Subscriber access provided by Northern Illinois University
Article
Computational Study of NOx Formations at Conditions Relevant to Gas Turbine Operation Part II: NOx in High Hydrogen Content Fuel Combustion at Elevated Pressure Sheikh F. Ahmed, Jeffrey Santner, Frederick L. Dryer, Bihter Padak, and Tanvir I. Farouk Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b00421 • Publication Date (Web): 12 Aug 2016 Downloaded from http://pubs.acs.org on August 14, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
Computational Study of NOx Formations at Conditions Relevant to Gas Turbine Operation Part II: NOx in High Hydrogen Content Fuel Combustion at Elevated Pressure Sheikh F. Ahmed1, Jeffrey Santner2, Frederick L. Dryer2, Bihter Padak1, Tanvir I. Farouk1,* 1
Department of Mechanical Engineering, University of South Carolina, Columbia, SC 29208, USA 2
Department of Mechanical and Aerospace Engineering, Princeton University, Princeton, NJ 08544, USA *
Corresponding Author Email:
[email protected] ABSTRACT: Part 1 of this two part series presented a computational study of NOx formation during methane, ethylene combustion - representative of small fuel fragments present in natural gas and chemical processing. The influence of fuel chemistry, reaction temperature history and inert dilution was examined using popular models present in the literature. The present work extends the study to hydrogen-rich conditions to remove the fuel variability dependency of NOx and identify possible inconsistencies in predicting NOx during high hydrogen content fuel combustion. A comprehensive chemical kinetic model is proposed consisting of CO/H2/NOx oxidation with the full implementation of thermal, N2O and NNH paths of NOx evolution. Predictions from the model are compared against multiple experimental datasets over a wide range of venues and operating conditions. The experimental venues include shock tube, plug 1 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 33
flow reactor, and stirred reactor experiments that cover pressures from 1 to 100 bar and equivalence ratios from 0.5 to 1.5. In general, the overall model predictions are in good agreement with global combustion targets, such as ignition delay time, as well as with more detailed measurements from flow reactors and stirred reactors. Simulations are conducted for a wide range of reacting mixtures (H2/O2/N2, CO/H2/O2 and CO/H2O/O2/N2) with initial NO and NO2 perturbations to consider exhaust gas recirculation (EGR) conditions. KEYWORDS: Synthesis Gas, NOx, Combustion Targets, EGR 1. INTRODUCTION Part 1 of this two part series primarily investigated the differences in predictions from various hydrocarbon-NOx models for methane and ethylene. The authors found significant differences not only in available hydrocarbon models in terms of their description of CH chemistry, but also in available nitrogen chemistry in their descriptions of Zel’dovich and Fenimore mechanisms. Since the hydrocarbon portion of any available kinetic mechanism affects NO production through the Fenimore routes, NOx prediction performances of existing HC-NOx models were found to differ significantly. The question remains whether NOx formation for hydrogen or synthesis gas (syngas, i.e. mixtures of H2/CO) where no Fenimore routes are present can be predicted with sufficient fidelity. This article focuses on assessing the performance of existing H2/CO – NOx models in predicting NOx formation and evolution under gas turbine-related conditions. A kinetic model is presented and its predictions are tested against data from a large number of fundamental experiments and differing venues. The present work is principally driven by the practical relevance of predicting NOx emissions for high hydrogen content (HHC) fuel and syngas combustion to meet stricter emissions
2 ACS Paragon Plus Environment
Page 3 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
regulations forthcoming from the Environmental Protection Agency (EPA) for gas turbine power generation. Syngas and HHC fuel combustion in gas turbines have been driven entirely by interest in gasifying a variety of feed stocks such as coal, biomass, and refuse to produce electric power at highest possible efficiency, with the prospects of reducing emissions of particulates and, in the longer term, capturing carbon dioxide (carbon sequestration). Gas turbine applications are especially suited for operating on natural gas 1, as well as gasification products from oxidative pyrolysis of coal
2
and other fossil resources, renewable biomass 3, and municipal
refuse to produce synthesis gas/syngas and even pure hydrogen 4. Hydrogen may be a long term replacement for carbon-containing fuels. However, the high adiabatic flame temperature and explosively fast chemical kinetics of pure hydrogen-air combustion lead to a need for significant exhaust gas recirculation requirements, unsteady combustion sensitivity, and combustor material degradation 5. The use of syngas rather than hydrogen continues to produce some emissions of carbon dioxide, but intrinsically obviates some of the economic issues associated with producing pure hydrogen, particularly if it can be implemented for a wide range of hydrogen/CO ratios 6. All of these options lead to potential means to reduce net carbon cycle emissions through the use of biomass and carbon sequestration principles. On the other hand, all of these solutions continue to evolve other air pollutants, notably NOx emissions. As a result, continuing research is directed toward achieving higher operating temperatures (higher thermal efficiencies) at reduced NOx and other air pollutant emissions. A major concern is that even when global combustion properties are well predicted, details such as speciation predictions vary significantly among the different fuel + NOx kinetic model constructs available in the literature. NOx speciation is of extreme relevance due to the strict emission standards enforced by the EPA 7. While the quantitative accuracy of a model can be 3 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 33
improved by optimization against experimental data, the fidelity of the model for predicting the trends of the speciated kinetic interactions with engineering design is essential. Figure 1 shows a comparison of ignition delay time and NOx speciation prediction using the Aramco model 8, integrated with the NOx subset of four prominent NOx mechanisms. Even though some similarities exists among the model predictions, there are significant mechanistic differences as well as differences in relative importance of various reaction pathway contributions to the global predictions, as highlighted in part 1. The NOx sub-mechanism of the CRECK model proposed
(a)
(b)
Figure 1. Numerical simulations of (a) ignition delay time (b) NOx evolution profiles for Aramco model 8, merged with different NOx subsets 9-12. Data for (a) and (b) are taken from the shock tube experiments of Mathieu et al. 13 and isothermal flow reactor experiments of Rasmussen et al. 9 respectively. The color bands of the ignition delay plot represent the variation in ignition delay associated with initial H atom impurities of 350 ppb.
by Ranzi et al. predicts the mutual sensitization of NO and hydrocarbons by their interactions during the oxidation of hydrocarbons at low temperatures
12
or during the high temperature
reburning process and a consequent promotion of NO/NO2 transformation 14. The kinetic model 4 ACS Paragon Plus Environment
Page 5 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
proposed by Dagaut et al.
10
takes into account the mutual sensitization of the oxidation of
methane and NO under jet stirred reactor and flow reactor operating conditions. Rasmussen et al. 9 proposed a detailed mechanism for the CO/H2/O2/NOx system that encompassed two diverse regimes of chemistry, low temperature atmospheric chemistry and high temperature combustion chemistry. The model of Konnov et al.
