We investigate the performance of single reactor and a cascade of two tanks for the production of ethanol through continuous fermentation.

The simulations use and experimentally validated model of Zymomonos mobilis cultures. The model possesses a rich array of complex dynamic behaviours. These include periodic and period doubling solutions. In certain parameter regions, the latter solutions have been shown to be a path to chaotic behaviour in the reactor.

Numerous studies have investigated these models using direct integration. This approach is time consuming as parameter regions of interest can only be determined through laborious and repetitive calculations. More importantly, crucial regions of parameter space (in terms of optimal performance) can easily be missed by using this approach. We show how techniques from nonlinear dynamical systems theory, in particular a combination of steady-state analysis and path following methods, can be used to gain practical insights into the best reactor design.

By incorporating capital expenditure, feedstock and energy costs the simulations can be used to design optimal fermenter configurations and to determine the likely benefits of new more efficient microorganisms.