Coupled hydrodynamic-metabolic simulations of
fermentation processes

C. Haringa, W. Tang, A.T. Deshmukh, M. Reuss, J.J.
Heijnen, H.J. Noorman and R.F. Mudde

The complex,
multiphase environment inside an industrial fermentor,
combined with a highly complex catalyst, makes fermentor
scale-up far from straightforward¹. Laboratory fermentors
are often poor representatives of their industrial counterparts, because they
fail to capture the significant concentration, temperature and shear rate
gradients existing on the industrial scale².

From the point of
view of the organism, the gradients encountered in industrial fermentation
environments translate to temporal fluctuations. Using a scale-down to scale-up
approach, various research groups experimentally subject organisms to temporal
fluctuations and assess their effect on the metabolic response³.
But, as the environment in large-scale fermentation processes is typically
poorly quantified, there exists a gap between the current experiments and
industrially relevant experiments⁴.

We propose to use
a CFD-approach towards quantifying the industrial environment, from the
organism's point of view, in order to design representative scale-down
reactors. In order to capture the complex response of the organism we use the
simulation approach pioneered by Lapin et al., where biomass is simulated as a Lagrangian phase coupled with a metabolic reaction model⁵.
A set of parameters representing their internal state is coupled to each
organism; the changes in internal parameters are coupled to the history of
external conditions that a particle encounters. Hence, each particle has a
response to the environment that is dependent on its path.

                    Figure 1: Left: Simulated
glucose concentration gradient in a 160 m³ penicillin fermentor.
                    Right: Metabolic model
used to study the biomass response to extracellular gradients.
                    The model contains 9
reactions (solid arrows). Dashed arrows show further dependencies.

We have adapted the
model of Lapin et al. for the organism Penicillium chrysogenum. In initial simulations only the influence
of an extracellular glucose concentration gradient has been considered. The
left of figure 1 shows a snapshot of the calculated local glucose concentration
in a 160 m³ high-cell density penicillin fermentation. The metabolic
model used to couple the extracellular glucose concentration to 4 intracellular
reactant pools is depicted graphically on the right of figure 1.

The use of a Lagrangian biomass phase allows to study the fermentation
process from the point of view of the organism. It allows to track the temporal
fluctuations encountered by an organism, as well as predict their metabolic
response to these fluctuations. Figure 2 shows how two organisms experience the glucose
concentration gradient. The uptake rate response of the organisms provides
insight in both the magnitude and frequency of external fluctuations
encountered. This data serves as a basis of design for experimental scale-down
simulators.

                    Figure 2: Simulated glucose uptake rate of versus
time for two organisms, travelling different
                    paths through the
reactor.

We use the
metabolic response to identify which fluctuations are most relevant, and which
process or metabolic bottlenecks may exist. Our model includes three-phase
(gas-liquid-biomass) flows, gas-liquid mass transfer, and multi-species
uptake/excretion. Furthermore heat transfer considerations and the impact of
hot and cold spot formation on the process will be investigated in the future.

Together, CFD
simulations and experimental data will be used to investigate the effect of
concentration gradients in fermentors on the
metabolism of organisms. These insights can be used for process optimization or
to identify metabolic engineering targets. Our aim is to combine CFD
simulations and the data from representative experiments towards a more
rational scale-up approach of fermentation processes.

This is a
multi-party research project, between DSM-Sinochem
Pharmaceuticals, TU Delft, East China University of Science and Technology and Guojia, subsidized by NWO and MoST.

1:            F.R. Schmidt, Optimization and scale-up of industrial
fermentation processes. Applied
microbiology and biotechnology, 68(4):425-435, 2005.

2:            S.-O. Enfors et al.,
Physiological responses to mixing in large scale bioreactors. Journal of Biotechnology, 85(2):175-185,
2001

3:            P. Neubauer and S. Junne, Scale-down simulators for metabolic analysis of
large-scale bioprocesses. Current opinion
in biotechnology, 21(1):114-121, 2010

4:            H.J. Noorman, An industrial perspective on bioreactor
scale-down: what we can learn from combined large-scale bioprocess and model
fluid-studies. Biotechnology Journal, 6(8):934-943,
2011

5:            A. Lapin et al., Modeling the
dynamics of E. coli populations in the three-dimensional turbulent field of a
stirred-tank bioreactor, a structured segregated approach. Chemical Engineering Science, 61(14):4784-4797, 2006