Oxy-fuel Flames Research Articles

Radiative heat transfer calculation of oxy-fuel flames using the multi-group full-spectrum K-distribution method

Oxy-fuel combustion has drawn much attention as a promising CO2 capture and storage/sequestration technology. Radiative heat transfer of oxy-fuel flames is more important than that of conventional flames owing to higher concentrations of radiating gases. The Full-Spectrum Correlated K-distribution (FSCK) methods with original correlated-K scheme (Modest and Zhang, 2002), the improved one (Cai and Modest, 2014), and the rank correlated one (Solovjov and Webb, 2018) provide obvious errors in the low-temperature regions of dry and wet oxy-fuel flames. To improve the accuracy of FSCK, the accuracy of the Multi-Group FSK (MGFSK) method with the three correlated-K schemes and two grouping strategies based on the scaling function and intervals of comonotonicity are investigated in two 1D, a 2D and a 3D oxy-fuel flames. The results show that MGFSK with different correlated-K schemes and grouping strategies demonstrate better accuracies than corresponding FSCK for both dry and wet oxy-fuel flames. MGFSK reduce the dependence of FSCK on the reference (or Planck) temperature and flame scale. The volume-averaged mole fractions of the maximum and minimum temperatures are recommend as the high- and low-temperature grouping states respectively. Dividing the spectral absorption coefficients into 2 to 4 groups is a good compromise between accuracy and efficiency.

Structures of Ethanol Spray Flames under CO2 Dilution of the Oxidizer in the Counterflow Configuration under MILD Combustion Conditions

Structures of both gaseous and liquid ethanol flames in different oxidizing gas environments in the axisymmetric counterflow configuration at atmospheric pressure are studied. Initially, ethanol/air gas flames are considered where pure ethanol is directed against air at initial temperatures of 400 K, and N2 is successively removed to obtain structures of ethanol/O2 gas flames. Furthermore, the addition of CO2 to the oxidizer side is carried out. Then, an ethanol spray is carried by air and directed against an air stream, and the same procedure is performed as described for the gas flames. The gas strain rate at the fuel side of the configuration is increased from low values of 55/s up to extinction, and the initial droplet diameter is varied. For the combustion of gaseous ethanol in air and in pure oxygen, the nitrogen removal results in an increase in the maximum flame temperature from 2010 K to 2920 K at a gas strain rate of 55/s on the fuel side of the configuration, and the extinction strain rates are 630/s and 26,000/s, respectively. It is confirmed that ethanol spray flames in air show two reaction zones at low strain whereas the lean ethanol spray flames in pure oxygen exhibit a single reaction zone in all situations studied. For increased liquid fuel mass flow rate to a global equivalence ratio of unity, two reaction zones are retrieved. An analysis regarding the addition of CO2 in both the ethanol/oxygen gas and spray flames is also discussed and is found that CO2 dilution of the carrier gas the spray is much more efficient than diluting the opposed gas stream in the counterflow configuration for the generation of MILD combustion conditions in oxy-fuel flames.

Flow pattern and behavior of 40 kWth pulverized torrefied biomass flames under atmospheric and oxy-fuel conditions

The combustion of torrefied biomass under oxy-fuel conditions is a promising technological alternative, which has the potential to transform electric and heat generation systems into more sustainable and carbon-neutral forms. This combustion approach combines two complex processes, oxy-fuel and biomass combustion, which still require significant understanding. For this, an experimental study has been conducted considering a series of 40 kWth aerodynamically stabilized pulverized torrefied biomass and coal flames. Given the increasing interest and advantages of substituting coal by biomass in different thermal industrial processes, a set of four experimental conditions was studied and compared. Two torrefied biomass flames, under conventional and oxy-fuel combustion atmospheres, are compared to two similar coal (lignite) operating conditions. The combustion behavior of the flames is assessed by conducting detailed measurements of flow fields, gas compositions, and radiative heat fluxes. Experiments are performed employing a combustion chamber with a simplified configuration. Combined results from the different measurements show that the overall combustion characteristics of torrefied biomass under the two atmospheres are fairly similar. Also, they do not significantly differ from the combustion characteristics observed for the coal flames. Comparatively, the radiative heat fluxes produced by the torrefied biomass flames were higher than those produced by coal. In addition, oxy-fuel flames produced 15–20% lower radiative heat fluxes than the air counterparts. Gas concentration measurements near the flames showed diverse results attributed to small differences in the turbulent flow fields and fuel composition. The presented analysis of results can also contribute to the retrofitting of conventional pulverized coal firing systems for operation under oxy-fuel conditions with torrefied biomass.

