Experimental and Modeling Study of C₁–C₃ Hydrocarbon Ignition in the Presence of Nitric Oxide

Chemical kinetic modeling of hydrocarbon fuel components in practical fuels (e.g., natural gas, kerosene, gasoline, or diesel) relies on the hierarchical nature of fuel decomposition [1] to form C₁–C₃ species. During the oxidation of long chain hydrocarbon fuels, the fuel molecules often undergo unimolecular and bimolecular decomposition to form alkyl radicals (CH₃, C₂H₅, C₃H₇, etc.) and alkenes (e.g., C₂H₄, C₃H₆, etc.) as intermediates. Figure 1 shows the major reaction pathways that involve formation of alkene and alkyl radicals from beta-scission reactions. This was further demonstrated by the shock tube experiments of Davidson et al. [2] in which the time-history profile of n-dodecane oxidation showed that C₂H₄ was formed as one of the major intermediate species. Moreover, the experimental data showed that there was a significant time lag (an order of magnitude difference) between the consumption of n-dodecane and C₂H₄ under combustion conditions. Therefore, the primary objective of the current work is to investigate the effect of nitric oxide (NO) on the ignition behavior of low molecular weight hydrocarbons that are relevant to the oxidation of practical fuels.

The presence of NO_x (NO and NO₂) in the oxidizer stream occurs in various practical devices where combustor design includes exhaust (or flue) gas recirculation (e.g., internal combustion engines and furnaces), vitiated air (e.g., gas turbine exit streams, and high-speed propulsion test rigs), and internal recirculation zones (e.g., gas turbines). Previous works by the authors [3–5] showed that NO_x plays a significant role in promoting ignition of jet fuels and its surrogate components. Experimental and direct numerical simulation studies of Lee et al. [6] have demonstrated the importance of NO on the reignition process of a vortex-perturbed counter flow flame. In the current work, experiments are carried out to investigate the effect of NO on the ignition of C₁–C₃ hydrocarbons, namely, CH₄, C₂H₄, C₂H_6, and C₃H₆. The current study builds upon previous work performed with C₃H₈ [7]. For the reasons discussed above, it is important to understand the interactions of these fuel molecules with NO_x in order to develop a detailed reaction mechanism for the combustion of jet fuels under vitiated conditions.

CH₄, C₂H_6, and C₃H₈ are primary fuel components in natural gas fuels. The influence of NO_x on CH₄ oxidation has been studied extensively [8–17], whereas limited work is reported on C₂H₄ [10,18,19], C₂H₆ [10,18,20,21], C₃H₆ [10], and C₃H₈ [7,10,22]. However, most of these studies focused on oxidation with heavily diluted fuel/oxidizer mixtures, which do not represent the scenario in practical combustion systems. Detailed information from the literature on NO_x interaction with C₁–C₃ hydrocarbon fuels is provided in Table 1. The table compares the current experimental conditions with previous work, most of which used heavily diluted fuel/oxidizer mixtures with low carbon content (<1 mole%) except for few studies related to ignition experiments [7,8,20]. Previous work by the authors [7] showed that the reaction pathways of NO in influencing hydrocarbon oxidation are sensitive to the ratio of NO to the total carbon present in the system. It was shown that experimental measurements from heavily diluted oxidizer systems are less influenced by the interaction between NO_x and the fuel molecule or its derivative radicals.

Recent work by the authors [7] also showed that the fuel-dependent reaction pathway—interactions between NO_x and a fuel molecule and its radicals—plays a crucial role in promoting oxidation. For example, H-atom abstraction by NO₂ from normal alkanes to form alkyl radicals plays a significant role in promoting oxidation. It was also shown [7] that contribution of this fuel-dependent reaction pathway becomes significant at conditions relevant to practical applications where the absolute value of the fuel concentrations is much higher than many reported studies listed in Table 1. Also, it should be noted that there is a lack of experimental data for the ignition of C₂H₄ and C₃H₆ in the presence of NO_x. Hence, in the current work, experiments were carried out to investigate the effect of NO_x on the ignition of CH₄, C₂H₆, C₂H₄, and C₃H₆ with varying levels of O₂ present in typical vitiated conditions. A detailed kinetic mechanism is used to investigate reaction pathways of NO_x in promoting ignition of these fuels.

