- Email: [email protected]
Inverse diffusion flame of CH4–O2 in hot syngas coflow
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Inverse diffusion flame of CH4eO2 in hot syngas coflow Xinyu Li, Zhenghua Dai*, Yueting Xu, Chao Li, Zhijie Zhou, Fuchen Wang** Key Laboratory of Coal Gasification and Energy Chemical Engineering of Ministry of Education, East China University of Science and Technology, Shanghai 200237, China
article info
abstract
Article history:
The structure and combustion mode of inverse diffusion flame of CH4 and O2 in hot syngas
Received 4 August 2015
coflow are numerically studied to gain a fundamental understanding of the flame in non-
Received in revised form
catalytic partial oxidation (NC-POX) reformer. The configuration is modified based on the
20 September 2015
burner system of Cabra et al. [Combust. Flame 2005, 143 (4), 491e506] to make the flame
Accepted 22 September 2015
representative of that in NC-POX reformer. The Eddy Dissipation Concept (EDC) model with
Available online xxx
the detailed GRI 3.0 mechanism is used to model the turbulenceereaction interactions. Results of the study show that the flame is stabilized by autoignition with a wide reaction
Keywords:
zone located far away from the stoichiometric line. Analyses on combustion mode show
Inverse diffusion flame
that the flame is established in Moderate and Intense Low-oxygen Dilution (MILD) mode.
Natural gas reformer
The inverse diffusion flame configuration which ensures a fully dilution of oxygen plays a
MILD combustion
key role in achieving MILD combustion in fuel rich coflow. The Increase of coflow tem-
Jet in hot coflow
perature or decrease of jet velocity within the range of this study can lead to an early autoignition, but doesn't change the combustion mode. Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction Non-catalytic partial oxidation (NC-POX) of natural gas is an important alternative technology used in the production of syngas (CO þ H2) for chemical and energetic use. Compared to other natural gas reforming technologies [1e3], the most significant advantage of NC-POX process is that the syngas H2/CO ratio is about 1.7e1.8, which is close to the desired ratio for Fischer-Tropsch synthesis, methanol synthesis and glycol synthesis [4,5]. The natural gas and pure oxygen jet into natural gas NC-POX reformer with high speed (>80 m/s) from a coaxial nozzle mounted at top of the reformer. And therefore a
jet zone, a recirculation zone and a reforming zone are formed in the reformer [6]. In most reported reformer designs [6,7], the oxygen stream jets from the central channel forming an inverse diffusion flame in jet zone. Slow reforming reactions between CH4 and CO2/H2O occur in recirculation zone and reforming zone. While commercial scale plants have been built using NCPOX technology, basic information of the NC-POX process still needs to be further studied. Experimental data available for reformers are mostly limited to outlet gas compositions. This is because measurement of the process in reformer is hard to conduct due to the high process temperature and pressure. Numerical study has been the main method in
* Corresponding author. Tel.: þ86 21 6425 0784; fax: þ86 21 6425 1312. ** Corresponding author. Tel.: þ86 21 6425 0784; fax: þ86 21 6425 1312. E-mail addresses: [email protected] (Z. Dai), [email protected] (F. Wang). http://dx.doi.org/10.1016/j.ijhydene.2015.09.073 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
understanding the physical processes in such systems [7,8]. Both kinetic models [9e11] and multi-dimension CFD models [4,12e14] have been carried out in the modeling of partial oxidation reformer, but these previous work mostly focus on the syngas yields, the information on the combustion process is limited. The characteristics of the combustion process and flame are important for the partial oxidation of natural gas, since the flame influences the temperature profile, oxidation zone volume and stability of the reformer, and therefore determines the overall natural gas conversion. This has attracted researcher's attention in recent years. Stelzner et al. [15] developed a laminar inverse diffusion flame model burner to study the rich inverse diffusion flame structure. They found that the use of pure O2 as oxidizer leads to extremely high temperatures (~3000 K) and the flame structure is strongly influenced by radiation and diffusion effects. Time scale analysis [16], flamelet modeling [17] and radiation modeling [18] were also performed to study this rich inverse diffusion flame. Li et al. [19] studied the hysteresis of the CH4eO2 inverse diffusion flame in laminar conditions. Many other studies on inverse diffusion flame which were not originated from the reformer applications were also conducted to study the flame behavior and structure [20e22]. The studies on the rich inverse diffusion flame were all conducted in low velocity and room temperature environment. The effects of high jet velocity and hot syngas recirculating flow were not considered. The flame behavior and stabilization mechanism in hot recirculated gas may be totally different from those in normal temperature environment [23e25]. Recently, a series of tests were conducted on a semi-industrial scale NC-POX reformer to study the flame structure using optical measurements [7]. The measured peak temperatures were ~2000 K. The extremely high temperatures (>3500 K) reported in previous numerical studies [4,10,13,14,26] and laminar flame studies [15] were not observed. The authors owed this big deviation to the limitation of the measurements. Although no study focused on the effect of hot syngas recirculating flow on flame in the research field of NC-POX, a good volume of work has been carried out in the fundamental study of turbulent flames with hot recirculating flow in the research field of combustion due to its wide application in many combustion systems. Because of the complex recirculation patterns within such systems, Jet-in-Hot-Coflow (JHC) flames have been developed to simulate recirculation. In such a simplified geometry, a fuel jet issues into a coflowing stream of hot combustion products, with a composition resembling that encountered in an actual recirculating system. Experiments using this configuration have been conducted by Cabra et al. [25,27] and Dally et al. [28] to investigate the autoignitive lifted flames and MILD combustion, respectively. Detailed information of temperature, velocity, and major species in these JHC flames had been measured to provide validation data for numerical study. Following these experimental studies, many numerical simulations had been conducted on JHC flames. Christo and Dally [29] studied the performance of various turbulence, combustion, and kinetic models in predicting the JHC flames, and found that the Eddy Dissipation Concept (EDC) model with a detailed kinetic scheme offers a practical and
reasonably accurate tool for predicting the characteristics of JHC configurations. Results of a similar work [30] also shows that the EDC model with detailed mechanism can well predict the combustion characteristics. Effects of coflow temperature, coflow velocity, oxygen level and fuel type [27,31e33]were numerically studied based on the validation of model using data from JHC experiments. These data were also used for validation in the studying of oxy-fuel combustion of a methane jet in hot O2/CO2 coflow [34]. Considering the lack of studies on the effects of high jet velocity and hot recirculated syngas flow on reformer flames, this study attempts to study the inverse diffusion flame in such conditions. To avoid the complex recirculation patterns in reformer, this study uses an inverse diffusion flame of CH4eO2 in hot syngas coflow to represent the flame established in a NC-POX reformer with an inverse diffusion flame burner. The well-proved k-ε model and the EDC model with detailed chemical kinetics are used to study the flame in hot syngas coflow. The aim of this work is to investigate the characteristics of the inverse diffusion flame in hot syngas coflow and provide insight into the combustion process in natural gas NC-POX reformers.
Model description Configuration of the burner system In order to investigate the inverse diffusion flame in hot syngas coflow, the present study uses a modified configuration of the JHC burner used by Cabra et al. [27]. Fig. 1(a) and (b) show the structure of the original JHC burner and the present configuration, respectively. The original JHC burner had been well described in earlier work [25,27,35], so only the modified configuration will be described in detail here. The single fuel tube in the cabra burner is replaced by two coaxial tubes. The oxygen tube has an inner diameter of Do ¼ 4 mm and a wall thickness of 1 mm. The fuel tube has an inner diameter of Df ¼ 8 mm. The two tubes are located in the center of a hot syngas coflow. The area ratio of the two channels is 0.57 which is similar to that of the industrial burner in natural gas NC-POX reformer. This designed system provides conditions that can simulate the flame in the NC-POX reformer.
Turbulence models According to the studies of Christo and Dally [29] on the performance of turbulence models in JHC flame modeling, the standard k-ε model with dissipation equation constant (Cε1) modified from 1.44 to 1.6 provided the best results. So the improved standard k-ε model with enhanced wall function is implemented to model the turbulent flow. The turbulence kinetic energy k and its dissipation rate ε are obtained from the following transport equations [36]: v v v ðrkÞ þ ðrkui Þ ¼ vt vxi vxj
m vk mþ t þ Gk þ Gb rε YM þ Sk sk vxj (1)
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
v v v ðrεÞ þ ðrεui Þ ¼ vt vxi vxj
m vε ε mþ t þ C1ε ðGk þ G3ε Gb Þ k sε vxj
C2ε r
ε2 þ Sε k
t* ¼ Ct ðn=εÞ1=2 (2)
The turbulent viscosity mt is represented by mt ¼ rCm k2 ε
(3)
Here, r is the density, ui is the velocity component in the ith direction, Gk means the generation of turbulence kinetic energy due to mean velocity gradients, Gb represents the generation of turbulence kinetic energy due to buoyancy. YM is the contribution of the fluctuating dilatation in compressible turbulence to the overall dissipation rate. C1ε, C2ε and C3ε are constants. sk and sε are the turbulent Prandtl numbers for k and ε, respectively. Sk and Sε are source terms.
