Design and performance characteristics of a continuous multistage tower fermentor

The design of a continuous multistage tower fermentor is described. The fermentor consists of five stages separated by perforated plates. Each stage includes mechanical mixing provided by two disc turbine impellers and has its own impeller shaft with bearing assembly and flexible coupling that enables the operation of an arbitrary number of stages. The normal operation of this system enables the co-current flow of gas and liquid, but the system can function countercurrently as well. The purpose of this study was to examine the hydrodynamic performance, i.e., the pressure gradient along the tower, the mixing time, gas holdup, the residence lime distribution of the continuous phase, the value of the backflow coefficient, and the oxygen transfer rate under conditions usually used during fermentations. From the interrelations between parameters influencing the proper performance of this system, an optimal design of plate geometry for processes requiring high oxygen transfer rate was formulated.     

Performance of a perforated plate column as a multistage continuous fermentor

Microorganisms were continuously cultivated in multistage column consisting of ten perforated plate sections to which medium and air were supplied concurrently from the bottom. At steady state the cell concentration in the various stages was gradationally differentiated from the bottom to the top in the direction of medium flow. RNA content per unit cell concentration at each sage was determined. The cells in the lower stages were higher in RNA content than those from the upper stages. Wash out was observed to occur in the column at dilution rates which do not result in wash out in a single stage chemostat system. A study of the flow characteristics revealed that the overall performance of the plate column was equivalent to that of a multistage system, when hole diameter and hole area to column cross sectional area ratio were properly selected. This was true even in highly aerated conditions. These results indicated that the perforated plates in the column hindred intermixing through the plates, and that each stage functioned as an independent stirred vessel. Industrial and research application of this type fermentor was discussed.     

Modelling mass transfer and agitator performance in multiturbine fermentors

A methodology for mathematically analyzing agitator performance and mass transfer in large multiturbine production fermentors is presented. The application of this approach provides a method for determining axial dissolved oxygen profiles under conditions of known mass transfer rates as a function of agitation-aeration characteristics. A stagewise approach is used which divides the fermentor into a series of mixing cells. This allows for each turbine and mixing cell to be individually optimized. The model also permits the determination of the mass transfer coefficient for each turbine based upon limited dissolved oxygen data. The primary limitation of this approach rests in the limited data and correlations available for multiturbine systems. The structure of the modelling approach can serve as a basis for testing single turbine correlations and adapting them to multiturbine systems. The step-by-step details of the mathematical analysis are presented and interpreted. A series of computer simulations demonstrate the effect of typical fermentor operating variables on the axial dissolved oxygen profile. Further simulations demonstrate the effect of modifying agitator blade numbers on the dissolved oxygen profile and agitator power requirement.     

Fermentors with the power introduced by aerating gas

In the present work methods and results of investigations on optimal construction of fermentor elements by multifactor planning at a random number of levels of any factor are presented. The optimal design of column fermentor of 0.02–100 m³ volume with sieve plates containing downcomers and with power introduced by aerating gas has been worked out. Several alternative designs have been compared by examining mass transfer rates, power requirements, and other operating characteristics. Several fermentor designs with the power introduced by aerating gas are discussed with respect to their performance for cultivating various microorganisms (yeast and bacteria).     

Design and physical characteristics of a multistage,continuous tower fermentor

A multistage tower laboratory fermentor has been constructed consisting of eight compartments separated by sieve plates. Flow of substrate and air is concurrent from the bottom to the top of the column. It, was hoped that this system could be used to reproduce, simultaneously on a continuous basis, eight distinct phases of a batch growth curve. It was believed that the extent of batch curve simulation would depend upon the character of hydraulic mean residence time of broth in the column and in the individual compartments. The expected relationship did not occur. Rather it was found that growth in the column involved residence time characteristics not only for the fluid but also for the microorganisms, and for the growth limiting substrate. Depending upon the column operation, these could be distinct and different. The purpose of this investigation was to study the residence time distribution (RTD) of the continous (fluid) and dispersed (microorganisms) phases for model systems as well as for a yeast fermentation. Various degrees of flow nonideality, i.e., fluid blackflow and dispersed phase sedimentation, were noticed. The former seems to be due to interaction of the concurrent gas and liquid flow; it is particularly dependent upon void area of the sieve plate holes. Sedimentation is probably a function of plate design as well as cell size and density. It wa concluded that for a particular plate design the gas hold-up wass controlled by superficial air velocity and was the main parameter governing the differences between dispersed and continous phase(Rt1). This conclusion was supported by a computeraided styudy utilizing a mathematical model of fluid flow to fit the growth kinetics and cell distribution observed experimentally throughout the fermentor. Some advantages of foam control in the tower fermentor by surface active compounds are mentioned. Also, suggestions are made for carrying out fermentations that have two liquid phases, such as a hydrocarbon fermentation. The possibility of closely approximating plug-flow conditions in the multistage tower fermentor, a necessary condition for batch growth simulation, is discussed from a practical point of view.     

