Date of Award
August 2020
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Electrical and Computer Engineering (Holcomb Dept. of)
Committee Member
Richard E Groff
Committee Member
Sarah Harcum
Committee Member
Adam Hoover
Abstract
Biopharmaceuticals produced in recombinant cells, such as antibodies or gene therapies, have the potential to treat diseases that were previously chronic, such as cancer or immune deficiencies. These products are produced in large scale fermentations where the cells are often limited by one nutrient, usually glucose or some other sugar. This allows one to use a single variable to control the growth of the cells. A feed controller has been developed that relies on feedback information about the oxygen consumption of the cells. Currently, there is not a standard means to calculate or measure the oxygen consumption of a cell culture online in a bioreactor. The oxygen transfer rate is directly related to the oxygen uptake rate of the cells. This quantity can be estimated by comparing the input and offgas oxygen concentrations, however, this estimation does not contain high-frequency information about the oxygen transfer rate. An online estimator has been developed that uses measurements of the oxygen concentration in the input gas, in the fluid of the bioreactor, and in the offgas exiting the bioreactor to predict the oxygen transfer rate to the fluid of a bioreactor. Recursive least squares is implemented to fit a model for the mass transfer coefficient, and from this, the oxygen transfer rate can be calculated. By using the dissolved oxygen measurements with the offgas measurements, the estimator is able to capture finer details about the oxygen transfer rate. The estimator was tested against a simulation of the Xu model for Escherichia coli and data from previous laboratory fermentations. The recursive least squares estimator leads the signals produced by standard industry calculations. The estimator developed can be applied to a wide variety of systems and is straightforward to tune. In simulation, the estimator was able to accurately track the mass transfer coefficient for oxygen and the oxygen transfer rate online and continuously. The performance of the estimator was shown to be acceptable for use in an online controller.
Recommended Citation
Trout, Marshall, "Development and Implementation of an Online Estimator for Oxygen Transfer in Bioreactors Using Recursive Least Squares" (2020). All Theses. 3368.
https://open.clemson.edu/all_theses/3368