Experimental investigation of the pressure influence on flame structure of fuel-rich oxy-fuel methane flames for synthesis gas production

The structure of fuel-rich CH4/O2-flames at equivalence ratios in the range of 2.5≤ϕ≤2.9, a preheating temperature of TP=573K and pressure in the range of 1bara≤p≤5bara was studied experimentally and compared with numerical results using detailed reaction mechanisms. Two measurement techniques, invasive gas sampling (GC/MS) and optical in-situ TDLAS measurements (tunable diode laser absorption spectroscopy), were used to determine temperature and major combustion species (CH4, O2, H2, H2O, CO and CO2). The results showed that the composition of the synthesis gas in the post flame zone strongly depends on ϕ, while p has only a minor influence. Furthermore, the influence of ϕ and p on the phenomena of super-adiabatic flame temperatures (SAFT) was determined by experimentally and numerically investigating temperature and H2O concentration. The results showed similar trends as for the synthesis gas composition: a strong dependence of ϕ and a minor influence of pressure on the degree of significance of the SAFT phenomena.

LES of oxy-fuel jet flames using the Eulerian Stochastic Fields method with differential diffusion

The Eulerian Stochastic Fields (ESF) Monte Carlo method to solve the transported PDF (TPDF) equation is extended to account for differential diffusion effects by incorporating species individual molecular diffusivities. The method has been applied in Large Eddy simulation (LES) to non-piloted oxy-fuel jet flames at different Reynolds numbers experimentally investigated by Sevault et al. [1]. Due to the high H2 content in the fuel stream and CO2 in the oxidizer these flames pose new challenges to combustion modeling as the flame structures are different compared to CH4/air flames. The simulations show very good agreement with the experiments in terms of mixture fraction conditional mean values for temperature and mayor species on the fuel lean side and the reaction zone, deviations on the fuel rich side are discussed. The trend and location of localized extinction is reproduced well in the simulations, as well as differential diffusion effects in the near field. Additionally, it is shown that a neglect of differential diffusion in the combustion model leads to a lifted flame.

Microstructure and mechanical property of high velocity oxy-fuel sprayed WC-Cr3C2-Ni coatings

WC-Cr3C2-12Ni coatings, with Cr3C2 weight content of 10% and 30%, were fabricated on steel substrate by high velocity oxy-fuel (HVOF) flame spraying technique, and microstructure, phase composition and mechanical properties of WC-Cr3C2-12Ni coatings prepared by different oxygen flows were investigated. The results reveal that phase compositions of WC-Cr3C2-12Ni coatings were mainly influenced by the Cr3C2 content rather than the oxygen flow. WC-Cr3C2-12Ni coatings demonstrated low decarburization and decomposition of WC due to coarse WC gains and high in-flight particle velocities during HVOF. WC-Cr3C2-12Ni coatings exhibited low porosities and typical microstructures, where a mixture of fine and coarse WC grains was uniformly embedded in the Ni/Cr3C2 matrix. The microhardness, indentation fracture toughness and elastic modulus of WC-Cr3C2-12Ni coatings were improved by coarse WC grains, Cr3C2 addition and dense microstructure, which resulted in high wear resistances of WC-Cr3C2-12Ni coatings. The abrasive wear features with groove, peel-off and exposed/fractured WC grains were observed in WC-Cr3C2-12Ni coatings after dry-sliding wear. The relationship between microstructure and mechanical property of WC-Cr3C2-12Ni coatings was discussed from the viewpoints of powder properties and process parameters.

