oxy-fuel combustion, the absence of nitrogen leads to higher flame temperatures and larger
concentration of major species as well as intermediate species. In the present work, freely propagating
Keywords:
methane-oxygen-nitrogen flames were numerically calculated using a 1D model from lean to rich condi-
Oxy-fuel
Premixed
tions in order to investigate the appearance of super-adiabatic flame temperatures (SAFT). The calcula-
Super adiabatic tions were performed for equivalence ratios of 0.5 < U < 3.0 with an increment of 0.1. Additionally, the
Flame temperature nitrogen content of the oxidizer was varied from air to pure oxy-fuel conditions. Furthermore, the influ-
Partial oxidation ence of preheat temperature and pressure on SAFT were analyzed for oxy-fuel flames.
The SAFT phenomenon was identified based on temperature and species profiles. Additionally, a
detailed analysis of convection, diffusion and chemical source of both, temperature and species, were per-
formed in order to investigate the role of physical transport processes on SAFT.
The results showed that the maximum flame temperature exceeds the equilibrium temperature for
equivalence ratios U > 0.9. Two different regimes were identified, where SAFT phenomenon appears.
The first regime was found in slightly rich conditions (1.0 < U < 2.0), whereas the second regime occurred
in ultra-rich regime (U > 2.0). A first maximum of temperature difference is observed at an equivalence
ratio of U = 1.5. Here, an exaggeration of approximately 120–180 K, depending on the applied reaction
mechanism, was found for standard conditions. The first maximum at U = 1.5 correlates with the maxi-
mum concentration of the H-radical, which plays a key role in the first SAFT regime. A minimum temper-
ature difference of 50 K was identified at an equivalence ratio of U = 2.1. While increasing the
equivalence ratio further, the maximum flame temperature exceeded the equilibrium up to almost
400 K at U = 3.0 in the second SAFT regime.
An increased preheating temperature enhanced the occurrence of SAFT in the first regime and degraded
it in the second regime. Elevated pressure leads to the opposite effects with decreased SAFT in the first
and increased SAFT in the second regime.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction icals or synthetic fuels, high-purity synthetic gas without nitrogen

As an alternative to usual combustion processes with air as oxi-
dant, oxy-fuel processes are attractive in high temperature thermal
or thermochemical processes, novel power plant concepts or in
gasification processes. Oxy-fuel processes provide significantly
higher flame temperatures and exhibit advantages as increased
thermal efficiency, higher processing rates, reduced flue gas vol-
umes and the potential for reduced pollutant emissions at high
oxygen concentrations [1]. Especially for production of basic chem-
⇑ Corresponding author.
E-mail address: bjoern.stelzner@kit.edu (B. Stelzner).
dilution is required for further industrial processing.
An interesting phenomenon in oxy-fuel combustion is the local
appearance of super-adiabatic flame temperatures (SAFT) which
has been identified and investigated in a few studies.
Meeks et al. [2] predicted in numerically calculated
C2H2/O2/H2-flames, that the local flame temperatures exceed sig-
nificantly the adiabatic equilibrium temperature, even in case of
including massive radiative heat losses of the gas phase in the
calculations. The occurrence of SAFT in this flame was demon-
strated experimentally via CARS spectroscopy by Bertagnolli and
Lucht [3] and Bertagnolli et al. [4].

http://dx.doi.org/10.1016/j.fuel.2017.01.025
0016-2361/Ó 2017 Elsevier Ltd. All rights reserved.
 B. Stelzner et al. / Fuel 201 (2017) 148–155
Ruf et al. [5] recognized SAFT in rich premixed C2H2-O2 flames.
Homogeneous calculation of the gas mixture showed an equilib-
rium H-concentration of 24%, whereas only 3% were experimen-
tally detected. They proposed that the reaction H2 + M M H + H
+ M is not fast enough in order to establish equilibrium at their
boundary conditions.
Liu et al. [6] performed a systematic study of SAFT in premixed
CH4/air-, different hydrocarbons/O2- and H2/O2-flames. They
showed that the occurrence of SAFT phenomena comes along with
super-equilibrium concentrations of some hydrocarbon species
and H2O in the flames and results in a local temperature overshoot.
It should be noted, that in premixed H2/O2-flames SAFT could not
be detected.
Zamashchikov et al. [7] studied numerically super-adiabatic
flame temperatures of CH4- and C3H8-air in fuel-rich regime. The
findings indicate that the degree of SAFT depends on the equiva-
lence ratio, while the phenomenon occurs in rich regime. They pro-
posed, that the diffusion of H2 from the reaction zone to the
preheat zone and its preferential oxidation is responsible for SAFT.
