Subscriber access provided by MT ROYAL COLLEGE
Article
What differences does LES find among traditional, high-temperature and MILD combustion processes of a CH4/H2 jet flame in hot oxidizer coflow? Guochang Wang, and Jianchun Mi Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b03874 • Publication Date (Web): 26 Feb 2018 Downloaded from http://pubs.acs.org on February 26, 2018
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 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 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 40 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
What differences does LES find among traditional, high-temperature and MILD combustion processes of a CH4/H2 jet flame in hot oxidizer coflow? Guochang Wang and Jianchun Mi* Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China *
Corresponding author: E-mail:
[email protected]; Fax: +86-10-62767074
Abstract Large Eddy Simulation (LES) of a CH4/H2 diffusion Jet flame in Hot Coflow (JHC) is undertaken to find distinct behaviors of Moderate or Intense Low oxygen Dilution (MILD) Combustion (MILDC), High Temperature Combustion (HTC) and Traditional Combustion (TC). These three JHC flames are realized by using different coflow temperatures and oxygen mass fractions: (1) 1300K and 9% for MILDC; (2) 1300K and 30% for HTC; (3) 600K and 30% for TC. The modeling of LES combining the eddy dissipation concept (EDC) with a global four-step reaction mechanism is validated by the JHC measurements of Dally et al. (Proc. Combust. Inst. 2002, 29, 1147-1154). The instantaneous and time-averaged velocities, temperatures and species concentrations such as carbon monoxide (CO) are presented and compared for the three cases.
It is demonstrated that the JHC flames of MILDC and HTC both develop from auto-ignition nearly immediately downstream of the nozzle exit, while the lift-off JHC flame of TC evolves from an induced-ignition with a significant delay. Manifestly, combustion reactions proceed gently in the MILDC case and highly aggressively in the TC case. In both MILDC and HTC cases, stable combustion ensues in the very near field. While most heat releases around the stoichiometric location and transfers away slowly, combustion species, e.g. CO, diffuse more rapidly across the jet flow. The
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
JHC flame for TC behaves completely differently. With a wobbling flame base, large-scale flame oscillations enhance crosswise turbulent mixing and heat transfer. Consequently, high temperatures and high CO concentrations concurrently emerge across the central region in the mid field. Besides, local extinction and re-ignition appear to occur frequently in the TC and do not happen in the HTC and MILDC.
ACS Paragon Plus Environment
Page 2 of 40
Page 3 of 40 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
1. Introduction Various new combustion technologies have been sought in past decades to achieve both clean and efficient combustion of fossil fuel. Moderate or intense low-oxygen diluted combustion (MILDC)1, also referred to as flameless combustion (FLC)2 or sometimes high temperature air combustion (HiTAC)3, is one of such promising combustion technologies. Cavaliere et al.1 defined MILDC occurring in well stir reactor (WSR) as a premixed combustion process that satisfies two conditions: (i) the inlet temperature of the reactants (Tin) is above the auto-ignition temperature (Tai), i.e. Tin > Tai; and (ii) the maximum temperature rise due to combustion ∆Tmax < Tai. In furnace, MILDC can be realized through external and/or internal exhaust gas recirculation that simultaneously dilutes and heats up the reactants, consequently reducing both peak flame temperatures and NOx formations. Note that MILDC has a substantially more uniform temperature and species concentration fields and behaves more stably relative to the traditional combustion (TC).
Previous studies4-10 have found that the morphology and scalar distribution in the reaction zone of MILDC differ considerably from those of TC. Laser measurements of Plessing et al.4 showed sharp rises of temperature and OH intensity occurring in the reaction zone of a traditional premixed flame, which are typical for the flamelet structure. Quite contrarily, the MILDC exhibits only a small temperature rise and a low OH intensity. Weber et al.5,
6
performed a series of experiments on
MILDC in a semi-industry scale furnace. These authors showed that, in MILDC cases of burning various fuels of different types (natural gas, light oil, and pulverized-coal), the mean temperature field is always quite uniform and no high temperature-gradient flame fronts are present. Li et al.7 studied the furnace MILDC through the reaction characteristics and found its reaction zone size being much larger than for the TC case. Nevertheless, the PLIF and Rayleigh scattering measurements of Duwig et al.8 appeared to demonstrate some different results. While the temperature measurements showed a uniform and wide combustion zone under MILDC, the PLIFs of OH and CH2O indicated the existence of thin reaction zones8. To resolve this contradiction, Minamoto et al.9,
10
performed Direct Numerical Simulations (DNS) of MILDC and traditional
premixed combustion simply in a 10-mm cube box. These investigaters9, 10 demonstrated that there
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
are indeed thin reaction regions with high reaction rates, thus leading to large scalar gradients and implying the presence of fine flamelets in the MILDC reaction zone. Those flamelets interact with one another very frequently in space and time, resulting in the thickening of reaction zone and enhanced reaction rates substantially. Therefore, apparently uniform and distributed combustion is expected even in practical cases under MILD condition. However, naturally limited by short time and small computational domain, the DNS calculations of Minamoto et al.9, 10 cannot help to understand the temporal/spatial formations and evolutions of fine flamelets and reaction zones over a sufficient large space, e.g., for a jet flame. As a result, dissimilar ignitions and stabilizations of JHC flames under MILDC, HTC and TC have yet to be explored.
