Deterministic global optimization in isothermal reactor network synthesis

The reactor network synthesis problem involves the simultaneous determination of the structure and operating conditions of a reactor system to optimize a given performance measure. This performance measure may be the yield of a given product, the selectivity between products, or the overall profitability of the process. The problem is formulated as a nonlinear program (NLP) using a superstructure based method in which plug flow reactors (PFRs) in the structure are modeled using differential-algebraic equations (DAEs). This formulation exhibits multiple local minima. To overcome this, a novel deterministic global optimization method tailored to the special structure and characteristics of this problem will be presented. Examples of isothermal networks will be discussed to show the nature of the local minima and illustrate various components of the proposed approach.