This contribution presents a multi-period synthesis of an optimally-integrated regional biorefinery's supply networks, based on a Mixed-Integer Linear Programming (MILP) model. The production processes from different sources of biomass include first, second, and third generations of biofuels such as bioethanol, biodiesel, hydrogen, Fischer-Tropsch (FT)-diesel, and green gasoline. The aim is to maximize the economically optimal utilization of seasonal and year-round continuously harvested raw materials from regionally-located available biomass resources, by considering the competition between fuels and food production. The proposed multi-period MILP model enables efficient bioenergy network synthesis and optimization. Economically optimal solutions are obtained, with optimal selection of technologies, raw materials, intermediate and final products, and the timely-optimal planning of harvesting, biofuels production, storage, and logistics.

This doctoral dissertation, which consists of four substantive wholes, presents several syntheses of sustainable bioprocesses using computer-aided process engineering. In the first part the synthesis of different integrated processes of ethanol production from the entire corn plant is presented. The synthesis of different processes is in the second part further extended to thesimplified and more comprehensive synthesis of bioproducts in the whole production supply chain network. Synthesis is based on the generic optimisation model of biomass production and supply chain networks. In the third part three methods for sustainable development assessment, suitable for multi-criteria optimisation, are presented: method of sustainability indexes, footprints and combined criteria, such as eco- and total profit. Methods are further upgraded with indirect effects in order to measure the unburdening theenvironment, associated with the use and replacement of environmentally-harmful products. Methods include the direct, indirect and total impacts on the environment. In the last part the methodology for reducing a large number of criteria within multi-objective optimisation to a small number of representative criteria is presented. This method is presentedon the case of environmental footprints.