15
is capable of simulating the oxidation of hydrogen,
carbon monoxide, formaldehyde, methanol, methane, C2-C3 hydrocarbon species and their oxygenated derivatives and also includes C/H/N/O reactions for in-flame NOx formation and reburning. Konnov published a revised mechanism 11 in 2009 with additional implementation of available kinetic pathways of the prompt NO route via NCN. The simulated ignition delay times are for systems where controlled trace NO2 was introduced in H2/O2 experiments. The different NOx models predict the ignition delay time reasonably well with the Dagaut et al.
10
and the CRECK model
12
predicting the longest and shortest ignition
delay times, respectively. Even then, the variance of the predictions using the different models is not large and the variations are further minimized by the addition of 350 ppb of initial H-atom as impurities, which could be a result of initial H-atom uncertainty in the system (details discussed in the following section). However, when specific species evolution (i.e. NO, NO2) experiments are simulated, significant inconsistencies among the predictions are observed. It is apparent that global combustion target of ignition delay is insufficient to provide the necessary constraining conditions for assessment and model development to predict NOx kinetic effects. The recent paper of Watson et al. 16 also highlights the importance of speciation data for the development of NOx kinetic models capable of predicting both global and speciated experimental behaviors with improved fidelity.
5 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 33
With this purpose in mind, the current work assembles and tests a comprehensive chemical kinetic reaction mechanism to describe NOx kinetics in CO/H2 oxidation. The mechanism is validated over a wide range of conditions and multiple experimental data sets including shock tube, plug flow reactor, and stirred reactor measurements that cover pressures from 1 to 100 bar and equivalence ratios from 0.5 to 1.5. In general, the overall model predictions are in good agreement with global combustion targets (shock tube ignition delay) as well as against detailed targets that include plug flow reactor species evolution, and stirred reactor measurements. A wide range of reacting mixtures with initial NO and NO2 perturbations are used in validating the present model, which not only take into account the coupled interactions of the fuel oxidation and NOx interactions, but encompass Exhaust Gas Recirculation (EGR) conditions. 2. DETAILED MECHANISM FORMULATION APPROACH The proposed CO/H2/NOx model with a limited consideration of small hydrocarbon species consists of several sub-mechanisms: a C0-C1 sub-mechanism, a NOx sub-mechanism, and a H/N/O sub-mechanism. The hydrocarbon portion of the present mechanism is adopted from Aramco Mech 8. The reaction mechanism reported by Konnov
11
served as the base set for NOx
kinetics with additional parameter revisions and elementary reaction inclusions. The NOx subset of the proposed model includes updated NxHy reaction paths as well as species such as HNO2 and HONO2 that have been found to significantly contribute to NOx production. Thermochemical parameters in this model are adopted from the Burcat database
17
. The details of each
sub-mechanism are presented in the following sections. 2.1. C0-C1 Sub-mechanism. The C0-C1 sub-mechanism consists of reactions involving the H2/O2 system, the CO/CO2 system and the C1 species. The present C0-C1 sub-mechanism is developed by the integration of the Burke C0
18
model and the C1 species and associated 6
ACS Paragon Plus Environment
Page 7 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
reactions of the Aramco 8 model. In order to compare the performance of the hydrogen subset of the Burke et al.
18
model and that of Kéromnès et al.19, the fuel and NOx species evolution for
three different operating pressures for a CO/H2/NOx system were simulated. Although similar performances were observed for both Burke et al.
18
and Kéromnès et al.19 hydrogen subsets at
low pressure condition (20 bar), the variation between the two prediction increases with increasing pressure showing further deviation of the Kéromnès
19
model predictions from
experimental data. The hydrogen model of Burke was therefore chosen as it showed better predictions at higher operating pressures. Details with representative comparisons of predictions and experiments are shown in the supporting information. The exothermic oxidation of carbon monoxide by hydroxyl radicals plays a significant role in combustion as the major pathway of CO-CO2 conversion and source of heat release. In general, CO oxidation by OH radicals
20
can
proceed through two paths, one forming HOCO, and the other forming CO2 and atomic hydrogen: CO + OH = HOCO
(R1a)
CO + OH = CO2+ H
(R1b)
Recent literature 20, 21 concludes that HOCO formation through reaction (R1a) is unimportant at pressures and temperatures relevant to combustion energy conversion processes; Rasmussen et al. 9 note as well that decomposition of HOCO to CO2 is very rapid. We also found the fuel and NOx speciation predictions to be insensitive to the inclusion of HOCO chemistry in the overall mechanism. Thus we neglect HOCO chemistry in the present modeling work. 2.2. NxHOy Sub-mechanism. The NOx kinetic components of the proposed model are developed based on a critical review of existing NOx formation, and NO-NO2 inter-conversion
7 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 33
sub-models available in the literature 9-11, 22 with the implementation of NOx evolution pathways which include thermal NOx 23, N2O, and NNH paths. Even in the absence of Fenimore prompt NOx an additional source of the prompt mechanism is due to the super equilibrium concentrations of O atoms and OH radicals in the flame zone which accelerates the Zel’dovich mechanism, which is inherently considered in the model. In lean and slightly rich flames the partial equilibrium assumption of O atoms and OH radicals in
1 1 1 O 2 → O, O 2 + H 2 → OH cease 2 2 2
to exist. Near the flame zone the ratio of the maximum concentration to the equilibrium concentration of both species can be different by an order of magnitude or more resulting in higher NOx formation rates. The super-equilibrium O atom concentration is nearly absent in higher hydrocarbons flames due to the presence of the reacting hydrocarbon fragments. The interconversion reactions between NO2 and NO (also referred to as the NOx recycling reactions) play a significant role in the NOx-related kinetics. In addition to the reactions NO+HO2 = NO2+OH and NO2+H = NO+OH,
interconversion also proceeds through
intermediate formation of HONO, HNO2 and HONO2 9, i.e. HxNOy reaction pathways. The existence of HONO, HNO2 and HONO2 intermediates in combustion systems has been verified experimentally
24, 25
. HNO2, a thermodynamically less stable isomer of nitrous acid (HONO),
appears to have a notable influence on predictions 9. The formation of HONO and HONO2 as intermediates in the NOx recycling process also play prominent roles in the atmospheric chemistry HOx cycle in terms of predicted consumption and production of OH radicals 9. Reactions R11 (Table 1) and R22 (Table 1) play the most significant role in this chemistry. Our model therefore includes the HxNOy reaction pathways with updates in the reaction rates 9 (Table 1). The detailed Konnov-mechanism 11 did not include the HNO2 and HONO2 reaction pathways shown in Table 1. 8 ACS Paragon Plus Environment
Page 9 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
Table 1. Reactions and Forward Rate Parameters for the HxNOy Reaction Pathways Reactions
A
b
E (cal/mole)
Reference
1.
