New Scale-Up Technologies for Multipurpose Pharmaceutical Production Plants: Use Case of a Heterogeneous Hydrogenation Process

Minimizing the effort associated with the pilot and laboratory-scale experiments needed for a successful scale-up of a process from laboratory to production scale is a significant challenge in process development. Efficient scale-up is becoming increasingly important in process development due to the growing pressure to reduce costs and timelines while achieving a first-time-right approach. This article describes innovative technologies that enable direct and efficient process scale-up from the laboratory to production scale, while concurrently optimizing scale-dependent parameters through in-depth process understanding. Those technologies include a dynamic process model (based on a digital twin) and a laboratory-scale imitation (Scale-Down-Reactor) of a specific production-scale reactor (4000 L). The core component of the Scale-Down-Reactor is a 3D-printed metallic insert (H/C-Finger), designed to replicate the heat transfer behavior of the production reactor by maintaining a similar heat transfer coefficient and surface-to-volume ratio. In order to maintain comparable gas–liquid mass transfer between the scales, the Scale-Down-Reactor was designed with geometric similarity to its large-scale counterpart. Both mass transfer and heat transfer were experimentally evaluated for the two scales, and the comparison demonstrated an excellent agreement. To finally prove and validate the concept, a hydrogenation process currently running at the production scale was conducted in the Scale-Down-Reactor. As a second technology, a dynamic process model is described that includes a kinetic model of the chemical reactions and a heat/mass transfer model (digital twin) of the aforementioned production-scale reactor. For the gas–liquid mass transfer model, an improved mathematical description (equation) was developed. Moreover, the production-scale hydrogenation process conditions were efficiently optimized using the dynamic process model. The measured reaction mixture temperature profile of the optimized production batch demonstrated excellent agreement with the profile predicted by the dynamic process model. By enabling direct and efficient process scale-up while concurrently optimizing scale-dependent parameters, the technologies described within this article offer a promising approach to reducing costs and timelines while improving process understanding.