(543e) Advanced Process Control Design for a Continuous API Synthesis Process

The Advanced Process Control, via model predictive control or complex cascade control loops, is a vital enabling technology to support a robust process with low downtime, high economic return, and consistence quality. Most of the traditional process control based on feedback loops and PAT are slow and inefficient, for example for plug flow reactors with RTD, dispersion, and lag time. The APC or MPC is based on mechanistic understanding of the process and equipment characterization also considering all the reaction kinetics and transport phenomena. These models will be connected to SCADA and can dynamically update, adapt, and optimize for moving horizon situations, fault detection and correction, and keep the process in the state of control for a wide variety of disturbances. This is also the highest level of the digital twin where the material, process, equipment, and control systems are connected and communicate dynamically.

In this presentation we will demonstrate a complex industrial scale case where two precursors are manufactured in parallel lines and a multi-step reaction system yields to final API product in a continuous fashion. The design of the digital twin, models, process map, and tuning of PID and control loops will be discussed. The benefit and steadiness of product quality (impurity concentration) will be demonstrated.