Systematic Design of the Green Ethylene Glycol Downstream Process under the Generalized Disjunctive Programming Framework

This work deals with the systematic design of the GReen Ethylene Glycol (GREG™) downstream process, currently under development at Biochemtex S.p.A. The GREG™ process produces bio-based glycols through the hydrogenation/hydrogenolysis catalytic conversion of 2nd generation C5 sugars. The aqueous liquid effluent from the catalytic reactor includes the target glycols along with mixed butanediols and heavy components. As a result, the bio-mixture has to be dehydrated and separated from the heavy cut to obtain market grade glycols. After these steps, a distillation sequence is required to get the main oxygenated chemicals at the desired purity. A systematic process design methodology is applied to the entire downstream process based on the formulation of a Generalized Disjunctive Programming problem. The solution of the reformulated MINLP optimization problem is approached with the Logic-Based Outer Approximation algorithm, coupling the simulation environment of PRO/II with the optimization environment of GAMS and C++. This optimization procedure selects the best technology, optimizes units’ topology, column design and operating parameters, and reduces the utility consumption by means of thermal integration.