To study the pre-treatment effect on thermal degradation of shredded oil palm empty fruit bunches by thermogravimetric analysis / Nur Fatihah Harun

Currently, there is abundance of waste from oil palm production which can be converted into valuable product and it is most effective and efficient solutions for sustainable energy supply correspond to the increasing of energy demand and also to reduce environmental pollution. Increasing in production of oil palm contribute to the increasing of biomass such as empty fruit bunch, palm kernel shell, and oil palm trunk. But then, some researchers found that the biomass has its own problem which is high moisture content, low density and low calorific value compared to fossil fuels to substitute as an alternative energy source. In this study, EFB is used and the pre-treatment is carried out at the earlier stage in order to determine the thermal decomposition of the sample. Initially, the sample was immersed using different method of pre-treatment which is conventional treatment and organoslv treatment. In order to study the decomposition temperature, thermogravimetric analysis (TGA) was used with heating rate of 20℃/min and the initial temperature is 30℃ and the final temperature is 900℃. In this experiment, 100 ml/min of nitrogen flow was used. During the process, the content in the EFB (hemicellulose, cellulose and lignin) affects the physical properties, texture and thermal reactivity. Based on the results, decomposition of EFB consist of three phase where drying period which water are liberated occur at temperature below 200℃, decomposition of hemicellulose and cellulose during devolatilization occur at the range temperature of 200℃ to 500℃, in which at this phase there is a significant drop of in weight loss of sample due to evaporation of volatile hydrocarbon, and lastly followed by steady decomposition of heavy component occur at temperature above 500℃. This study reveal that the pre-treatment by using distilled water is a good pre-treatment because the decomposition of sample is faster compared to others pre-treatment where it shows a high peak on TGA graph.