Characterization of a pilot plant airlift tower loop reactor: III. Evaluation of local properties of the dispersed gas phase during yeast cultivation and in model media

The local properties of the dispersed gas phase (gasholdup, bubble diamater, and bubble velocity) were measured and evaluated at different positions in the riser and downcomer of a pilot plant reactor and, for comparison, in a laboratory reactor. These were described in Parts I and II of this series of articles during yeast cultivation and with model media. In the riser of the pilot plant reactor, the local gas holdup and bubble velocities varied only slightly in axial direction. The gas holdup increased considerably, while the bubble velocity increased only slightly with aeration rate. The bubble size diminished with increasing distance from the aerator in the riser, since the primary bubble size was larger than the equilibrium bubble size. In the downcomer, the mean bubble size was smaller than in the riser. The mean bubble size varied only slightly, the bubble velocity was accelerated, and the gas holdup decreased from top to bottom in the downcomer. In pilot plant at constant aeration rate, the properties of the dispersed phase were nearly constant during the batch cultivation, i.e., they depended only slightly on the cell concentration. In the laboratory reactor, the mean bubble sizes were much larger than in the pilot plant reactor. In the laboratory reactor, the bubble velocities in the riser and downcomer increased, and the mean gas holdup and bubble diameter in the downcomer remained constant as the aeration rate was increased.     

Gas–liquid dispersion in a fibrous fixed bed biofilm reactor at growth and non-growth conditions

There is limited data on gas dispersion characteristics of fixed bed biofilm reactors under growth and non-growth conditions. In this paper, the gas–liquid dispersion of a bubble bed packed with a fibrous structured packing for biofilm application is studied. The reactor is operated with Pseudomonas putida aimed at aniline degradation in wastewater. Gas hold-up and bubble size distribution are determined. Running gas–liquid reaction conditions as well as non-reactive flow gas hold-up and bubble size distribution in the presence of surface-active and viscous components were measured. The properties of the gas dispersion proved to be stabilized by the fibrous bed presence and showed improvement of the dispersion parameter by the packing. Gas hold-up was found to increase monotonously with the rise of gas superficial velocity and viscosity and with surface tension fall. Liquid superficial velocity showed marginal effect. Apart from showing high gas hold-up and low bubble size due to surface-active and viscous dissolved elements, the biochemical reaction did not pose any significant additional effect. In agreement with the expected lack of bubble coalescence and break-up in the highly ionic solution practiced, the population size distribution and average bubble size were found to vary with the major operation factors opposite to their gas hold-up contribution. Gas hold-up was correlated with the specific bubble-to-channel size ratio and further with the variables considered. An empirical equation is proposed that relates gas hold-up with all studied variables. Assuming geometric similarity of the prototype and the real vessels, the equation as well as its corresponding range of fluid velocities can be used for bioreactor design and scale-up. The results concerning the gas hold-up are shown to be comparable with previous studies of mesh wire packing.     

Absorption of oxygen in highly viscous newtonian and non-Newtonian fermentation model media in bubble column bioreactors

Summary Mean relative gas holdup, slip velocity, bubble size distribution, mean specific interfacial area, and volumetric mass transfer coefficient of oxygen were estimated in sparged columns 14 cm in diameter and 380 and/or 390 cm high with two different aerator types (porous plate and injector nozzle) in highly viscous Newtonian (glycerol solutions) and non-Newtonian (CMC solutions) fluids.For the Newtonian liquids the above properties were estimated as function of the viscosity of the liquid. For the non-Newtonian liquids they were determined as function of the fluid consistency index and flow behavior index. Significant differences between Newtonian and non-Newtonian systems appear. In Newtonian medium k_L a drops with increasing viscosity and already approaches a constant value at =40 cP. In pseudoplastic medium k_L a varies with the fluid consistency and flow behavior indexes in the entire investigated range.In both of these systems the primary bubble population changes into two or three populations along the reactor: the medium bubbles gradually disappear and small and large bubbles are formed.     

