Thomas H. Fletcher

Progress in coal pyrolysis

Abstract The heterogeneous nature of coal and the complexity of the pyrolysis process have made it very difficult to perform unambiguous experiments to determine the rates and mechanisms in coal pyrolysis. However, recent years have seen a number of new experimental and theoretical approaches which shed new light on the subject. This paper considers the recent progress on kinetics, the formation of volatile products, network models, cross-linking, rank effects, and the ‘two-component’ model of coal structure. Recent experiments which measured coal particle temperatures at high heating rates provide reasonable agreement on kinetic rate constants. These rates also agree with those derived from experiments at low heating rates. In tar formation and transport, a consensus is being reached on the central role of the volatility of tar molecules in explaining the variation with operating conditions (pressure, heating rate, particle size, etc.) of the amounts and molecular weight distributions of tars. Progress in the quantitative prediction of tar and char yields is being made through recently developed models for the fragmentation of the macromolecular coal network. These models, which provide quantitative descriptions of the relations between the chemical structure of the coal and the physical and chemical properties of the pyrolysis products (gas, tar, soot, and char), are an exciting advance in the understanding of the pyrolysis process. Such models are linking the occurrence of the plastic phase of the coal with the ‘liquid’ fragments formed during pyrolysis. On the subject of retrogressive cross-linking reactions, both solvent swelling and n.m.r. measurements confirm important rank-dependent differences in reaction rates; these appear to be related to the oxygen functionalities. Reasonable agreement is also seen for variations with coal rank of kinetic rates derived from measurements at low heating rates. Experiments suggest that the recently revived ‘two-component’ hypothesis of coal structure has application to low-rank coals, which are mixtures of two distinct components: polymethylenes and a more aromatic network. Bituminous coals, however, appear far more homogeneous. Although experiments can distinguish loosely and tightly bound fractions these fractions appear to consist of similar materials and are differentiated primarily in their molecular weight and degree of connection to the network. These coals appear to behave in a manner that is described by the network decomposition models.

Soot in coal combustion systems

Soot is generated from coal when volatile matter, tar in particular, undergoes secondary reactions at high temperatures. A description of soot in coal flames allows better calculations of radiative transfer and temperatures in near-burner regions, which in turn allows more accurate predictions of NOx formation in coal-fired furnaces. Experiments are reviewed that examine the formation, agglomeration and properties of coal-derived soot, including pyrolysis experiments and combustion experiments. This review includes the types of experiments performed, the soot yields obtained, the size of the soot particles and agglomerates, the optical properties of soot, the relationship between coal-derived soot and soot from simple hydrocarbons, and attempts to model soot in coal flames.

Decreases in the Swelling and Porosity of Bituminous Coals during Devolatilization at High Heating Rates

Abstract Concern about comparability and validity of different methods for producing coal chars for reactivity experiments has led to research on the effect of devolatilization conditions on the char physical and chemical structure. Particle diameter and porosity changes during devolatilization significantly affect char oxidation rates. In particular, physical properties of chars prepared in drop tube reactors differ greatly from chars prepared in flat flame burner experiments. Recent data indicate that the presence of oxygen in the gas atmosphere has no effect on swelling until char oxidation has begun. The present research concentrates on the effects of heating rate, particle temperature and residence time on the swelling and porosity of a plastic coal, and compares these results with a nonplastic coal. The heating rate at which the transition from increasing swelling to decreasing swelling occurs is approximately 5 × 103 K/s for swelling coals. Swelling coals also reach a maximum porosity near this heating rate. At low particle heating rates swelling gradually increases versus heating rate in contrast to a decline in the swelling at high heating rates in a narrow heating rate region of 2 × 104 to 7 × 104 K/s. Nonswelling bituminous and lignite coals continue to increase in porosity beyond the heating rate of 2 × 104 K/s.

Time-resolved particle temperature and mass loss measurements of a bituminous coal during devolatilization

Abstract Time-resolved measurements of the rate of devolatilization of high volatile bituminous coal particles in nitrogen at gas temperatures of 1050 and 1250 K are presented. Size fractions studied here are 63–75 and 106–125 μm in diameter, with particle heating rates approaching 10 4 K/s. Particle samples are collected in a helium-quench probe, and a multielement tracer technique is used to determine overall mass loss due to devolatilization. An in situ method is used for simultaneously measuring the size, temperature, and velocity of individual coal particles during pyrolysis. The particle temperature measurement is performed using infrared pyrometry at wavelengths of 1.3 and 2.2 μm. The minimum detection threshold for a 100-μm particle in this system is 850 K. The measured particle temperature history is combined with the mass loss determinations in order to study overall coal devolatilization rates. Experimentally determined coal devolatilization rates obtained using this facility are compared with published engineering kinetic models. One-step Arrhenius kinetic coefficients indicated by these data are A = 2.3 × 10 14 s −1 and E = 55 kcal/mol.