Combustion model The EDC model has been widely used in the modeling of JHC flames and performs reasonably well. Thus, the EDC model with detailed chemical kinetic mechanism (GRI-Mech 3.0) [37] is presently used for the modeling of reactions. The GRI-Mech 3.0 mechanism containing 53 species and 325 reversible reactions was optimized for methane combustion and performs rather well in JHC flame modeling [29,32,34,38]. The EDC model assumes that reaction occurs in small turbulent structures, called the fine scales. The length fraction of the fine scales x* and the chemical residence time scale of fluid in the fine structures t* are modeled as [36]: 1=2 x ¼ Cx nε k2 *
(4)
3
(5)
Where Cx (¼2.1377) is the volume fraction constant, Ct (¼0.4082) is the time scale constant, n is the kinematic viscosity. Combustion at the fine scales is assumed to occur as a constant pressure reactor, with initial conditions taken as the current species and temperature in the cell. The evolution of species concentrations is then computed by integrating the chemistry within these fine scales. The in situ adaptive tabulation (ISAT) model of Pope [39] is used to reduce the computational cost of time integration. In the EDC model, transport equations of mass fraction for each species in the chemical mechanism are modeled as [36]: v ! ðrYi Þ þ V$ðr! v Yi Þ ¼ V$ Ji þ Ri þ Si vx
(6)
Where Yi is the local mass fraction of species i, v is the overall velocity, Ji is the diffusion flux of species i due to concentration gradients, Si is the source terms and Ri is the net rate of production of species i by chemical reaction, given by Ref. [36]: Ri ¼
2 * r x* * 3 i Yi Yi * t 1 x h
(7)
Where Yi* is the fine-scale species mass fraction after reacting over the time t*.
Mixture fraction model The hot syngas coflow in this configuration mainly consist of H2 and CO, so the coflow should be treated as a second fuel. Such a situation clearly exceeds the scope of the standard
Fig. 1 e Geometry of the origin burner [27] (a) and the model configuration (b). Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
mixture fraction model which was originally derived for twofeed systems with one oxidizer stream and one fuel stream. To describe such a situation with two fuel inlets, the multipleinlet mixture fraction model proposed by Gomet et al. [40] is used in this study. For the configuration in this study, two transport equations for the inlet tracers zj and one for mixture fraction x is needed to model the mixture fraction. The inlet tracer which is defined as the mass fraction issued from the jth inlet is given by Ref. [40]:
zj ¼ mj mtot
(8)
The transport equations for inlet tracers are given by Ref. [40]: v v nt vzj rzj þ rui zj r ¼0 vt vxi Sct vxi
(9)
The transport equation for mixture fraction is given by Ref. [40]: v v nt vx ðrxÞ þ rui x r ¼0 vt vxi Sct vxi
(10)
Here, nt is the turbulent kinematic viscosity, Sct ¼ 0.7 is the turbulent Schmidt number. These transport equations are implemented in FLUENT using User-Defined Scalars.
reformer. Boundary conditions for the cabra case are also shown in Table 1 for comparison.
Results and discussion Model validation Model validation is conducted by predicting the flame characteristics under the same conditions as those for the experiments of Cabra et al. [27]. Figs. 2 and 3 compare the predictions and the measurements of the radial distributions of the mixture fraction (f) and mean temperature (T) obtained at x ¼ 1D, 15D, 30D and 40D. Here, x is the axial distance from the nozzle exit; D is the diameter of the fuel tube. The mixture fraction was obtained from the Bilger formulation [41]. It is shown that the predictions of mixture fraction and mean temperature agree well with the measurements. These good agreements between measurements and simulation results indicate that the present CFD model can correctly capture the features of the JHC combustion processes. Hence, the use of the same model is appropriate for investigating the combustion characteristics of the JHC flames surrounded by a hot flue-gas coflow.
Structure of the flame Radiation model The discrete ordinates (DO) radiation model is used to model the radiation in this problem because this model spans the entire range of optical thickness. The DO radiation model solves the radiative transfer equation (RTE) for a finite number of discrete solid angles across the computational domain [36]. The absorption coefficient of gas mixture is calculated by the weighted-sum-of-gray-gases model (WSGGM). The basic assumption of WSGGM is that the total emissivity can be computed as a function of mixture composition and temperature. The WSGGM is a reasonable compromise between the oversimplified gray gas model and a complete model which takes into account particular absorption bands.
To illustrate the overall flame structure, plots of the computed profiles of temperature (T), heat release (QHR), OH mole fraction (XOH) and CH2O mole fraction (XCH2 O ) for the flame are presented in Figs. 4 and 5, respectively. The stoichiometric mixture fraction lines xst based on the multi-inlet mixture fraction model are also shown in the figures. Because of the inverse diffusion configuration, the stoichiometric mixture fraction line locates in the central low temperature region of the jet flow. It can be seen that no reaction occurs near the stoichiometric line, the temperature rise and heat release occur in the fuel rich area in the coflow side. The range of the high temperature region (~100 mm) is rather wide compared to the diameter of the oxygen nozzle (d ¼ 4 mm). The peak temperature of the flame is 1797 K, which is far lower than the adiabatic stoichiometric temperature of the CH4eO2 mixture (3052 K). The flame structure obtained here is totally different
Simulation details Due to the symmetry of the system, a 2D axisymmetric computational model was constructed. The computational domain extends to 1500 mm in the axial direction and 120 mm in the radial direction, as shown in Fig. 1(b). A structured mesh with about 50,000 cells is used after grid independence study. Boundary conditions at the three inlets are set to velocity inlet, and symmetry condition and pressure outlet are used for the outside and downstream, respectively. The differential diffusion effects are considered in the model. Pressure velocity coupling is solved using SIMPLE algorithm. The secondorder upwind scheme is employed for discretizing the equations. All computations presented here use the Commercial CFD code ANSYS FLUENT 12.1. The boundary conditions for the present jet in hot syngas coflow case are shown in Table 1. These conditions are similar to that in a natural gas NC-POX
Table 1 e Boundary conditions for the cabra case [27] and jet in hot syngas coflow case. Cabra Jet Temperature(K) 320 Velocity (m/s) 100 Species mole fractions 0.33 CH4 0.15 O2 0.0001 H2 CO e H2O 0.0029 OH e Balance N2
Coflow 1355 5.4 0.0003 0.12 0.0001 e 0.15 0.0002 Balance
Jet in hot syngas coflow Oxidant
Fuel
Coflow
300 100
300 100
1600 5.4
0 1 0 0 0 0 0
1 0 0 0 0 0 0
0 0 0.56 0.32 0.12 0 0
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
5
Fig. 2 e Radial profiles of mixture fraction at various axial positions; Circles: measurements [27], Solid lines: predictions.
from that of a rich laminar inverse diffusion flame conducted without a hot coflow [15]. As shown in Ref. 15, the peak temperature of the rich laminar inverse diffusion flame is close to 3000 K and the high temperature region in the flame locates near the burner exit along with the stoichiometric line. The differences in flame structure between the two flames indicate that the inverse diffusion flame in this study is not stabilized by an edge flame as the rich laminar IDF but by autoignition which occurs after the mixing of fuel and oxidizer.
Ref. [35] shown that ‘time history’ of intermediates (such as CH2O and OH) can be regarded as an indicator of flame stabilization mechanism: autoignition or premixed flame propagation. The autoignition process is characterized by the buildup of CH2O before OH, while the premixed flame propagation process is characterized by a simultaneous buildup of CH2O and OH. For the studied flame, profiles of OH and CH2O are presented in Fig. 5. Profiles of OH are similar in shape and location to those of temperature, while the CH2O are mainly located at upstream side of the OH location. The CH2O is
Fig. 3 e Radial profiles of mean temperature at various axial positions; Circles: measurements [27], Solid lines: predictions. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
Fig. 4 e Contour plots of temperature T and Heat release QHR.
produced by fuel oxidation in low temperature region and increase gradually before autoignition. Few heat is released in this process and this process is assisted by the hot coflow. It can be seen that the peak concentration of CH2O is reached just before the initial OH position and then quickly decreases. At the same time, the OH is quickly produced. Large heat is released in this autoignition process, and the flame is established. This behavior indicates that autoignition is the stabilizing mechanism. This result is consistent with the previous studies on flame in hot coflow [23,35,42]. The most reactive mixture fraction lines xMR are also shown in Figs. 4 and 5. The values of xMR are obtained from a series of homogeneous reactor calculations with initial conditions of species and temperature corresponding to frozen mixing between the O2, CH4 and hot syngas coflow. The most
reactive fraction xMR is defined as the mixture fraction with the shortest autoignition time [43] which is defined based on the maximum of OH mass fraction in this study. At central region where the mixture fraction is relatively small, the oxygen mole fraction XO2 is high but the temperature and hydrogen mole fraction is low. As mixture fraction x increases with radial distance, the temperature and hydrogen mole fraction increase but the oxygen concentration decreases. The competition of these two effects results in an optimum composition, giving the fastest autoignition time. The xMR provides an estimate of where autoignition will occur. It can be seen from the figures that the xMR is not equal to the xst, and is located in fuel rich side. Though the composition and temperature change with positions in the flow field, the xMR doesn't change much with a value varying between 0.96 and 0.98. This is consistent with the findings of several direct numerical simulation (DNS) results of autoignition sites as reviewed by Ref. [43]. The origin of the reaction zone, as shown in Figs. 4 and 5, agree well with the xMR line, which is calculated from homogeneous autoignition. The reaction zone is mainly located inside the xMR line. The radial profiles of mixture fraction are shown in Fig. 6. At x ¼ 10d and 20d, the mixture fraction near the axis is below xst, but increases along the radial direction fast. At x > 40d, the x is larger than xst even at r ¼ 0, and increases slowly along the radial direction. Fig. 7 shows the radial profiles of T, XO2 , XOH and XCH2 O in mixture fraction space. The highest temperature and OH mole fraction across mixture fraction space evolves from a very rich mixture fraction (xMR) to smaller mixture fraction with downstream distance. But different from the flames in hot air coflow [38,43], the location of the highest temperature and OH mole fraction can never reach the stoichiometric mixture fraction position. XO2 decreases rapidly after leaving the nozzle exit. This decay is mainly due to the dilution of fuel and entrained coflow gas.