Liu and Gülder [8] analyzed the reaction flux in premixed CH4/
air and CH4/O2-flames by artificially changing diffusion properties
of H2 and H radicals in their calculations. In contrast to [7], they
concluded that the relative scarcity of H-radicals at the end of
major heat release reactions and not the preferential diffusion of
H2 causes SAFT phenomenon. The H-radicals at high temperatures
enhance the radical chain branching reaction H + O2 , O + OH. The
reduction of the H-radical in the reaction zone is correlated to a
temperature overshoot.
The effects of pressure and preheating on SAFT in rich premixed
CH4/air flames were studied for equivalence ratios of / = 1.4 to /
= 1.8 by Liu and Gülder [9]. Preheating leads to elevated combus-
tion temperatures, which finally suppress a H2O overshoot in the
flame and hence, the SAFT phenomena. In pressurized systems,
the degree of SAFT is enhanced.
Kumagami et al. [10] calculated laminar premixed CH4/O2/H2O-
flames for fuel rich conditions under elevated pressures up to
3 MPa. They identified two different SAFT regimes: the first regime
occurs for equivalence ratios of 1.0–2.0 and low water dilutions.
The SAFT phenomenon gets less pronounced with increased pres-
sure in the first regime, whereas it occurs in the second regime
under all pressure conditions. Furthermore, it was shown, that
the laminar burning velocity decreases with increased pressure
for the first regime, whereas a slight increase with pressure was
found under more fuel-rich conditions.
Takagi and Xu [11] showed numerically that SAFT is correlated
by preferential species diffusion of H and H2 in case of non-
premixed combustion.
The literature review revealed that all studies focused on a
small range of equivalence ratios and either on air or oxy-fuel com-
bustion. The present study focuses on the SAFT phenomenon in
gasification and oxy-fuel processes. Therefore, CH4-O2/N2 flames
are systematically investigated with established kinetics sets.
Hereby, the oxygen-content of the oxidizer is varied from air to
pure oxy-fuel combustion. All conditions were investigated for a
wide range of equivalence ratios from / = 0.5 to / = 3.0 in order
to exhibit global coherence.
Since industrial processes are operated with preheat and ele-
vated pressure, their influence on SAFT is additionally analyzed
in pure oxy-fuel flames.
2. Numerical approach
In the present study, the numerical calculations were carried
out with the reactor for freely propagating premixed flames of
the software package CHEMKIN 4.1. The PREMIX code was origi-
nally developed by Kee et al. [12,13].
149
The GRI 3.0 was used as a reference kinetic mechanism in the
calculations [14], consisting of 53 species and 325 reactions. Addi-
tionally, the mechanism of Appel et al. [15] (99 species and 544
reactions) with detailed C2-pathway chemistry was used for
selected cases. Unless otherwise stated, all presented results were
carried out with the GRI 3.0 mechanism.
An initial grid of 24 points was used, which was adaptively
refined to approximately 250 grid points. Grid independency was
ensured in preliminary investigations. Full multicomponent and
thermal diffusion were considered. A total reactor length of
10 cm was chosen for all calculations. The feed consists of CH4-
O2/N2 mixtures.
The comparison of air to oxy-fuel conditions were performed
for standard thermodynamic inlet conditions (STC, 300 K, 1 atm).
The oxygen content of the oxidizer was varied from 0.21 to 1.0
with in increment of 0.1. The flames were studied for a range of
equivalence ratios of 0.5 < / < 3.0 with an increment of 0.1.
Additionally, the oxy-fuel flames were calculated with three dif-
ferent inlet temperatures of 300 K, 500 K and 700 K, respectively,
and the pressure was varied from 0.1 MPa to 1 MPa for / = 1.0 to
/ = 3.0 with an increment of 0.5.
3. Results and discussions
Since oxy-fuel combustion leads to significant higher flame
temperatures and species concentrations than in air combustion,
substantial dissociation and hence amount of radicals may
expected. Selected species of the H-O-system and the temperature
in equilibrium state for methane-oxygen are presented in Fig. 1 for
various equivalence ratios at standard inlet conditions.
The maximum equilibrium temperature is found at slightly
rich conditions at equivalence ratios of / = 1.0 to / = 1.2. Here,
the temperature varies in a range of approximately 3050 K only
a few Kelvins. Those high temperatures lead to notable dissocia-
tion products. An amount of 8% molecular oxygen is found at
stoichiometry and drops at an equivalence ratio of / = 1.5 below
1%. The peak of the O-radical (5%) and OH radical (10%) are
located on the lean side, whereas the H-radical finds its maxi-
mum of 6.5% in rich conditions at an equivalence ratio of /
= 1.4. Here, still an extremely high equilibrium temperature of
2973 K and subsequently large H-mole fraction are present and
hence, a large diffusion flux of the H-radical may be expected
due to high concentration and temperature gradients under
those conditions.
