Includes bibliographical references. / In this thesis it was demonstrated that the current L-lysine fermentation technology can be enhanced by continuously withdrawing spent medium while recycling the biomass in the culture suspension to the bioreactor. The biomass in the reactor outlet stream is separated from the spent medium using cross-flow filtration. The objective of this thesis was to study, understand, model and optimise the performance of the L-Lysine fermentation with biomass retention using cross flow filtration. Following a review of the factors affecting cross-flow filtration and modelling approaches available, the most suitable filtration flux estimation equation was selected. The impact of filtration on microbial performance was assessed and approaches to modelling the lysine fermentation overviewed, leading to the selection of an appropriate model. Thereafter a rigorous approach to the optimisation of the biomass recycling system for lysine production was conducted and experimentally validated. A generic form of Hermia's blocking laws was found to be well suited to the description of the initial stages of cross-flow microfiltration. A constant term (the pseudo steady state flux) has been included to provide a semi-empirical correlation of the cross flow filtration flux. The pseudo steady state flux is based on Darcy's law and a combination of the shear induced diffusion and surface transport models. The presented model adequately described the experimental data. The qualitative effects of the increased hydrodynamic shear stress experienced in the filtration recycling loop on the growth, metabolism and morphology of Corynebacterium glutamicum cells have been investigated. It was found that the cell volume increases under increased hydrodynamic shear although increased shear does not alter the cell shape. The apparent specific growth rate, the yield of biomass from threonine and the specific lysine productivity of the cells exposed to hydrodynamic shear in the filtration system decreases at increased hydrodynamic shear. Using a bioreaction network (BRN) model, it was postulated that increased hydrodynamic shear causes a shift in cellular metabolism from oxidative phosphorylation to substrate level phosphorylation and glycolysis. Furthermore it is postulated that increased hydrodynamic shear causes an increase in the flux of carbon towards the cell wall to either repair or strengthen the cell wall. Fermentation models were developed based on mass and volume balances coupled to either a set of empirical correlations of the cellular metabolism developed from experimental data or a bioreaction network. The impact of filtration-associated hydrodynamic stress on the cellular metabolism was modelled based on a linear relationship between the metabolic impact and the average energy dissipation rate per unit cell mass. A critical average energy dissipation rate was identified below which no impact on the fermentation performance relative to conventional batch fermentations was detected. The fed batch fermentation with biomass recycling using cross flow filtration was optimised using an equation-based dynamic simulation package (gPROMS). The predicted optimum represented a 26% reduction in variable cost of production compared to the conventional fed-batch fermentation technology (R14.50/kg vs. R10.65/kg). The predicted optimum was physically achievable and the experimental results obtained when a fermentation was conducted at the optimal conditions corresponded well with that predicted by the proposed model. The model parameters were re-established for the industrial lysine producing strain (AEC94). At the optimum conditions the model predicted a 12% improvement in variable cost of production while a 14% improvement was realised from experimental data.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/8786 |
Date | January 2002 |
Creators | Potgieter, Thomas |
Publisher | University of Cape Town, Faculty of Engineering and the Built Environment, Centre for Bioprocess Engineering Research |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Doctoral Thesis, Doctoral, PhD |
Format | application/pdf |
Page generated in 0.0085 seconds