Study on ignition delay of multicomponent syngas using shock tube

Study on Ignition Delay of Multi‐Component Syngas Using Shock Tube Luong Dinh Thi,1 Yingjia Zhang,1* Jin Fu,1 Zuohua Huang1* and Yang Zhang2 1. State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, Xi'an 710049, People's Republic of China 2. Combustion and Flow Diagnostics Research Group, Department of Mechanical Engineering, University of Sheffield, Sheffield S1 3JD, UK

Ignition delays of four typical syngas mixtures were investigated using both experimental and simulated methods. The shock tube experiments were conducted behind the reflected shock waves at temperature ranges from 870 to 1 350 K, pressures of 0.2, 1.0 and 2.0 MPa. Six available kinetic models were evaluated by comparing them to the new ignition delay data obtained. NUIG C3 model was in good agreement with the experimental results. The effect of the pressure on ignition delay of all four syngas mixtures was similar to that of hydrogen. The promoting and inhibiting ignition were experimentally observed at high and moderate temperatures, respectively. Sensitivity analyses were conducted at different temperatures and pressures to identify the dominant elementary reactions in the ignition process of the syngas. The results indicated that the ignition delay was generally most sensitive to the chain branching reaction H þ O2 ¼ O þ OH at both high and low temperatures. Decomposition reaction H2O (þM) ¼ H þ OH (þM) was of importance for the syngas with relative inert gas CO2 and N2 at a high temperature. The reaction CO þ OH ¼ CO2 þ H involving CO kinetic dominated the ignition at a high temperature for the syngas with a high CO level. The inhibiting effect of pressure was attributed to the decrease of sensitivity of the reactions H þ O2 ¼ O þ OH and O þ H2 ¼ H þ OH and the increase of sensitivity of the reaction H þ O2 (þM) ¼ HO2 (þM) at a moderate‐low temperature. Keywords: syngas, shock tube, ignition delay, kinetic analysis

INTRODUCTION

A

s the main coal gasification process produces, syngas has been expected to play an important role in future energy production due to its attractive fuel supply flexibility and wide source. Currently, the syngas is being widely used to the integrated gasification combined cycle (IGCC) which is an advanced combustion technology with high efficiency and low emissions. Although syngas is mainly composed of H2 and CO, potential variables need to be considered in direct combustion of syngas mixtures including the level of CO2, N2, H2O, CH4 and other impurities.[1] Thus, the combustor used in practical equipment on syngas must be fuel‐flexible as the syngas composition can vary greatly. Variations in its composition can have considerable influence on the performance and operability of the combustion device.[2] It is well known that auto‐ignition delay is closely related to flashback and knocking. To develop a new concept combustor, the ignition characteristics of syngas with various compositions must be further understood. Furthermore, ignition delay is traditionally a key parameter for the validation and modification of chemical kinetic models. For these practical reasons and its fundamental interest in the combustion models of hydrocarbon fuels[3], many experimental and modelling studies have been conducted for syngas over a wide range of conditions using different facilities. A shock tube has been used to perform the measurements of ignition delays at a high temperature and low pressure in previous studies on CO/H2 with diluted gas.[4,5] Also, work for auto‐ignition of syngas/air mixtures with various CO/H2 blending ratios was carried out by Kalitan et al.[6] at temperatures from 890 to 1 300 K and pressures from 0.1 to 1.5 MPa. Herzler and Naumann[7] measured the ignition delays of six different lean H2/CO/O2/Ar mixtures including 5%H2/95% CO and 50% H2/50% CO with three dilutions at temperatures from 1 020 to 1 260 K and pressure of about 1.6 MPa. Petersen et al.[8]

