(420b) Synthesis of Sustainable Processing Networks: Location-Dependent Biorefinery Models

The current and projected increase in demand of commodities, water and energy caused by the growing world population motivates innovation in process synthesis-design towards more sustainable processes. This is achieved by using alternative (renewable) raw materials, such as biomass, incorporating new process technologies and satisfying new design objectives and constraints, including sustainability.

This work describes the development, modeling and optimization of a comprehensive biorefinery superstructure with multiple product and feedstocks. Over 100 processing alternatives for converting the mentioned feedstocks into the various products are included, leading to over a hundred trillion theoretical alternative solutions.

A systematic process synthesis framework for biorefineries has been developed (Bertran et al, 2017), alongside the necessary methods and tools (such as a comprehensive database of biomass, chemicals, reaction paths, technologies, prices and many more), and implemented in a software interface named Super-O. This framework has been applied in this contribution to solve various problems derived from the biorefinery superstructure, under different scenarios. These include: (i) selection of a product or a set of products given the feedstock mix; (ii) selection of a feedstock or set of feedstocks given the product mix; (iii) selection of the processing route for a given feedstock-product combination; (iv) selection of the location given the feedstock, product and route; (v) all possible combinations of the aforementioned options. Moreover, multiple location options are investigated, by incorporating transportation options between each location, to determine the optimal distributed location-configuration for the entire value chain. Using integer cuts, a number of top-ranked solutions are generated and analyzed.

In this contribution, a strong emphasis is made on the management of the problem data through a data structure implemented as a database. Upon a specific problem definition, the relevant data is retrieved from the database and the problem is then solved via optimization.

The production of bioethanol alongside co-products (such as succinic acid, lactic acid, and lysine) from various lignocellulosic feedstocks is used as a case study to determine the optimal feedstock mix, processing route, set of locations and configuration.

References:

• M.-O. Bertran, R. Frauzem, A.-S. Sanchez-Arcilla, L. Zhang, R. Gani, Comput. Chem. Eng. (2017) DOI: 10.1016/j.compchemeng.2017.01.030.