“…Furthermore, many of these factors are not typically measured or reported in the literature or may be qualitative, such as the type of impeller/sparger. A comprehensive literature review was therefore conducted to identify factors most crucial to scale up for possible inclusion in a machine learning algorithm: - Parameters including power input per volume, gas mass transfer coefficient (k L a), gas flow rates, impeller tip speed, mixing time, and Reynolds number are often analyzed to obtain insights for matching key performance indicators across bioreactor scales [63–65]. These scale‐specific parameters can represent many qualitative and quantitative variables pertaining to mixing, mass transfer, and shear damage to cells.
- There are many factors affecting power consumption in a stirred‐tank bioreactor including impeller design and configuration, sparger design and location, fluid properties, and vessel baffling.
…”