• Energy Research
• Open Access close

The effect of biomass size and aspect ratio on intra-particle tar decomposition has been investigated both numerically and experimentally to achieve a high rate of intra-particle tar decomposition. In one experiment, wood cylinders with a diameter of 8 mm and lengths of 2, 5, and 9 mm were pyrolyzed in an infrared reactor in an argon environment. The final reactor temperature was 973 K and the heating rate was 30 K/s. To make a calculation, a two-dimensional, unsteady state, single particle model was used, and the same convective and radiative heat fluxes were given to the top and side surface of all wood cylinders. Both calculation and experiment showed that tar yield when the length of the biomass was increased and the diameter was kept constant. The calculation showed that, first, tar was formed in the wood cylinder, and then it moved outwards during decomposition. To find an effective aspect ratio of the wood cylinder for further tar decomposition, calculations were also performed in which the aspect ratio (D/L) varied from 0.4 to 6.9 and the wood volume was fixed. As a result, a low aspect ratio was suited for intra-particle tar decomposition because of the difference in the thermal conductivity along the grain and radial directions, although there is an optimum aspect ratio because of the change of residence time. It is well known that the thermal conductivity of unpyrolyzed wood in the radial direction is much lower than that along the grain. By decreasing the aspect ratio, the ratio of the side surface area to total surface area increases. This means that more heat entered from side surface, and low thermal conductivity in the radial direction caused a temperature gradient in the cylinder. When the intra-particle temperature gradient was large, primary tar, which has been formed in the biomass with a relatively low temperature, passed through the side surface layer at high temperatures, enough to advance intra-particle tar decomposition before the tar was released. This resulted in the enhancement of intra-particle tar decomposition.

In this study, tar formation and steam-reforming mechanisms are discussed by separating the tars into heavy, middle, and light tars. Cellulose was heated in a drop-tube furnace under an Ar or Ar/steam atmosphere. After the tars were passed through the furnace for thermal cracking and polymerization, they were trapped by filters set at different temperatures (573, 393, and 273 K), and were respectively defined as heavy, middle, and light tars. Incondensable volatiles and gaseous products were measured using gas chromatography with thermal conductivity (GC-TCD), and flame ionization (GC-FID) detectors. The middle and light tars obtained under an Ar atmosphere were first characterized using time-of-flight mass spectrometry (TOF-MS). The analysis showed that the middle tar did not contain any low-boiling-point light tar components, while the light tar did contain them. It was also found that complex species in the tars were separated to a certain degree by changing the trap temperature. Moreover, the formation of heavy tar was quite different from that of the light tar. With increasing temperature, the formation of heavy tar was inhibited, while that of the light tar was enhanced during pyrolysis. The steam-reforming characteristics of these tars were also different. The heavy tar was barely reformed at a low temperature of 873 K, even with a long residence time, while the middle tar was well reformed by steam. While it was difficult to describe the tar formation and steam-reforming characteristics when the tar was considered as a single condensable matter, the tar formation and steam-reforming characteristics were clarified by separating the tars. This study shows that, to prevent tar emissions, the formation of heavy tar, which barely reacts with steam, should be inhibited during pyrolysis by controlling the heating.

Pyrolysis of cylindrical woody biomass has been investigated both numerically and experimentally with emphasis on intra-particle heat transfer and tar decompostion. In experiment, wood cylinder of 8 mm diameter and 9 mm length was pyrolyzed in an infrared reactor exposed to both convective and radiative heat fluxes in argon environment. The final reactor temperature was 973 K, and heating rate was 5, 10 and 30 K/s. Three K-type thermocouples were located in the sample to measure intra-particle temperature history. The weight fraction history and intra-particle temperature profiles were measured at different runs. Tar was obtained at a cold trap. In calculation, a two-dimensional, unsteady state single particle model was developed and used to simulate the pyrolysis process. Wood cylinder was modeled as an isotropic porous solid. Solid mass conservation equations were solved by using first-order Euler Implicit Method. Gas phase mass conservation equations and energy conservation equation were discretised by finite volume method. In order to investigate the effect of intra-particle heat transfer, simulations were carried out, first, by considering temperature gradient and second, by assuming uniform temperature within the sample. When temperature gradient was considered, simulation results were in good agreement with experimental data. When uniform intra-particle temperature was used in the simulation, simulation results were quite different from experimental measurements, the degree of difference increasing with increase in heating rate. Both calculation and experiment showed tar yield decreased with increasing heating rate. This was because tar formation reaction and intra-particle tar decomposition reactions were enhanced by increase in heating rate but the latter was dominant. It was shown that intra-particle heat transfer and tar decomposition played an important role in pyrolysis characteristics of wood cylinder.