(4c) From Process Understanding to Efficiency Improvement: Optimization to search optimum operating conditions

Computational Fluid Dynamics (CFD) and Discrete Element Modeling (DEM) is increasingly used in the Chemical & Process industry to design, analyze and trouble-shoot components. Recent advances in modeling, computational power and automation opens the door for automated design space exploration, which can improve efficiency in terms of throughput, energy savings or process efficiency. However, once a design is committed, there is little flexibility to operators and plant engineers to change the asset easily. They are limited to only changing a few of the operating parameters (rather than the core design of the asset) in order to optimize and improve efficiency.

In this study we show how CFD in combination with a hybrid adaptive optimization algorithm can be used to search a given operating space efficiently to find better designs or better operating conditions based on two mixing vessel examples. A first study, classically done during design stages, identifies a number of different geometric designs, where the mixing time can be reduced by approximately 50% for a given power input compared to literature results. A second study, the core message of this work, explores operating conditions for an aerated fermentation reactor with the goal to minimize the power requirement for the aeration as well as for the agitation while maintaining optimal conditions for the microorganisms.

These two case studies show the applicability of CFD and hybrid multi-objective optimization algorithms for an automated design space exploration (MDX) or operating condition optimization (OCX). The geometric optimization results show a decrease in mixing time up to 50% compared to literature correlations while based on the results of the operating condition exploration the best suited combination which fulfills the case specific requirements can be selected