Experiments were performed using an atmospheric pressure flow reactor to investigate the effect of NO on the ignition delay time (IDT) of C₁–C₃ hydrocarbon fuels. The species examined in this study included common natural gas components (CH₄ and C₂H₆) and olefin intermediates (C₂H₄ and C₃H₆). The flow reactor consists of a premixing section and a test section. The premixing section, shown in Fig. 2, consists of radial injectors and a swirler to achieve near-perfect mixing of the fuel and oxidizer at inlet of the flow reactor. The portion of the premixing section located downstream of the swirler is an expanding duct that gradually increases the reactor diameter from 1.3 cm to 5.1 cm. The test section consists of an alumina tube with a nominal diameter of 5.1 cm and an overall length of approximately 240 cm.

For each test condition, the mass flow rate of the oxidizer stream was maintained nominally at 2 g/s. The fuel flow was independently varied to provide test conditions at the targeted fuel equivalence ratios. The gaseous fuel was premixed with nitrogen (N₂ to fuel molar ratio is 5) and preheated up to 675 K prior to radial injection upstream of the swirler. The fuel was premixed with nitrogen in order to improve jet penetration and mixing of the fuel and oxidizer. The oxidizer stream was preheated (up to 1000 K) so that the temperature of fuel/oxidizer mixture reaches the temperature of the test section prior to entering the mixing section. A uniform temperature in the test section was maintained by independently controlled electrical heating. A more detailed description of the flow reactor apparatus can be found in Ref. [7].

Ignition delay time is determined by measuring the time delay between when the fuel/oxidizer mixture enters the diffuser section and the time of ignition. The time at which the fuel/oxidizer mixture enters the diffuser is determined using the activation time of the fuel injection solenoid and infrared laser absorption measurements to determine the presence of hydrocarbons entering the diffuser. Ignition of the fuel/oxidizer mixture is measured by the detection of the chemiluminescence emission of the OH* radical that occurs during ignition. At the downstream end of the flow reactor test section, a photomultiplier tube (PMT) is placed behind a quartz window with direct line of site down the axis of the entire flow reactor test and premixing sections. A 310 nm narrow bandpass filter (±5 nm) is used to filter out light that is not emitted by OH*. The difference between the time at which the fuel/oxidizer mixture enters the diffuser section and the time of initial PMT excitation due to OH* emissions is the measured ignition delay time. An example of the time-history profile of the signals used to determine the ignition delay time is provided in Fig. 3.

The experimental data reported in this paper are the average of at least three measurements. The experimental uncertainty based on two standard deviations (95% confidence interval) of the measured values are provided in the included figures.

The scope of the current experimental work is comprised of the following:

• Fuels: CH₄, C₂H₄, C₂H_6, and C₃H₆

• O₂ in the oxidizer stream: 15 and 21 mole%

• Reactor temperature: 900 and 950 K

• Equivalence Ratio: 1.0

• NO_x in the oxidizer stream: 50, 100, 250, 500, and 750 ppm

Standard gas cylinders with 0.1% and 0.5% (nominally) NO in N₂ were used in the experiments to make desired concentrations of NO in the oxidizer (O₂/N₂) using rotameters. Testing has shown that mixing of NO with the oxidizer converts only a very small amount of NO into NO₂ depending on the NO concentration. Measurements at room temperature showed that mixing of NO with the oxidizer produced 0 ppm NO₂ with 50 ppm NO, whereas only 5% of the NO was converted to NO₂ for the mixture with 750 ppm NO. Measurements of NO and NO₂ in the oxidizer were made using a thermo environmental chemiluminescence NO_x analyzer.

The detailed chemical kinetic mechanism used in the current work is based on the natural gas mechanism developed by the authors [7,23–25] that was validated for CH₄, C₂H₄, C₂H₆, C₃H_6, and C₃H₈ oxidation chemistry over a wide range of conditions. This mechanism also includes a detailed reaction subset for NO_x chemistry that has been validated for high-temperature NO_x emissions [23,24] as well as low-temperature vitiated kinetics [7]. In the current work, the mechanism is validated for the kinetics of NO_x interactions with CH₄, C₂H₄, C₂H_6, and C₃H₆ under vitiated conditions. Most of the rate parameters for the reactions of hydrocarbon species with NO_x were adopted from the chemical kinetic mechanism developed for the influence of NO_x on C₃H₈ oxidation [7] as well as the work of Mendiara and Glarborg [26]. The rate parameters of some of the important reactions are listed in Table 2. Most of the reaction rate parameters for C₂ and C₃ hydrocarbon species with NO_x are based on the values recommended for C₁ species. Therefore, there is a considerable uncertainty in the reaction rates due to the lack of experimental or theoretical work in the literature for the rate estimation of these reactions [20]. The reaction rate parameters for H-atom abstraction of C₂H₆ and C₂H₅ molecules by NO₂ were reduced by a factor of 2 (within the uncertainty limits of the published values) in order to accurately predict the effect of NO_x on the ignition delay time of C₂H₆ obtained in the present work. The current model was able to capture the effect of NO_x on the ignition of C₂ and C₃ hydrocarbon fuels as discussed below.