Vortex behavior in the waldhof fermentor

This is the first part of a systematic study on the mechanisms of the Waldhof fermentor. When an agitator is rotating in a liquid, a vortex will develop on the surface. Air is dispersed into the liquid when the vortex is deep enough to reach the agitator. This is the basic mechanism of air dispersion in a Waldhof fermentor. In this work, experimental results, empirical correlations, and theoretical equations were obtained to relate the vortex depth to a number of physical factors including agitator diameter, agitator speed, tank diameter, liquid depth, liquid viscosity, and so on. The vortex depth was found a function of both Reynolds number and Froude Number.     

Disintegration of microorganisms in an industrial horizontal mill of novel design

A continuous high-speed horizontal colloid mill of novel design for use in the microbiological and food industries was tested for the disintegration of cells of Saccharomyces cerevisiae and Candida utilis. The mill consists of a horizontal vessel with round or oval cross sections fitted with a high-speed longitudinal agitator shaft on which are mounted disk agitators, alternating radially and obliquely to the shaft. The mill is partly filled with freely moving grinding elements which, during a continuous operation, are maintained in the vessel by a vibrating annular slot separator. Highly efficient cooling is provided by circulation of cooling fluid through a jacket surrounding the vessel as well as through the agitator shaft and disks. The radial agitator disks impart a radial motion to the grinding elements, while the oblique disks give rise to the axial movement of a substantial part of the elements. The crossing of paths thus achieved gives the mill a very high efficiency. Using a mill of 20 liter nominal capacity, the effects of agitator design, agitator speed, flow rate, and concentration of the cell suspension on the disintegration efficiency and heat production were studied. Ninety per cent of S. cerevisiae cells in a 15% suspension could be broken at a residence time of 2.5 min. The temperature rise did not exceed 8° C. The corresponding figure for C. utilis was 84%. The maximal flow rate was 400 liter/hr. Extrapolation indicates that available industrial mills of 300 liter capacity based on the same design can handle flows of 2000 liter/hr.     

Continuous alpha-amylase production using Bacillus amyloliquefaciens adsorbed on an ion exchange resin

Summary A method for the continuous production of extracellular alpha amylase by surface immobilized cells of Bacillus amyloliquefaciens NRC 2147 has been developed. A large-pore, macroreticular anionic exchange resin was capable of initially immobilizing an effective cell concentration of 17.5 g DW/1 (based on a total reactor volume of 160 ml). The reactor was operated continuously with a nutrient medium containing 15 g/l soluble starch, as well as yeast extract and salts. Aeration was achieved by sparging oxygen enriched air into the column inlet. Fermentor plugging by cells was avoided by periodically substituting the nutrient medium with medium lacking in both soluble starch and yeast extract. This fermentor was operated for over 200 h and obtained a steady state enzyme concentration of 18700 amylase activity units per litre (18.7 kU/l), and an enzyme volumetric productivity of 9700 amylase activity units per litre per hour (9.7 kU/l-h). Parallel fermentations were performed using a 2 l stirred vessel fermentor capable of operation in batch and continuous mode. All fermentation conditions employed were identical to those of the immobilized cell experiments in order to assess the performance of the immobilized cell reactor. Batch stirred tank operation yielded a maximum amylase activity of 150 kU/l and a volumetric productivity of 2.45 kU/l-h. The maximum cell concentration obtained was 5.85 g DW/l. Continuous stirred tank fermentation obtained a maximum effluent amylase activity of 6.9 kU/l and a maximum enzyme volumetric productivity of 2.73 kU/l-h. Both of these maximum values were observed at a dilution rate of 0.345 l/h. The immobilized cell reactor was observed to achieve larger volumetric productivities than either mode of stirred tank fermentation, but achieved an enzyme activity concentration lower than that of the batch stirred tank fermentor.     

Determination of the interparticular effective diffusion coefficient for CO(2) and O(2) in solid state fermentation

A simple experimental diffusion controlled fermentor (DCF), coupled with the use of a mathematical model based on mass balance, is proposed to measure the variation of the gas (CO(2) and O(2)) diffusion coefficients in solid state fermentation. The DCF was packed with an ion-exchange resin impregnated with a nutritive medium and inoculated with Aspergillus niger. The growth conditions in the DCF were very similar to those found in equipment operated with convective oxygen supply. The diffusion coefficient was shown to be very dependent on the biomass concentration within the solid state fermentor, and attained values of less than 5% of the molecular diffusion in air when the biomass in the fermentor reached 27 mg dry/g dry support.     

