This study investigates by experiment the global characteristics of both moderate or intense low-oxygen dilution (MILD) oxy-combustion and air combustion of firing light oil and pulverized coal in a pilot-scale furnace. There are three burner configurations used, i.e., (I) central straight (primary) jet + swirl (secondary) jet, (II) central straight (primary) jet + two side symmetrical (secondary) jets, and (III) central straight (primary) jet + side asymmetrical jet. The furnace centerline temperature, species concentrations, and exhaust emissions are measured and compared for the MILD and conventional combustion cases. For light oil and pulverized coal, the MILD air combustion or oxy-combustion occurs with burner II or III, while the conventional combustion takes place when using burner I. For the light oil, the MILD oxy-combustion can be reached even using pure oxygen. As the MILD combustion is reached, a fairly uniform temperature distribution and low emissions of NO and CO are obtained. Note that burner III produces the largest internal recirculation of the flue gas, lowest peak temperature, and most uniform temperature, whereas the opposite occurs for burner I. Importantly, the MILD combustion is found to reduce the NO emission much more effectively in the oxy-combustion case than in the air combustion case. Moreover, the appearance of the MILD combustion of light oil and pulverized coal differs from the invisible MILD combustion of gaseous fuels. Dark sparks from burning oil droplets or char particles are present in the MILD combustion of light oil or pulverized coal. It is also revealed that the char burnout under the MILD combustion is weaker than that under the conventional combustion.

No abstract

In this paper, a classification of combustion regimes is investigated for the diffusion flame of a hydrocarbon fuel jet in hot flue-gas co-flow (JHC) with varying oxygen fraction (up to 100%). Numerical simulations by computational fluid dynamics (CFD) are performed to obtain both forced-ignition and autoignition temperatures for the present investigation. All calculations use the Eddy Dissipation Concept (EDC) model with the well-known detailed chemistry-reaction mechanism of methane combustion (GRI-Mech 3.0). To validate the modeling, the predicted JHC flame characteristics are compared with those measured previously [Dally et al., Proc. Combust. Inst. 2002, 29, 1147–1154]. It is found that the predictions agree well with the measurements. Use of the predicted ignition temperatures can qualitatively classify the JHC combustion, based on previous suggestions [Cavaliere and de Joannon, Prog. Energy Combust. Sci. 2004, 30, 329–366] for combustion in a well-stirred reactor (WSR), into three distinct regimes: traditional combustion (TC), high-temperature combustion (HTC), and flameless combustion (FLC). The FLC regime can be further divided into three distinct zones: MILD (moderate or intense low-oxygen dilution), MILD-like, and quasi-MILD. The MILD and MILD-like combustion regimes share the same necessary conditions proposed by Cavaliere and de Joannon while the quasi-MILD combustion does not. It is found that theoretically the MILD-like combustion should occur at any oxygen fraction, as long as the preheating temperature of co-flow prior to their reactions is sufficiently high. By comparison, all previous diffusion MILD combustions were established only for highly diluted reactants at an oxygen fraction of <10%. In this paper, a fundamental analysis of combustion regimes is provided generally for any combustion configuration.

Through experiment and numerical modeling, this study investigated the establishment of moderate or intense low-oxygen dilution (MILD) combustion in a laboratory-scale furnace when fuel and air are fully premixed (FP), partially premixed (PP), or non-premixed (NP). Experiments were carried out at firing rates from 7.5 to 15 kW and equivalence ratios (Φ) ranging from 0.5 to 1. The furnace thermal fields and exhaust NO x emissions for the three mixing patterns were compared. Validated computational fluid dynamics was used to aid in better understanding the flow and compositional structures in the furnace. Natural gas was used as the fuel. The eddy dissipation concept (EDC) model and the GRI-Mech 3.0 mechanism were used. Additional chemical kinetics calculations were also performed to examine reaction pathways under the MILD combustion regime. Moreover, the characteristics of the reaction regime of MILD combustion were examined and are discussed in detail. Estimation of the initial jet momentum rate (J) showed that J FP > J NP > J PP , and consistently the recirculating rate of internal flue gas (K v ) was found to be in the order K v,FP > K v,NP > K v,PP . Correspondingly, the highest values of both furnace temperature and NO x emission were experimentally measured in the PP case, while the lowest values were found in the FP case. The measured NO x emission was negligibly low for the FP case. Numerical results revealed that in all the three cases of firing natural gas (FP, PP, NP), more than 80% of the total NO formation results from the N 2 O intermediate route while other NO mechanisms are unimportant. As Φ is increased from 0.5 to 1.0, both the measured and simulated NO emissions in the three cases initially increase and then decrease. Moreover, for Φ > 0.9, the NO-reburning reaction becomes significant and the resulting reduction of NO is notable. The rates of both turbulent mixing and chemical reaction were found to play a significant role in the structure and establishment of MILD combustion, with estimated Damkoḧler numbers in the range Da = 0.01−5.35.

No abstract