TIME-RESOLVED KINETICS OF RAPID COAL DEVOLATILIZATION

An apparatus has been developed for kinetic studies of coal devolatilization under closely controlled heating rates, temperatures, and ambient pressures over a wide range of solids reaction time. With the electronic control and quenching introduced here, the heating rates are uniform and the crossover to constant reaction temperature is abrupt. A non-destructive nitrogen quench eliminates decomposition as the sample cools, so that isothermal reaction times are variable in 0.1 s increments. Temperature records produced by the fixed thermocouple at the support are usually reliable; when they are not, analyses of the heat transfer between the support and thermocouple can be interpreted to indicate reliably the sample temperature. These precise temperature programs were imposed between 0.1 torr and 100 atm, at heating rates up to several thousand degrees per second, and at reaction temperatures up to 1000(DEGREES)C. The uncertainties for all experimental conditions in all of our kinetic studies are small.; Several new facets of rapid pyrolysis are evident in the transient weight loss data from an HVA bituminous coal over this experimental domain. During primary devolatilization up to half the sample volatilizes during a few seconds of isothermal heating, and the yields reach constant values which depend on the temperature, heating rate, pressure, and particle size. The reaction time scales depend on temperature and heating rate but not pressure. Below 600(DEGREES)C two distinct stages are evident, one for the evolution of light gases and the other for the co-liberation of gas and liquids. At higher temperatures these stages coalesce.; An intrinsic coupling between temperature and time is fundamental to primary devolatilization. As the isothermal reaction period increases, the yields from a wide range of temperatures shift to significantly greater fractional conversions at low temperatures. The yields after a brief isothermal stage, however, exhibit an acute sensitivity at high temperatures. Based on the transient yields we have reconciled quantitatively incongruous temperature dependences at atmospheric pressure from two earlier microsample strip furnaces and two entrained-flow reactors. Trends from a laboratory fluidized bed are also consistent, albeit qualitatively, once the consequences of disparate reaction times are recognized.