Microstructure and mechanical property of high velocity oxy-fuel sprayed multimodal WC-Cr3C2–Co–Y2O3 cermet coating

To investigate the effect of Cr3C2 and Y2O3 additions on cermet coatings, agglomerated and sintered multimodal WC-10Cr3C2–12Co and WC-10Cr3C2–12Co–1Y2O3 cermet powders were deposited as new type of WC-based cermet coatings on plain carbon steel substrate by high velocity oxy-fuel flame spraying. Microstructures, phases, and mechanical properties of both WC-10Cr3C2–12Co and WC-10Cr3C2–12Co–1Y2O3 coatings were examined and compared with those of WC-12Co coating. Results showed that WC-10Cr3C2–12Co and WC-10Cr3C2–12Co–1Y2O3 cermet coatings presented homogeneous microstructures with multimodal carbide particles well distributed in binder matrix, and exhibited phases similar to the feedstock without decomposed phase. However, WC-12Co coating showed the existence of some serious decomposed phases including W2C and amorphous Co3W3C. Morphologies of individual splats on WC-Co substrate clearly demonstrated liquid–solid two phase deposition behaviors of three types of powders in which carbide particles were mainly present in solid state and Co in liquid state. Furthermore, fracture toughness, elastic modulus, and wear resistance of WC-10Cr3C2–12Co and WC-10Cr3C2–12Co–1Y2O3 coatings were found to be higher than those of WC-12Co coating owing to synergistic effects of Cr3C2, Y2O3, and multimodal carbides. Surface morphologies of worn out WC-12Co, WC-10Cr3C2–12Co, and WC-10Cr3C2–12Co–1Y2O3 coatings indicated that wear mechanism mainly included gouging and peeling off.

Effects of the K-value solution schemes on radiation heat transfer modelling in oxy-fuel flames using the full-spectrum correlated K-distribution method

Radiation heat transfer in oxy-fuel flames is more important than in conventional fuel-air flames. The Full-Spectrum Correlated K-distribution methods (FSCK) with the original correlated-K solution scheme (Modest and Zhang, 2002) and a newly proposed one (Cai and Modest, 2014), and the Rank Correlated Full-Spectrum K-distribution method (RC-FSK) are used in radiative calculations of oxy-fuel flames. Twelve one-dimensional flames, including fuel-air, dry and wet oxy-fuel flames with various length scales, as well as a two-dimensional dry oxy-fuel flame are studied. The results show that the reference temperature has a non-negligible impact on the accuracy of original scheme and the emission-weighted temperature leads to a good accuracy. The accuracy of the new scheme is almost unaffected by the reference temperature, except for small-scale dry oxy-fuel flames. The error of the new scheme is mainly in the low-temperature region, and its accuracy depends on the length scale of computational domain. A hybrid correlated-K scheme using either the original or the new scheme according to local temperature is proposed. It combines the advantages of two schemes in different temperature regions, and is independent of the reference temperature. In addition, the RC-FSK demonstrates almost the same accuracy as FSCK with the hybrid scheme when the Planck temperature is set equal to the reference temperature.

Determination of temperature and water-concentration in fuel-rich oxy-fuel methane flames applying TDLAS

Combustion processes with pure oxygen (oxy-fuel) instead of air as oxidant are attractive for high temperature thermal or thermochemical and gasification processes. The absence of nitrogen in such applications leads to higher temperature and species concentrations, which can stabilize even extremely rich flames. Despite their benefits, there is lack of knowledge concerning the internal structure of rich oxy-fuel flames, which feature reactions with largely diverging chemical time scales, namely, the fast oxidation reactions and the slow endothermic formation of synthesis gas. In order to get a better insight, the scope of this study was to determine axial H2O- and temperature profiles of flat, fuel-rich methane-oxygen flames with equivalence ratios from 2.5 ≤ ϕ ≤ 2.9. A Heat-Flux-burner was used to stabilize quasi-adiabatic one-dimensional flames. The inlet temperature of the gas mixture was kept constant at TP = 300 K and the inlet velocity equal to the laminar burning velocity, which was determined in a preceding experimental study. The in-situ temperature and H2O-concentration measurements were performed using Tunable Diode Laser Absorption Spectroscopy (TDLAS). Laser measurements were carried out with three different diode lasers at center wavelengths λcw = 1344.5 nm, 1392.3 nm and 1853.5 nm, respectively, where multiple absorption peaks of the water molecule were investigated. Additionally, one-dimensional calculations with detailed chemistry were performed using the PREMIX code together with the GRI3.0 and CalTec2.3 mechanisms and compared with the experimental data. The results of the temperature measurements showed temperature peaks in the flame zone and a temperature decrease in the endothermic post flame zone, where synthesis gas is formed. The measured peak temperatures exceed the calculated equilibrium temperatures by approximately 100–400 K indicating super-adiabatic flame temperatures (SAFT). Both reaction mechanisms showed similar trends with respect to the decrease of the temperature in the post flame zone and were in line with the measured temperature. In contrast, the calculated decomposition of water in the post flame zone highly depends on the applied chemistry scheme. Here, the CalTech2.3 mechanism showed excellent performance in comparison to the experimental data for ϕ > 2.5. For ϕ = 2.5 the GRI3.0 performed better.

