Systematic errors in data evaluation due to ethanol stripping and water vaporization.

Systematic errors due to the neglect of water and/or ethanol partition between liquid and gaseous phases are discussed for bioreactors equipped with or without a condenser. Both water vapor and ethanol vapor are present in the off-gas leaving the condenser. Presence of residual water vapor largely influences the gas measurements by dilution. As a consequence, the oxygen consumption rate can be overestimated by a factor of 3 if calculations are not corrected for water vapor content or if no additional device is implemented after the condenser to completely dry the off-gases. The mass balance and partition equations predict that the condenser has only a small effect on reduction of the ethanol vapor content of the off-gas. The reason is the high ethanol concentration of the condensate droplets on the condenser wall in contact with the off-gases. Model predictions as well as experimental results show that ethanol evaporation represents a large fraction of the ethanol production rate and influences greatly the elemental recoveries. For a reactor working at 30 degrees C without condensation of the vapors and for a volumetric aeration rate of 0.63vvm, stripping of ethanol resulted in a gaseous dilution rate of 0.016 h-1 for ethanol. The dilution rate by stripping was reduced to 0.014 h-1 when a condenser at 12 degrees C was implemented. The fraction of ethanol that is stripped is mainly dependent on the ratio D/vvm (liquid to gaseous flow rates), and the effect is only slightly influenced by low condenser temperature. The evaporation of ethanol may account for more than 20% of the ethanol formation rate. Therefore, the condenser does not succeed to reflux all ethanol to the reactor broth. In terms of a unit operation, ethanol vapor can be efficiently reduced by absorption instead of condensation. To demonstrate the feasibility, a simple modification of the reactor was tested for continuous cultures: the feed port was changed from the top-plate to the top of the condenser, which was used as an absorption column. Ethanol stripping was reduced by a factor of 4 as compared to the condensation setup (at 12 degrees C): it accounted for 2% of the ethanol production rate as compared to 8.2% at D = 0.19 h-1 and 0.63vvm.