NO2 + H2 = HONO + H
1.30E+04
2.76
29770
9
2.
NO2 + HO2 = HONO + O2
1.91E+00
3.32
3044
9
3.
HNO + NO2 = HONO + NO
4.42E+04
2.64
4040
9
4.
HONO + O = NO2 + OH
1.20E+13
0
5960
9
5.
HONO + OH = NO2 + H2O
1.70E+12
0
-520
9
6.
HONO + HONO = NO + NO2 + H2O
3.49E-01
3.64
12140
9
7.
H2NO + NO2 = HONO + HNO
6.00E+11
0
2000
26
8.
HNOH + NO2 = HONO + HNO
6.00E+11
0
2000
26
9.
HONO + H = HNO + OH
5.64E+10
0.86
5000
26
10.
HONO + H = NO + H2O
8.12E+06
1.89
3850
26
11.
NO + OH(+M) = HONO(+M)
1.10E+14
-0.3
0
9
Low-pressure limit:
3.39E+23
-2.5
0
Troe parameters: 0.75 1E-30 1E+30 1E+30 12.
NO2 + CH2O = HONO + HCO
1.42E-07
5.64
9220
9
13.
NO2 + HCO = HONO + CO
4.95E+12
0
0
9
14.
HONO(+M) = HNO2(+M)
2.50E+14
0
32300
9
Low-pressure limit:
3.10E+18
0
31500
Troe parameters: 1.149 1E-30 3125 1E+30 15.
NO2 + H2 = HNO2 + H
2.43E+00
3.73
32400
9
16.
NO2 + HO2 = HNO2 + O2
1.85E+01
3.26
4983
9
17.
NO2 + OH = HNO2 + O
1.70E+08
1.5
2000
9
18.
NO2 + H2O = HNO2 + OH
4.00E+13
0
0
9
19.
NO2 + CH2O = HNO2 + HCO
1.07E-01
4.22
19850
9
20.
HONO + NO2 = HONO2 + NO
2.00E+11
0
32700
9
21.
HONO + OH = HONO2 + H
3.82E+05
2.3
6976
9
22.
NO2 + OH(+M) = HONO2(+M)
3.00E+13
0
0
9
Low-pressure limit:
2.94E+25
-3
0
Troe parameters: 0.4 1E-30 1e+30 1E+30 23.
HONO2 + H = H2 + NO3
5.56E+08
1.5
16400
9
24.
HONO2 + H = H2O + NO2
6.08E+01
3.3
6285
9
25.
HONO2 + OH = H2O + NO3
1.03E+10
0
-1240
9
9 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 33
2.3. NxHyO Sub-mechanism. The NxHyO reaction pathways typically found in ammonia oxidation models may also participate in the interconversion of NO/NO2 as well as production of N2 in flames, and we incorporate also these reaction pathways from the recent ammonia oxidation model of Skreiberg et al. 27. Updated rate constants of several reactions involving NH2, HNO and NH2OH species are proposed based on the detailed NH3-oxidation and thermal DeNOx model of Klippenstein et al. 26. The updated reactions of this sub-mechanism are shown in Table 2. Table 2. Reactions and Forward Rate Parameters for Selected NxHyO Reaction Pathways Reactions
A
b
E (cal/mol)
Reference
1.
NH2 + NO = N2 + H2O
1.300E+16
-1.250
0
26
2.
NH2 + NO = N2 + H2O
-3.100E+13
-0.480
1180
26
3.
NH2 + NO = NNH + OH
3.100E+13
-0.480
1180
26
4.
NH2 + NO2 = N2O + H2O
3.100E+14
-0.770
242
26
5.
NH2 + NO2 = H2NO + NO
1.300E+15
-0.770
242
26
6.
HNO + NO = N2O + OH
1.200E-04
4.330
25080
26
7.
NH2OH + OH = HNOH + H2O
1.500E+04
2.610
-3537
26
8.
NH + NO2 = HNO + NO
5.900E+12
0.0
0
26
9.
NH2OH + H = HNOH + H2
4.800E+08
1.500
6249
26
10.
NH2OH + H = H2NO + H2
2.400E+08
1.500
5067
26
11.
NH2OH + O = HNOH + OH
3.300E+08
1.500
3865
26
12.
NH2OH + O = H2NO + OH
1.700E+08
1.500
3010
26
13.
NH2OH + OH = H2NO + H2O
1.500E+05
2.280
-1296
26
14.
NH2OH + HO2 = HNOH + H2O2
2.900E+04
2.690
9557
26
15.
NH2OH + HO2 = H2NO + H2O2
1.400E+04
2.690
6418
26
3. MODEL PERFORMANCE In order to assess the assembled model, a wide range of predictions of experimental data from shock tube, plug flow reactor, and stirred reactor configurations are considered. The
10 ACS Paragon Plus Environment
Page 11 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
experimental conditions cover a pressure range of 1 to 100 bar and an equivalence ratio range of 0.5 to 1.5. The Chemkin-II package 28 is used for all the simulations of this study. The shock tube experimental conditions are simulated using the SENKIN code 29 with constant volume and zerodimensional approximations. In order to simulate the experiments of perfectly stirred reactor, the PSR code
30
is used in the present study. The SENKIN code
29
is also used to simulate the
adiabatic, zero-dimensional plug flow reactor experiments. 3.1. Ignition Delay Times. The performance of the model is first compared against global combustion targets; ignition delay times over a wide range of pressure and NO2 loading. Shock tube ignition delay measurements of Mathieu et al.
13
are utilized. In their shock tube
experiments, Mathieu et al. 13 added increasing amounts of NO2 (100, 400 and 1600 ppm) to an H2/O2/Ar mixture, observing a strong dependence of the ignition delay on the initial NO2 concentration. Model predictions are compared with data for three different initial pressures (1.66, 13.0, 33.6 atm) appear in Fig. 2. At different initial pressure, the effects of initial NO2 loading in the mixture show significantly different behavior. At the lowest pressure, small additions of NO2 (100 ppm) cause no measurable change in ignition delay time, while at intermediate addition (400 ppm), an increase in ignition delay is observed for initial reaction temperatures below 1285 K. At the highest addition (1600 ppm), a significant increase in ignition delay is observed for reaction temperatures below approximately 1540 K. At higher reaction pressures (13.0 and 33.6 atm), the non-monotonic dependence of ignition delay on initial NO2 addition is more emphasized. At 13.0 atm pressure, the ignition delay decreases with NO2 addition for reaction temperatures below 1175 K, but with a non-monotomic dependence on the amount of added NO2.
