Etude expérimentale de la dévolatilisation rapide de charbons pulvérisés

Experimental study of the rapid devolatilization of pulverized coals. Rapid devolatilization of various pulverized coals have been studied in a laboratory bench constituted by a flat flame burner of propane which reproduces thermal conditions of an industrial flame. The particles, which undergo a heating rate of 6.10⁶ K⁻¹.s⁻¹ with a peak temperature of 1 100 °C, are completely devolatilized within 24 ms. Fifteen coals, included in a wide range (anthracite to subbituminous coal) have been tested. The coal weight loss is globally proportional to the normalized volatile matter content with some exceptions which confirm the advantage of this laboratory bench. The formation of tars or hydrocarbons has been related to the coal weight loss. Carbon, hydrogen and nitrogen devolatilized fraction have been followed with total mass weight loss. In spite of the fact that hydrogen and carbon devolatilized fractions present a good correlation with the total mass weight loss, the nitrogen devolatilized fraction have an anarchic evolution. For high heating rates, the devolatilization of tars observed for the bituminous coals seems to explain this unpredictable phenomenon. These results will be valorized in comparison with those obtained in industrial flame conditions.     

Predictive one step kinetic model of coal pyrolysis for CFD applications

This work’s aim is the development of a simplified kinetic model for coal devolatilization, suitable for CFD applications. The detailed model of coal devolatilization, already developed and tested against a very large set of experiments and conditions, is too large to be implemented and used inside a CFD code.An automatic procedure is developed to estimate the kinetic parameters and stoichiometric coefficients of a one step model. An optimization technique manages to minimize an objective function which measures the distance between the simple one step kinetic mechanism and the results of the detailed model of coal volatilization. The results show that rate parameters can be expressed as function of the coal rank, or better of its carbon content. Despite its simplicity, the one step model is able to characterize not only the weight loss, but also the relative yields of gaseous species (CO, H₂O, CO₂, H₂, C₂H₄, HCN, H₂S), tar species and char. The coal database discussed in this work includes 13 coals of different elemental composition, from lignite to anthracite. The agreement of the one step model with the detailed model of coal volatilization is satisfactory for both evolution rates and cumulative values. Finally, it is important to newly underline that the proposed model, based on the detailed model, is predictive and only needs the coal elemental composition (coal rank) as an input. Two sets of correlation for the distribution of the nitrogen and sulfur compounds into solid, tar and gas phases are also proposed.     

Numerical study of coal particle ignition in air and oxy-atmosphere

The ignition and combustion of coal particles are investigated numerically under conventional and oxy-fuel atmospheres. Devolatilization is computed using the chemical percolation devolatilization (CPD) model. The CPD model is coupled with a Lagrangian particle tracking method in the framework of a multiphysics, multiscale Navier–Stokes solver. Combustion in the gas phase is described using finite rate chemistry. The numerical results for ignition are compared with available experimental data and a remarkably good agreement is observed. The effect on flame ignition of the different phases characterizing the release of volatile gases is assessed. These different phases manifest themselves in two distinct peaks in the devolatilization rate and it is observed that ignition can occur during the first volatile release or on the onset of the second, depending on the particle size and gas temperature. It is found that an increase of ignition delay time in oxy-atmosphere compared to the air case is related to the depletion of radicals that react with the abundant carbon dioxide of the oxy-atmosphere, while the increased heat capacity of the mixture does not play a role.     

Swelling properties and intrinsic reactivities of coal chars produced at elevated pressures and high heating rates

A high-temperature, high-pressure flat-flame burner reactor was developed to prepare char at different pressures. This system achieves particle heating rates of 10⁵ K/s, which better mimics industrial conditions than conventional drop tube or radiative flow reactors. Previous data at atmospheric pressure demonstrated a significant decrease in particle swelling during devolatilization as heating rates increased from 10⁴ K/s (the typical drop tube heating rate) to 10⁵ K/s. Pyrolysis experiments were performed at pressures from 1 to 15 atm at 1300 °C for two bituminous coals and a lignite. Average swelling was determined from a combination of the mass release and the average density. The results indicate significantly lower swelling ratios at elevated pressures than reported in the literature. Scanning electron micrographs show that the bubbles in the bituminous coal particles popped before significant swelling at these elevated heating rates. Lignite particles exhibited shrinkage rather than swelling, but still showed a small effect of pressure. TGA oxidation reactivities were determined for the Pitt #8 and Knife River lignite char samples at their respective char preparation pressures. The oxidation reactivities of both the bituminous and lignite chars decreased with increasing pressure.     

