OUR data have been shown to correlate well with the heat generation rate from cellular metabolism in bacterial fermentation
processes, with 1 mole of oxygen being consumed during cellular metabolism to generate ~115 kcal of heat.21 The following heat-based OUR can be derived from the correlation and used for comparing to OUR determined from the mass spectrometer:
in which units for OUR are mmol/kg–hr and for Hmetab, J/kg–hr.
Case Study: Scale-Up of a Microbial Fermentation Unit Operation
In a case study of microbial fermentation, inlet gas flow rates and the molar concentrations of the outlet gases were measured
by a volumetric gas flow meter and an in-line mass spectrometer (Perkin-Elmer MGA1200). Data were captured continuously using
the PI data historian software, and the heat generation rate profile from the bioreactor was calculated as described above
and compared to the OUR profile. Relevant OUR and RQ profiles from three microbial runs performed in a 300-L production fermenter
are shown in Figures 5 and 6, respectively. The peak OUR value from the spectrometer was found to be approximately 300 mmol/kg/hr
(Figure 5), which matched the maximum value of the OUR based off the heat removal rate measurements. The RQ was found to be
~1.0 during the first half of the induction phase, which was also consistent with the target for an aerobic E.coli process.
Figure 5. Oxygen uptake rate (OUR) measurements using mass spectroscopy for a high-cell density microbial process for three
runs at 2,000-L scale. The peak OUR is observed to be 300 mmol/kg/hr.
CONCLUSION AND RECOMMENDATIONS
There are a number of process modeling tools that can be useful in facilitating scale-up and technology transfer of microbial
and mammalian processes. Application of scale-up predictors using empirical calculations should be used to assess large-scale
bioreactor capability with respect to gas flows and agitation speeds in relation to constant agitation (P/V), tip speeds,
superficial gas velocities, and predicted kLa. Information pertaining to integrated shear factors, shear rates, and Kolmogorov eddy size (B5m) also provides useful information
when comparing bioreactor conditions at different scales. Calculations of oxygen uptake rate (OUR) and carbon dioxide evolution
rate (CER) for high cell density microbial processes can be useful to ensure that the process requirements (peak OUR, peak
heat generation rate) can be met by the facility capabilities (maximum oxygen transfer rate, maximum heat removal rate).
Figure 6. Ratio of CER to OUR (RQ) measurements using mass spectroscopy for a high-cell density microbial process. The RQ
of ~1.0 is consistent with the target for an aerobic E.coli process.
Anurag S.Rathore, PhD, is the director of process development, Ken Green is a principal scientist, and Yas Hashimura is a senior engineer, all at Amgen,Inc., Thousand Oaks, CA, 805.447.1000, email@example.com
Greg Nyberg is a principal scientist at Amgen,Inc., Boulder, CO.