Fig. 1. Equilibrium temperatures and selected species at STC depending on
equivalence ratio.
150 B. Stelzner et al. / Fuel 201 (2017) 148–155

3.1. The appearance of SAFT

In the following, the super-adiabatic flame temperature (SAFT)

is defined as difference between local maximum temperature in
the flame and equilibrium temperature calculated with STANJAN
code. All three values are shown in Fig. 2 for pure CH4-O2-
compositions at STC and calculated with different reaction mecha-
nisms in a range of an equivalence ratio of / = 0.5 to / = 3.0.
SAFT is observed under fuel rich conditions for / > 1, slightly
depending on the applied reaction mechanism. The first peak of
the difference of maximum flame temperature and equilibrium
temperature is found with approximately 150 K at an equivalence
ratio of / = 1.5. Similar values were found in [8]. At equivalence
ratios of / > 1.5, no significant molecular oxygen exists at equilib-
rium, thus most of the oxygen will be consumed in the oxidation
process (compare Fig. 1). Additionally, the highest H-
concentration in equilibrium is found at this equivalence ratio.
The results are more affected by the chosen reaction mechanism
in the region of / = 1.0 to / = 1.5.
At an equivalence ratio of / = 2.1 the still evident over-
temperature decreases to a minimum of 50 K. An increase of the
temperature difference, up to 400 K at / = 3.0, follows in the
ultra-rich regime.
It should be noted that the degree in the first SAFT regime
Fig. 3. SAFT for different oxygen content depending on the equivalence ratio at STC.
depends on the applied reaction mechanism, whereas the mini-
mum / = 2.1 as well as the second SAFT regime shows no signifi-
cant influence by the used kinetic set. The second regime shows similar behavior and magnitude in all
flames. At low oxygen concentrations, the appearance of SAFT
starts at an equivalence ratio of U = 1.4 and shifts towards richer
3.2. Influence of oxygen-content conditions with increasing oxygen content. In the transition of
both SAFT regimes a local minimum is found.
In the following, the oxygen-content of the oxidizer was varied Liu and Gülder [8] showed for CH4-air-flames in a range of
stepwise from air conventional combustion (x = 0.21) to pure oxy- U = 1.4 to U = 1.75 that the occurrence of SAFT-phenomenon is
fuel (x = 1.0). Here, the maximum equilibrium temperature is regardless of the treatment of species diffusion modeling. In con-
changed from approximately 2200 K to 3050 K at stoichiometry trary, the degree of SAFT in CH4-O2 flames is strongly affected by
and hence, diffusion transport and kinetics are strongly modified. the chosen diffusion model in a range of U = 1.0 to U = 2.0.
The results of the SAFT are presented in Fig. 3 for standard inlet The presented results indicate that two SAFT regimes with dif-
conditions. ferent physical reasons exist. The first regime (sensitive to diffu-
The results reveal that the first and second SAFT-regime show sion modeling [8]) occurs only above a certain temperature level
different behavior while varying the oxygen content of the oxi- and its magnitude increases in high temperature flames, whereas
dizer: at concentrations of 0.21 to 0.4, no SAFT is recognized and the second regime (insensitive to diffusion modeling and hence
is first notable at xO2,Ox = 0.5 at U = 1.7. With increasing the oxygen not related to preferential diffusion of H or H2 [8]) is independent
content further, the degree of the first regime enhances, where the of oxygen content and/or flame temperature. In the following, the
peak of the SAFT shifts towards stoichiometry to U = 1.4 in a pure scope is to apply the findings of Liu and Gülder on the two different
oxy-fuel flame. SAFT regimes.

3.3. Identification of SAFT regimes

Liu et al. [6] stated that SAFT is strongly correlated by a local

Fig. 2. Maximum flame temperature, equilibrium temperature and their difference
depending on the equivalence ratio at STC.
H2O-overshoot in the flame and its degree is roughly proportional
to the degree of SAFT. In the following, two representative flames
for each regime were selected, namely pure oxy-fuel flames at
U = 1.5 and U = 3.0. The profiles of the main species are plotted
in Figs. 4 and 5, respectively. Additionally, the equilibrium values
for temperature and the concentration of H2O, H2, CO and CO2
are indicated in the plots.
