Abstract
Industrial microbial bioprocesses are an important subset of the world-wide chemical industry, contributing to the production of pharmaceuticals, chemicals, biocatalysts, and fuels alike. The conditions met by the microorganisms in industrial-scale reactors differ from those encountered at laboratory scale, decreasing titer, yield, and productivity achieved in the process.
Modeling has been used to characterize large-scale reactors, as experiments are challenging and costly. Thus far, the reactor models that admit physico-chemical heterogeneity have been numerical. Analytical, simple, and generalized models would be preferable for preliminary investigations. The first aim of this study was to develop a comprehensive but analytically solvable bioreactor model encompassing mixing times, concentrations of substrate and dissolved oxygen, and profiles of pH, temperature, and carbon dioxide. The second objective was to study means to improve the mixing and to homogenize the relevant quantities in large-scale bioreactors operated in the fed-batch mode.
To achieve the goals, analytical solutions to axial diffusion equations were developed and validated against a large set of literature data, bioreactors were characterized using analytical and numerical models, optimal feed point placements were derived, and a stable bacterial co-culture capable of homogenizing the substrate profiles experienced by the constituent strains was constructed. The derived feed point placements and co-cultures were also modeled and simulated in large-scale bioreactors.
As a conclusion to the conducted modeling, great improvements in mixing should be achievable if the optimal feed arrangements could be implemented. The shared carbon flow in a co-culture also homogenized the substrate profiles experienced by the microorganisms. The study demonstrated a simple yet spatially accurate model of heterogeneous bioreactors and also two potential approaches to recover the reactor performance by efficient homogenization.
Modeling has been used to characterize large-scale reactors, as experiments are challenging and costly. Thus far, the reactor models that admit physico-chemical heterogeneity have been numerical. Analytical, simple, and generalized models would be preferable for preliminary investigations. The first aim of this study was to develop a comprehensive but analytically solvable bioreactor model encompassing mixing times, concentrations of substrate and dissolved oxygen, and profiles of pH, temperature, and carbon dioxide. The second objective was to study means to improve the mixing and to homogenize the relevant quantities in large-scale bioreactors operated in the fed-batch mode.
To achieve the goals, analytical solutions to axial diffusion equations were developed and validated against a large set of literature data, bioreactors were characterized using analytical and numerical models, optimal feed point placements were derived, and a stable bacterial co-culture capable of homogenizing the substrate profiles experienced by the constituent strains was constructed. The derived feed point placements and co-cultures were also modeled and simulated in large-scale bioreactors.
As a conclusion to the conducted modeling, great improvements in mixing should be achievable if the optimal feed arrangements could be implemented. The shared carbon flow in a co-culture also homogenized the substrate profiles experienced by the microorganisms. The study demonstrated a simple yet spatially accurate model of heterogeneous bioreactors and also two potential approaches to recover the reactor performance by efficient homogenization.
| Original language | English |
|---|---|
| Place of Publication | Tampere |
| Publisher | Tampere University |
| ISBN (Electronic) | 978-952-03-3226-6 |
| ISBN (Print) | 978-952-03-3225-9 |
| Publication status | Published - 2024 |
| Publication type | G5 Doctoral dissertation (articles) |
Publication series
| Name | Tampere University Dissertations - Tampereen yliopiston väitöskirjat |
|---|---|
| Volume | 932 |
| ISSN (Print) | 2489-9860 |
| ISSN (Electronic) | 2490-0028 |