Bubble characteristics and mass transfer in an airlift reactor with multiple net draft tubes

A pilot scale airlift reactor with multiple net draft tubes was developed. The reactor, 29?cm in diameter and 300?cm height, had four modules of double net draft tubes. Bubble size, bubble number, gas holdup, and volumetric mass transfer coefficient were measured under different superficial air velocities. The air velocity had little effect on bubble size but had significant influence on bubble number. A bubble column was also investigated for comparison. The airlift reactor had a higher gas holdup and volumetric mass transfer coefficient than those in the bubble column. The draft tubes in the airlift reactor substantially improved the reactor performance.     

Generation of small bubbles and small bubble-liquid mass transfer in airlift reactors containing highly viscous liquids

The time-dependent gas hold-up is investigated during the aeration of the Saccharomyces cerevisiae suspension, the aqueous saccharose solutions and the glycerol solutions in the external loop airlift reactor. Due to the time-dependent bubble size distribution the fraction of the small bubble hold-up in the total gas hold-up decreases with an increase of the gas flow rate and with a decrease of the viscosity. The course of the accumulation process of the small bubbles is described by the first-order kinetic equation. The small bubble accumulation rate is investigated in the airlift reactor and the bubble column. It is showed that the small bubbles form and disappear exclusively in the riser of the airlift reactor. It is found that the small bubble-liquid mass transfer coefficient is several times larger than the overall oxygen transfer coefficient.     

Stochastic modelling of axial dispersion in external-loop airlift bioreactors

The paper presents a model of the motion of a particle subjected to several transport processes in connection with mixing in two phase flow. A residence time distribution technique coupled with a one-dimensional dispersion model was used to obtain the axial dispersion coefficient in the liquid phase, D_ax. The proposed model of D_ax for an external-loop airlift bioreactor is based on the stochastic analysis of the two-phase flow in a cocurrent bubble column and modified for the specific flow in the airlift reactor. The model takes into account the riser gas superficial velocity, the riser liquid superficial velocity, the Sauter bubble diameter, the riser gas hold-up, the downcomer-to-riser cross sectional area ratio. The proposed model can be applied with an average error of ᆨ.     

Characterization of a pilot plant airlift tower loop bioreactor. I: evaluation of the phase properties with model media

Investigations were carried out in a 9 m high, 4 m(3) volume, pilot plant airlift tower loop bioreactor with a draft tube. The reactor was characterized by measuring residence time distributions of the gas phase using pseudostochastic tracer signals and a mass spectrometer and by evaluating the mixing in the liquid phase with single-pulse tracer inputs. The local gas holdup and the bubble size (piercing length) were measured with two-channel electrical conductivity probes. The mean residence times and the intensities of the axial mixing in the riser and downcomer and the circulation times of the phases as well as the fraction of the recirculated gas phase were evaluated. The gas holdup in the riser is nearly uniform along the reactor. In the downcomer, it diminishes from top to bottom. The liquid phase dispersion coefficients, D(L), are smaller than those measured in the corresponding bubble columns. In the pilot plant with tap water the following relationship was found: D(Lr) = cw(SG) (n); with c = 203.4; n = 0.5;D(Lr)(cm(2) s(-1);) and W(SG)(cm s(-1)) where D(Lr) is the longitudinal dispersion coefficient in the riser and W(SG) is the superficial gas velocity. The gas phase dispersion coefficients in the riser of the pilot plant, D(Gr), are also enlarged with increasing superficial gas velocity, W(SG), however, no simple relationship exists. Parameter D(Gr) is the highest in the presence of antifoam agents, intermediate in tap water, and the smallest in ethanol solution.     