In both cases, the CH4 and O2 are rapidly consumed and the for-
mation of the products of the synthesis gas starts. Both flames
show an overshoot in the H2O-production and hence the tempera-
ture. The H2O is finally consumed further downstream in endother-
mic reactions (mainly due reversed #84: OH + H2 , H + H2O),
which lowers subsequently the flame temperature in the post
flame zone to equilibrium state.
In the first SAFT-regime an overshoot in the H2- and CO-
production is recognized. Additionally, it should be noted that in
 B. Stelzner et al. / Fuel 201 (2017) 148–155
Fig. 4. Main species of pure CH4-O2 at an equivalence ratio of U = 1.5 at STC.
Fig. 5. Main species of pure CH4-O2 at an equivalence ratio of U = 3.0 at STC.
the second regime the CH4 is not completely consumed and still
present in the post flame zone and equilibrium state still not
reached.
A comparison of the main species between equilibrium state
and local concentration at the position of maximum temperature
in the flame are presented in Figs. 6 and 7 for CH4-air- and CH4-
O2-flames.
In CH4-air flames, SAFT arises from U > 1.4. Under this condi-
tion, overshoot in the local H2O-production can be observed. The
trends of SAFT and the local H2O overshoot are in agreement with
the relevant literature: similar behavior was found by Liu and
Gülder [6,8] and Zamashchikov [7] (here in C3H8-air flames).
The H2O-concentration stays almost constant at approximately
x = 0.17, even when the equivalence ratio is increased (an analysis
of the H2O formation and decomposition is given for both SAFT
regimes in Section 3.5). At the same time, the equilibrium state
drops constantly, so that the relative difference enhances the local
degree of SAFT. The minor decrease of the H2O may be attributed to
the presence of non-converted CH4 at high equivalence ratios. In
this range, the concentrations of H2 and CO are under their equilib-
rium values, whereas the CO2 almost reaches this level.
In case of pure oxy-fuel combustion, a slight overshoot in the
H2O-production is observed from 1.0 < U < 2.0. Additionally, the
CO overshoots and the CO2 has not reached equilibrium state.
These phenomena are not detected in air flames. At an equivalence
ratio of U > 2.0, the H2O formation transpose to an almost constant
value of x = 0.32, whereas the CO- and the H2-production are also
still below equilibrium state. Whatever the case, it can be noted
151
Fig. 6. Comparison of main species concentration between equilibrium state and
the position of maximum temperature in CH4-air flames at STC.
Fig. 7. Comparison of main species concentration between equilibrium state and
the position of maximum temperature in CH4-O2 flames at STC.
that both SAFT regimes are correlated to a local overshoot in the
H2O-formation in the flames.
An analysis of the data revealed (not shown here) that in all
cases the oxygen is bound almost equally in H- and C-
components (mainly H2O, CO and CO2) at the major heat release
zone of the flame, whereas the H2O-production is slightly preferred
in magnitude as well as in time scale with higher equivalence ratio.
However, Figs. 8 and 9 show that the H-atoms are mostly not oxi-
dized but rather bound in H2 in equilibrium state while increasing
the equivalence ratio. Subsequently, the water gas shift reaction
has to take place in the post flame zone and shifts the O-atom to
C-components in the form of CO.
3.4. Role of the H-radical
As stated by Zamashchikov et al. [7], the local element compo-
sition is significantly changed in the main reaction zone due to
preferential diffusion. The local element ratio and the resulting
equivalence ratio are presented for CH4-O2-flames at / = 1.5 and
/ = 3.0 in the following figures. Both cases show an increased
H-content upstream of the temperature peak in the flame. In com-
bination with the decreased C-content, the H/C-ratio is increased
152 B. Stelzner et al. / Fuel 201 (2017) 148–155
Fig. 8. Local element ratio and equivalence ratio (normalized to inlet) for CH4-O2-
flame at / = 1.5 at STC.
Fig. 9. Local element ratio and equivalence ratio (normalized to inlet) for CH4-O2-
flame at / = 3.0 at STC.
significantly. The mixtures become leaner and are followed by a
shortage of O, which leads to a locally richer mixture than the
fed. The magnitude of the change in the element composition is
similar in both flames.
As a characteristic of the process, the equilibrium temperature
Tequi, the local maximum temperature in the flame Tmax and the
maximum of the local equilibrium temperatures Tequi,local,max
(based on the change of local element balance due to preferential
diffusion) are plotted in Fig. 10 for oxy-fuel flames at varying
equivalence ratio. The scope is to verify that SAFT is not only a phe-
nomenon of a shift in the element balance.
