Microwave pyrolysis of forestry waste

This thesis reports a fundamental study of the unassisted pyrolysis of wood using microwave energy for the production of bio-oils. The majority of previous work on the microwave pyrolysis of woody biomass to produce bio-oils has been performed in domestic type multimode cavity based microwaves ovens and has concluded that attaining temperatures required for pyrolysis (500°C) is not possible due to the microwave transparent nature of the material. To overcome this, many researchers have resorted to adding microwave susceptible doping agents to stimulate heating of the wood through conductive heat transfer. Although this method generates overall process effects, it does not realise the unique heating characteristics that microwave energy may offer, such as volumetric and highly selective heating. An in-depth study of the effect of temperature on the dielectric properties concluded that at room temperature, wood is a relatively good microwave absorber with a loss tangent (tan δ) of 0.20 at 2.142 GHz compared to water which has a tan δ of 0.15 under the same conditions. However, as temperature is increased, wood starts to become microwave transparent as the inherent moisture (the microwave significant material from a microwave heating point of view) evaporates causing a decrease in tan 15. Dielectric property results indicated that wood can be classified as a microwave transparent cellular matrix of cellulose, hemicellulose and lignin containing a microwave absorbing phase (water). Selective heating of the bound water before evaporation may be used to heat the remaining bulk to pyrolysis temperatures of circa. 500°C It has been demonstrated that the rate of heating has a marked effect on the microwave susceptibility of the wood above 300°C. As heating rate increases, wood remains microwave responsive up to pyrolysis temperatures of 500°C. At a heating rate of 2°C/min, tan δ was measured at 0.03 at 2.142 GHz, whilst at a heating rate of 15°C/min, tan δ increased more than six fold to 0.19 under the same conditions. A TMo1n applicator was designed and fabricated for the pyrolysis of wood, based directly upon the dielectric properties of the wood feed. This, coupled with automatic tuning to minimise reflected power and increase energy efficiency, ensured a high bulk power density of ~108 W/m3 with 1kW of microwave power compared to a domestic microwave oven which would only generate ~104 W/m3 in the wood under the same conditions. Such a high power density leads to a high heating rate which is required to overcome the decrease in tan 0 shown at lower heating rates in the earlier work. As opposed to the majority of literature, this work has categorically shown that the unassisted microwave pyrolysis of wood to produce bio-oil is technically feasible. This could lead to the full utilisation of the benefits of microwave heating and the unique heating gradients generated that may be beneficial for this process. To test the benefits microwave heating may offer, a matrix of batch pyrolysis tests was carried out to determine the effect of power density, particle size, moisture content and residence time on the yield of bio-oil, char and gas produced from pine and spruce samples. An increase in bulk power density from 1.7x107 to 7.5x107 W/m3 increased bio-oil yield from 29% to 55%. A further increase in power density had no effect on the yield of bio-oil. This body of work showed that the dimensions and geometry of the sample are important factors affecting the yields of products produced. Even though microwave energy heats volumetrically, sample cooling is still constrained to conventional heat loss models (conduction, convection and radiation). Results showed that minimising heat loss and maximising bulk power density can lead to higher bio-oil yields. It was demonstrated that as residence time increased (using a constant power of 1kW), the yield of bio-oil also increased from 13% at 90 seconds to 39% at 180 seconds. These results are the opposite to those observed from conventionally heated pyrolysis experiments as the longer residence promotes cracking of the bio-oil into incondensable gases, causing a decrease in bio-oil yield. This may lead to a potential benefit in utilising microwave heating in this process as expensive rapid quenchers need not be designed. From the particle size range tested, an optimum particle size of 25 mm was found to maximise bio-oil yield. This is much greater than the optimum particle size in conventionally heated pyrolysis (