Analysis of the combustion mode As motioned above, the peak temperature of the oxy-flame in hot syngas coflow is only 1797 K. This temperature is
Fig. 5 e Contour plots of XOH and XCH2 O .
Fig. 6 e Radial profiles of mixture fraction.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
7
Fig. 7 e Radial profiles of T, XO2 , XOH and XCH2 O in mixture fraction space.
consistent with the measured peak temperature in a bench scale NC-POX reformer [7], but not consistent with the previous idea that the flame temperature in natural gas reformer is over 3000 K. A further analysis is conducted here to explain this unexpected observation. To illustrate the process more clearly and directly, profiles of XOH, XO2 , XCH2 O and T are shown in Fig. 8. The profile of XOH is used to indicate the flame. Note that the x: r scales of the figure are 3:1, so that the structure can be seen more clearly. The jet zone can be divided into three zones according to the local temperature, O2 mole fraction and OH mole fraction, as
Fig. 8 e Profiles of the XOH (left contour), XCH2 O (right contour), XO2 (left lines) and T (right lines).
shown in Fig. 8. In the near nozzle region (Zone A), though the oxygen mole fraction is high, the flame still cannot be established because of the low mixture temperature (<900 K) and high gas velocity. In zone B which is located downstream and outside of zone A, pre-ignition reactions occur due to the higher temperature (900e1200 K), CH2O produced by preignition reactions is predominant in this temperature region. In further outside or downstream where the temperature is higher than the autoignition temperature and the gas velocity is relatively low, the ignition occurs. But the oxygen mole fraction has decreased to a level below 10% in these areas due to the dilution effect. This region in Fig. 8 is marked as zone C. The low oxygen concentration leads to a low temperature increase in the flame. As a result the peak temperature is only 1797 K. The combustion reactions occur in a place where the local temperature is higher than the autoignition temperature and the O2 mole fraction is lower than 10%. So it can be concluded that MILD combustion occurs in Zone C, according to the studies on MILD combustion [32,44]. To identify the combustion mode of the studied flame in a rigorous way, the definition of MILD combustion given by Cavaliere and de Joannon [44] is used as a criterion. In their definition, a combustion process is named Mild when the inlet temperature of reactant mixture (Tin) is higher than mixture self-ignition temperature (Tig) whereas the maximum allowable temperature increase with respect to inlet temperature during combustion (DT) is lower than mixture self-ignition temperature (Tig). This definition is based on WSR (Well Stirred Reactor) calculations. In this study, each computational cell is regarded as a WSR, so the combustion mode in each position can be decided. The temperature of a cell in frozen mixing conditions is regarded as the inlet temperature Tin, the self-ignition temperature is obtained from WSR calculations using CHEMKIN with mixture corresponding to the frozen
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
Fig. 9 e Combustion mode classifications of the flame at axis and different radial positions.
mixing conditions. This calculated Tig is in fact lower than the actual self-ignition temperature because the long residence time (1s) and zero scalar dissipation rate used in the WSR calculations. With the resulting data of Tig, Tin and DT at axis and different radial positions, a map can be drawn in Fig. 9 following the method proposed by Cavaliere and de Joannon. It can be seen that all the data are located in the No Combustion regime or MILD combustion regime. So it can be concluded that the combustion mode of the studied flame is MILD combustion.
Effect of the diffusion flame configuration Fig. 10 shows the radial profiles of mean temperature at different axial locations for both the normal diffusion flame (NDF) configuration and inverse diffusion flame (IDF) configuration. In the NDF configuration, the fuel stream jets from
the central channel and the oxygen stream jets from the outer channel. The NDF case and the IDF case use the same burner. The flow rates of all streams in NDF case are kept the same as those in IDF case. It can be seen from Fig. 10 that the temperature profiles of the two configurations are significantly different. The temperature of the NDF configuration at x ¼ 10 mm is about 3000 K, this indicates that the pure oxygen reacts with the hot syngas coflow immediately after leaving the nozzle exit. While the temperature of the IDF case doesn't show an increase until x ¼ 200 mm and the temperature increase is rather small compared to the NDF case. This result shows that the IDF configuration is an essential configuration in the establishment of MILD combustion in hot rich coflow conditions. This can be explained. In IDF configuration, the oxygen stream flows through the central channel and the fuel stream flows through the outer channel. Due to this special configuration, the outer cold fuel stream prevents the inner oxygen from directly reacting with the hot syngas coflow, the oxygen has to go through a mixing and preheat process before reacting with the fuel mixture. As a result, the oxygen mole fraction has been decreased to about 10% when the mixture is preheated above its autoignition temperature. The low oxygen concentration leads to a low temperature increase, so the MILD combustion is establishment. While for the NDF configuration where the oxygen flows through the outer channel, the pure oxygen can react with hot syngas coflow immediately after leaving the nozzle exit due to the low dissipation rate in coflow. The IDF configuration ensures a fully dilution of oxygen before autoignition, this is a key point to the establishment of MILD combustion.
Effect of the coflow temperature Fig. 11 shows the axial and radial distributions of mean temperature and OH mole fraction for different coflow temperature Tcoflow. The temperature range considered here is 1300 Ke1700 K, which is the common operating range of a natural gas NC-POX reformer. It can be seen that the temperature and XOH is higher as the Tcoflow increases. The position of the peak temperature and XOH shift to lower mixture fraction as the Tcoflow increases. This means that autoignition occurs early as the Tcoflow increases. This may be caused by the increase of the mixture temperature at lower mixture fraction due to the higher Tcoflow. Fig. 12 shows the effects of Tcoflow on the combustion mode. This map is calculated and drawn in the same method as Fig. 9. Only the calculations for axis is presented here. These results show that the coflow temperature has little effect on the combustion mode among the range of 1300 Ke1700 K, the flame is still in MILD mode in the studied temperature range.
Effect of jet velocity
Fig. 10 e Radial profiles of mean temperature for NDF configuration and IDF configuration at various axial positions.
Jet momentum is a crucial parameter in the establishment of MILD combustion in non-preheat cases. The effect of jet momentum on the flame is also studied. The jet velocity studied here varying from 100 m/s to 40 m/s. This range is considered to be a possible change interval in the operation of a NC-POX natural gas reformer. Fig. 13 shows the profiles of T and XOH in mixture fraction space for different jet velocities. As the jet
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
9
Fig. 11 e Distributions of mean temperature and OH mass fraction for different Tcoflow, left: axial distributions; right: radial distributions at x ¼ 50d.
far lower than the Tig. The flame is still in MILD combustion mode even when the jet velocity decreases to 40 m/s.
Conclusions
Fig. 12 e Combustion mode classification map at axis for different Tcoflow.
velocity decreases, the max temperature and OH location shift to lower mixture fraction. This means that the autoignition occurs early and the reaction zone shifts to upstream and central area of the jet flow in physical space. Due to the early autoignition at lower mixture fraction, the peak temperature increases from 1797 K to 1819 K as the jet velocity decreases from 100 m/s to 40 m/s. But the maximum DT (~500 K) is still
The present study has numerically investigated the inverse diffusion flame of CH4eO2 in hot syngas coflow. The EDC model with detailed GRI 3.0 mechanism is implemented for the simulation. The multi-inlet mixture fraction model is built within the numerical model. The structure and combustion mode of the flame are analyzed, and the effects of diffusion flame configuration, coflow temperature and jet momentum on flame are studied. The results show that the flame characteristics are totally different from that of the rich inverse diffusion flames without hot syngas coflow. The temperature profile in reaction zone is uniform with a maximum temperature of 1797 K. The reaction zone is located in the coflow side and is away from the stoichiometric mixture fraction line. The flame structure indicates that the flame is stabilized by autoignition. Analyses of data in mixture fraction space show that autoignition starts at the very rich side of mixture near the most reactive mixture fraction, then the peak temperature location shift to lower mixture fraction with downstream distance, but cannot reach the stoichiometric mixture fraction. Combustion mode of the studied flame is analyzed using different criteria. Results show that the flame is established in MILD combustion mode.