Time-Resolved Temperature Measurements of Individual Coal Particles During Devolatilization

Abstract The wide difference in reported coal devolatilization rates during rapid heating can be attributed to inadequate determinations of particle time-temperature profiles. The sensitivity of predicted particle temperature in a rapid heating environment is studied as a function of heat capacity, treatment of local gas temperature, and diameter. An in situ method for measuring simultaneously the size, temperature, and velocity of individual coal particles during pyrolysis is presented. Thermal emission from moving particles is collected using a reflecting microscope assembly and passes through a specially designed aperture, yielding a double pulse signal. The aperture is designed such that the ratio of peak heights yields particle diameter. The emitted light from the particle is split and filtered at wavelengths of 1.36 and 2.2 μm, and particle temperature is determined by applying two-color pyrometry. Particle velocity is obtained from the same emission signals using transit timing. Particle temperatur...

Impact of coal pyrolysis on combustion

The pyrolysis process has impacts throughout coal combustion. The roles of pyrolysis in various aspectsof the coal combustion process are described, including the devolatilization yield, nitrogen release, softening and swelling, soot formation, and char reactivity. These processes can be understood and quantitatively predicted using recently developed network pyrolysis models that describe the transformation of the coals chemical structure. The models are described and examples of their predictive ability for important coal combustion phenomena are presented.

Solid-state 13C NMR characterization of matched tars and chars from rapid coal devolatilization

Matehed tar/char sets were prepared by pyrolysis of a lignite and a bituminous coal in two entrained flow reactors at temperatures between 900 K and 1650K and heating rates of 10 4 –10 5 K/s. Detailed chemical structural characterization of these tars and chars was performed using elemental analysis and solid-state 13 C NMR. This is the first set of detailed solid-state 13 C nuclear magnetic resonance (NMR) data on coal tar samples. The average aromatic cluster sizes of the primary tars from these experiments are quite similar to those of their parent coals, confirming an assumption often made in network devolatilization models. Carbon aromaticities increase in char and tar samples until about 1250 K, after which line broadening in the NMR signal is observed. This line broadening is interpreted as formation of large aromatic radicals. Increases in bridges and loops per cluster are evidence for increased crosslinking above 1250 K. The measured molecular weights per cluster of the primary tars are lower than expected, indicating that some multiple cluster molecules (i.e., dimers) may exist in the tar. Tar and char nitrogen chemical structure is shown to correlate with changes in the carbon aromaticity, which may have implications for nitrogne release models that treat secondary reactions in the tar.

Effects of Moisture on Ignition Behavior of Moist California Chaparral and Utah Leaves

Abstract Individual cuttings from eight plant species native to California chaparral or Utah were burned in a well-controlled, well-instrumented facility. Gas temperatures above a flat-flame burner were controlled at 987 ± 12°C and 10 ± 0.5 mol% O2, resulting in a heat flux at the leaf surface varying from 80–140 kW/m2. High moisture leaves were observed to burst due to the rapid escape of vapor from the leaf interior. Bubbles in or on the leaf surface were observed for leaves with moderate moisture contents. A large number of leaf temperature measurements were made, along with measurements of the ignition time and temperature, flame height, and flame duration. Average ignition temperatures were species dependent, ranging from 227°C to 453°C, with a large degree of scatter from leaf to leaf. Correlations of time to ignition and ignition temperature were made, but showed only a weak dependence on leaf thickness and almost no dependence on mass of moisture in the leaf. Leaf samples with similar mass showed that Utah juniper took longer to burn than the other species, and that the Utah broadleaf species burned more rapidly than all the other species.

Structural evolution of matched tar-char pairs in rapid pyrolysis experiments☆

Abstract Solid-state 13C and 1H nuclear magnetic resonance (n.m.r.) spectroscopy techniques were used to investigate the relationship between the chemical structure of coal and the char particles and condensed tar vapours produced from coals of various ranks at rapid heating conditions. The 13C n.m.r. analysis of the coal chars indicated that significant amounts of aliphatic material were released from the coal during devolatilization with little initial change to the aromatic cluster size or number of cross-links per cluster. The evolution of the char structure following tar release was a function of the temperature history of the char. The structure of the primary tars were compared to the parent coal, and the gas phase evolution of the tar structure was followed with time.

Simulated Land-Based Turbine Deposits Generated in an Accelerated Deposition Facility

This report presents a validation of the design and operation of an accelerated testing facility for the study of foreign deposit layers typical to the operation of land-based gas turbines. This facility was designed to produce turbine deposits in a 4-h test that would simulate 10 000 h of turbine operation. This is accomplished by matching the net foreign particulate throughput of an actual gas turbine. Flow Mach number, temperature and particulate impingement angle are also matched. Validation tests were conducted to model the ingestion of foreign particulate typically found in the urban environment. The majority of this particulate is ceramic in nature and smaller than 10 microns in size, but varies up to 80 microns. Deposits were formed for flow Mach number and temperature of 0.34 and 1150° C, respectively, using MCrAlY coated coupons donated from industry. Investigations over a range of impingement angles yielded samples with deposit thicknesses from 10 to 50 microns in 4h , accelerated-service simulations. Deposit thickness increased substantially with temperature and was roughly constant with impingement angle when the deposit thickness was measured in the direction of the impinging flow. Test validation was achieved using direct comparison with deposits from service hardware. Deposit characteristics affecting blade heat transfer via convection and conduction were assessed. Surface topography analysis indicated that the surface structure of the generated deposits were similar to those found on actual turbine blades. Scanning electron microscope (SEM) and x-ray spectroscopy analyses indicated that the deposit microstructures and chemical compositions were comparable to turbine blade deposit samples obtained from industry. DOI: 10.1115/1.1860380