A GPU-based line-by-line method for thermal radiation transfer of H2O, CO2, and H2O/CO2 mixture

The thermal radiation characteristic of H2O/CO2 mixture in combustion system under high pressure was complex. In order to character its radiation heat transfer, a parallel Line-by-Line method was established to study the thermal radiation characteristic by GPU. The radiation heat transfer was investigated with HITRAN2012 database under the isothermal homogeneous and non-isothermal non-uniform conditions in one-dimensional planar enclosure. The thermal radiation of the pressurized oxy-fuel flame with the dry flue gas recirculation under different pressure were discussed. The results showed that the parallel calculation method had significant accuracy and high computational efficiency. The maximum relative error of computing total emissivity for gas mixture was 6.36%. The parallel method can provide a speedup ratio of 80 by GPU. The difference between the HITRAN and HITEMP was small. With the pressure went up, the parallel method had better computational efficiency than LBL model based on CPU for thermal radiation transfer. The speedup ratio reached 100. The parallel calculation method was completely applicable to study radiation heat transfer of combustion system or model in CFD.

Effects of gas and particle radiation on IFRF 2.5 MW swirling flame under oxy-fuel combustion

This paper reports a comprehensive computational fluid dynamics simulation of the IFRF’s swirling oxy-fuel flame with pulverized coal using wet flue gas recycle. Special attention is given to the influence of particle radiative property model. Based on the burnout rate of pulverized coal, a recent proposed nonlinear conversion-dependent particle radiative property model (Guo et al., 2018) is used in the simulations. The influences of the high CO2 concentration on the sub-models are also considered in detail. The gas-phase global reaction mechanism, kinetic parameters of char oxidation and gasification, and gas radiative property model are modified for oxy-fuel condition. The results show the prediction results agree well with the measurements, in terms of the temperature and composition of the flue gas. The particle radiation has an important effect on the flame temperature. Even in the small furnace studied here (optical length is approximately 1.6 m), the radiation in the flame zone is still dominated by particle radiation. Linear variation assumption (Yin et al., 2015) under-predicts the particle absorption in combustion zone. Ignoring the effect of burnout rate on particle radiative property will 10 times over-predict the particle absorption after burnout.

Electrical characteristics of the oxyfuel flame while cutting steel

We present the first published record of electrical phenomena in the oxyfuel cutting torch flame while cutting a steel work piece. Measurements of the voltage-current characteristic between the torch and the work piece demonstrate three piecewise-linear regimes in the ±10 V range. Flame resistance measurements in the ohmic regime are studied using three candidate excitation signals while varying fuel/oxygen ratio, feed rate, and standoff distance. Electrical characteristics are also observed during successful and unsuccessful pierce and loss-of-cut events. The flame’s electrical resistance is found to be 2.8 kΩ per mm of length; a factor of 3 smaller than the same conditions without cutting oxygen. We present an argument that the drastic increase in plasma conductivity is due to an abundance of electrons produced at the work piece. We draw preliminary conclusions about the potential applicability of these measurements for the in-process detection of standoff, pierce success, and loss-of-cut events.