11 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
(a)
(b)
Page 12 of 33
(c)
12 ACS Paragon Plus Environment
Page 13 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Energy & Fuels
(d)
Figure 2. Effect of initial NO2 concentration on τign for H2/O2 mixtures at (a) 1.66 atm, (b) 13.0 atm (c) 33.6 atm pressure (Lines represent numerical simulations and symbols represent measurements behind reflected shock waves 13. The color bands represent the variation in ignition delay associated with initial H atom impurities of 350 ppb) and (d) Temporal evolution of OH concentration as a function of H atom impurities at P = 13 atm, T = 1100 K, φ = 0.5, for 0 ppm, 100 ppm, 400 ppm and 1600 ppm of NO2 doping.
13 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 33
The addition of 100 ppm of NO2 removes the sharp increase in ignition delay noted for the pure H2-O2 case with decreasing temperatures below 1185 K, but slightly increases the ignition delay for higher reaction temperatures. At each reaction temperature below 1185 K, The magnitude of decrease in the ignition delay in comparison to the pure H2-O2 case, increases with increasing NO2 addition from 100 to 400 ppm, but eventually decreases at higher NO2 addition (1600 ppm). Above 1175 K the addition of NO2, the ignition delay increases with increasing NO2 addition for all cases, in comparison to the pure H2-O2 case. Similar non-monotonic behavior is observed at the highest investigated pressure (33.6 atm) but with a slightly higher transition temperature (1275 K versus 1175 K). In other work published since Mathieu et al.
13
, Urzay et al.
31
investigated the effects of
residual impurities in the shock tube ignition delay experiments conducted at Stanford on H2 oxidation. Urzay et al.
31
embodied these effects in comparing predictions with experiments by
assuming the presence of small amounts of H atoms in the initial reactant composition. The addition of small H atom levels were used to adjust the reaction times at which water production profiles were predicted so as to align them with experimental measurements
32
. However, the
addition also affects comparisons of predictions with experimental ignition delay data
33
. The
effect of H atom addition to pure and NO2 doped H2-O2 reaction predictions at the three different pressures are presented as color bands in Figure 2. For the pure H2/O2 cases, the addition of as little as 350 ppb of H atom to the initial mixtures reduces the predicted ignition delay time significantly for all the pressures, yielding a substantially improved agreement between the experimental data and predictions. Our analysis shows that initial H atom seeding levels to 200 ppb lead to significant reductions in predicted ignition delay times (Figure S1 in Supplementary Materials), with a much reduced relative influence for amounts above ~ 350 ppb. The inflection 14 ACS Paragon Plus Environment
Page 15 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
in behavior at ~ 350 ppb, led to the selection of this level of perturbation in demonstrating the bandwidth of H impurity seeding. Despite the decrease in the ignition delay time due to initial H atom seeding, the temporal evolution of OH concentrations shown in Figure 2d remain unaltered by H atom addition and are only “time shifted”. For the NO2 doped cases, the effect of adding initial trace H-atom concentrations decreases with increasing initial NO2 concentration and the effects become discernible at the maximum NO2 loading. This result is also apparent in the color bandwidths in Figure 2 a, b and c. Furthermore, the NOx kinetic model predictions of reaction gradients after the induction period are not affected by any assumed H atom seeding. With increasing NO2 concentrations, H atoms are more readily consumed by NO2 (NO2 + H = NO + OH), reducing the impact of other reactions which limit building the active radical pool (see discussions below). It is also observed in Fig. 1 that the addition of H atom impurity has the most prominent effect on the CRECK
model
14
predictions and with reduced impact
using other models. In particular, the least sensitivity occurs for predictions based upon the Dagaut NOx model
10
. A flux analysis at the corresponding experimental conditions shows a
significant HONO-NO2 route in the CRECK NOx model 14 by the reaction- HONO + H = NO2 + H2, with HONO coming from the initial NO2 doping through the paths- NO2 + H = NO + OH, NO + OH(+M) = HONO(+M). The reaction rate constant of this HONO-NO2 reaction for CRECK NOx model model
10
14
is approximately two orders of magnitude larger than in Dagaut NOx
. The impurity effects demonstrated in predictions using the Dagaut model
10
are
recovered using the CRECK model 14 but with rate parameters for HONO + H = NO2 + H2 taken from Dagaut et al. 10 In order to identify the dominant reactions that dictate the ignition delay observations, first order logarithmic sensitivity analyses at every pressure and NO2-perturbation were performed for 15 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 33
conditions within the high and low-temperature regions of Figure 2. Each reaction in the mechanism is increased and decreased by a factor of 2 to calculate the ignition delay times and
τ log 1 τ2 the sensitivity coefficient σ; σ = , where the symbols τ1 and τ2 represents the 2.0 log 0.5 calculated ignition delay times with the reaction rates increased and decreased, respectively. An example of the sensitivity analyses for low (1100 K) and high (1220 K) temperature zones at 13.0 atm for different NO2-perturbations is shown in Fig. 3 and 4 respectively.
(a)
(b)
(c)
(d)
Figure 3. First order ignition sensitivity analysis at 13.0 atm and 1100 K for (a) neat mixture of H2/O2, (b) with 100, (c) 400 and, (d)1600 ppm of NO2. The directions of all the reactions in
16 ACS Paragon Plus Environment
Page 17 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
these sensitivity charts are forward in nature.
