SESSION 5

COMBUSTION OF SOLID FUELS: PF,FBC AND WASTE

ABSTRACT

Three lots of ash from a small bubbling fluidized bed combustor, each made up of three fractions and two lots of ash from a large industrial scale circula6ng fluidized bed installation were examined. Limestone sorbents were used in both cases. Ash samples were subjected to phase analysis "as received" and after prehydration at ambient conditions and under pressure. Determinations of the proportion of free CaO were also made for selected samples, including some to which a controlled amount of CaO or silica had been added before hydration. When possible, the thermal effect associated with mixing the samples with an excess of water was measured. It was concluded that during the prehydration process CaO can be liberated from Ca-containing coal ash components but it can also be consumed in reactions with silica or silicates. These processes can affect the exothermicity.

ABSTRACT

An improved version of a combustion-desulfurization model for fluidized bed combustors burning high ash and sulfur content low quality lignite in its own ash with the addition of limestone is presented. The desulfurization model considers sulfur retention by both limestone and ash and is based on first order catalyst deactivation where the kinetics of desulfurization is described by two rate constants; one for the initial surface reaction, the other for the rate of pore plugging or deactivation. The accuracy of the complete model with experimentally determined deactivation rate constants was tested by applying it to a 0.3 MWt AFBC test rig and comparing its predictions with measurements. Comparisons show that the model produces reasonably accurate predictions of sulfur retention provided that the rate constants are determined experimentally under the fluidized bed combustor operating conditions.

ABSTRACT

Bagasse fired furnaces have long been used in the sugar industry but suffer from stability problems due to the high moisture content of the fuel (45-50% as fired). For a wide range of furnace conditions a cyclic pattern of material accumulating, drying and then burning can be observed at various places on the grate. For some furnace conditions the pattern occurs simultaneously over the entire grate leading to large fluctuations in furnace pressure, reductions in steam output and generally unstable furnace behaviour. In extreme cases explosions have occurred, leading to severe damage to the furnace itself. A time-dependant computational fluid dynamics model of the combustion processes is developed. In particular, the accumulation and burnout of material deposited on the grate has an important influence on the stability of the furnace. The model has an ignition criterion and together with the drying and burnout rates this controls the frequency of the combustion fluctuations. FURNACE predictions are compared with gas temperature measurements at three separate points 0.72 meters above the grate. The qualitative behaviour of the gas temperature is predicted well at all three positions.

ABSTRACT

Extensive studies have documented the strong dependence of char reactivity on heat treatment history, but there have been relatively few studies focusing on the short time scales relevant to combustion processes. This paper presents measurements of reactivity loss in coal and alternate fuel chars as a function of heat treatment under combustion-like conditions. An all-graphite transient heat treatment device has been developed capable of temperatures from 700 - 3000 K, maximum heating rates of 2*10³ K/s and hold times at peak temperature from 2 - 120 seconds. The experiments were augmented by TGA analysis of very young chars generated by dilute-phase oxidative pyrolysis in a flame-supported entrained flow reactor to explore the role of annealing at very short residence times (< 50 msec).

The low temperature (500 - 800°C) oxidation reactivity of chars was observed to vary by almost six orders of magnitude as a function of organic precursor and heat treatment conditions. Rapid heat treatment (~2 seconds) between 700°C and 2400°C has a very large effect on reactivity, causing a variation in reactivity by a factor that ranges from 30-30,000 depending on precursor. The propensity to deactivate shows a strong rank dependence (the low-rank, reactive chars deactivate more), but there are significant fuel-to-fuel differences in addition to the overall rank trend. An example is the high annealing propensity of the anthracite (despite its low initial reactivity) which can be rationalized by structural changes observed by high resolution TEM fringe imaging.

ABSTRACT

The background and current state of biomass resources and utilization in Romania is presented. Important quantities result annually especially from agriculture and wood processing factories, but most of the biomass is not used at all or is burned inefficiently in old boilers.

Researches begun at ICPET Cercetare to develop a more efficient and environment friendly technology for biomass combustion, suitable for local conditions. The experiments were performed on two laboratory-scale facilities - one fluidized bed and one circulating fluidized bed (CFB) combustor - and on a 1 MWt CFB pilot plant. All three facilities are described, with a special attention conferred to the fuel feeding systems.

Experimental results are discussed for several biomass types tested - wood chips, sawdust and sorghum waste. Figures, tables and diagrams are included.

ABSTRACT

The objective of this study is to investigate the effect of heat treatment on the evolution of reactivity of highly porous chars, without the presence of mineral matter, from Raman scattering measurements. Pure synthetic char particles (Spherocarb) were used as a model material for a class of highly porous coal-chars. Two sets of experiments were carried out: A) Samples were oxidized in air at 600°C without heat treatment and Raman spectra were measured as a function of conversion. B) Samples were heat-treated in the range 900-1400°C, in nitrogen, then characterized by micro-Raman spectroscopy, in the range 800-1800 cm^-1, and then oxidized by air at 500 °C and by CO₂ at 900 °C in a thermogravimetric balance. In all Raman scattering measurements, three-band spectra were obtained. The G and D bands were dominant and a weak "1180 cm^-1" band was apparent. The spectra were fitted to a 3-Lorentzian function from which width, peak position and intensity of each band were determined and used for analysis of the evolution of reactivity: The following conclusions can be drawn from results of experiments A and B respectively:
A) Throughout reaction with oxygen the most significant change in system characteristics (reactivity, Raman features) occurs in the conversion range 0-30%. This is consistent with previous measurements of thermal resistivity and intrinsic reaction rate.
B) Heat treatment up to 900 °C at time duration up to 300 minutes did not show any effect on reactivity. Decrease in reactivity was apparent at 1200 °C for 2-min treatment, above which the effect was more pronounced.

