Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

Kinetic study of aviation fuels oxidation in a JSR: Jet-A1 and Bio-Kerosene

S. Gal, 1, P. Dagaut1
Centre National de la Recherche Scientifique 1C, Avenue de la recherche scientifique 45071 Orlans cedex 2 France Abstract Bio-diesel has the potential to be used as a kerosene extender by blending it with conventional kerosene up to approximately 10% - 20% by volume to produce bio-kerosene. The oxidation of kerosene (Jet-A1) and bio-kerosene (kerosene-rapeseed oil methyl esters 80/20 in mole) was studied experimentally in a jet-stirred reactor at 10 atm and constant residence time (0.5s), over the high temperature range 740-1200 K, and for variable equivalence ratios in the range 0.5-1.5. Concentration profiles of reactants, stable intermediates, and final products were obtained by probe sampling followed by on-line and off-line GC analyses. The oxidation of these fuels in these conditions was modeled using a detailed kinetic reaction mechanism consisting of 2027 reversible reactions and 263 species. The surrogate bio-kerosene model-fuel used consisted of a mixture of n-hexadecane, n-propylcyclohexane, npropylbenzene, and n-decane. The methyl ester fraction was simply represented by n-hexadecane. The proposed kinetic reaction mechanism used in the modeling yielded a good representation of the kinetics of oxidation of kerosene and bio-kerosene under JSR conditions. The data and the model showed the bio-kerosene (Jet A-1/RME mixture) has a slightly higher reactivity than Jet A-1 whereas no major modification of the products distribution was observed besides the formation of small methyl esters from RMEs oxidation. Introduction The gap between the growth rate of oil production and demand is increasing as well as environmental concerns such as global warming and air pollution. Therefore, sustainable and environment-friendly fuels are needed for the future. Biofuels derived from vegetable oils or animal fat may be considered sustainable if sufficient quantities of plants can be grown. Furthermore, this can be viewed as a step toward a carbon neutral fuel economy. Biofuels, could be mixed in small quantities (5-20%) with current kerosene [1-3] as it is already done with bio-Diesel [4]. Biodiesel is a mixture of mono-alkyl esters of long carbon-chain fatty acids obtained from renewable lipid feedstock (vegetable, animal, waste). The reported reduced emissions of carbon oxides and PAH make bio-diesel useful for preserving our environment [4-8]. Alkyl esters from vegetable oils are obtained by transesterification with mostly methanol, but also ethanol. The alkyl esters made from different vegetable oils have already been successfully tested in conventional Diesel engines as well as in directinjection engines [8-12]. Rapeseed oil methyl ester (RME) derives from one of the main crop growing Europe. RME is a complex mixture of C14-C22 esters with highly saturated carbon chain (ca. 94% wt of mostly mono-unsaturated esters.). A previous kinetic study of rapeseed oil methyl ester (RME) combustion showed a strong similitude between the oxidation of nhexadecane and that of RME [13] allowing the use of nhexadecane as chemical surrogate model-fuel for modeling RME kinetic of oxidation. Kerosene (Jet A, Jet A-1, JP-8, TR0) is a complex mixture of alkanes (50-65% vol.), mono- and polyaromatics (10-20% vol.) and cycloalkanes or naphtenes (mono- and polycyclic, 20-30% vol.). The average Corresponding author: [email protected] Proceedings of the European Combustion Meeting 2007
1

chemical formula for kerosene (Jet A, Jet A-1, TR0, JP8) differs from one source to another [14]. We adopted the formula C11H22. Due to the complex composition of this fuel, it is necessary to use a surrogate model-fuel for simulating its oxidation. Under high-pressure JSR conditions, the detailed kinetic modeling of kerosene oxidation was initially performed using n-decane as a model-fuel [15], since n-decane and kerosene showed very similar oxidation rates under JSR [15-17] and premixed flame conditions [18]. It was previously shown [13] that n-decane is a reasonable model-fuel for modeling kerosene oxidation if the formation of aromatics is not a major issue since the oxidation of ndecane yields much less aromatics that kerosene. Therefore, more complex model-fuels are necessary to model the formation of aromatics from the oxidation of kerosene [19-23]. Surrogate model-fuels consisting of n-decane and mixtures of n-decane with simple aromatic hydrocarbons and cycloalkanes were recently tested, showing good kerosene oxidation modeling can be achieved using the 3-component model-fuel surrogate mixture of n-decane, n-propylbenzene, and npropyl-cyclohexane [23]. In this paper, we present new experimental results obtained in a JSR for the oxidation of Jet A-1 and RMEJet A-1 mixtures at 10 atm, over a wide range of equivalence ratio (0.5 to 1.5), for temperatures in the range 740-1200 K. The oxidation of kerosene oxidation under JSR and premixed flame conditions is modeled as well as that of the RME-kerosene blends. Experimental The experimental set-up used here is that used earlier [15-17,23]. The JSR consisted of a small sphere of 4 cm diameter (39 cm3) made of fused-silica (to minimize wall catalytic reactions), equipped with 4

Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

nozzles of 1 mm I.D. for the admission of the gases which are achieving the stirring. A nitrogen flow of 100 L/h was used to dilute the fuel. As before [15-17,23], all the gases were preheated before injection to minimize temperature gradients inside the reactor. A regulated heating wire of ca. 1.5 kW maintained the temperature of the reactor at the desired working temperature. The reactants were diluted by nitrogen (<50ppm of O2; <1000ppm of Ar; <5ppm of H2), and mixed at the entrance of the injectors. High purity oxygen (99.995% pure) was used in these experiments. Kerosene Jet A-1 and RME were sonically degassed before use. A Shimadzu LC10 AD VP pump with online degasser (Shimadzu DGU-20 A3) was used to deliver the fuel to an atomizer-vaporizer assembly maintained at 230C. Good thermal homogeneity along the vertical axis of the reactor was observed for each experiment by thermocouple (0.1mm Pt-Pt/Rh 10% located inside a thin-wall silica tube) measurements (gradients of ca.1 K/cm). The reacting mixtures were probe sampled by means of a fused-silica low pressure sonic probe. The samples (ca. 4-6 kPa) were taken at steady temperature and residence time. They were analyzed on-line by GCFID and off-line after collection and storage in 1 L Pyrex bulbs. High vapor-pressure species and permanent gases were analyzed off-line whereas low vapor-pressure compounds were analyzed on-line. The experiments were performed at steady state, at a constant mean residence time, the reactants continually flowing in the reactor. The temperature of the gases inside the JSR was varied stepwise. A high degree of dilution was used, reducing temperature gradients in the JSR and heat release (no flame occurred in the JSR). Gas chromatographs (GC), equipped with capillary columns (Poraplot-U, Molecular Sieve-5A, DB-5ms, DB-624, Plot Al2O3/KCl, Carboplot-P7), TCD, and FID were used for measuring stable species. Compounds identifications were made through GC/MS analyses of the samples. A quadrupole mass spectrometer operating in electron impact ionization mode (GC/MS Varian 1200) was used. As before [23], CH2O and CO2 were measured by FID after hydrogenation on a Ni/H2 catalyst connected to the exit of the GC column. A good repeatability of the measurements and a good carbon balance (100+/-10%) were obtained in this series of experiments. The composition of RME was 0.1% C14, 5.4% C16, 92.0% C18, 2.0% C20, and 0.5% C22, with mostly one double bond on the acid chain. The equation for the oxidation of RME can be written as follows: C17.92H33O2 + 25.17 O2 = 17.92 CO2 + 16.5 H2O. Modeling For the kinetic modeling of premixed flames, we used the Premix computer code [24]. For simulating the ignition delays, we used the SENKIN code [25]. For the JSR computations, we used the PSR computer code [26] that computes species concentrations from the balance between the net rate of production of each species by chemical reactions and the difference between the input

