Modeling for Quality and Safety in Biopharmaceutical Production Processes

Abstract: To design efficient and safe biopharmaceutical production processes, understanding the process and how it is affected by changes in the process conditions is essential. Experimental data form the basis of this understanding which can be enhanced by modeling and simulations of the physical and chemical processes known to control the process. This thesis describes modeling of preparative chromatography and how these models can be applied for the design and characterization of separation steps in biopharmaceutical processing. The process’ robustness to disturbances is studied by simulation of variations in the operating conditions, such as concentrations of target protein, impurities and modifying agents. The sensitivity to the process parameters and the overall risk of producing a product of low quality are determined. These simulation studies are used to find the process parameter variations most relevant to study experimentally. The model-based approach is demonstrated for two industrial case studies. Process simulations are also used to design and analyze control strategies for the preparative chromatography steps. Specifically, an algorithm is designed which selects the optimal cut points used to pool the product in the outlet flow in each batch. Flexible processes which can adapt to disturbances in the process inputs are shown to give higher yield and productivity and less variations in the output.