Bioreactor Modeling
Bioreactor Modeling
Bioreactor Modeling
6. Bioreactor Modeling
OBJECTIVES
The word 'model' (derived from the Latin word " modus") means to measure some
quantity in a sizeable representation of a planned or existing "object". Modeling is
the mathematical representations of a system. More precisely, it is the
representation of the phenomena in a set of mathematical equations. A
comprehensive list of different tasks of modeling is given in Figure 6.1.
For bioreactor operation, control of heat is one of the important factors. This is
a combination of heat generated by metabolic reaction as well as the
contribution from mechanical sources (like agitation, etc.). The control of heat
in the bioreactor can be done by controlling coolant temperature.
Generally, models are developed from the first principles of physics and
chemistry. They are classified into following categories (Thilakavathi et al., 2007).
They are developed from the principles of basic sciences. For example, the Henri-
Michaelis-Menten kinetic model is derived from the first principles.
This is a compromise between type (a) and type (b) models. One or two
parameters can be evaluated from plant data. The modification of Han and
Levenspiel model for the growth of Trichoderma harzianum growing on chitin to
produce chitinase is an interesting example (Felse, 1999). Other examples are
mixed vessels in series, plug flow with dispersion, etc. These models are
proposed, but the validity of the models on a production scale is questionable.
The basis of these models is RTD.
These models can be applied under dynamic (unsteady state) and steady state
approaches.
Therefore,
(6.1)
Unsteady state (Dynamic) approach
(6.2)
(6.3)
Equating the above two expressions for the rate of energy accumulation,
(6.4)
No assumption, except the constant specific heat, has been made for the
variables on the right hand side of the equation.
It assumes variation with time but not with space. For example, in unsteady state
CFSTBR, concentration varies with time, but not with space. The model is
represented by ordinary differential equations (ODEs).
In micro concept, a single unit operation is a system having inputs, states, and
outputs.
Simulation helps to understand the system behavior and must make sense to be
useful for analysis.
The mathematical tools that are used to study the problems of linear dynamic
systems are known as system analysis techniques.
1. Laplace transform
It is applied to analyze single, linear and the nth order ODEs. Annexure to this
chapter summarizes a few Laplace transform of standard useful functions.
2. State-space techniques
This is used to analyze the behavior of multiple first order linear differential
equations. If the differential equations are non-linear, they can be linearized at a
desired steady state point.
6.4. MODELING PRINCIPLES
Material/energy balances
Force/momentum balances
Real model should include the important dynamic effects not that much
complicated than needed and keeps the minimal number of equations and
parameters. The model equation must provide unique relationship among all
input and output variables. Degree of freedom is one of such approaches.
Cases:
10. Solve the necessary equations and do the parameter sensitivity analysis.
(6.5)
At constant density,
(6.6)
The energy balance equation is
(6.7)
where, Ti, Tref , and T are the inlet, reference and outlet temperatures,
respectively.
respectively.
Equations (6.5) and (6.7) are the dynamic model systems. One needs to specify
the conditions for volume and temperature, input variables (wi, w, Q and T) as
a function of time and relationships of an algebraic function (ρ, c).
From Equation (6.7)
(6.8)
Since
Dividing by ρcV
(6.9)
The reaction is .
Solution
2. Assumptions
a. Both the reactor contents and cooling jacket are perfectly mixed.
d. Jacket wall has negligible thermal inertia (assuming thin wall in the
jacket).
(6.10)
(6.11)
where k is the reaction rate constant. kCnA = rate of reaction /unit
volume.
c. Energy balance equation for the reactor is described below.
(6.12)
where h is the enthalpy Q is the heat removed from the system.
Balances for jacket
i. Energy balance
Enthalpy accumulation,
(6.13)
(6.14)
where AH is the heat transfer area.
4. Other balances
i. Expression for rate constant
(6.15)
(6.16)
where kv is valve coefficient and (V – Vmin) drop in volume across the
valve.
This is not required if a pump is used for feeding.
Assuming enthalpy
(6.17)
(6.18)
(6.19)
(6.20)
(6.21)
Equation of motion
Rate of reaction
Thermodynamic equilibrium
(6.22)
Component balance is
(6.23)
At steady state,
(6.24)
At steady state
(6.25)
(6.26)
(6.27)
where,
(6.28)
Then
(6.29)
(6.30)
This comes from a concept of two reactors in series. In this case, the overall
transfer function of the system will be the product of each transfer function. The
scheme is given in Figure 6.5.
where
k, R2 are gain.
