Lakeland Steel Limited developed a pilot plant for biomass pyrolysis based on sawdust. The pilot plant was based on an auger screw design which was indirectly heated using a double pipe heat exchange configuration to prevent oxidation (combustion) of the feedstock. This thesis covers the preliminary assessment of sawdust pyrolysis for R.H Tregoweth sawmills on behalf of Lakeland Steel Limited. Proximate and ultimate analyses were deployed on the sawdust to determine its composition. Proximate analysis results gave a moisture content of 60%. The dry solids had an organic matter content of 99.22% with ash making the balance. Ultimate analysis was used to determine content levels of elemental carbon, hydrogen, nitrogen sulphur and oxygen. The results on a dry basis were 47.2%, 6.5%, 0.3%, 0.3%, and 44.9 % respectively. Drying models were also used to analyse the sawdust drying characteristics. Drying curves were obtained experimentally and four models: Newton; Page; Henderson and Pabis; and Simpson and Tschernitz, were fitted to the data and their accuracy of fit was determined using residual squared sum of errors. Page’s model was used to describe the sawdust behaviour in the dryer design as it had the highest accuracy. The sawdust reaction kinetics were determined using data from thermogravimetric analysis (TGA) and analysed using distributed action energy model. The kinetics were observed at three heating rates of 10, 20 and 30 °C/min with a maximum temperature of 900°C under an argon atmosphere. Sawdust was modelled as a mixture of water, hemicellulose, cellulose and lignin. Good agreement between Gaussian distribution functions for each component and experimental data were observed. Pilot plant trials were performed using a three factor-three level design of experiment. The factors under investigation were; feedstock moisture content with levels at 15, 30 and 60; reaction temperature with levels at 400, 450 and 500°C; and reactor auger speed with levels at 15, 20 and 25 rpm. Experiments at 60% moisture could not be performed to completion as the auger blocked repeatedly. The other two moisture contents showed that moisture content enhanced heat exchange properties of the feedstock and this generally increased the amount of volatile organic matter released. It was observed that for 15% moisture sawdust increase of temperature did not consistently exhibit an increase in degree of devolatilisation of organic matter. However, the 30% moisture sawdust showed an increase in devolatilisation with increase in temperature. The effects of increasing reactor auger speed had the most consistent trend with which an increase in speed showed a decrease in degree of devolatilisation thus increasing char yield. The empirical data collected from lab scale and pilot plant experiments were used to create mass and energy balances. These were the basis of the large scale mobile pyrolysis plant which was designed to process 3.45 tonnes per hour. Due to size restrictions the large scale dryer was not fitted in the container. It was then determined that the feedstock would either be dried using an onsite kiln or the reactor would process green sawdust. A preliminary economic feasibility assessment was performed for the base case scenario which processed pre-dried sawdust of 15% moisture content at 400°C and a retention time of 45 minutes. A sensitivity analysis based on predicted optimistic and pessimistic conditions showed that automation of the plant had the potential to increase the economic viability of the large scale process.