Date of Award:

5-2010

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Biological Engineering

Committee Chair(s)

Timothy Taylor

Committee

Timothy Taylor

Committee

Ronald Sims

Committee

Heng Ban

Abstract

Accurate experimental oxygen mass transfer coefficient, a measure of how quickly oxygen travels from a gas bubble to the bulk liquid, is important for comparing performance and for evaluating the oxygen transfer capability of a fermentor. Delays in probe response and changing gas volumes upon start-up of gassing affect the accuracy of oxygen transfer measurements. To mitigate these inaccuracies, a standard correction procedure for oxygen mass transfer data was established for highly oxygenated, well-mixed fermentation systems. Probe response time correction was generated by applying a second-order response model to dissolved oxygen probes and shown to be effective within 4%. By using a derived model for transient volume rise, the effect of changing gas volume at start-up was shown to cause very minimal error (1-2%) in kLa.

The unsteady-state method of kLa determination was used to compare design aspects of a hypothetical fermentor, including gas sparging devices and locations, baffle geometries and quantities, and impeller configurations. It was shown that locating the sparging device in the center of the tank, directly below the drive shaft and bottom impeller, is optimal for oxygen mass transfer. Sparger type was shown to have little effect on oxygen mass transfer values, although an open-pipe sparger was shown to provide slightly more oxygen mass transfer than a ring sparger. The use of rounded baffles in place of traditional rectangular baffles resulted in a 67-80% decrease in oxygen mass transfer coefficient. A comparison of three and four traditional baffles showed that three baffles produced a higher oxygen mass transfer than four. Correlation of baffle ratio and oxygen mass transfer coefficient indicated that the optimum baffle ratio is approximately one. Radial impellers were observed to provide better mixing, and thus higher oxygen mass transfer coefficients than axial impellers. In seven of ten comparisons, an impeller quantity ratio of 1.33 instead of 1.00 provided significant improvement in kLa. Additionally, only two of ten comparisons showed a difference between traditional Rushton turbine impellers and Smith turbine impellers, indicating that the difference in oxygen mass transfer capability of the two is negligible.

Checksum

4a7a5b05c5d938ece43ab1b0fe9bd6a3

Comments

This work made publicly available electronically on August 30, 2010.

Share

COinS