H1 and H2 are height of the fluid in the first and second reactor, respectively.
(6.31)
(6.32)
(6.33)
(6.34)
(6.35)
where,
(6.36)
Rearranging,
(6.37)
(6.38)
For step response, values of the roots depend on ξ, which are discussed by
Coughanour (1991) with proper Equations (6.39)–(6.43).
(6.39)
The response of the system is critically damped.
(6.40)
For ξ > 1, the roots of the equation are real, resulting in overdamped or non-
oscillatory response.
(6.41)
(6.42)
(6.43)
Simple model is better to understand and is easily applied in modeling. This may
not accommodate enough information of the system. For this reason, models are
sometimes complicated which results in cumbersome solution. Such type of
diffcult model may not find real applications. For examples, most of the models
related to fungal morphological parameters are complicated models.
Step 1: To get first estimate of the values of the parameters in the model.
Step 2: Substitute these parameter values into the model and solve.
Step 3: Vary the values of the parameters and record the influence on the most
important output variable of the model.
Kossen and Oosterhuis (1985) defined parameter sensitivity (PS) in the following
way.
Example 6.3 Batch growth with oxygen limitation
Considerations are listed below.
2. Oxygen is transferred from the gas phase to the liquid phase by inter-phase
mass transfer.
(6.44)
where
(6.45)
Upon integration
(6.46)
This equation suggests that the growth is linear. In real system, exponential
growth occurs as the growth is limited by reactants.
(6.47)
Reactant balance equation is
(6.48)
Oxygen balance equation is
(6.49)
where,
(6.50)
As
(6.51)
From oxygen balance, we get
(6.52)
where C* = Concentration of oxygen in the bulk (0 ≤ C* ≤ CL). CL is the
saturation level of oxygen in reaction phase.
∴ From Equations (6.47), (6.49), and (6.52), Equation (6.53) is obtained
(6.53)
Dialysis culture
Cell concentration,
(6.55)
(6.56)
(6.57)
Under quasi-steady state conditions,
(6.58)
In a fed-batch reactor, the rate of consumption of reactants = rate of reactant
addition.
Under quasi-steady state condition,
(6.59)
The dilution rate, D and, therefore, μ decreases with time in a fed batch
culture.
To compare with other basic reactors let us consider the following figures.
Figure 6.7. μ vs. time plot for various reactors. (a) Batch bioreactor, (b)
CFSTBR (c) Fed-batch bioreactor.
Cell Concentration
(6.60)
where
where
Product Concentration
Two expressions are suggested here.
a. YP/S is constant
Therefore,
(6.61)
and hence,
(6.62)
Total product is
(6.63)
(6.64)
We know
(6.65)
Therefore, the product per unit volume of culture from Equations (6.61) and
(6.63)
(6.66)
where,
6.11. EXERCISES
The CSFTBR is filled with gases. The forward and reverse reactions are of
different orders. The mole fraction of A is ξ. Both pressure and ξ vary with
time. The product stream flows into another reactor through a restriction valve.
The second reactor is maintained at pressure of P2 (absolute). The feed stream
has a density of ρo and a mole fraction of yi of reactant A. Assume both the
gases are perfectly mixed and the system is isothermal. Molecular weights of A
and B are MA and MB, respectively. Calculate degrees of freedom in the system.
6.2 Consider the statement given in worked out Example 6.2, but flow in the
jacket is plug flow. Write the model equations and estimate the degrees of
freedom.
where subscript 'o' for initial/input conditions and R for recycle condition.
where
Discuss on the difficulties you expect in the estimation of the yield coefficients.
6.12. REFERENCES
Coughanour DR (Ed.) (1991) Process System Analysis and Control, 2nd edn.,
McGraw-Hill International edn.
6.13. APPENDIX 6
Let us consider a time domain function f(t). Laplace transform of f(t) = L(f(t)).
6.13.1.1. APPLICATION
Laplace transform is used for linear operation, to solve linear partial differential
equations, exponential function, step function, time delay, integral and impulse
functions. Some standard Laplace transforms are summarized in Table 6.1 to help
solve the problems.
Table 6.1. Some standard Laplace transforms
Citation
EXPORT
Professor Tapobrata Panda: Bioreactors: Analysis and Design. Bioreactor
Modeling, Chapter (McGraw-Hill Professional, 2011), AccessEngineering
This product incorporates part of the open source Protégé system. Protégé is
available at http://protege.stanford.edu//