Sustainability assessment of lignocellulosic biorefineries under uncertainty

Biorefinery concepts refer to the integrated production of food, fuel, power, commodity chemicals, and polymers from biomass, mimicking the structure of a fossil-based refinery. The use of lignocellulosic biomass as feedstock for biorefineries has been promoted as a long-term strategy to mitigate climate change and foster food security. One of the critical bottlenecks for the development of these systems is the recalcitrance of the feedstock. Costly and energy intensive physicochemical and/or enzymatic pretreatment are therefore needed to release simple sugars from the lignocellulose matrix. This and other challenges would penalise the conversion efficiency or the environmental impact. This thesis puts the focus on the techno-economic and environmental assessment of sustainability of biorefinery concepts. On the one hand, it proposes an alternative approach to the conventional economic performance indicators for biorefineries. The proposed economic evaluation method addresses the lack of a formal market for crop residues and other intermediate agro-industrial streams. The approach allows for the comparison of process variants in terms of the maximum willingness-to-pay of a process unit for the bioresource that enters the system. This is calculated as the global biorefinery revenues, estimated from the definition of competitive target prices, minus the processing costs. This thesis also investigated the inclusion of parameter and scenario uncertainty into the LCA of lignocellulosic biorefineries. Scale and state-of-technology issues behind modelling assumptions are not always explicitly presented and justified. In order to facilitate decision-making, a robust comparison among different biorefinery options must include an explicit estimation of the uncertainty associated to process design parameters. The sensitivity of the LCA indicators to methodological choices such as allocation of impacts among co-products, choice of system boundaries and uncertainty in the values of background flows to the system has been evaluated by using systematic process simulation and global sensitivity analysis techniques.The proposed methods are illustrated by two case studies based on process flowsheets adapted from the second generation (2G) ethanol model developed by the US National Renewable Energy Laboratory (NREL). Different alternatives for feedstock, biomass pretreatment and valorization of extracted sugars were considered. They were simulated with Aspen Plus® process simulator. Case study 1 was used to illustrate the techno-economic assessment method. It consists in a basic 2G ethanol facility, refining sugarcane bagasse. Two technological variants were considered: The first corresponded to the NREL benchmark design, using dilute acid pretreatment and producing ethanol (only fuel-OF). The second variant consisted of a low-severity liquid hot water (LHW) pretreatment producing ethanol from cellulose and a pentoses syrup (Feed and Fuel – FF). [...]