Fig. 13 e Profiles of T and XOH for different jet velocities. Left: axial profiles; Right: radial profiles at x ¼ 50d. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
10
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
The IDF configuration is an essential configuration in the establishment of MILD combustion for diffusion flames in hot rich coflow. Increase of coflow temperature or decrease of jet momentum lead to an early autoignition and cause the reaction zone shifting to upstream and central area. But the combustion mode remains unchanged. This paper provides a view to better understanding of the combustion process in NC-POX reformer. The results indicate that the effect of hot syngas recirculating flow must be considered in the study of flame in natural gas reformers, and more attention should be paid to the MILD nature of the combustion process in natural gas reformers in the future.
Acknowledgments This work was supported by the Coal-based Key Science and Technology Program Of Shanxi Province, China (MH2014-01) and the Fundamental Research Fund for the Central Universities, China (WB1213004).
references
[1] Zahedinezhad M, Rowshanzamir S, Eikani M. Autothermal reforming of methane to synthesis gas: modeling and simulation. Int J Hydrogen Energy 2009;34(3):1292e300. [2] Ebrahimi H, Mohammadzadeh JSS, Zamaniyan A, Shayegh F. Effect of design parameters on performance of a top fired natural gas reformer. Appl Therm Eng 2008;28(17e18):2203e11. [3] Bharadwaj SS, Schmidt LD. Catalytic partial oxidation of natural gas to syngas. Fuel Process Technol 1995;42(2e3):109e27. [4] Xu Y, Dai Z, Li C, Li X, Zhou Z, Yu G, et al. Numerical simulation of natural gas non-catalytic partial oxidation reformer. Int J Hydrogen Energy 2014;39(17):9149e57. [5] Li CE, Burke N, Gerdes K, Patel J. The undiluted, non-catalytic partial oxidation of methane in a flow tube reactor e an experimental study using indirect induction heating. Fuel 2013;109:409e16. [6] Fuchen W, Weifeng L, Zhenghua D, Xueli C, Haifeng L, Zunhong Y. Preparation of syngas from natural gas by noncatalytic partial oxidation. Petrochem Technol 2006;35(1):47e51. [7] Richter A, Seifert P, Compart F, Tischer P, Meyer B. A largescale benchmark for the CFD modeling of non-catalytic reforming of natural gas based on the Freiberg test plant HP POX. Fuel 2015;152:110e21. gorre B, Gicquel O. Modeling [8] Caudal J, Fiorina B, Labe interactions between chemistry and turbulence for simulations of partial oxidation processes. Fuel Process Technol 2015;134:231e42. [9] Lutz AE, Bradshaw RW, Bromberg L, Rabinovich A. Thermodynamic analysis of hydrogen production by partial oxidation reforming. Int J Hydrogen Energy 2004;29(8):809e16. [10] Zhou X, Chen C, Wang F. Modeling of non-catalytic partial oxidation of natural gas under conditions found in industrial reformers. Chem Eng Process Process Intensif 2010;49(1):59e64.
[11] Zhu J, Zhang D, King KD. Reforming of CH4 by partial oxidation: thermodynamic and kinetic analyses. Fuel 2001;80(7):899e905. [12] Rehm M, Seifert P, Meyer B. Theoretical and numerical investigation on the EDC-model for turbulenceechemistry interaction at gasification conditions. Comput Chem Eng 2009;33(2):402e7. [13] Vegendla SNP, Hasse C. Steady Flamelet Progress-Variable (FPV) modeling and simulation of a high-pressure gasifier. Energy Fuels 2013;27(12):7772e7. [14] Zhou X, Chen C, Wang F. Multi-dimensional modeling of non-catalytic partial oxidation of natural gas in a high pressure reformer. Int J Hydrogen Energy 2010;35(4):1620e9. [15] Stelzner B, Hunger F, Voss S, Keller J, Hasse C, Trimis D. Experimental and numerical study of rich inverse diffusion flame structure. Proc Combust Inst 2013;34(1):1045e55. [16] Pru¨fert U, Hunger F, Hasse C. The analysis of chemical time scales in a partial oxidation flame. Combust Flame 2014;161(2):416e26. [17] Hunger F, Stelzner B, Trimis D, Hasse C. Flamelet-modeling of inverse rich diffusion flames. Flow Turbul Combust 2012;90(4):833e57. [18] Garten B, Hunger F, Messig D, Stelzner B, Trimis D, Hasse C. Detailed radiation modeling of a partial-oxidation flame. Int J Therm Sci 2015;87:68e84. [19] Li X, Dai Z, Xu Y, Li C, Wang F. Mode and hysteresis of laminar methane/oxygen inverse diffusion flame. Acta Phys Sin 2015;64(2). 024704. [20] Elbaz AM, Roberts WL. Flame structure of methane inverse diffusion flame. Exp Therm Fluid Sci 2014;56:23e32. [21] Elbaz AM, Roberts WL. Experimental characterization of methane inverse diffusion flame. Combust Sci Technol 2014;186(9):1249e72. [22] Mahesh S, Mishra DP. Dynamic sensing of blowout in turbulent CNG inverse jet flame. Combust Flame 2015;162(8):3046e52. [23] Mir Najafizadeh SM, Sadeghi MT, Sotudeh-Gharebagh R, Roekaerts DJEM. Chemical structure of autoignition in a turbulent lifted H2/N2 jet flame issuing into a vitiated coflow. Combust Flame 2013;160(12):2928e40. [24] Medwell PR, Dally BB. Experimental observation of lifted flames in a heated and diluted coflow. Energy Fuels 2012;26(9):5519e27. [25] Cabra R, Myhrvold T, Chen JY, Dibble RW, Karpetis AN, Barlow RS. Simultaneous laser raman-rayleigh-lif measurements and numerical modeling results of a lifted turbulent H2/N2 jet flame in a vitiated coflow. Proc Combust Inst 2002;29(2):1881e8. [26] Vegendla SNP, Messig D, Weise S, Hasse C. Flamelet-based time-scale analysis of a high-pressure gasifier. Energy Fuels 2011;25(9):3892e9. [27] Cabra R, Chen J, Dibble R, Karpetis A, Barlow R. Lifted methaneeair jet flames in a vitiated coflow. Combust Flame 2005;143(4):491e506. [28] Dally BB, Karpetis AN, Barlow RS. Structure of turbulent nonpremixed jet flames in a diluted hot coflow. Proc Combust Inst 2002;29(1):1147e54. [29] Christo FC, Dally BB. Modeling turbulent reacting jets issuing into a hot and diluted coflow. Combust Flame 2005;142(1e2):117e29. [30] Myhrvold T, Ertesv AG* IS, Gran IR, Cabra R, Chen JY. A numerical investigation of a lifted H2/N2 turbulent jet flame in a vitiated coflow. Combust Sci Technol 2006;178(6):1001e30. [31] Mardani A, Tabejamaat S. Effect of hydrogen on hydrogenemethane turbulent non-premixed flame under MILD condition. Int J Hydrogen Energy 2010;35(20):11324e31. [32] Wang F, Mi J, Li P. Combustion regimes of a jet diffusion flame in hot co-flow. Energy Fuels 2013;27(6):3488e98.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
[33] Wang F, Mi J, Li P, Zheng C. Diffusion flame of a CH4/H2 jet in hot low-oxygen coflow. Int J Hydrogen Energy 2011;36(15):9267e77. [34] Mei Z, Mi J, Wang F, Zheng C. Dimensions of CH4-jet flame in hot O2/CO2 coflow. Energy Fuels 2012;26(6):3257e66. [35] Gordon RL, Masri AR, Pope SB, Goldin GM. Transport budgets in turbulent lifted flames of methane autoigniting in a vitiated co-flow. Combust Flame 2007;151(3):495e511. [36] AnsysInc. Ansys fluent 12.0 theory guide. 2009. [37] Smith, G. P., Golden, D. M., Frenklach, M., Moriarty, N. W., Eiteneer, B., Goldenberg, M., et al http://www.me.berkeley. edu/gri_mech/. ndez-Pe rez F, van Oijen J, [38] Abtahizadeh E, Sepman A, Herna Mokhov A, de Goey P, et al. Numerical and experimental investigations on the influence of preheating and dilution on transition of laminar coflow diffusion flames to Mild combustion regime. Combust Flame 2013;160(11):2359e74.