C-shaped extinction curves and lean fuel limits of methane oxy-fuel diffusion flames at different oxygen concentrations

The extinction characteristics and lean fuel limits of methane oxy-fuel diffusion flames with oxygen mole fractions of 0.35 and 0.4 were numerically studied by using two different radiation models, namely optically thin radiation model (OTM) and discrete ordinates method (DOM). Experimental studies on extinction limits of the oxy-fuel diffusion flames under normal- and micro- gravity conditions were conducted. The results of methane air diffusion flame from present normal-gravity and previous micro-gravity experiments were used for comparison and analysis. C-shaped extinction curves, which consist of the stretch and the radiation extinction branches, were obtained by numerical computations using OTM and DOM for the oxy-fuel diffusion flame. The computational results by using the two different radiation models were compared with the experimental results. The computational lean fuel limit by using the DOM agrees well with the experimental result, while the result by OTM is slightly higher. The primary reason is due the effect of radiation reabsorption, which is not considered in the OTM. The computational and experimental results show that the effects of oxygen concentration of the oxidizer on the combustion region and lean fuel limit of the oxy-fuel diffusion flame are significant. The lean fuel limit of oxy-fuel diffusion flame increases with increasing oxygen concentration. Both computational and experimental results suggest that the lean fuel limit of the methane oxy-fuel diffusion flame with oxygen mole fraction of 0.4 is higher than that of the methane air diffusion flame.

WC10Co4Cr coatings deposited by HVOF on martensitic stainless steel for use in hydraulic turbines: Resistance to corrosion and slurry erosion

The corrosion resistance and slurry erosion wear of WC-10Co-4Cr coatings sprayed by the High Velocity Oxy-Fuel Flame (HVOF) method on an AISI 410 stainless steel substrate, usually employed in hydraulic turbines, were studied. Studies on the materials tested in this work in relation to dry wear and cavitation have been found. However, they do not represent the conditions found in hydraulic turbines. In this work, the wear behavior of the coatings was assessed by erosion tests in aqueous medium containing erodent particles simulating environmental and operating conditions of hydraulic turbines. By using different thermal spray parameters, different levels of porosity were generated in the coated layer. Particle velocity and temperature were monitored during deposition. From the characterization by X-ray diffraction and EDS the phases WC, W2C and the CoCr metal matrix were detected in the layer. The mechanism of material removal by wear was evaluated by mass loss as a function of time and coating properties. Different wear regimes were observed as the thickness of the layer was progressively reduced. The erosion depletion in aqueous medium occurs preferentially in the metal matrix of the coating. Corrosion behavior in aqueous medium containing 3.5 wt% NaCl was also investigated by potentiodynamic polarization tests. Results show that in the coating, corrosion occurred preferentially on the metal matrix - essentially cobalt - the most active element of the coating, while the nonmetallic phases (WC, W2C) are site of the cathodic reaction. The stainless steel substrate is cathodically protected by the metal matrix of the coating.

Numerical analysis of turbulence radiation interaction effect on radiative heat transfer in a swirling oxyfuel furnace

With the application of oxyfuel combustion technology, the status of radiative heat transfer in oxyfuel flames is prominent, and accurate description of it is essential. This study simulates the turbulence radiation interaction (TRI) in a swirling gas-fired furnace to understand the thermal radiation behavior under oxycombustion conditions. Employing the non-gray weighted sum of gray gases model calculates the radiative properties of the participating media and the TRI is modeled. The effects of TRI on the radiative heat transfer in the furnace under different oxycombustion conditions and swirl numbers are analyzed. The results show that the effects of thermal radiation on temperature fields are critical in the gas-fired furnace, especially for oxycombustion condition with high oxygen concentration, and the same is true for TRI effect. With the increase of O2 concentration in O2/CO2 oxidizer, the radiative heat flux on the chamber wall, the radiative heat flux contribution to total heat flux, and the net radiative heat loss are increased. Increasing swirl number can weaken the TRI effect. It is observed that with 35% O2 in oxidizer, the radiation contribution to heat flux along the entire wall and local increase of radiative heat flux due to TRI effect are almost than 45% and 18%, respectively.

