Agitation, mixing and mass transfer in simulated high viscosity fermentation broths

Gas-liquid mass transfer, agitator power consumption, rheology, gas-liquid mixing and gas hold-up have been studied in an agitated, sparged vessel of diameter, T = 0.3 m, with a liquid capacity of 0.02 m\(^3\), unaerated liquid height = 0.3 m. The solutions of sodium carboxymethylcellulose used exhibit moderate viscoelasticity and shear thinning behaviour, obeying the power law over the range of shear rates studied. The gas-liquid mass transfer was studied using a steady state technique. This involves monitoring the gas and liquid phase oxygen concentrations when a microorganism (yeast) is cultured in the solutions of interest. Agitator power consumption was measured using strain gauges mounted on the impeller shaft. Various agitator geometries were used. These were: Rushton turbines ( D = T/3 and D = T/2 ), used singly and in pairs; Intermig impellers ( D = 0.58T ), used as a pair; and a 45° pitched blade turbine ( D = T/2 ), used in combination with a Rushton turbine. Gas hold-up and gas-liquid flow patterns were visually observed. In addition, the state of the culture variables, (oxygen uptake rate and carbon dioxide production rate), were used to provide a respiratory quotient, the value of which can be linked to the degree of gas-liquid mixing in the vessel. Measurement of point values of the liquid phase oxygen concentration is also used to indicate the degree of liquid mixing attained. The volumetric mass transfer coefficient, k\(_L\)a, was found to be dependent on the conditions in which the yeast was cultivated, as well as being a function of time. These variations were associated with variations in solution composition seen over the course of each experiment. Steps were taken to ensure that further k\(_L\)a values were measured under identical conditions of the culture variables, in order to determine the effect on k\(_L\)a of varying viscosity, agitator speed and type and air flow rate. Increasing solution viscosity results in poorer gas-liquid mixing and a reduction in k\(_L\)a, as has been found by earlier workers. Thus high agitator speeds and power inputs are required to maintain adequate mass transfer rates. In the more viscous solutions used, large diameter dual impeller systems were required, to mix the gas and liquid phases. Of these a pair of Rushton turbines ( D = T/2 ) gave the highest k\(_L\)a values at a given power input. In these solutions the dependence of k\(_L\)a on the gassing rate, which is seen in intermediate and low viscosity solutions, virtually disappears, with k\(_L\)a highly dependent on the power input and the apparent viscosity. At intermediate viscosities a smaller pair of Rushton turbines showed the most efficient mass transfer characteristics, here k\(_L\)a is dependent on the power input and the gassing rate, but independent of viscosity. This is linked to the flow regime force in the vessel, which at intermediate viscosities lies in the transition region between the laminar and turbulent flow regimes. Variations in gas hold-up, rising then falling with increasing impeller speed, were linked to variations in the gassed power number, falling then rising with increasing impeller speed. These effects are considered to be due to variations in the size of the gas filled cavities behind the impeller blades