11
[39] Pope SB. Computationally efficient implementation of combustion chemistry using in situ adaptive tabulation. Combust Theory Model 1997;1(1):41e63. [40] Gomet L, Robin V, Mura A. A multiple-inlet mixture fraction model for nonpremixed combustion. Combust Flame 2015;162:668e87. [41] Bilger RW, Sta˚rner SH, Kee RJ. On reduced mechanisms for methane air combustion in nonpremixed flames. Combust Flame 1990;80(2):135e49. [42] Jangi M, Zhao X, Haworth DC, Bai X-S. Stabilization and liftoff length of a non-premixed methane/air jet flame discharging into a high-temperature environment: an accelerated transported PDF method. Combust Flame 2015;162(2):408e19. [43] Mastorakos E. Ignition of turbulent non-premixed flames. Prog Energy Combust Sci 2009;35(1):57e97. [44] Cavaliere A, de Joannon M. Mild combustion. Prog Energy Combust Sci 2004;30(4):329e66.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Inverse diffusion flame of CH4eO2 in hot syngas coflow Xinyu Li, Zhenghua Dai*, Yueting Xu, Chao Li, Zhijie Zhou, Fuchen Wang** Key Laboratory of Coal Gasification and Energy Chemical Engineering of Ministry of Education, East China University of Science and Technology, Shanghai 200237, China
article info
abstract
Article history:
The structure and combustion mode of inverse diffusion flame of CH4 and O2 in hot syngas
Received 4 August 2015
coflow are numerically studied to gain a fundamental understanding of the flame in non-
Received in revised form
catalytic partial oxidation (NC-POX) reformer. The configuration is modified based on the
20 September 2015
burner system of Cabra et al. [Combust. Flame 2005, 143 (4), 491e506] to make the flame
Accepted 22 September 2015
representative of that in NC-POX reformer. The Eddy Dissipation Concept (EDC) model with
Available online xxx
the detailed GRI 3.0 mechanism is used to model the turbulenceereaction interactions. Results of the study show that the flame is stabilized by autoignition with a wide reaction
Keywords:
zone located far away from the stoichiometric line. Analyses on combustion mode show
Inverse diffusion flame
that the flame is established in Moderate and Intense Low-oxygen Dilution (MILD) mode.
Natural gas reformer
The inverse diffusion flame configuration which ensures a fully dilution of oxygen plays a
MILD combustion
key role in achieving MILD combustion in fuel rich coflow. The Increase of coflow tem-
Jet in hot coflow
perature or decrease of jet velocity within the range of this study can lead to an early autoignition, but doesn't change the combustion mode. Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction Non-catalytic partial oxidation (NC-POX) of natural gas is an important alternative technology used in the production of syngas (CO þ H2) for chemical and energetic use. Compared to other natural gas reforming technologies [1e3], the most significant advantage of NC-POX process is that the syngas H2/CO ratio is about 1.7e1.8, which is close to the desired ratio for Fischer-Tropsch synthesis, methanol synthesis and glycol synthesis [4,5]. The natural gas and pure oxygen jet into natural gas NC-POX reformer with high speed (>80 m/s) from a coaxial nozzle mounted at top of the reformer. And therefore a
jet zone, a recirculation zone and a reforming zone are formed in the reformer [6]. In most reported reformer designs [6,7], the oxygen stream jets from the central channel forming an inverse diffusion flame in jet zone. Slow reforming reactions between CH4 and CO2/H2O occur in recirculation zone and reforming zone. While commercial scale plants have been built using NCPOX technology, basic information of the NC-POX process still needs to be further studied. Experimental data available for reformers are mostly limited to outlet gas compositions. This is because measurement of the process in reformer is hard to conduct due to the high process temperature and pressure. Numerical study has been the main method in
* Corresponding author. Tel.: þ86 21 6425 0784; fax: þ86 21 6425 1312. ** Corresponding author. Tel.: þ86 21 6425 0784; fax: þ86 21 6425 1312. E-mail addresses: [email protected] (Z. Dai), [email protected] (F. Wang). http://dx.doi.org/10.1016/j.ijhydene.2015.09.073 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
2
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
understanding the physical processes in such systems [7,8]. Both kinetic models [9e11] and multi-dimension CFD models [4,12e14] have been carried out in the modeling of partial oxidation reformer, but these previous work mostly focus on the syngas yields, the information on the combustion process is limited. The characteristics of the combustion process and flame are important for the partial oxidation of natural gas, since the flame influences the temperature profile, oxidation zone volume and stability of the reformer, and therefore determines the overall natural gas conversion. This has attracted researcher's attention in recent years. Stelzner et al. [15] developed a laminar inverse diffusion flame model burner to study the rich inverse diffusion flame structure. They found that the use of pure O2 as oxidizer leads to extremely high temperatures (~3000 K) and the flame structure is strongly influenced by radiation and diffusion effects. Time scale analysis [16], flamelet modeling [17] and radiation modeling [18] were also performed to study this rich inverse diffusion flame. Li et al. [19] studied the hysteresis of the CH4eO2 inverse diffusion flame in laminar conditions. Many other studies on inverse diffusion flame which were not originated from the reformer applications were also conducted to study the flame behavior and structure [20e22]. The studies on the rich inverse diffusion flame were all conducted in low velocity and room temperature environment. The effects of high jet velocity and hot syngas recirculating flow were not considered. The flame behavior and stabilization mechanism in hot recirculated gas may be totally different from those in normal temperature environment [23e25]. Recently, a series of tests were conducted on a semi-industrial scale NC-POX reformer to study the flame structure using optical measurements [7]. The measured peak temperatures were ~2000 K. The extremely high temperatures (>3500 K) reported in previous numerical studies [4,10,13,14,26] and laminar flame studies [15] were not observed. The authors owed this big deviation to the limitation of the measurements. Although no study focused on the effect of hot syngas recirculating flow on flame in the research field of NC-POX, a good volume of work has been carried out in the fundamental study of turbulent flames with hot recirculating flow in the research field of combustion due to its wide application in many combustion systems. Because of the complex recirculation patterns within such systems, Jet-in-Hot-Coflow (JHC) flames have been developed to simulate recirculation. In such a simplified geometry, a fuel jet issues into a coflowing stream of hot combustion products, with a composition resembling that encountered in an actual recirculating system. Experiments using this configuration have been conducted by Cabra et al. [25,27] and Dally et al. [28] to investigate the autoignitive lifted flames and MILD combustion, respectively. Detailed information of temperature, velocity, and major species in these JHC flames had been measured to provide validation data for numerical study. Following these experimental studies, many numerical simulations had been conducted on JHC flames. Christo and Dally [29] studied the performance of various turbulence, combustion, and kinetic models in predicting the JHC flames, and found that the Eddy Dissipation Concept (EDC) model with a detailed kinetic scheme offers a practical and
reasonably accurate tool for predicting the characteristics of JHC configurations. Results of a similar work [30] also shows that the EDC model with detailed mechanism can well predict the combustion characteristics. Effects of coflow temperature, coflow velocity, oxygen level and fuel type [27,31e33]were numerically studied based on the validation of model using data from JHC experiments. These data were also used for validation in the studying of oxy-fuel combustion of a methane jet in hot O2/CO2 coflow [34]. Considering the lack of studies on the effects of high jet velocity and hot recirculated syngas flow on reformer flames, this study attempts to study the inverse diffusion flame in such conditions. To avoid the complex recirculation patterns in reformer, this study uses an inverse diffusion flame of CH4eO2 in hot syngas coflow to represent the flame established in a NC-POX reformer with an inverse diffusion flame burner. The well-proved k-ε model and the EDC model with detailed chemical kinetics are used to study the flame in hot syngas coflow. The aim of this work is to investigate the characteristics of the inverse diffusion flame in hot syngas coflow and provide insight into the combustion process in natural gas NC-POX reformers.
Model description Configuration of the burner system In order to investigate the inverse diffusion flame in hot syngas coflow, the present study uses a modified configuration of the JHC burner used by Cabra et al. [27]. Fig. 1(a) and (b) show the structure of the original JHC burner and the present configuration, respectively. The original JHC burner had been well described in earlier work [25,27,35], so only the modified configuration will be described in detail here. The single fuel tube in the cabra burner is replaced by two coaxial tubes. The oxygen tube has an inner diameter of Do ¼ 4 mm and a wall thickness of 1 mm. The fuel tube has an inner diameter of Df ¼ 8 mm. The two tubes are located in the center of a hot syngas coflow. The area ratio of the two channels is 0.57 which is similar to that of the industrial burner in natural gas NC-POX reformer. This designed system provides conditions that can simulate the flame in the NC-POX reformer.
Turbulence models According to the studies of Christo and Dally [29] on the performance of turbulence models in JHC flame modeling, the standard k-ε model with dissipation equation constant (Cε1) modified from 1.44 to 1.6 provided the best results. So the improved standard k-ε model with enhanced wall function is implemented to model the turbulent flow. The turbulence kinetic energy k and its dissipation rate ε are obtained from the following transport equations [36]: v v v ðrkÞ þ ðrkui Þ ¼ vt vxi vxj
m vk mþ t þ Gk þ Gb rε YM þ Sk sk vxj (1)
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
v v v ðrεÞ þ ðrεui Þ ¼ vt vxi vxj
m vε ε mþ t þ C1ε ðGk þ G3ε Gb Þ k sε vxj
C2ε r
ε2 þ Sε k
t* ¼ Ct ðn=εÞ1=2 (2)
The turbulent viscosity mt is represented by mt ¼ rCm k2 ε
(3)
Here, r is the density, ui is the velocity component in the ith direction, Gk means the generation of turbulence kinetic energy due to mean velocity gradients, Gb represents the generation of turbulence kinetic energy due to buoyancy. YM is the contribution of the fluctuating dilatation in compressible turbulence to the overall dissipation rate. C1ε, C2ε and C3ε are constants. sk and sε are the turbulent Prandtl numbers for k and ε, respectively. Sk and Sε are source terms.