A grouping strategy for the multi-group full-spectrum K-distribution method in gas mixtures

The Multi-Group Full-Spectrum K-distribution method (MGFSK) is a promising method for efficient and accurate thermal radiation calculations in participating media with strong temperature inhomogeneity. The current MGFSK method cannot be used for modeling radiative transfer in gas mixtures since it is applicable only to a single radiating gas. To make MGFSK applicable to mixtures without losing efficiency, a new grouping strategy for gas mixtures is proposed based on a scaling function of the mixture. Several one-dimensional two-layer systems with large temperature gradient and a two-dimensional oxy-fuel flame are simulated to demonstrate the accuracy of the modified MGFSK method (MGFSK-M). The simulation results are compared with those of the Line-by-Line model, the Full-Spectrum Correlated K-distribution method (FSCK) and the Rank Correlated Full-Spectrum K-distribution method (RC-FSK). MGFSK-M significantly outperforms both FSCK and RC-FSK in the two-layer systems. For the oxy-fuel flame, RC-FSK and MGFSK-M offer similar accuracy and both perform better than FSCK.

Mathematical modelling of swirl oxy-fuel burner flame characteristics

Abstract Oxy-fuel combustion is the most promising carbon capture and storage technology, which eliminates carbon dioxide emissions into the atmosphere and also decreases nitrogen oxides emissions thereby lowering global warming potential. In order to implement oxy-fuel combustion technology in full scale power plants, its costs, mainly connected with the amount of pure oxygen produced, must be lowered. The main hypothesis is that it is possible to maintain similar velocity and heat transfer distribution while maintaining stable and efficient burner operation during both combustion technologies modifying burner aerodynamics. Excess oxygen is chosen as a representative parameter of burner’s performance and investigation is carried out for four different oxy-fuel burner oxygen excess ratios (λ: 0.8, 0.98, 1.07, and 1.24) together with reference air combustion case. This study suggests a workflow, based on semi-industrial experimental investigations and computational fluid dynamics model composed of advanced sub-models for different combustion phases for development of real scale dual-mode coal swirl burners able for efficient operation during both combustion regimes. The results show that the temperature in near-burner zone and nitrogen oxides emissions increase, while carbon monoxide emissions decrease with the increase of burner oxygen excess ratio, and stable combustion with similar velocity and temperature distributions for both combustion modes is achieved for oxygen excess ratio of 1.07, with decrease in nitrogen oxides and carbon monoxide emissions during oxy-fuel combustion. The performed study demonstrates that it is possible to choose the appropriate burner settings regarding nitrogen oxides and carbon monoxide emissions and burner’s ability to operate stably in both air and oxy-fuel combustion modes.

Combustion Characteristics of a Non-Premixed Oxy-Flame Applying a Hybrid Filtered Eulerian Stochastic Field/Flamelet Progress Variable Approach

The oxidation of methane under oxy-fuel combustion conditions with carbon capture is attractive and deserves huge interest towards reducing CO2 and NOx emissions. The current paper reports on the predictions and analysis of combustion characteristics of a turbulent oxy-methane non-premixed flame operating under highly diluted conditions of CO2 and H2 in oxidizer and fuel streams, respectively. These are achieved by applying a novel, well-designed numerical combustion model. The latter consists of a large eddy simulation (LES) extension of a recently suggested hybrid model in Reynolds averaging-based numerical simulation (RANS) context by the authors. It combines a transported joint scalar probability density function (T-PDF) following the Eulerian Stochastic Field methodology (ESF) on the one hand, and a flamelet progress variable (FPV) turbulent combustion model under consideration of detailed chemical reaction mechanism on the other hand. This novel hybrid ESF/FPV approach removes the weaknesses of the presumed-probability density function (P-PDF)-based FPV modeling, along with the solving of associated additional modeled transport equations while rendering the T-PDF computationally less affordable. First, the prediction capability of the LES hybrid ESF/FPV was appraised on the well-known air-piloted methane jet flame (Sandia Flame D). Then, it was assessed in analyzing the combustion properties of a non-premixed oxy-flame and in capturing the CO2 dilution effect on the oxy-fuel flame behavior. To this end, the so-called oxy-flame B3, already numerically investigated in a RANS context, was analyzed. Comparisons with experimental data in terms of temperature, scalar distributions, and scatter plots agree satisfactorily. Finally, the impact of generating the FPV chemistry table under condition of unity Lewis number, even with CO2 dilution, was investigated on the general prediction of the oxy-fuel flame structure, stability and emissions. In particular, it turns out that 68% molar percentage of CO2 leads to 0.39% of CO formation near the burner fuel nozzle and 0.62% at 10 dfuel above the nozzle.