At the high temperature condition, the ignition delay becomes more sensitive to the chain branching reaction- O + H2 = H + OH (R2), whereas the chain propagation reaction H2O2 + H = H2 + HO2 (R3) becomes more sensitive in case of lower temperature, which causes the increase in ignition delay time at lower temperatures. With an addition of small amounts of NO2 (100 ppm) to the mixture, a decrease in ignition delay time is observed for both temperature zones. With 100 ppm of NO2, the NO-NO2 inter-conversion cycle causes the formation of highly reactive OH, H and HONO radicals through the reactions: NO2 + H = NO + OH (R4), NO + HO2 = NO2 + OH (R5) and NO2 + H2 = HONO + H (R1, Table 1). Flux analysis of NO and NO2 at 1100 K and 13.0 atm shows that the majority of NO2 is consumed through reaction (R4) producing NO and OH radicals. Again, most of the NO is recycled to NO2 through (R5). The formation of OH radicals through all the above reactions causes a significant decrease in the overall ignition delay of the mixture. Fig. 2b shows that the ignition delay time decreases in the low temperature cases with the addition of 400 ppm of NO2. With the increase in initial NO2 concentration, the chain branching reaction (R2) becomes more sensitive, and the reaction NO2 + H2 = HONO + H (R1, Table 1) leads to increases H radical formation rate. The higher H concentration consequently increases the OH concentration through HO2 + H = 2OH (R6), which further accentuates the reactivity when the NO2 is increased from 100 to 400 ppm. A significant change in the most sensitive reactions and consequently, a much higher ignition delay is observed for the overall temperature range in Fig. 2b, when the initial NO2 concentration further increases from 400 to 1600 ppm. At lower temperature, the sensitivity analyses shows that the sensitivity of the propagation reaction
17 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 33
OH + H2 = H + H2O (R7) increases for the 1600 ppm case in comparison to the 400 ppm case. Flux analysis at lower temperatures identifies that the NO-NO2 inter-conversion by the reactions (R5) and (R4) is dominated by (R4) at lower temperatures, which eventually increases the sensitivity of H atom producing reaction (R7). Additionally, a number of inhibiting reactions, such as, H + NO (+M) = HNO (+M) (R9), HO2 + OH = H2O + O2 (R10) become significant at higher NO2 loading, i.e. the 1600 ppm case. A similar sensitivity analysis was conducted earlier by Mathieu et al.
13
. However the current analysis identifies a number of additional sensitive
reactions (H+HO2=H2+O2, HO2+OH=H2O+O2 for 100 ppm NO2, HO2+OH=H2O+O2, H+HO2=H2+O2 for 400 ppm NO2, and H+NO(+M)=HNO(+M), NO+HO2=NO2+OH for 1600 ppm NO2), which do not appear in the analyses of Mathieu et al.
13
. In addition, a different
sensitivity rankings are observed here.
(a)
(b)
(c)
(d)
Figure 4. First order ignition sensitivity analysis at 13.0 atm and 1220 K for (a) neat mixture of
18 ACS Paragon Plus Environment
Page 19 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
H2/O2, (b) with 100, (c) 400 and, (d) 1600 ppm of NO2. The directions of all the reactions in these sensitivity charts are forward in nature. 3.2. Plug Flow Reactor Experiments under Dilute Conditions. The model performance is further compared against detailed plug flow reactor species evolution with time and temperature for a wide range of reacting mixtures. As a first step, simulations are conducted for H2/O2/N2 mixture with initial NO2 perturbations and compared to the adiabatic flow reactor experiments of Mueller et al. 34, shown in Fig. 5a. It can be seen that H2 is consumed to produce H2O and NO2 is converted to NO to increase the amounts of NO over the course of the residence time. Improvement in the NO-NO2 inter-conversion, relative to the experimental data is observed with the NxHyO reaction rate updates highlighted in Table 2. A flux analysis for the NO-NO2 recycling process is shown in Fig 5b. The conversion of NO to NO2 at the investigated conditions occurs through the formation of intermediate HNO3 and a consequent effect on the yield of OH radicals – NO + HO2 (+M) = HNO3 (+M) (R11), HNO3(+M) = NO2 + OH(+M) (R12). NO2 can be converted to NO through three different reaction paths- (i) directly by (R4), (ii) through intermediate formation of HONO by the reaction of HO2 radical (R2, Table 1), followed by the formation of NO and OH by reaction (R11), Table 1, (iii) through intermediate formation and isomerization of HNO2 by the reactions (R16), (R14) of Table 1. HONO is then converted to NO by the pressure dependent reactions noted as (ii) above.
19 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 33
(a)
(b)
Figure 5. (a) Time histories of species concentrations for H2/O2/N2 mixture, perturbed with 85 ppm of NO2 at 10.0 atm and Tin = 780 K. Symbols represent experimental data from Mueller et al. 34. The initial mole fractions for the H2/O2/NO2 mixture are 0.98%/0.5%/85 ppm, balance N2. (b) Major reaction pathways of NO-NO2 conversion. The '+' and '-' symbols in the flux analysis represent respectively, the formation and consumption of the species associated with the symbol. The different colors are used to show the reaction paths of different species. The performance of the present model in simulating experiments with a wide range of pressure, temperature, and NO concentrations on CO/H2O/NO oxidation is also investigated in this study. The moist CO oxidation experiments of Mueller et al.
34
are considered in this case.
Figure 6 shows the temporal evolution of CO, NO, and NO2 at 950 K and at pressures ranging from 1.2 to 10.0 atm. The model accurately predicts the experiments for both fuel oxidation and NO-NO2 conversion processes. An increase in pressure is found to inhibit fuel oxidation and to promote NO-NO2 conversion. With an increase in pressure, the three body recombination reactions (e.g. H + O2(+M) = HO2(+M), NO + O(+M) = NO2(+M), CO + O(+M) = CO2(+M) etc.) become dominant over the branching reactions (e.g. H + O2 = O + OH, H2O + O = OH + OH etc.), that lead to the inhibiting effect of pressure on fuel oxidation. The higher HO2 20 ACS Paragon Plus Environment
Page 21 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
formation at higher pressure leads to a faster NO-NO2 recycling process by the consumption of HO2 by NO to form NO2 (HO2 + NO = NO2 + OH). Figure 6 illustrates that under these conditions a complete NO-NO2 inter-conversion takes place without any formation of molecular nitrogen.
Figure 6. Effect of pressure on the reaction profiles for CO/H2O/O2/NO/N2 mixture at Tin = 950 K. Symbols represent experimental data from Mueller et al. 34 and lines represent model predictions. The initial mole fractions for the three different pressure conditions are CO/H2O/O2/NO = 0.53%/0.49%/0.76%/107 ppm, balance N2.
Experimental data of CO/H2O/NO oxidation
34
for a fixed pressure and different initial
reaction temperature together with model predictions are presented in Figure 7. The current model is found to predict the experimental trends with reasonable accuracy. The predicted extent of CO consumption with different initial NO concentrations is compared in Fig. 8 for a series of experiments
34
at 3.0 atm, 950 K and with initial NO mole fraction of 54 to 508 ppm. The
experimental trend of the strongest CO consumption at intermediate NO levels is well captured by the model. The reactivity measurements of Mueller et al.
