CONICET | Buscador de Institutos y Recursos Humanos

Integrated biorefineries based on cyanobacteria are a promising alternative to achieve net-zero carbon and sustainability goals. High value-added products such as photosynthetic pigments can be used in pharmaceutical, nutraceutical and food industries. Furthermore, these microorganisms can produce fourth generation bioethanol, when genetically modified. In this work we propose a mixed integer nonlinear programming (MINLP) model for the optimal design of an integrated cyanobacteria-based biorefinery and its heat exchanger network (HEN), for the production of pigments (phycocyanin), biopolymers (PHA), biofuels (bioethanol) and chemicals (diethyl ether, DEE). The main objective is to assess the efficiency of in-silico genetically engineered strains against in-vivo ones, within the framework of a large-scale integrated biorefinery.The proposed production process includes different technology alternatives, embedded within a superstructure. Five Synechocystis sp. strains are considered: three in-silico strains (a genome-scale wild type, two genetically modified to produce ethanol and PHA, respectively) [1] and two in-vivo strains (a wild type and a genetically modified one to produce ethanol). The idea behind evaluating the performance of the developed in-silico strains is to determine if it is worth theeffort to design them based on metabolic engineering. Technologies related with phase isolation, cell harvesting, disruption, and purification are included in the pigments and bioethanol production step. The proposed PHAs production process embeds four alternatives for the biopolymer extraction. An anaerobic digestor is included to produce biogas and fertilizers, and to recycle N and P to the cultivation stage. The superstructure includes DEE and hydrogen production from bioethanol by a tubular packed-bed reactor and steam reforming, respectively. It also includes the HEN design equations. The objective function is the sustainability net present value (SNPV), a multi-criteria assessment of sustainable systems [2], where economic, environmental and social net present values are considered.The proposed superstructure model for the production of 180 t/y of phycocyanin (current industrial scale) and 1700 t/y of bioethanol is formulated as an MINLP problem and implemented in GAMS for SNPV maximization. Model equality constraints include mass and energy balances, yield equations and detailed capital cost for process equipment, while inequality constraints include process and product specifications and operating bounds on process units. The resultingMINLP model has 42407 continuous variables, 9128 discrete variables and 61493 constraints. Numerical results provide a positive SNPV, which gives a promising insight into cyanobacteria biorefineries, based on one of the proposed in-silico strains. Further comparison shows that SNPV is increased in 41% when the in-silico engineered strain is selected, against the in-vivo engineered one.

enviar mensaje