Candidate Attainable Regions for the Oxidative Dehydrogenation of n-Butane using the Recursive Constant Control (RCC) Policy Algorithm

Abstract

Attainable region (AR) ideas have previously been used to identify candidate attainable regions (AR C s) for the oxidative dehydrogenation (ODH) of n-butane to butenes and butadiene and in so doing to identify the maximum possible yields of different hydrocarbon product. Because of the large dimensionality of the problem, it was not possible then to do a complete AR analysis. Among the configurations considered, it was found that the reactor configuration for the respective AR C s in all instances was an inert membrane reactor (IMR) functioning as a differential side-stream reactor in which one of the reactants, oxygen, was introduced along the length of the reactor so as to maintain a very low and constant value of its partial pressure. Nevertheless, despite producing high yields of product, extremely large and impractical residence times ensued. In this paper, a new tool, the recursive convex control (RCC) policy, is used to identify the AR C s in the full dimensional space. These AR C s showed excellent agreement with those previously published, and the optimal reactor structures presented in those publications have been confirmed albeit with different oxygen control parameters. The maximum yields are now achieved with very much lower residence times. These results also confirm the benefit from using the AR approach on problems where a full AR analysis is not possible.