Control of packed column fouling in the continuous fermentation and stripping of ethanol

By recycling the contents of a 14 L fermentor through a stripping column to continuously remove ethanol and reduce product inhibition, continuous complete conversion of nutrient feed containing 600 g/L glucose was achieved in a small pilot plant. Ethanol was recovered from the carbon dioxide stripping gas in a refrigerated condenser, and the gas was reheated with steam and recycled by a blower. Productivity of ethanol in the fermentor as high as 15.8 g/L/h and condensate production of up to 10 L/day of almost 50% by volume ethanol were maintained for up to 60 days of continuous operation. Weekly washing of the column packing in situ was required to prevent loss of performance caused by attached growth of yeast cells, which restricts the gas flow rate through the stripping column. (c) 1996 John Wiley & Sons, Inc.     

Continuous production of extracellular protease by Bacillus subtilis in a two-stage fermentor

Growth and protease production of Bacillus subtilis in semisynthetic and synthetic media were studied in batch culture and in a two-stage, laboratory scale, continuous fermentor. The amount, of extracellular protease production was measured under specific growth conditions in both stages of the ferment or. At the dilution rates employed, the cells in the first stage of the ferment or produced negligible quantities of protease, and the culture primarily functioned as a continuous inoculum for the second stage of the fermentor. The culture effluent from the second stage of the fermentor contained extracellular protease, on the average, equal to 60 per cent, of the activity that had been found in the supernatant of a 48-hr batch culture grown in a medium having the same composition as that in the continuous fermentor. Extracellular protease was produced in semisynthetic medium by B. subtilis in the two-stage fermentor for as long as 20 days without culture degeneration. Additional studies indicated that continuous protease production could also be achieved in a synthetic medium. The RNA/ protein ratios of cells grown in semisynthetic medium in batch culture and in each stage of the two-stage fermentor were examined. There was a positive correlation between the amount of protease produced by the cells and their RNA/ protein ratio. Techniques employed for continuous production of protease by B. subtilis and the potential use of the method for investigating the control of secondary metabolite synthesis are discussed.     

Rapid large-scale growth of Helicobacter pylori in flasks and fermentors.

We developed procedures for large-scale cultivation of Helicobacter pylori in flasks and fermentors. Flasks incubated closed under a microaerophilic gas phase with a cotton plug covered by a plastic bag, followed by removal of the bag after 8 h, gave excellent growth. Growth in a 10-liter fermentor led to excessive foaming if the medium was sparged with gas; silicone- or polyglycol-based antifoaming agents were severely inhibitory. Use of fermentor surface gassing, first with a microaerophilic 6% oxygen gas mixture, then with air, and then with 95% oxygen, allowed the culture to grow to an A600 of 2.5 in < 24 h. This method was modified for scale-up to a 100-liter fermentor.     

Effect of interstage mixing in multistage culture systems on continuous biomass production

The significance of the interstage mixing on important process parameters of biomass production was studied. The experiments were performed in a multistage tower fermentor and in fermentors in series. The interstage mixing effect can be evaluated under conditions of geometrical similarity, identity of oxygen transfer rate, and identity of dilution rate per stage in the individual stages of both culture systems. Candida utilis was cultivated on a synthetic medium with ethanol as the sole carbon and energy source in the concentration range 10–100 g/liter. Dilution rate, temperature, and pH in each stage of both culture systems were kept constant. It was demonstrated that in the multistage tower fermentor the definite backflow which ensures the permanent reinoculation by adapted cells significantly decreases the inhibitory effect of higher ethanol concentrations on the cell growth and on the rate of ethanol utilization.     

Spore production ofPenicillium roqueforti in fermentors filled with buckwheat seeds: batch and semi-continuous cultivation

Summary The behaviour of a drum fermentor and a column fermentor during the sporulation ofPenicillium roqueforti on buckwheat seeds is presented. The main problem encountered during the course of a cultivation is the free water released (about 0.1 ml/g dry matter) which must be removed from the medium. The rotation of the drum fermentor may disturb the growth and the sporulation. The column fermentor thus represents the best way to perform batch cultivation of the fungus: 10⁹ external spores/g dry matter are obtained.Semi-continous cultivation, with sequential emptying and filling, is performed in 1-liter bottles. This kind of cultivation may give a maximal average productivity close to 9.2·10⁶ external spores/g dry matter per hour. A drum fermentor, rotading only when emptying and filling, could represent an alternative to perform this kind of cultivation.     