A positive sensitivity coefficient indicates that the reaction inhibits ignition, whereas a negative value means that the reaction promotes ignition.

Figures 4 and 5 compare the current model predictions with the experimental data of Dagaut et al. [18] for the effect of NO on the oxidation of C₂H₄ and C₂H₆. These experiments were conducted in an atmospheric pressure jet-stirred reactor (JSR) with fuel/O₂ in ∼99 mole% N₂ at 0.5 equivalence ratio with and without the NO addition. The reactor residence time was maintained at 120 ms. The JSR experiments were modeled using perfectly stirred reactor simulations in Cantera [27]. The experimental data and the model predictions show that the effect of NO in reducing the temperature of the onset of fuel decomposition is more pronounced for C₂H₄ (from 950 K to 800 K) than C₂H₆ (from 975 K to 875 K). A similar trend was observed for reducing the ignition delay time with NO addition in the current work as discussed in the Results and Discussion section. However, the reaction pathways for the effect of NO in promoting the oxidation have significant differences for the conditions of Dagaut et al. [18] (Figs. 4 and 5) and the current work.

Figure 6 shows the reaction sensitivity coefficients for important reactions computed with respect to C₂H₆ concentration for JSR experimental conditions (ϕ = 0.5; 0.875% O₂) in Fig. 5 at 950 K with 220 ppm NO. Figure 6 also shows the reaction sensitivity coefficients with respect to ignition delay time computed for C₂H6/air (ϕ = 0.5) at 950 K with 220 ppm NO. Because of the differences in the dilution, the carbon to NO molar ratio is approximately 11 for the experimental conditions in Fig. 5 and 210 for the current work, at an equivalence ratio of 0.5.

It should also be noted that reactions (R9) and (R10) have the third and fifth largest negative sensitivity coefficients for the ignition of C₂H₆, respectively. However, reaction (R9) does not have any significant influence on C₂H₆ decomposition at JSR conditions while a positive sensitivity coefficient for reaction (R10) indicates that it hinders C₂H₆ oxidation at these conditions. Nevertheless, fuel-independent reactions (R2) and (R4) are the two most important NO_x reactions that promote the oxidation at JSR conditions.

Figure 7 shows the effect of NO_x on ignition delay time of a stoichiometric CH₄/air mixture at 900 K and 950 K. It should be noted that ignition was not recorded in the absence of NO_x as the ignition delay time is longer than the maximum reactor residence time at these temperatures. The detailed chemical kinetic mechanism predicts an ignition delay time of around 2 s at 950 K for the CH₄/air mixture, whereas the reactor residence time is approximately 1 s at these conditions. Figure 7 shows that an addition of 100 ppm NO_x reduced the ignition delay time significantly (more than 75% based on the model predictions) as compared to 0 ppm NO addition. The effect of NO_x in promoting the ignition becomes marginal for NO_x addition above 500 ppm. It is also interesting to note that the addition of NO_x reduces the effect of temperature on ignition delay time. For example, the ignition delay time at 900 K is almost three times that of 950 K with 100 ppm NO_x, whereas it is only twice as long with 750 ppm NO_x addition.

Methane initiation reactions with OH, O, and H species generate CH₃ radicals with the consumption of approximately 90% of the fuel molecules. However, the chain-branching reaction (R12) has to compete with the termination reactions (R11) and (R13) to convert relatively less reactive CH₃ into CH₃O and OH. This results in a longer chemical induction period for the ignition of CH₄ at these conditions in the absence of NO.