34
for fuel-lean CO/H2O/O2/NO/N2
mixtures at various pressures show a strong dependence of CO2 production on the initial NO 21 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 33
concentration. Figure 8 also shows the simulated NO2 profiles with variable initial NO concentration. In accordance with the NO-NO2 conversion process, increase in initial NO in the mixture increases the NO2 concentrations. Additionally, with increasing the NO concentration the initial ramping to NO2 conversion is decreased significantly. Similar experiments investigating the impact of NO and H2O perturbations on the oxidation of CO are available in the literature
35
and the proposed model is validated based on those experiments as well, showing a
reasonable agreement with the experimental measurements. The comparisons are presented as supporting information. Model validation is also performed based on the isothermal tubular reactor experiments of Arai et al.
36
to predict thermal NOx formation in a binary N2-O2 system
as a function of temperature, which shows reasonable agreement with the data; the major variation is observed at the highest temperature and for O2 fraction more than 30%. Similar variance was observed by Abian et al. 37 and the disagreement at high temperature was attributed to be an experimental artifact. Such comparisons are also illustrated in supporting information.
Figure 7. Effect of initial temperature on the reaction profiles for CO/H2O/O2/NO/N2 mixture at P = 10 atm and xNO = 41.0 ppm. Symbols represent experimental data from Mueller et al. 34 and
22 ACS Paragon Plus Environment
Page 23 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
solid lines represent model predictions. The initial mole fractions for the three different temperatures are CO/H2O/O2/NO = 0.54%/0.50%/0.75%/71 ppm, balance N2.
The model was also employed to simulate the experimental measurements of Rasmussen et al. 9
where species concentration for a CO/H2/NOx system was measured in a flow tube at various
initial temperatures and pressures under a prescribed temperature distribution with isothermal conditions being maintained in a designated test section. Unlike the flow reactor measurements mentioned above
34
where species temporal evolution was measured for a constant residence
time, in the aforementioned experiments the overall residence time changes with each specified reaction temperature and pressure at each measured point in this case. Figure 9 presents the experimental data and the model prediction for all the different pressures. Five simulations using the reported experimental temperature profiles with initial ramp-up at the inlet, isothermal reaction zone, and ramp-down at the outlet of the reactor tube 9 were performed. The fuel, CO, is being oxidized to CO2 and follows the experimental trends for all three experiments. The onset of CO consumption (the initiation temperature) decreases from 800 K to 700 K when the pressure increases from 20 to 100 bar. The pressure-dependence of the CO initiation temperature decreases with increasing pressure and the most significant decrease (75 K) occurs when the pressure increases from 20 to 50 bar. A decrease in initiation temperature of only 25 K is observed when the pressure increases from 50 to 100 bar. Isothermal simulations reasonably predict the experimental data for the lowest pressure case of 20 bar.
23 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 33
Figure 8. Effect of initial NO mole fractions on the reaction profiles for CO/H2O/O2/NO/N2 mixture at P = 3.0 atm and Tin = 950 K. Symbols represent experimental data from Mueller et al. 34 and solid lines represent model predictions. The initial mole fractions for different NO concentrations are CO/H2O/O2 = 0.50%/0.48%/0.74%, balance N2.
Simulations using the reported experimental temperature profile and the isothermal simulations deviate significantly at higher pressures (50 and 100 bar) and at reaction temperatures higher than 750 K. A substantial conversion of NO to NO2 outside the isothermal zone at high pressure and temperature is observed, which is discussed below in flux analysis. Despite a temperature distribution existed in the entire flow reactor length (i.e. initial ramp up, isothermal zone, ramp down) the temperature in the species evolution corresponds to the temperature in the isothermal zone only. Additionally the NOx evolution at high pressure and temperature cases can only be simulated accurately if the complete experimental temperature profiles are considered in the simulation.
24 ACS Paragon Plus Environment
Page 25 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
(a)
(b)
(c) 9
Figure 9. Experimental data and model predictions of CO/H2/NOx oxidation for 0.063 equivalence ratio at (a) 20 bar, (b) 50 bar and (c) 100 bar. The closed symbols represent experimental data. The solid lines and the open symbols represent simulations with
25 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 33
isothermal assumption and with the complete experimental temperature profile respectively. The initial mole fractions for 20, 50 and 100 bar conditions are respectivelyCO/H2/O2/NO/NO2 = 0.0518%/0.0453%/1.53%/113 ppm/36 ppm, 0.0518%/0.0446%/1.54%/26 ppm/125 ppm, 0.0502%/0.044%/1.48%/ 6 ppm/ 145 ppm, balance N2. Flux analyses at different pressures for the aforementioned experiments are performed to explain the NO-NO2 inter-conversion paths in the present model. Figure 10 shows the NO-NO2 recycling process at 20 and 50 bar. At 20 bar, the recycling can take place (i) directly by the reactions NO + HO2 = NO2 + OH (R5) and NO2 + H = NO + OH (R4) (ii) by the intermediate formation of HONO or (iii) by the formation and isomerization of HNO2. However, at 50 bar, two additional NO-NO2 recycling pathways appear due to the dominance of a couple of pressuredependent reactions: (a) the addition reaction of OH with NO2 to form HONO2 and its subsequent oxidation to NO3 and NO by the reactions NO2 + OH(+M) = HONO2(+M) (R22, Table 1), HONO2 + OH = H2O + NO3 (R25, Table 1) and NO3 = NO + O2 (R13), and (b) the addition reaction of OH with NO2 to form HNO3 and then the subsequent oxidation to NO3 and NO by the reactions: NO2 + OH(+M) = HNO3(+M) (R14), HNO3 + OH = NO3 + H2O (R15) and NO3 = NO + O2 (R13).
26 ACS Paragon Plus Environment
Page 27 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
(a)
(b)
Figure 10. Major reaction pathways for NO-NO2 conversion at (a) 20 bar, (b) 50 bar for CO/H2/NOx oxidation at 0.063 equivalence ratio. The '+' and '-' symbols in the flux analysis represent formation and consumption of the species associated with the symbol. '(+M)' represents pressure-dependent reaction. The different colors are used to show the paths of different species. 3.3. Stirred Reactor Experiments. The ability of the proposed kinetic mechanism to predict stirred reactor experiments is further tested by comparing simulation results to experiments performed by Dayma et al.
38
. They conducted rich and lean oxidation of NOx-perturbed
hydrogen over a temperature range of 700-1200 K and pressures of 1-10 atm in a fused silica jetstirred reactor. Figure 11 represents some exemplar cases. According to the measurements 38, the hydrogen reactivity and the extent of NO-NO2 conversion decreases from fuel-lean to fuel-rich mixtures, which is captured with reasonable accuracy by the present model. Flux analysis of NOx species at 10 atm pressure and rich conditions shows the importance of the reaction of H2NO with nitric oxide (H2NO + NO = HNO + HNO (R16)), the rate constant of which is updated based on the detailed NH3-oxidation and thermal DeNOx model of Klippenstein et al. 26. For the rich case at atmospheric pressure, despite the decrease in the NO concentration no increase in NO2 is observed suggesting that the decrease in NO is not related to NO-NO2 inter conversion.