VOLUME 92, MAY 2014

measured experimentally the ignition delays of CO/H2/air mixtures under practical conditions (940–1 148 K, 1.7–3.3 MPa). In the latest study reported by Mathieu et al.[9] the ignition delays were measured, and the effects of NH3 impurity on the ignition were investigated for the real syngas produced from biomass (0.29659% CO/0.29659% H2/0.15748% CO2/0.08924 CH4/ 0.20997% H2O/0.95013% O2 in 98% Ar) at temperatures from 990 to 2000 K and pressures of 0.16, 1.25 and 3.2 MPa. They explored the influence of the different compositions on the syngas ignition, and suggested the CO2 addition has a negligible effect on the ignition delays, while the CH4 addition presents a noticeable promoting effect. Furthermore, a rapid compression machine has been also used to conduct the auto‐ignition behaviour of syngas mixtures at moderate‐low temperature. Mittal et al.[10] studied the auto‐ ignition of H2/CO/O2 mixtures containing between 0% and 80% CO at temperatures from 950 to 1 100 K, pressures from 1.5 to 5.0 MPa, and equivalence ratios from 0.36 to 1.6. They suggested that the reactions involving chemical properties of HO2 and H2O2 were considerably important for H2/CO oxidation at intermediate temperature and high pressure. Walton et al.[11] investigated the ignition of simulated syngas mixtures (H2/CO/O2/N2/CO2) with H2:CO ratios from 0.25 to 4.0 at conditions that are relevant to

Additional supporting information may be found in the online version of this article at the publisher’s web‐site. *Author to whom correspondence may be addressed. E‐mail address: [email protected] (Yingjia Zhang), [email protected] (Zuohua Huang) Can. J. Chem. Eng. 92:861–870, 2014 © 2013 Canadian Society for Chemical Engineering DOI 10.1002/cjce.21938 Published online 19 November 2013 in Wiley Online Library (wileyonlinelibrary.com).

THE CANADIAN JOURNAL OF CHEMICAL ENGINEERING

861

gas‐turbine technologies. They provided a simplified Arrhenius type correlation to compute ignition delays for designing combustor with syngas fuel. Fotache et al.[12] reported a non‐premixed ignition in counter‐flow CO/H2 versus heated air jets. They indicated that the temperature range of the glow regime and the glow intensity decreased with the increase of H2, and the glow could not be detected when the H2 fraction is over 73%. From the literature survey on the ignition of syngas, it is evident that little work on measuring ignition delay has been conducted for real syngas mixtures at wide conditions, as well as assessing systematically chemical kinetic models. For these reasons, the aim of this study has three sections. The first is to provide new shock tube ignition delay data for four typical syngas mixtures at wide conditions. The second is to evaluate the adaptability of six available kinetic models through a comparison between the measurement and simulation. The final is to interpret the effects of temperature and pressure on auto‐ignition, and identify the key elementary reactions in the ignition kinetics of four syngas mixtures using selected kinetic model based sensitivity analysis. EXPERIMENTAL SETUP AND SIMULATED PROCEDURE Experimental Setup All measurements were carried out in a steel shock tube described in detail previously by Zhang et al.[13] The schematic diagram is given in Figure 1. Briefly, a double diaphragm divides the shock tube into a driver section of 4.0 m and a driven section of 5.3 m in length. To obtain a longer testing time, tailored conditions were created by using the mixtures with adjustable He/N2 ratios as the driver gas. Before each experiment, the driven section was evacuated to pressure below 0.02 Pa. The velocities of incident shock wave were measured by four fast‐response pressure transducers and three time counters located on 30 cm intervals. The temperature behind the reflected shock wave was calculated using the chemical equilibrium software GASEQ.[14] The uncertainty of temperature calculation is 6 K. Test mixtures were prepared according to Dalton’s law of partial pressure in a 128 L stainless‐steel tank, which was evacuated to pressure below 0.02 Pa before introducing the testing mixtures. The purity of H2 is 99.999%, of CO is 99.99%, of CO2 is 99.99%, of N2 is 99.999%, of O2 is 99.99% and of Ar is 99.995%. Detailed compositions and content of the syngas mixtures are given in

