Process development for platform chemical production from agricultural and forestry residues

Doctoral thesis, 2021

As part of a bio-based economy, biorefineries are envisaged to sustainably produce platform chemicals via biochemical conversion of agricultural and forestry residues. However, supply risks, the recalcitrance of lignocellulosic biomass, and inhibitor formation during pre­treatment impair the economic feasibility of such biorefineries. In this thesis, process design and assessment were developed with the aim of addressing these hurdles and improving the cost-effectiveness of lignocellulose-derived platform chemicals.

To expand the feedstock base and reduce operational costs, logging residues served as underutilised and inexpensive raw material. The major impediment in converting logging residues was their high recalcitrance and low cellulose content, which resulted in low attainable ethanol titres during simultaneous saccharification and co-fermentation (SSCF). Pretreatment optimisation reduced inhibitor formation and recalcitrance, and led to enzymatic hydrolysis yields at par with those obtained for stem wood, despite the less favourable chemical composition. Upgrading logging residues with carbohydrate-rich oat hulls increased ethanol titres to >50 g/L using batch SSCF at 20% WIS loadings, demonstrating the potential to further decrease downstream processing costs.

To alleviate the toxicity of inhibitors generated during pretreatment, preadaptation was applied to Saccharomyces cerevisiae. Exposure to the inhibitors in the pretreated liquid fraction improved ethanol production during subsequent fermentation. Transferring the concept of preadaptation to lactic acid production by Bacillus coagulans cut the process times by half and more than doubled the average specific lactic acid productivity, showcasing how preadaptation could decrease operational costs.

To assess the performance and robustness of process designs against process input variations, a multi-scale variability analysis framework was developed. The framework included models for bioprocess, flowsheet, techno-economic, and life cycle assessment. In a case study, multi-feed processes, in which solids and cells are fed to the process using model-based predictions, were more robust against variable cellulolytic activities than batch SSCFs in a wheat straw-based ethanol biorefinery. The developed framework can be used to identify robust biorefinery process designs, which simultaneously meet technological, economic, and environmental goals.