Protease production by fermentation of fish solubles from salmon canning processes

Production of protease by fermentation, using Sorangium 495, of a substrate based on condensed fish solubles is demonstrated. The effects of carbohydrate addition, pH, fish solubles concentration, scale-up, agitation, and air flow rate on protease yields are described. While the fish solubles medium alone could give rise to measurable yields of protease, these were, at worst, doubled when 1% glucose was added to the medium. pH 7 was optimal for protease yield. Although the concentration of fish solubles in the basic medium showed no significant effect on cell yield, maximum protease yield was observed at a protein concentration equivalent to 3.85 mg/mL of bovine serum albumin. Protease production rates decreased as medium protein fermentor showed no significant effect on maximum protease yields. The effects of agitator speed and air flow rate on protease yield suggested that the rate of O2 transfer from air to medium could limit the rate of protease production. It was also noted that protease production is not growth associated.     

Fluid dynamics in bubble column bioreactors: experiments and numerical simulations

Detailed measurements of multiphase flows that prevail in bioreactors tell us that different transport mechanisms are dominating on different observation scales. The consequence in terms of reactor modeling is that the processes on different scales can be treated independently. A three-dimensional, dynamical model is presented that can be used to describe bubble column bioreactors on the reactor scale. It is based on the Navier-Stokes equation system. On the next smaller scale, the dynamics of the gas phase is described in a Lagrangian way, by tracking many bubble clusters or bubbles simultaneously on their way through the reactor. The model is capable of describing bubble columns of different size and can thus be used for scale-up. Its performance is demonstrated with a production-scale beer fermentor. (c) 1996 John Wiley & Sons, Inc.     

Agitator speed and dissolved oxygen effects in xanthan fermentations

Agitation speed affects both the extent of motion in Xanthan fermentation broths because of their rheological complexity and the rate of oxygen transfer. The combination of these two effects causes the dissolved oxygen concentration and its spatial uniformity also to change with agitator speed. Separating these complex interactions has been achieved in this study in the following way. First, the influence of agitation speeds of 500 and 1000 rpm has been investigated at a constant nonlimiting dissolved oxygen concentration of 20% of air saturation using gas blending. Under these controlled dissolved oxygen conditions, the results demonstrate that the biological performance of the culture was independent of agitation speed as long as broth homogeneity could be ensured. With the development of increasing rheological complexity lending to stagnant regions at Xanthan concentrations >20 g/L, it is shown that the superior bulk mixing achieved at 1000 rpm, compared with 500 rpm, leading to an increased proportion of the cells in the fermentor to be metabolically active and hence higher microbial oxygen uptake rates, was responsible for the enhanced performance. Second, the effects of varying dissolved oxygen are compared with a control in each case with an agitator speed of 1000 rpm to ensure full motion, but with a fixed, nonlimiting dissolved oxygen of 20% air saturation. The specific oxygen uptake rate of the culture in the exponential phase, determined using steady-state gas analysis data, was found to be independent of dissolved oxygen above 6% air saturation, whereas the specific growth rate of the culture was not influenced by dissolved oxygen, even at levels as low as 3%, although a decrease in Xanthan production rate could be measured. In the production phase, the critical oxygen level was determined to be 6% to 10%, so that, below this value, both specific Xanthan production rate as well as specific oxygen uptake rate decreased significantly. In addition, it is shown that the dynamic method of oxygen uptake determination is unsuitable even for moderately viscous Xanthan broths. Copyright 1998 John Wiley & Sons, Inc.     

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.     

Xanthan production in bubble column and air-lift reactors

A bubble column (0.05 m(3)) and an air-lift fermentor (1.2 m(3)) were used for the production of the exocellular microbial polysaccharide xanthan with Xanthomonas campestris in a synthetic medium. Upon oxygen depletion in the liquid, the xanthan production rate dropped sharply and then became a linear function of the oxygen transfer rate. The volumetric mass transfer coefficients for oxygen conformed to the correlation of Suh et al. Using this correlation in combination with the model for xanthan batch fermentation suggested by Peters et al., the xanthan fermentations in the bubble column were well described. The model also correctly predicted the time course of the molecular weight of the polysaccharide even when a complex medium was used. In the air-lift fermentor, however, the xanthan production rate and the xanthan yields with respect to oxygen and glucose were lower than expected at the overall oxygen transfer rate. The poor performance of the air lift was traced back to the lack of any oxygen supply in the downcomer.     

Measurement of dissolved carbon dioxide.

Several probes for measuring dissolved carbon dioxide (CO2) concentration were installed in a 68-litre fermentor and their effectiveness compared. Submerged silastic rubber tubing gave reproducible results over a wide range of operating conditions and was generally superior to all other probes evaluated. The silastic rubber probe was used to compare the partial pressure of CO2 in viscous fermentation media with that in the fermentor exhaust gas. No significant difference was found. Results show that determination of the CO2 partial pressure in the exhaust gas gives an excellent approximation of the partial pressure of dissolved CO2 in the liquid medium, eliminating the need for measurement of CO2 concentration in the broth.