The shift in the element balance leads to a higher possible equi-
librium temperature than the one based on the inlet fed in all cases
and increases with increasing equivalence ratio. In the first SAFT
regime at approximately / = 1.5 as well for flames / > 2.5, the
maximum flame temperature exceeds the possible equilibrium
temperature and SAFT occurrence is not only based on local shift
in the element balance. However, it should be mentioned that let-
ting the local state reach its equilibrium, which is represented by
Tequi,local, implies a closed system and, hence, the fluxes occurring
in the flames are neglected.
An analysis of the profiles of the H-radical in the flames (not
shown here) confirm the conclusion of Fig. 1 that the highest
Fig. 10. Comparison of equilibrium temperature including local element balance
with flame temperatures for oxy-fuel flames.
H-concentration is present in slightly rich flames. For equivalence
ratios of / = 1.0 and / = 1.5, the H-radical reaches equilibrium in
the flame front without any overshoot. This behavior was found
to be characteristic for this regime. A small peak is recognized at
/ = 2.0, where the H-radical is consumed shortly after till it finally
reaches equilibrium state. In ultra-rich cases of equivalence ratios
of / = 2.5 and / = 3.0, the H-radical reaches super-equilibrium and
drops in the post flame zone to equilibrium. In the second regime
(as well in conditions where no SAFT occurs at all) a typical over-
shoot is recognized in the reaction zone.
The convection, source and diffusion term of the temperature
and the species equation for the H-radical are separately plotted
for each regime (CH4-O2 at / = 1.5 and / = 3.0) in Figs. 11 and 12,
respectively. It should be noted that the diffusion term of the H-
radical is plotted inverted in order to have better comparison with
its source term.
The terms of the temperature equation in the flame of / = 3.0
show a relative wide non-reactive preheat zone, where convection
and diffusion keep the balance. In the reaction zone, the tempera-
ture source is mainly distributed by diffusion. The post flame zone
shows the negative endothermic reactions. In contrast, the flame of
the first SAFT regime (at / = 1.5) exhibits no distinct preheat zone
and the temperature convection plays an important role in the
main heat release domain.
With respect to the species equation of the H-radical, the spe-
cies convection is negligible for the H-balance at / = 3.0; the source
term is almost completely balanced by diffusion and distributes
Fig. 11. Convection, source and diffusion term of the temperature equation and
species equation of the H-radical for CH4-O2-flame at / = 1.5.
 B. Stelzner et al. / Fuel 201 (2017) 148–155
Fig. 12. Convection, source and diffusion term of the temperature equation and
species equation of the H-radical for CH4-O2-flame at / = 3.0.
the H-radical mainly to the preheat zone and slightly to the post
flame zone. Furthermore, the H-diffusion and H-consumption are
strongly attached to the heat release (temperature source term)
in the beginning of the reaction zone which suggests that temper-
ature limits the H-source. Since the H-radical is not transported by
convection, only the diffusion is active for balancing.
On the contrary, the flame at / = 1.5 shows a significant species
convection upstream of the flame, but also a strong diffusion flux
upstream to the preheat zone before the temperature source. This
implies that the kinetics is not limited to the supply by diffusion.
Furthermore, there is still a formation of the H-radical in the
post flame zone. Additionally, the diffusion flux is much higher
than the consumption compared to the case at / = 3.0, which
finally provides a sufficient H-pool in order to form H2O.
In order to evaluate the origin of the H-radical in the flames, its
rate of production was examined. The rate of production of the H-
radical is presented for the two selected cases for equivalence
ratios of / = 1.0 and / = 3.0 and are presented in Figs. 13 and 14,
respectively.
The analysis of the production rates of the H-radical revealed
that in all cases substantial amount of the H is formed by reaction
#84 (OH + H2 , H + H2O). The main consumption of the H-radical
is required in all cases for both oxygen destruction reactions #38
(H + O2 , O + OH) and the fuel destruction reaction #53 (H + CH4 -
, CH3 + H2), respectively, where the consumption is shifted to
reaction #53 with increased equivalence ratio.
Fig. 13. Rate of production of the H-radical for an equivalence ratio of / = 1.0 at
STC.
153
Fig. 14. Rate of production of the H-radical for an equivalence ratio of / = 3.0 at
STC.
Since substantial amount of oxygen is still present under low
rich conditions (compare Fig. 1), high temperature oxidation reac-
tions (formaldehyde ? formyl ? carbon monoxide) produce and
consume the H-radical in the first SAFT regime. In ultra-rich
flames, the production and consumption shifts to the C2-pathway
in absence of oxygen. A balance of high temperature oxidation
reactions and C2-pathway-reactions were found at an equivalence
ratio of approximately / = 2.0. The turn over point of the reaction
path way correlates with the minimum of SAFT phenomenon
(see Fig. 2).