The characteristics (width, peak location, and intensity) of each of the three Raman bands showed clear correlation with reactivity changes both in air and CO₂. These were attributed to morphological (dimensions and ordering) changes in the rearrangement of the carbon network structure, as previously suggested. We propose in the present study an interpretation of these changes and their effect on reactivity.

Key Words: Char, Reactivity, Porous Structure, Oxidation, Annealing, Raman Scattering.

ABSTRACT

This paper discusses the results obtained from an experimental combustion work undertaken to investigate the behaviour of multicomponent briquettes, prepared by mixing two different particle sizes of coal and two different types of binder species.

Single briquettes were burned over a wide range of temperatures in a laboratory scale fluidised bed combustor facility.

Nitrogen (NO_x and N₂0) and Sulphur (SO₂) oxides emissions resulting from the combustion of these briquettes were constantly monitored during the time of burning. The levels of O₂, CO₂ and CO were also recorded during the same period.

Experimental results showed that coal particle size influenced burn-out times and emission levels of some of gaseous species. The binder type was also found to have a major influence on the emissions of different pollutants.

The temperature was observed to significantly influence the extent of the effects of the other operating parameters studied.

ABSTRACT

Surface area of Tarfaya shale particles was measured for large and small particles formed by attrition and fragmentation during combustion in fluidized reactor. Bed temperature was held constant between 350°C and 750°C. During fluidization, each particle size changes continuously and results in a particle size distribution. Parent particle fragments to complete destruction after a period of 4 hours for bed temperature equal to 350°C. This time was reduced to few second (20 - 40 s) if bed temperature exceeded 750°C. For each experiment, particles were collected out of the reactor, screened and weighed at different times of combustion. The specific area of each class of particle was determined by nitrogen adsorption desorption. Results obtained show that surface area of bed material depends on particles size distribution which range from the parent to the fine particles. Bed temperature and initial size particle act together to change the shale porosity at any stage of the combustion process. Particles processed at 400°C present a surface area 1.5 order of magnitude greater than the original ones. It was found that fine particles, generally formed by abrasion, present a higher surface area for bed temperature greater than 450°C. This is due to the fact that oxidation occurs mainly on external part of the particle and they are free of organic matter once detached from the parent particle. Surface area of parent, particle, not fragmented during fluidization varies slightly from surface area of the original particle. For bed temperature exceeding 600°C, all the class of particles present in the bed have surface area greater than 4 m²/g and tend to rapidly devolatilise and change their surface area to reach 7 m²/g.

ABSTRACT

The sound-assisted fluidized bed combustion of powders made of tyre derived fuel (TDF) and of a biomass (Robinia Pseudoacacia) has been studied to assess the benefits deriving from the application of acoustic fields of different frequency on the performance of the combustor. The parameters monitored were: the carbon elutriation rate, the carbon loading in the bed and the degree of fixed carbon conversion. Both fuels are characterized by strong propensity to generate fines during pyrolysis and/or combustion. Elutriable particles are mainly formed during devolatilization for TDF, during char burn-off for Robinia. Experiments consisted of steady combustion, either ordinary or sound-assisted, of TDF and Robinia fine particles. Acoustic fields of 150dB and frequency of 80, 120 and 240Hz were used in sound assisted experiments. High intensity acoustic fields of 120Hz reduce carbon elutriation rates by a factor of about 1.5, regardless of the fuel used. Correspondingly, fixed carbon conversion efficiency increases by 5-8% and 2-3% for TDF and Robinia respectively. Sound frequencies either lower or higher than 120Hz reduce the effectiveness of sound, approaching the performances of ordinary fluidized bed combustors.

Results have been analyzed considering that fine particles are present in the bed either as fines freely moving in the interstices of the bed (free fines) or as fines attached onto coarse inert bed particles (attached fines). Accordingly the inventory of carbon present under steady state conditions in the bed depends on the competition between the following processes: elutriation of free fines, adhesion and/or cohesion of elutriable fines onto bed material, generation of elutriable fines by attrition of fine-coarse aggregates and combustion of fine particles. In this framework, acoustic fields have the property of enhancing the formation of fine-coarse aggregates.

ABSTRACT

In order to optimize the combustion process in small-scale wood heaters with a thermal capacity of up to 15 kW using natural wood logs an important task is to investigate the mixing and combustion processes within the reaction zones of such firing systems. For a basic understanding of the formation and decomposition of different gas components and the dependency on mixing intensity and therefore on turbulent behaviour of the reacting flow numerical studies were carried out using the simulation program AIOLOS. A comparison of results obtained during the numerical modeling studies applying the k,e -, low Reynolds- number k,e - and the Reynolds-stress turbulence model on different grids with detailed experimental data shows a good correspondence both for isothermal conditions as well as during combustion. The main aim is to validate the numerical model enabling its use as an engineering tool to investigate the influence of parameters such as mixing conditions, combustion air distribution and furnace geometry on the combustion process within the wood heater on a complete combustion for reduced emission of unburned components, such as CO, hydrocarbons and soot.

ABSTRACT

Combustion studies of five coals of different origin were carried out in a laboratory scale fluidized bed combustor. Five blends prepared by mixing two coals based on their petrological characterisation, in varying amounts, were selected to study the possibility of reduction NO_x, N₂O and SO₂ emissions.

The results showed that some blends had the opposite behaviour concerning the release of NO_x and SO₂ in relation to parent coals, and the emissions were higher than expected. The N₂O amounts observed were, however, in almost all blends tested, lower than predicted values.

With some blends, the mixing levels intended to reduce SO₂ were not always found to correspond to those for simultaneous decrease of NO_x.

Most of the blends studied showed some evidence of interaction between them. Varying the proportion of the blend components was observed to alter the temperatures at which interactions were stronger.