Bubble bed reactor: A reactor design to minimize the damage of bubble aeration on animal cells

A new bubble aeration system was designed to minimize cell killing and cellular damage due to sparging. The residence time of the bubbles in the developed bubble bed reactor was prolonged dramatically by floating them in a countercurrent produced by an impeller. The performance of the new reactor bubble aeration system, implemented in a laboratory reactor, was tested in dynamic aeration experiments with an without cells. An efficiency up to 95% in oxygen transfer could be achieved, which enables a much lower gas flow rate compared with conventional bubble aeration reactors. The low gas flow rate is important to keep cell damage by bubbles as low as possible. A laser light sheet technique used to find the optimal flow pattern in the reactor. The specific power dissipation of the impeller is a good measure to predict cell damage in a turbulent flow. Typical values for the power dissipation measured in the bubble bed reactor were in the range of 0.002 to 0.013 W/kg, which is far below the critical limit for animal cells. The growth of a hybridoma cell line was studied in cell cultivation experiments. A protein-free medium without supplements such as serum or Pluronic F68 was used to exclude any effect of cell-protecting factors, No difference in the specific growth rate and the yield of the antibodies was observed in cell grown in the bubble free surface aeration in the spinner flask. In contrast to the spinner flask, however, the bubble bed reactor design could be scaled up. (c) 1994 John Wiley & Sons, Inc.     

Investigation of the structure of two-phase flow fermentation model media in bubble-column bioreactors

Summary Electrical conductivity microprobes have been used to estimate the transverse variation of bubble size, local gas holdup and local specific gas/liquid interfacial area in bench scale bubble column bioreactors containing fermentation model media. Inserted O₂-electrodes and plane parallel windows alter the structure of the two phase flow. Even slight tilting of the column strongly influences the transverse profiles of the bubble size and local gas holdup. The larger bubbles are collected at the wall, where they can be redispersed. These observations open up new possibilities for the construction of bubble column bioreactors.     

Saccharomyces cerevisiae fermentations in a pilot scale airlift bioreactor: Comparison of air sparger configurations

Hydrodynamic and oxygen transfer comparisons were made between two ring sparger locations, draft tube and annulus, in a concentric pilot scale airlift reactor with a baker's yeast suspension. Sectional hydrodynamic measurements were made and a mobile DOT probe was used to characterise the oxygen transfer performance through the individual sections of the reactor. The hydrodynamic performance of the reactor was improved by using a draft tube ring sparger rather than the annulus ring sparger. This was due to the influence of the ratio of the cross sectional area of the downcomer and riser (A _D/A_R) in conjunction with the effect of liquid velocity and a parameter,C ₀, describing the distribution of the liquid velocity and gas holdup across the riser on the bubble coalescence rates. The mixing performance of the reactor was dominated by the frequency of the passage of the broth through the end sections of the reactor. An optimum liquid height above the draft tube, for liquid mixing was demonstrated, above which no further improvement in mixing occurred. The liquid velocity and degree of gas entrainment showed little dependency on top section size for both sparger configurations. Extreme dissolved oxygen heterogeneity was demonstrated around the vessel with both sparger configurations and was shown to be detrimental to the oxygen uptake rate of the baker's yeast. Dissolved oxygen tensions below 1% air saturation occurred along the length of the riser and then rose in the downcomer. The greater oxygen transfer rate in the downcomer than in the riser was caused by the combined effects of a larger slip velocity in the downcomer which enhancedk _La and gas residence time, high downcomer gas holdup, and the change in bubble size distribution between the riser and downcomer. The position of greatest oxygen transfer rate in the downcomer was shown to be affected by the reactor from the influence on downcomer liquid linear velocity. UCL is the Biotechnology and Biological Sciences Research Council sponsored Advanced Centre for Biochemical Engineering and the Council's support is greatly acknowledged.     

Hydrodynamics in bubble column bioreactors with fermentation broths having a yield stress

Summary The hydrodynamics in a bubble column bioreactor with fermentation broths having a yield stress are studied. Specifically, the liquid velocity at the reactor axis, the axial dispersion coefficient, and the gas hold-up are examined. The liquid velocity at the reactor axis and the gas hold-up are measured in a 40-1 bench-scale bubble column fermentor using carboxypolymethylene (Carbopol) aqueous solutions as simulated broths. Theoretical correlations for the liquid velocity at the reactor axis, the axial dispersion coefficient, and the gas hold-up are derived on the basis of an energy balance and the mixing length theory. The correlations are compared with the present data and a reasonable agreement is found. The theoretical predictions are also in satisfactory agreement with the re-examined data for actual fermentation broths which are Chaetomium cellulolyticum and Neurospora sitophila cultured in a 1000-1 pilot-plant scale airlift fermentor.     

