Michael P. Heap

The thermal decomposition of pulverized coal particles

The physical, thermal, and chemical behavior of pulverized coal particles during thermal decomposition are examined for five coal types and two particle sizes for one of the bituminous coals. Particles were injected axially into a lean (35% excess air) methane/air flat flame with a nominal peak temperature of 1750°K. The significant events observed are classified by three time scales. Particles heat to the gas temperature in less than 10 msec, devolatilization occurs between 10 and 75 msec and, under the appropriate conditions, large soot particles are formed WRS and grow for times exceeding 75 msec. The events that accompany devolatilization are dependent upon coal type and particle size. For large bituminous particles (ca., 80 μm) a significant volatile fraction is ejected from the particle as a jet. This volatile jet reacts close to the particle producing a trail of small solid particles. The local heat released during the reaction of the volatiles, in combination with heterogeneous oxidation, increases the particle, temperature and raises it above that of the bulk gas stream. At later times, large soot structures, are formed which are attributed to the agglomeration of small, homogeneously formed soot on the volatile trail structures. Small bituminous particles (ca., 40 μm) burn with a higher intensity (i.e., higher temperature and more rapidly) with few trails and do not produce soot structures probably because of the more diffuse nature of the devolatilization process. Other ranks of coal exhibit different physical, thermal, and chemical behavior. For example, neither the lignites nor the anthracite produce volatile trials. Further, the particle temperature for the lignites is only slightly shifted above the bulk gas temperature in the devolatilization region while anthracite takes 50 msec to reach the bulk gas temperature level. This is attributable to the relatively low heat content of the volatiles in the former case and the low volatile content in the latter. The impact of the above observations on the formation of fuel NO is discussed.

The burning velocity of hydrogen-air flames

Conical hydrogen-air flames have been stabilized on a 10.3-mm-diameter water-cooled nozzle burner designed so as to minimize the lightback tendency and to maintain laminar flow at very high rates. Measurements of the cone half-angle, ϑ , by schlieren optical techniques and the unburned gas velocity, V , by particle-tracking techniques have been used to determine burning velocity from the equation S u = V sin ϑ . A value of 296 cm sec −1 is obtained for a 50% hydrogen-air flame, whereas Jahn and also Scholte and Vaags reported the burning velocity to be approximately 250 cm sec −1 . This discrepancy is attributed to errors in the earlier measurements caused by the use of averaging methods with very small burners. There is remarkably good agreement between the present experimental value of burning velocity and the value computed for a 50% hydrogen-air flame by Dixon-Lewis.

The burning velocity of methane-air flames inhibited by methyl bromide

Abstract A nozzle burner, schlieren cone angle method of burning velocity determination, involving measurement of unburnt gas velocity under flame conditions, has been developed to permit precise measurements of the burning velocity of inhibited flames. Data for methane-air flames inhibited by methyl bromide indicate the following significant features: (1) Addition of methyl bromide causes a shift of maximum burning velocity towards leaner conditions; (2) The effectiveness of methyl bromide as an inhibitor increases as the methane content of the mixture increases; (3) Successive, equal additions of methyl bromide causes progressively smaller reductions in burning velocity. These features are shown to be consistent with the inhibition mechanism proposed by Rosser, Wise and Miller: decomposition of methyl bromide is considered to be practically complete prior to combustion and inhibition is considered to depend on removal, by bromine substitution, of chain carriers essential to the critical stages of combustion. The alternative chain breaking/chain branching competition mechanism of inhibition has not yet been developed sufficiently to permit a reasonable comparison between predicted and measured trends in burning velocity.

The correlation of burning velocity and blowoff data by the flame stretch concept

Abstract Standardized burning velocity and blowoff measurements for methane-air, ethane-air, propane-air, butane-air, and ethylene-air flames are reported. The experimental data relate to inverted flames stabilized on thin plates and to flames stabilized on cylindrical burner rims. The equation K b = kg b pcS u 2 is used to interpret these data in terms of the flame stretch concept. Blowoff of inverted flames from a very thin (0.030-cm) stabilization plate occurs at an approximately constant value of K b = 0.95 ± 0.15 q irrespective of mixture composition. It thus seems probable that blowoff of these inverted flames occurs as a result of excessive flame stretch in the stabilization zone. Blowoff of cylindrical burner flames does not apparently occur at a constant value of K b , even when secondary combustion is eliminated. However, with these flames there are systematic errors in the determination of S u and g h , in the stabilization zone. The local burning velocity will tend to be increased by secondary combustion and decreased by dilution with the atmosphere. Thus S u may be overestimated or underestimated by intdeterminate amounts, and g h is overestimated because flattening of the velocity gradient downstream of the burner rim is neglected. These systematic errors prevent a satisfactory test of the flame stretch theory of blowoff.

