Using a flame ionisation detector to measure the rate and duration of pyrolysis of a biomass particle

Abstract

Single cubes and spheres of spruce wood have been heated in beds of inert sand, fluidised by nitrogen and heated electrically to 500–700 °C. The release of volatile matter from these pyrolysing particles, submerged inside a cage in the bed, was monitored by continuously sampling the off-gas from the bed into a rapid, flame ionisation detector (FID). This instrument's output was shown to be proportional to the rate at which carbonaceous volatile species were produced by a wooden particle, when thermally decomposing in the hot fluidised bed. The FID's rapid response revealed that some volatiles were released in brief, explosive bursts (lasting ~ 1 s), probably after local build-ups of pressure inside the biomass. Interestingly, the production of volatiles continued after the centre of a decomposing particle had reached the bed's temperature. Thus, the FID provided good measurements of pyrolysis times. The measurements also indicated that volatiles appeared in a fluidised bed as a cloud of bubbles rising around a decomposing particle. The bubbles pushed away the hot sand and so markedly reduced the rate of heat transfer from the bed to a particle. This had unexpected consequences. In a hot bed (700 °C), the duration of pyrolysis for a cube of spruce was proportional to the length, L, of the cube's side, for L ≤ 7 mm. This means that external heat transfer, which included radiation from the bed, then controlled the rate of thermal decomposition. In a cooler bed (500 °C), the duration of pyrolysis depended on a mix of L and L2, indicating that control was then by both internal and external heat transfer. Thus, from 500 to 700 °C bubbles of volatiles increasingly inhibited external heat transfer from the bed to a pyrolysing particle. Also, at 700 °C, the bed's radiation was largely absorbed by the products of pyrolysis inside the bubbles. After a particle's centre had reached the bed's temperature, volatiles continued to appear slowly at a rate, probably controlled by chemical kinetics. The identity of the rate-determining step is discussed for spruce particles of different sizes and beds at various temperatures. However, it is clear that the FID, with its rapid response and sensitivity, revealed new details of the pyrolysis of small particles (2 – 7 mm) of wood in a fluidised bed. For example, the thermal decomposition of spruce involves at least two separate, endothermic reactions and a final, exothermic step.