Predicting nitrogen release during coal tar decomposition

This paper develops a detailed phenomenological reaction mechanism for N-species transformations throughout tar decomposition, including tar-N sequestration into soot. It expands the previously validated mechanism for tar decomposition based on FLASHCHAIN^® theory to cover N-transformations during pulverized fuel firing. Tar-N transformations are described by two distinctive features: (1) An elimination reaction that produces HCN governs the decay in the average moles of nitrogen per aromatic nucleus throughout tar decomposition; and (2) Empirical observations determine the fraction of tar-N incorporated into soot. Validation cases represent heating rates of at least several thousand degrees per second; temperatures from 600 to 1100 °C; tar contact times from 75?ms through 2?s; and coal ranks from subbituminous through low volatile bituminous. The predicted partitioning of coal-N into tar-N, HCN, soot-N, and char-N was within measurement uncertainties for all coals for simulated p. f. firing conditions, including the variation in fractional char-N levels from 0.4 to 0.8 across this domain. Since primary tar-N levels are directly proportional to fractional primary tar yields, and since ultimate soot-N levels account for one-third of tar-N with any coal type, the ultimate coal-N partitioning for CFD furnace simulations can be accurately described with two analyses: (1) A primary devolatilization mechanism to predict primary tar yields under rapid heating conditions; and (2) A submechanism to predict HCN release from char throughout devolatilization up to the point of char ignition. Dynamics may be resolved with either global reactions or the full tar decomposition mechanism, depending on the impact of the lag between tar decomposition and soot production in the subject application.     

Predicting ultimate soot yields from any coal

This study presents a simple prediction method for soot yields from any coal, provided that the thermal processing is severe enough to completely incorporate all the aromatics in primary tars and oils into a so-called ultimate soot yield, and given the yield and elemental composition of primary tar. It also introduces a 3-step quasi-global reaction scheme to describe the dynamic approach to ultimate soot yields for deployment in CFD furnace simulations, where the rate parameters are evaluated to match the predicted dynamics from a comprehensive tar conversion mechanism. The predicted ultimate soot yields are validated with measurements for coal ranks from brown coals to anthracites, for heating rates of at least several thousand degrees per second; temperatures from 900 to 1300 °C; and tar contact times from 75?ms through 14?s. Primary tar, defined as the aggregate amount of all condensed aromatics from primary devolatilization, determines the ultimate soot yields from any coal at any operating conditions. Ultimate soot yields from any coal can be evaluated from the yield and FLASHCHAIN^®-based elemental composition of primary tar as accurately as they have been measured. On average, only 18% of the primary tar mass is excluded from the ultimate soot product, which mostly reflects the exchange of C₂H₂ into soot for the CO that carries away tar-oxygen. This exchange nearly compensates for the much more extensive elimination of heteroatoms and aliphatics from tar during soot production. The proposed three-step global scheme accurately depicts the reaction dynamics for tar decomposition at elevated temperatures from the full tar decomposition mechanism, including the maximum and ultimate saturation value of secondary tar. Whereas the kinetic parameters for primary and secondary tar display the established tendencies for different operating conditions and coals, those for soot production are remarkably uniform across this domain.     