When NO is present, the reaction between CH₃ and NO₂ via (R4) shifts the balance toward CH₃O formation. This is further enhanced by reaction (R2) in which the NO is converted to NO₂ while HO₂ is converted to the more reactive OH radical. In the process, the chain branching reaction (R6) becomes the most sensitive reaction for the ignition of CH₄ with NO_x addition as shown in Fig. 8. Species flux analysis for the conditions in Fig. 8 shows that reactions (R2) and (R5) are responsible for 92% and 8% of the NO₂ generated during the induction period, respectively. Meanwhile, 88% of the NO₂ is consumed by reaction (R4) to convert CH₃ radical into CH₃O, whereas around 10% of the NO₂ takes part in reaction (R3).

Figure 9 shows the influence of NO_x addition on the ignition delay time of stoichiometric C₂H₆/air mixture at 900 and 950 K. Figure 10 shows the effect of NO_x on C₂H₆ ignition at 950 K with 15 mole% and 21 mole% O₂. The presence of NO_x in the inlet stream reduces the ignition delay time at these conditions. However, the effect of NO_x in promoting the ignition of C₂H₆ is not as pronounced as in CH₄ cases shown in Fig. 7. For example, 100 ppm of NO_x addition reduces the ignition delay by less than 25% based on the experimental data in Figs. 9 and 10. As expected, ignition delay time of mixtures with 21% O₂ is shorter than that of 15% O₂ level in the absence of NO_x. However, as the NO_x level is increased above 250 ppm, the ignition delay times are almost the same for 21% O₂ and 15% O₂ levels. A similar trend was observed for C₃H₈ in a previous study by the authors [7].

Figure 11 shows the sensitivity coefficients for some of the most important reactions for the ignition of C₂H₆ with and without NO addition. The sensitivity coefficients were computed for stoichiometric C₂H₆/air mixture at 950 K with 0 and 100 ppm NO using the detailed kinetic mechanism. Reactions (R7) and (R8) have the two largest negative sensitivity coefficients. Species flux analysis shows that the initiation reaction (R8) is responsible for approximately 50% of the H₂O₂ produced while chain branching reaction (R7) generates approximately 80% of the OH radical produced at 950 K with 0 ppm NO.

With the addition of 100 ppm NO, reaction (R9) becomes an important initiation step for C₂H₆ ignition with the third largest negative sensitivity coefficient. H-atom abstraction of C₂H₆ by NO₂, which was generated from reaction (R2), produces C₂H₅ radical and HNO₂. Also, the C₂H₅ radical reacts with NO₂ to regenerate NO while producing C₂H₅O via reactions (R10). In order to demonstrate the importance of reactions (R9) and (R10) for C₂H₆ ignition, Fig. 9 also shows the model predictions without these reactions at 900 K.

Figure 12 shows the ignition delay time of stoichiometric C₂H₄/air mixture as a function of NO_x addition at 900 K and 950 K. Figure 13 shows the effect of NO_x on C₂H₄ ignition at 950 K with 15 mole% and 21 mole% O₂ levels. The presence of NO_x in the inlet reduces the ignition delay time of C₂H₄ at these conditions. For example, 100 ppm of NO_x reduced the ignition delay time by approximately 40%. This puts the effectiveness of NO_x in promoting C₂H₄ ignition between CH₄ and C₂H₆: CH₄ > C₂H₄ > C₂H₆. It should also be noted that the effect of NO_x in promoting ignition appears to be less sensitive to O₂ level for C₂H₄ than C₂H₆ based on the experimental data. For example, 100 ppm NO_x reduced the ignition delay times of stoichiometric C₂H₆ mixtures by approximately 25% and 10% at 15 mole% and 21 mole% O₂ levels, respectively, as shown in Fig. 10. However, 100 ppm of NO_x addition reduced the ignition delay time of C₂H₄ by approximately 40% at 15 mole% and 21 mole% O₂ levels as shown in Fig. 13.

Species flux analysis for the conditions in Fig. 14 shows that approximately 70% of the C₂H₃ radical is produced via initiation reaction (R18) while approximately 50% of it is consumed by reaction (R16). Production of O and OH species via reactions (R16) and (R17) makes C₂H₄ a more reactive fuel with shorter ignition delay times than C₂H₆.