Table 1. Reliability of the apparatus has been validated in the previous works for various fuel mixtures and under different conditions.[15] The ignition was monitored by measuring the pressure with a piezoelectric pressure transducer located at the end‐wall. Also, OH emission (wavelength of 307 nm) at the same position was measured with a photomultiplier through a narrow band pass filter. The ignition delay was defined as the time interval between the initiation of the reflected shock wave and the onset of ignition, which was determined by the extrapolation of the steepest rise in the OH chemi‐luminescence signal to the zero baseline, as shown in Figure 2. Due to the effect of practical conditions, the pressure behind the reflected shock wave increased linearly at a rate of 4.22% per millisecond (i.e. dp/dt ¼ 4.22%/ms), it was considerably important to calculate the ignition delay. Simulated Procedure Calculations on ignition delays and kinetic analyses of the syngas were made using the CHEMKIN II[16] program with SENKIN[17] package. According to the shock wave theory, the pressure and temperature behind the reflected shock wave are maintained in a steady state under ideal conditions. However, using a common constant volume (V) system with a constant internal energy (U) assumption (i.e. constant U, V) is unable to reproduce the longer ignition.[18] Many previous studies[19–22] indicated that an obvious rise of pressure behind the reflected shock waves will be presented in the practical shock tube experiments due to the boundary layer growth and non‐ideal effects, which lead to a significant decrease in the experimental ignition delays. In the current study, a similar observation was also found in the shock tube experiments, as shown in Figure 2. In this case, it was necessary to concert the effect of the dp/dt (4.22%/ms) on the calculated ignition delays. Chaos and Dryer[20] confirmed that the SENKIN/VTIM approach can provide much improved predictions in ignition delays and yield very similar values to those computed using the CHEMSHOCK model proposed by Li et al.[23] for low concentration mixtures at moderate‐low temperatures. Therefore, the SENKIN/VTIM proposed by Chaos and Dryer[20] was used to calculate the ignition delays. To further confirm the reasonability of the SENKIN/VTIM used, a comparison between the predicted and measured ignition delays was carried out for SH70 mixture at pressure of 1.0 MPa, as shown in Figure 3. The calculations were performed using both SENKIN/

Figure 1. Schematic diagram of shock tube.

862


VOLUME 92, MAY 2014

Table 1. Composition and fraction of syngas mixtures (% volume). Mixture SH70 SH33 SH50C30 SH26N58

% H2

% CO

% CO2

% N2

% H2/ (H2 þ CO)

70 33 35 11

30 67 35 31

0 0 30 0

0 0 0 58

70 33 50 26

VTIM approach and constant U, V assumption with NUIG C3 mechanism.[24] The result indicated that the measured ignition delays with temperatures above 1 100 K were consistent with those calculated using SENKIN/VTIM and constant U, V. However, the measured ignition delays with temperatures below 1 100 K were shorter than those predicted with the constant U, V assumption, and were in good agreement with those predicted with the SENKIN/VTIM. RESULTS AND DISCUSSION Kinetic Model Evaluation To perform an evaluation of the current kinetic models, the six following mechanisms were considered: (1) NUIG C3,[24] developed by Combustion Chemistry Center at National University of Ireland, consists of 118 different chemical species and 663 elementary reactions. (2) GRI 3.0,[25] built by Gas Research Institute, consists of 53 species and 325 reactions. (3) USC‐Davis, developed by Davis et al.[26] at Combustion Kinetics Laboratory of University of Southern California, consists of 14 species and 30 reactions. (4) UCSD,[27] developed by Combustion Research at University of California at San Diego, includes 50 species and 244 elementary reactions. (5) Princeton‐Li, developed by Li et al.[28] at Fuels and Combustion Research Laboratory of Princeton University, takes 127 species and 1 027 reactions into account. (6) Leeds 1.5 developed by Hughes et al.[29] at University of Leeds, involves 37 species and 175 reactions.

Figure 3. Comparison between calculations performed using SENKIN/ VTIM with dp/dt ¼ 4.22%/ms and constant U, V assumption. Measured ignition delays of SH70 mixture are also presented at pressure of 1.0 MPa.

All the evaluations were performed under different initial pressures for four typical syngas mixtures, that is SH70, SH33, SH50C30 and SH26C58. The measured and predicted results using above six kinetic mechanisms ignition delays were shown in Figures 4–6. Study at pressure of 0.2 MPa Measured and calculated ignition delays at a pressure of 0.2 MPa for SH70, SH33, SH50C30 and SH26N58 mixtures are shown in Figure 4. As can be seen, obvious discrepancies between six models were presented for all the investigated syngas mixtures. Note that the agreement of the different mechanisms is strongly dependent on the H2 content in the syngas mixtures. The results indicated that for the syngas mixtures with a high H2 content, such as SH70 (Figure 4a), the experimental data were well captured over the entire range of conditions by the NUIG C3 and Leeds 1.5 mechanisms, while Princeton‐Li, GRI 3.0 and UCSD mechanisms over‐predicted with the temperature below 1 050 K. Nonetheless, USC‐Davis mechanism under‐predicted with the temperature below 950 K. For the syngas mixtures with relative low H2 content, such as SH33 (Figure 4b), SH50C30 (Figure 4c) and SH26N58 (Figure 4d), however, USC‐Davis mechanism gave the results which were in good agreement with the experimental data, while the other five mechanisms showed a lack of reactivity, especially with the temperature below 950 K. Study at pressure of 1.0 MPa

Figure 2. Definition of ignition delay. Mixture composed of 2.812% H2/ 1.188% CO/2% O2/94%Ar. Conditions: P ¼ 1.0 MPa, T ¼ 1 021 K and w ¼ 1.0.