Combustion model The EDC model has been widely used in the modeling of JHC flames and performs reasonably well. Thus, the EDC model with detailed chemical kinetic mechanism (GRI-Mech 3.0) [37] is presently used for the modeling of reactions. The GRI-Mech 3.0 mechanism containing 53 species and 325 reversible reactions was optimized for methane combustion and performs rather well in JHC flame modeling [29,32,34,38]. The EDC model assumes that reaction occurs in small turbulent structures, called the fine scales. The length fraction of the fine scales x* and the chemical residence time scale of fluid in the fine structures t* are modeled as [36]: 1=2 x ¼ Cx nε k2 *
(4)
3
(5)
Where Cx (¼2.1377) is the volume fraction constant, Ct (¼0.4082) is the time scale constant, n is the kinematic viscosity. Combustion at the fine scales is assumed to occur as a constant pressure reactor, with initial conditions taken as the current species and temperature in the cell. The evolution of species concentrations is then computed by integrating the chemistry within these fine scales. The in situ adaptive tabulation (ISAT) model of Pope [39] is used to reduce the computational cost of time integration. In the EDC model, transport equations of mass fraction for each species in the chemical mechanism are modeled as [36]: v ! ðrYi Þ þ V$ðr! v Yi Þ ¼ V$ Ji þ Ri þ Si vx
(6)
Where Yi is the local mass fraction of species i, v is the overall velocity, Ji is the diffusion flux of species i due to concentration gradients, Si is the source terms and Ri is the net rate of production of species i by chemical reaction, given by Ref. [36]: Ri ¼
2 * r x* * 3 i Yi Yi * t 1 x h
(7)
Where Yi* is the fine-scale species mass fraction after reacting over the time t*.
Mixture fraction model The hot syngas coflow in this configuration mainly consist of H2 and CO, so the coflow should be treated as a second fuel. Such a situation clearly exceeds the scope of the standard
Fig. 1 e Geometry of the origin burner [27] (a) and the model configuration (b). Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
4
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
mixture fraction model which was originally derived for twofeed systems with one oxidizer stream and one fuel stream. To describe such a situation with two fuel inlets, the multipleinlet mixture fraction model proposed by Gomet et al. [40] is used in this study. For the configuration in this study, two transport equations for the inlet tracers zj and one for mixture fraction x is needed to model the mixture fraction. The inlet tracer which is defined as the mass fraction issued from the jth inlet is given by Ref. [40]:
zj ¼ mj mtot
(8)
The transport equations for inlet tracers are given by Ref. [40]: v v nt vzj rzj þ rui zj r ¼0 vt vxi Sct vxi
(9)
The transport equation for mixture fraction is given by Ref. [40]: v v nt vx ðrxÞ þ rui x r ¼0 vt vxi Sct vxi
(10)
Here, nt is the turbulent kinematic viscosity, Sct ¼ 0.7 is the turbulent Schmidt number. These transport equations are implemented in FLUENT using User-Defined Scalars.
reformer. Boundary conditions for the cabra case are also shown in Table 1 for comparison.
Results and discussion Model validation Model validation is conducted by predicting the flame characteristics under the same conditions as those for the experiments of Cabra et al. [27]. Figs. 2 and 3 compare the predictions and the measurements of the radial distributions of the mixture fraction (f) and mean temperature (T) obtained at x ¼ 1D, 15D, 30D and 40D. Here, x is the axial distance from the nozzle exit; D is the diameter of the fuel tube. The mixture fraction was obtained from the Bilger formulation [41]. It is shown that the predictions of mixture fraction and mean temperature agree well with the measurements. These good agreements between measurements and simulation results indicate that the present CFD model can correctly capture the features of the JHC combustion processes. Hence, the use of the same model is appropriate for investigating the combustion characteristics of the JHC flames surrounded by a hot flue-gas coflow.
Structure of the flame Radiation model The discrete ordinates (DO) radiation model is used to model the radiation in this problem because this model spans the entire range of optical thickness. The DO radiation model solves the radiative transfer equation (RTE) for a finite number of discrete solid angles across the computational domain [36]. The absorption coefficient of gas mixture is calculated by the weighted-sum-of-gray-gases model (WSGGM). The basic assumption of WSGGM is that the total emissivity can be computed as a function of mixture composition and temperature. The WSGGM is a reasonable compromise between the oversimplified gray gas model and a complete model which takes into account particular absorption bands.
To illustrate the overall flame structure, plots of the computed profiles of temperature (T), heat release (QHR), OH mole fraction (XOH) and CH2O mole fraction (XCH2 O ) for the flame are presented in Figs. 4 and 5, respectively. The stoichiometric mixture fraction lines xst based on the multi-inlet mixture fraction model are also shown in the figures. Because of the inverse diffusion configuration, the stoichiometric mixture fraction line locates in the central low temperature region of the jet flow. It can be seen that no reaction occurs near the stoichiometric line, the temperature rise and heat release occur in the fuel rich area in the coflow side. The range of the high temperature region (~100 mm) is rather wide compared to the diameter of the oxygen nozzle (d ¼ 4 mm). The peak temperature of the flame is 1797 K, which is far lower than the adiabatic stoichiometric temperature of the CH4eO2 mixture (3052 K). The flame structure obtained here is totally different
Simulation details Due to the symmetry of the system, a 2D axisymmetric computational model was constructed. The computational domain extends to 1500 mm in the axial direction and 120 mm in the radial direction, as shown in Fig. 1(b). A structured mesh with about 50,000 cells is used after grid independence study. Boundary conditions at the three inlets are set to velocity inlet, and symmetry condition and pressure outlet are used for the outside and downstream, respectively. The differential diffusion effects are considered in the model. Pressure velocity coupling is solved using SIMPLE algorithm. The secondorder upwind scheme is employed for discretizing the equations. All computations presented here use the Commercial CFD code ANSYS FLUENT 12.1. The boundary conditions for the present jet in hot syngas coflow case are shown in Table 1. These conditions are similar to that in a natural gas NC-POX
Table 1 e Boundary conditions for the cabra case [27] and jet in hot syngas coflow case. Cabra Jet Temperature(K) 320 Velocity (m/s) 100 Species mole fractions 0.33 CH4 0.15 O2 0.0001 H2 CO e H2O 0.0029 OH e Balance N2
Coflow 1355 5.4 0.0003 0.12 0.0001 e 0.15 0.0002 Balance
Jet in hot syngas coflow Oxidant
Fuel
Coflow
300 100
300 100
1600 5.4
0 1 0 0 0 0 0
1 0 0 0 0 0 0
0 0 0.56 0.32 0.12 0 0
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
5
Fig. 2 e Radial profiles of mixture fraction at various axial positions; Circles: measurements [27], Solid lines: predictions.
from that of a rich laminar inverse diffusion flame conducted without a hot coflow [15]. As shown in Ref. 15, the peak temperature of the rich laminar inverse diffusion flame is close to 3000 K and the high temperature region in the flame locates near the burner exit along with the stoichiometric line. The differences in flame structure between the two flames indicate that the inverse diffusion flame in this study is not stabilized by an edge flame as the rich laminar IDF but by autoignition which occurs after the mixing of fuel and oxidizer.
Ref. [35] shown that ‘time history’ of intermediates (such as CH2O and OH) can be regarded as an indicator of flame stabilization mechanism: autoignition or premixed flame propagation. The autoignition process is characterized by the buildup of CH2O before OH, while the premixed flame propagation process is characterized by a simultaneous buildup of CH2O and OH. For the studied flame, profiles of OH and CH2O are presented in Fig. 5. Profiles of OH are similar in shape and location to those of temperature, while the CH2O are mainly located at upstream side of the OH location. The CH2O is
Fig. 3 e Radial profiles of mean temperature at various axial positions; Circles: measurements [27], Solid lines: predictions. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
6
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
Fig. 4 e Contour plots of temperature T and Heat release QHR.
produced by fuel oxidation in low temperature region and increase gradually before autoignition. Few heat is released in this process and this process is assisted by the hot coflow. It can be seen that the peak concentration of CH2O is reached just before the initial OH position and then quickly decreases. At the same time, the OH is quickly produced. Large heat is released in this autoignition process, and the flame is established. This behavior indicates that autoignition is the stabilizing mechanism. This result is consistent with the previous studies on flame in hot coflow [23,35,42]. The most reactive mixture fraction lines xMR are also shown in Figs. 4 and 5. The values of xMR are obtained from a series of homogeneous reactor calculations with initial conditions of species and temperature corresponding to frozen mixing between the O2, CH4 and hot syngas coflow. The most
reactive fraction xMR is defined as the mixture fraction with the shortest autoignition time [43] which is defined based on the maximum of OH mass fraction in this study. At central region where the mixture fraction is relatively small, the oxygen mole fraction XO2 is high but the temperature and hydrogen mole fraction is low. As mixture fraction x increases with radial distance, the temperature and hydrogen mole fraction increase but the oxygen concentration decreases. The competition of these two effects results in an optimum composition, giving the fastest autoignition time. The xMR provides an estimate of where autoignition will occur. It can be seen from the figures that the xMR is not equal to the xst, and is located in fuel rich side. Though the composition and temperature change with positions in the flow field, the xMR doesn't change much with a value varying between 0.96 and 0.98. This is consistent with the findings of several direct numerical simulation (DNS) results of autoignition sites as reviewed by Ref. [43]. The origin of the reaction zone, as shown in Figs. 4 and 5, agree well with the xMR line, which is calculated from homogeneous autoignition. The reaction zone is mainly located inside the xMR line. The radial profiles of mixture fraction are shown in Fig. 6. At x ¼ 10d and 20d, the mixture fraction near the axis is below xst, but increases along the radial direction fast. At x > 40d, the x is larger than xst even at r ¼ 0, and increases slowly along the radial direction. Fig. 7 shows the radial profiles of T, XO2 , XOH and XCH2 O in mixture fraction space. The highest temperature and OH mole fraction across mixture fraction space evolves from a very rich mixture fraction (xMR) to smaller mixture fraction with downstream distance. But different from the flames in hot air coflow [38,43], the location of the highest temperature and OH mole fraction can never reach the stoichiometric mixture fraction position. XO2 decreases rapidly after leaving the nozzle exit. This decay is mainly due to the dilution of fuel and entrained coflow gas.