In a combustion process, the flame stability is important. Dally et al.11 experimented the JHC flames of CH4/H2. They measured the instantaneous temperature and major/minor species concentrations using the single-point Raman-Rayleigh-laser-induced fluorescence technique and found that the fluctuations of temperature and species concentrations are fairly small, indicating a well stabilized JHC flame for MILDC. Medwell et al.12 investigated the spatial distribution of hydroxyl radical (OH) and formaldehyde (H2CO) to further understand the local extinction and re-ignition phenomenon in the JHC flame. Their results showed that the local extinction allows the premixing of the fuel and oxidant, highlighting the need for homogeneous mixing to maintain the MILDC process. Oldenhof et al.13, 14 studied the influence of entrainment on the MILDC establishment and the JHC flame stability. Their OH-PLIF measurements and the quantitative analysis indicated that the quantity of hot oxidant traveling toward the fuel jet edge is an important factor in the flame stability, leading to the dominant role of auto-ignition for the stabilization of JHC flame. Yet, the JHC flame stability in either HTC or TC mode has not been studied in any extent. Thus, their stability characteristics are still unknown. Likewise, this question is new: i.e., how distinct are the stability characteristics of HTC and TC from those of MILDC? So it is significant to investigate distinct flame stability characteristics of the JHC flames for MILDC, HTC and TC.
However, the above issues neither can be addressed currently through experiments because of the
ACS Paragon Plus Environment
Page 4 of 40
Page 5 of 40 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
difficulty of obtaining the instantaneous field measurements of high reaction temperatures and species concentrations. Nor they can be well resolved by DNS due to too high cost and too much time of computations. The method of LES is nevertheless more applicable. To simulate turbulent JHC flames, Ihme et al.15, 16 successfully used LES and steady flamelet/progress variable (FPV) model, Afarin et al.17, 18 adopted LES and CPaSR model while Bhaya et al.19 employed LES and PDF-based chemistry model. These studies have confirmed the capability of LES to capture the transient flame structures and reaction zone evolutions of JHC flames. Importantly, the applications of LES with different chemical mechanisms have been well validated by the previous works15,16 and also, late, by the present work for a global four-step reaction mechanism, using the JHC measurements of Dally et al.11
In this context, using the LES modeling, the present study is aimed at characterizing distinct JHC flames under the regimes of MILDC, HTC and TC. The following special coflow conditions are designated to achieve the three regimes: (1) Tc = 1300K and YO2,c = 9% for MILDC, (2) Tc = 1300K and YO2,c = 30% for HTC, (3) Tc = 600K and YO2,c = 30% for TC; where Tc denotes the coflow temperature and YO2,c is the coflow oxygen level in mass fraction. Of note, a relatively high oxygen level of YO2,c = 30% is selected for the HTC and TC regimes in order to effectively characterize the aspect of high temperature for the HTC case and that of traditional flame for the TC case. This oxygen level is more than three times as high as that for the MILDC case so that the HTC case will be highly distinct from the MILDC. The coflow temperature of 600 K for the TC case is set not only to make it below the auto-ignition temperature but also to have a much lower environmental temperature than that for the HTC case, so that the TC heat can transfer out of the combustion zone far more rapidly than does the HTC heat. Also, worth noting is that LES is implemented to simulate combustion aerodynamics of those JHC flames, while the turbulence-chemistry interaction is approximately modeled using eddy dissipation concept (EDC) with a global four-step chemistry mechanism (see Section 2 for more details). The previous numerical studies20-23 by RANS (Reynolds-averaged Navier Stokes) modeling have well validated EDC for the JHC flame. It can simulate the uniform and distributed reactions in the MILD combustion22, compared to other chemistry models, e.g., steady flamelet model or ξ/PDF model. Crucially, the previous works7,20 have
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 40
also confirmed that the EDC model is applicable to the HTC and TC cases.
The rest of the paper is organized as follows. Computational models of the present modeling are given in Section 2 while Section 3 offers the numerical setup for present calculations. In Section 4, an experimental validation of the LES modeling is made first, and then various computational results for the MILDC, HTC and TC cases follow orderly as combustion regimes of the three cases, instantaneous JHC flame structures / combustion aerodynamics and statistical analysis. To understand the JHC distinctions of MILDC, HTC and TC, the flow structures and reaction zones in the three cases are specified by distinguishing their instantaneous and time-averaged distributions of velocity, mixture fraction, temperature and concentrations of the combustion species CO and CO2. Finally, several findings and conclusions from the present work are summarized in Section 5.
2. Computational models Large eddy simulation (LES) is employed to model the turbulence in JHC flame. In LES, large unsteady turbulent motions are directly represented whereas small turbulent motions are modeled. Therefore, a filter operation is needed to decompose the dependent variable f into the filtered (resolved) component and the sub-grid scale (SGS) component. The Favre (density averaged) filter operator is usually adopted, i.e.,
=
(1)
where ̅ is the average density; for incompressible flow, ̅ = ρ, so = ̅. The filtered mass, momentum, species and energy conservation equations are as follows:
+
+
+
=0
(2) =
= (
− −
(3)
− [ ( ) + ! " − # " )]
ACS Paragon Plus Environment
(4)
Page 7 of 40 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
& %
+
& %
=
+ '
* + () + + * − [ , ! ℎ. − # ℎ. /]
(5)
In Eqs. (2) - (5), , 1, " , ℎ. , 2 are the Favre-filtered velocity, pressure, mass fraction of kth species,
sensible enthalpy and temperature, respectively; 3#' is the stress tensor due to molecular viscosity defined by 3#' ≡ [ 5 6
+
8
7 − 5 9
: :
;#' ]
(6)