VOLUME 92, MAY 2014

At the intermediate pressure of 1.0 MPa, as shown in Figure 5, the difference of the ignition delays predicted by six kinetic mechanisms was significant. The results indicated that the Princeton‐Li mechanism exhibited the poorest reactivity, while USC‐Davis mechanism showed the fastest ignition for all four syngas mixtures. For SH70 mixture with the highest H2 content (Figure 5a), the measured ignition delays were captured by the NUIG C3 mechanism. However, Princeton‐Li and GRI 3.0 mechanisms showed the over‐predictions, while USC‐Davis mechanism gave the under‐predictions in the whole temperature range. Note that UCSD and Leeds 1.5 mechanisms presented similar predictions to the NUIG C3 mechanism and which were consistent with the experiments for only a high temperature range (T > 1 160 K). For 1 080 K < T < 1 160 K, the measured ignition delays were slightly shorter than the predicted ones, however they were well reproduced with T < 1 080 K using Leeds 1.5 mechanism. For SH33 mixture with a low H2 content (Figure 5b) and


863

Figure 4. Evaluation of six kinetic models using measured ignition delay data for SH70, SH33, SH50C30 and SH26N58 at pressure of 0.2 MPa. (a) SH70 as syngas, (b) SH33 as syngas, (c) SH50C30 as syngas, (d) SH26N58 as syngas.

SH50C30 mixture containing CO2 (Figure 5c), NUIG C3, UCSD and Leeds 1.5 mechanisms showed an acceptable prediction with T > 1 150 K, whereas only using Leeds 1.5 mechanism could yield satisfactory ignition prediction with T < 1 150. Furthermore, for SH26N58 containing N2 (Figure 5d), the results predicted by using NUIG C3 and Leeds 1.5 mechanisms exhibited a good agreement with the experimental data with T > 1 150 K, but only using USC‐ Davis could well reproduce the experimental observation, while the other five mechanisms showed generally too long ignition delays with T < 1 150 K. Study at about 2.0 MPa The comparisons between measured and calculated ignition delays for four syngas mixtures at a high pressure of 2.0 MPa were visible in Figure 6. Compared to the predictions at low and intermediate pressures, the differences of predicted performance amongst six kinetic mechanisms were clearly presented. It can be seen that five mechanisms exhibited generally failed predictions in comparing with the measured ignition delays at whole investigated temperature range, except NUIG C3. It is worth emphasising that the Leeds 1.5 mechanism gave acceptable predicted capability for SH70 (Figure 6a), SH50C30 (Figure 6c) and SH26N58 (Figure 6d) at a temperature range from 1 100 K to 1 250 K. Even so, Leeds 1.5 mechanism could not capture well the activation energy characteristics at the whole investigated temperature range, and the global

864


predicted activation energy was slightly lower than that of the measurements, especially for SH70, SH33 (Figure 6b) and SH50C30. Actually, NUIG C3, USC‐Davis and UCSD mechanisms exhibited similar simulations of global activation energy, while the GRI 3.0 mechanism showed a similar calculation to Leeds 1.5 mechanism. Effect of Pressure on Ignition Delay Figure 7 shows the measured and calculated ignition delays for all four syngas at three pressure conditions (0.2, 1.0 and 2.0 MPa). The results indicated that the effect of pressure was significant on the ignition of these syngas mixtures, and the observations were consistent with the conclusions in the previous studies on the ignition of neat H2 and H2/CO blends (Kalitan et al.[6], Zhang et al.[13], Herzler et al.[30]). It is interesting that the ignition delays of the syngas with different compositions exhibited almost the same dependence of pressure at the whole temperature range. Moreover, there were three crossing points resembling H2 ignition behaviour which were almost identical in temperature corresponding to every crossing point for four syngas mixtures. They were 1 210 K (the crossing between 1.0 and 2.0 MPa), 1 150 K (the crossing between 0.2 and 2.0 MPa), and 1 090 K (the crossing between 0.2 and 1.0 MPa), respectively. It was therefore reasonable that the effect of the pressures on the ignition delays was discussed together according to the above temperature ranges