Mechanism on the contribution of coal/char fragmentation to fly ash formation during pulverized coal combustion

In this paper, the correlations between coal/char fragmentation and fly ash formation during pulverized coal combustion are investigated. We observed an explosion-like fragmentation of Zhundong coal in the early devolatilization stage by means of high-speed photography in the Hencken flat-flame burner. While high ash-fusion (HAF) bituminous and coal-derived char samples only undergo gentle perimeter fragmentation in the char burning stage. Simultaneously, combustion experiments of two kinds of coals were conducted in a 25?kW down-fired combustor. The particle size distributions (PSDs) of both fine particulates (PM_1-10) and bulk fly ash (PM₁₀₊) were measured by Electrical Low Pressure Impactor (ELPI) and Malvern Mastersizer 2000, respectively. The results show that the mass PSD of residual fly ash (PM₁₊) from Zhundong coal exhibits a bi-modal shape with two peaks located at 14?µm and 102?µm, whereas that from HAF coal only possesses a single peak at 74?µm. A hybrid model accounting for multiple-route ash formation processes is developed to predict the PSD of fly ash during coal combustion. By incorporating coal/char fragmentation sub-models, the simulation can quantitatively reproduce the measured PM₁₊ PSDs for different kinds of coals. The sensitivity analysis further reveals that the bi-modal mass distribution of PM₁₊ intrinsically results from the coal fragmentation during devolatilization.     

Measurement and kinetics of elemental and atomic potassium release from a burning biomass pellet

Combining polarizing-filtered planar laser-induced fluorescence (PLIF) with simultaneous laser absorption, quantitative laser-induced breakdown spectroscopy (LIBS) and two-color pyrometry, the potassium release during the combustion of biomass fuels (corn straw and poplar) has been investigated. The temporal release profiles of volatile atomic potassium and potassium compounds from a corn straw show a single peak. The woody biomass, poplar, produces a dual-maxima distribution for potassium and potassium compounds. For both biomass samples, the highest concentrations of released atomic potassium and potassium compounds occur in the devolatilization stage. The mass ratios between volatile atomic potassium and potassium compounds in the corn straw and poplar cases are 0.77% and 0.79%, respectively. These values agree well with chemical equilibrium predictions that 0.68% of total potassium will be in atomic form. A two-step kinetic model of potassium release has been developed, which gives better predictions during the devolatilization stage than the existing single-step model. Finally, a map of potassium transformation processes during combustion is developed. Starting with inorganic and organic potassium, there are eight proposed transformation pathways including five proposed release pathways that occur during the combustion. The pathways describe the transformation of potassium between the fuel volatile matter, char, and ash. Potassium release during the devolatilization stage is due to pyrolysis and evaporation; during the char burnout stage, potassium release is due to char oxidation and decomposition; and during the ash cooking stage, potassium release is caused by reactions between the ash and H₂O in the co-flow.     

Flamelet modeling of coal particle ignition

This paper numerically investigates the ignition of a single coal particle during the devolatilization phase in a laminar entrained-flow reactor, for which experimental data are available from Molina and Shaddix [3]. Different numerical approaches are combined to evaluate the non-premixed flamelet approach for coal particle ignition. First, the particle trajectory and the particle heating are simulated with a Lagrangian–Eulerian approach using a detailed pyrolysis model. In a second step, these results are used as transient boundary conditions for a simulation fully resolving the flow, the mixing field and the chemical reactions around the particle. Finally, in combination with the boundary conditions the time-dependent scalar dissipation rate profiles from the resolved particle calculation are used in a flamelet calculation for the particle up- and downstream directions. Very good agreement is obtained in terms of ignition delay as well as temperature and chemical species distributions in the mixture fraction space when the resolved particle calculation and the unsteady flamelet calculation are compared in the downstream direction. Good agreement is obtained when the numerical results for the ignition time and the time-averaged OH distribution are compared with the available experimental data. The results show the capability of the laminar flamelet approach to correctly predict coal particle ignition during devolatilization using accurate scalar dissipation rate profiles.     