and output flow rates of species. These rates are computed from the kinetic reaction mechanism and the rate constants of the elementary reactions calculated at the experimental temperature, using the modified Arrhenius equation. The reaction mechanism used here has a strong hierarchical structure. It is based on the comprehensive kerosene and RME oxidation mechanisms [13,23] developed earlier. The reaction mechanism used here consisted of 263 species and 2027 reversible reactions. This mechanism, including references and thermochemical data, is available from the authors ([email protected]). The rate constants for reverse reactions are computed from the corresponding forward rate constants and the appropriate equilibrium constants, Kc = kforward / kreverse calculated using thermochemical data [27-29]. Results and Discussion The oxidation of RME-kerosene was studied at 10 atm in a JSR over the temperature range 740 to 1200 K, for a residence time of 0.5 s. The experiments were performed at three equivalence ratios = 0.5, 1, and 1.5. The initial fuel mole fraction was 0.0595-0.1%. In these conditions, the fuel reacted rapidly, yielding hydrocarbon intermediates (mostly 1-olefins and methane) and oxygenates (mainly formaldehyde and CO). Mole fractions were measured for oxygen, hydrogen, carbon monoxide, carbon dioxide, formaldehyde, methane, ethane, ethene, acetylene, propene, 1-butene, 1-pentene, 1-hexene, 1-heptene, higher alkanes (C3-C10), methyl esters, and simple aromatics. A good repeatability of the results was observed. The accuracy of the mole fractions was typically +/-10% and better than 15% whereas the uncertainty on the experimental temperature was +/-5K. The comparison of these experimental results with those obtained previously for the oxidation of kerosene [23] in similar conditions indicated a strong similitude. This new set of experimental data was used to validate a detailed chemical kinetic reaction mechanism for the oxidation of kerosene-RME blends. This similitude results from the oxidation of the long saturated alkyl chain present in RME and large n-alkanes present in kerosene. However, one should note the reactivity of the RME-kerosene blend is noticeably higher than that of kerosene alone, particularly in the lower temperature range of this study. However, the oxidation of the RMEkerosene blend and kerosene produced very similar concentration of hydrocarbon intermediate. The most noticeable difference between the products of oxidation of the two fuels is the formation of light methyl esters deriving from the oxidation of RME. A comparison between the bio-kerosene blend and kerosene oxidation data is presented in Fig. 1. Specific products of RME oxidation, consisting of methyl esters, were measured. The results are presented in Fig. 2. The oxidation of Jet A-1 was modeled (Fig. 3). As can be seen from this figure, the agreement between the data and the modeling is similar to that previously obtained [14,23].

Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

1e-4

10-2

methyl-2-propenoate methyl-3-butenoate methyl-4-pentenoate methyl-5-hexenoate

Mole Fraction

Mole Fraction

2e-5 1e-5

10-3 O2 H2 CO CO2 CH2O 800 900 1000 1100 1200 1300

10-4

2e-6 1e-6 750

10-5

800

850

900

950

1000 1050

T/K

T/K
1e-4
10-3 CH4 C2H4 C2H2 C2H6 C3H6
methyl-2-propenoate methyl-3-butenoate methyl-4-pentenoate methyl-5-hexenoate

(a)

Mole Fraction

10-4

Mole Fraction

2e-5 1e-5

10-5

2e-6
10-6 800 900 1000 1100 1200 1300

T/K
1e-4
1,3C4H6 1C4H8 1C5H10 1C6H12

1e-6 750

800

850

900

950

1000 1050

T/K

(b)

Mole Fraction

2e-5 1e-5

Figure 2: Methyl esters formed from the oxidation of bio-kerosene (Jet A-1/RME 80/20 in mole) in a JSR at 10 atm, =0.5s; (a) =0.5 and (b) =1.0. The model-fuel used here for modeling the kinetics of oxidation of the Jet A-1/RME blend consisted of a mixture of n-decane/ n-propylbenzene/ n-propylcyclohexane/ n-hexadecane. The inclusion of nhexadecane in the composition of the present modelfuel was motivated by our previous modeling of RME oxidation performed in similar conditions [13] where nhexadecane was used as model-fuel. The model-fuel used for modeling the oxidation of the kerosene-RME blend (80/20 in mole) was n-decane (53.5% in mole), npropylbenzene (17.2%), n-propylcyclohexane (9.5%), and n-hexadecane (19.8%). Comparisons between experimental and computational results are presented in Fig. 4. As can be seen from this figure, the proposed model represents reasonably well the kinetic of oxidation of the bio-kerosene (Jet A-1 / RME blend). A kinetic analysis of the reaction paths during the oxidation at 10 atm of the stoichiometric bio-kerosene mixture indicates that the overall oxidation of the fuel is driven by n-decane. According to the model, at 900 K, where the oxidation of the fuel starts, hydroxyl radicals are the main species involved in the oxidation of the fuel mixture. The oxidation of n-hexadecane and n-decane are responsible for the production of these radicals via a complex reaction scheme that can be summarized as follows: n-C16H34 => x-C16H33 (x=1-8) n-C10H22 => 3-C10H21, 4-C10H21, and 5-C10H21

