Scale‐Up, Stirred Tank Reactors

Abstract

Scale‐up is best approached by considering the need to provide the optimum environment for the cell. It is inevitably a compromise and needs appropriate small‐scale data as discussed here. However, arbitrary and inflexible “scale‐up rules” are not useful. The most critical aspects of scale‐up are oxygen and carbon dioxide mass transfer and their concentration in the broth, all of which depend on the air sparge rate and power input from the selected impeller(s). The Rushton turbine is discussed followed by the now‐preferred hollow‐blade and up‐pumping, high solidity ratio hydrofoils. Bulk blending/homogeneity (especially at the high aspect ratios found on the commercial scale), holdup, and foaming and heat transfer in turbulent conditions as found with small‐sized microorganisms are considered, followed by the additional complexity that arises with these parameters when dealing with high viscosity, rheologically complex broths with mycelial and biopolymer fermentations. The impact of scale‐up on two aspects of bioperformance is considered, namely, mechanical stress from agitation and stresses from the inevitable poorer homogeneity. It is concluded that mechanical stresses are only important with mycelial fermentations and the impact depends on the organism. Stresses due to spatial differences in dO 2 and in pH batch fermentations, as well as nutrients in fed batch, cause a poorer performance at the commercial scale. Such information is vital for enabling a realistic appreciation beforehand of the likely large‐scale bioperformance, and for guiding design. Finally, with the present understanding of the physical and biological aspects of bioprocesses, computational fluid dynamics (CFD) is unlikely to be able to significantly improve it.

The physical characterisation of a microscale parallel bioreactor platform with an industrial CHO cell line expressing an IgG4

“…When sparged, an up-pumping agitator loses very little if any power even in bacterial fermentations when the high aeration rate is introduced directly under the impeller, which then effectively disperses the air [10]. Here, the gases are not introduced directly under the impeller, though there is some dispersion, and the sparge rate is very low (at the speeds and flow rates normally used, the gas flow number, Q G /ND 3 = ~ 6 x 10 -4 ).…”

“…Though fully instrumented shake flask systems, designed to measure and potentially control pH and dO 2 levels online have become available [8,9], they are relatively complicated to operate and have not become routinely used by industry. Certainly, there are particular physical aspects such as the impact of fluid dynamic stresses from agitation and sparging as well as estimates of k L a that will be found in large scale stirred bioreactors which can only be effectively obtained in stirred bench scale bioreactors of a few litres and larger, hence, a critical aspect of upstream process development is still done in 1 -30 L stirred tank bioreactors [10].…”

Optimization of a multi‐stage, multi‐subunit malaria vaccine candidate for the production in Pichia pastoris by the identification and removal of protease cleavage sites

“…Even with P. pastoris KM71H, a strain that grows slowly on methanol (Mut s , methanol utilization slow phenotype), we reached a maximum OTR of 127 mmol L −1 h −1 at the end of the induction phase that resulted in a volumetric oxygen transfer coefficient (k L a) of 370 h −1 (data not shown). Taking into account that in large scale processes energy dissipation rates (ε T ) for high cell density fermentations are normally applied only up to 5 W/kg because of power restrictions (Nienow, ) and ε T correlates with k L a (Van't Riet, ) we already operated at the upper limit of the OTR (Fuchs and Ryu, ). Higher k L a values are only feasible when additional gassing with cost‐intensive pure oxygen is applied (Fu et al, ).…”

A simple model for power consumption in gassed and boiling stirred vessels

“…The gas flow rate, Q g , was varied from 1.4 to 14 × 10 −4 m 3 /s, which is equivalent to a gas flow number, Fl g , variation from 0.01 to 0.09. Fl g is defined as the ratio between Q g and ND 3. The upper limit of gas flow rate was due to the necessity to obtain clear images for the measurement of the cavity dimensions.…”

“…In the application of stirred vessels to gas–liquid systems, the proper estimation of the hydrodynamics of the two‐phase mixture in the turbine region is a critical task. In fact, the formation of ventilated cavities behind the blades strongly influences the apparatus performance, due to their impact on power consumption, gas hold‐up distribution, interfacial area, and interphase mass transfer rate 1–6. So far, several works concerned the study of gas cavity formation mainly in Rushton turbines, while limited information is available on other impeller types 7, 8…”