Furthermore, the competition of H-production by reaction #84
and H-consumption by reaction #38 and #53 is indicated by the
peak position of #84 in Figs. 13 and 14. The first SAFT regime is
dominated by rapid H-consumption, where reactions with high
activation energies like #38 take place earlier. The total reaction
rate shows a distinct negative production rate in the preheat zone
and has to be fed by diffusive H-transport. In the second regime at
higher equivalence ratios (and hence lower flame temperatures)
the degree of H-consumption is reduced in the preheat zone. The
preferential diffusion of the H-radical prompts an overshoot in
the H2O-production [7].
3.5. Analysis of H2O-formation
An overshoot of the H2O-production is noticed in all cases,
where SAFT occurs. It becomes more pronounced, compared to
equilibrium state, with increased equivalence ratio. The local
super-equilibrium production of H2O is closely related to the
degree of SAFT [8]. Further downstream in the flames, the con-
sumption of H2O takes place by endothermic reactions, which low-
ers subsequently the flame temperature in the post flame zone.
The rate of production of H2O is shown in Fig. 15 for / = 1.5
(first SAFT regime) and Fig. 16 for / = 3.0 (second SAFT regime)
at STC.
The results show that the reactions #84 and #98 are notewor-
thy and responsible for the main water production, especially in
ultra-rich conditions. In the first SAFT regime (Fig. 15), the oxida-
tion pathway of CH4 (reactions #96, #97, #98) produces an addi-
tional amount of water. Furthermore, H2O-dissocition to OH is
observed in the main reaction zone. In the post flame zone, the
H2O-consumption occurs due to third body reaction #43 and
#84. Additionally, the HO2- and H2O2-radical are involved in the
H2O-degradation (#89 and #287). Equilibrium state is reached rel-
atively fast due to high temperatures.
In the second SAFT regime, H2O is mainly produced by both
reactions # 84 and #96 due to the consumption of OH-radical
154 B. Stelzner et al. / Fuel 201 (2017) 148–155
Fig. 15. Rate of production of H2O for an equivalence ratio of / = 1.5 at STC.
Fig. 16. Rate of production of H2O for an equivalence ratio of / = 3.0 at STC.
upstream the position of the maximum flame temperature. After-
wards downstream the location of the maximal temperature, the
H2O is mainly consumed and H2 is produced due to the reaction
#84 in reversed direction. Here, a concentration of almost 8%
methane is still present in the flame and is slowly consumed by
reaction #53, which provides the required H-radical in order to
feed the H2O-consumption in the post flame zone. The released
OH-radical of reaction #84 is partly consumed in reaction #99
(OH + CO2 , H + CO) for CO-production.
3.6. Influence of preheating temperature on SAFT
The influence of preheating temperature was investigated for
the pure oxy-fuel flames. Increased preheat alters the radical com-
positions of the H-O-system to higher concentrations due to stron-
ger dissociation. The maximum H-concentration changes from
6.5% to 7.5% in equilibrium state when the preheat temperature
is increased from 300 K to 700 K at an equivalence ratio of /
= 1.4. In contrast to the radicals, there is no clear effect on the main
species; the H2O concentration for example is increased for equiv-
alence ratios / < 2.3 and slightly decreased for / > 2.3, both with
increasing preheating temperature.
The local maximum flame temperature, the equilibrium tem-
perature and their difference is shown in Fig. 17 for different inlet
temperatures at atmospheric pressure.
Fig. 10 shows that the first SAFT peak at an equivalence ratio of
/ = 1.5 increases with increased preheat. The minimum SAFT shifts
slightly from / = 2.1 at 300 K to richer conditions of / = 2.3 at
Fig. 17. Maximum flame temperature, equilibrium temperature and their differ-
ence depending on the equivalence ratio for different inlet temperatures at a
pressure of 0.1 MPa.
700 K. In the second SAFT regime, an increase of the inlet temper-
ature leads to a smaller temperature difference and reduces the
SAFT phenomenon.
In the first SAFT regime, the increased temperature level and
radical concentration lead to higher gradients and hence diffusion
effects. The enlarged H-pool in the preheat zone has been
observed. Additionally, the H2O concentration in equilibrium state
decreases slightly with preheating. The H2O overshoot in the flame
increases relative to the maximum concentration found in the
flame. On the contrary, the H2O concentration increases with pre-
heat in equilibrium for the second SAFT regime and leads to a rel-
ative lower H2O overshoot. The trends of the relative H2O
overshoot are in accordance to the SAFT behavior with preheat.
