Large scale fermentations face challenges in mixing and mass transfer as well as in the design and construction of the equipment. Scale-up from laboratory and pilot scale experiments is difficult because different phenomena-such as mixing times and mass transfer conditions-scale in a different way. We study the effect of mass transfer, reactor type and scale on the growth of Pichia pastoris yeast. Batch cultivation experiments monitoring the cell growth and ethanol formation are conducted in laboratory scale in two reactor types-stirred tank and an Outotec OKTOP®9000 draft tube reactor. Model for the yeast growth-including respirative and fermentative metabolism and the effect of dissolved oxygen-is formed based on literature. For scale-up studies, the growth model is used along with one dimensional reactor model that accounts for liquid mixing, gas phase dynamics and local gas holdup and mass transfer coefficient. By using a realistic growth model along with the reactor model, the simulated effects of scale-up are presented in terms of cell yield. A decrease in yield is noticed due to oxygen depletion in gas and insufficient liquid mixing. Potential improvements are related to the gas handling capacity and liquid mixing of the reactor.

Gas-liquid mass transfer is an important phenomenon in aerobic microbial cultivations, and the mass transfer performance of an industrial reactor strongly affects the overall process economics. Traditionally, industrial and laboratory bioreactors have been agitated with flat disc turbines (Rushton turbines) although there are many variants to this design. In addition, pneumatically agitated reactors such as bubble columns and airlift reactors have been studied and used by the industry. In this study we utilize an agitated draft tube reactor in cell cultivation and mass transfer studies. A standard reactor geometry agitated with three Rushton turbines was compared to Outotec OKTOP®9000 reactor which is a draft tube reactor agitated with a single impeller located just below the draft tube. The experiments included cell cultivation with Pichia pastoris yeast, determination of overall mass transfer coefficient by dynamic gassing in method and measurement of local gas holdup by electrical impedance tomography (EIT). In addition, agitation power was estimated from the power consumption of the DC-motor. OKTOP®9000 reactor was found to have higher kLa values than the STR with similar agitation power and gas flowrate. The overall gas holdup was similar in both geometries at same power inputs and gas flow rates. However, some significant differences were detected in the distribution of gas phase between the two geometries especially in the axial direction. Also changes in the gas dispersion regime can be detected from the spatial distribution of the gas holdup measured by EIT. The cell cultivation experiments showed the applicability of this type of agitated draft tube reactor to bioprocesses although a direct comparison with Rushton geometry is not straightforward.

In the enzymatic hydrolysis of cellulose, several phenomena have been proposed to cause a decrease in the reaction rate with increasing conversion. The importance of each phenomenon is difficult to distinguish from batch hydrolysis data. Thus, kinetic models for the enzymatic hydrolysis of cellulose often suffer from poor parameter identifiability. This work presents a model that is applicable to fed-batch hydrolysis by discretizing the substrate based on the feeding time. Different scenarios are tested to explain the observed decrease in reaction rate with increasing conversion, and comprehensive assessment of the parameter sensitivities is carried out. The proposed model performed well in the broad range of experimental conditions used in this study and when compared to literature data. Furthermore, the use of data from fed-batch experiments and discretization of the model substrate to populations was found to be very informative when assessing the importance of the rate-decreasing phenomena in the model.

5Different approaches have been adopted to quantify the performance of stirred vessels in 6 suspending sinking solids into liquid phase. In this study we used electrical impedance 7 tomography (EIT) to estimate the solids distribution in a lab-scale stirred vessel with a diameter 8 of 362 mm. Also visual measurements were made to determine the cloud height and just 9 suspended impeller speed. Quartz sand with a density of 2650 kg/m 3 was employed as the solid 10 phase with different particle size fractions from 50 to 180 µm and solids volume fractions of 7.5 11 and 15 %. The effect of impeller type was studied by using two axial flow impellers, a pitched 12 blade turbine and a hydrofoil impeller. 13Two different states -partial and homogeneous suspension -were defined from the EIT data in 14 addition to visual measurement of complete off-bottom suspension and cloud height. Partial 15 suspension was determined from the EIT data, and it was reached at relatively low agitation 16 rates. Visual measurements and data from the literature also support this observation, and EIT 17 was proved to be a suitable method to quantify a repeatable partial suspension criterion. 18Complete off-bottom suspension was measured visually by determining the agitation rate at 19 which there were no stationary solid particles at the vessel bottom for longer than 2 seconds. 20However, the applicability of this widely used criterion was questioned in the case of dense 21 suspensions of small particles. Homogeneous suspension was estimated from the EIT data, and

Computational fluid dynamics is a powerful method for scale-up of reactors although it is still challenging to fully embrace hydrodynamics and biological complexities. In this article, an aerobic fermentation of Pichia pastoris cells is modeled in a batch OKTOP®9000 reactor. The 800 m3 industrial scale reactor is equipped with a radial impeller, designed by Outotec Oy for gas dispersion in the draft tube reactor. Measured Np of the impeller is used in hydrodynamics validation. The resolved energy dissipation rate is compensated, and its influence on mass transfer is analyzed and discussed. Gas–liquid drag force is modified to simulate effects of liquid turbulence and bubble swarms. Resolved steady state multiphase hydrodynamics is used to simulate the fermentation process. Temporal evolution of species concentrations is compared to experimental data measured in a small copy of the reactor at lab scale (14 L). The effect of oxygenation on the P. pastoris cells cultivation is considered.