VOLUME 92, MAY 2014


for all syngas mixtures. From fast to slow ignition delays: at T > 1 210 K, the increasing pressure produces a greater reactivity which leads to the decrease in ignition delays. At 1 150 K < T < 1 210 K, the ignition data followed the trend t1.0 MPa < t2.0 MPa < t0.2 MPa, but at 1 090 K < T < 1 150 K, the fastest ignition still presented at 1.0 MPa and the slowest ignition exhibited at 2.0 MPa. At T < 1 090 K, however, quite a reversed observation was given compared to that at a high temperature (T > 1 210 K), increasing the pressure yielded the inhibiting effects which lead to the increase in ignition delays. It is clearly known that the increase in pressure and/or decrease in the temperature will lead to a large amount of production of metastable HO2 radicals which dominate the ignition chemistry at high pressure and moderate‐low temperature conditions. Note that, at a lower temperature, the ignition was only slightly inhibited for all syngas mixtures at pressure over 1.0 MPa, but the inhibiting effect was considerably obvious at pressure below 1.0 MPa. It can be seen that SH70 having the highest H2 content showed the strongest dependence of the pressure, and SH26N58 having a high N2 content gave moderate dependence of the pressure at whole temperature range. Furthermore, the results predicted by NUIG C3 mechanism have captured well the change in the activation energy due to the reaction system transitions from high temperature kinetics to moderate‐low‐temperature reactions.

VOLUME 92, MAY 2014

SENSITIVITY ANALYSIS Normalised sensitivity analyses of the ignition delay were conducted for all four syngas mixtures at both lower (900 K) and higher (1 250 K) temperatures, and pressure of 0.2 MPa using NUIG C3 mechanism, as shown in Figure 8. For each reaction i in this mechanism, the ignition delay was calculated with reaction rates of 0.5 ki and 2.0 ki. Subsequently, the normalised sensitivity of ignition delay is defined as: S¼

tð2:0ki Þ tð0:5ki Þ 1:5tðki Þ

ð1Þ

where: ki is the rate constant of the reaction ith, t is the ignition delay value. A negative value of sensitivity implies promoting ignition, and vice versa. It is clear from Figure 8 that the chain branching reaction R1, H þ O2 ¼ O þ OH

ðR1Þ

has the largest influence on the ignition delay at the entire range of test conditions herein except that for SH26N58 (containing high fraction N2) at a high temperature (1 250 K‐case). Although this


865


important reaction R1 had a large sensitivity coefficient and influences strongly on the global reaction kinetics, other reactions, depended on temperature and composition in the syngas mixture, also considerably affect the ignition delay properties. For SH70 having a higher H2 content (Figure 8a), at a lower temperature (900 K‐case), three promoting‐ignition reactions with relatively high sensitivity to ignition delay were presented in addition to R1, they are the reactions R11, R25, and the reverse reaction R‐18.

experiments and simulations indicated that longer ignition delays were presented at a lower temperature. The increase in the ignition delay could be attributed to H‐elimination reaction R10, HO2 recombination reaction R‐14/R‐15, and the formation reaction of HO2 via R9 and R‐12,

HO2 þ H ¼ OH þ OH

OH þ O2 ¼ HO2 þ O ðR‐12Þ

ðR11Þ

CO þ HO2 ¼ CO2 þ OH ðR25Þ H2 þ HO2 ¼ H2 O2 þ H ðR‐18Þ

HO2 þ H ¼ H2 þ O2

ðR10Þ

HO2 þ HO2 ¼ H2 O2 þ O2

ðR‐14=R‐15Þ

However, for the same mixture at a higher temperature (1 250 K‐ case), the sensitivity analysis indicated that the chain branching reactions (R1 and R2) and the propagation reaction (R24) dominated the promoting‐ignition kinetics,

Note that, all these three reactions involve HO2 radical, and this implies that the chain termination R9,