27 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(a)
Page 28 of 33
(b)
Figure 11. Experimental and modeling results of the concentration profiles as a function of reactor temperature for H2/O2/NO/N2 system at 1.0 atm and (a) fuel lean, (b) fuel rich conditions. Symbols represent data for JSR experiments of Dayma et al. 38 at fixed residence time (τ) and solid lines represent model predictions. The initial mole fractions for the H2/O2/H2O/NO mixture are- 1%/5%/0.015%/250 ppm, 1%/0.3333%/0%/235 ppm, balance N2. For this particular fuel-rich atmospheric condition, flux analysis confirms that the NO-NO2 inter-conversion paths are not dominant, rather the N-atom balance occurs through an NO-N2 conversion path with intermediate formation of HNO, NH and N by the reactions- NO + H(+M) = HNO(+M) (R17), HNO + H = NH + OH (R18), NH + H = N + H2 (R19), N + NO = N2 + O 28 ACS Paragon Plus Environment
Page 29 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
(R20). Model predictions for 10 atm pressure for both fuel rich and lean conditions are shown in supporting information. In addition to the NOx-perturbed H2 oxidation experiments, the predictability of the model is also compared against NOx-perturbed H2/CO oxidation experiments in a stirred reactor arrangement at atmospheric pressure condition
39
, showing reasonable agreements with the
experimental data. The comparisons for both lean and stoichiometric conditions are illustrated in supporting information. 4. SUMMARY AND CONCLUSIONS As an extension of part I of this two part series we investigated NOx formation for hydrogen and CO/hydrogen mixture oxidation. A specific target was to assess and identify if possible discrepancy in predicting NOx concentration remain when Fenimore NOx reaction kinetic pathways are absent. The results show that even though global combustion targets (e.g. ignition delay time) can be predicted well by different models available in the literature, the nitrogen containing species predictions varied by factors through significantly different evolutionary pathways. A comprehensive detailed chemical kinetic model was developed through assembly of updated literature sub-mechanisms to describe the oxidation of CO/H2/NOx mixtures with particular focus on the detailed implementation of NOx evolution pathways. The construct consists of a C0-C1 sub-mechanism, NOx sub-mechanism and H/N/O sub-mechanism. The model also emphasizes incorporation of NxHy reaction paths as well as species, such as HNO2 and HONO2, which were found to play a decisive role in improving NOx predictions. HONO, HNO2 and HONO2 (thermodynamically less stable isomer of nitrous acid, HONO), has a notable influence on the net NOx formation through modifying the net active radical pool concentration behavior. 29 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 30 of 33
The overall model predictions are in good agreement with multiple experimental datasets over a wide range of venues and operating conditions, including shock tube, plug flow reactor and stirred reactor experiments that cover pressures from 1 to 100 bar and equivalence ratios from 0.5 to 1.5, and temperatures from 600 to 1000 K. In order to replicate the EGR condition, the present study considers a wide range of NOx-perturbed reacting mixtures, such as H2/O2/N2, CO/H2/O2, and CO/H2O/O2/N2. Simulations suggest that the performance of the current model describes the effects of EGR over a wide range of conditions relevant to practical combustion. Supporting Information Initial mole fractions of reactant species and the physical conditions of different experiments used as validation targets, additional graphs with model validation, comparison of the performance of the proposed model with other recent and widely accepted models to predict different experimental targets. Funding The present study is based upon the research work funded by the Department of Energy under Award Number DE-FE0012005. The authors gratefully acknowledge this financial support.
REFERENCES 1. Wang, X.; Law, C. K., An analysis of the explosion limits of hydrogen-oxygen mixtures. The Journal of Chemical Physics 2013, 138, (13), 134305. 2. Bolland, O.; Undrum, H., A novel methodology for comparing CO2 capture options for natural gas-fired combined cycle plants. Advances in Environmental Research 2003, 7, (4), 901911. 3. Zheng, X.; Mantzaras, J.; Bombach, R., Homogeneous combustion of fuel-lean syngas mixtures over platinum at elevated pressures and preheats. Combustion and Flame 2013, 160, (1), 155-169. 30 ACS Paragon Plus Environment
Page 31 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
4. Minchener, A. J., Coal gasification for advanced power generation. Fuel 2005, 84, (17), 2222-2235. 5. Krejci, M. C.; Mathieu, O.; Vissotski, A. J.; Ravi, S.; Sikes, T. G.; Petersen, E. L.; Kérmonès, A.; Metcalfe, W.; Curran, H. J., Laminar Flame Speed and Ignition Delay Time Data for the Kinetic Modeling of Hydrogen and Syngas Fuel Blends. Journal of Engineering for Gas Turbines and Power 2013, 135, (2), 021503-021503. 6. Chacartegui, R.; Sánchez, D.; de Escalona, J. M.; Jimenez-Espadafor, F.; Munoz, A.; Sánchez, T., SPHERA project: Assessing the use of syngas fuels in gas turbines and combined cycles from a global perspective. Fuel Processing Technology 2012, 103, 134-145. 7. Alternative Control Techniques Documents- NOx Emissions from Stationary Gas Turbines, US EPA Office of Air and Radiation, Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina 27711. In 1993. 8. Metcalfe, W. K.; Burke, S. M.; Ahmed, S. S.; Curran, H. J., A hierarchical and comparative kinetic modeling study of C1− C2 hydrocarbon and oxygenated fuels. International Journal of Chemical Kinetics 2013, 45, (10), 638-675. 9. Rasmussen, C. L.; Hansen, J.; Marshall, P.; Glarborg, P., Experimental measurements and kinetic modeling of CO/H2/O2/NOX conversion at high pressure. International Journal of Chemical Kinetics 2008, 40, (8), 454-480. 10. Dagaut, P.; Nicolle, A., Experimental study and detailed kinetic modeling of the effect of exhaust gas on fuel combustion: mutual sensitization of the oxidation of nitric oxide and methane over extended temperature and pressure ranges. Combustion and Flame 2005, 140, (3), 161-171. 11. Konnov, A. A., Implementation of the NCN pathway of prompt-NO formation in the detailed reaction mechanism. Combustion and Flame 2009, 156, (11), 2093-2105. 12. Faravelli, T.; Frassoldati, A.; Ranzi, E., Kinetic modeling of the interactions between NO and hydrocarbons in the oxidation of hydrocarbons at low temperatures. Combustion and Flame 2003, 132, (1–2), 188-207. 13. Mathieu, O.; Levacque, A.; Petersen, E. L., Effects of NO2 addition on hydrogen ignition behind reflected shock waves. Proceedings of the Combustion Institute 2013, 34, (1), 633-640. 14. Frassoldati, A.; Faravelli, T.; Ranzi, E., Kinetic modeling of the interactions between NO and hydrocarbons at high temperature. Combustion and Flame 2003, 135, (1–2), 97-112. 15. Konnov, A. A. Detailed Reaction Mechanism for Small Hydrocarbon Combustion Version 0.5. www.homepages.vub.ac.be/~akonnov/ 16. Watson, G.; Munzar, J.; Bergthorson, J., Diagnostics and modeling of stagnation flames for the validation of thermochemical combustion models for NOx predictions. Energy and Fuels 2013, 27, 7031 - 7043. 17. Burcat, A. Burcat's Thermodynamic Data, Laboratory for Chemical Kinetics. www.garfield.chem.elte.hu/Burcat/burcat.html 18. Burke, M. P.; Chaos, M.; Ju, Y.; Dryer, F. L.; Klippenstein, S. J., Comprehensive H2/O2 kinetic model for high-pressure combustion. International Journal of Chemical Kinetics 2012, 44, (7), 444-474. 19. Kéromnès, A.; Metcalfe, W. K.; Heufer, K. A.; Donohoe, N.; Das, A. K.; Sung, C.-J.; Herzler, J.; Naumann, C.; Griebel, P.; Mathieu, O.; Krejci, M. C.; Petersen, E. L.; Pitz, W. J.; Curran, H. J., An experimental and detailed chemical kinetic modeling study of hydrogen and syngas mixture oxidation at elevated pressures. Combustion and Flame 2013, 160, (6), 9951011.