Study of a full scale oxy-fuel cement rotary kiln

Oxy-fuel combustion has been shown to be an attractive technology to be implemented in cement production with CO2 capture from a process point of view. Due to the market situations and future perspectives it is necessary that the technology can be retrofitted into existing plants. As part of the H2020 EU project CEMCAP, the study presented here investigates the retrofitting of oxy-fuel combustion technology on a typical full-scale cement rotary kiln of 3000 ton/year clinker capacity. The flame generated by a commercial burner has been modelled by CFD with a seven steps combustion reaction model, the k-omega SST turbulence model, and the discrete ordinates radiation model with the Weighted Sum Grey Gas model. The burner is composed of an annular outlet where coal is fed and nozzles from which angled high velocity jets generate a swirling motion for flame stabilization and mixing purposes. As in all rotary cement kilns, a high temperature low velocity secondary oxidizer stream coming from the clinker cooler flows coaxially to the burner in the kiln for heat efficiency purposes. A flame developing in air has been used as a reference case and several oxy-fuel flames have been characterized and compared. The oxy-fuel inlet parameters that are varied in the study are the oxidizer composition and flow rates in both the primary and secondary streams. The work presented is part of an optimization procedure based on iterative interaction with the process modelling of the whole cement plant with full oxy-fuel capture (not presented in this article). With retrofit application in mind, the objective was to reproduce a heat radiation profile to the clinker equivalent to that obtained in the air reference case. The optimized oxy-fuel case is characterized by a primary burner oxidizer of lower flow rate, but with higher oxygen concentration of up to 52.1% to generate the necessary heat in the near burner region.

Oxyfuel combustion and reactants preheating to enhance turbulent flame stabilization of low calorific blast furnace gas

The need for the reduction of the environmental impact of combustion systems in terms of pollutant emissions and preservation of non-renewable resources forces to consider the improvement of energy efficiency and to turn towards alternative fuels in industrial combustion furnaces. However, their use may be an issue because of their low calorific value (LCV) compared to traditional natural gas (NG). The present work considers the combination of oxyfuel combustion with fuel and/or oxygen preheating in order to increase thermal efficiency by heat recovery and enhance LCV oxyfuel flame stabilization, without using a supporting fuel as NG. The study is focused on Blast Furnace Gas (BFG) which has one-tenth the heating value of NG.This assessment starts with thermochemical calculations of major flame properties at several reactant temperatures. Two fundamental flame configurations are simulated: a fully premixed 1D flame for the determination of adiabatic temperature, thermal thickness and laminar burning velocity, and a counter-flow diffusion flame for the determination of extinction strain rates. The results show that significant enhancement of oxyfuel flame properties can be obtained thanks to the preheating of BFG and oxygen. The effect of such preheating is then experimentally studied at laboratory-scale (25 kW) using a tri-coaxial burner generating a non-premixed turbulent BFG-O2 flame. The burner geometry consists of an annular BFG injection surrounded by an inner central oxygen injection and an external annular oxygen injection. Based on a critical Damköhler number, a theoretical analysis of the stabilization limit for a turbulent diffusion BFG-O2 flame with preheated reactant is described and used as a criterion to calculate the burner dimensions. Detailed flame characteristics are investigated from measurements of flue gas emissions and OH* chemiluminescence imaging. From these, the analysis of flame stability diagrams as function of reactant velocities, thermal power, oxygen distribution and preheating temperatures points out the limits of flame stability and the stable combustion regimes achieved at the different operational conditions. Regarding the emissions, very low levels of pollutant emissions such as CO and NOx are achieved in most cases. Further analysis of the results shows that transitions between the various types of flames are controlled by a critical convection velocity in the BFG – oxygen mixing layer. This is quantified from the measurements and matches with the theoretical prediction.Complementary large-scale experiments are performed on a semi-industrial facility, with identical burner geometry, scaled-up to 180 kW using the velocity criterion. The flames show similar structures as obtained at laboratory scale, demonstrating the benefit of preheated oxyfuel combustion for the stabilization of LCV flames. These results validate the analysis of the physical phenomena controlling the limit of stability of BFG oxyfuel flames, as well as the burner design strategy of preheated oxyfuel combustion adapted to low calorific fuels, and the scale-up criteria used for this particular case.