Research On Kinetic Product Model For Ethanol Fermentation

With the increasing scarcity of oil resources, the fuel ethanol-renewable energy, has beenmore national attention and development. The research of ethanol fermentation kinetic isbenefit for understanding the fermentation process. It provides a theoretical basis for theamplification and optimization of the fermentation industry. Ethanol fermentation kinetics inmany literatures are based on cell growth models which include the models of sugarconsumption, ethanol production and cell growth. These kind models have more variables. Itneeds to standardize the variables, when it is used for calculation of the bioreactor, so thecalculation is very complicated. Based on the above considerations, this paper attempts toestablish a kinetic product model considering the process of the ethanol formation.In this paper the process of cell growth, ethanol production and substrate consumptionwas analyzed. Dissolved oxygen(DO), pH and temperature in the fermentation was alsodiscussed.Gompertz model was first used to simulating the process of alcoholic fermentation.Gompertz model simulated fermentation experiments with the initial sugar condition150g/L,200g/L,230g/L and250g/L. The correlation coefficient is0.991,0.989,0.994and0.981respectively. The result shows that the model fits the experimental data better. In addition, therelationship of parameters and initial sugar was analyzed, but there is no obvious regularrelationship.Logistic modified model simulated fermentation experiments with the initial sugarcondition150g/L,200g/L,230g/L and250g/L. The correlation coefficient is0.974,0.966,0.984and0.961respectively. The modified Logistic model fits the experimental data better.The relationships of two parameters with initial sugar have analyzed which finds that theyshowed a good linear relationship. The total Logistic model was acquired with considering theeffect of initial sugar. This model fits the fermentation process with initial sugar130g/L verywell. Relative error of the predicted value and experimental value is very small. In particular,it shows a better prediction in the mid-late ethanol fermentation that the relative error is evenless than5%.