Hydrodynamic study of an internal airlift reactor for microalgae culture

Internal airlift reactors are closed systems considered today for microalgae cultivation. Several works have studied their hydrodynamics but based on important solid concentrations, not with biomass concentrations usually found in microalgae cultures. In this study, an internal airlift reactor has been built and tested in order to clarify the hydrodynamics of this system, based on microalgae typical concentrations. A model is proposed taking into account the variation of air bubble velocity according to volumetric air flow rate injected into the system. A relationship between riser and downcomer gas holdups is established, which varied slightly with solids concentrations. The repartition of solids along the reactor resulted to be homogenous for the range of concentrations and volumetric air flow rate studied here. Liquid velocities increase with volumetric air flow rate, and they vary slightly when solids are added to the system. Finally, liquid circulation time found in each section of the reactor is in concordance with those employed in microalgae culture.     

Response of <Emphasis Type="Italic">Tisochrysis lutea</Emphasis> [Prymnesiophycidae] to aeration conditions in a bench-scale photobioreactor

The effects of superficial gas velocity (U_gr), gas entrance velocity (ν), and bubble size on the growth of Tisochrysis lutea was investigated in 600-mL photobioreactors operated with airlift pumps. Superficial gas velocities, calculated from measured air flow rates, ranging from 7 to 93 mm s^?1 were created using a 1.6-mm diameter syringe. We tested the effects of sparger velocity over a range of 2.48 to 73.4 m s^?1 and the effects of bubble size by using two styles of air stones and an open glass pipette, which created a bubble sizes in the range of 0.5 to 5 mm. We calculated oxygen mass transfer coefficient, k_La, values for all experimental conditions. Cell growth increased linearly with increased superficial gas velocity and decreased with increased sparger velocity. Results indicated that smaller bubble size leads to some initial cell damage, but after time, the increased gas transfer as reflected by the k_La value produced higher growth than larger bubbles. Two mechanisms were observed to correlate with cell damage in T. lutea: increasing velocity at the sparger tip and bubble bursting at the surface. These results demonstrate a method to test sensitivity of T. lutea to aeration, which is important for the design of airlift systems.     

Integration of Gas Enhanced Oil Recovery in Multiphase Fermentations for the Microbial Production of Fuels and Chemicals

In multiphase fermentations where the product forms a second liquid phase or where solvents are added for product extraction, turbulent conditions disperse the oil phase as droplets. Surface‐active components (SACs) present in the fermentation broth can stabilize the product droplets thus forming an emulsion. Breaking this emulsion increases process complexity and consequently the production cost. In previous works, it has been proposed to promote demulsification of oil/supernatant emulsions in an off‐line batch bubble column operating at low gas flow rate. The aim of this study is to test the performance of this recovery method integrated to a fermentation, allowing for continuous removal of the oil phase. A 500 mL bubble column is successfully integrated with a 2 L reactor during 24 h without affecting cell growth or cell viability. However, higher levels of surfactants and emulsion stability are measured in the integrated system compared to a base case, reducing its capacity for oil recovery. This is related to release of SACs due to cellular stress when circulating through the recovery column. Therefore, it is concluded that the gas bubble‐induced oil recovery method allows for oil separation and cell recycling without compromising fermentation performance; however, tuning of the column parameters considering increased levels of SACs due to cellular stress is required for improving oil recovery.     