O þ H2 ¼ H þ OH

H þ O2 ðþMÞ ¼ HO2 ðþMÞ

CO þ OH ¼ CO2 þ H ðR24Þ

ðR9Þ

became more important because the levels of HO2 were increased at a lower temperature via this reaction path. Both

866


ðR2Þ

However, R1 still has major effect on the ignition delay compared to the above two reactions. The analysis is almost

VOLUME 92, MAY 2014

Figure 7. Effect of pressure on ignition delays for four syngas mixtures. Open symbols: experiments, squares: 0.2 MPa, circles: 1.0 MPa, triangles: 2.0 MPa. Solid lines are simulations calculated using NUIG C3 mechanism at different pressures. (a) SH70 as syngas, (b) SH33 as syngas, (c) SH50C30 as syngas, (d) SH26N58 as syngas.

consistent with the previous study of Kalitan et al. [6] on the ignition and oxidation of lean CO/H2 fuel blends (20% CO and 80% H2) in air. For SH33 having a lower H2 content (Figure 8b), at a lower temperature (900 K‐case), overall, the sensitivity analysis presented a similar description to that for SH70 at the same temperature condition. Although most of the sensitive reactions to the ignition delay were similar, SH33 case also gave some different responses to the changing reaction rate compared to SH70. For example, for higher and lower temperatures an obvious discrepancy of the sensitivity exhibited when the R2 reaction rate of SH33 was changed, while a negligible change was presented at the same condition for SH70. Most notable was that the termination reactions R22 and R‐26 also played a considerably important role in ignition, CO þ OðþMÞ ¼ CO2 ðþMÞ ðR22Þ H þ COðþMÞ ¼ HCOðþMÞ

ðR‐26Þ

Note that, above these two reactions involve CO species. Furthermore, at higher temperature (1 250 K‐case), R24 plays an important role like R1 for the promoting‐ignition kinetics. In fact, the reactions involving CO species became more important in

VOLUME 92, MAY 2014

higher levels of CO of syngas mixture and they could considerably contribute to the ignition process. For SH50C30 containing CO2 content (Figure 8c), there were large differences compared to SH77 and SH33 that decomposition reactions R7 and R8, OHðþMÞ ¼ O þ HðþMÞ

ðR7Þ

H2 OðþMÞ ¼ H þ OHðþMÞ

ðR8Þ

became the main promoting‐ignition reaction in addition to R1 and R2 at a high temperature (1 250 K‐case). However, for SH26N58 containing N2 content (Figure 8d), at a high temperature (1 250 K‐case), R8 and R24 dominated the ignition kinetic instead of R1 and R2 due to less H2 fraction in the syngas mixture, but R1 still played a dominant role in the ignition process at a lower temperature (900 K‐case). Because almost identical dependencies on pressure were exhibited for all syngas mixtures at the experimental conditions, only SH33 was selected as the object of the research to examine the effect of pressure on ignition delay sensitivity. Thus, an analysis on SH33 was performed for both pressures of 0.2 and 2.0 MPa at 1 000 K. The results are shown in Figure 9. As expected, the chain


867

Figure 8. Normalised sensitivity with respect to individual reaction rates for four syngas mixtures at lower (900 K) and higher (1 250 K) temperature and 0.2 MPa using NUIG C3 mechanism. (a) SH70 as syngas, (b) SH33 as syngas, (c) SH50C30 as syngas, (d) SH26N58 as syngas.

branching reactions R1 and R2 were still two of the most dominant reactions for promoting ignition. It is observed clearly that increasing pressure significantly reduced the sensitivities of the reactions R1 and R2 and this led to inhibiting the global reaction. This explained the increase of ignition delays with increasing pressure in the experimental observations at corresponding temperature. Moreover, the CO2 formation reactions R24 and R25 involving in CO kinetics worked to promote ignition, but an opposite trend of the sensitivity was presented as the pressure increased due to R25 competing with R24 for CO. Moreover, R25 had a negligible sensitivity coefficient at a lower pressure (0.2 MPa‐case) while it had a more obvious negative sensitivity at a higher pressure (2.0 MPa‐ case), and this implied a little contribution to the ignition process from this channel at a higher pressure. For the conditions of this study, third body (M ¼ Ar) was the dominant collision carrier because of the high level Ar (94%) in the syngas mixture before and during the ignition process. For 2.0 MPa‐case, the ignition showed more sensitivity to the all third‐body reactions such as R16, R9 and R22 than that of 0.2 MPa‐case, H2O2 (þM) ¼ OH þ OH (þM) (R16). Figure 9. Normalised sensitivity with respect to individual reaction rates for SH33 at both lower (0.2 MPa) and higher (2.0 MPa) pressures and 1 000 K using NUIG C3 mechanism.

