Kinetic Modeling of Lignocellulosic Ethanol Production from Pretreated Birch by SSF

Biological conversion of lignocellulosic biomass to fuels and chemicals provides recognized benefits and has huge potential. To reduce cost of such a bioprocess, e.g. simultaneous saccharification and fermentation (SSF), it is necessary to use high density raw materials to achieve high concentration of end product. But unfavorable mixing and concentrated inhibitors come along with high gravity substrate loads. This thesis tried to address these problems by developing fedbatch strategies for SSF processes, including substrate, enzyme and cell feedings. The experimental results, however, indicated that the proposed feeding strategies did not result in significant improvement in enzymatic hydrolysis, glucose/xylose turnover and ethanol production at 10% water insoluble solids (WIS) loading. An integrated kinetic model was developed to simulate lignocellulosic ethanol production from pretreated birch slurry, through batch SSF and fedbatch SSF with discrete feeds of solid substrates and active cells. A concept of particle conversion rather than reactor conversion was applied to describe the declining substrate reactivity. Modified Langmuir equation and adsorption kinetics were proposed and tested for determining enzyme adsorption. Growth based and glucose allocation based fermentation models were developed and fitted to experimental data. After these, the overall SSF model was constructed upon the selected adsorption kinetics for cellulytic enzyme and semi-growth associated fermentation model for cells. It could reproduce the major profiles of substrate hydrolysis, glucose turnover, yeast growth and ethanol production. It could also reflect the influences of feeding strategies in these observed profiles. Further studies dedicated to expanding the modeling of intracellular reactions by developing a structured fermentation model are suggested. Application of modeling for optimization of feeding strategies at even higher WIS loading is also suggeste