Simultaneous Synthesis And Design of Multiproduct Biotechnological Plants

CARLOS MARTÍNEZ RIASCOS; JORGE MONTAGNA; JOSÉ PINTO

Rio de Janeiro

Congreso; 4th Mercosur Congress on Process Systems Engineering; 2005

Recent advances in biotechnology have allowed the development of many products with high- added value. After laboratory development, clinical trials and regulatory agency approval, the production process must be addressed. The synthesis and design of biotechnological plants is a complex problem that is often approached as two hierarchical problems. The synthesis problem defines the host to be used for the production and consequently the stages involved in the process, whereas in the design problem the goal is to establish the configuration and to size the units in order to satisfy production targets at the lowest cost. This work considers the simultaneous synthesis and design of a multiproduct biotechnological plant. The main decisions involve the selection of the hosts for each product, as well as their corresponding stages; at each stage, parallel units working in-phase or out-of-phase may be selected and storage tanks may be assigned, which allow for a reduction in equipment volume, cost and idle time. The resulting mixed-integer nonlinear problem (MINLP) presents a large number of binary variables and nonlinear design equations. The case study is a multiproduct batch plant for the production and purification of four proteins with several alternative hosts for each one. The process is composed of fifteen stages, and in some of them there are several unit operations that can perform the tasks. Big-M and convex hull formulations were tested in the problem. The latter generates a larger number of variables and constraints, but its computational effort is smaller, due to its lower relaxation gap.