Design and Control Integration of a Reactive Distillation Column for Ethyl Lactate Production

Abstract

ABSTRACT: Nowadays, the worldwide tendency to obtain environmentally friendly products through the use of safe and stable production processes, minimizing the energy consumption (i.e. using energy integration), and avoiding products out of specification, are an important motivation for applying a process design methodology that incorporates controllability issues since the earliest design stages. Although the topic of design-control integration has been a research topic investigated from different fronts for more than thirty-five years, it was in 2005 where a methodology incorporating local practical controllability issues for nonlinear systems was proposed. Such methodology allows designing processes that fulfill some controllability criteria, which assures that the resulted design will be controllable from the modern control theory. The mentioned design-control integration methodology was applied in this work for designing a reactive distillation column for producing ethyl lactate, an important green solvent. Production of this green solvent has gained great attention worldwide since it is seen as an excellent alternative for replacing petroleum-based solvents. As with any green product that intends to replace oil-based products, ethyl lactate production needs to be improved (in terms of its economic feasibility) to have an actual chance for replacing the petroleum-based solvents at a worldwide scale. One of the proposals for improving the economic feasibility of this green solvent, is to produce it in a reactive distillation column system, which would reduce the energy consumption, increasing the process profit. The design-control methodology applied here involved several steps. First, the development of a first principles-based model is required. Unfortunately, experimental data for a reactive distillation system for ethyl lactate production are scarce. Therefore, the model was identified and validated using data generated by running simulations in Aspen Plus. After model validation, simulated data were used in conjunction with knowledge of the process (obtained from technical literature) to select the state variables to be controlled. Then the manipulated and controlled variables were paired by applying digraphs theory, which avoids linearization of the nonlinear model. After this, local practical controllability metrics were formulated for being used as constraints during the optimization step of the design-control methodology. Besides the controllability metrics, physical constraints as well as product specifications constraints were included in the optimization. To compare the integrated design methodology with a traditional design methodology, the optimization was also run but considering only the physical and product specifications as constraints, but not the controllability metrics. Results of the comparison of the integrated design and the traditional design methodologies have shown that the design obtained by using the design control methodology leads to a higher profit while fulfilling all the constraints. A key factor in the design of the reactive distillation column is the ratio between the number of trays in the rectification zone and the stripping zone. Therefore, the optimization was run for several values of this ratio. Then the best case for this ratio was used for finally designing the column under the design–control methodology. Furthermore, as defining a ratio between the column length and column diameter is a common practice in the traditional design of distillation columns, in this work, such ratio was also included as a constraint in the optimization problem, to investigate how it impacted the optimal design results. It was observed that such type of constraint is not suitable for being included in the design of the reactive distillation column for the analyzed case study.