868


CONCLUSIONS Auto‐ignition characteristics were investigated behind the reflected shock waves for four typical syngas mixtures with 94% Ar at

VOLUME 92, MAY 2014

temperatures from 870 to 1 350 K with pressures of 0.2 MPa, 1.0 MPa and 2.0 MPa. Six kinetic mechanisms (NUIG C3, GRI 3.0, USC‐Davis, USCD, Princeton‐Li and Leeds 1.5) were used in this study to simulate and compare with the shock tube ignition delays. The main conclusions are summarised as follows: (1) Predictions with six kinetic mechanisms present significant discrepancy to the measured ignition delays, especially at higher pressure and low temperature. Agreement of the results was strongly dependent on the H2 fraction in the syngas mixtures. At the pressure of 0.2 MPa, only USC‐Davis mechanism showed a reasonable agreement with the experimental data for the syngas mixtures with relative low levels of H2 such as SH33, SH50C30 and SH26N58, while NUIG C3 and Leeds 1.5 mechanisms exhibited qualitatively well predictions for SH70 with the highest level of H2. At a pressure of 1.0 MPa, only NUIG C3 well reproduced the ignition data and the global activation energy transition for SH70 and SH50C30 mixtures. Only Leeds 1.5 mechanism showed a good agreement with the measured values for the syngas mixture with a lower level of H2, as SH33. USC‐Davis model predicted well the ignition delays for the syngas mixture with N2, as SH26N58. At a pressure of 2.0 MPa, only NUIG C3 qualitatively well reproduced the experimental data for all four syngas mixtures. (2) Effect of pressure on the auto‐ignition behaviours was considerably significant for the syngas mixtures, and it was similar to that of neat H2. The increase in pressure could promote the auto‐ignition at a high temperature but inhibit it at a low temperature due to different chemical kinetic regime which dominated the ignition process at different temperatures. All four syngas mixtures presented almost identical dependencies of the effect of pressure. (3) Sensitivity analyses indicated that the chain branching reaction H þ O2 ¼ O þ OH dominated the ignition kinetic of the syngas mixtures at both high (1 250 K‐case) and low (900 K‐case) temperatures, except for SH26N58 at 1 250 K. The H2O decomposition reaction H2O (þM) ¼ H þ OH (þM) exhibited the most sensitivity to ignition delay. For the syngas mixtures SH50C30 and SH26N58 which contain the inert gas CO2 and N2, the pressure‐dependent reactions became important at a high temperature in the syngas ignition kinetics. For the syngas mixture SH33 which contained a higher level of CO, auto‐ignition was dominated by the reaction CO þ OH ¼ CO2 þ H involving CO kinetic at a high temperature. At a temperature of 1 000 K, the inhibiting role of pressure increasing on the global reaction rate was attributed to the decrease of the sensitivity of the reactions H þ O2 ¼ O þ OH and O þ H2 ¼ H þ OH, and the increase of the sensitivity of the reactions involving HO2 chemistry. NOTATION List of Symbols IGCC H2 CO CO2 N2 O2 Ar

integrated gasification combined cycle hydrogen carbon monoxide carbon dioxide nitrogen oxygen argon