Foam fractionation of globular proteins

Foam fractionation of bovine serum albumin (BSA) was studied as a model system for potato wastewater. The effects of feed concentration, superficial gas velocity, feed flow rate, bubble size, pH, and ionic strength on the enrichment and recovery of BSA were investigated in a single-stage continuous foam fractionation column. Enrichments ranged from 1.5 to 6.0 and recoveries from 5 to 85%. The feed concentrations were varied from 0.01 to 0.2 wt %, and enrichments were found to increase with lower feed concentrations. Enrichments also increased with lower superficial gas velocities and larger bubble sizes. At sufficiently low feed flow rates, enrichment was found to increase with an increase in the flow rate, eventually becoming insensitive to the feed flow rate at higher values. The pH was varied from 3.5 to 7.0 and ionic strength from 0.001M to 0.2M. The effects of pH and ionic strength were found to be coupled with bubble size. A minimum bubble size was found at pH 4.8, the isoelectric point of BSA, resulting in a minimum in the enrichment. Bubble size, and thus enrichment, was found to increase as the ionic strength decreased from 0.2M to 0.01M. Previous models(1,2) for the hydrodynamics of foam column were extended for a singlestage continuous foam fractionation column for the prediction of enrichment and recovery. The model assumed adsorption equilibrium, infinite surface viscosity, and bubbles of the same size. Though coalescence was formally accounted for in the model by considering bubble size as a function of foam height, calculations for the experimental runs were performed only for the case of no coalescence. Quantitative predictions of enrichment and recovery could not be made with a single representative bubble size because of the broad inlet bubble size distribution as well as broadening of the distribution as a result of coalescence. The experimental enrichments were higher and recoveries were lower than the model predictions, the discrepancy being more pronounced at lower feed concentrations because of increased coalescence. The higher enrichments are due to the predominant effect of internal reflux as a result of coalescence whereas the lower recoveries are a result of detrimental effects of broadening bubble size distributions.     

External-circulation-loop airlift bioreactors: study of the liquid circulating velocity in highly viscous non-Newtonian liquids

In order to obtain further information on the behavior and optimal design of external-circulation-loop airlift (ECL-AL) bioreactors, the liquid circulating velocity, gas holdup and average bubble diameter in the downcomer were studied using highly viscous pseudoplastic solutions of various types of CMC. A few comparative measurements also were made using a viscous Newtonian aqueous sucrose solution. For the liquid velocity measurements, an ultrasonic flow meter (Doppler frequency shift principle) was applied for the first time to the gas/non-Newtonian liquid dispersion in downward flow and satisfactory results were obtained. For viscous liquids, the circulating liquid velocity in the riser section of an ECL-AL (u(LR)) is shown to be dependent mainly on the downcomer-to-riser cross-sectional area ratio (A(d)/A(r)), the effective viscosity (eta(eff)) and the gas superficial velocity (u(GR)) as described by the following equation \documentclass{article}\pagestyle{empty}\begin{document}$$ u_{LR} = 0.23u_{GR};{0.32} (A_d /A_r);{0.97} \eta _{eff};{ - 0.39} $$\end{document} The circulating liquid velocity exerts opposing effects on the mass transfer and liquid-phase mixing performances of ECL-AL fermentors. Therefore, it is proposed that the optimum operating conditions for a given fermentation may be best achieved by means of independently regulating the circulating liquid velocity.     

Fluiddynamic characterization of airlift tower loop bioreactors with and without motionless mixtures

In a 60 l airlift tower reactor with outer loop fluiddynamical measurements were carried out in presence of three motionless mixer modules (Type SMV, Sulzer) in water, 0.6%, 0.9% and 1.2% CMC solutions. The global mixing properties were determined in the liquid and gas phases by tracers. Detailed local measurements revealed differences of local flow patterns and mixing properties in the unhindered riser and in the immediate wake of the mixer module. The local liquid velocities were measured by the pseudorandom heat pulse technique. The local bubble velocities were determined by the ultrasound Doppler technique. The dependence of liquid velocity, gas velocity and gas holdup on the superficial gas velocity were determined. The radial profiles of the mean liquid velocities and their standard deviations were evaluated in water and CMC solutions upstream and downstream of the motionless mixer modules. The radial profiles of the mean large-bubble velocities and mean small-bubble velocities and their standard deviations were determined as well. The influence of the presence of the motionless mixers in the riser on these properties was evaluated.List of Symbols Bo _L w _L L/D _L liquid Bo number - Bo _G w _G L/D _G ,gas Bo number - c tracer concentration - CMC carboxymelthylcellulose - D _G gas dispersion coefficient - D ₁ local liquid dispersion coefficient - D _L liquid dispersion coefficient - D _r riser diameter - d distance between transmitter and detector of the heat pulse probe - E _G gas holdup - HBV horizontal bubble velocity - HLBV horizontal large-bubble velocity - HSBV horizontal small-bubble velocity - L length of the column - l relative distance of the detector position from the tracer injection with respect to L - LBV large-bubble velocity - n number of circulations - OHBV overall horizontal bubble velocity - OVBV overall vertical bubble velocity - P power input - Pe ₁ w ₁d/D₁, local liquid Peclet number - SBV small-bubble-velocity - V _L liquid volume - VBV vertical bubble velocity - VLBV vertical large-bubble velocity - VSBV vertical small-bubble velocity - w _G gas velocity - w ₁ local liquid velocity - w _L liquid velocity - w _SG superficial gas velocity - w _SL superficial liquid velocity - mean residence time of the liquid in the riser H.M.R. thank the Verband der Chemischen Industrie for a Fund der Chemie scholarship     