3.7. Influence of pressure on SAFT
With increasing the pressure, the main species composition
alters in equilibrium near stoichiometric conditions. Such high
temperatures of approximately 3000 K lead to significant radical
pool due to dissociation, which combines while increasing the sys-
tem pressure. Subsequently, the concentration of H2O and CO2 and
hence the adiabatic equilibrium temperature increase. In rich mix-
tures, no significant change of the equilibrium gas composition and
temperature are observed while increasing the pressure.
An analysis of the equilibrium temperatures revealed that an
increase of the temperature is most pronounced at stoichiometric
conditions. A pressure increase to 1 MPa elevates the temperature
by approximately 300 K. In ultra-rich flames (/ = 2.5 and / = 3.0)
the effect becomes negligible. Furthermore, it should be noted, that
the pressure elevation also leads to a substantial reduced H-radical
pool in equilibrium. The degree of SAFT, the difference of maxi-
mum flame temperature and equilibrium temperature, is pre-
sented in Fig. 18 for selected equivalence ratios in a pressure
range of 0.1 MPa to 1 MPa.
In the first regime (1.0 < / < 2.0), the temperature difference
decreases with increased pressure. At / = 2.0 an almost constant
temperature difference independent of pressure is observed. A bal-
ance of the SAFT decrease of the first and the SAFT increase of the
second regime may be assumed. An increase of the temperature
difference is found under ultra-rich conditions (from approxi-
mately 380 K to 500 K at U = 3.0).
The tendency of temperature overshoot in the flames is also vis-
ible in the overshoot of the H2O-production. The analysis of the
H2O-formation revealed that similar to the temperature
dependency on pressure, the H2O-concentration is increased near
 B. Stelzner et al. / Fuel 201 (2017) 148–155
Fig. 18. Difference of maximum flame temperature and equilibrium temperature
depending on the pressure for different equivalence ratios at an inlet temperature
of 300 K.
stoichiometry in equilibrium state and is unaffected in ultra-rich
mixtures with increased pressure. In the flame, the maximum
H2O-concentration increases constantly with elevated pressure so
that the relative H2O-overshoot increases. Approaching stoichiom-
etry (/ = 1.0 and / = 1.5) the H2O-overshoot vanishes and there-
with the SAFT phenomenon. Since the first SAFT regime is
strongly influence by preferential diffusion of the H-radical, the
result is reasonable due to a reduced diffusion flux. In ultra-rich
regime (the second SAFT regime) with slower kinetics, the ten-
dency is contrary and SAFT is enhanced.
4. Conclusions
The SAFT phenomenon was parametrically investigated for pre-
mixed CH4-N2/O2-flames for different equivalence ratios from con-
ventional air to oxy-fuel combustion. Additionally, the preheat
temperatures and pressures for the pure oxy-fuel cases were
examined.
The results revealed that SAFT occurs for all fuel rich conditions
in premixed CH4-O2-flames. Two different SAFT regimes were
identified: in the slightly rich regime, a local maximum of the
flame temperature occurs at an equivalence ratio of / = 1.5 and
exceeds by about 150 K the overall equilibrium state temperature
(SAFT regime 1). A minimum of 50 K over-temperature was found
at an equivalence ratio of / = 2.1, while at higher stoichiometry the
SAFT effect increases again (SAFT regime 2). A turn-over point in
the kinetics from the oxidation- to the C2-pathway at this equiva-
lence ratio is supported by the presented results.
With lowering the oxygen content of the oxidizer and hence the
temperature level of the flame, the degree of SAFT in regime 1
decreases and disappears at an oxygen concentration of approxi-
mately 0.4. The SAFT regime 2 is in magnitude insensitive to the
oxygen-content, but shifting slightly towards stoichiometry while
transposing to air combustion.
In the present study, the rate of production of the H-radical and
H2O-molecule were analyzed in order to identify the mechanisms
of the two different SAFT regimes. The SAFT phenomenon occurs
in both regimes due to an overshoot of H2O-production.
The analysis of the calculated data revealed that in both SAFT
regimes, the preferential diffusion of the H-radical plays a key role
and alters the element composition in the main reaction domain.
The results support the findings of Liu and Gülder [8]. However,
in the first regime, the kinetics is very fast and locally almost in
155
equilibrium, whereas in the second regime the time scales of the
fuel oxidation and endothermic synthesis gas production diverge.
