Methodology for a Variable Rate Control Strategy Development in Continuous Manufacturing Applied to Twin-screw Wet-Granulation and Continuous Fluid-bed Drying

AbstractPurposeContinuous manufacturing (CM) promises unrestricted scale up of production volume through runtime, together with lower product quality variability through stricter process control. However, unforeseen incidents that disrupt the process’ steady state or even force temporary shutdown of the line may threaten this promise. A thorough, variable rate control strategy that assures constant product quality through model-based adaption of process parameters in real time while scaling material throughput together with appropriate buffers, can buy time for incident handling and mitigate the risk of complete shutdown. As experience on development of such a variable rate control strategy is limited up to this date, the presented paper is intended as a general guideline that is transferable to other CM processes.MethodsRisk assessment of process-units twin-screw wet-granulation and continuous fluid-bed drying and subsequent design of experiment testing of factor-response relationships was used to identify and quantify critical process parameters in regard to dried granules’ critical quality attributes moisture content (LOD) and particle size distribution. Quantitative findings were transferred into real-time process control actions to demonstrate validity.ResultsThe described methodology facilitated science-based control strategy development, and the implemented compensatory control actions ensured consistent product quality within specified limits at varying process conditions. Especially dryer rotation speed proved to be a beneficial control parameter of LOD at varying total material throughput.ConclusionA variable rate control strategy can mitigate economical losses during continuous production by avoiding process shutdown while managing unforeseen incidents. Although the quantitative findings of this study are distinctively related to the investigated process, the demonstrated methods can be widely transferred to simplify future CM-development. Graphical Abstractᅟ