Experimental and modeling assessment of sulfur release from coal under low and high heating rates

Coal combustion releases elevated amounts of pollutants to the atmosphere including SO_X. During the pyrolysis step, sulfur present in the coal is released to the gas phase as many different chemical species such as H₂S, COS, SO₂, CS₂, thiols and larger tars, also called SO_X precursors, as they form SO_X during combustion. Understanding the sulfur release process is crucial to the development of reliable kinetic models, which support the design of improved reactors for cleaner coal conversion processes. Sulfur release from two bituminous coals, Colombian hard coal (K1) and American high sulfur coal (U2), were studied in the present work. Low heating rate (LHR) experiments were performed in a thermogravimetric analyzer coupled with mass spectrometry (TG-MS), allowing to track the mass loss and the evolution of many volatile species (CO, CO₂, CH₄, SO₂, H₂S, COS, HCl and H₂O). High heating rate (HHR) experiments were performed in an entrained flow reactor (drop-tube reactor – DTR), coupled with MS and nondispersive infrared sensor (NDIR). HHR experiments were complemented with CFD simulation of the multidimentional reacting flow field. A kinetic model of coal pyrolysis is employed to reproduce the experiments allowing a comprehensive assessment of the process. The suitability of this model is confirmed for LHR. The combination of HHR experiments with CFD simulations and kinetic modeling revealed the complexity of sulfur chemistry in coal combustion and allowed to better understand of the individual phenomena resulting in the formation of the different SO_X precursors. LHR and HHR operating conditions lead to different distribution of sulfur species released, highly-dependent on the gas-phase temperature and residence time. Higher retention of total sulfur in char is observed at LHR (63%) when compared to HHR (37–44%), at 1273 K. These data support the development of reliable models with improved predictability.     

Experimental and modeling study of H2S formation and evolution in air staged combustion of pulverized coal

High-concentration H₂S formed in the reduction zone of pulverized coal air-staged combustion can result into the high temperature corrosion of water wall tube of boiler, so it is of great importance to accurately predict H₂S concentration for the safe operation of boilers and burners. H₂S formation and evolution depends on two steps: the sulfur release from coal conversion and gas-phase reactions of sulfur species. In this study, the sulfur release characteristics from the pyrolysis of 17 coals, including 5 lignite, 9 bituminous coals and 3 anthracites, are investigated in a drop tube furnace (DTF). Sulfur release model is developed to describe the relationship between sulfur release and coal types. A global gas-phase reaction mechanism of sulfur species composed of ten reactions is used to calculate and predict the formation and evolution of H₂S, COS and SO₂ in the reduction zone of pulverized coal air-staged combustion. A wide range of air-staged combustion experiments of 17 coals are conducted in the DTF at different temperatures and stoichiometric ratios to validate the developed model. The results show that the prediction errors of sulfur species, including SO₂, H₂S and COS, are within ± 30%, which indicates that the developed prediction model of sulfur species is of great assistance for CFD modeling of actual engineering application.     

Particle imaging of ignition and devolatilization of pulverized coal during oxy-fuel combustion

Oxy-fuel combustion of coal is a promising technology for cost-effective power production with carbon capture and sequestration that has ancillary benefits of emission reductions and lower flue gas cleanup costs. To fully understand the results of pilot-scale tests of oxy-fuel combustion and to accurately predict scale-up performance through CFD modeling, fundamental data are needed concerning coal and coal char combustion properties under these unconventional conditions. In the work reported here, the ignition and devolatilization characteristics of both a high-volatile bituminous coal and a Powder River Basin subbituminous coal were analyzed in detail through single-particle imaging at a gas temperature of 1700 K over a range of 12–36 vol % O₂ in both N₂ and CO₂ diluent gases. The bituminous coal images show large, hot soot cloud radiation whose size and shape vary with oxygen concentration and, to a lesser extent, with the use of N₂ versus CO₂ diluent gas. Subbituminous coal images show cooler, smaller emission signals during devolatilization that have the same characteristic size as the coal particles introduced into the flow (nominally 100 μm). The measurements also demonstrate that the use of CO₂ diluent retards the onset of ignition and increases the duration of devolatilization, once initiated. For a given diluent gas, a higher oxygen concentration yields shorter ignition delay and devolatilization times. The effect of CO₂ on coal particle ignition is explained by its higher molar specific heat and its tendency to reduce the local radical pool. The effect of O₂ on coal particle ignition results from its effect on the local mixture reactivity. CO₂ decreases the rate of devolatilization because of the lower mass diffusivity of volatiles in CO₂ mixtures, whereas higher O₂ concentrations increase the mass flux of oxygen to the volatiles flame and thereby increase the rate of devolatilization.     