2e-6 1e-6

800

900

1000

1100

1200

1300

T/K
1e-4
C6H6 benzene

Mole Fraction

2e-5 1e-5

2e-6 1e-6

800

900

1000

1100

1200

1300

T/K

Figure 1: Compared concentration profiles obtained from the oxidation of Jet A-1 (closed symbols) and biokerosene (Jet A-1/RME 80/20 in mole, open symbols) in a JSR (10 atm, =0.5s, 12384 ppm of carbon, =1). In these computations, the jet-fuel was represented by a mixture of n-decane, n-propylbenzene, and n-propylcyclohexane (69%, 20%, and 11% in mole). Since the RME global formula is C17.92H33O2 and that of kerosene is C11H22, the resulting global formula for the biokerosene mixture (80% Jet A-1, 20% RME in mole) is C12.384H24.2O0.4.

Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

10-2

10-2

Mole Fraction

10-3 O2 H2 CO CO2 CH2O 800 900 1000 1100 1200 1300

Mole Fraction

10-3 O2 H2 CO CO2 CH2O 800 900 1000 1100 1200 1300

10-4

10-4

10-5

10-5

T/K

T/K

10-3

Mole Fraction

10-4

Mole Fraction

CH4 C2H4 C2H2 C2H6 C3H6

10-3

10-4

CH4 C2H4 C2H2 C2H6 C3H6

10-5

10-5

10-6

800

900

1000

1100

1200

1300

10-6

800

900

1000

1100

1200

1300

T/K
1e-4
1,3C4H6 1C4H8 1C5H10 1C6H12

T/K
1e-4
1,3C4H6 1C4H8 1C5H10 1C6H12

Mole Fraction

2e-5 1e-5

Mole Fraction
1300

2e-5 1e-5

2e-6 1e-6

2e-6 1e-6

800

900

1000

1100

1200

800

900

1000

1100

1200

1300

T/K
1e-4 1e-4

T/K

Mole Fraction

2e-5 1e-5
C6H6 Toluene

Mole Fraction

2e-5 1e-5
C6H6

2e-6 1e-6

2e-6 1e-6

800

900

1000

1100

1200

1300

800

900

1000

1100

1200

1300

T/K

T/K

Figure 3: The oxidation of Jet A-1 in a JSR ( = 0.75, 10 atm, =0.5 s, 0.067% of fuel, 1.474% of O2, total initial carbon concentration of 7370 ppm). The data (large symbols) are compared to the modeling (lines and small symbols). n-C10H22 => 1-C8H17, 4-C8H17, 2-C8H17, and 3-C8H17. These alkyl radicals isomerize and decompose.

Figure 4: The oxidation of a Jet A-1/RME mixture 80/20 mole in a JSR ( = 0.5, 10 atm, =0.5 s, 806<T/K<1200, 0.089% of fuel, 3.246% of O2, total initial C concentration of 11022 ppm). Their decomposition yields ethylene via -scission, and smaller alkyl radicals such as 1-butyl radical that in turn decomposes. The further reactions yield OH radicals. 1-C4H9 +M => C2H5 + C2H4 + M C2H5 + O2 => C2H4 + HO2

Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

2 HO2 => H2O2 + O2 H2O2 + M => OH+ OH + M CH3 + HO2 => OH + CH3O. The kerosene premixed flame of Dout et al. [18] studied at atmospheric pressure was also simulated in the present study to further verify the validity of the proposed kinetic model. We used the experimental temperature profile reported by the authors [18] in our computations. As can be seen from Fig. 5, the model represents fairly well the experimental concentration profiles of this atmospheric pressure flame. Actually, the computed mole fraction profiles are in very close agreement with those previously computed using an earlier version of the kinetic scheme [23]. The ignition delays have been simulated before for kerosene-air mixtures [14,23]. Here, we compared the predicted kerosene-air and bio-kerosene-air ignition delays for stoichiometric mixtures. According to the present kinetic model, the ignition delays of biokerosene are undistinguishable from those of Jet A-1. This is an interesting result that should be verified experimentally. Unfortunately, no ignition data are presently available for the ignition of bio-kerosene. Conclusion The oxidation of a kerosene Jet A-1 and a biokerosene blend (Jet A-1/RME 80/20 mole) was studied experimentally in a jet-stirred reactor at 10 atm and constant residence time, over the temperature range 740-1200 K, and for variable equivalence ratios in the range 0.5-1.5. Concentration profiles of reactants, stable intermediates, and final products were obtained by probe sampling followed by on-line and off-line GC analyses. No significant variation of the concentration of the products was observed by changing the fuel from Jet A-1 to bio-kerosene besides the formation of monounsaturated methyl esters (<50 ppm) produced from the oxidation of RME. These experiments were simulated using a detailed chemical kinetic reaction mechanism consisting of 2027 reversible reactions and 263 species. The surrogate bio-kerosene model-fuel used consisted of a mixture of n-hexadecane, n-propylcyclohexane, npropylbenzene, and n-decane. For bio-kerosene, the methyl ester fraction was represented by n-hexadecane. The proposed chemical kinetic reaction mechanism used in the modeling yielded a good representation of the kinetics of oxidation of kerosene and bio-kerosene under JSR conditions. The data and the model showed the RME-Jet A-1 mixture has a slightly higher reactivity than Jet A-1 at low temperature. The ignition of Jet A-1 and that of biokerosene were simulated. The kinetic model does not predict a significant effect of RME on the ignition of kerosene. Therefore, it seems that switching from kerosene to bio-kerosene (<20% RME) has no major impact on the kinetics of oxidation of the fuel, which is an interesting result allowing the use of relatively simple kinetic schemes. However, regarding the ignition, this result needs experimental confirmation.

Further studies are also needed to test the similitude between kerosene and bio-kerosene blends under other conditions. Such work is underway for the formation of PAH in diffusion flames.
0.2 O2 CO H2 CO2

Mole fraction

0.15

0.1

0.05

0.5

1.5

2.5

z/mm
0.03 0.025 C2H4 C2H2 C2H6 CH4 C3H6

Mole fraction

0.02 0.015 0.01 0.005 0

0.5

1.5

2.5

z/mm
0.0012 0.001 Allene Benzene

Mole fraction

8e-4 6e-4 4e-4 2e-4 0

0.5

1.5

2.5

z/mm

Figure 5: The oxidation of kerosene under premixed flame conditions (1 atm, 0.010739794 g/cm2/s, initial mole fractions: 0.0319 of kerosene, 0.28643 of oxygen). The data of [9] (symbols) are compared to the modeling (lines). The initial mole fractions used in the modeling were: n-decane, 0.02463685; n-propylbenzene, 0.004993912; n-propylcyclo-hexane, 0.003662271, O2, 0.28643; N2, 0.680276967. References [1] Shauck, M.E., Zanin, G., and Alvarez, S., IASH 2005, 9th International Conference on Stability, Handling and Use of Liquid Fuels, Sitges, Spain, 18-22 Sept. 2005. [2] Delafontaine, M., Dunn, R.O., Hubert, M., and Krishna, C.R., IASH 2005, 9th International

