Dynamic modeling and simulation of continuous airlift bioreactors

Abstract

For dynamic behaviors of continuous airlift bioreactors, a mathematical model based on a tanks-inseries model with back¯ow has been developed. The equations describing the dynamics of airlift bioreactors are material balances for micro-organism, substrate, dissolved oxygen and oxygen in gas-phase and heat balances. Nonideal mixing of liquid and gas phases is taken into account using a tanks-in-series model with back¯ow. The batch operation, startup operation and the consequence of plant failure were simulated and the effects of design and operating parameters for an airlift bioreactor on its dynamic behaviors were discussed. The concentration pro®les of micro-organism, substrate, dissolved oxygen and oxygen in gas-phase and the temperature pro®le in an airlift bioreactors and their dynamics were obtained. The computational results indicate that the transients of a chemostat in the case of bubble column bioreactor are slower compared with those in the case of airlift bioreactor. The proposed simulator is more precise as compared with models published previously in the literature and therefore provides more reliable and rational examination of continuous airlift bioreactor performance. List of symbolsA m 2 surface area for heat transfer b back¯ow parameter C kg/m 3 dissolved oxygen concentration C dimensionless dissolved oxygen concentration (=C=C in ) C p J/kgK constant-pressure speci®c heat Da Damko Èhler number (=Y X=S S in ÀDH=T a C p q) E a J/mol activation energy K C kg/m 3 saturation constant for oxygen K C dimensionless saturation constant for oxygen (=K C =C in ) K L a L 1/s volumetric gas-liquid mass transfer coef®cient K S kg/m 3 saturation constant K S dimensionless saturation constant (=K S =S in ) M number of stages in the riser N total number of stages in the airlift bioreactor P ratio of the number of stages for the gas phase to that for the liquid phase in the riser P d ratio of the number of stages for the gas phase to that for the liquid phase in the downcomer Q LC m 3 /s liquid circulation velocity Q G m 3 /s inlet gas¯ow rate Q G dimensionless inlet gas¯ow rate (=Q G =Q LC ) Q GC m 3 /s gas circulation velocity Q GC dimensionless gas circulation velocity (=Q GC =Q LC ) Q L m 3 /s inlet liquid¯ow rate Q O 2 kg/m 3 s speci®c rate of oxygen consumption Q O 2 dimensionless speci®c rate oxygen consumption (=Q O 2 s C =C in ) R recycle ratio (=Q LC =Q L ) R g J/kgK gas constant r x kg/m 3 s reaction rate S kg/m 3 concentration of substrate S dimensionless substrate concentration (=S=S in ) St C Stanton number for mass transfer (=K L a L s C ) St h Stanton number for heat transfer (=UA=Q C C p q) St o Stanton number for oxygen consumption Q O 2 Y X=S S in s C =C in T K temperature T a K coolant temperature T dimensionless temperature (=T=T a ) t s time t dimensionless time (=t=s C ) U W/m 2 K overall heat transfer coef®cient V L m 3 liquid volume X kg/m 3 concentration of micro-organism X dimensionless micro-organism concentration (=X=Y X=S S in ) Y Pa oxygen concentration in the gas phase Y dimension...

Oxygen mass transfer and hydrodynamics in a multi-phase airlift bioscrubber system

“…For larger Pe, the airlift system becomes closer to plug flow conditions and when Pe is smaller it indicates that the system is closer to being well-mixed. Within concentric tube airlift bioreactors, it is often assumed that the downcomer section is plug flow (Kanai et al, 2000;Znad et al, 2004), with the riser column being more well-mixed in nature. It should be noted that this study provides an overall Pe for the airlift vessel, which is commonly reported in studies comparing hydrodynamic parameters in systems under different conditions (Giovannettone and Gulliver, 2008;Merchuk et al, 1998).…”

Dynamic simulation of multicomponent gas separation by hollow-fiber membrane module: Nonideal mixing flows in permeate and residue sides using the tanks-in-series model

“…5. The fiber-side pressure change can be written as: P r,j = P r,j−1 + P j (16) where Fig. 7 illustrates dynamic performance of methane recovery with the operating and feed conditions given in Table 4.…”

“…An understanding of dynamic performance of gas membrane separation processes is essential for rational design and operation of industrial scale gas separation membrane modules. A dynamic model taking account of nonideal mixing can provide practical information for startup and shutdown operations, the transition between two stationary states, optimal design and operability studies besides a steady-state performance [16,17]. To the best of our knowledge, on the basis of the tanks-in-series concept the unsteady-state phenomena of the membrane separation have not yet been modeled in the open literature.…”

“…The axial dispersion model, which has been frequently applied to represent in mixing, can describe satisfactorily only mixing deviating not largely from the plug flow. Furthermore, a set of differential equations with initial and boundary conditions obtained for the axial dispersion model has to be solved by rather complicated numerical techniques [16,17]. A tanks-in-series model is more realistic and advantageous as compared with the axial dispersion model.…”

Model for a solid‐liquid airlift two‐phase partitioning bioscrubber for the treatment of BTEX

“…However, it is known that the downcomer section in airlift systems is nearly plug flow relative to the riser, and has been represented as 10 tanks-in-series by other authors in studies of airlifts of similar size to that used in the current study. 9,11 It has been stated that the number of tanks in series does not have a strong effect on simulation results, 10 and therefore, the number of tanksin-series used to represent all flow rates was selected to be 11. These 11 well-mixed tanks were distributed as shown in Fig.…”