Ambient atmosphere effects in flat-flame measurements of burning velocity

A 1-in,-diameter water-cooled porous metal flat-flame burner has been used to determine burning velocities for a wide range of methane-air mixtures, including lean mixtures for which reliable data were not previously available. Provision was made for surrounding the burner with gases other than atmospheric air. Significantly different burning velocity curves were obtained with each different ambient atmosphere used (oxygen, air, nitrogen, and carbon dioxide). The atmosphere dependence is attributed to edge mixing, possibly under the influence of convective entrainment. The measurements of burning velocity in an atmosphere of air (maximum approximately 36.5 cm/sec at approximately 10.25% methane in air) are generally smaller than recent measurements by alternative methods. Edge mixing may well cause an absolute reduction of measured burning velocity additional to the relative effects detected in the between-atmosphere experiments. However, Botha and Spalding [8] have reported satisfactory agreement between the burning velocities of propane-air flames obtained by the cooled-flat-flame method and by other methods. Further work to test the validity of the cooled-flat-flame method of burning velocity measurement thus seems to be necessary.

Behavior of N2O in staged pulverized coal combustion

An experimental study is reported, the objeN/sub 2/Ot of which was to examine the influence of NOx control technology concepts on N/sub 2/O emissions from pulverized coal flames. The work concentrated on the effects of burner design, staged combustion and thermal environment on nitrogen oxides. The results showed that N/sub 2/O emissions from pulverised coal flames can from a significant portion of the total nitrogen oxides (up to 25). The application of combustion modification technology to control NO emissions does not result in significantly higher levels of N/sup 2/O.

Blowoff of inverted flames

Blowoff data have been obtained for methane-air flames stabilized on the edge of a thin plate forming a common long side of twin rectangular slits. Good correlations of the data, using the concepts of critical boundary velocity gradient and Karlovitz flame stretch factor, are possible, but different correlations are obtained for each stabilization plate thickness. The values obtained with very thin plates probably approximate the true values most closely, and with these plates blowoff tends to occur at a constant value of flame stretch factor independent of mixture composition. The results thus provide additional support for the flame stretch theory of blowoff. It is suggested that twin-slit inverted flame burners with extremely thin stabilization plates should permit the accurate determination of critical values of Karlovitz flame stretch factor.

Flame-mode destruction of hazardous waste compounds

Incineration is a promising technique for the disposal of organic hazardous wastes. However, the waste destruction characteristics of turbulent spray flames have not been characterized. In the present research two reactors are used to simulate various aspects of liquid injection incinerator flame zones. The following questions are addressed: (1) Under what conditions do flames quantitatively destroy waste compounds, and (2) how must the flame be perturbed to cause it to fail to quantitatively destroy wastes. The two reactors operated on a simulated waste stream consisting of acrylonitrile, benzene, chlorobenzene, and chloroform. A microspray reactor was used to investigate destruction processes associated with individual droplets of waste compounds. A turbulent flame reactor used a heptane-fueled waste-doped turbulent spray flame to simulate incinerator flame-zone processes. The flames were found to be capable of quantitative waste destruction without the necessity of using common post-flame processes such as afterburners. Furthermore, the high waste destruction efficiency conditions corresponded to high combustion efficiency conditions (i.e., minimum CO and hydrocarbon emissions). Failure to achieve high destruction efficiency resulted from the perturbation of flame parameters. Failure conditions were identified with high and low theoretical air, low temperature, poor atomization quality, and flame impingement on a cold surface. Each failure condition also resulted in elevated CO and hydrocarbon emissions. Thus, the results suggest that CO and hydrocarbon measurements can be used as an indirect, continuous means of monitoring incinerator flame-zone performance.

Observation of the behavior of coal particles during thermal decomposition

Abstract This paper reports on the observed behavior of coal particles during thermal decomposition. The data presented are the first from a study initiated to address both the physical and chemical behavior of pulverized coal from the time of initial heating through the evolution of soot particulate. In the present study, pulverized (high volatile bituminous) coal was injected through a slit centered in a methaneair flat flame burner, and high resolution holography was employed to record the evolution of volatiles and the structure of soot particulate. Volatiles were observed to evolve in a variety of shapes that range from individual jets to uniformly distributed clouds. Coal particle fragments and/or incipient soot nodules (∼ 3 micron) were found to be present within the volatile gases during the evolution. Further from the burner, stringlike soot particles approaching 1600 microns in length were observed.