Back to Table of Contents

THIRD EUROPEAN COMBUSTION MEETING ECM 2007

Conference on Stability, Handling and Use of Liquid Fuels, Sitges, Spain, 18-22 Sept. 2005. [3] Daggett, D., Hadaller, O., Hendricks, R., and Walther, R., 25th International Congress of the Aeronautical Sciences, Hamburg, Germany, 3-8 Sept. 2006. [4] Ryan, L., Convery, F., and Ferreira, S., Energy Policy, 34 (2006) 31843194 [5] US Environmental Protection Agency, A comprehensive analysis of biodiesel impacts on exhaust emissions, Report EPA420-P-02-001, Oct. 2002. [6] Garrett, T.K., Automotive fuels and fuel systems. Vol.2: Diesel, Pentech Press, London (1994). [7] Sheehan, J., Camobreco, V., Duffield, J., Graboski, M., and Shapouri, H., Life cycle inventory of biodiesel and petroleum diesel for use in an urban bus, National Renewable Energy Laboratory Report NREL/SR-580-24089, May 1998. [8] Schrder, O., Frahl, J., Munack, A., Krahl, J., and Bnger, J., SAE 1999-01-3561. [9] Montagne, X., SAE 962065 (1996) 247-256. [10] Krahl, J., Munack, A., Bahadir, M., Schumacher, L., and Elser, N., SAE 962096 (1996) 319-338. [11] Graboski, M.S. and McCormick, R.L., Prog. Energ. Combust. Sci, 24 (1998) 125-164. [12] Ramadhas, A.S., Jayaraj, S., and Muraleedharan, C., Renewable Energy, 29 (2004) 727-742. [13] Dagaut, P., Gal, S., and Sahasrabudhe, M., Proc. Combust. Inst., 31 (2007) 2955-2961. [14] Dagaut, P. and Cathonnet, M., Prog. Energy Combust. Sci., 32 (2006) 48-92. [15] Dagaut, P., Reuillon, M., Boettner, J.-C., and Cathonnet, M., Proc. Combust. Inst., 25 (1994) 919-926. [16] Dagaut, P., Reuillon, M., Cathonnet, M., and Voisin, D., J. Chim. Phys. Phys.-Chim. Biol. 92 (1995) 47-76. [17] Cathonnet, M., Voisin, D., Etsouli, A., Sferdean, C., Reuillon, M., Boettner, J.-C., and Dagaut, P., Symposium of the Applied Vehicle Technology Panel on Gas Turbine Engine Combustion, Emissions and Alternative Fuels, Lisbon, Portugal, 12-16 October 1998. RTO Meeting Proceedings 14 (1999) 1-9. [18] Dout, C., Delfau, J.-L., Akrich, R., and Vovelle, C., Combust. Sci. and Technol., 106 (1995) 327344. [19] Mawid, M.A., Park, T.W., Sekar, B., and Arana, C., AIAA 2002-3876 (2002) 38th Joint Propulsion Conference and Exhibit, Indianapolis, Indiana, 710 July 2002. [20] Dagaut, P., Phys. Chem. Chem. Phys., 4 (2002) 2079-2094. [21] Mawid, M.A., Park, T.W., Sekar, B., and Arana, C., AIAA 2003-4938 (2003) 39th Joint Propulsion Conference and Exhibit, Huntsville, Alabama, 2023 July 2003. [22] Mawid, M.A., Park, T.W., Sekar, B., and Arana, C., AIAA 2004-4207 (2004) 40th Joint Propulsion

Conference and Exhibit, Fort Luderdale, FL, 11-14 July 2004. [23] Dagaut, P., Proc. ASME Turbo Expo 2006: Power for Land, Sea, and Air, May 8-11, 2006, Barcelona, Spain. [24] Kee, R.J., Grcar, J.F., Smooke, M.D., and Miller, J.A., Premix: A Fortran program for modeling steady laminar one-dimensional premixed flame, Sandia Report No. SAND85-8240 (1985) Sandia National Laboratories, Livermore, CA. [25] Lutz, A.E, Kee, R.J., and Miller, J.A., Senkin: A Fortran program for predicting homogeneous gas phase chemical kinetics with sensitivity analysis, Sandia Report No SAND87-8248 (1988) Sandia National Laboratories, Livermore, CA. [26] Glarborg, P., Kee, R.J., Grcar, J.F., and Miller, J.A, PSR: A FORTRAN program for modeling wellstirred reactors, Sandia Report No. SAND86-8209 (1986) Sandia National Laboratories. [27] Tan, Y., Dagaut, P., Cathonnet, M., and Boettner, J.-C., Combust Sci and Technol 102 (1994) 21-55. [28] Kee, R.J., Rupley, F.M., and Miller, J.A., Thermodynamic data base, Sandia Report SAND87-8215 (1991) Sandia National Laboratories, Livermore, CA, 1991. [29] Muller, C., Michel, V., Scacchi, G., and Cme, G.M., J. Chim. Phys. Phys.-Chim. Biol., 92 (1995) 1154-1178.