Analysis of the combustion mode As motioned above, the peak temperature of the oxy-flame in hot syngas coflow is only 1797 K. This temperature is
Fig. 5 e Contour plots of XOH and XCH2 O .
Fig. 6 e Radial profiles of mixture fraction.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
7
Fig. 7 e Radial profiles of T, XO2 , XOH and XCH2 O in mixture fraction space.
consistent with the measured peak temperature in a bench scale NC-POX reformer [7], but not consistent with the previous idea that the flame temperature in natural gas reformer is over 3000 K. A further analysis is conducted here to explain this unexpected observation. To illustrate the process more clearly and directly, profiles of XOH, XO2 , XCH2 O and T are shown in Fig. 8. The profile of XOH is used to indicate the flame. Note that the x: r scales of the figure are 3:1, so that the structure can be seen more clearly. The jet zone can be divided into three zones according to the local temperature, O2 mole fraction and OH mole fraction, as
Fig. 8 e Profiles of the XOH (left contour), XCH2 O (right contour), XO2 (left lines) and T (right lines).
shown in Fig. 8. In the near nozzle region (Zone A), though the oxygen mole fraction is high, the flame still cannot be established because of the low mixture temperature (<900 K) and high gas velocity. In zone B which is located downstream and outside of zone A, pre-ignition reactions occur due to the higher temperature (900e1200 K), CH2O produced by preignition reactions is predominant in this temperature region. In further outside or downstream where the temperature is higher than the autoignition temperature and the gas velocity is relatively low, the ignition occurs. But the oxygen mole fraction has decreased to a level below 10% in these areas due to the dilution effect. This region in Fig. 8 is marked as zone C. The low oxygen concentration leads to a low temperature increase in the flame. As a result the peak temperature is only 1797 K. The combustion reactions occur in a place where the local temperature is higher than the autoignition temperature and the O2 mole fraction is lower than 10%. So it can be concluded that MILD combustion occurs in Zone C, according to the studies on MILD combustion [32,44]. To identify the combustion mode of the studied flame in a rigorous way, the definition of MILD combustion given by Cavaliere and de Joannon [44] is used as a criterion. In their definition, a combustion process is named Mild when the inlet temperature of reactant mixture (Tin) is higher than mixture self-ignition temperature (Tig) whereas the maximum allowable temperature increase with respect to inlet temperature during combustion (DT) is lower than mixture self-ignition temperature (Tig). This definition is based on WSR (Well Stirred Reactor) calculations. In this study, each computational cell is regarded as a WSR, so the combustion mode in each position can be decided. The temperature of a cell in frozen mixing conditions is regarded as the inlet temperature Tin, the self-ignition temperature is obtained from WSR calculations using CHEMKIN with mixture corresponding to the frozen
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
8
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
Fig. 9 e Combustion mode classifications of the flame at axis and different radial positions.
mixing conditions. This calculated Tig is in fact lower than the actual self-ignition temperature because the long residence time (1s) and zero scalar dissipation rate used in the WSR calculations. With the resulting data of Tig, Tin and DT at axis and different radial positions, a map can be drawn in Fig. 9 following the method proposed by Cavaliere and de Joannon. It can be seen that all the data are located in the No Combustion regime or MILD combustion regime. So it can be concluded that the combustion mode of the studied flame is MILD combustion.
Effect of the diffusion flame configuration Fig. 10 shows the radial profiles of mean temperature at different axial locations for both the normal diffusion flame (NDF) configuration and inverse diffusion flame (IDF) configuration. In the NDF configuration, the fuel stream jets from
the central channel and the oxygen stream jets from the outer channel. The NDF case and the IDF case use the same burner. The flow rates of all streams in NDF case are kept the same as those in IDF case. It can be seen from Fig. 10 that the temperature profiles of the two configurations are significantly different. The temperature of the NDF configuration at x ¼ 10 mm is about 3000 K, this indicates that the pure oxygen reacts with the hot syngas coflow immediately after leaving the nozzle exit. While the temperature of the IDF case doesn't show an increase until x ¼ 200 mm and the temperature increase is rather small compared to the NDF case. This result shows that the IDF configuration is an essential configuration in the establishment of MILD combustion in hot rich coflow conditions. This can be explained. In IDF configuration, the oxygen stream flows through the central channel and the fuel stream flows through the outer channel. Due to this special configuration, the outer cold fuel stream prevents the inner oxygen from directly reacting with the hot syngas coflow, the oxygen has to go through a mixing and preheat process before reacting with the fuel mixture. As a result, the oxygen mole fraction has been decreased to about 10% when the mixture is preheated above its autoignition temperature. The low oxygen concentration leads to a low temperature increase, so the MILD combustion is establishment. While for the NDF configuration where the oxygen flows through the outer channel, the pure oxygen can react with hot syngas coflow immediately after leaving the nozzle exit due to the low dissipation rate in coflow. The IDF configuration ensures a fully dilution of oxygen before autoignition, this is a key point to the establishment of MILD combustion.
Effect of the coflow temperature Fig. 11 shows the axial and radial distributions of mean temperature and OH mole fraction for different coflow temperature Tcoflow. The temperature range considered here is 1300 Ke1700 K, which is the common operating range of a natural gas NC-POX reformer. It can be seen that the temperature and XOH is higher as the Tcoflow increases. The position of the peak temperature and XOH shift to lower mixture fraction as the Tcoflow increases. This means that autoignition occurs early as the Tcoflow increases. This may be caused by the increase of the mixture temperature at lower mixture fraction due to the higher Tcoflow. Fig. 12 shows the effects of Tcoflow on the combustion mode. This map is calculated and drawn in the same method as Fig. 9. Only the calculations for axis is presented here. These results show that the coflow temperature has little effect on the combustion mode among the range of 1300 Ke1700 K, the flame is still in MILD mode in the studied temperature range.
Effect of jet velocity
Fig. 10 e Radial profiles of mean temperature for NDF configuration and IDF configuration at various axial positions.
Jet momentum is a crucial parameter in the establishment of MILD combustion in non-preheat cases. The effect of jet momentum on the flame is also studied. The jet velocity studied here varying from 100 m/s to 40 m/s. This range is considered to be a possible change interval in the operation of a NC-POX natural gas reformer. Fig. 13 shows the profiles of T and XOH in mixture fraction space for different jet velocities. As the jet
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
9
Fig. 11 e Distributions of mean temperature and OH mass fraction for different Tcoflow, left: axial distributions; right: radial distributions at x ¼ 50d.
far lower than the Tig. The flame is still in MILD combustion mode even when the jet velocity decreases to 40 m/s.
Conclusions
Fig. 12 e Combustion mode classification map at axis for different Tcoflow.
velocity decreases, the max temperature and OH location shift to lower mixture fraction. This means that the autoignition occurs early and the reaction zone shifts to upstream and central area of the jet flow in physical space. Due to the early autoignition at lower mixture fraction, the peak temperature increases from 1797 K to 1819 K as the jet velocity decreases from 100 m/s to 40 m/s. But the maximum DT (~500 K) is still
The present study has numerically investigated the inverse diffusion flame of CH4eO2 in hot syngas coflow. The EDC model with detailed GRI 3.0 mechanism is implemented for the simulation. The multi-inlet mixture fraction model is built within the numerical model. The structure and combustion mode of the flame are analyzed, and the effects of diffusion flame configuration, coflow temperature and jet momentum on flame are studied. The results show that the flame characteristics are totally different from that of the rich inverse diffusion flames without hot syngas coflow. The temperature profile in reaction zone is uniform with a maximum temperature of 1797 K. The reaction zone is located in the coflow side and is away from the stoichiometric mixture fraction line. The flame structure indicates that the flame is stabilized by autoignition. Analyses of data in mixture fraction space show that autoignition starts at the very rich side of mixture near the most reactive mixture fraction, then the peak temperature location shift to lower mixture fraction with downstream distance, but cannot reach the stoichiometric mixture fraction. Combustion mode of the studied flame is analyzed using different criteria. Results show that the flame is established in MILD combustion mode.