VOLUME 92, MAY 2014

ACKNOWLEDGEMENTS This work is supported by National Natural Science Foundation of China (51136005 51206132 and 51006080), National Basic Research Program (2013CB228406) and China Postdoctoral Science Foundation (2012M512001). REFERENCES [1] R. Chacartegui, M. Torres, D. Sanchez, F. Jimenez, A. Munoz, T. Sanchez, Fuel Process Technol. 2011, 92, 213. [2] P. Glarborg, Proc. Combust. Inst. 2007, 31, 77. [3] C. K. Westbrook, F. L. Dryer, Proc. Energy Combust. Sci. 1984, 10, 1. [4] W. C. Gardiner, Jr., M. McFarland, K. Morinaga, T. Takeyama, B. F. Walker, J. Phys. Chem. A 1971, 75, 1504. [5] A. M. Dean, D. C. Steiner, E. E. Wang, Combust. Flame 1978, 32, 73. [6] D. M. Kalitan, J. D. Mertens, M. W. Crofton, E. L. Petersen, J. Propul. Power 2007, 23, 1291. [7] J. Herzler, C. Naumann, Combust. Sci. Technol. 2008, 180, 2015. [8] E. L. Petersen, D. M. Kalitan, A. B. Barrett, S. C. Reehal, J. D. Mertens, D. J. Beerer, R. L. Hack, V. G. McDonell, Combust. Flame 2007, 149, 244. [9] O. Mathieu, M. M. Kopp, E. L. Petersen, Proc. Combus. Inst. 2013, 34, 3211. [10] G. Mittal, C. J. Sung, R. A. Yetter, Int. J. Chem. Kinet. 2006, 38, 516. [11] S. M. Walton, X. He, B. T. Zigler, M. S. Wooldridge, Proc. Combus. Inst. 2007, 31, 3147. [12] C. G. Fotache, Y. Tan, C. J. Sung, C. K. Law, Combust. Flame 2000, 120, 417. [13] Y. J. Zhang, Z. H. Huang, L. J. Wei, J. X. Zhang, C. K. Law, Combust. Flame 2012, 159, 918. [14] C. Morley, Gaseq. http://www.c.morley.dsl.pipex.com/ [15] J. X. Zhang, S. D. Niu, Y. J. Zhang, C. L. Tang, X. Jiang, E. J. Hu, Z. H. Huang, Combust. Flame 2012, 160, 31. [16] R. J. Kee, F. M. Rupley, J. A. Miller, CHEMKIN II. A fortran chemical kinetics package for the analysis of gas phase chemical kinetics, Sandia National Laboratory, Albuquerque, NM, Report SAND 1989, p. 89. [17] A. E. Lutz, R. J. Kee, J. A. Miller, Senkin: A Fortran program for predicting homogeneous gas phase chemical kinetics with sensitivity analysis, Sandia National Laboratories, Albuquerque, NM, Report SAND 1987, p. 87. [18] Y. J. Zhang, X. Jiang, L. J. Wei, J. X. Zhang, C. L. Tang, Z. H. Huang, Int. J. Hydrogen Energy 2012, 37, 19168. [19] G. A. Pang, D. F. Davidson, R. K. Hanson, Proc. Combust. Inst. 2009, 32, 181. [20] M. Chaos, F. L. Dryer, Int. J. Chem. Kinet. 2010, 42, 143. [21] D. F. Davidson, R. K. Hanson, Int. J. Chem. Kinet. 2004, 36, 510. [22] E. L. Petersen, R. K. Hanson, Shock Waves 2001, 10, 405. [23] H. Li, Z. Owens, D. F. Davidson, R. K. Hanson, Int. J. Chem. Kinet. 2008, 40, 189. [24] E. L. Petersen, D. M. Kalitan, S. Simmons, G. Bourque, H. J. Curran, J. M. Simmie, Proc. Combust. Inst. 2007, 31, 447.


869

[25] G. P. Smith, D. M. Golden, M. Frenklach, N. W. Moriarty, B. Eiteneer, M. Goldenberg, C. T. Bowman, R. K. Hanson, S. Song, Jr., W. C. Gardiner, GRI‐Mech 3.0. http://www.me. berkeley.edu/gri_mech. 2000. [26] S. G. Davis, A. V. Joshi, H. Wang, F. Egolfopoulos, Pro. Combust. Inst. 2005, 30, 1283. [27] P. Saxena, F. A. Williams, USCD, Chemical‐Kinetic Mechanisms for Combustion Applications. http://web.eng.ucsd. edu/mae/groups/combustion/mechanism.html 2012. [28] J. Li, Z. Zhao, A. Kazakov, M. Chaos, F. L. Dryer, Jr., J. J. Scire, Int. J. Chem. Kinet. 2007, 39, 109. [29] K. J. Hughes, T. Turányi, A. R. Clague, M. J. Pilling, Int. J. Chem. Kinet. 2001, 33, 513. [30] J. Herzler, C. Naumann, Proc. Combust. Inst. 2009, 32, 213.

Manuscript received April 18, 2013; revised manuscript received May 29, 2013; accepted for publication June 02, 2013.

870


VOLUME 92, MAY 2014