Lignocellulosic processes are highly prone to batch-to batch variability, e.g. of raw materials and enzyme activities. This variability can be propagated throughout system scales during process development and optimization, influencing the outputs of bioreaction models, techno-economic analyses and life cycle assessments. As these outputs are the main decision variables for designing and developing lignocellulose-based processes, tools are required to evaluate the influences of process variation at different system scales.

Uncertainty analysis quantifies the effects of model input variations on model outputs. It is an effective tool to consistently propagate process variation throughout scales and analyse its influence on model outputs. As an example, we use a model describing multi-feed simultaneous saccharification and co-fermentation (SSCF) of wheat straw. During the process enzymes hydrolyse the lignocellulosic material to release glucose which can be converted by microorganisms into ethanol. To investigate the impact of batch-to-batch variability in enzyme cocktails, we collected literature data on the enzymatic activity of Cellic CTec2. Retrieved data were propagated in models at bioreactor, techno-economic analysis and life cycle assessment scale. We show how uncertainty analysis can be used to guide process development by comparing different modes of operation. The method can identify economically feasible process ranges with low environmental impact while increasing the robustness of bioprocesses with high variation in raw material inputs. Furthermore, uncertainty analysis could help to identify relevant parameters to choose as response variables in experimental designs.