A portion of the NO₂ is regenerated from HONO produced in reaction (R19) via reaction (R15). It is also interesting to note that the consumption of HO₂ by NO to form NO₂ and OH via reaction (R2) reduces the importance of reaction (R17) while increasing for the importance of reaction (R18). The sensitivity coefficient for reaction (R17) is reduced by about 65% in the presence of NO while the value for reaction (R18) is increased by about 50% as shown in Fig. 14.

Species flux analysis shows that 98% of NO₂ is produced by reaction (R2) in the induction period while about 35% of it is converted to HONO via reaction (R19). Most of the HONO generated in reaction (R19) is decomposed into NO and OH via reaction (R15). Meanwhile, reactions (R3) and (R4) convert about 10% and 35% of NO₂ into NO, respectively, while generating OH and CH₃O radicals in the process.

Figure 15 shows the ignition delay times for stoichiometric C₃H₆/air as a function of NO_x at 900 K and 950 K. Figure 16 shows the effect of NO_x on C₃H₆ ignition delay time at 950 K with 15 mole% and 21 mole% O₂ levels. The addition of NO_x reduces the ignition delay time C₃H₆ significantly. For example, 100 ppm NO_x reduced the ignition delay time by about 60%, whereas 500 ppm NO_x reduced it by about 90% at 950 K.

It should also be noted that reaction (R2) has the largest negative sensitive coefficient in the presence of NO_x as shown in Fig. 17. Species flux analysis for the conditions in Fig. 17 shows that reactions (R2) and (R4) generate 95% and 5% of the NO₂, respectively, in the induction period while around 80% of it is consumed by reaction (R24). It is also interesting to note that reaction (R24) becomes an important pathway for allyl radical consumption with 23% compared to 19% for reaction (R22).

Figure 18 compares the effectiveness of NO_x in promoting the ignition of the fuels investigated in the current work. The change in ignition delay time relative to the change in NO_x addition (i.e., Δτ_idt/ΔX_NOx) is provided in Fig. 18 as a function of NO_x in the inlet. The effectiveness of NO_x addition in promoting the ignition is ranked based on the overall average percent reduction in the ignition delay time observed for each fuel: CH₄ > C₃H₆ > C₃H₈ > C₂H₆ > C₂H₄. It should be noted that the negative value for C₂H₆ at 50 ppm is due to a small increase in the ignition delay time, shown in Fig. 9, which lies within the experimental uncertainty of the measurement.

An experimental and modeling study was performed to investigate the effect of nitric oxide on the ignition of C₁–C₃ fuels, which are important intermediate species generated during the decomposition of long chain hydrocarbon fuel components typically present in jet fuels. CH₄ and C₂H₆ are also the major components of natural gas fuels. Ignition delay time measurements were carried out using an atmospheric pressure flow reactor facility with alkanes (CH₄ and C₂H₆) and olefins (C₂H₄ and C₃H₆) at 900 K and 950 K temperatures with 15 mole% and 21 mole% O₂. Based on the experimental data, the overall ranking of the effectiveness of NO_x in promoting the ignition is: CH₄ > C₃H₆ > C₃H₈ > C₂H₆ > C₂H₄. For example, 500 ppm of NO_x reduced the ignition delay time of CH₄ by approximately 90%.

A detailed kinetic mechanism, developed for natural gas fuels ranging from C₁ to C₃, was used for model predictions as well as for the sensitivity and species flux analyses. The reaction between HO₂ and NO plays a critical role in promoting the ignition by generating OH radical while converting NO into NO₂. Furthermore, fuel-dependent reaction pathways also play important role in promoting the ignition of these fuels. NO₂ reacts with C₂H₆ to enhance the ignition of C₂H_6, whereas its reactions with CH₂CHO and aC₃H₅ radicals are important rate-limiting steps for the ignition of C₂H₄ and C₃H₆, respectively. In addition, NO₂ reactions with C₂H₄ and C₂H₅ also play important roles in promoting the ignition of C₂H₄ and C₂H₆, respectively. Reaction between NO₂ and the relatively unreactive CH₃ radical to produce CH₃O plays an important role in reducing the ignition delay time of CH₄ compared to the other fuels examined.

The material was cleared for public release by the U.S. Air Force (Case No. 88ABW‐2016‐6489).

• United States Air Force (Grant No. FA8650-15-C-2531).

IDT =

ignition delay time

JSR =

jet-stirred reactor

k =

reaction rate constant

T_ign =

ignition delay time

X =

mole fraction

ϕ =

equivalence ratio