In the first regime, the H-radical is rapidly consumed by both
reactions #53 (H + CH4 , CH3 + H2) and #84 (OH + H2 , H + H2O)
faster than produced by reaction #38 and hence, the maximum
level of H-concentration does not exceed equilibrium state. The
sufficient level of oxygen leads to relative high participation of
reactions in the high temperature fuel oxidation pathways.
In the second regime, the degree of rapid H-consumption is
reduced due to lower flame temperatures, so that an adequate H-
pool higher than at equilibrium state is established. Next to the
main reactions #38 (H + O2 , O + OH), #53 and #84, the H-pool
is strongly influence by pyrolysis reaction of the C2-pathway. The
H2O is mainly formed by reaction #53 and #84, where the H2O-
consumption in the post flame zone takes place only by reaction
#84 in reversed direction.
Preheat increases SAFT in the first regime and decreases in the
second regime, where the turnover point of both regimes is shifted
to higher equivalence ratios. The equilibrium temperature and
concentrations are affected by elevated pressure most in the near
stoichiometry equivalence ratios, while the pressure influence on
equilibrium temperature and concentrations is negligible in
ultra-rich flames. Pressure increase leads to a decrease of SAFT in
an equivalence ratio of / = 1.0 to / = 2.0 and to an increase for /
> 2.0.
Acknowledgments
The present research work contributes to the Research Field
Energy of Helmholtz Association (HGF).
References
[1] Baukal Jr CE. Oxygen-enhanced combustion. Boca Florida: CRC Press LLC; 1998.
[2] Meeks E, Kee RJ, Dandy DS, Coltrin ME. Computational simulation of diamond
chemical vapor deposition in premixed C2H2/O2/H2 and CH4/O2-strained
flames. Comb Flame 1993;92:144.
[3] Bertagnolli KE, Lucht RP. Temperature profile measurements in stagnation-
flow diamond-forming flames using hydrogen CARS spectroscopy. Proc
Combust Inst 1996;26:1825–33.
[4] Bertagnolli KE, Lucht RP, Bui-Pham MN. Atomic hydrogen concentration profile
measurements in stagnation-flow diamond-forming flames using three-
photon excitation laser-induced fluorescence. J Appl Phys 1998;83:2315–26.
[5] Ruf B, Behrendt F, Deutschmann O, Kleditzsch S, Warnatz J. Modeling of
chemical vapor deposition of diamond films from acetylene-oxygen flame.
Proc Combst Inst 2000;28:1455–61.
[6] Liu F, Guo H, Smallwood GJ, Gülder ÖL. Numerical study of the superadiabatic
flame temperature phenomenon in hydrocarbon premixed flames. Proc
Combst Inst 2002;29:1543–50.
[7] Zamashchikov VV, Namyatov IG, Bunev VA, Babkin VS. On the nature of
superadiabatic temperatures in premixed rich hydrocarbon flames. Combust
Explos Shock Waves 2004;40:32–5.
[8] Liu F, Gülder ÖL. Effects of H2 and H preferential diffusion and unity Lewis
number on superadiabatic flame temperatures in rich premixed methane
flames. Comb Flame 2005;143:264–81.
[9] Liu F, Gülder ÖL. Effects of pressure and preheat on super-adiabatic flame
temperatures in rich premixed methane/air flames. Combust Sci Technol
2008;180:437–52.
[10] Kumagami M, Ogami Y, Tamaki Y, Kobayashi H. Numerical analysis of
extremely-rich CH4/O2/H2O premixed flames at high pressure and high
temperature considering production of higher hydrocarbons. J Therm Sci
Technol 2010;5:109–23.
[11] Takagi T, Xu Z. Numerical analysis of laminar diffusion flames – effects of
preferential diffusion of heat and species. Comb Flame 1994;96:50–9.
[12] Kee RJ, Grcar JF, Smooke MD, Miller JA. PREMIX, Report No. SAND85-8240,
Sandia National Laboratories; 1985.
[13] Kee RJ, Rupley FM, Miller JA. CHEMKIN-II: A Fortan chemical kinetics package
for the analysis of gas phase chemical kinetics, Report No. SAND 89–8009B,
Sandia National Laboratories; 1989.
[14] Smith GP, Golden DM, Frenklach M, Moriarty NW, Eiteneer B, Goldenberg M,
et al. <http://www.me.berkley.edu/gri_mech/>.
[15] Appel J, Bockhorn H, Frenklach M. Kinetic modeling of soot formation with
detailed chemistry and physics: laminar premixed flames of C2 hydrocarbons.
Comb Flame 2000;121:122–36.