Modeling and control of non-ideally mixed bioreactors

Mixing plays a substantial role in determining the overall performance of a bioreactor. Well mixing in bioreactor, especially for ethanolic fermentation process is important for the homogenization of miscible and immiscible liquids, gas dispersion and suspension of solid particles. Improper mixing will eventually affect the biological and kinetics reactions occurring in the bioreactor and subsequently deteriorate the bioreactor performance. Currently, most modeling and control applications of bioreactors have been devoted to ideally mixed assumption, for simplicity. This is not realistic in practical applications. Furthermore, the strength and accuracy of the bioreactor models reflect their performance and subsequently its control strategy. Therefore, it is vital to consider the imperfect mixing for the control of bioreactor.In this study, a batch, micro-aerobic bioreactor for ethanolic fermentation process will be considered for modeling. Up to date, not much study has been conducted in exploiting the mixing mechanism for controlling this type of bioreactor. Traditionally, only the bioreactor conditions such as temperature and pH are controlled for such a batch bioreactor. Other parameters, such as aeration rate and stirrer speed are not used to control the bioreactor. Thus, it is difficult to improve the bioreactor performance as the bioreactor performance is less sensitive to both temperature and pH than to the mixing mechanism. However, the mixing behaviour of the bioreactor needs to be captured if we are to employ both aeration rate and stirrer speed for the control of such a batch bioreactor.It is known that aeration rate and stirrer speed could significantly affect the biological and kinetics reactions. Therefore, both aeration rate and stirrer speed are suggested in this work as manipulated variables in the modeling of batch bioreactor. Thus, with this approach the ideally mixed assumption will be relaxed.The models proposed will be implemented for control s