On the rank-dependence of coal tar secondary reactions

The secondary reactions of volatile compounds, including coal tar and light gases, accounts for a great portion of soot formation and the subsequent heat release and pollutant emissions in the combustion zone. While coal primary pyrolysis has been extensively studied over the last few decades and several network pyrolysis models has been developed to describe this process, coal secondary pyrolysis is still not well understood. The Babcock and Wilcox Company has been investigating coal secondary pyrolysis in order to develop a comprehensive mechanism for inclusion in predictive computational fluid dynamics and coal combustion models. Supportive experiments were carried out in an entrained-flow reactor. Tar was extracted from the pyrolysis byproducts of seven coals of widely-distributed rank at temperatures ranging from 923 to 1473 K, and analyzed by ¹³C NMR. Tars formed from higher rank coals generally demonstrated higher sooting propensities. This rank-dependent sooting propensity is associated with tar’s chemical structure properties. With increased heat treatment severity, tar molecules lose a substantial amount of aliphatic attachments, and the average size of substitution per cluster decreases. Compared to tars formed from high-rank bituminous coals, those formed from low-rank sub-bituminous coals have a larger attachment portion, higher averaged substitution, and higher oxygen-containing functional groups. These differences contribute to the higher cracking propensity observed for low-rank coal tars.     

Characterization of high heating rate chars of biomass fuels

Data on biomass chars obtained under conditions similar to those of practical applications (high heating rate and low residence time) are required for co-combustion and gasification plants. A methodological procedure is developed and applied to two biomass fuels (cacao shells and olive cake) for producing high heating rate chars and characterizing their reactivity and morphology after the first steps of devolatilization. Different chars are produced in a drop tube reactor (rapid pyrolysis) by varying the nominal temperature and the residence time. Oxidation in air is performed to compare typical temperatures and kinetic parameters and evaluate the effect of the operating conditions on char reactivity. A detailed SEM analysis allows to assess the structural variations during the pyrolysis and detect the main phenomena (softening, swelling, melting, formation of bubbles). A quantitative morphological study is also performed to provide size and shape (important for biomasses) distributions of the parent fuel and the chars. These data are more significant than average values in advanced model to correctly simulate the fluid dynamic behaviour of each dimensional class of particles in large scale furnaces and gasifiers and predict a more reliable residence time of the particles.     

Kinetics of methane and tar evolution during coal pyrolysis

The first objective of this work was to compare the pyrolysis behavior of coals coming from different geographic locations (South Africa, South America, Europe, Australia, and North America). This preliminary study was limited to the kinetics of methane and tar evolution, with data on additional species to be reported in a separate publication. The second objective was to examine the possible relationship between tar and methane evolution during pyrolysis. This study was done by employing a thermogravimetric analyzer coupled with a Fourier-transform infrared spectrometer (TG-FTIR). The evolution curves for 35 coals of different elemental compositions were measured at three different heating rates (10, 30, and 100 K/min). Pyrolysis kinetics were described using a simple first-order reaction model. The technique, first proposed by Kissinger, is based on the variation of the temperature at which a volatile species evolution rate is a maximum (T_max) as a function of the heating rate. The TG-FTIR data for tar evolution reveal a generally consistent behavior for coals from different parts of the world, showing increasing activation energies with increasing coal rank. The same correlation is also true for methane, although the slope of the activation energy versus carbon content curve is rather flat, at least up to about 90% carbon content. The values of activation energies for methane evolution were found to be lower in the case of the Argonne coals, as compared with the non-US coals. A study of the temperatures at which the evolution of methane and tar begins (T_ini), and the temperatures at which the evolution rates reach a maximum (T_max), reveals a correlation between the T_ini for methane and T_max for tar. This may be due to the fact that both tar and methane evolve as a result of similar reactions involved in the breakup and recombination of the coal macromolecular network.     

