(679d) A Mechanistic Approach to Predicting Oxygen Mass Transfer in Bioreactors.

Stirred tank bioreactors are widely used in biopharmaceutical manufacturing unit operations. The quality of the drug substance produced during the operation is a function of the fluid mechanical environment realized within the vessel (e.g. dissolved gas concentrations, nutrient availability, and fluid shear rates etc.). In principle, this environment is a direct function of the vessel operating conditions. In practice, however, drug production is a complex convolution of the impeller speed, gas flow rate, fluid fill level, and reaction chemistry. Understanding the link between these operating conditions and the resulting environment inside the vessel is key to optimizing production. This understanding is particularly relevant to new bioreactor systems, where are not yet well characterized.

In this presentation, we propose a modeling capability in which hydrodynamics of a bioreactor along with its capability in transferring oxygen to the working fluid is assessed and validated against a set of published experimental data in literature. Specifically, we employ first principle mechanistic modeling to model oxygen mass transfer in a widely used lab-scale bioreactor (AMBR 15). For the first time, to the best of our knowledge, we show how this application allows the prediction of oxygen mass transfer to the fluid from both the open surface and sparger. This methodology is based on sound first principles and thus is not bioreactor specific, in other words, it can be applied across a wide range of scales and bioreactors.