31 ACS Paragon Plus Environment
Energy & Fuels
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 32 of 33
20. Senosiain, J. P.; Klippenstein, S. J.; Miller, J. A., A complete statistical analysis of the reaction between OH and CO. Proceedings of the Combustion Institute 2005, 30, (1), 945-953. 21. Joshi, A. V.; Wang, H., Master equation modeling of wide range temperature and pressure dependence of CO + OH → products. International Journal of Chemical Kinetics 2006, 38, (1), 57-73. 22. Tian, Z.; Li, Y.; Zhang, L.; Glarborg, P.; Qi, F., An experimental and kinetic modeling study of premixed NH3/CH4/O2/Ar flames at low pressure. Combustion and Flame 2009, 156, (7), 1413-1426. 23. Zeldovich, Y. B., Acta Physicochim 1946, 21. 24. Chen, W. J.; Lo, W. J.; Cheng, B. M.; Lee, Y. P., Photolysis of nitric acid in solid nitrogen. The Journal of chemical physics 1992, 97, (10), 7167-7173. 25. Koch, T. G.; Sodeau, J. R., Photochemistry of Nitric Acid in Low-Temperature Matrixes. The Journal of Physical Chemistry 1995, 99, (27), 10824-10829. 26. Klippenstein, S. J.; Harding, L. B.; Glarborg, P.; Miller, J. A., The role of NNH in NO formation and control. Combustion and Flame 2011, 158, (4), 774-789. 27. Skreiberg, Ø.; Kilpinen, P.; Glarborg, P., Ammonia chemistry below 1400 K under fuelrich conditions in a flow reactor. Combustion and Flame 2004, 136, (4), 501-518. 28. Kee, R. J.; Rupley, F. M.; Miller, J. A. Chemkin-II: A Fortran chemical kinetics package for the analysis of gas-phase chemical kinetics; Sandia National Labs., Livermore, CA (USA): 1989. 29. Lutz, A. E.; Kee, R. J.; Miller, J. A. SENKIN: A FORTRAN program for predicting homogeneous gas phase chemical kinetics with sensitivity analysis; Sandia National Labs., Livermore, CA (USA): 1988. 30. Glarborg, P.; Kee, R.; Grcar, J.; Miller, J. PSE: a Fortran program for modeling wellstirred reactors; Sandia National Labs., Livermore, CA (USA); Danmarks Tekniske Hoejskole, Lyngby. Lab. of Heating and Air Conditioning: 1986. 31. Urzay, J.; Kseib, N.; Davidson, D. F.; Iaccarino, G.; Hanson, R. K., Uncertaintyquantification analysis of the effects of residual impurities on hydrogen–oxygen ignition in shock tubes. Combustion and Flame 2014, 161, (1), 1-15. 32. Hong, Z.; Davidson, D. F.; Hanson, R. K., An improved H2/O2 mechanism based on recent shock tube/laser absorption measurements. Combustion and Flame 2011, 158, (4), 633644. 33. Dryer, F. L.; Haas, F. M.; Santner, J.; Farouk, T. I.; Chaos, M., Interpreting chemical kinetics from complex reaction–advection–diffusion systems: Modeling of flow reactors and related experiments. Progress in Energy and Combustion Science 2014, 44, 19-39. 34. Mueller, M. A.; Yetter, R. A.; Dryer, F. L., Flow reactor studies and kinetic modeling of the H2/O2/NOX and CO/H2O/O2/NOX reactions. International Journal of Chemical Kinetics 1999, 31, (10), 705-724. 35. Roesler, J. F.; Yetter, R. A.; Dryer, F. L., Kinetic interactions of CO, NOx, and HCI emissions in postcombustion gases. Combustion and Flame 1995, 100, (3), 495-504. 36. Arai, N.; Higashi, T.; Hasatani, M.; Sugiyama, S., Formation of thermal NOx in a binarysystem of nitrogen and oxygen. Internation al Chemical Engineering 1978, 18, (4), 661-665. 37. Abian, M.; Alzueta, M. U.; Glarborg, P., Formation of NO from N2/O2 Mixtures in a Flow Reactor: Toward an Accurate Prediction of Thermal NO. International Journal of Chemical Kinetics 2015, 47, (8), 518-532.
32 ACS Paragon Plus Environment
Page 33 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Energy & Fuels
38. Dayma, G.; Dagaut, P., Effects of air contamination on the combustion of hydrogen effect of NO and NO2 addition on hydrogen ignition and oxidation kinetics. Combustion Science and Technology 2006, 178, (10-11), 1999-2024. 39. Dagaut, P.; Lecomte, F.; Mieritz, J.; Glarborg, P., Experimental and kinetic modeling study of the effect of NO and SO2 on the oxidation of CO H2 mixtures. International journal of chemical kinetics 2003, 35, (11), 564-575.
33 ACS Paragon Plus Environment