Issue link: https://www.e-digitaleditions.com/i/988595
24 Pharmaceutical Technology BIOLOGICS AND STERILE DRUG MANUFACTURING 2018 P h a r mTe c h . c o m The dual A340 configuration exhibited zon- ing with each impeller creating its own f low loop whereas the triple A340 configuration established a single flow loop and maintained a more uniform velocity distribution. Discussion Blending is an important task performed by agitators in bioreactors and is responsible for maintaining cell concentration, nutrient, and temperature uniformity. The visual blend testing suggested that all impeller configurations had statistically similar blend times to the dual down-pumping A320 base case. This re- sult was due to a large confidence interval for the base case. Additional trials or a more quantitative blend test (e.g., conductivity measurement with tracer ad- dition) would likely reduce uncertainty. When the agitator was removed from the sys- tem, the rising gas formed a bubble column, and the blend time increased by no less than 50%. The observed increase in blend time suggested that agi- tators should be used in bioreactors to reduce blend time and minimize the potential for cell concen- tration, nutrient, and temperature striations. Mass transfer is the primary task performed by agitators in bioreactors and is crucial for cell viability. Mass-transfer testing indicated the dual down-pump- ing A320 and triple up-pumping A340 configurations generated the highest mean values of k L a 20 and were statistically similar. Meanwhile, the single down- pumping A320 and dual up-pumping A340 configu- ration mean k L a 20 values trailed. The 3D CFD analy- sis demonstrated that the dual A320 and triple A340 cases created a single-flow loop and maintained more uniform velocity distributions than the single A320 and dual A340 cases. A more uniform velocity dis- tribution was inferred to coincide with a reduction in droplet coalescence, which ultimately led to increased k L a 20 values via increased interfacial area. Therefore, appropriately spaced multiple impeller configurations are recommended for use in bioreactors. The power and gas input sensitivity test conducted on the dual down-pumping A320 impeller configura- tion revealed that superficial gas velocity was a more pertinent factor than power per unit volume for biore- actors. However, when the agitator was removed from the system, the k L a 20 was even lower than when the gas flow rate was halved. This result meant that agitators are required to disperse gas to maximize mass transfer, and that the k L a 20 of an agitated bioreactor becomes more dependent on the gas flow rate than the agitator power input as long as the gas is adequately dispersed by the agitator (i.e., not flooding). Conclusion Bioreactors should be equipped with agitators to improve blending and mass transfer, consequently increasing the likelihood for cell viability and pro- cess success. Impeller configuration selection for the bioreactor agitator is an impactful decision. Al- though blend time was not found to be a function of impeller configuration in this study (likely due to experimental error), mass transfer was affected by the selection. This study was performed at a single tank diameter and liquid level, but bioreactor de- signs vary. Optimum impeller configuration varies with bioreactor design, but the benefit of properly spaced, multiple impeller configurations holds. References 1. D.A.R. Brown et al., "Chapter 4: Experimental Methods," in Handbook of Industrial Mixing, E. Paul et al., Eds. pp. 169–173 (John Wiley & Sons, Inc., 2004). 2. B. Gigas and T.A. Post. "Application of Batch Method k L a Determination in Mechanically Agitated Vessels," Presentation at the AIChE Annual Meeting (Florida, 1998). PT Bioreactor Design