程序升温气化过程中氮的释放规律研究

采用 U 型管反应系统研究了不同煤种在程序升温条件下在四种气氛中反应时燃料氮的释放规律.研究发现热解时主要的含氮产物为 HCN,如果煤中含有较多的氧,即使在惰性气氛下也会释放出 NO.没有检测到 NH3 的生成,证实NH3 来源于 HCN 在焦表面的二次反应.N2O 仅在有氧气氛中气化时生成,说明氧气是 N2O 生成的必要条件.气化时的总固定氮 TFN 高于热解时,且气化剂中 O2 含量越高则总固定氮 TFN 越高.     

Gamma-iron formation by reduction of iron compounds inherent in brown coal

The reduction of the iron compounds inherent in Victorian brown coal to form α-iron, γ-iron and cementite following heating to high temperatures under reducing conditions has been observed. It is probable that the γ-iron phase is responsible for the singlet reported in heat-treated coals.     

The formation of mesophase composites under influence of the coal-tar pitch during coal carbonization

The influence of coal tar pitch on the formation of the mesophase composites from coal and pitch blends, the development of porous structure during carbonization, and the texture of matter in the pore walls has been studied using X-rays and scanning electronic microscopy (SEM). The influence of the coal-tar pitch on the structure of a coal plastic layer formed during the pyrolysis process of four Polish coals has been monitored. It was determined that, during pyrolysis of three mixtures of Z, S and K coals with coal-tar pitch, more porous composites are formed, and a denser composite is formed from P coal–pitch mixture. Based on a study of the texture of composites and carbonizates, it is suggested that during carbonization of coal Z with a coal-pitch in the plastic state there is a nematic mesophase formed, and when heating P coal–pitch mixtures, a cholesteric one is formed.     

Evaluation of char combustion models: measurement and analysis of variability in char particle size and density

Char samples representing a range of combustion conditions and extents of burnout were obtained from a well-characterized laminar flow combustion experiment. Individual particles from the parent coal and char samples were characterized to determine distributions in particle volume, mass, and density at different extent of burnout. The data were then compared with predictions from a comprehensive char combustion model referred to as the char burnout kinetics model (CBK). The data clearly reflect the particle-to-particle heterogeneity of the parent coal and show a significant broadening in the size and density distributions of the chars resulting from both devolatilization and combustion. Data for chars prepared in a lower oxygen content environment (6% oxygen by vol.) are consistent with zone II type combustion behavior where most of the combustion is occurring near the particle surface. At higher oxygen contents (12% by vol.), the data show indications of more burning occurring in the particle interior. The CBK model does a good job of predicting the general nature of the development of size and density distributions during burning but the input distribution of particle size and density is critical to obtaining good predictions. A significant reduction in particle size was observed to occur as a result of devolatilization. For comprehensive combustion models to provide accurate predictions, this size reduction phenomenon needs to be included in devolatilization models so that representative char distributions are carried through the calculations.     

Reaction/crystallization kinetics studied via in situ XRD: experimental conditions versus methods of kinetic analysis

ABSTRACT

Theoretically simulated kinetic data were used to evaluate the errors associated with the common issue of evaluating the in situ non-isothermal X-ray diffraction data, where the complex multi-step temperature program (alternating the non-isothermal heating steps with isothermal steps during which the diffraction patterns are collected) is for the purposes of evaluation replaced by a simple non-isothermal heating performed at the reduced/effective heating rate. The kinetic analysis has shown that, in general, best results are provided by the non-linear optimisation methods simultaneously evaluating the data-curves obtained for all the different heating rates. For the nucleation growth (KMJMA) kinetics the distortive influence of the temperature program parameters increases as follows: heating rate during non-isothermal segments?<?duration of the isothermal segment?<?temperature interval between the isothermal segments. The non-optimisation methods of kinetic analysis (integral isoconversional methods for evaluation of activation energy E and master plots for determining the appropriate kinetic model) were found to perform inaccurately, with large degree of randomness based on the selection of starting temperature, and are not recommended for evaluation of the in situ XRD data – the only exception seem to be the differential isoconversional methods that provided accurate E values. Generalisation of the present conclusions for all KMJMA processes is suggested and discussed.