Fig. 13 e Profiles of T and XOH for different jet velocities. Left: axial profiles; Right: radial profiles at x ¼ 50d. Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
10
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
The IDF configuration is an essential configuration in the establishment of MILD combustion for diffusion flames in hot rich coflow. Increase of coflow temperature or decrease of jet momentum lead to an early autoignition and cause the reaction zone shifting to upstream and central area. But the combustion mode remains unchanged. This paper provides a view to better understanding of the combustion process in NC-POX reformer. The results indicate that the effect of hot syngas recirculating flow must be considered in the study of flame in natural gas reformers, and more attention should be paid to the MILD nature of the combustion process in natural gas reformers in the future.
Acknowledgments This work was supported by the Coal-based Key Science and Technology Program Of Shanxi Province, China (MH2014-01) and the Fundamental Research Fund for the Central Universities, China (WB1213004).
references
[1] Zahedinezhad M, Rowshanzamir S, Eikani M. Autothermal reforming of methane to synthesis gas: modeling and simulation. Int J Hydrogen Energy 2009;34(3):1292e300. [2] Ebrahimi H, Mohammadzadeh JSS, Zamaniyan A, Shayegh F. Effect of design parameters on performance of a top fired natural gas reformer. Appl Therm Eng 2008;28(17e18):2203e11. [3] Bharadwaj SS, Schmidt LD. Catalytic partial oxidation of natural gas to syngas. Fuel Process Technol 1995;42(2e3):109e27. [4] Xu Y, Dai Z, Li C, Li X, Zhou Z, Yu G, et al. Numerical simulation of natural gas non-catalytic partial oxidation reformer. Int J Hydrogen Energy 2014;39(17):9149e57. [5] Li CE, Burke N, Gerdes K, Patel J. The undiluted, non-catalytic partial oxidation of methane in a flow tube reactor e an experimental study using indirect induction heating. Fuel 2013;109:409e16. [6] Fuchen W, Weifeng L, Zhenghua D, Xueli C, Haifeng L, Zunhong Y. Preparation of syngas from natural gas by noncatalytic partial oxidation. Petrochem Technol 2006;35(1):47e51. [7] Richter A, Seifert P, Compart F, Tischer P, Meyer B. A largescale benchmark for the CFD modeling of non-catalytic reforming of natural gas based on the Freiberg test plant HP POX. Fuel 2015;152:110e21. gorre B, Gicquel O. Modeling [8] Caudal J, Fiorina B, Labe interactions between chemistry and turbulence for simulations of partial oxidation processes. Fuel Process Technol 2015;134:231e42. [9] Lutz AE, Bradshaw RW, Bromberg L, Rabinovich A. Thermodynamic analysis of hydrogen production by partial oxidation reforming. Int J Hydrogen Energy 2004;29(8):809e16. [10] Zhou X, Chen C, Wang F. Modeling of non-catalytic partial oxidation of natural gas under conditions found in industrial reformers. Chem Eng Process Process Intensif 2010;49(1):59e64.
[11] Zhu J, Zhang D, King KD. Reforming of CH4 by partial oxidation: thermodynamic and kinetic analyses. Fuel 2001;80(7):899e905. [12] Rehm M, Seifert P, Meyer B. Theoretical and numerical investigation on the EDC-model for turbulenceechemistry interaction at gasification conditions. Comput Chem Eng 2009;33(2):402e7. [13] Vegendla SNP, Hasse C. Steady Flamelet Progress-Variable (FPV) modeling and simulation of a high-pressure gasifier. Energy Fuels 2013;27(12):7772e7. [14] Zhou X, Chen C, Wang F. Multi-dimensional modeling of non-catalytic partial oxidation of natural gas in a high pressure reformer. Int J Hydrogen Energy 2010;35(4):1620e9. [15] Stelzner B, Hunger F, Voss S, Keller J, Hasse C, Trimis D. Experimental and numerical study of rich inverse diffusion flame structure. Proc Combust Inst 2013;34(1):1045e55. [16] Pru¨fert U, Hunger F, Hasse C. The analysis of chemical time scales in a partial oxidation flame. Combust Flame 2014;161(2):416e26. [17] Hunger F, Stelzner B, Trimis D, Hasse C. Flamelet-modeling of inverse rich diffusion flames. Flow Turbul Combust 2012;90(4):833e57. [18] Garten B, Hunger F, Messig D, Stelzner B, Trimis D, Hasse C. Detailed radiation modeling of a partial-oxidation flame. Int J Therm Sci 2015;87:68e84. [19] Li X, Dai Z, Xu Y, Li C, Wang F. Mode and hysteresis of laminar methane/oxygen inverse diffusion flame. Acta Phys Sin 2015;64(2). 024704. [20] Elbaz AM, Roberts WL. Flame structure of methane inverse diffusion flame. Exp Therm Fluid Sci 2014;56:23e32. [21] Elbaz AM, Roberts WL. Experimental characterization of methane inverse diffusion flame. Combust Sci Technol 2014;186(9):1249e72. [22] Mahesh S, Mishra DP. Dynamic sensing of blowout in turbulent CNG inverse jet flame. Combust Flame 2015;162(8):3046e52. [23] Mir Najafizadeh SM, Sadeghi MT, Sotudeh-Gharebagh R, Roekaerts DJEM. Chemical structure of autoignition in a turbulent lifted H2/N2 jet flame issuing into a vitiated coflow. Combust Flame 2013;160(12):2928e40. [24] Medwell PR, Dally BB. Experimental observation of lifted flames in a heated and diluted coflow. Energy Fuels 2012;26(9):5519e27. [25] Cabra R, Myhrvold T, Chen JY, Dibble RW, Karpetis AN, Barlow RS. Simultaneous laser raman-rayleigh-lif measurements and numerical modeling results of a lifted turbulent H2/N2 jet flame in a vitiated coflow. Proc Combust Inst 2002;29(2):1881e8. [26] Vegendla SNP, Messig D, Weise S, Hasse C. Flamelet-based time-scale analysis of a high-pressure gasifier. Energy Fuels 2011;25(9):3892e9. [27] Cabra R, Chen J, Dibble R, Karpetis A, Barlow R. Lifted methaneeair jet flames in a vitiated coflow. Combust Flame 2005;143(4):491e506. [28] Dally BB, Karpetis AN, Barlow RS. Structure of turbulent nonpremixed jet flames in a diluted hot coflow. Proc Combust Inst 2002;29(1):1147e54. [29] Christo FC, Dally BB. Modeling turbulent reacting jets issuing into a hot and diluted coflow. Combust Flame 2005;142(1e2):117e29. [30] Myhrvold T, Ertesv AG* IS, Gran IR, Cabra R, Chen JY. A numerical investigation of a lifted H2/N2 turbulent jet flame in a vitiated coflow. Combust Sci Technol 2006;178(6):1001e30. [31] Mardani A, Tabejamaat S. Effect of hydrogen on hydrogenemethane turbulent non-premixed flame under MILD condition. Int J Hydrogen Energy 2010;35(20):11324e31. [32] Wang F, Mi J, Li P. Combustion regimes of a jet diffusion flame in hot co-flow. Energy Fuels 2013;27(6):3488e98.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e1 1
[33] Wang F, Mi J, Li P, Zheng C. Diffusion flame of a CH4/H2 jet in hot low-oxygen coflow. Int J Hydrogen Energy 2011;36(15):9267e77. [34] Mei Z, Mi J, Wang F, Zheng C. Dimensions of CH4-jet flame in hot O2/CO2 coflow. Energy Fuels 2012;26(6):3257e66. [35] Gordon RL, Masri AR, Pope SB, Goldin GM. Transport budgets in turbulent lifted flames of methane autoigniting in a vitiated co-flow. Combust Flame 2007;151(3):495e511. [36] AnsysInc. Ansys fluent 12.0 theory guide. 2009. [37] Smith, G. P., Golden, D. M., Frenklach, M., Moriarty, N. W., Eiteneer, B., Goldenberg, M., et al http://www.me.berkeley. edu/gri_mech/. ndez-Pe rez F, van Oijen J, [38] Abtahizadeh E, Sepman A, Herna Mokhov A, de Goey P, et al. Numerical and experimental investigations on the influence of preheating and dilution on transition of laminar coflow diffusion flames to Mild combustion regime. Combust Flame 2013;160(11):2359e74.
11
[39] Pope SB. Computationally efficient implementation of combustion chemistry using in situ adaptive tabulation. Combust Theory Model 1997;1(1):41e63. [40] Gomet L, Robin V, Mura A. A multiple-inlet mixture fraction model for nonpremixed combustion. Combust Flame 2015;162:668e87. [41] Bilger RW, Sta˚rner SH, Kee RJ. On reduced mechanisms for methane air combustion in nonpremixed flames. Combust Flame 1990;80(2):135e49. [42] Jangi M, Zhao X, Haworth DC, Bai X-S. Stabilization and liftoff length of a non-premixed methane/air jet flame discharging into a high-temperature environment: an accelerated transported PDF method. Combust Flame 2015;162(2):408e19. [43] Mastorakos E. Ignition of turbulent non-premixed flames. Prog Energy Combust Sci 2009;35(1):57e97. [44] Cavaliere A, de Joannon M. Mild combustion. Prog Energy Combust Sci 2004;30(4):329e66.
Please cite this article in press as: Li X, et al., Inverse diffusion flame of CH4eO2 in hot syngas coflow, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.09.073