Investigation of the structure of two-phase flow model-media in bubble-column bioreactors

Summary By applying photographic, electrical conductivity, and electrooptical methods, the transverse variation of bubble size and velocity, the local gas holdup, and the local specific gas/liquid interfacial area were estimated in a bench scale bubble-column bioreactor containing distilled water. The liquid velocity profile, the transverse turbulence intensity variations, and the turbulence energy dissipation scale were also measured by a hot film turbulence probe and constant temperature anemometer technique.     

Comments on the communication: on the calculation of shear rate and apparent viscosity in airlift and bubble column bioreactors

The determination of the shear rate in bubble column and airlift bioreactors is an important question from both the perspective of cell damage and the correlation of hydrodynamic parameters in non-Newtonian fluids in these contractors. In the context of correlating hydrodynamic parameters in non-Newtonian fluids, a common approach involves assuming that there exists an average shear rate in the column that is proportional to the superficial gas velocity. This average shear rate is then used to evaluate an effective viscosity of the non-Newtonian fluid that is subsequently used to quantify the fluid's rheological behavior in correlation. Contrary to a recent communication, this report illustrates that this approach, which has mainly been applied to bubble columns, can also be applied to external loop airlift contractors, replacing the superficial gas velocity by the superficial gas velocity by the superficial gas velocity supplied to the riser of the contractor. This extension is based upon consideration of the relevant characteristic velocity in the active zone (i.e., the riser section) of the reactor.     

Incorporating hydrodynamics into spatiotemporal metabolic models of bubble column gas fermentation

Gas fermentation has emerged as a technologically and economically attractive option for producing renewable fuels and chemicals from carbon monoxide (CO) rich waste streams. LanzaTech has developed a proprietary strain of the gas fermentating acetogen Clostridium autoethanogenum as a microbial platform for synthesizing ethanol, 2,3-butanediol, and other chemicals. Bubble column reactor technology is being developed for the large-scale production, motivating the investigation of multiphase reactor hydrodynamics. In this study, we combined hydrodynamics with a genome-scale reconstruction of C. autoethanogenum metabolism and multiphase convection–dispersion equations to compare the performance of bubble column reactors with and without liquid recycle. For both reactor configurations, hydrodynamics was predicted to diminish bubble column performance with respect to CO conversion, biomass production, and ethanol production when compared with bubble column models in which the gas phase was modeled as ideal plug flow plus axial dispersion. Liquid recycle was predicted to be advantageous by increasing CO conversion, biomass production, and ethanol and 2,3-butanediol production compared with the non-recycle reactor configuration. Parametric studies performed for the liquid recycle configuration with two-phase hydrodynamics showed that increased CO feed flow rates (more gas supply), smaller CO gas bubbles (more gas–liquid mass transfer), and shorter column heights (more gas per volume of liquid per time) favored ethanol production over acetate production. Our computational results demonstrate the power of combining cellular metabolic models and two-phase hydrodynamics for simulating and optimizing gas fermentation reactors.